diff options
194 files changed, 4601 insertions, 23288 deletions
@@ -64,7 +64,7 @@ without having to use hardware. The virtio drivers are added via KERNEL_FEATURES_INTEL_COMMON. This prevents them from being added to custom kernels by default. They can be removed by adding the following to a conf or kernel bbappend file: - KERNEL_FEATURES_INTEL_COMMON_remove = “cfg/virtio.scc” + KERNEL_FEATURES_INTEL_COMMON:remove = “cfg/virtio.scc” OVMF firmware is also built and can be used in order to emulate a UEFI environment. A full runqemu command line for intel-corei7-64 could look like this: @@ -84,4 +84,4 @@ setting, it will not work in as the primary MACHINE tune as the bootloader needs to be built as a 64bit binary. The setup for this would be as follows: require conf/multilib.conf MULTILIBS = "multilib:libx32 - DEFAULTTUNE_virtclass-multilib-libx32 = "corei7-64-x32" + DEFAULTTUNE:virtclass-multilib-libx32 = "corei7-64-x32" @@ -206,15 +206,15 @@ based installer image instead of default wic image. To build it, you need to add below configuration to local.conf : WKS_FILE = "image-installer.wks.in" - IMAGE_FSTYPES_append = " ext4" - IMAGE_TYPEDEP_wic = "ext4" + IMAGE_FSTYPES:append = " ext4" + IMAGE_TYPEDEP:wic = "ext4" INITRD_IMAGE_LIVE="core-image-minimal-initramfs" do_image_wic[depends] += "${INITRD_IMAGE_LIVE}:do_image_complete" do_rootfs[depends] += "virtual/kernel:do_deploy" - IMAGE_BOOT_FILES_append = "\ + IMAGE_BOOT_FILES:append = "\ ${KERNEL_IMAGETYPE} \ microcode.cpio \ - ${IMGDEPLOYDIR}/${IMAGE_BASENAME}-${MACHINE}.ext4;rootfs.img \ + ${IMGDEPLOYDIR}/${IMAGE_BASENAME}-${MACHINE}.rootfs.ext4;rootfs.img \ ${@bb.utils.contains('EFI_PROVIDER', 'grub-efi', 'grub-efi-bootx64.efi;EFI/BOOT/bootx64.efi', '', d)} \ ${@bb.utils.contains('EFI_PROVIDER', 'grub-efi', '${IMAGE_ROOTFS}/boot/EFI/BOOT/grub.cfg;EFI/BOOT/grub.cfg', '', d)} \ ${@bb.utils.contains('EFI_PROVIDER', 'systemd-boot', 'systemd-bootx64.efi;EFI/BOOT/bootx64.efi', '', d)} \ @@ -394,9 +394,20 @@ Note that both 64-bit and 32-bit firmware is available for the MinnowBoard Turbot, so it is tested against both intel-corei7-64 and intel-core2-32. intel-corei7-64: - NUC6i5SYH - NUC7i7BNH - Coffee Lake-H + Alder Lake-P + Alder Lake-S + Elkhart Lake + Kaby Lake + Raptor Lake-P + Tiger Lake + +intel-skylake-64: + Alder Lake-P + Alder Lake-S + Elkhart Lake + Kaby Lake + Raptor Lake-P + Tiger Lake intel-core2-32: MinnowBoard Turbot diff --git a/README.sources b/README.sources deleted file mode 100644 index 9b483ee0..00000000 --- a/README.sources +++ /dev/null @@ -1,17 +0,0 @@ -The sources for the packages comprising the images shipped with this -BSP can be found at the following location: - -http://downloads.yoctoproject.org/mirror/sources/ - -The metadata used to generate the images shipped with this BSP, in -addition to the code contained in this BSP, can be found at the -following location: - -http://downloads.yoctoproject.org/releases/yocto/yocto-2.7/poky-warrior-21.0.0.tar.bz2 - -The metadata used to generate the images shipped with this BSP, in -addition to the code contained in this BSP, can also be found at the -following locations: - -git://git.yoctoproject.org/poky.git -git://git.yoctoproject.org/meta-intel diff --git a/SECURITY.md b/SECURITY.md new file mode 100644 index 00000000..e70fcb0f --- /dev/null +++ b/SECURITY.md @@ -0,0 +1,6 @@ +# Security Policy +Intel is committed to rapidly addressing security vulnerabilities affecting our customers and providing clear guidance on the solution, impact, severity and mitigation. + +## Reporting a Vulnerability +Please report any security vulnerabilities in this project [utilizing the guidelines here](https://www.intel.com/content/www/us/en/security-center/vulnerability-handling-guidelines.html). + diff --git a/classes/uefi-comboapp.bbclass b/classes/uefi-comboapp.bbclass index 4ecc5535..a05e0ca0 100644 --- a/classes/uefi-comboapp.bbclass +++ b/classes/uefi-comboapp.bbclass @@ -4,8 +4,8 @@ # it with the efi stub obtained from systemd-boot. # Don't add syslinux or build an ISO -PCBIOS_forcevariable = "0" -NOISO_forcevariable = "1" +PCBIOS:forcevariable = "0" +NOISO:forcevariable = "1" # image-live.bbclass will default INITRD_LIVE to the image INITRD_IMAGE creates. # We want behavior to be consistent whether or not "live" is in IMAGE_FSTYPES, so @@ -138,7 +138,7 @@ build_efi_cfg() { : } -populate_kernel_append() { +populate_kernel:append() { # The kernel and initrd are built into the app, so we don't need these if [ -f $dest/initrd ]; then rm $dest/initrd diff --git a/conf/include/maintainers.inc b/conf/include/maintainers.inc index 90c39e4e..9849d0d8 100644 --- a/conf/include/maintainers.inc +++ b/conf/include/maintainers.inc @@ -1,44 +1,54 @@ # This file contains a list of recipe maintainers for meta-intel -RECIPE_MAINTAINER_pn-backport-iwlwifi = "Naveen Saini <naveen.kumar.saini@intel.com>" -RECIPE_MAINTAINER_pn-core-image-rt = "Anuj Mittal <anuj.mittal@intel.com>" -RECIPE_MAINTAINER_pn-core-image-rt-sdk = "Anuj Mittal <anuj.mittal@intel.com>" -RECIPE_MAINTAINER_pn-core-image-tiny = "Naveen Saini <naveen.kumar.saini@intel.com>" -RECIPE_MAINTAINER_pn-core-image-minimal-initramfs = "Anuj Mittal <anuj.mittal@intel.com>" -RECIPE_MAINTAINER_pn-gmmlib = "Anuj Mittal <anuj.mittal@intel.com>" -RECIPE_MAINTAINER_pn-hdcp = "Naveen Saini <naveen.kumar.saini@intel.com>" -RECIPE_MAINTAINER_pn-intel-graphics-compiler = "Naveen Saini <naveen.kumar.saini@intel.com>" -RECIPE_MAINTAINER_pn-intel-compute-runtime = "Naveen Saini <naveen.kumar.saini@intel.com>" -RECIPE_MAINTAINER_pn-intel-media-driver = "Anuj Mittal <anuj.mittal@intel.com>" -RECIPE_MAINTAINER_pn-intel-mediasdk = "Anuj Mittal <anuj.mittal@intel.com>" -RECIPE_MAINTAINER_pn-intel-microcode = "Anuj Mittal <anuj.mittal@intel.com>" -RECIPE_MAINTAINER_pn-intel-vaapi-driver = "Anuj Mittal <anuj.mittal@intel.com>" -RECIPE_MAINTAINER_pn-isa-l = "Naveen Saini <naveen.kumar.saini@intel.com>" -RECIPE_MAINTAINER_pn-itt = "Naveen Saini <naveen.kumar.saini@intel.com>" -RECIPE_MAINTAINER_pn-ixgbe = "Naveen Saini <naveen.kumar.saini@intel.com>" -RECIPE_MAINTAINER_pn-ixgbevf = "Naveen Saini <naveen.kumar.saini@intel.com>" -RECIPE_MAINTAINER_pn-iucode-tool = "Anuj Mittal <anuj.mittal@intel.com>" -RECIPE_MAINTAINER_pn-jhi = "Naveen Saini <naveen.kumar.saini@intel.com>" -RECIPE_MAINTAINER_pn-level-zero = "Naveen Saini <naveen.kumar.saini@intel.com>" -RECIPE_MAINTAINER_pn-libipt = "Naveen Saini <naveen.kumar.saini@intel.com>" -RECIPE_MAINTAINER_pn-libyami = "Anuj Mittal <anuj.mittal@intel.com>" -RECIPE_MAINTAINER_pn-libyami-utils = "Anuj Mittal <anuj.mittal@intel.com>" -RECIPE_MAINTAINER_pn-libxcam = "Naveen Saini <naveen.kumar.saini@intel.com>" -RECIPE_MAINTAINER_pn-linux-intel = "Anuj Mittal <anuj.mittal@intel.com>" -RECIPE_MAINTAINER_pn-linux-intel-rt = "Anuj Mittal <anuj.mittal@intel.com>" -RECIPE_MAINTAINER_pn-linux-intel-dev = "Naveen Saini <naveen.kumar.saini@intel.com>" -RECIPE_MAINTAINER_pn-lms = "Anuj Mittal <anuj.mittal@intel.com>" -RECIPE_MAINTAINER_pn-metee = "Naveen Saini <naveen.kumar.saini@intel.com>" -RECIPE_MAINTAINER_pn-metrics-discovery = "Naveen Saini <naveen.kumar.saini@intel.com>" -RECIPE_MAINTAINER_pn-onednn = "Naveen Saini <naveen.kumar.saini@intel.com>" -RECIPE_MAINTAINER_pn-open-model-zoo = "Anuj Mittal <anuj.mittal@intel.com>" -RECIPE_MAINTAINER_pn-opencl-clang = "Naveen Saini <naveen.kumar.saini@intel.com>" -RECIPE_MAINTAINER_pn-openvino-inference-engine = "Anuj Mittal <anuj.mittal@intel.com>" -RECIPE_MAINTAINER_pn-openvino-model-optimizer = "Anuj Mittal <anuj.mittal@intel.com>" -RECIPE_MAINTAINER_pn-ovmf-shell-image-enrollkeys = "Naveen Saini <naveen.kumar.saini@intel.com>" -RECIPE_MAINTAINER_pn-sbsigntool-native = "Anuj Mittal <anuj.mittal@intel.com>" -RECIPE_MAINTAINER_pn-secureboot-selftest-image-signed = "Anuj Mittal <anuj.mittal@intel.com>" -RECIPE_MAINTAINER_pn-secureboot-selftest-image-unsigned = "Anuj Mittal <anuj.mittal@intel.com>" -RECIPE_MAINTAINER_pn-thermald = "Anuj Mittal <anuj.mittal@intel.com>" -RECIPE_MAINTAINER_pn-xf86-video-ast = "Anuj Mittal <anuj.mittal@intel.com>" -RECIPE_MAINTAINER_pn-zlib-intel = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-backport-iwlwifi = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-core-image-rt = "Anuj Mittal <anuj.mittal@intel.com>" +RECIPE_MAINTAINER:pn-core-image-rt-sdk = "Anuj Mittal <anuj.mittal@intel.com>" +RECIPE_MAINTAINER:pn-core-image-tiny = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-core-image-minimal-initramfs = "Anuj Mittal <anuj.mittal@intel.com>" +RECIPE_MAINTAINER:pn-embree = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-gmmlib = "Lim Siew Hoon <siew.hoon.lim@intel.com>" +RECIPE_MAINTAINER:pn-hdcp = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-intel-cmt-cat = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-intel-compute-runtime = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-intel-crypto-mb = "Anuj Mittal <anuj.mittal@intel.com>" +RECIPE_MAINTAINER:pn-intel-graphics-compiler = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-intel-media-driver = "Lim Siew Hoon <siew.hoon.lim@intel.com>" +RECIPE_MAINTAINER:pn-intel-mediasdk = "Lim Siew Hoon <siew.hoon.lim@intel.com>" +RECIPE_MAINTAINER:pn-intel-microcode = "Anuj Mittal <anuj.mittal@intel.com>" +RECIPE_MAINTAINER:pn-intel-vaapi-driver = "Lim Siew Hoon <siew.hoon.lim@intel.com>" +RECIPE_MAINTAINER:pn-ipmctl = "Anuj Mittal <anuj.mittal@intel.com>" +RECIPE_MAINTAINER:pn-isa-l = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-ispc = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-itt = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-ixgbe = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-ixgbevf = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-iucode-tool = "Anuj Mittal <anuj.mittal@intel.com>" +RECIPE_MAINTAINER:pn-level-zero = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-libipt = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-libva-intel = "Anuj Mittal <anuj.mittal@intel.com>" +RECIPE_MAINTAINER:pn-libva-intel-utils = "Anuj Mittal <anuj.mittal@intel.com>" +RECIPE_MAINTAINER:pn-libxcam = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-linux-intel = "Anuj Mittal <anuj.mittal@intel.com>" +RECIPE_MAINTAINER:pn-linux-intel-rt = "Anuj Mittal <anuj.mittal@intel.com>" +RECIPE_MAINTAINER:pn-linux-intel-dev = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-lms = "Anuj Mittal <anuj.mittal@intel.com>" +RECIPE_MAINTAINER:pn-metee = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-metrics-discovery = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-oidn = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-onednn = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-onedpl = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-onevpl = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-onevpl-intel-gpu = "Yew Chang Ching <chang.ching.yew@intel.com>" +RECIPE_MAINTAINER:pn-open-model-zoo = "Anuj Mittal <anuj.mittal@intel.com>" +RECIPE_MAINTAINER:pn-opencl-clang = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-openvino-inference-engine = "Anuj Mittal <anuj.mittal@intel.com>" +RECIPE_MAINTAINER:pn-openvino-model-optimizer = "Anuj Mittal <anuj.mittal@intel.com>" +RECIPE_MAINTAINER:pn-openvkl = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-ospray = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-ovmf-shell-image-enrollkeys = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-rkcommon = "Naveen Saini <naveen.kumar.saini@intel.com>" +RECIPE_MAINTAINER:pn-sbsigntool-native = "Anuj Mittal <anuj.mittal@intel.com>" +RECIPE_MAINTAINER:pn-secureboot-selftest-image-signed = "Anuj Mittal <anuj.mittal@intel.com>" +RECIPE_MAINTAINER:pn-secureboot-selftest-image-unsigned = "Anuj Mittal <anuj.mittal@intel.com>" +RECIPE_MAINTAINER:pn-thermald = "Anuj Mittal <anuj.mittal@intel.com>" +RECIPE_MAINTAINER:pn-xf86-video-ast = "Anuj Mittal <anuj.mittal@intel.com>" diff --git a/conf/layer.conf b/conf/layer.conf index c4781ffd..97dfb897 100644 --- a/conf/layer.conf +++ b/conf/layer.conf @@ -10,15 +10,16 @@ BBFILE_PATTERN_intel := "^${LAYERDIR}/" BBFILE_PRIORITY_intel = "5" # Additional license directories. -LICENSE_PATH += "${LAYERDIR}/custom-licenses" +CUSTOM_LICENSES_PATH = "${LAYERDIR}/custom-licenses" +LICENSE_PATH += "${CUSTOM_LICENSES_PATH}" LAYERDEPENDS_intel = "core" -LAYERRECOMMENDS_intel = "dpdk intel-qat" +LAYERRECOMMENDS_intel = "dpdk" # This should only be incremented on significant changes that will # cause compatibility issues with other layers LAYERVERSION_intel = "5" -LAYERSERIES_COMPAT_intel = "dunfell gatesgarth hardknott" +LAYERSERIES_COMPAT_intel = "kirkstone scarthgap" BBFILES_DYNAMIC += " \ clang-layer:${LAYERDIR}/dynamic-layers/clang-layer/*/*/*.bb \ @@ -31,10 +32,6 @@ BBFILES_DYNAMIC += " \ require ${LAYERDIR}/conf/include/maintainers.inc -PREFERRED_PROVIDER_zlib ?= "zlib" -PREFERRED_PROVIDER_zlib-native ?= "zlib-native" -PREFERRED_PROVIDER_nativesdk-zlib ?= "nativesdk-zlib" - # Use the libva from OE-Core when layer is included but no MACHINE # from meta-intel is being used. PREFERRED_PROVIDER_libva ?= "libva" @@ -44,3 +41,5 @@ PREFERRED_PROVIDER_nativesdk-libva ?= "nativesdk-libva" PREFERRED_PROVIDER_libva-utils ?= "libva-utils" PREFERRED_PROVIDER_libva-utils-native ?= "libva-utils-native" PREFERRED_PROVIDER_nativesdk-libva-utils ?= "nativesdk-libva-utils" + +# addpylib ${LAYERDIR}/lib oeqa diff --git a/conf/machine/include/intel-common-pkgarch.inc b/conf/machine/include/intel-common-pkgarch.inc index d0f208bd..94d77e06 100644 --- a/conf/machine/include/intel-common-pkgarch.inc +++ b/conf/machine/include/intel-common-pkgarch.inc @@ -1,15 +1,15 @@ INTEL_COMMON_PACKAGE_ARCH ?= "${TUNE_PKGARCH}-intel-common" -PACKAGE_ARCH_pn-linux-intel = "${INTEL_COMMON_PACKAGE_ARCH}" -PACKAGE_ARCH_pn-linux-intel-rt = "${INTEL_COMMON_PACKAGE_ARCH}" -PACKAGE_ARCH_pn-linux-intel-tiny = "${INTEL_COMMON_PACKAGE_ARCH}" -PACKAGE_ARCH_pn-linux-intel-dev = "${INTEL_COMMON_PACKAGE_ARCH}" -PACKAGE_ARCH_pn-linux-yocto = "${INTEL_COMMON_PACKAGE_ARCH}" -PACKAGE_ARCH_pn-linux-yocto-rt = "${INTEL_COMMON_PACKAGE_ARCH}" -PACKAGE_ARCH_pn-linux-yocto-tiny = "${INTEL_COMMON_PACKAGE_ARCH}" -PACKAGE_ARCH_pn-linux-yocto-dev = "${INTEL_COMMON_PACKAGE_ARCH}" -PACKAGE_ARCH_pn-intel-microcode = "${INTEL_COMMON_PACKAGE_ARCH}" -PACKAGE_ARCH_pn-backport-iwlwifi = "${INTEL_COMMON_PACKAGE_ARCH}" -PACKAGE_ARCH_pn-ixgbe = "${INTEL_COMMON_PACKAGE_ARCH}" -PACKAGE_ARCH_pn-ixgbevf = "${INTEL_COMMON_PACKAGE_ARCH}" -PACKAGE_EXTRA_ARCHS_append += "${INTEL_COMMON_PACKAGE_ARCH}" +PACKAGE_ARCH:pn-linux-intel = "${INTEL_COMMON_PACKAGE_ARCH}" +PACKAGE_ARCH:pn-linux-intel-rt = "${INTEL_COMMON_PACKAGE_ARCH}" +PACKAGE_ARCH:pn-linux-intel-tiny = "${INTEL_COMMON_PACKAGE_ARCH}" +PACKAGE_ARCH:pn-linux-intel-dev = "${INTEL_COMMON_PACKAGE_ARCH}" +PACKAGE_ARCH:pn-linux-yocto = "${INTEL_COMMON_PACKAGE_ARCH}" +PACKAGE_ARCH:pn-linux-yocto-rt = "${INTEL_COMMON_PACKAGE_ARCH}" +PACKAGE_ARCH:pn-linux-yocto-tiny = "${INTEL_COMMON_PACKAGE_ARCH}" +PACKAGE_ARCH:pn-linux-yocto-dev = "${INTEL_COMMON_PACKAGE_ARCH}" +PACKAGE_ARCH:pn-intel-microcode = "${INTEL_COMMON_PACKAGE_ARCH}" +PACKAGE_ARCH:pn-backport-iwlwifi = "${INTEL_COMMON_PACKAGE_ARCH}" +PACKAGE_ARCH:pn-ixgbe = "${INTEL_COMMON_PACKAGE_ARCH}" +PACKAGE_ARCH:pn-ixgbevf = "${INTEL_COMMON_PACKAGE_ARCH}" +PACKAGE_EXTRA_ARCHS:append = " ${INTEL_COMMON_PACKAGE_ARCH}" MACHINEOVERRIDES =. "${INTEL_COMMON_PACKAGE_ARCH}:" diff --git a/conf/machine/include/intel-core2-32-common.inc b/conf/machine/include/intel-core2-32-common.inc index 8c469499..ce34d9bb 100644 --- a/conf/machine/include/intel-core2-32-common.inc +++ b/conf/machine/include/intel-core2-32-common.inc @@ -3,5 +3,5 @@ # DEFAULTTUNE ?= "core2-32" -require conf/machine/include/tune-core2.inc -require conf/machine/include/x86-base.inc +require conf/machine/include/x86/tune-core2.inc +require conf/machine/include/x86/x86-base.inc diff --git a/conf/machine/include/intel-corei7-64-common.inc b/conf/machine/include/intel-corei7-64-common.inc index 00282569..3498ebe3 100644 --- a/conf/machine/include/intel-corei7-64-common.inc +++ b/conf/machine/include/intel-corei7-64-common.inc @@ -3,5 +3,5 @@ # DEFAULTTUNE ?= "corei7-64" -require conf/machine/include/tune-corei7.inc -require conf/machine/include/x86-base.inc +require conf/machine/include/x86/tune-corei7.inc +require conf/machine/include/x86/x86-base.inc diff --git a/conf/machine/include/meta-intel.inc b/conf/machine/include/meta-intel.inc index 19f7857d..230d0253 100644 --- a/conf/machine/include/meta-intel.inc +++ b/conf/machine/include/meta-intel.inc @@ -5,17 +5,11 @@ # PREFERRED_PROVIDER_virtual/kernel ?= "linux-intel" -PREFERRED_PROVIDER_virtual/kernel_poky-tiny ?= "linux-intel" +PREFERRED_PROVIDER_virtual/kernel:poky-tiny ?= "linux-intel" -# Only use the Intel-tuned zlib for target builds to improve reuse -PREFERRED_PROVIDER_zlib = "zlib-intel" -PREFERRED_PROVIDER_zlib-native = "zlib-native" -PREFERRED_PROVIDER_nativesdk-zlib = "nativesdk-zlib" - -PREFERRED_VERSION_linux-intel ?= "5.10%" -PREFERRED_VERSION_linux-intel-rt ?= "5.10%" -PREFERRED_VERSION_linux-intel_poky-altcfg ?= "5.4%" -PREFERRED_VERSION_linux-intel-rt_poky-altcfg ?= "5.4%" +PREFERRED_VERSION_linux-intel ?= "6.6%" +PREFERRED_VERSION_linux-intel-rt ?= "6.6%" +PREFERRED_VERSION_linux-intel:poky-altcfg ?= "6.8%" # Need to point to latest version of libva needed for media components PREFERRED_PROVIDER_libva = "libva-intel" @@ -26,29 +20,28 @@ PREFERRED_PROVIDER_libva-utils = "libva-intel-utils" PREFERRED_PROVIDER_libva-utils-native = "libva-intel-utils-native" PREFERRED_PROVIDER_nativesdk-libva-utils = "nativesdk-libva-intel-utils" -PREFERRED_VERSION_opencl-clang ?= "${@bb.utils.contains('LLVMVERSION', '10.0.1', '10.0.0', \ - bb.utils.contains('LLVMVERSION', '11.1.0', '11.0.0', '12.0.0', d), d)}" -PREFERRED_VERSION_opencl-clang-native ?= "${@bb.utils.contains('LLVMVERSION', '10.0.1', '10.0.0', \ - bb.utils.contains('LLVMVERSION', '11.1.0', '11.0.0', '12.0.0', d), d)}" +LLVM_MAJOR_VERSION = "${@d.getVar('LLVMVERSION').split('.')[0]}" +PREFERRED_VERSION_opencl-clang ?= "${@bb.utils.contains('LLVM_MAJOR_VERSION', '14', '14.0.0', '15.0.0', d)}" +PREFERRED_VERSION_opencl-clang-native ?= "${@bb.utils.contains('LLVM_MAJOR_VERSION', '14', '14.0.0', '15.0.0', d)}" XSERVER_X86_ASPEED_AST = "xf86-video-ast \ " # include the user space intel microcode loading support in the generated images. -MACHINE_ESSENTIAL_EXTRA_RDEPENDS_append = "${@bb.utils.contains('MACHINE_FEATURES', 'intel-ucode', ' intel-microcode', '', d)}" +MACHINE_ESSENTIAL_EXTRA_RDEPENDS:append = "${@bb.utils.contains('MACHINE_FEATURES', 'intel-ucode', ' intel-microcode', '', d)}" # recommended extra packages common to all intel machines -MACHINE_EXTRA_RRECOMMENDS_append = " kernel-modules linux-firmware" -MACHINE_ESSENTIAL_EXTRA_RRECOMMENDS_append = " kernel-module-i915 linux-firmware-i915" +MACHINE_EXTRA_RRECOMMENDS:append = " kernel-modules linux-firmware" +MACHINE_ESSENTIAL_EXTRA_RRECOMMENDS:append = " kernel-module-i915 linux-firmware-i915 kernel-module-igc kernel-module-r8152" # for the early boot time kernel microcode loading support, # merge the microcode data in the final initrd image. -INITRD_LIVE_prepend = "${@bb.utils.contains('MACHINE_FEATURES', 'intel-ucode', '${DEPLOY_DIR_IMAGE}/microcode.cpio ', '', d)}" +INITRD_LIVE:prepend = "${@bb.utils.contains('MACHINE_FEATURES', 'intel-ucode', '${DEPLOY_DIR_IMAGE}/microcode.cpio ', '', d)}" DISK_SIGNATURE_UUID ??= "deadbeef-dead-beef-dead-beefdeadbeef" EFI_PROVIDER ?= "systemd-boot" -EFI_PROVIDER_x86-x32 = "grub-efi" +EFI_PROVIDER:x86-x32 = "grub-efi" # Add general MACHINEOVERRIDE for meta-intel MACHINEOVERRIDES =. "intel-x86-common:" diff --git a/conf/machine/include/qemu-intel.inc b/conf/machine/include/qemu-intel.inc index 81510404..4608db99 100644 --- a/conf/machine/include/qemu-intel.inc +++ b/conf/machine/include/qemu-intel.inc @@ -2,7 +2,7 @@ # distro content (in particular the kernel) less than qemu.inc. # Ensure that qemu gets built when building images. -EXTRA_IMAGEDEPENDS += "qemu-native qemu-helper-native" +EXTRA_IMAGEDEPENDS += "qemu-system-native qemu-helper-native:do_addto_recipe_sysroot" # Build ovmf firmware for uefi support in qemu. EXTRA_IMAGEDEPENDS += "ovmf" diff --git a/conf/machine/include/qemuboot-intel.inc b/conf/machine/include/qemuboot-intel.inc index befcc250..17a1234d 100644 --- a/conf/machine/include/qemuboot-intel.inc +++ b/conf/machine/include/qemuboot-intel.inc @@ -1,19 +1,18 @@ # For runqemu IMAGE_CLASSES += "qemuboot" -QB_SYSTEM_NAME_intel-core2-32 = "qemu-system-i386" -QB_CPU_intel-core2-32 = "-cpu coreduo" -QB_CPU_KVM_intel-core2-32 = "-cpu kvm32" +QB_SYSTEM_NAME:intel-core2-32 = "qemu-system-i386" +QB_CPU:intel-core2-32 = "-cpu coreduo" +QB_CPU_KVM:intel-core2-32 = "-cpu kvm32" -QB_SYSTEM_NAME_intel-corei7-64 = "qemu-system-x86_64" -QB_CPU_intel-corei7-64 = "-cpu Nehalem" -QB_CPU_KVM_intel-corei7-64 = "-cpu kvm64" +QB_SYSTEM_NAME:intel-corei7-64 = "qemu-system-x86_64" +QB_CPU:intel-corei7-64 = "-cpu Nehalem" +QB_CPU_KVM:intel-corei7-64 = "-cpu kvm64" -QB_SYSTEM_NAME_intel-skylake-64 = "qemu-system-x86_64" -QB_CPU_intel-skylake-64 = "-cpu Skylake-Client" -QB_CPU_KVM_intel-skylake-64 = "-cpu Skylake-Client" +QB_SYSTEM_NAME:intel-skylake-64 = "qemu-system-x86_64" +QB_CPU:intel-skylake-64 = "-cpu Skylake-Client" +QB_CPU_KVM:intel-skylake-64 = "-cpu Skylake-Client" QB_AUDIO_DRV = "alsa" -QB_AUDIO_OPT = "-soundhw ac97,es1370" -QB_KERNEL_CMDLINE_APPEND = "vga=0 uvesafb.mode_option=640x480-32 oprofile.timer=1 uvesafb.task_timeout=-1" -# Add the 'virtio-rng-pci' device otherwise the guest may run out of entropy -QB_OPT_APPEND = "-vga vmware -show-cursor -usb -usbdevice tablet -device virtio-rng-pci" +QB_AUDIO_OPT = "-device AC97" +QB_KERNEL_CMDLINE_APPEND = " oprofile.timer=1" +QB_OPT_APPEND = " -usb -usbdevice tablet " diff --git a/conf/machine/include/tune-skylake.inc b/conf/machine/include/tune-skylake.inc deleted file mode 100644 index 4882600b..00000000 --- a/conf/machine/include/tune-skylake.inc +++ /dev/null @@ -1,48 +0,0 @@ -# Settings for the GCC(1) cpu-type "skylake": -# -# Intel Skylake CPU with 64-bit extensions, MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, -# SSE4.2, AVX, AVX2 and POPCNT instruction set support. -# -# This tune is recommended for Intel Skylake CPU (and beyond). -# -DEFAULTTUNE ?= "skylake-64" - -# Include the previous tune to pull in PACKAGE_EXTRA_ARCHS -require conf/machine/include/tune-corei7.inc - -# Extra tune features -TUNEVALID[skylake] = "Enable skylake specific processor optimizations" -TUNE_CCARGS .= "${@bb.utils.contains('TUNE_FEATURES', 'skylake', ' -march=skylake ${SKYLAKE_TUNE} -mfpmath=sse', '', d)}" - -# Extra tune selections - -AVAILTUNES += "skylake-64" -TUNE_FEATURES_tune-skylake-64 = "${TUNE_FEATURES_tune-x86-64} skylake" -BASE_LIB_tune-skylake-64 = "lib64" -TUNE_PKGARCH_tune-skylake-64 = "skylake-64" -PACKAGE_EXTRA_ARCHS_tune-skylake-64 = "${PACKAGE_EXTRA_ARCHS_tune-core2-64} skylake-64" -QEMU_EXTRAOPTIONS_skylake-64 = " -cpu Skylake-Client" - - -# Disable QEMU usermode by default (get avx2) -MACHINE_FEATURES_remove = "qemu-usermode" - -# If qemu-usermode is enabled, we have to disable avx2 ISA extensions, but we can keep mtune as skylake vs generic -SKYLAKE_TUNE .= "${@bb.utils.contains('MACHINE_FEATURES', 'qemu-usermode', ' -mtune=skylake ${QEMU_UNAVAILABLE_ISA}', '-mtune=generic -mavx2', d)}" - -QEMU_UNAVAILABLE_ISA = " \ --mno-avx \ --mno-avx2 \ --mno-avx512f \ --mno-avx512er \ --mno-avx512cd \ --mno-avx512pf \ --mno-avx512dq \ --mno-avx512bw \ --mno-avx512vl \ --mno-avx512ifma \ --mno-avx512vbmi \ --mno-avx512vbmi2 \ --mno-avx512vnni \ --mno-avx512bitalg \ -" diff --git a/conf/machine/intel-core2-32.conf b/conf/machine/intel-core2-32.conf index 384ad1e9..5b134dc3 100644 --- a/conf/machine/intel-core2-32.conf +++ b/conf/machine/intel-core2-32.conf @@ -13,9 +13,6 @@ MACHINE_FEATURES += "intel-ucode" MACHINE_HWCODECS ?= "intel-vaapi-driver gstreamer1.0-vaapi" -# Enable optional QAT items: -COMPATIBLE_MACHINE_pn-qat17 = "intel-core2-32" - XSERVER ?= "${XSERVER_X86_BASE} \ ${XSERVER_X86_EXT} \ ${XSERVER_X86_FBDEV} \ @@ -32,4 +29,4 @@ APPEND += "rootwait console=ttyS0,115200 console=ttyPCH0,115200 console=tty0" IMAGE_FSTYPES += "wic" WKS_FILE ?= "${@bb.utils.contains_any("EFI_PROVIDER", "systemd-boot", "systemd-bootdisk-microcode.wks.in", "grub-bootdisk-microcode.wks.in", d)}" -WKS_FILE_DEPENDS_append = " intel-microcode" +WKS_FILE_DEPENDS:append = " intel-microcode" diff --git a/conf/machine/intel-corei7-64.conf b/conf/machine/intel-corei7-64.conf index 033b1aa9..968395fe 100644 --- a/conf/machine/intel-corei7-64.conf +++ b/conf/machine/intel-corei7-64.conf @@ -14,12 +14,8 @@ MACHINE_FEATURES += "intel-ucode" MACHINE_HWCODECS ?= "${@bb.utils.contains('TUNE_FEATURES', 'mx32', '', 'intel-media-driver intel-mediasdk', d)} gstreamer1.0-vaapi" # Enable optional dpdk: -COMPATIBLE_MACHINE_pn-dpdk = "intel-corei7-64" - -# Enable optional QAT items: -COMPATIBLE_MACHINE_pn-qat16 = "intel-corei7-64" -COMPATIBLE_MACHINE_pn-qat17 = "intel-corei7-64" -COMPATIBLE_MACHINE_pn-zlib-qat = "intel-corei7-64" +COMPATIBLE_MACHINE:pn-dpdk = "intel-corei7-64" +COMPATIBLE_MACHINE:pn-dpdk-module = "intel-corei7-64" XSERVER ?= "${XSERVER_X86_BASE} \ ${XSERVER_X86_EXT} \ @@ -37,4 +33,4 @@ APPEND += "rootwait console=ttyS0,115200 console=tty0" IMAGE_FSTYPES += "wic" WKS_FILE ?= "${@bb.utils.contains_any("EFI_PROVIDER", "systemd-boot", "systemd-bootdisk-microcode.wks.in", "grub-bootdisk-microcode.wks.in", d)}" -WKS_FILE_DEPENDS_append = " intel-microcode" +WKS_FILE_DEPENDS:append = " intel-microcode" diff --git a/conf/machine/intel-skylake-64.conf b/conf/machine/intel-skylake-64.conf index eccd59b8..a2b392c2 100644 --- a/conf/machine/intel-skylake-64.conf +++ b/conf/machine/intel-skylake-64.conf @@ -4,8 +4,8 @@ #@DESCRIPTION: Machine configuration for 64 bit Intel Skylake CPU (and later) with MMX, SSE, SSE2, SSE3, SSE4.1, SSE4.2, AVX, and AVX2 instruction set support. Supports a moderately wide range of drivers that should boot and be usable on "typical" hardware. require conf/machine/include/meta-intel.inc -require conf/machine/include/x86-base.inc -require conf/machine/include/tune-skylake.inc +require conf/machine/include/x86/x86-base.inc +require conf/machine/include/x86/tune-x86-64-v3.inc require conf/machine/include/intel-common-pkgarch.inc MACHINE_FEATURES += "efi" @@ -14,7 +14,8 @@ MACHINE_FEATURES += "intel-ucode" MACHINE_HWCODECS ?= "intel-media-driver intel-mediasdk gstreamer1.0-vaapi" -COMPATIBLE_MACHINE_pn-dpdk = "intel-skylake-64" +COMPATIBLE_MACHINE:pn-dpdk = "intel-skylake-64" +COMPATIBLE_MACHINE:pn-dpdk-module = "intel-skylake-64" XSERVER ?= "${XSERVER_X86_BASE} \ ${XSERVER_X86_EXT} \ @@ -32,4 +33,4 @@ APPEND += "rootwait console=ttyS0,115200 console=tty0" IMAGE_FSTYPES += "wic" WKS_FILE ?= "${@bb.utils.contains_any("EFI_PROVIDER", "systemd-boot", "systemd-bootdisk-microcode.wks.in", "grub-bootdisk-microcode.wks.in", d)}" -WKS_FILE_DEPENDS_append = " intel-microcode" +WKS_FILE_DEPENDS:append = " intel-microcode" diff --git a/custom-licenses/EULA b/custom-licenses/EULA new file mode 100644 index 00000000..549f78e1 --- /dev/null +++ b/custom-licenses/EULA @@ -0,0 +1,392 @@ +Intel End User License Agreement for Developer Tools (Version October 2021) + +IMPORTANT NOTICE - PLEASE READ AND AGREE BEFORE DOWNLOADING, INSTALLING, COPYING +OR USING + +This Agreement is between you, or the company or other legal entity that you +represent and warrant you have the legal authority to bind, (each, "You" or +"Your") and Intel Corporation and its subsidiaries (collectively, "Intel") +regarding Your use of the Materials. By downloading, installing, copying or +using the Materials, You agree to be bound by the terms of this Agreement. If +You do not agree to the terms of this Agreement, or do not have legal authority +or required age to agree to them, do not download, install, copy or use the +Materials. + +1. LICENSE DEFINITIONS. + + A. "Cloud Provider" means a third party service provider offering a cloud-based + platform, infrastructure, application or storage services, such as Microsoft + Azure or Amazon Web Services, which You may only utilize to host the + Materials subject to the restrictions set forth in Section 2.3 B. + + B. "Derivative Work" means a derivative work, as defined in 17 U.S.C. 101, of + the Source Code. + + C. "Executable Code" means computer programming code in binary form suitable + for machine execution by a processor without the intervening steps of + interpretation or compilation. + + D. "Materials" mean the software, documentation, the software product serial + number, and other collateral, including any updates, made available to You + by Intel under this Agreement. Materials include Redistributables, + Executable Code, Source Code, Sample Source Code, and Pre-Release Materials, + but do not include Third Party Software. + + E. "Pre-Release Materials" mean the Materials, or portions of the Materials, + that are identified (in the product release notes, on Intel's download + website for the Materials or elsewhere) or labeled as pre-release, + prototype, alpha or beta code and, as such, are deemed to be pre-release + code (i) which may not be fully functional or tested and may contain bugs or + errors; (ii) which Intel may substantially modify in its development of a + production version; or (iii) for which Intel makes no assurances that it + will ever develop or make a production version generally available. + Pre-Release Materials are subject to the terms of Section 3.2. + + F. "Reciprocal Open Source Software" means any software that is subject to a + license which requires that (i) it must be distributed in source code form; + (ii) it must be licensed under the same open source license terms; and (iii) + its derivative works must be licensed under the same open source license + terms. Examples of this type of license are the GNU General Public License + or the Mozilla Public License. + + G. "Redistributables" mean the files (if any) listed in the "redist.txt," + "redist-rt.txt" or similarly-named text files that may be included in the + Materials. Redistributables include Sample Source Code. + + H. "Sample Source Code" means those portions of the Materials that are Source + Code and are identified as sample code. Sample Source Code may not have been + tested or validated by Intel and is provided purely as a programming example. + + I. "Source Code" means the software portion of the Materials provided in human + readable format. + + J. "Third Party Software" mean the files (if any) listed in the + "third-party-software.txt" or other similarly-named text file that may be + included in the Materials for the applicable software. Third Party Software + is subject to the terms of Section 2.2. + + K. "Your Product" means one or more applications, products or projects + developed by or for You using the Materials. + +2. LICENSE GRANTS. + +2.1 License to the Materials. Subject to the terms and conditions of this + Agreement, Intel grants You a non-exclusive, worldwide, non-assignable, + non-sublicensable, limited right and license under its copyrights, to: + + A. reproduce internally a reasonable number of copies of the Materials for Your + personal or business use; + + B. use the Materials solely for Your personal or business use to develop Your + Product, in accordance with the documentation included as part of the + Materials; + + C. modify or create Derivative Works only of the Redistributables, or any + portions, that are provided to You in Source Code; + + D. distribute (directly and through Your distributors, resellers, and other + channel partners, if applicable), the Redistributables, including any + modifications to or Derivative Works of the Redistributables or any portions + made pursuant to Section 2.1.C subject to the following conditions: + + (1) Any distribution of the Redistributables must only be as part of Your + Product which must add significant primary functionality different than + that of the Redistributables themselves; + + (2) You must only distribute the Redistributables originally provided to You + by Intel only in Executable Code subject to a license agreement that + prohibits reverse engineering, decompiling or disassembling the + Redistributables; + + (3) This distribution right includes a limited right to sublicense only the + Intel copyrights in the Redistributables and only to the extent necessary + to perform, display, and distribute the Redistributables (including Your + modifications and Derivative Works of the Redistributables provided in + Source Code) solely as incorporated in Your Product; and + + (4) You: (i) will be solely responsible to Your customers for any update, + support obligation or other obligation or liability which may arise from + the distribution of Your Product, (ii) will not make any statement that + Your Product is "certified" or that its performance is guaranteed by Intel + or its suppliers, (iii) will not use Intel's or its suppliers' names or + trademarks to market Your Product, (iv) will comply with any additional + restrictions which are included in the text files with the + Redistributables and in Section 3 below, (v) will indemnify, hold + harmless, and defend Intel and its suppliers from and against any claims + or lawsuits, costs, damages, and expenses, including attorney's fees, that + arise or result from (a) Your modifications or Derivative Works of the + Materials or (b) Your distribution of Your Product. + +2.2 Third Party Software. Third Party Software, even if included with the + distribution of the Materials, may be governed by separate license terms, + including without limitation, third party license terms, open source + software notices and terms, and/or other Intel software license terms. These + separate license terms solely govern Your use of the Third Party Software. + +2.3 Third Party Use. + + A. If You are an entity, Your contractors may use the Materials under the + license specified in Section 2, provided: (i) their use of the Materials is + solely on behalf of and in support of Your business, (ii) they agree to the + terms and conditions of this Agreement, and (iii) You are solely responsible + for their use, misuse or disclosure of the Materials. + + B. You may utilize a Cloud Provider to host the Materials for You, provided: + (i) the Cloud Provider may only host the Materials for Your exclusive use + and may not use the Materials for any other purpose whatsoever, including the + restriction set forth in Section 3.1(xi); (ii) the Cloud Provider's use of + the Materials must be solely on behalf of and in support of Your Product, and + (iii) You will indemnify, hold harmless, and defend Intel and its suppliers + from and against any claims or lawsuits, costs, damages, and expenses, + including attorney's fees, that arise or result from Your Cloud Provider's + use, misuse or disclosure of the Materials. + +3. LICENSE CONDITIONS. + +3.1 Restrictions. Except as expressly provided in this Agreement, You may NOT: + (i) use, reproduce, disclose, distribute, or publicly display the + Materials; (ii) share, publish, rent or lease the Materials to any third + party; (iii) assign this Agreement or transfer the Materials; (iv) modify, + adapt, or translate the Materials in whole or in part; (v) reverse engineer, + decompile, or disassemble the Materials, or otherwise attempt to derive the + source code for the software; (vi) work around any technical limitations in + the Materials; (vii) distribute, sublicense or transfer any Source Code, + modifications or Derivative Works of any Source Code to any third party; + (viii) remove, minimize, block or modify any notices of Intel or its + suppliers in the Materials; (ix) include the Redistributables in malicious, + deceptive, or unlawful programs or products or use the Materials in any way + that is against the law; (x) modify, create a Derivative Work, link, or + distribute the Materials so that any part of it becomes Reciprocal Open + Source Software; (xi) use the Materials directly or indirectly for SaaS + services or service bureau purposes (i.e., a service that allows use of or + access to the Materials by a third party as part of that service, such as + the salesforce.com service business model). + +3.2 Pre-Release Materials. If You receive Pre-Release Materials, You may + reproduce a reasonable number of copies and use the Pre-Release Materials + for evaluation and testing purposes only. You may not (i) modify or + incorporate the Pre-Release Materials into Your Product; (ii) continue to + use the Pre-Release Materials once a commercial version is released; or + (iii) disclose to any third party any benchmarks, performance results, or + other information relating to the Pre-Release Materials. Intel may waive + these restrictions in writing at its sole discretion; however, if You decide + to use the Pre-Release Materials in Your Product (even with Intel's waiver), + You acknowledge and agree that You are fully responsible for any and all + issues that result from such use. + +3.3 Safety-Critical, and Life-Saving Applications; Indemnity. The Materials may + provide information relevant to safety-critical applications + ("Safety-Critical Applications") to allow compliance with functional safety + standards or requirements. You acknowledge and agree that safety is Your + responsibility. To the extent You use the Materials to create, or as part + of, products used in Safety-Critical Applications, it is Your responsibility + to design, manage, and ensure that there are system-level safeguards to + anticipate, monitor, and control system failures, and You agree that You are + solely responsible for all applicable regulatory standards and + safety-related requirements concerning Your use of the Materials in Safety + Critical Applications. + + Should You use the Materials for Safety-Critical Applications or in any type + of a system or application in which the failure of the Materials could + create a situation where personal injury or death may occur (e.g., medical + systems, life-sustaining or life-saving systems) ("Life-Saving + Applications"), You agree to indemnify, defend, and hold Intel and its + representatives harmless against any claims or lawsuits, costs, damages, and + expenses, including reasonable attorney fees, arising in any way out of Your + use of the Materials in Safety-Critical Applications or Life-Saving + Applications and claims of product liability, personal injury or death + associated with those applications; even if such claims allege that Intel + was negligent or strictly liable regarding the design or manufacture of the + Materials or its failure to warn regarding the Materials. + +3.4 Media Format Codecs and Digital Rights Management. You acknowledge and agree + that Your use of the Materials or distribution of the Redistributables with + Your Product as permitted by this Agreement may require You to procure + license(s) from third parties that may hold intellectual property rights + applicable to any media decoding, encoding or transcoding technology (e.g., + the use of an audio or video codec) and/or digital rights management + capabilities of the Materials, if any. Should any such additional licenses + be required, You are solely responsible for obtaining any such licenses and + agree to obtain any such licenses at Your own expense. + +4. DATA COLLECTION AND PRIVACY. + +4.1 Data Collection. The Materials may generate and collect anonymous data + and/or provisioning data about the Materials and/or the development + environment and transmit the data to Intel as a one-time event during + installation. Optional data may also be collected by the Materials, however, + You will be provided notice of the request to collect optional data and no + optional data will be collected without Your consent. All data collection by + Intel is performed pursuant to relevant privacy laws, including notice and + consent requirements. + +4.2 Intel's Privacy Notice. Intel is committed to respecting Your privacy. To + learn more about Intel's privacy practices, please visit + http://www.intel.com/privacy. + +5. OWNERSHIP. Title to the Materials and all copies remain with Intel or its + suppliers. The Materials are protected by intellectual property rights, + including without limitation, United States copyright laws and international + treaty provisions. You will not remove any copyright or other proprietary + notices from the Materials. Except as expressly provided herein, no license + or right is granted to You directly or by implication, inducement, estoppel + or otherwise; specifically, Intel does not grant any express or implied right + to You under Intel patents, copyrights, trademarks, or trade secrets. + +6. NO WARRANTY AND NO SUPPORT. + +6.1 No Warranty. Disclaimer. Intel disclaims all warranties of any kind and the + terms and remedies provided in this Agreement are instead of any other + warranty or condition, express, implied or statutory, including those + regarding merchantability, fitness for any particular purpose, + non-infringement or any warranty arising out of any course of dealing, usage + of trade, proposal, specification or sample. Intel does not assume (and does + not authorize any person to assume on its behalf) any liability. + +6.2 No Support; Priority Support. Intel may make changes to the Materials, or to + items referenced therein, at any time without notice, but is not obligated + to support, update or provide training for the Materials under the terms of + this Agreement. Intel offers free community and paid priority support + options. More information on these support options can be found at: + https://software.intel.com/content/www/us/en/develop/support/priority-support.html. + +7. LIMITATION OF LIABILITY. + +7.1 Intel will not be liable for any of the following losses or damages (whether + such losses or damages were foreseen, foreseeable, known or otherwise): (i) + loss of revenue; (ii) loss of actual or anticipated profits; (iii) loss of + the use of money; (iv) loss of anticipated savings; (v) loss of business; + (vi) loss of opportunity; (vii) loss of goodwill; (viii) loss of use of the + Materials; (ix) loss of reputation; (x) loss of, damage to, or corruption of + data; or (xi) any indirect, incidental, special or consequential loss of + damage however caused (including loss or damage of the type specified in + this Section 7). + +7.2 Intel's total cumulative liability to You, including for direct damages for + claims relating to this Agreement, and whether for breach of contract, + negligence, or for any other reason, will not exceed $100. + +7.3 You acknowledge that the limitations of liability provided in this Section 7 + are an essential part of this Agreement. You agree that the limitations of + liability provided in this Agreement with respect to Intel will be conveyed + to and made binding upon any customer of Yours that acquires the + Redistributables. + +8. USER SUBMISSIONS. Should you provide Intel with comments, modifications, + corrections, enhancements or other input ("Feedback") related to the + Materials, Intel will be free to use, disclose, reproduce, license or + otherwise distribute or exploit the Feedback in its sole discretion without + any obligations or restrictions of any kind, including without limitation, + intellectual property rights or licensing obligations. If You wish to provide + Intel with information that You intend to be treated as confidential + information, Intel requires that such confidential information be provided + pursuant to a non-disclosure agreement ("NDA"); please contact Your Intel + representative to ensure the proper NDA is in place. + +9. NON-DISCLOSURE. Information provided by Intel to You may include information + marked as confidential. You must treat such information as confidential under + the terms of the applicable NDA between Intel and You. If You have not + entered into an NDA with Intel, You must not disclose, distribute or make use + of any information marked as confidential, except as expressly authorized in + writing by Intel. Intel retains all rights in and to its confidential + information specifications, designs, engineering details, discoveries, + inventions, patents, copyrights, trademarks, trade secrets, and other + proprietary rights relating to the Materials. Any breach by You of the + confidentiality obligations provided for in this Section 9 will cause + irreparable injury to Intel for which money damages may be inadequate to + compensate Intel for losses arising from such a breach. Intel may obtain + equitable relief, including injunctive relief, if You breach or threaten to + breach Your confidentiality obligations. + +10. TERM AND TERMINATION. This Agreement becomes effective on the date You + accept this Agreement and will continue until terminated as provided for in + this Agreement. The term for any Pre-Release Materials terminates upon + release of a commercial version. This Agreement will terminate if You are in + breach of any of its terms and conditions. Upon termination, You will + promptly destroy the Materials and all copies. In the event of termination of + this Agreement, Your license to any Redistributables distributed by You in + accordance with the terms and conditions of this Agreement, prior to the + effective date of such termination, will survive any such termination of this + Agreement. Sections 1, 2.1.D(4)(v), 2.2, 2.3.A(iii), 2.3.B(iii), 3.3, 5, 6, + 7, 8, 9, 10 (with respect to these survival provisions in the last sentence), + and 12 will survive expiration or termination of this Agreement. + +11. U.S. GOVERNMENT RESTRICTED RIGHTS. The technical data and computer software + covered by this license is a "Commercial Item," as such term is defined by + the FAR 2.101 (48 C.F.R. 2.101) and is "commercial computer software" and + "commercial computer software documentation" as specified under FAR 12.212 + (48 C.F.R. 12.212) or DFARS 227.7202 (48 C.F.R. 227.7202), as applicable. + This commercial computer software and related documentation is provided to + end users for use by and on behalf of the U.S. Government with only those + rights as are granted to all other end users pursuant to the terms and + conditions of this Agreement. + +12. GENERAL PROVISIONS. + +12.1 ENTIRE AGREEMENT. This Agreement contains the complete and exclusive + agreement and understanding between the parties concerning the subject + matter of this Agreement, and supersedes all prior and contemporaneous + proposals, agreements, understanding, negotiations, representations, + warranties, conditions, and communications, oral or written, between the + parties relating to the same subject matter. Each party acknowledges and + agrees that in entering into this Agreement it has not relied on, and will + not be entitled to rely on, any oral or written representations, + warranties, conditions, understanding, or communications between the + parties that are not expressly set forth in this Agreement. The express + provisions of this Agreement control over any course of performance, course + of dealing, or usage of the trade inconsistent with any of the provisions + of this Agreement. The provisions of this Agreement will prevail + notwithstanding any different, conflicting, or additional provisions that + may appear on any purchase order, acknowledgement, invoice, or other + writing issued by either party in connection with this Agreement. No + modification or amendment to this Agreement will be effective unless in + writing and signed by authorized representatives of each party, and must + specifically identify this Agreement by its title and version (e.g., "Intel + End User License Agreement for Developer Tools (Version October 2021)"); + except that Intel may make changes to this Agreement as it distributes new + versions of the Materials. When changes are made, Intel will make a new + version of the Agreement available on its website. If You received a copy + of this Agreement translated into another language, the English language + version of this Agreement will prevail in the event of any conflict between + versions. + +12.2 EXPORT. You acknowledge that the Materials and all related technical + information are subject to export controls and you agree to comply with all + laws and regulations of the United States and other applicable governments + governing export, re-export, import, transfer, distribution, and use of the + Materials. In particular, but without limitation, the Materials may not be + exported or re-exported (i) into any U.S. embargoed countries or (ii) to + any person or entity listed on a denial order published by the U.S. + government or any other applicable governments. By using the Materials, You + represent and warrant that You are not located in any such country or on + any such list. You also agree that You will not use the Materials for, or + sell or transfer them to a third party who is known or suspected to be + involved in, any purposes prohibited by the U.S. government or other + applicable governments, including, without limitation, the development, + design, manufacture, or production of nuclear, missile, chemical or + biological weapons. + +12.3 GOVERNING LAW, JURISDICTION, AND VENUE. All disputes arising out of or + related to this Agreement, whether based on contract, tort, or any other + legal or equitable theory, will in all respects be governed by, and + construed and interpreted under, the laws of the United States of America + and the State of Delaware, without reference to conflict of laws + principles. The parties agree that the United Nations Convention on + Contracts for the International Sale of Goods (1980) is specifically + excluded from and will not apply to this Agreement. All disputes arising + out of or related to this Agreement, whether based on contract, tort, or + any other legal or equitable theory, will be subject to the exclusive + jurisdiction of the courts of the State of Delaware or of the Federal + courts sitting in that State. Each party submits to the personal + jurisdiction of those courts and waives all objections to that jurisdiction + and venue for those disputes. + +12.4 SEVERABILITY. The parties intend that if a court holds that any provision + or part of this Agreement is invalid or unenforceable under applicable law, + the court will modify the provision to the minimum extent necessary to make + it valid and enforceable, or if it cannot be made valid and enforceable, + the parties intend that the court will sever and delete the provision or + part from this Agreement. Any change to or deletion of a provision or part + of this Agreement under this Section will not affect the validity or + enforceability of the remainder of this Agreement, which will continue in + full force and effect. diff --git a/custom-licenses/ISSL b/custom-licenses/ISSL index 5435ee04..f987e502 100644 --- a/custom-licenses/ISSL +++ b/custom-licenses/ISSL @@ -1,63 +1,73 @@ -Intel Simplified Software License (Version April 2018) +Intel Simplified Software License (Version August 2021) -Copyright (c) 2018 Intel Corporation. +Use and Redistribution. You may use and redistribute the software (the +"Software"), without modification, provided the following conditions are met: -Use and Redistribution. You may use and redistribute the software (the "Software"), without modification, -provided the following conditions are met: +* Redistributions must reproduce the above copyright notice and the following + terms of use in the Software and in the documentation and/or other materials + provided with the distribution. +* Neither the name of Intel nor the names of its suppliers may be used to + endorse or promote products derived from this Software without specific + prior written permission. +* No reverse engineering, decompilation, or disassembly of this Software is + permitted. -* Redistributions must reproduce the above copyright notice and the following terms of use in the Software -and in the documentation and/or other materials provided with the distribution. +No other licenses. Except as provided in the preceding section, Intel grants no +licenses or other rights by implication, estoppel or otherwise to, patent, +copyright, trademark, trade name, service mark or other intellectual property +licenses or rights of Intel. -* Neither the name of Intel nor the names of its suppliers may be used to endorse or promote products derived -from this Software without specific prior written permission. +Third party software. The Software may contain Third Party Software. "Third +Party Software" is open source software, third party software, or other Intel +software that may be identified in the Software itself or in the files (if any) +listed in the "third-party-software.txt" or similarly named text file included +with the Software. Third Party Software, even if included with the distribution +of the Software, may be governed by separate license terms, including without +limitation, open source software license terms, third party software license +terms, and other Intel software license terms. Those separate license terms +solely govern your use of the Third Party Software, and nothing in this license +limits any rights under, or grants rights that supersede, the terms of the +applicable license terms. -* No reverse engineering, decompilation, or disassembly of this Software is permitted. +DISCLAIMER. THIS SOFTWARE IS PROVIDED "AS IS" AND ANY EXPRESS OR IMPLIED +WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF +MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, AND NON-INFRINGEMENT ARE +DISCLAIMED. THIS SOFTWARE IS NOT INTENDED FOR USE IN SYSTEMS OR APPLICATIONS +WHERE FAILURE OF THE SOFTWARE MAY CAUSE PERSONAL INJURY OR DEATH AND YOU AGREE +THAT YOU ARE FULLY RESPONSIBLE FOR ANY CLAIMS, COSTS, DAMAGES, EXPENSES, AND +ATTORNEYS' FEES ARISING OUT OF ANY SUCH USE, EVEN IF ANY CLAIM ALLEGES THAT +INTEL WAS NEGLIGENT REGARDING THE DESIGN OR MANUFACTURE OF THE SOFTWARE. -Limited patent license. Intel grants you a world-wide, royalty-free, non-exclusive license under patents -it now or hereafter owns or controls to make, have made, use, import, offer to sell and sell ("Utilize") -this Software, but solely to the extent that any such patent is necessary to Utilize the Software alone. -The patent license shall not apply to any combinations which include this software. No hardware per se -is licensed hereunder. +LIMITATION OF LIABILITY. IN NO EVENT WILL INTEL BE LIABLE FOR ANY DIRECT, +INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, +BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF +LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE +OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF +ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. -Third party and other Intel programs. "Third Party Programs" are the files listed in the -"third-party-programs.txt" text file that is included with the Software and may include Intel programs -under separate license terms. Third Party Programs, even if included with the distribution of the Materials, -are governed by separate license terms and those license terms solely govern your use of those programs. +No support. Intel may make changes to the Software, at any time without notice, +and is not obligated to support, update or provide training for the Software. -DISCLAIMER. THIS SOFTWARE IS PROVIDED "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, -BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, -AND NON-INFRINGEMENT ARE DISCLAIMED. THIS SOFTWARE IS NOT INTENDED FOR USE IN SYSTEMS OR APPLICATIONS -WHERE FAILURE OF THE SOFTWARE MAY CAUSE PERSONAL INJURY OR DEATH AND YOU AGREE THAT YOU ARE FULLY RESPONSIBLE -FOR ANY CLAIMS, COSTS, DAMAGES, EXPENSES, AND ATTORNEYS’ FEES ARISING OUT OF ANY SUCH USE, EVEN IF ANY CLAIM -ALLEGES THAT INTEL WAS NEGLIGENT REGARDING THE DESIGN OR MANUFACTURE OF THE MATERIALS. +Termination. Your right to use the Software is terminated in the event of your +breach of this license. -LIMITATION OF LIABILITY. IN NO EVENT WILL INTEL BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, -EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR -SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF -LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN -ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. YOU AGREE TO -INDEMNIFY AND HOLD INTEL HARMLESS AGAINST ANY CLAIMS AND EXPENSES RESULTING FROM YOUR USE OR UNAUTHORIZED -USE OF THE SOFTWARE. +Feedback. Should you provide Intel with comments, modifications, corrections, +enhancements or other input ("Feedback") related to the Software, Intel will be +free to use, disclose, reproduce, license or otherwise distribute or exploit the +Feedback in its sole discretion without any obligations or restrictions of any +kind, including without limitation, intellectual property rights or licensing +obligations. -No support. Intel may make changes to the Software, at any time without notice, and is not obligated to -support, update or provide training for the Software. +Compliance with laws. You agree to comply with all relevant laws and regulations +governing your use, transfer, import or export (or prohibition thereof) of the +Software. -Termination. Intel may terminate your right to use the Software in the event of your breach of this Agreement -and you fail to cure the breach within a reasonable period of time. - -Feedback. Should you provide Intel with comments, modifications, corrections, enhancements or other input -("Feedback") related to the Software Intel will be free to use, disclose, reproduce, license or otherwise -distribute or exploit the Feedback in its sole discretion without any obligations or restrictions of any -kind, including without limitation, intellectual property rights or licensing obligations. - -Compliance with laws. You agree to comply with all relevant laws and regulations governing your use, -transfer, import or export (or prohibition thereof) of the Software. - -Governing law. All disputes will be governed by the laws of the United States of America and the State -of Delaware without reference to conflict of law principles and subject to the exclusive jurisdiction of -the state or federal courts sitting in the State of Delaware, and each party agrees that it submits to -the personal jurisdiction and venue of those courts and waives any objections. The United Nations -Convention on Contracts for the International Sale of Goods (1980) is specifically excluded and will -not apply to the Software. - -*Other names and brands may be claimed as the property of others.
\ No newline at end of file +Governing law. All disputes will be governed by the laws of the United States of +America and the State of Delaware without reference to conflict of law +principles and subject to the exclusive jurisdiction of the state or federal +courts sitting in the State of Delaware, and each party agrees that it submits +to the personal jurisdiction and venue of those courts and waives any +objections. The United Nations Convention on Contracts for the International +Sale of Goods (1980) is specifically excluded and will not apply to the +Software. diff --git a/custom-licenses/syrah b/custom-licenses/syrah new file mode 100644 index 00000000..5cd7f4e6 --- /dev/null +++ b/custom-licenses/syrah @@ -0,0 +1,32 @@ +Copyright (c) 2009, Stanford University, and authors listed below. +All rights reserved. + +Original authors: + Solomon Boulos + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are +met: + +Redistributions of source code must retain the above copyright notice, +this list of conditions and the following disclaimer. + +Redistributions in binary form must reproduce the above copyright +notice, this list of conditions and the following disclaimer in the +documentation and/or other materials provided with the distribution. + +Neither the name of Stanford University nor the names of its +contributors may be used to endorse or promote products derived from +this software without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
\ No newline at end of file diff --git a/documentation/dpcpp-compiler.md b/documentation/dpcpp-compiler.md new file mode 100644 index 00000000..b709a685 --- /dev/null +++ b/documentation/dpcpp-compiler.md @@ -0,0 +1,107 @@ +Intel(R) oneAPI DPC++/C++ Compiler (ICX) toolchain +========================================================================== + +Get Started with the Intel oneAPI DPC++/C++ Compiler: + +https://www.intel.com/content/www/us/en/developer/tools/oneapi/dpc-compiler.html# + + +Getting Started +=============== + +Clone the required layers and include them in bblayers.conf: + +``` +git clone https://git.openembedded.org/openembedded-core +git clone https://git.openembedded.org/bitbake +git clone https://git.openembedded.org/meta-openembedded +git clone https://github.com/kraj/meta-clang.git +git clone https://git.yoctoproject.org/meta-intel + +$ source openembedded-core/oe-init-build-env + +$ bitbake-layers add-layer ../meta-openembedded/meta-oe/ +$ bitbake-layers add-layer ../meta-intel +$ bitbake-layers add-layer ../meta-clang +``` + +Distro +====== + +Note that oneAPI DPC++/C++ compiler currently only works when the vendor string is "oe". + +``` +DISTRO ?= "nodistro" +``` + +MACHINE configuration +===================== + +``` +MACHINE ?= "intel-skylake-64" +``` + +Package installation +==================== + +``` +# To include OpenCL driver that might be needed when compiling SYCL programs, include: +IMAGE_INSTALL:append = " intel-compute-runtime intel-graphics-compiler" + +# To install only runtime libraries, include: +IMAGE_INSTALL:append = " intel-oneapi-dpcpp-cpp-runtime intel-oneapi-dpcpp-cpp-runtime-dev" + +# To install the toolchain, include: +IMAGE_INSTALL:append = " intel-oneapi-dpcpp-cpp intel-oneapi-dpcpp-cpp-dev" +``` +in local.conf. + +Build an image +============== + +``` +$ bitbake core-image-minimal +``` + +Including oneAPI C++/DPC++ compiler in generated SDK toolchain +============================================================== + +The compiler is not included in the generated SDK by default. If it is expected to be part of SDK, add ICXSDK = "1" in local.conf: + +``` +ICXSDK = "1" +``` + +Generate SDK: +``` +bitbake core-image-minimal -c populate_sdk +``` + + +To setup PATH variables on target +================================= + +Once image is booted successfully, some variables would need to be exported to make sure compiler can be used: + +``` +$ source /opt/intel/oneapi/compiler/2022.1.0/env/vars.sh + +$ mkdir -p /lib64 + +$ ln -sf /lib/ld-linux-x86-64.so.2 /lib64/ld-linux-x86-64.so.2 +``` + +Build application and run +========================= + +To compile a sycl application, for example: + +``` +$ icpx --target=x86_64-oe-linux -fsycl simple-sycl-app.c -o simple-sycl-app +``` + +To run: + +``` +$ ./simple-sycl-app +``` diff --git a/dynamic-layers/clang-layer/recipes-core/dnn/onednn_2.1.1.bb b/dynamic-layers/clang-layer/recipes-core/dnn/onednn_3.4.bb index 5b96ea0e..afea70df 100644 --- a/dynamic-layers/clang-layer/recipes-core/dnn/onednn_2.1.1.bb +++ b/dynamic-layers/clang-layer/recipes-core/dnn/onednn_3.4.bb @@ -3,19 +3,20 @@ SUMMARY = "Deep Neural Network Library" DESCRIPTION = "This software is a user mode library that accelerates\ deep-learning applications and frameworks on Intel architecture." LICENSE = "Apache-2.0 & BSD-3-Clause & BSL-1.0" -LIC_FILES_CHKSUM = "file://LICENSE;md5=8e17c0f9656ebaf0c380d9b22707c846 \ +LIC_FILES_CHKSUM = "file://LICENSE;md5=3b64000f6e7d52516017622a37a94ce9 \ file://tests/gtests/gtest/LICENSE;md5=cbbd27594afd089daa160d3a16dd515a \ file://src/cpu/x64/xbyak/COPYRIGHT;md5=3b9bf048d063d54cdb28964db558bcc7 \ - file://src/cpu/x64/jit_utils/jitprofiling/LICENSE.BSD;md5=e671ff178b24a95a382ba670503c66fb \ - file://doc/assets/mathjax/MathJax.js;endline=17;md5=25a014ad78c3d72a0e15d15f1d007c20 \ + file://src/common/ittnotify/LICENSE.BSD;md5=e671ff178b24a95a382ba670503c66fb \ " SECTION = "lib" inherit pkgconfig cmake ptest +DNN_BRANCH = "rls-v${@'.'.join(d.getVar('PV').split('.')[0:2])}" + S = "${WORKDIR}/git" -SRCREV = "53f53c2bb859d386a40e03f7623a4f9af758c1b3" -SRC_URI = "git://github.com/oneapi-src/oneDNN.git;branch=rls-v2.1 \ +SRCREV = "ecd7fb6d5a0df6503d1691c1754a684b9c769c16" +SRC_URI = "git://github.com/oneapi-src/oneDNN.git;branch=${DNN_BRANCH};protocol=https \ file://run-ptest \ " @@ -24,7 +25,7 @@ UPSTREAM_CHECK_GITTAGREGEX = "^v(?P<pver>(\d+(\.\d+)+))$" CVE_PRODUCT = "intel:math_kernel_library" COMPATIBLE_HOST = '(x86_64).*-linux' -COMPATIBLE_HOST_libc-musl = 'null' +COMPATIBLE_HOST:libc-musl = 'null' EXTRA_OECMAKE += " \ -DDNNL_LIBRARY_TYPE=SHARED \ @@ -33,12 +34,13 @@ EXTRA_OECMAKE += " \ -DDNNL_CPU_RUNTIME=OMP \ -DDNNL_ARCH_OPT_FLAGS="" \ -DCMAKE_SKIP_RPATH=ON \ + -DONEDNN_BUILD_GRAPH=OFF \ " -PACKAGECONFIG ??= "" -PACKAGECONFIG[gpu] = "-DDNNL_GPU_RUNTIME=OCL, , opencl-headers ocl-icd, intel-compute-runtime" +PACKAGECONFIG ??= "gpu" +PACKAGECONFIG[gpu] = "-DDNNL_GPU_RUNTIME=OCL, , opencl-headers virtual/opencl-icd, intel-compute-runtime" -do_install_append () { +do_install:append () { install -d ${D}${bindir}/mkl-dnn/tests/benchdnn/inputs install -m 0755 ${B}/tests/benchdnn/benchdnn ${D}${bindir}/mkl-dnn/tests/benchdnn cp -r ${B}/tests/benchdnn/inputs/* ${D}${bindir}/mkl-dnn/tests/benchdnn/inputs @@ -52,4 +54,4 @@ do_install_ptest () { PACKAGES =+ "${PN}-test" -FILES_${PN}-test = "${bindir}/mkl-dnn/*" +FILES:${PN}-test = "${bindir}/mkl-dnn/*" diff --git a/dynamic-layers/clang-layer/recipes-devtools/clang/files/0001-Fix-debug-info-of-work-item-builtin-translation-745.patch b/dynamic-layers/clang-layer/recipes-devtools/clang/files/0001-Fix-debug-info-of-work-item-builtin-translation-745.patch deleted file mode 100644 index 923b871f..00000000 --- a/dynamic-layers/clang-layer/recipes-devtools/clang/files/0001-Fix-debug-info-of-work-item-builtin-translation-745.patch +++ /dev/null @@ -1,119 +0,0 @@ -From 200c200eb19602ffd7c8f29d0b2df9df1fd311bf Mon Sep 17 00:00:00 2001 -From: Naveen Saini <naveen.kumar.saini@intel.com> -Date: Wed, 7 Apr 2021 17:44:20 +0800 -Subject: [PATCH] Fix debug info of work-item builtin translation (#745) - -debug info of work-item builtins are lost in both llvm IR -> spirv and -spirv -> llvm IR translations. See #744 - -Upstream-Status: Backport [https://github.com/KhronosGroup/SPIRV-LLVM-Translator/commit/c734c5c8bbd3012a09c610e4be68e90cc603c580] -Signed-off-by: Wenju He <wenju.he@intel.com> -Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> ---- - lib/SPIRV/OCL20ToSPIRV.cpp | 5 ++- - lib/SPIRV/SPIRVReader.cpp | 1 + - test/DebugInfo/builtin-get-global-id.ll | 60 +++++++++++++++++++++++++ - 3 files changed, 65 insertions(+), 1 deletion(-) - create mode 100644 test/DebugInfo/builtin-get-global-id.ll - -diff --git a/lib/SPIRV/OCL20ToSPIRV.cpp b/lib/SPIRV/OCL20ToSPIRV.cpp -index 1262c48c..a742c8cf 100644 ---- a/lib/SPIRV/OCL20ToSPIRV.cpp -+++ b/lib/SPIRV/OCL20ToSPIRV.cpp -@@ -1297,11 +1297,14 @@ void OCL20ToSPIRV::transWorkItemBuiltinsToVariables() { - for (auto UI = I.user_begin(), UE = I.user_end(); UI != UE; ++UI) { - auto CI = dyn_cast<CallInst>(*UI); - assert(CI && "invalid instruction"); -- Value *NewValue = new LoadInst(BV, "", CI); -+ const DebugLoc &DLoc = CI->getDebugLoc(); -+ Instruction *NewValue = new LoadInst(BV, "", CI); -+ NewValue->setDebugLoc(DLoc); - LLVM_DEBUG(dbgs() << "Transform: " << *CI << " => " << *NewValue << '\n'); - if (IsVec) { - NewValue = - ExtractElementInst::Create(NewValue, CI->getArgOperand(0), "", CI); -+ NewValue->setDebugLoc(DLoc); - LLVM_DEBUG(dbgs() << *NewValue << '\n'); - } - NewValue->takeName(CI); -diff --git a/lib/SPIRV/SPIRVReader.cpp b/lib/SPIRV/SPIRVReader.cpp -index 16a3dd38..528f6663 100644 ---- a/lib/SPIRV/SPIRVReader.cpp -+++ b/lib/SPIRV/SPIRVReader.cpp -@@ -307,6 +307,7 @@ bool SPIRVToLLVM::transOCLBuiltinFromVariable(GlobalVariable *GV, - auto Replace = [&](std::vector<Value *> Arg, Instruction *I) { - auto Call = CallInst::Create(Func, Arg, "", I); - Call->takeName(I); -+ Call->setDebugLoc(I->getDebugLoc()); - setAttrByCalledFunc(Call); - SPIRVDBG(dbgs() << "[transOCLBuiltinFromVariable] " << *I << " -> " << *Call - << '\n';) -diff --git a/test/DebugInfo/builtin-get-global-id.ll b/test/DebugInfo/builtin-get-global-id.ll -new file mode 100644 -index 00000000..a4a00e63 ---- /dev/null -+++ b/test/DebugInfo/builtin-get-global-id.ll -@@ -0,0 +1,60 @@ -+; Check debug info of builtin get_global_id is preserved from LLVM IR to spirv -+; and spirv to LLVM IR translation. -+ -+; Original .cl source: -+; kernel void test() { -+; size_t gid = get_global_id(0); -+; } -+ -+; Command line: -+; ./clang -cc1 1.cl -triple spir64 -cl-std=cl2.0 -emit-llvm -finclude-default-header -debug-info-kind=line-tables-only -O0 -+ -+; RUN: llvm-as %s -o %t.bc -+; RUN: llvm-spirv %t.bc -spirv-text -o - | FileCheck %s --check-prefix CHECK-SPIRV -+; RUN: llvm-spirv %t.bc -o %t.spv -+; RUN: llvm-spirv -r %t.spv -o - | llvm-dis -o - | FileCheck %s -+ -+target datalayout = "e-i64:64-v16:16-v24:32-v32:32-v48:64-v96:128-v192:256-v256:256-v512:512-v1024:1024" -+target triple = "spir64" -+ -+; CHECK-SPIRV: ExtInst {{.*}} DebugScope -+; CHECK-SPIRV-NEXT: Line {{[0-9]+}} 2 16 -+; CHECK-SPIRV-NEXT: Load {{[0-9]+}} [[LoadRes:[0-9]+]] -+; CHECK-SPIRV-NEXT: CompositeExtract {{[0-9]+}} {{[0-9]+}} [[LoadRes]] 0 -+ -+; Function Attrs: convergent noinline norecurse nounwind optnone -+define spir_kernel void @test() #0 !dbg !7 !kernel_arg_addr_space !2 !kernel_arg_access_qual !2 !kernel_arg_type !2 !kernel_arg_base_type !2 !kernel_arg_type_qual !2 { -+entry: -+ %gid = alloca i64, align 8 -+ %call = call spir_func i64 @_Z13get_global_idj(i32 0) #2, !dbg !10 -+; CHECK: %call = call spir_func i64 @_Z13get_global_idj(i32 0) #1, !dbg [[DBG:![0-9]+]] -+ store i64 %call, i64* %gid, align 8, !dbg !11 -+ ret void, !dbg !12 -+} -+ -+; Function Attrs: convergent nounwind readnone -+declare spir_func i64 @_Z13get_global_idj(i32) #1 -+ -+attributes #0 = { convergent noinline norecurse nounwind optnone "correctly-rounded-divide-sqrt-fp-math"="false" "disable-tail-calls"="false" "frame-pointer"="none" "less-precise-fpmad"="false" "min-legal-vector-width"="0" "no-infs-fp-math"="false" "no-jump-tables"="false" "no-nans-fp-math"="false" "no-signed-zeros-fp-math"="false" "no-trapping-math"="true" "stack-protector-buffer-size"="8" "uniform-work-group-size"="false" "unsafe-fp-math"="false" "use-soft-float"="false" } -+attributes #1 = { convergent nounwind readnone "correctly-rounded-divide-sqrt-fp-math"="false" "disable-tail-calls"="false" "frame-pointer"="none" "less-precise-fpmad"="false" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "no-signed-zeros-fp-math"="false" "no-trapping-math"="true" "stack-protector-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="false" } -+attributes #2 = { convergent nounwind readnone } -+ -+!llvm.dbg.cu = !{!0} -+!llvm.module.flags = !{!3, !4} -+!opencl.ocl.version = !{!5} -+!opencl.spir.version = !{!5} -+!llvm.ident = !{!6} -+ -+!0 = distinct !DICompileUnit(language: DW_LANG_C99, file: !1, producer: "clang version 12.0.0 (https://github.com/llvm/llvm-project.git b5bc56da8aa23dc57db9d286b0591dbcf9b1bdd3)", isOptimized: false, runtimeVersion: 0, emissionKind: LineTablesOnly, enums: !2, nameTableKind: None) -+!1 = !DIFile(filename: "<stdin>", directory: "") -+!2 = !{} -+!3 = !{i32 2, !"Debug Info Version", i32 3} -+!4 = !{i32 1, !"wchar_size", i32 4} -+!5 = !{i32 2, i32 0} -+!6 = !{!"clang version 12.0.0 (https://github.com/llvm/llvm-project.git b5bc56da8aa23dc57db9d286b0591dbcf9b1bdd3)"} -+!7 = distinct !DISubprogram(name: "test", scope: !8, file: !8, line: 1, type: !9, scopeLine: 1, flags: DIFlagPrototyped, spFlags: DISPFlagDefinition, unit: !0, retainedNodes: !2) -+!8 = !DIFile(filename: "1.cl", directory: "") -+!9 = !DISubroutineType(types: !2) -+!10 = !DILocation(line: 2, column: 16, scope: !7) -+!11 = !DILocation(line: 2, column: 10, scope: !7) -+!12 = !DILocation(line: 3, column: 1, scope: !7) --- -2.17.1 - diff --git a/dynamic-layers/clang-layer/recipes-devtools/clang/files/0001-Memory-leak-fix-for-Managed-Static-Mutex.patch b/dynamic-layers/clang-layer/recipes-devtools/clang/files/0001-Memory-leak-fix-for-Managed-Static-Mutex.patch deleted file mode 100644 index 103dad5e..00000000 --- a/dynamic-layers/clang-layer/recipes-devtools/clang/files/0001-Memory-leak-fix-for-Managed-Static-Mutex.patch +++ /dev/null @@ -1,35 +0,0 @@ -From c86c43b70e029b102543e8a85d269cbeb5c00279 Mon Sep 17 00:00:00 2001 -From: juanrod2 <> -Date: Tue, 22 Dec 2020 08:33:08 +0800 -Subject: [PATCH] Memory leak fix for Managed Static Mutex - -Upstream-Status: Backport [Taken from opencl-clang patches; https://github.com/intel/opencl-clang/blob/ocl-open-100/patches/llvm/0001-Memory-leak-fix-for-Managed-Static-Mutex.patch] - -Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> - -Cleaning a mutex inside ManagedStatic llvm class. ---- - llvm/lib/Support/ManagedStatic.cpp | 6 +++++- - 1 file changed, 5 insertions(+), 1 deletion(-) - -diff --git a/llvm/lib/Support/ManagedStatic.cpp b/llvm/lib/Support/ManagedStatic.cpp -index 053493f72fb5..6571580ccecf 100644 ---- a/llvm/lib/Support/ManagedStatic.cpp -+++ b/llvm/lib/Support/ManagedStatic.cpp -@@ -76,8 +76,12 @@ void ManagedStaticBase::destroy() const { - - /// llvm_shutdown - Deallocate and destroy all ManagedStatic variables. - void llvm::llvm_shutdown() { -- std::lock_guard<std::recursive_mutex> Lock(*getManagedStaticMutex()); -+ getManagedStaticMutex()->lock(); - - while (StaticList) - StaticList->destroy(); -+ -+ getManagedStaticMutex()->unlock(); -+ delete ManagedStaticMutex; -+ ManagedStaticMutex = nullptr; - } --- -2.29.2 - diff --git a/dynamic-layers/clang-layer/recipes-devtools/clang/files/0002-Add-cl_khr_extended_subgroup-extensions.patch b/dynamic-layers/clang-layer/recipes-devtools/clang/files/0002-Add-cl_khr_extended_subgroup-extensions.patch deleted file mode 100644 index cbe492c4..00000000 --- a/dynamic-layers/clang-layer/recipes-devtools/clang/files/0002-Add-cl_khr_extended_subgroup-extensions.patch +++ /dev/null @@ -1,812 +0,0 @@ -From 27d47f1a17c8921b07acc8cdc26e38cc609de4a9 Mon Sep 17 00:00:00 2001 -From: Naveen Saini <naveen.kumar.saini@intel.com> -Date: Wed, 7 Apr 2021 16:38:38 +0800 -Subject: [PATCH 2/2] Add cl_khr_extended_subgroup extensions. - -Added extensions and their function declarations into -the standard header. - -Patch by Piotr Fusik! - -Tags: #clang - -Upstream-Status: Backport [https://github.com/llvm/llvm-project/commit/4a4402f0d72167477a6252e4c3daf5089ebc8f9a] -Signed-off-by: Anastasia Stulova <anastasia.stulova@arm.com> -Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> ---- - .../include/clang/Basic/OpenCLExtensions.def | 7 + - clang/lib/Headers/opencl-c.h | 668 ++++++++++++++++++ - clang/test/SemaOpenCL/extension-version.cl | 83 +++ - 3 files changed, 758 insertions(+) - -diff --git a/clang/include/clang/Basic/OpenCLExtensions.def b/clang/include/clang/Basic/OpenCLExtensions.def -index 608f78a13eef..d1574164f9b2 100644 ---- a/clang/include/clang/Basic/OpenCLExtensions.def -+++ b/clang/include/clang/Basic/OpenCLExtensions.def -@@ -74,6 +74,13 @@ OPENCLEXT_INTERNAL(cl_khr_mipmap_image_writes, 200, ~0U) - OPENCLEXT_INTERNAL(cl_khr_srgb_image_writes, 200, ~0U) - OPENCLEXT_INTERNAL(cl_khr_subgroups, 200, ~0U) - OPENCLEXT_INTERNAL(cl_khr_terminate_context, 200, ~0U) -+OPENCLEXT_INTERNAL(cl_khr_subgroup_extended_types, 200, ~0U) -+OPENCLEXT_INTERNAL(cl_khr_subgroup_non_uniform_vote, 200, ~0U) -+OPENCLEXT_INTERNAL(cl_khr_subgroup_ballot, 200, ~0U) -+OPENCLEXT_INTERNAL(cl_khr_subgroup_non_uniform_arithmetic, 200, ~0U) -+OPENCLEXT_INTERNAL(cl_khr_subgroup_shuffle, 200, ~0U) -+OPENCLEXT_INTERNAL(cl_khr_subgroup_shuffle_relative, 200, ~0U) -+OPENCLEXT_INTERNAL(cl_khr_subgroup_clustered_reduce, 200, ~0U) - - // Clang Extensions. - OPENCLEXT_INTERNAL(cl_clang_storage_class_specifiers, 100, ~0U) -diff --git a/clang/lib/Headers/opencl-c.h b/clang/lib/Headers/opencl-c.h -index 93a946cec5b1..67d900eb1c3d 100644 ---- a/clang/lib/Headers/opencl-c.h -+++ b/clang/lib/Headers/opencl-c.h -@@ -17530,6 +17530,674 @@ double __ovld __conv sub_group_scan_inclusive_max(double x); - - #endif //cl_khr_subgroups cl_intel_subgroups - -+#if defined(cl_khr_subgroup_extended_types) -+char __ovld __conv sub_group_broadcast( char value, uint index ); -+char2 __ovld __conv sub_group_broadcast( char2 value, uint index ); -+char3 __ovld __conv sub_group_broadcast( char3 value, uint index ); -+char4 __ovld __conv sub_group_broadcast( char4 value, uint index ); -+char8 __ovld __conv sub_group_broadcast( char8 value, uint index ); -+char16 __ovld __conv sub_group_broadcast( char16 value, uint index ); -+ -+uchar __ovld __conv sub_group_broadcast( uchar value, uint index ); -+uchar2 __ovld __conv sub_group_broadcast( uchar2 value, uint index ); -+uchar3 __ovld __conv sub_group_broadcast( uchar3 value, uint index ); -+uchar4 __ovld __conv sub_group_broadcast( uchar4 value, uint index ); -+uchar8 __ovld __conv sub_group_broadcast( uchar8 value, uint index ); -+uchar16 __ovld __conv sub_group_broadcast( uchar16 value, uint index ); -+ -+short __ovld __conv sub_group_broadcast( short value, uint index ); -+short2 __ovld __conv sub_group_broadcast( short2 value, uint index ); -+short3 __ovld __conv sub_group_broadcast( short3 value, uint index ); -+short4 __ovld __conv sub_group_broadcast( short4 value, uint index ); -+short8 __ovld __conv sub_group_broadcast( short8 value, uint index ); -+short16 __ovld __conv sub_group_broadcast( short16 value, uint index ); -+ -+ushort __ovld __conv sub_group_broadcast( ushort value, uint index ); -+ushort2 __ovld __conv sub_group_broadcast( ushort2 value, uint index ); -+ushort3 __ovld __conv sub_group_broadcast( ushort3 value, uint index ); -+ushort4 __ovld __conv sub_group_broadcast( ushort4 value, uint index ); -+ushort8 __ovld __conv sub_group_broadcast( ushort8 value, uint index ); -+ushort16 __ovld __conv sub_group_broadcast( ushort16 value, uint index ); -+ -+// scalar int broadcast is part of cl_khr_subgroups -+int2 __ovld __conv sub_group_broadcast( int2 value, uint index ); -+int3 __ovld __conv sub_group_broadcast( int3 value, uint index ); -+int4 __ovld __conv sub_group_broadcast( int4 value, uint index ); -+int8 __ovld __conv sub_group_broadcast( int8 value, uint index ); -+int16 __ovld __conv sub_group_broadcast( int16 value, uint index ); -+ -+// scalar uint broadcast is part of cl_khr_subgroups -+uint2 __ovld __conv sub_group_broadcast( uint2 value, uint index ); -+uint3 __ovld __conv sub_group_broadcast( uint3 value, uint index ); -+uint4 __ovld __conv sub_group_broadcast( uint4 value, uint index ); -+uint8 __ovld __conv sub_group_broadcast( uint8 value, uint index ); -+uint16 __ovld __conv sub_group_broadcast( uint16 value, uint index ); -+ -+// scalar long broadcast is part of cl_khr_subgroups -+long2 __ovld __conv sub_group_broadcast( long2 value, uint index ); -+long3 __ovld __conv sub_group_broadcast( long3 value, uint index ); -+long4 __ovld __conv sub_group_broadcast( long4 value, uint index ); -+long8 __ovld __conv sub_group_broadcast( long8 value, uint index ); -+long16 __ovld __conv sub_group_broadcast( long16 value, uint index ); -+ -+// scalar ulong broadcast is part of cl_khr_subgroups -+ulong2 __ovld __conv sub_group_broadcast( ulong2 value, uint index ); -+ulong3 __ovld __conv sub_group_broadcast( ulong3 value, uint index ); -+ulong4 __ovld __conv sub_group_broadcast( ulong4 value, uint index ); -+ulong8 __ovld __conv sub_group_broadcast( ulong8 value, uint index ); -+ulong16 __ovld __conv sub_group_broadcast( ulong16 value, uint index ); -+ -+// scalar float broadcast is part of cl_khr_subgroups -+float2 __ovld __conv sub_group_broadcast( float2 value, uint index ); -+float3 __ovld __conv sub_group_broadcast( float3 value, uint index ); -+float4 __ovld __conv sub_group_broadcast( float4 value, uint index ); -+float8 __ovld __conv sub_group_broadcast( float8 value, uint index ); -+float16 __ovld __conv sub_group_broadcast( float16 value, uint index ); -+ -+char __ovld __conv sub_group_reduce_add( char value ); -+uchar __ovld __conv sub_group_reduce_add( uchar value ); -+short __ovld __conv sub_group_reduce_add( short value ); -+ushort __ovld __conv sub_group_reduce_add( ushort value ); -+ -+char __ovld __conv sub_group_reduce_min( char value ); -+uchar __ovld __conv sub_group_reduce_min( uchar value ); -+short __ovld __conv sub_group_reduce_min( short value ); -+ushort __ovld __conv sub_group_reduce_min( ushort value ); -+ -+char __ovld __conv sub_group_reduce_max( char value ); -+uchar __ovld __conv sub_group_reduce_max( uchar value ); -+short __ovld __conv sub_group_reduce_max( short value ); -+ushort __ovld __conv sub_group_reduce_max( ushort value ); -+ -+char __ovld __conv sub_group_scan_inclusive_add( char value ); -+uchar __ovld __conv sub_group_scan_inclusive_add( uchar value ); -+short __ovld __conv sub_group_scan_inclusive_add( short value ); -+ushort __ovld __conv sub_group_scan_inclusive_add( ushort value ); -+ -+char __ovld __conv sub_group_scan_inclusive_min( char value ); -+uchar __ovld __conv sub_group_scan_inclusive_min( uchar value ); -+short __ovld __conv sub_group_scan_inclusive_min( short value ); -+ushort __ovld __conv sub_group_scan_inclusive_min( ushort value ); -+ -+char __ovld __conv sub_group_scan_inclusive_max( char value ); -+uchar __ovld __conv sub_group_scan_inclusive_max( uchar value ); -+short __ovld __conv sub_group_scan_inclusive_max( short value ); -+ushort __ovld __conv sub_group_scan_inclusive_max( ushort value ); -+ -+char __ovld __conv sub_group_scan_exclusive_add( char value ); -+uchar __ovld __conv sub_group_scan_exclusive_add( uchar value ); -+short __ovld __conv sub_group_scan_exclusive_add( short value ); -+ushort __ovld __conv sub_group_scan_exclusive_add( ushort value ); -+ -+char __ovld __conv sub_group_scan_exclusive_min( char value ); -+uchar __ovld __conv sub_group_scan_exclusive_min( uchar value ); -+short __ovld __conv sub_group_scan_exclusive_min( short value ); -+ushort __ovld __conv sub_group_scan_exclusive_min( ushort value ); -+ -+char __ovld __conv sub_group_scan_exclusive_max( char value ); -+uchar __ovld __conv sub_group_scan_exclusive_max( uchar value ); -+short __ovld __conv sub_group_scan_exclusive_max( short value ); -+ushort __ovld __conv sub_group_scan_exclusive_max( ushort value ); -+ -+#if defined(cl_khr_fp16) -+// scalar half broadcast is part of cl_khr_subgroups -+half2 __ovld __conv sub_group_broadcast( half2 value, uint index ); -+half3 __ovld __conv sub_group_broadcast( half3 value, uint index ); -+half4 __ovld __conv sub_group_broadcast( half4 value, uint index ); -+half8 __ovld __conv sub_group_broadcast( half8 value, uint index ); -+half16 __ovld __conv sub_group_broadcast( half16 value, uint index ); -+#endif // cl_khr_fp16 -+ -+#if defined(cl_khr_fp64) -+// scalar double broadcast is part of cl_khr_subgroups -+double2 __ovld __conv sub_group_broadcast( double2 value, uint index ); -+double3 __ovld __conv sub_group_broadcast( double3 value, uint index ); -+double4 __ovld __conv sub_group_broadcast( double4 value, uint index ); -+double8 __ovld __conv sub_group_broadcast( double8 value, uint index ); -+double16 __ovld __conv sub_group_broadcast( double16 value, uint index ); -+#endif // cl_khr_fp64 -+ -+#endif // cl_khr_subgroup_extended_types -+ -+#if defined(cl_khr_subgroup_non_uniform_vote) -+int __ovld sub_group_elect(void); -+int __ovld sub_group_non_uniform_all( int predicate ); -+int __ovld sub_group_non_uniform_any( int predicate ); -+ -+int __ovld sub_group_non_uniform_all_equal( char value ); -+int __ovld sub_group_non_uniform_all_equal( uchar value ); -+int __ovld sub_group_non_uniform_all_equal( short value ); -+int __ovld sub_group_non_uniform_all_equal( ushort value ); -+int __ovld sub_group_non_uniform_all_equal( int value ); -+int __ovld sub_group_non_uniform_all_equal( uint value ); -+int __ovld sub_group_non_uniform_all_equal( long value ); -+int __ovld sub_group_non_uniform_all_equal( ulong value ); -+int __ovld sub_group_non_uniform_all_equal( float value ); -+ -+#if defined(cl_khr_fp16) -+int __ovld sub_group_non_uniform_all_equal( half value ); -+#endif // cl_khr_fp16 -+ -+#if defined(cl_khr_fp64) -+int __ovld sub_group_non_uniform_all_equal( double value ); -+#endif // cl_khr_fp64 -+ -+#endif // cl_khr_subgroup_non_uniform_vote -+ -+#if defined(cl_khr_subgroup_ballot) -+char __ovld sub_group_non_uniform_broadcast( char value, uint index ); -+char2 __ovld sub_group_non_uniform_broadcast( char2 value, uint index ); -+char3 __ovld sub_group_non_uniform_broadcast( char3 value, uint index ); -+char4 __ovld sub_group_non_uniform_broadcast( char4 value, uint index ); -+char8 __ovld sub_group_non_uniform_broadcast( char8 value, uint index ); -+char16 __ovld sub_group_non_uniform_broadcast( char16 value, uint index ); -+ -+uchar __ovld sub_group_non_uniform_broadcast( uchar value, uint index ); -+uchar2 __ovld sub_group_non_uniform_broadcast( uchar2 value, uint index ); -+uchar3 __ovld sub_group_non_uniform_broadcast( uchar3 value, uint index ); -+uchar4 __ovld sub_group_non_uniform_broadcast( uchar4 value, uint index ); -+uchar8 __ovld sub_group_non_uniform_broadcast( uchar8 value, uint index ); -+uchar16 __ovld sub_group_non_uniform_broadcast( uchar16 value, uint index ); -+ -+short __ovld sub_group_non_uniform_broadcast( short value, uint index ); -+short2 __ovld sub_group_non_uniform_broadcast( short2 value, uint index ); -+short3 __ovld sub_group_non_uniform_broadcast( short3 value, uint index ); -+short4 __ovld sub_group_non_uniform_broadcast( short4 value, uint index ); -+short8 __ovld sub_group_non_uniform_broadcast( short8 value, uint index ); -+short16 __ovld sub_group_non_uniform_broadcast( short16 value, uint index ); -+ -+ushort __ovld sub_group_non_uniform_broadcast( ushort value, uint index ); -+ushort2 __ovld sub_group_non_uniform_broadcast( ushort2 value, uint index ); -+ushort3 __ovld sub_group_non_uniform_broadcast( ushort3 value, uint index ); -+ushort4 __ovld sub_group_non_uniform_broadcast( ushort4 value, uint index ); -+ushort8 __ovld sub_group_non_uniform_broadcast( ushort8 value, uint index ); -+ushort16 __ovld sub_group_non_uniform_broadcast( ushort16 value, uint index ); -+ -+int __ovld sub_group_non_uniform_broadcast( int value, uint index ); -+int2 __ovld sub_group_non_uniform_broadcast( int2 value, uint index ); -+int3 __ovld sub_group_non_uniform_broadcast( int3 value, uint index ); -+int4 __ovld sub_group_non_uniform_broadcast( int4 value, uint index ); -+int8 __ovld sub_group_non_uniform_broadcast( int8 value, uint index ); -+int16 __ovld sub_group_non_uniform_broadcast( int16 value, uint index ); -+ -+uint __ovld sub_group_non_uniform_broadcast( uint value, uint index ); -+uint2 __ovld sub_group_non_uniform_broadcast( uint2 value, uint index ); -+uint3 __ovld sub_group_non_uniform_broadcast( uint3 value, uint index ); -+uint4 __ovld sub_group_non_uniform_broadcast( uint4 value, uint index ); -+uint8 __ovld sub_group_non_uniform_broadcast( uint8 value, uint index ); -+uint16 __ovld sub_group_non_uniform_broadcast( uint16 value, uint index ); -+ -+long __ovld sub_group_non_uniform_broadcast( long value, uint index ); -+long2 __ovld sub_group_non_uniform_broadcast( long2 value, uint index ); -+long3 __ovld sub_group_non_uniform_broadcast( long3 value, uint index ); -+long4 __ovld sub_group_non_uniform_broadcast( long4 value, uint index ); -+long8 __ovld sub_group_non_uniform_broadcast( long8 value, uint index ); -+long16 __ovld sub_group_non_uniform_broadcast( long16 value, uint index ); -+ -+ulong __ovld sub_group_non_uniform_broadcast( ulong value, uint index ); -+ulong2 __ovld sub_group_non_uniform_broadcast( ulong2 value, uint index ); -+ulong3 __ovld sub_group_non_uniform_broadcast( ulong3 value, uint index ); -+ulong4 __ovld sub_group_non_uniform_broadcast( ulong4 value, uint index ); -+ulong8 __ovld sub_group_non_uniform_broadcast( ulong8 value, uint index ); -+ulong16 __ovld sub_group_non_uniform_broadcast( ulong16 value, uint index ); -+ -+float __ovld sub_group_non_uniform_broadcast( float value, uint index ); -+float2 __ovld sub_group_non_uniform_broadcast( float2 value, uint index ); -+float3 __ovld sub_group_non_uniform_broadcast( float3 value, uint index ); -+float4 __ovld sub_group_non_uniform_broadcast( float4 value, uint index ); -+float8 __ovld sub_group_non_uniform_broadcast( float8 value, uint index ); -+float16 __ovld sub_group_non_uniform_broadcast( float16 value, uint index ); -+ -+char __ovld sub_group_broadcast_first( char value ); -+uchar __ovld sub_group_broadcast_first( uchar value ); -+short __ovld sub_group_broadcast_first( short value ); -+ushort __ovld sub_group_broadcast_first( ushort value ); -+int __ovld sub_group_broadcast_first( int value ); -+uint __ovld sub_group_broadcast_first( uint value ); -+long __ovld sub_group_broadcast_first( long value ); -+ulong __ovld sub_group_broadcast_first( ulong value ); -+float __ovld sub_group_broadcast_first( float value ); -+ -+uint4 __ovld sub_group_ballot( int predicate ); -+int __ovld __cnfn sub_group_inverse_ballot( uint4 value ); -+int __ovld __cnfn sub_group_ballot_bit_extract( uint4 value, uint index ); -+uint __ovld __cnfn sub_group_ballot_bit_count( uint4 value ); -+ -+uint __ovld sub_group_ballot_inclusive_scan( uint4 value ); -+uint __ovld sub_group_ballot_exclusive_scan( uint4 value ); -+uint __ovld sub_group_ballot_find_lsb( uint4 value ); -+uint __ovld sub_group_ballot_find_msb( uint4 value ); -+ -+uint4 __ovld __cnfn get_sub_group_eq_mask(void); -+uint4 __ovld __cnfn get_sub_group_ge_mask(void); -+uint4 __ovld __cnfn get_sub_group_gt_mask(void); -+uint4 __ovld __cnfn get_sub_group_le_mask(void); -+uint4 __ovld __cnfn get_sub_group_lt_mask(void); -+ -+#if defined(cl_khr_fp16) -+half __ovld sub_group_non_uniform_broadcast( half value, uint index ); -+half2 __ovld sub_group_non_uniform_broadcast( half2 value, uint index ); -+half3 __ovld sub_group_non_uniform_broadcast( half3 value, uint index ); -+half4 __ovld sub_group_non_uniform_broadcast( half4 value, uint index ); -+half8 __ovld sub_group_non_uniform_broadcast( half8 value, uint index ); -+half16 __ovld sub_group_non_uniform_broadcast( half16 value, uint index ); -+ -+half __ovld sub_group_broadcast_first( half value ); -+#endif // cl_khr_fp16 -+ -+#if defined(cl_khr_fp64) -+double __ovld sub_group_non_uniform_broadcast( double value, uint index ); -+double2 __ovld sub_group_non_uniform_broadcast( double2 value, uint index ); -+double3 __ovld sub_group_non_uniform_broadcast( double3 value, uint index ); -+double4 __ovld sub_group_non_uniform_broadcast( double4 value, uint index ); -+double8 __ovld sub_group_non_uniform_broadcast( double8 value, uint index ); -+double16 __ovld sub_group_non_uniform_broadcast( double16 value, uint index ); -+ -+double __ovld sub_group_broadcast_first( double value ); -+#endif // cl_khr_fp64 -+ -+#endif // cl_khr_subgroup_ballot -+ -+#if defined(cl_khr_subgroup_non_uniform_arithmetic) -+char __ovld sub_group_non_uniform_reduce_add( char value ); -+uchar __ovld sub_group_non_uniform_reduce_add( uchar value ); -+short __ovld sub_group_non_uniform_reduce_add( short value ); -+ushort __ovld sub_group_non_uniform_reduce_add( ushort value ); -+int __ovld sub_group_non_uniform_reduce_add( int value ); -+uint __ovld sub_group_non_uniform_reduce_add( uint value ); -+long __ovld sub_group_non_uniform_reduce_add( long value ); -+ulong __ovld sub_group_non_uniform_reduce_add( ulong value ); -+float __ovld sub_group_non_uniform_reduce_add( float value ); -+ -+char __ovld sub_group_non_uniform_reduce_mul( char value ); -+uchar __ovld sub_group_non_uniform_reduce_mul( uchar value ); -+short __ovld sub_group_non_uniform_reduce_mul( short value ); -+ushort __ovld sub_group_non_uniform_reduce_mul( ushort value ); -+int __ovld sub_group_non_uniform_reduce_mul( int value ); -+uint __ovld sub_group_non_uniform_reduce_mul( uint value ); -+long __ovld sub_group_non_uniform_reduce_mul( long value ); -+ulong __ovld sub_group_non_uniform_reduce_mul( ulong value ); -+float __ovld sub_group_non_uniform_reduce_mul( float value ); -+ -+char __ovld sub_group_non_uniform_reduce_min( char value ); -+uchar __ovld sub_group_non_uniform_reduce_min( uchar value ); -+short __ovld sub_group_non_uniform_reduce_min( short value ); -+ushort __ovld sub_group_non_uniform_reduce_min( ushort value ); -+int __ovld sub_group_non_uniform_reduce_min( int value ); -+uint __ovld sub_group_non_uniform_reduce_min( uint value ); -+long __ovld sub_group_non_uniform_reduce_min( long value ); -+ulong __ovld sub_group_non_uniform_reduce_min( ulong value ); -+float __ovld sub_group_non_uniform_reduce_min( float value ); -+ -+char __ovld sub_group_non_uniform_reduce_max( char value ); -+uchar __ovld sub_group_non_uniform_reduce_max( uchar value ); -+short __ovld sub_group_non_uniform_reduce_max( short value ); -+ushort __ovld sub_group_non_uniform_reduce_max( ushort value ); -+int __ovld sub_group_non_uniform_reduce_max( int value ); -+uint __ovld sub_group_non_uniform_reduce_max( uint value ); -+long __ovld sub_group_non_uniform_reduce_max( long value ); -+ulong __ovld sub_group_non_uniform_reduce_max( ulong value ); -+float __ovld sub_group_non_uniform_reduce_max( float value ); -+ -+char __ovld sub_group_non_uniform_scan_inclusive_add( char value ); -+uchar __ovld sub_group_non_uniform_scan_inclusive_add( uchar value ); -+short __ovld sub_group_non_uniform_scan_inclusive_add( short value ); -+ushort __ovld sub_group_non_uniform_scan_inclusive_add( ushort value ); -+int __ovld sub_group_non_uniform_scan_inclusive_add( int value ); -+uint __ovld sub_group_non_uniform_scan_inclusive_add( uint value ); -+long __ovld sub_group_non_uniform_scan_inclusive_add( long value ); -+ulong __ovld sub_group_non_uniform_scan_inclusive_add( ulong value ); -+float __ovld sub_group_non_uniform_scan_inclusive_add( float value ); -+ -+char __ovld sub_group_non_uniform_scan_inclusive_mul( char value ); -+uchar __ovld sub_group_non_uniform_scan_inclusive_mul( uchar value ); -+short __ovld sub_group_non_uniform_scan_inclusive_mul( short value ); -+ushort __ovld sub_group_non_uniform_scan_inclusive_mul( ushort value ); -+int __ovld sub_group_non_uniform_scan_inclusive_mul( int value ); -+uint __ovld sub_group_non_uniform_scan_inclusive_mul( uint value ); -+long __ovld sub_group_non_uniform_scan_inclusive_mul( long value ); -+ulong __ovld sub_group_non_uniform_scan_inclusive_mul( ulong value ); -+float __ovld sub_group_non_uniform_scan_inclusive_mul( float value ); -+ -+char __ovld sub_group_non_uniform_scan_inclusive_min( char value ); -+uchar __ovld sub_group_non_uniform_scan_inclusive_min( uchar value ); -+short __ovld sub_group_non_uniform_scan_inclusive_min( short value ); -+ushort __ovld sub_group_non_uniform_scan_inclusive_min( ushort value ); -+int __ovld sub_group_non_uniform_scan_inclusive_min( int value ); -+uint __ovld sub_group_non_uniform_scan_inclusive_min( uint value ); -+long __ovld sub_group_non_uniform_scan_inclusive_min( long value ); -+ulong __ovld sub_group_non_uniform_scan_inclusive_min( ulong value ); -+float __ovld sub_group_non_uniform_scan_inclusive_min( float value ); -+ -+char __ovld sub_group_non_uniform_scan_inclusive_max( char value ); -+uchar __ovld sub_group_non_uniform_scan_inclusive_max( uchar value ); -+short __ovld sub_group_non_uniform_scan_inclusive_max( short value ); -+ushort __ovld sub_group_non_uniform_scan_inclusive_max( ushort value ); -+int __ovld sub_group_non_uniform_scan_inclusive_max( int value ); -+uint __ovld sub_group_non_uniform_scan_inclusive_max( uint value ); -+long __ovld sub_group_non_uniform_scan_inclusive_max( long value ); -+ulong __ovld sub_group_non_uniform_scan_inclusive_max( ulong value ); -+float __ovld sub_group_non_uniform_scan_inclusive_max( float value ); -+ -+char __ovld sub_group_non_uniform_scan_exclusive_add( char value ); -+uchar __ovld sub_group_non_uniform_scan_exclusive_add( uchar value ); -+short __ovld sub_group_non_uniform_scan_exclusive_add( short value ); -+ushort __ovld sub_group_non_uniform_scan_exclusive_add( ushort value ); -+int __ovld sub_group_non_uniform_scan_exclusive_add( int value ); -+uint __ovld sub_group_non_uniform_scan_exclusive_add( uint value ); -+long __ovld sub_group_non_uniform_scan_exclusive_add( long value ); -+ulong __ovld sub_group_non_uniform_scan_exclusive_add( ulong value ); -+float __ovld sub_group_non_uniform_scan_exclusive_add( float value ); -+ -+char __ovld sub_group_non_uniform_scan_exclusive_mul( char value ); -+uchar __ovld sub_group_non_uniform_scan_exclusive_mul( uchar value ); -+short __ovld sub_group_non_uniform_scan_exclusive_mul( short value ); -+ushort __ovld sub_group_non_uniform_scan_exclusive_mul( ushort value ); -+int __ovld sub_group_non_uniform_scan_exclusive_mul( int value ); -+uint __ovld sub_group_non_uniform_scan_exclusive_mul( uint value ); -+long __ovld sub_group_non_uniform_scan_exclusive_mul( long value ); -+ulong __ovld sub_group_non_uniform_scan_exclusive_mul( ulong value ); -+float __ovld sub_group_non_uniform_scan_exclusive_mul( float value ); -+ -+char __ovld sub_group_non_uniform_scan_exclusive_min( char value ); -+uchar __ovld sub_group_non_uniform_scan_exclusive_min( uchar value ); -+short __ovld sub_group_non_uniform_scan_exclusive_min( short value ); -+ushort __ovld sub_group_non_uniform_scan_exclusive_min( ushort value ); -+int __ovld sub_group_non_uniform_scan_exclusive_min( int value ); -+uint __ovld sub_group_non_uniform_scan_exclusive_min( uint value ); -+long __ovld sub_group_non_uniform_scan_exclusive_min( long value ); -+ulong __ovld sub_group_non_uniform_scan_exclusive_min( ulong value ); -+float __ovld sub_group_non_uniform_scan_exclusive_min( float value ); -+ -+char __ovld sub_group_non_uniform_scan_exclusive_max( char value ); -+uchar __ovld sub_group_non_uniform_scan_exclusive_max( uchar value ); -+short __ovld sub_group_non_uniform_scan_exclusive_max( short value ); -+ushort __ovld sub_group_non_uniform_scan_exclusive_max( ushort value ); -+int __ovld sub_group_non_uniform_scan_exclusive_max( int value ); -+uint __ovld sub_group_non_uniform_scan_exclusive_max( uint value ); -+long __ovld sub_group_non_uniform_scan_exclusive_max( long value ); -+ulong __ovld sub_group_non_uniform_scan_exclusive_max( ulong value ); -+float __ovld sub_group_non_uniform_scan_exclusive_max( float value ); -+ -+char __ovld sub_group_non_uniform_reduce_and( char value ); -+uchar __ovld sub_group_non_uniform_reduce_and( uchar value ); -+short __ovld sub_group_non_uniform_reduce_and( short value ); -+ushort __ovld sub_group_non_uniform_reduce_and( ushort value ); -+int __ovld sub_group_non_uniform_reduce_and( int value ); -+uint __ovld sub_group_non_uniform_reduce_and( uint value ); -+long __ovld sub_group_non_uniform_reduce_and( long value ); -+ulong __ovld sub_group_non_uniform_reduce_and( ulong value ); -+ -+char __ovld sub_group_non_uniform_reduce_or( char value ); -+uchar __ovld sub_group_non_uniform_reduce_or( uchar value ); -+short __ovld sub_group_non_uniform_reduce_or( short value ); -+ushort __ovld sub_group_non_uniform_reduce_or( ushort value ); -+int __ovld sub_group_non_uniform_reduce_or( int value ); -+uint __ovld sub_group_non_uniform_reduce_or( uint value ); -+long __ovld sub_group_non_uniform_reduce_or( long value ); -+ulong __ovld sub_group_non_uniform_reduce_or( ulong value ); -+ -+char __ovld sub_group_non_uniform_reduce_xor( char value ); -+uchar __ovld sub_group_non_uniform_reduce_xor( uchar value ); -+short __ovld sub_group_non_uniform_reduce_xor( short value ); -+ushort __ovld sub_group_non_uniform_reduce_xor( ushort value ); -+int __ovld sub_group_non_uniform_reduce_xor( int value ); -+uint __ovld sub_group_non_uniform_reduce_xor( uint value ); -+long __ovld sub_group_non_uniform_reduce_xor( long value ); -+ulong __ovld sub_group_non_uniform_reduce_xor( ulong value ); -+ -+char __ovld sub_group_non_uniform_scan_inclusive_and( char value ); -+uchar __ovld sub_group_non_uniform_scan_inclusive_and( uchar value ); -+short __ovld sub_group_non_uniform_scan_inclusive_and( short value ); -+ushort __ovld sub_group_non_uniform_scan_inclusive_and( ushort value ); -+int __ovld sub_group_non_uniform_scan_inclusive_and( int value ); -+uint __ovld sub_group_non_uniform_scan_inclusive_and( uint value ); -+long __ovld sub_group_non_uniform_scan_inclusive_and( long value ); -+ulong __ovld sub_group_non_uniform_scan_inclusive_and( ulong value ); -+ -+char __ovld sub_group_non_uniform_scan_inclusive_or( char value ); -+uchar __ovld sub_group_non_uniform_scan_inclusive_or( uchar value ); -+short __ovld sub_group_non_uniform_scan_inclusive_or( short value ); -+ushort __ovld sub_group_non_uniform_scan_inclusive_or( ushort value ); -+int __ovld sub_group_non_uniform_scan_inclusive_or( int value ); -+uint __ovld sub_group_non_uniform_scan_inclusive_or( uint value ); -+long __ovld sub_group_non_uniform_scan_inclusive_or( long value ); -+ulong __ovld sub_group_non_uniform_scan_inclusive_or( ulong value ); -+ -+char __ovld sub_group_non_uniform_scan_inclusive_xor( char value ); -+uchar __ovld sub_group_non_uniform_scan_inclusive_xor( uchar value ); -+short __ovld sub_group_non_uniform_scan_inclusive_xor( short value ); -+ushort __ovld sub_group_non_uniform_scan_inclusive_xor( ushort value ); -+int __ovld sub_group_non_uniform_scan_inclusive_xor( int value ); -+uint __ovld sub_group_non_uniform_scan_inclusive_xor( uint value ); -+long __ovld sub_group_non_uniform_scan_inclusive_xor( long value ); -+ulong __ovld sub_group_non_uniform_scan_inclusive_xor( ulong value ); -+ -+char __ovld sub_group_non_uniform_scan_exclusive_and( char value ); -+uchar __ovld sub_group_non_uniform_scan_exclusive_and( uchar value ); -+short __ovld sub_group_non_uniform_scan_exclusive_and( short value ); -+ushort __ovld sub_group_non_uniform_scan_exclusive_and( ushort value ); -+int __ovld sub_group_non_uniform_scan_exclusive_and( int value ); -+uint __ovld sub_group_non_uniform_scan_exclusive_and( uint value ); -+long __ovld sub_group_non_uniform_scan_exclusive_and( long value ); -+ulong __ovld sub_group_non_uniform_scan_exclusive_and( ulong value ); -+ -+char __ovld sub_group_non_uniform_scan_exclusive_or( char value ); -+uchar __ovld sub_group_non_uniform_scan_exclusive_or( uchar value ); -+short __ovld sub_group_non_uniform_scan_exclusive_or( short value ); -+ushort __ovld sub_group_non_uniform_scan_exclusive_or( ushort value ); -+int __ovld sub_group_non_uniform_scan_exclusive_or( int value ); -+uint __ovld sub_group_non_uniform_scan_exclusive_or( uint value ); -+long __ovld sub_group_non_uniform_scan_exclusive_or( long value ); -+ulong __ovld sub_group_non_uniform_scan_exclusive_or( ulong value ); -+ -+char __ovld sub_group_non_uniform_scan_exclusive_xor( char value ); -+uchar __ovld sub_group_non_uniform_scan_exclusive_xor( uchar value ); -+short __ovld sub_group_non_uniform_scan_exclusive_xor( short value ); -+ushort __ovld sub_group_non_uniform_scan_exclusive_xor( ushort value ); -+int __ovld sub_group_non_uniform_scan_exclusive_xor( int value ); -+uint __ovld sub_group_non_uniform_scan_exclusive_xor( uint value ); -+long __ovld sub_group_non_uniform_scan_exclusive_xor( long value ); -+ulong __ovld sub_group_non_uniform_scan_exclusive_xor( ulong value ); -+ -+int __ovld sub_group_non_uniform_reduce_logical_and( int predicate ); -+int __ovld sub_group_non_uniform_reduce_logical_or( int predicate ); -+int __ovld sub_group_non_uniform_reduce_logical_xor( int predicate ); -+ -+int __ovld sub_group_non_uniform_scan_inclusive_logical_and( int predicate ); -+int __ovld sub_group_non_uniform_scan_inclusive_logical_or( int predicate ); -+int __ovld sub_group_non_uniform_scan_inclusive_logical_xor( int predicate ); -+ -+int __ovld sub_group_non_uniform_scan_exclusive_logical_and( int predicate ); -+int __ovld sub_group_non_uniform_scan_exclusive_logical_or( int predicate ); -+int __ovld sub_group_non_uniform_scan_exclusive_logical_xor( int predicate ); -+ -+#if defined(cl_khr_fp16) -+half __ovld sub_group_non_uniform_reduce_add( half value ); -+half __ovld sub_group_non_uniform_reduce_mul( half value ); -+half __ovld sub_group_non_uniform_reduce_min( half value ); -+half __ovld sub_group_non_uniform_reduce_max( half value ); -+half __ovld sub_group_non_uniform_scan_inclusive_add( half value ); -+half __ovld sub_group_non_uniform_scan_inclusive_mul( half value ); -+half __ovld sub_group_non_uniform_scan_inclusive_min( half value ); -+half __ovld sub_group_non_uniform_scan_inclusive_max( half value ); -+half __ovld sub_group_non_uniform_scan_exclusive_add( half value ); -+half __ovld sub_group_non_uniform_scan_exclusive_mul( half value ); -+half __ovld sub_group_non_uniform_scan_exclusive_min( half value ); -+half __ovld sub_group_non_uniform_scan_exclusive_max( half value ); -+#endif // cl_khr_fp16 -+ -+#if defined(cl_khr_fp64) -+double __ovld sub_group_non_uniform_reduce_add( double value ); -+double __ovld sub_group_non_uniform_reduce_mul( double value ); -+double __ovld sub_group_non_uniform_reduce_min( double value ); -+double __ovld sub_group_non_uniform_reduce_max( double value ); -+double __ovld sub_group_non_uniform_scan_inclusive_add( double value ); -+double __ovld sub_group_non_uniform_scan_inclusive_mul( double value ); -+double __ovld sub_group_non_uniform_scan_inclusive_min( double value ); -+double __ovld sub_group_non_uniform_scan_inclusive_max( double value ); -+double __ovld sub_group_non_uniform_scan_exclusive_add( double value ); -+double __ovld sub_group_non_uniform_scan_exclusive_mul( double value ); -+double __ovld sub_group_non_uniform_scan_exclusive_min( double value ); -+double __ovld sub_group_non_uniform_scan_exclusive_max( double value ); -+#endif // cl_khr_fp64 -+ -+#endif // cl_khr_subgroup_non_uniform_arithmetic -+ -+#if defined(cl_khr_subgroup_shuffle) -+char __ovld sub_group_shuffle( char value, uint index ); -+uchar __ovld sub_group_shuffle( uchar value, uint index ); -+short __ovld sub_group_shuffle( short value, uint index ); -+ushort __ovld sub_group_shuffle( ushort value, uint index ); -+int __ovld sub_group_shuffle( int value, uint index ); -+uint __ovld sub_group_shuffle( uint value, uint index ); -+long __ovld sub_group_shuffle( long value, uint index ); -+ulong __ovld sub_group_shuffle( ulong value, uint index ); -+float __ovld sub_group_shuffle( float value, uint index ); -+ -+char __ovld sub_group_shuffle_xor( char value, uint mask ); -+uchar __ovld sub_group_shuffle_xor( uchar value, uint mask ); -+short __ovld sub_group_shuffle_xor( short value, uint mask ); -+ushort __ovld sub_group_shuffle_xor( ushort value, uint mask ); -+int __ovld sub_group_shuffle_xor( int value, uint mask ); -+uint __ovld sub_group_shuffle_xor( uint value, uint mask ); -+long __ovld sub_group_shuffle_xor( long value, uint mask ); -+ulong __ovld sub_group_shuffle_xor( ulong value, uint mask ); -+float __ovld sub_group_shuffle_xor( float value, uint mask ); -+ -+#if defined(cl_khr_fp16) -+half __ovld sub_group_shuffle( half value, uint index ); -+half __ovld sub_group_shuffle_xor( half value, uint mask ); -+#endif // cl_khr_fp16 -+ -+#if defined(cl_khr_fp64) -+double __ovld sub_group_shuffle( double value, uint index ); -+double __ovld sub_group_shuffle_xor( double value, uint mask ); -+#endif // cl_khr_fp64 -+ -+#endif // cl_khr_subgroup_shuffle -+ -+#if defined(cl_khr_subgroup_shuffle_relative) -+char __ovld sub_group_shuffle_up( char value, uint delta ); -+uchar __ovld sub_group_shuffle_up( uchar value, uint delta ); -+short __ovld sub_group_shuffle_up( short value, uint delta ); -+ushort __ovld sub_group_shuffle_up( ushort value, uint delta ); -+int __ovld sub_group_shuffle_up( int value, uint delta ); -+uint __ovld sub_group_shuffle_up( uint value, uint delta ); -+long __ovld sub_group_shuffle_up( long value, uint delta ); -+ulong __ovld sub_group_shuffle_up( ulong value, uint delta ); -+float __ovld sub_group_shuffle_up( float value, uint delta ); -+ -+char __ovld sub_group_shuffle_down( char value, uint delta ); -+uchar __ovld sub_group_shuffle_down( uchar value, uint delta ); -+short __ovld sub_group_shuffle_down( short value, uint delta ); -+ushort __ovld sub_group_shuffle_down( ushort value, uint delta ); -+int __ovld sub_group_shuffle_down( int value, uint delta ); -+uint __ovld sub_group_shuffle_down( uint value, uint delta ); -+long __ovld sub_group_shuffle_down( long value, uint delta ); -+ulong __ovld sub_group_shuffle_down( ulong value, uint delta ); -+float __ovld sub_group_shuffle_down( float value, uint delta ); -+ -+#if defined(cl_khr_fp16) -+half __ovld sub_group_shuffle_up( half value, uint delta ); -+half __ovld sub_group_shuffle_down( half value, uint delta ); -+#endif // cl_khr_fp16 -+ -+#if defined(cl_khr_fp64) -+double __ovld sub_group_shuffle_up( double value, uint delta ); -+double __ovld sub_group_shuffle_down( double value, uint delta ); -+#endif // cl_khr_fp64 -+ -+#endif // cl_khr_subgroup_shuffle_relative -+ -+#if defined(cl_khr_subgroup_clustered_reduce) -+char __ovld sub_group_clustered_reduce_add( char value, uint clustersize ); -+uchar __ovld sub_group_clustered_reduce_add( uchar value, uint clustersize ); -+short __ovld sub_group_clustered_reduce_add( short value, uint clustersize ); -+ushort __ovld sub_group_clustered_reduce_add( ushort value, uint clustersize ); -+int __ovld sub_group_clustered_reduce_add( int value, uint clustersize ); -+uint __ovld sub_group_clustered_reduce_add( uint value, uint clustersize ); -+long __ovld sub_group_clustered_reduce_add( long value, uint clustersize ); -+ulong __ovld sub_group_clustered_reduce_add( ulong value, uint clustersize ); -+float __ovld sub_group_clustered_reduce_add( float value, uint clustersize ); -+ -+char __ovld sub_group_clustered_reduce_mul( char value, uint clustersize ); -+uchar __ovld sub_group_clustered_reduce_mul( uchar value, uint clustersize ); -+short __ovld sub_group_clustered_reduce_mul( short value, uint clustersize ); -+ushort __ovld sub_group_clustered_reduce_mul( ushort value, uint clustersize ); -+int __ovld sub_group_clustered_reduce_mul( int value, uint clustersize ); -+uint __ovld sub_group_clustered_reduce_mul( uint value, uint clustersize ); -+long __ovld sub_group_clustered_reduce_mul( long value, uint clustersize ); -+ulong __ovld sub_group_clustered_reduce_mul( ulong value, uint clustersize ); -+float __ovld sub_group_clustered_reduce_mul( float value, uint clustersize ); -+ -+char __ovld sub_group_clustered_reduce_min( char value, uint clustersize ); -+uchar __ovld sub_group_clustered_reduce_min( uchar value, uint clustersize ); -+short __ovld sub_group_clustered_reduce_min( short value, uint clustersize ); -+ushort __ovld sub_group_clustered_reduce_min( ushort value, uint clustersize ); -+int __ovld sub_group_clustered_reduce_min( int value, uint clustersize ); -+uint __ovld sub_group_clustered_reduce_min( uint value, uint clustersize ); -+long __ovld sub_group_clustered_reduce_min( long value, uint clustersize ); -+ulong __ovld sub_group_clustered_reduce_min( ulong value, uint clustersize ); -+float __ovld sub_group_clustered_reduce_min( float value, uint clustersize ); -+ -+char __ovld sub_group_clustered_reduce_max( char value, uint clustersize ); -+uchar __ovld sub_group_clustered_reduce_max( uchar value, uint clustersize ); -+short __ovld sub_group_clustered_reduce_max( short value, uint clustersize ); -+ushort __ovld sub_group_clustered_reduce_max( ushort value, uint clustersize ); -+int __ovld sub_group_clustered_reduce_max( int value, uint clustersize ); -+uint __ovld sub_group_clustered_reduce_max( uint value, uint clustersize ); -+long __ovld sub_group_clustered_reduce_max( long value, uint clustersize ); -+ulong __ovld sub_group_clustered_reduce_max( ulong value, uint clustersize ); -+float __ovld sub_group_clustered_reduce_max( float value, uint clustersize ); -+ -+char __ovld sub_group_clustered_reduce_and( char value, uint clustersize ); -+uchar __ovld sub_group_clustered_reduce_and( uchar value, uint clustersize ); -+short __ovld sub_group_clustered_reduce_and( short value, uint clustersize ); -+ushort __ovld sub_group_clustered_reduce_and( ushort value, uint clustersize ); -+int __ovld sub_group_clustered_reduce_and( int value, uint clustersize ); -+uint __ovld sub_group_clustered_reduce_and( uint value, uint clustersize ); -+long __ovld sub_group_clustered_reduce_and( long value, uint clustersize ); -+ulong __ovld sub_group_clustered_reduce_and( ulong value, uint clustersize ); -+ -+char __ovld sub_group_clustered_reduce_or( char value, uint clustersize ); -+uchar __ovld sub_group_clustered_reduce_or( uchar value, uint clustersize ); -+short __ovld sub_group_clustered_reduce_or( short value, uint clustersize ); -+ushort __ovld sub_group_clustered_reduce_or( ushort value, uint clustersize ); -+int __ovld sub_group_clustered_reduce_or( int value, uint clustersize ); -+uint __ovld sub_group_clustered_reduce_or( uint value, uint clustersize ); -+long __ovld sub_group_clustered_reduce_or( long value, uint clustersize ); -+ulong __ovld sub_group_clustered_reduce_or( ulong value, uint clustersize ); -+ -+char __ovld sub_group_clustered_reduce_xor( char value, uint clustersize ); -+uchar __ovld sub_group_clustered_reduce_xor( uchar value, uint clustersize ); -+short __ovld sub_group_clustered_reduce_xor( short value, uint clustersize ); -+ushort __ovld sub_group_clustered_reduce_xor( ushort value, uint clustersize ); -+int __ovld sub_group_clustered_reduce_xor( int value, uint clustersize ); -+uint __ovld sub_group_clustered_reduce_xor( uint value, uint clustersize ); -+long __ovld sub_group_clustered_reduce_xor( long value, uint clustersize ); -+ulong __ovld sub_group_clustered_reduce_xor( ulong value, uint clustersize ); -+ -+int __ovld sub_group_clustered_reduce_logical_and( int predicate, uint clustersize ); -+int __ovld sub_group_clustered_reduce_logical_or( int predicate, uint clustersize ); -+int __ovld sub_group_clustered_reduce_logical_xor( int predicate, uint clustersize ); -+ -+#if defined(cl_khr_fp16) -+half __ovld sub_group_clustered_reduce_add( half value, uint clustersize ); -+half __ovld sub_group_clustered_reduce_mul( half value, uint clustersize ); -+half __ovld sub_group_clustered_reduce_min( half value, uint clustersize ); -+half __ovld sub_group_clustered_reduce_max( half value, uint clustersize ); -+#endif // cl_khr_fp16 -+ -+#if defined(cl_khr_fp64) -+double __ovld sub_group_clustered_reduce_add( double value, uint clustersize ); -+double __ovld sub_group_clustered_reduce_mul( double value, uint clustersize ); -+double __ovld sub_group_clustered_reduce_min( double value, uint clustersize ); -+double __ovld sub_group_clustered_reduce_max( double value, uint clustersize ); -+#endif // cl_khr_fp64 -+ -+#endif // cl_khr_subgroup_clustered_reduce -+ - #if defined(cl_intel_subgroups) - // Intel-Specific Sub Group Functions - float __ovld __conv intel_sub_group_shuffle( float x, uint c ); -diff --git a/clang/test/SemaOpenCL/extension-version.cl b/clang/test/SemaOpenCL/extension-version.cl -index 0e6bbb7d3bcd..86c78143a0eb 100644 ---- a/clang/test/SemaOpenCL/extension-version.cl -+++ b/clang/test/SemaOpenCL/extension-version.cl -@@ -333,3 +333,86 @@ - #endif - #pragma OPENCL EXTENSION cl_intel_device_side_avc_motion_estimation : enable - -+#if (defined(__OPENCL_CPP_VERSION__) || __OPENCL_C_VERSION__ >= 200) -+#ifndef cl_khr_subgroup_extended_types -+#error "Missing cl_khr_subgroup_extended_types" -+#endif -+#else -+#ifdef cl_khr_subgroup_extended_types -+#error "Incorrect cl_khr_subgroup_extended_types define" -+#endif -+// expected-warning@+2{{unsupported OpenCL extension 'cl_khr_subgroup_extended_types' - ignoring}} -+#endif -+#pragma OPENCL EXTENSION cl_khr_subgroup_extended_types : enable -+ -+#if (defined(__OPENCL_CPP_VERSION__) || __OPENCL_C_VERSION__ >= 200) -+#ifndef cl_khr_subgroup_non_uniform_vote -+#error "Missing cl_khr_subgroup_non_uniform_vote" -+#endif -+#else -+#ifdef cl_khr_subgroup_non_uniform_vote -+#error "Incorrect cl_khr_subgroup_non_uniform_vote define" -+#endif -+// expected-warning@+2{{unsupported OpenCL extension 'cl_khr_subgroup_non_uniform_vote' - ignoring}} -+#endif -+#pragma OPENCL EXTENSION cl_khr_subgroup_non_uniform_vote : enable -+ -+#if (defined(__OPENCL_CPP_VERSION__) || __OPENCL_C_VERSION__ >= 200) -+#ifndef cl_khr_subgroup_ballot -+#error "Missing cl_khr_subgroup_ballot" -+#endif -+#else -+#ifdef cl_khr_subgroup_ballot -+#error "Incorrect cl_khr_subgroup_ballot define" -+#endif -+// expected-warning@+2{{unsupported OpenCL extension 'cl_khr_subgroup_ballot' - ignoring}} -+#endif -+#pragma OPENCL EXTENSION cl_khr_subgroup_ballot : enable -+ -+#if (defined(__OPENCL_CPP_VERSION__) || __OPENCL_C_VERSION__ >= 200) -+#ifndef cl_khr_subgroup_non_uniform_arithmetic -+#error "Missing cl_khr_subgroup_non_uniform_arithmetic" -+#endif -+#else -+#ifdef cl_khr_subgroup_non_uniform_arithmetic -+#error "Incorrect cl_khr_subgroup_non_uniform_arithmetic define" -+#endif -+// expected-warning@+2{{unsupported OpenCL extension 'cl_khr_subgroup_non_uniform_arithmetic' - ignoring}} -+#endif -+#pragma OPENCL EXTENSION cl_khr_subgroup_non_uniform_arithmetic : enable -+ -+#if (defined(__OPENCL_CPP_VERSION__) || __OPENCL_C_VERSION__ >= 200) -+#ifndef cl_khr_subgroup_shuffle -+#error "Missing cl_khr_subgroup_shuffle" -+#endif -+#else -+#ifdef cl_khr_subgroup_shuffle -+#error "Incorrect cl_khr_subgroup_shuffle define" -+#endif -+// expected-warning@+2{{unsupported OpenCL extension 'cl_khr_subgroup_shuffle' - ignoring}} -+#endif -+#pragma OPENCL EXTENSION cl_khr_subgroup_shuffle : enable -+ -+#if (defined(__OPENCL_CPP_VERSION__) || __OPENCL_C_VERSION__ >= 200) -+#ifndef cl_khr_subgroup_shuffle_relative -+#error "Missing cl_khr_subgroup_shuffle_relative" -+#endif -+#else -+#ifdef cl_khr_subgroup_shuffle_relative -+#error "Incorrect cl_khr_subgroup_shuffle_relative define" -+#endif -+// expected-warning@+2{{unsupported OpenCL extension 'cl_khr_subgroup_shuffle_relative' - ignoring}} -+#endif -+#pragma OPENCL EXTENSION cl_khr_subgroup_shuffle_relative : enable -+ -+#if (defined(__OPENCL_CPP_VERSION__) || __OPENCL_C_VERSION__ >= 200) -+#ifndef cl_khr_subgroup_clustered_reduce -+#error "Missing cl_khr_subgroup_clustered_reduce" -+#endif -+#else -+#ifdef cl_khr_subgroup_clustered_reduce -+#error "Incorrect cl_khr_subgroup_clustered_reduce define" -+#endif -+// expected-warning@+2{{unsupported OpenCL extension 'cl_khr_subgroup_clustered_reduce' - ignoring}} -+#endif -+#pragma OPENCL EXTENSION cl_khr_subgroup_clustered_reduce : enable --- -2.17.1 - diff --git a/dynamic-layers/clang-layer/recipes-devtools/clang/files/BasicBlockUtils-Add-metadata-fixing-in-SplitBlockPre.patch b/dynamic-layers/clang-layer/recipes-devtools/clang/files/BasicBlockUtils-Add-metadata-fixing-in-SplitBlockPre.patch deleted file mode 100644 index cd519971..00000000 --- a/dynamic-layers/clang-layer/recipes-devtools/clang/files/BasicBlockUtils-Add-metadata-fixing-in-SplitBlockPre.patch +++ /dev/null @@ -1,111 +0,0 @@ -From eeb816d95f0910bd246e37bb2bb3923acf0edf6b Mon Sep 17 00:00:00 2001 -From: Aleksander Us <aleksander.us@intel.com> -Date: Mon, 26 Aug 2019 15:47:41 +0300 -Subject: [PATCH] [BasicBlockUtils] Add metadata fixing in - SplitBlockPredecessors. - -In case when BB is header of some loop and predecessor is latch of -this loop, metadata was not attached to newly created basic block. -This led to loss of loop metadata for other passes. - -Upstream-Status: Submitted [https://reviews.llvm.org/D66892] - -https://github.com/intel/llvm-patches/commit/8af4449e2d201707f7f2f832b473a0439e255f32 - -Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> ---- - lib/Transforms/Utils/BasicBlockUtils.cpp | 23 ++++++++---- - test/Transforms/LoopSimplify/loop_metadata.ll | 36 +++++++++++++++++++ - 2 files changed, 52 insertions(+), 7 deletions(-) - create mode 100644 test/Transforms/LoopSimplify/loop_metadata.ll - -diff --git a/lib/Transforms/Utils/BasicBlockUtils.cpp b/lib/Transforms/Utils/BasicBlockUtils.cpp -index 5fa371377c8..3a90ae061fb 100644 ---- a/lib/Transforms/Utils/BasicBlockUtils.cpp -+++ b/lib/Transforms/Utils/BasicBlockUtils.cpp -@@ -579,24 +579,33 @@ BasicBlock *llvm::SplitBlockPredecessors(BasicBlock *BB, - - // The new block unconditionally branches to the old block. - BranchInst *BI = BranchInst::Create(BB, NewBB); -+ bool IsBBHeader = LI && LI->isLoopHeader(BB); -+ Loop *BBLoop = LI ? LI->getLoopFor(BB) : nullptr; - // Splitting the predecessors of a loop header creates a preheader block. -- if (LI && LI->isLoopHeader(BB)) -+ if (IsBBHeader) - // Using the loop start line number prevents debuggers stepping into the - // loop body for this instruction. -- BI->setDebugLoc(LI->getLoopFor(BB)->getStartLoc()); -+ BI->setDebugLoc(BBLoop->getStartLoc()); - else - BI->setDebugLoc(BB->getFirstNonPHIOrDbg()->getDebugLoc()); - - // Move the edges from Preds to point to NewBB instead of BB. -- for (unsigned i = 0, e = Preds.size(); i != e; ++i) { -+ for (BasicBlock *Pred : Preds) { -+ Instruction *PI = Pred->getTerminator(); - // This is slightly more strict than necessary; the minimum requirement - // is that there be no more than one indirectbr branching to BB. And - // all BlockAddress uses would need to be updated. -- assert(!isa<IndirectBrInst>(Preds[i]->getTerminator()) && -+ assert(!isa<IndirectBrInst>(PI) && - "Cannot split an edge from an IndirectBrInst"); -- assert(!isa<CallBrInst>(Preds[i]->getTerminator()) && -- "Cannot split an edge from a CallBrInst"); -- Preds[i]->getTerminator()->replaceUsesOfWith(BB, NewBB); -+ assert(!isa<CallBrInst>(PI) && "Cannot split an edge from a CallBrInst"); -+ if (IsBBHeader && BBLoop->contains(Pred) && BBLoop->isLoopLatch(Pred)) { -+ // Update loop metadata if it exists. -+ if (MDNode *LoopMD = PI->getMetadata(LLVMContext::MD_loop)) { -+ BI->setMetadata(LLVMContext::MD_loop, LoopMD); -+ PI->setMetadata(LLVMContext::MD_loop, nullptr); -+ } -+ } -+ PI->replaceUsesOfWith(BB, NewBB); - } - - // Insert a new PHI node into NewBB for every PHI node in BB and that new PHI -diff --git a/test/Transforms/LoopSimplify/loop_metadata.ll b/test/Transforms/LoopSimplify/loop_metadata.ll -new file mode 100644 -index 00000000000..c15c92fe3ae ---- /dev/null -+++ b/test/Transforms/LoopSimplify/loop_metadata.ll -@@ -0,0 +1,36 @@ -+; RUN: opt -S -loop-simplify < %s | FileCheck %s -+ -+; CHECK: for.cond.loopexit: -+; CHECK: br label %for.cond, !llvm.loop !0 -+; CHECK: br i1 %cmp1, label %for.body1, label %for.cond.loopexit -+ -+define void @foo() { -+entry: -+ br label %for.cond -+ -+for.cond: ; preds = %for.cond1, %entry -+ %j = phi i32 [ 0, %entry ], [ %add, %for.cond1 ] -+ %cmp = icmp ult i32 %j, 8 -+ br i1 %cmp, label %for.body, label %for.end -+ -+for.body: ; preds = %for.cond -+ %dummy1 = add i32 1, 1 -+ %add = add nuw nsw i32 %j, 1 -+ br label %for.cond1 -+ -+for.cond1: ; preds = %for.body1, %for.body -+ %i.0 = phi i32 [ 1, %for.body ], [ %inc, %for.body1 ] -+ %cmp1 = icmp ult i32 %i.0, 8 -+ br i1 %cmp1, label %for.body1, label %for.cond, !llvm.loop !0 -+ -+for.body1: ; preds = %for.cond1 -+ %dummy2 = add i32 1, 1 -+ %inc = add nuw nsw i32 %i.0, 1 -+ br label %for.cond1 -+ -+for.end: ; preds = %for.cond -+ ret void -+} -+ -+!0 = distinct !{!0, !1} -+!1 = !{!"llvm.loop.unroll.full"} --- -2.18.0 - diff --git a/dynamic-layers/clang-layer/recipes-devtools/clang/files/IndVarSimplify-Do-not-use-SCEV-expander-for-IVCount-.patch b/dynamic-layers/clang-layer/recipes-devtools/clang/files/IndVarSimplify-Do-not-use-SCEV-expander-for-IVCount-.patch deleted file mode 100644 index 48307deb..00000000 --- a/dynamic-layers/clang-layer/recipes-devtools/clang/files/IndVarSimplify-Do-not-use-SCEV-expander-for-IVCount-.patch +++ /dev/null @@ -1,146 +0,0 @@ -From 35e218a886f4c066eabd18685240d55270bd5a6d Mon Sep 17 00:00:00 2001 -From: Aleksander Us <aleksander.us@intel.com> -Date: Mon, 26 Aug 2019 15:45:47 +0300 -Subject: [PATCH] [IndVarSimplify] Do not use SCEV expander for IVCount in - LFTR when possible. - -SCEV analysis cannot properly cache instruction with poison flags -(for example, add nsw outside of loop will not be reused by expander). -This can lead to generating of additional instructions by SCEV expander. - -Example IR: - - ... - %maxval = add nuw nsw i32 %a1, %a2 - ... -for.body: - ... - %cmp22 = icmp ult i32 %ivadd, %maxval - br i1 %cmp22, label %for.body, label %for.end - ... - -SCEV expander will generate copy of %maxval in preheader but without -nuw/nsw flags. This can be avoided by explicit check that iv count -value gives the same SCEV expressions as calculated by LFTR. - -Upstream-Status: Submitted [https://reviews.llvm.org/D66890] - -https://github.com/intel/llvm-patches/commit/fd6a6c97341a56fd21bc32bc940afea751312e8f - -Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> ---- - lib/Transforms/Scalar/IndVarSimplify.cpp | 12 +++++++++- - test/Transforms/IndVarSimplify/add_nsw.ll | 23 ++++++++++++++++++++ - test/Transforms/IndVarSimplify/lftr-reuse.ll | 9 +++----- - test/Transforms/IndVarSimplify/udiv.ll | 1 + - 4 files changed, 38 insertions(+), 7 deletions(-) - create mode 100644 test/Transforms/IndVarSimplify/add_nsw.ll - -diff --git a/lib/Transforms/Scalar/IndVarSimplify.cpp b/lib/Transforms/Scalar/IndVarSimplify.cpp -index f9fc698a4a9..5e04dac8aa6 100644 ---- a/lib/Transforms/Scalar/IndVarSimplify.cpp -+++ b/lib/Transforms/Scalar/IndVarSimplify.cpp -@@ -2375,6 +2375,17 @@ static Value *genLoopLimit(PHINode *IndVar, BasicBlock *ExitingBB, - if (UsePostInc) - IVLimit = SE->getAddExpr(IVLimit, SE->getOne(IVLimit->getType())); - -+ // If computed limit is equal to old limit then do not use SCEV expander -+ // because it can lost NUW/NSW flags and create extra instructions. -+ BranchInst *BI = cast<BranchInst>(ExitingBB->getTerminator()); -+ if (ICmpInst *Cmp = dyn_cast<ICmpInst>(BI->getOperand(0))) { -+ Value *Limit = Cmp->getOperand(0); -+ if (!L->isLoopInvariant(Limit)) -+ Limit = Cmp->getOperand(1); -+ if (SE->getSCEV(Limit) == IVLimit) -+ return Limit; -+ } -+ - // Expand the code for the iteration count. - assert(SE->isLoopInvariant(IVLimit, L) && - "Computed iteration count is not loop invariant!"); -@@ -2383,7 +2394,6 @@ static Value *genLoopLimit(PHINode *IndVar, BasicBlock *ExitingBB, - // SCEV expression (IVInit) for a pointer type IV value (IndVar). - Type *LimitTy = ExitCount->getType()->isPointerTy() ? - IndVar->getType() : ExitCount->getType(); -- BranchInst *BI = cast<BranchInst>(ExitingBB->getTerminator()); - return Rewriter.expandCodeFor(IVLimit, LimitTy, BI); - } - } -diff --git a/test/Transforms/IndVarSimplify/add_nsw.ll b/test/Transforms/IndVarSimplify/add_nsw.ll -new file mode 100644 -index 00000000000..abd1cbb6c51 ---- /dev/null -+++ b/test/Transforms/IndVarSimplify/add_nsw.ll -@@ -0,0 +1,23 @@ -+; RUN: opt -indvars -S %s | FileCheck %s -+ -+target datalayout = "e-p:32:32-i64:64-n8:16:32" -+ -+; CHECK: for.body.preheader: -+; CHECK-NOT: add -+; CHECK: for.body: -+ -+define void @foo(i32 %a1, i32 %a2) { -+entry: -+ %maxval = add nuw nsw i32 %a1, %a2 -+ %cmp = icmp slt i32 %maxval, 1 -+ br i1 %cmp, label %for.end, label %for.body -+ -+for.body: ; preds = %entry, %for.body -+ %j.02 = phi i32 [ 0, %entry ], [ %add31, %for.body ] -+ %add31 = add nuw nsw i32 %j.02, 1 -+ %cmp22 = icmp slt i32 %add31, %maxval -+ br i1 %cmp22, label %for.body, label %for.end -+ -+for.end: ; preds = %for.body -+ ret void -+} -diff --git a/test/Transforms/IndVarSimplify/lftr-reuse.ll b/test/Transforms/IndVarSimplify/lftr-reuse.ll -index 14ae9738696..509d662b767 100644 ---- a/test/Transforms/IndVarSimplify/lftr-reuse.ll -+++ b/test/Transforms/IndVarSimplify/lftr-reuse.ll -@@ -67,11 +67,9 @@ define void @expandOuterRecurrence(i32 %arg) nounwind { - ; CHECK-NEXT: [[CMP1:%.*]] = icmp slt i32 0, [[SUB1]] - ; CHECK-NEXT: br i1 [[CMP1]], label [[OUTER_PREHEADER:%.*]], label [[EXIT:%.*]] - ; CHECK: outer.preheader: --; CHECK-NEXT: [[TMP0:%.*]] = add i32 [[ARG]], -1 - ; CHECK-NEXT: br label [[OUTER:%.*]] - ; CHECK: outer: --; CHECK-NEXT: [[INDVARS_IV:%.*]] = phi i32 [ [[TMP0]], [[OUTER_PREHEADER]] ], [ [[INDVARS_IV_NEXT:%.*]], [[OUTER_INC:%.*]] ] --; CHECK-NEXT: [[I:%.*]] = phi i32 [ [[I_INC:%.*]], [[OUTER_INC]] ], [ 0, [[OUTER_PREHEADER]] ] -+; CHECK-NEXT: [[I:%.*]] = phi i32 [ [[I_INC:%.*]], [[OUTER_INC:%.*]] ], [ 0, [[OUTER_PREHEADER]] ] - ; CHECK-NEXT: [[SUB2:%.*]] = sub nsw i32 [[ARG]], [[I]] - ; CHECK-NEXT: [[SUB3:%.*]] = sub nsw i32 [[SUB2]], 1 - ; CHECK-NEXT: [[CMP2:%.*]] = icmp slt i32 0, [[SUB3]] -@@ -81,14 +79,13 @@ define void @expandOuterRecurrence(i32 %arg) nounwind { - ; CHECK: inner: - ; CHECK-NEXT: [[J:%.*]] = phi i32 [ 0, [[INNER_PH]] ], [ [[J_INC:%.*]], [[INNER]] ] - ; CHECK-NEXT: [[J_INC]] = add nuw nsw i32 [[J]], 1 --; CHECK-NEXT: [[EXITCOND:%.*]] = icmp ne i32 [[J_INC]], [[INDVARS_IV]] -+; CHECK-NEXT: [[EXITCOND:%.*]] = icmp ne i32 [[J_INC]], [[SUB3]] - ; CHECK-NEXT: br i1 [[EXITCOND]], label [[INNER]], label [[OUTER_INC_LOOPEXIT:%.*]] - ; CHECK: outer.inc.loopexit: - ; CHECK-NEXT: br label [[OUTER_INC]] - ; CHECK: outer.inc: - ; CHECK-NEXT: [[I_INC]] = add nuw nsw i32 [[I]], 1 --; CHECK-NEXT: [[INDVARS_IV_NEXT]] = add i32 [[INDVARS_IV]], -1 --; CHECK-NEXT: [[EXITCOND1:%.*]] = icmp ne i32 [[I_INC]], [[TMP0]] -+; CHECK-NEXT: [[EXITCOND1:%.*]] = icmp ne i32 [[I_INC]], [[SUB1]] - ; CHECK-NEXT: br i1 [[EXITCOND1]], label [[OUTER]], label [[EXIT_LOOPEXIT:%.*]] - ; CHECK: exit.loopexit: - ; CHECK-NEXT: br label [[EXIT]] -diff --git a/test/Transforms/IndVarSimplify/udiv.ll b/test/Transforms/IndVarSimplify/udiv.ll -index b3f2c2a6a66..3530343ef4a 100644 ---- a/test/Transforms/IndVarSimplify/udiv.ll -+++ b/test/Transforms/IndVarSimplify/udiv.ll -@@ -133,6 +133,7 @@ declare i32 @printf(i8* nocapture, ...) nounwind - ; CHECK-LABEL: @foo( - ; CHECK: for.body.preheader: - ; CHECK-NOT: udiv -+; CHECK: for.body: - - define void @foo(double* %p, i64 %n) nounwind { - entry: --- -2.18.0 - diff --git a/dynamic-layers/clang-layer/recipes-devtools/clang/files/fix-shared-libs.patch b/dynamic-layers/clang-layer/recipes-devtools/clang/files/fix-shared-libs.patch deleted file mode 100644 index d69d2a97..00000000 --- a/dynamic-layers/clang-layer/recipes-devtools/clang/files/fix-shared-libs.patch +++ /dev/null @@ -1,30 +0,0 @@ -From a6d4ccf082858e63e139ca06c02a071c343d2657 Mon Sep 17 00:00:00 2001 -From: Andrea Bocci <andrea.bocci@cern.ch> -Date: Sun, 15 Mar 2020 17:35:44 +0100 -Subject: [PATCH] Fix building in-tree with cmake -DLLVM_LINK_LLVM_DYLIB=ON - -Building in-tree with LLVM 11.0 master with the LLVM_LINK_LLVM_DYLIB -cmake flag fails to link with the LLVMSPIRVLib library. - -Add an explicit dependency to force the correct build order and linking. - -Signed-off-by: Andrea Bocci <andrea.bocci@cern.ch> -Upstream-Status: Backport -Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> ---- - tools/llvm-spirv/CMakeLists.txt | 2 +- - 1 file changed, 1 insertion(+), 1 deletion(-) - -diff --git a/tools/llvm-spirv/CMakeLists.txt b/tools/llvm-spirv/CMakeLists.txt -index 9aa96d9c..501c0daf 100644 ---- a/tools/llvm-spirv/CMakeLists.txt -+++ b/tools/llvm-spirv/CMakeLists.txt -@@ -14,7 +14,7 @@ add_llvm_tool(llvm-spirv - NO_INSTALL_RPATH - ) - --if (LLVM_SPIRV_BUILD_EXTERNAL) -+if (LLVM_SPIRV_BUILD_EXTERNAL OR LLVM_LINK_LLVM_DYLIB) - target_link_libraries(llvm-spirv PRIVATE LLVMSPIRVLib) - endif() - diff --git a/dynamic-layers/clang-layer/recipes-devtools/clang/files/llvm10-OpenCL-3.0-support.patch b/dynamic-layers/clang-layer/recipes-devtools/clang/files/llvm10-OpenCL-3.0-support.patch deleted file mode 100644 index 53395ea0..00000000 --- a/dynamic-layers/clang-layer/recipes-devtools/clang/files/llvm10-OpenCL-3.0-support.patch +++ /dev/null @@ -1,8259 +0,0 @@ -From 31ec702cb365f4d02dd2146fb4329d642b8fc30b Mon Sep 17 00:00:00 2001 -From: Naveen Saini <naveen.kumar.saini@intel.com> -Date: Wed, 7 Apr 2021 16:36:10 +0800 -Subject: [PATCH 1/2] OpenCL 3.0 support - -Upstream-Status: Backport [Taken from opencl-clang patches, https://github.com/intel/opencl-clang/blob/ocl-open-100/patches/clang/0001-OpenCL-3.0-support.patch] -Signed-off-by: Anton Zabaznov <anton.zabaznov@intel.com> -Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> ---- - clang/include/clang/Basic/Builtins.def | 65 +- - clang/include/clang/Basic/Builtins.h | 13 +- - .../clang/Basic/DiagnosticParseKinds.td | 2 + - .../clang/Basic/DiagnosticSemaKinds.td | 7 + - clang/include/clang/Basic/LangOptions.def | 2 + - clang/include/clang/Basic/LangStandards.def | 4 + - .../include/clang/Basic/OpenCLExtensions.def | 15 + - clang/include/clang/Basic/OpenCLOptions.h | 175 +- - clang/include/clang/Driver/Options.td | 2 +- - clang/include/clang/Sema/Sema.h | 9 + - clang/lib/AST/ASTContext.cpp | 3 +- - clang/lib/Basic/Builtins.cpp | 28 +- - clang/lib/Basic/TargetInfo.cpp | 11 + - clang/lib/Basic/Targets.cpp | 1 - - clang/lib/CodeGen/CodeGenFunction.cpp | 6 +- - clang/lib/Frontend/CompilerInvocation.cpp | 22 +- - clang/lib/Frontend/InitPreprocessor.cpp | 6 +- - clang/lib/Headers/opencl-c-base.h | 75 +- - clang/lib/Headers/opencl-c.h | 3228 ++++++++++++++--- - clang/lib/Parse/ParseDecl.cpp | 12 +- - clang/lib/Parse/ParsePragma.cpp | 10 +- - clang/lib/Sema/OpenCLBuiltins.td | 49 +- - clang/lib/Sema/Sema.cpp | 47 +- - clang/lib/Sema/SemaChecking.cpp | 38 +- - clang/lib/Sema/SemaDecl.cpp | 15 +- - clang/lib/Sema/SemaDeclAttr.cpp | 9 +- - clang/lib/Sema/SemaDeclCXX.cpp | 10 + - clang/lib/Sema/SemaLookup.cpp | 19 +- - clang/lib/Sema/SemaType.cpp | 14 +- - .../CodeGenOpenCL/addr-space-struct-arg.cl | 7 +- - .../address-spaces-conversions.cl | 2 + - .../CodeGenOpenCL/address-spaces-mangling.cl | 3 +- - clang/test/CodeGenOpenCL/address-spaces.cl | 4 + - .../amdgcn-automatic-variable.cl | 1 + - .../CodeGenOpenCL/amdgpu-sizeof-alignof.cl | 21 +- - .../CodeGenOpenCL/arm-integer-dot-product.cl | 1 + - .../test/CodeGenOpenCL/cl-uniform-wg-size.cl | 2 + - clang/test/CodeGenOpenCL/fpmath.cl | 2 + - .../generic-address-space-feature.cl | 28 + - .../intel-subgroups-avc-ext-types.cl | 1 + - .../kernels-have-spir-cc-by-default.cl | 3 + - clang/test/CodeGenOpenCL/logical-ops.cl | 1 + - clang/test/CodeGenOpenCL/no-half.cl | 1 + - clang/test/CodeGenOpenCL/pipe_builtin.cl | 3 + - clang/test/CodeGenOpenCL/pipe_types.cl | 1 + - clang/test/CodeGenOpenCL/printf.cl | 2 + - clang/test/CodeGenOpenCL/unroll-hint.cl | 1 + - clang/test/Driver/autocomplete.c | 2 + - clang/test/Driver/opencl.cl | 2 + - clang/test/Driver/unknown-std.cl | 1 + - clang/test/Frontend/stdlang.c | 1 + - clang/test/Headers/opencl-c-header.cl | 7 +- - clang/test/Index/pipe-size.cl | 7 + - clang/test/Preprocessor/predefined-macros.c | 13 + - .../Sema/feature-extensions-simult-support.cl | 75 + - clang/test/Sema/features-ignore-pragma.cl | 24 + - clang/test/Sema/opencl-features-pipes.cl | 18 + - clang/test/Sema/opencl-features.cl | 128 + - clang/test/Sema/pipe_builtins_feature.cl | 21 + - .../address-spaces-conversions-cl2.0.cl | 3 + - clang/test/SemaOpenCL/address-spaces.cl | 1 + - .../SemaOpenCL/cl20-device-side-enqueue.cl | 16 +- - .../SemaOpenCL/forget-unsupported-builtins.cl | 22 + - clang/test/SemaOpenCL/image-features.cl | 20 + - .../SemaOpenCL/invalid-pipe-builtin-cl2.0.cl | 1 + - clang/test/SemaOpenCL/storageclass-cl20.cl | 1 + - .../TableGen/ClangOpenCLBuiltinEmitter.cpp | 35 +- - 67 files changed, 3656 insertions(+), 723 deletions(-) - create mode 100644 clang/test/CodeGenOpenCL/generic-address-space-feature.cl - create mode 100644 clang/test/Sema/feature-extensions-simult-support.cl - create mode 100644 clang/test/Sema/features-ignore-pragma.cl - create mode 100644 clang/test/Sema/opencl-features-pipes.cl - create mode 100644 clang/test/Sema/opencl-features.cl - create mode 100644 clang/test/Sema/pipe_builtins_feature.cl - create mode 100644 clang/test/SemaOpenCL/forget-unsupported-builtins.cl - create mode 100644 clang/test/SemaOpenCL/image-features.cl - -diff --git a/clang/include/clang/Basic/Builtins.def b/clang/include/clang/Basic/Builtins.def -index 1a6c85ce2dd3..b76e277f0337 100644 ---- a/clang/include/clang/Basic/Builtins.def -+++ b/clang/include/clang/Basic/Builtins.def -@@ -106,6 +106,10 @@ - # define LANGBUILTIN(ID, TYPE, ATTRS, BUILTIN_LANG) BUILTIN(ID, TYPE, ATTRS) - #endif - -+#if defined(BUILTIN) && !defined(OPENCLBUILTIN) -+# define OPENCLBUILTIN(ID, TYPE, ATTRS, BUILTIN_LANG, FEATURE) BUILTIN(ID, TYPE, ATTRS) -+#endif -+ - // Standard libc/libm functions: - BUILTIN(__builtin_atan2 , "ddd" , "Fne") - BUILTIN(__builtin_atan2f, "fff" , "Fne") -@@ -1514,50 +1518,54 @@ BUILTIN(__builtin_coro_param, "bv*v*", "n") - - // OpenCL v2.0 s6.13.16, s9.17.3.5 - Pipe functions. - // We need the generic prototype, since the packet type could be anything. --LANGBUILTIN(read_pipe, "i.", "tn", OCLC20_LANG) --LANGBUILTIN(write_pipe, "i.", "tn", OCLC20_LANG) -+OPENCLBUILTIN(read_pipe, "i.", "tn", OCLC2P_LANG, "__opencl_c_pipes") -+OPENCLBUILTIN(write_pipe, "i.", "tn", OCLC2P_LANG, "__opencl_c_pipes") - --LANGBUILTIN(reserve_read_pipe, "i.", "tn", OCLC20_LANG) --LANGBUILTIN(reserve_write_pipe, "i.", "tn", OCLC20_LANG) -+OPENCLBUILTIN(reserve_read_pipe, "i.", "tn", OCLC2P_LANG, "__opencl_c_pipes") -+OPENCLBUILTIN(reserve_write_pipe, "i.", "tn", OCLC2P_LANG, "__opencl_c_pipes") - --LANGBUILTIN(commit_write_pipe, "v.", "tn", OCLC20_LANG) --LANGBUILTIN(commit_read_pipe, "v.", "tn", OCLC20_LANG) -+OPENCLBUILTIN(commit_write_pipe, "v.", "tn", OCLC2P_LANG, "__opencl_c_pipes") -+OPENCLBUILTIN(commit_read_pipe, "v.", "tn", OCLC2P_LANG, "__opencl_c_pipes") - --LANGBUILTIN(sub_group_reserve_read_pipe, "i.", "tn", OCLC20_LANG) --LANGBUILTIN(sub_group_reserve_write_pipe, "i.", "tn", OCLC20_LANG) -+OPENCLBUILTIN(sub_group_reserve_read_pipe, "i.", "tn", OCLC2P_LANG, "__opencl_c_pipes") -+OPENCLBUILTIN(sub_group_reserve_write_pipe, "i.", "tn", OCLC2P_LANG, "__opencl_c_pipes") - --LANGBUILTIN(sub_group_commit_read_pipe, "v.", "tn", OCLC20_LANG) --LANGBUILTIN(sub_group_commit_write_pipe, "v.", "tn", OCLC20_LANG) -+OPENCLBUILTIN(sub_group_commit_read_pipe, "v.", "tn", OCLC2P_LANG, "__opencl_c_pipes") -+OPENCLBUILTIN(sub_group_commit_write_pipe, "v.", "tn", OCLC2P_LANG, "__opencl_c_pipes") - --LANGBUILTIN(work_group_reserve_read_pipe, "i.", "tn", OCLC20_LANG) --LANGBUILTIN(work_group_reserve_write_pipe, "i.", "tn", OCLC20_LANG) -+OPENCLBUILTIN(work_group_reserve_read_pipe, "i.", "tn", OCLC2P_LANG, "__opencl_c_pipes") -+OPENCLBUILTIN(work_group_reserve_write_pipe, "i.", "tn", OCLC2P_LANG, "__opencl_c_pipes") - --LANGBUILTIN(work_group_commit_read_pipe, "v.", "tn", OCLC20_LANG) --LANGBUILTIN(work_group_commit_write_pipe, "v.", "tn", OCLC20_LANG) -+OPENCLBUILTIN(work_group_commit_read_pipe, "v.", "tn", OCLC2P_LANG, "__opencl_c_pipes") -+OPENCLBUILTIN(work_group_commit_write_pipe, "v.", "tn", OCLC2P_LANG, "__opencl_c_pipes") - --LANGBUILTIN(get_pipe_num_packets, "Ui.", "tn", OCLC20_LANG) --LANGBUILTIN(get_pipe_max_packets, "Ui.", "tn", OCLC20_LANG) -+OPENCLBUILTIN(get_pipe_num_packets, "Ui.", "tn", OCLC2P_LANG, "__opencl_c_pipes") -+OPENCLBUILTIN(get_pipe_max_packets, "Ui.", "tn", OCLC2P_LANG, "__opencl_c_pipes") - - // OpenCL v2.0 s6.13.17 - Enqueue kernel functions. - // Custom builtin check allows to perform special check of passed block arguments. --LANGBUILTIN(enqueue_kernel, "i.", "tn", OCLC20_LANG) --LANGBUILTIN(get_kernel_work_group_size, "Ui.", "tn", OCLC20_LANG) --LANGBUILTIN(get_kernel_preferred_work_group_size_multiple, "Ui.", "tn", OCLC20_LANG) --LANGBUILTIN(get_kernel_max_sub_group_size_for_ndrange, "Ui.", "tn", OCLC20_LANG) --LANGBUILTIN(get_kernel_sub_group_count_for_ndrange, "Ui.", "tn", OCLC20_LANG) -+OPENCLBUILTIN(enqueue_kernel, "i.", "tn", OCLC2P_LANG, -+ "__opencl_c_device_enqueue") -+OPENCLBUILTIN(get_kernel_work_group_size, "Ui.", "tn", OCLC2P_LANG, -+ "__opencl_c_device_enqueue") -+OPENCLBUILTIN(get_kernel_preferred_work_group_size_multiple, "Ui.", "tn", -+ OCLC2P_LANG, "__opencl_c_device_enqueue") -+OPENCLBUILTIN(get_kernel_max_sub_group_size_for_ndrange, "Ui.", "tn", -+ OCLC2P_LANG, "__opencl_c_device_enqueue") -+OPENCLBUILTIN(get_kernel_sub_group_count_for_ndrange, "Ui.", "tn", OCLC2P_LANG, "__opencl_c_device_enqueue") - - // OpenCL v2.0 s6.13.9 - Address space qualifier functions. - // FIXME: Pointer parameters of OpenCL builtins should have their address space - // requirement defined. --LANGBUILTIN(to_global, "v*v*", "tn", OCLC20_LANG) --LANGBUILTIN(to_local, "v*v*", "tn", OCLC20_LANG) --LANGBUILTIN(to_private, "v*v*", "tn", OCLC20_LANG) -+OPENCLBUILTIN(to_global, "v*v*", "tn", OCLC2P_LANG, "__opencl_c_generic_address_space") -+OPENCLBUILTIN(to_local, "v*v*", "tn", OCLC2P_LANG, "__opencl_c_generic_address_space") -+OPENCLBUILTIN(to_private, "v*v*", "tn", OCLC2P_LANG, "__opencl_c_generic_address_space") - - // OpenCL half load/store builtin --LANGBUILTIN(__builtin_store_half, "vdh*", "n", ALL_OCLC_LANGUAGES) --LANGBUILTIN(__builtin_store_halff, "vfh*", "n", ALL_OCLC_LANGUAGES) --LANGBUILTIN(__builtin_load_half, "dhC*", "nc", ALL_OCLC_LANGUAGES) --LANGBUILTIN(__builtin_load_halff, "fhC*", "nc", ALL_OCLC_LANGUAGES) -+OPENCLBUILTIN(__builtin_store_half, "vdh*", "n", ALL_OCLC_LANGUAGES, "") -+OPENCLBUILTIN(__builtin_store_halff, "vfh*", "n", ALL_OCLC_LANGUAGES, "") -+OPENCLBUILTIN(__builtin_load_half, "dhC*", "nc", ALL_OCLC_LANGUAGES, "") -+OPENCLBUILTIN(__builtin_load_halff, "fhC*", "nc", ALL_OCLC_LANGUAGES, "") - - // Builtins for os_log/os_trace - BUILTIN(__builtin_os_log_format_buffer_size, "zcC*.", "p:0:nut") -@@ -1578,3 +1586,4 @@ BUILTIN(__builtin_ms_va_copy, "vc*&c*&", "n") - #undef BUILTIN - #undef LIBBUILTIN - #undef LANGBUILTIN -+#undef OPENCLBUILTIN -diff --git a/clang/include/clang/Basic/Builtins.h b/clang/include/clang/Basic/Builtins.h -index e4ed482d9068..713ea4bc267d 100644 ---- a/clang/include/clang/Basic/Builtins.h -+++ b/clang/include/clang/Basic/Builtins.h -@@ -33,13 +33,13 @@ enum LanguageID { - CXX_LANG = 0x4, // builtin for cplusplus only. - OBJC_LANG = 0x8, // builtin for objective-c and objective-c++ - MS_LANG = 0x10, // builtin requires MS mode. -- OCLC20_LANG = 0x20, // builtin for OpenCL C 2.0 only. -+ OCLC2P_LANG = 0x20, // builtin for OpenCL C 2.0+ versions. - OCLC1X_LANG = 0x40, // builtin for OpenCL C 1.x only. - OMP_LANG = 0x80, // builtin requires OpenMP. - ALL_LANGUAGES = C_LANG | CXX_LANG | OBJC_LANG, // builtin for all languages. - ALL_GNU_LANGUAGES = ALL_LANGUAGES | GNU_LANG, // builtin requires GNU mode. - ALL_MS_LANGUAGES = ALL_LANGUAGES | MS_LANG, // builtin requires MS mode. -- ALL_OCLC_LANGUAGES = OCLC1X_LANG | OCLC20_LANG // builtin for OCLC languages. -+ ALL_OCLC_LANGUAGES = OCLC1X_LANG | OCLC2P_LANG // builtin for OCLC languages. - }; - - namespace Builtin { -@@ -228,6 +228,10 @@ public: - /// for non-builtins. - bool canBeRedeclared(unsigned ID) const; - -+ bool requiresFeatures(unsigned ID) const { -+ return requiresFeatures(getRecord(ID)); -+ } -+ - private: - const Info &getRecord(unsigned ID) const; - -@@ -235,6 +239,11 @@ private: - bool builtinIsSupported(const Builtin::Info &BuiltinInfo, - const LangOptions &LangOpts); - -+ bool OclBuiltinIsSupported(const Builtin::Info &BuiltinInfo, -+ const LangOptions &LangOpts) const; -+ -+ bool requiresFeatures(const Builtin::Info &BuiltinInfo) const; -+ - /// Helper function for isPrintfLike and isScanfLike. - bool isLike(unsigned ID, unsigned &FormatIdx, bool &HasVAListArg, - const char *Fmt) const; -diff --git a/clang/include/clang/Basic/DiagnosticParseKinds.td b/clang/include/clang/Basic/DiagnosticParseKinds.td -index 33adf093693f..43ea63586f91 100644 ---- a/clang/include/clang/Basic/DiagnosticParseKinds.td -+++ b/clang/include/clang/Basic/DiagnosticParseKinds.td -@@ -1199,6 +1199,8 @@ def warn_pragma_unsupported_extension : Warning< - "unsupported OpenCL extension %0 - ignoring">, InGroup<IgnoredPragmas>; - def warn_pragma_extension_is_core : Warning< - "OpenCL extension %0 is core feature or supported optional core feature - ignoring">, InGroup<DiagGroup<"pedantic-core-features">>, DefaultIgnore; -+def warn_opencl_pragma_feature_ignore : Warning< -+ "OpenCL feature support can't be controlled via pragma, ignoring">, InGroup<IgnoredPragmas>; - - // OpenCL errors. - def err_opencl_taking_function_address_parser : Error< -diff --git a/clang/include/clang/Basic/DiagnosticSemaKinds.td b/clang/include/clang/Basic/DiagnosticSemaKinds.td -index 917377420505..91648aa27820 100644 ---- a/clang/include/clang/Basic/DiagnosticSemaKinds.td -+++ b/clang/include/clang/Basic/DiagnosticSemaKinds.td -@@ -9378,6 +9378,13 @@ def ext_opencl_ext_vector_type_rgba_selector: ExtWarn< - def err_openclcxx_placement_new : Error< - "use of placement new requires explicit declaration">; - -+def err_opencl_pipes_require_feat : Error< -+ "usage of OpenCL pipes requires feature support">; -+def err_opencl_memory_scope_require_feat : Error< -+ "usage of memory scope requires feature support">; -+def err_opencl_memory_ordering_require_feat : Error< -+ "usage of memory ordering requires feature support">; -+ - // MIG routine annotations. - def warn_mig_server_routine_does_not_return_kern_return_t : Warning< - "'mig_server_routine' attribute only applies to routines that return a kern_return_t">, -diff --git a/clang/include/clang/Basic/LangOptions.def b/clang/include/clang/Basic/LangOptions.def -index 3319a3123976..406f45c0be69 100644 ---- a/clang/include/clang/Basic/LangOptions.def -+++ b/clang/include/clang/Basic/LangOptions.def -@@ -203,6 +203,8 @@ LANGOPT(OpenCL , 1, 0, "OpenCL") - LANGOPT(OpenCLVersion , 32, 0, "OpenCL C version") - LANGOPT(OpenCLCPlusPlus , 1, 0, "C++ for OpenCL") - LANGOPT(OpenCLCPlusPlusVersion , 32, 0, "C++ for OpenCL version") -+LANGOPT(OpenCLGenericKeyword , 1, 0, "OpenCL generic keyword") -+LANGOPT(OpenCLPipeKeyword , 1, 0, "OpenCL pipe keyword") - LANGOPT(NativeHalfType , 1, 0, "Native half type support") - LANGOPT(NativeHalfArgsAndReturns, 1, 0, "Native half args and returns") - LANGOPT(HalfArgsAndReturns, 1, 0, "half args and returns") -diff --git a/clang/include/clang/Basic/LangStandards.def b/clang/include/clang/Basic/LangStandards.def -index 7f1a24db7e9b..69aaba3ff5a2 100644 ---- a/clang/include/clang/Basic/LangStandards.def -+++ b/clang/include/clang/Basic/LangStandards.def -@@ -167,6 +167,9 @@ LANGSTANDARD(opencl12, "cl1.2", - LANGSTANDARD(opencl20, "cl2.0", - OpenCL, "OpenCL 2.0", - LineComment | C99 | Digraphs | HexFloat | OpenCL) -+LANGSTANDARD(opencl30, "cl3.0", -+ OpenCL, "OpenCL 3.0", -+ LineComment | C99 | Digraphs | HexFloat | OpenCL) - LANGSTANDARD(openclcpp, "clc++", - OpenCL, "C++ for OpenCL", - LineComment | CPlusPlus | CPlusPlus11 | CPlusPlus14 | CPlusPlus17 | -@@ -176,6 +179,7 @@ LANGSTANDARD_ALIAS_DEPR(opencl10, "CL") - LANGSTANDARD_ALIAS_DEPR(opencl11, "CL1.1") - LANGSTANDARD_ALIAS_DEPR(opencl12, "CL1.2") - LANGSTANDARD_ALIAS_DEPR(opencl20, "CL2.0") -+LANGSTANDARD_ALIAS_DEPR(opencl30, "CL3.0") - LANGSTANDARD_ALIAS_DEPR(openclcpp, "CLC++") - - // CUDA -diff --git a/clang/include/clang/Basic/OpenCLExtensions.def b/clang/include/clang/Basic/OpenCLExtensions.def -index 517481584313..608f78a13eef 100644 ---- a/clang/include/clang/Basic/OpenCLExtensions.def -+++ b/clang/include/clang/Basic/OpenCLExtensions.def -@@ -93,6 +93,21 @@ OPENCLEXT_INTERNAL(cl_intel_subgroups, 120, ~0U) - OPENCLEXT_INTERNAL(cl_intel_subgroups_short, 120, ~0U) - OPENCLEXT_INTERNAL(cl_intel_device_side_avc_motion_estimation, 120, ~0U) - -+OPENCLEXT_INTERNAL(__opencl_c_pipes, 200, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_generic_address_space, 200, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_work_group_collective_functions, 200, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_atomic_order_acq_rel, 200, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_atomic_order_seq_cst, 200, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_atomic_scope_device, 200, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_atomic_scope_all_devices, 200, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_subgroups, 200, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_3d_image_writes, 100, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_device_enqueue, 200, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_read_write_images, 200, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_program_scope_global_variables, 200, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_fp64, 120, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_int64, 100, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_images, 100, ~0U) - #undef OPENCLEXT_INTERNAL - - #ifdef OPENCLEXT -diff --git a/clang/include/clang/Basic/OpenCLOptions.h b/clang/include/clang/Basic/OpenCLOptions.h -index 15661154eab5..9a3a8e33cadd 100644 ---- a/clang/include/clang/Basic/OpenCLOptions.h -+++ b/clang/include/clang/Basic/OpenCLOptions.h -@@ -16,11 +16,16 @@ - - #include "clang/Basic/LangOptions.h" - #include "llvm/ADT/StringMap.h" -+#include "llvm/ADT/StringSwitch.h" - - namespace clang { - - /// OpenCL supported extensions and optional core features - class OpenCLOptions { -+ // OpenCL Version -+ unsigned CLVer = 120; -+ bool IsOpenCLCPlusPlus = false; -+ - struct Info { - bool Supported; // Is this option supported - bool Enabled; // Is this option enabled -@@ -31,7 +36,38 @@ class OpenCLOptions { - :Supported(S), Enabled(E), Avail(A), Core(C){} - }; - llvm::StringMap<Info> OptMap; -+ - public: -+ void setOpenCLVersion(const LangOptions &LO) { -+ IsOpenCLCPlusPlus = LO.OpenCLCPlusPlus; -+ CLVer = IsOpenCLCPlusPlus ? 200 : LO.OpenCLVersion; -+ } -+ -+ // Get extension which is semantically equivalent to a given feature -+ // if exists (e.g. __opencl_c_subgroups -> cl_khr_subgroups) -+ llvm::Optional<StringRef> getEquivalentExtension(StringRef Feature) const { -+ return llvm::StringSwitch<llvm::Optional<StringRef>>(Feature) -+ .Case("__opencl_c_3d_image_writes", -+ Optional<StringRef>("cl_khr_3d_image_writes")) -+ .Case("__opencl_c_subgroups", Optional<StringRef>("cl_khr_subgroups")) -+ .Case("__opencl_c_fp64", Optional<StringRef>("cl_khr_fp64")) -+ .Default(Optional<StringRef>()); -+ } -+ -+ // Same as above but for extensions -+ llvm::Optional<StringRef> getEquivalentFeature(StringRef Extension) const { -+ return llvm::StringSwitch<llvm::Optional<StringRef>>(Extension) -+ .Case("cl_khr_3d_image_writes", -+ Optional<StringRef>("__opencl_c_3d_image_writes")) -+ .Case("cl_khr_subgroups", Optional<StringRef>("__opencl_c_subgroups")) -+ .Case("cl_khr_fp64", Optional<StringRef>("__opencl_c_fp64")) -+ .Default(Optional<StringRef>()); -+ } -+ -+ bool isFeature(llvm::StringRef Ext) const { -+ return Ext.startswith("__opencl_c"); -+ } -+ - bool isKnown(llvm::StringRef Ext) const { - return OptMap.find(Ext) != OptMap.end(); - } -@@ -42,32 +78,88 @@ public: - - // Is supported as either an extension or an (optional) core feature for - // OpenCL version \p CLVer. -- bool isSupported(llvm::StringRef Ext, const LangOptions &LO) const { -+ bool isSupported(llvm::StringRef Ext) const { - // In C++ mode all extensions should work at least as in v2.0. -- auto CLVer = LO.OpenCLCPlusPlus ? 200 : LO.OpenCLVersion; - auto I = OptMap.find(Ext)->getValue(); - return I.Supported && I.Avail <= CLVer; - } - - // Is supported (optional) OpenCL core features for OpenCL version \p CLVer. - // For supported extension, return false. -- bool isSupportedCore(llvm::StringRef Ext, const LangOptions &LO) const { -+ bool isSupportedCore(llvm::StringRef Ext) const { - // In C++ mode all extensions should work at least as in v2.0. -- auto CLVer = LO.OpenCLCPlusPlus ? 200 : LO.OpenCLVersion; - auto I = OptMap.find(Ext)->getValue(); - return I.Supported && I.Avail <= CLVer && I.Core != ~0U && CLVer >= I.Core; - } - - // Is supported OpenCL extension for OpenCL version \p CLVer. - // For supported (optional) core feature, return false. -- bool isSupportedExtension(llvm::StringRef Ext, const LangOptions &LO) const { -+ bool isSupportedExtension(llvm::StringRef Ext) const { - // In C++ mode all extensions should work at least as in v2.0. -- auto CLVer = LO.OpenCLCPlusPlus ? 200 : LO.OpenCLVersion; - auto I = OptMap.find(Ext)->getValue(); - return I.Supported && I.Avail <= CLVer && (I.Core == ~0U || CLVer < I.Core); - } - -+ // Support features whose support is directly related to the -+ // specific OpenCL version. For example, OpenCL 2.0 supports -+ // all features that are optional in 3.0 -+ void adjustFeatures() { -+ // Support int64 by default (assume compiling for FULL profile) -+ OptMap["__opencl_c_int64"].Supported = true; -+ -+ if (CLVer >= 300) { -+ // Simultaneously support extension and corresponding feature -+ for (llvm::StringRef F : -+ {"__opencl_c_subgroups", "__opencl_c_3d_image_writes", -+ "__opencl_c_fp64"}) { -+ auto Ext = getEquivalentExtension(F); -+ OptMap[*Ext].Supported = OptMap[F].Supported; -+ } -+ -+ // OpenCL C compilers that define the feature macro __opencl_c_pipes or -+ // or __opencl_c_device_enqueue must also define the -+ // feature macro __opencl_c_generic_address_space. -+ if (OptMap["__opencl_c_pipes"].Supported || -+ OptMap["__opencl_c_device_enqueue"].Supported) -+ OptMap["__opencl_c_generic_address_space"].Supported = true; -+ -+ // OpenCL C compilers that define the feature macro -+ // __opencl_c_3d_image_writes or __opencl_c_read_write_images must also -+ // define the feature macro __opencl_c_images -+ if (OptMap["__opencl_c_3d_image_writes"].Supported || -+ OptMap["__opencl_c_read_write_images"].Supported) -+ OptMap["__opencl_c_images"].Supported = true; -+ -+ // All other features are already supported with options -+ // or in target settings -+ return; -+ } -+ -+ auto FeaturesRange = llvm::make_filter_range( -+ OptMap, [&](llvm::StringMapEntry<Info> &OptVal) { -+ auto Opt = OptVal.getKey(); -+ return isFeature(Opt); -+ }); -+ -+ for (auto &It : FeaturesRange) { -+ auto &Info = It.getValue(); -+ // For OpenCL version less then 3.0 some -+ // features should be supported simulateneously -+ // with specific extension -+ if (Optional<llvm::StringRef> Ext = getEquivalentExtension(It.getKey())) -+ Info.Supported = Info.Enabled = OptMap[*Ext].Supported; -+ else if (Info.Avail <= CLVer) -+ Info.Supported = Info.Enabled = true; -+ } -+ } -+ - void enable(llvm::StringRef Ext, bool V = true) { -+ // Ignore disabling extensions if corresponding features -+ // already supported for OpenCL version higher then 3.0 -+ if (CLVer >= 300) -+ if (Optional<llvm::StringRef> F = getEquivalentFeature(Ext)) -+ if (V != OptMap[*F].Enabled) -+ return; - OptMap[Ext].Enabled = V; - } - -@@ -96,7 +188,7 @@ public: - OptMap[Ext].Supported = V; - } - -- OpenCLOptions(){ -+ OpenCLOptions() { - #define OPENCLEXT_INTERNAL(Ext, AvailVer, CoreVer) \ - OptMap[#Ext].Avail = AvailVer; \ - OptMap[#Ext].Core = CoreVer; -@@ -104,35 +196,86 @@ public: - } - - void addSupport(const OpenCLOptions &Opts) { -+ assert(IsOpenCLCPlusPlus == Opts.IsOpenCLCPlusPlus && CLVer == Opts.CLVer); - for (auto &I:Opts.OptMap) -- if (I.second.Supported) -+ if (I.second.Supported) { - OptMap[I.getKey()].Supported = true; -+ // All features are enabled as they are supported -+ if (isFeature(I.getKey())) -+ OptMap[I.getKey()].Enabled = true; -+ } -+ if (CLVer >= 300) { -+ // Enabling extensions with respect to features -+ for (llvm::StringRef Ext : -+ {"cl_khr_3d_image_writes", "cl_khr_subgroups", "cl_khr_fp64"}) { -+ auto Feature = getEquivalentFeature(Ext); -+ enable(Ext, OptMap[*Feature].Enabled); -+ } -+ } - } - - void copy(const OpenCLOptions &Opts) { -+ CLVer = Opts.CLVer; -+ IsOpenCLCPlusPlus = Opts.IsOpenCLCPlusPlus; - OptMap = Opts.OptMap; - } - - // Turn on or off support of all options. - void supportAll(bool On = true) { -- for (llvm::StringMap<Info>::iterator I = OptMap.begin(), -- E = OptMap.end(); I != E; ++I) -- I->second.Supported = On; -+ for (llvm::StringMap<Info>::iterator I = OptMap.begin(), E = OptMap.end(); -+ I != E; ++I) -+ if (!isFeature(I->getKey())) -+ I->second.Supported = On; - } - - void disableAll() { -- for (llvm::StringMap<Info>::iterator I = OptMap.begin(), -- E = OptMap.end(); I != E; ++I) -- I->second.Enabled = false; -+ for (llvm::StringMap<Info>::iterator I = OptMap.begin(), E = OptMap.end(); -+ I != E; ++I) { -+ auto Ext = I->getKey(); -+ if (!isFeature(Ext)) -+ enable(Ext, false); -+ } - } - -- void enableSupportedCore(LangOptions LO) { -+ void enableSupportedCore() { - for (llvm::StringMap<Info>::iterator I = OptMap.begin(), E = OptMap.end(); - I != E; ++I) -- if (isSupportedCore(I->getKey(), LO)) -+ if (isSupportedCore(I->getKey())) - I->second.Enabled = true; - } - -+ // This enum specifies how OpenCL versions map into values -+ // for encoding. This is used when generating built-ins -+ // from tablegen -+ enum OpenCLVersionsEncodings : unsigned short { -+ OPENCL_C_100_CODE = 0x1, -+ OPENCL_C_110_CODE = 0x2, -+ OPENCL_C_120_CODE = 0x4, -+ OPENCL_C_200_CODE = 0x8, -+ OPENCL_C_300_CODE = 0x10, -+ OPENCL_C_ALL_CODE = 0x1f -+ }; -+ -+ // Encode version into single integer -+ static unsigned short EncodeVersion(unsigned OpenCLVersion) { -+ switch (OpenCLVersion) { -+ default: -+ llvm_unreachable("Unknown OpenCL version"); -+ case 0: -+ return OpenCLVersionsEncodings::OPENCL_C_ALL_CODE; -+ case 100: -+ return OpenCLVersionsEncodings::OPENCL_C_100_CODE; -+ case 110: -+ return OpenCLVersionsEncodings::OPENCL_C_110_CODE; -+ case 120: -+ return OpenCLVersionsEncodings::OPENCL_C_120_CODE; -+ case 200: -+ return OpenCLVersionsEncodings::OPENCL_C_200_CODE; -+ case 300: -+ return OpenCLVersionsEncodings::OPENCL_C_300_CODE; -+ } -+ } -+ - friend class ASTWriter; - friend class ASTReader; - }; -diff --git a/clang/include/clang/Driver/Options.td b/clang/include/clang/Driver/Options.td -index 391c895a453b..e03c22c749ad 100644 ---- a/clang/include/clang/Driver/Options.td -+++ b/clang/include/clang/Driver/Options.td -@@ -522,7 +522,7 @@ def cl_mad_enable : Flag<["-"], "cl-mad-enable">, Group<opencl_Group>, Flags<[CC - def cl_no_signed_zeros : Flag<["-"], "cl-no-signed-zeros">, Group<opencl_Group>, Flags<[CC1Option]>, - HelpText<"OpenCL only. Allow use of less precise no signed zeros computations in the generated binary.">; - def cl_std_EQ : Joined<["-"], "cl-std=">, Group<opencl_Group>, Flags<[CC1Option]>, -- HelpText<"OpenCL language standard to compile for.">, Values<"cl,CL,cl1.1,CL1.1,cl1.2,CL1.2,cl2.0,CL2.0,clc++,CLC++">; -+ HelpText<"OpenCL language standard to compile for.">, Values<"cl,CL,cl1.1,CL1.1,cl1.2,CL1.2,cl2.0,CL2.0,cl3.0,CL3.0,clc++,CLC++">; - def cl_denorms_are_zero : Flag<["-"], "cl-denorms-are-zero">, Group<opencl_Group>, Flags<[CC1Option]>, - HelpText<"OpenCL only. Allow denormals to be flushed to zero.">; - def cl_fp32_correctly_rounded_divide_sqrt : Flag<["-"], "cl-fp32-correctly-rounded-divide-sqrt">, Group<opencl_Group>, Flags<[CC1Option]>, -diff --git a/clang/include/clang/Sema/Sema.h b/clang/include/clang/Sema/Sema.h -index 842e49602274..d8ac91bc8a31 100644 ---- a/clang/include/clang/Sema/Sema.h -+++ b/clang/include/clang/Sema/Sema.h -@@ -9632,6 +9632,10 @@ public: - /// \return true if type is disabled. - bool checkOpenCLDisabledDecl(const NamedDecl &D, const Expr &E); - -+ bool checkOpenCLSubgroupExtForCallExpr(CallExpr *Call); -+ -+ bool isSupportedOpenCLOMemoryOrdering(int64_t Ordering) const; -+ - //===--------------------------------------------------------------------===// - // OpenMP directives and clauses. - // -@@ -11102,6 +11106,11 @@ public: - /// that the user intended an assignment used as condition. - void DiagnoseEqualityWithExtraParens(ParenExpr *ParenE); - -+ template <typename DiagLocT, typename DiagInfoT> -+ void DiagnoseOpenCLRequiresOption(llvm::StringRef OpenCLOptName, -+ DiagLocT DiagLoc, DiagInfoT DiagInfo, -+ unsigned Selector, SourceRange SrcRange); -+ - /// CheckCXXBooleanCondition - Returns true if conversion to bool is invalid. - ExprResult CheckCXXBooleanCondition(Expr *CondExpr, bool IsConstexpr = false); - -diff --git a/clang/lib/AST/ASTContext.cpp b/clang/lib/AST/ASTContext.cpp -index 1be72efe4de8..2a81fbcde79d 100644 ---- a/clang/lib/AST/ASTContext.cpp -+++ b/clang/lib/AST/ASTContext.cpp -@@ -1490,7 +1490,8 @@ void ASTContext::InitBuiltinTypes(const TargetInfo &Target, - ObjCSuperType = QualType(); - - // void * type -- if (LangOpts.OpenCLVersion >= 200) { -+ if (Target.getSupportedOpenCLOpts().isEnabled( -+ "__opencl_c_generic_address_space")) { - auto Q = VoidTy.getQualifiers(); - Q.setAddressSpace(LangAS::opencl_generic); - VoidPtrTy = getPointerType(getCanonicalType( -diff --git a/clang/lib/Basic/Builtins.cpp b/clang/lib/Basic/Builtins.cpp -index 0cd89df41b67..7a3067345098 100644 ---- a/clang/lib/Basic/Builtins.cpp -+++ b/clang/lib/Basic/Builtins.cpp -@@ -23,6 +23,8 @@ static const Builtin::Info BuiltinInfo[] = { - { #ID, TYPE, ATTRS, nullptr, ALL_LANGUAGES, nullptr }, - #define LANGBUILTIN(ID, TYPE, ATTRS, LANGS) \ - { #ID, TYPE, ATTRS, nullptr, LANGS, nullptr }, -+#define OPENCLBUILTIN(ID, TYPE, ATTRS, LANGS, FEATURE) \ -+ {#ID, TYPE, ATTRS, nullptr, LANGS, FEATURE}, - #define LIBBUILTIN(ID, TYPE, ATTRS, HEADER, LANGS) \ - { #ID, TYPE, ATTRS, HEADER, LANGS, nullptr }, - #include "clang/Basic/Builtins.def" -@@ -69,16 +71,17 @@ bool Builtin::Context::builtinIsSupported(const Builtin::Info &BuiltinInfo, - bool ObjCUnsupported = !LangOpts.ObjC && BuiltinInfo.Langs == OBJC_LANG; - bool OclC1Unsupported = (LangOpts.OpenCLVersion / 100) != 1 && - (BuiltinInfo.Langs & ALL_OCLC_LANGUAGES ) == OCLC1X_LANG; -- bool OclC2Unsupported = -- (LangOpts.OpenCLVersion != 200 && !LangOpts.OpenCLCPlusPlus) && -- (BuiltinInfo.Langs & ALL_OCLC_LANGUAGES) == OCLC20_LANG; -+ bool OclC2PUnsupported = -+ (BuiltinInfo.Langs & ALL_OCLC_LANGUAGES) == OCLC2P_LANG && -+ ((LangOpts.OpenCLVersion < 200 && !LangOpts.OpenCLCPlusPlus) || -+ !OclBuiltinIsSupported(BuiltinInfo, LangOpts)); - bool OclCUnsupported = !LangOpts.OpenCL && - (BuiltinInfo.Langs & ALL_OCLC_LANGUAGES); - bool OpenMPUnsupported = !LangOpts.OpenMP && BuiltinInfo.Langs == OMP_LANG; - bool CPlusPlusUnsupported = - !LangOpts.CPlusPlus && BuiltinInfo.Langs == CXX_LANG; - return !BuiltinsUnsupported && !MathBuiltinsUnsupported && !OclCUnsupported && -- !OclC1Unsupported && !OclC2Unsupported && !OpenMPUnsupported && -+ !OclC1Unsupported && !OclC2PUnsupported && !OpenMPUnsupported && - !GnuModeUnsupported && !MSModeUnsupported && !ObjCUnsupported && - !CPlusPlusUnsupported; - } -@@ -191,3 +194,20 @@ bool Builtin::Context::canBeRedeclared(unsigned ID) const { - (!hasReferenceArgsOrResult(ID) && - !hasCustomTypechecking(ID)); - } -+ -+bool Builtin::Context::OclBuiltinIsSupported( -+ const Builtin::Info &BuiltinInfo, const LangOptions &LangOpts) const { -+ if (!requiresFeatures(BuiltinInfo)) -+ return true; -+ -+ return llvm::StringSwitch<bool>(BuiltinInfo.Features) -+ .Case("__opencl_c_device_enqueue", LangOpts.Blocks) -+ .Case("__opencl_c_generic_address_space", LangOpts.OpenCLGenericKeyword) -+ .Case("__opencl_c_pipes", LangOpts.OpenCLPipeKeyword) -+ .Default(false); -+} -+ -+bool Builtin::Context::requiresFeatures( -+ const Builtin::Info &BuiltinInfo) const { -+ return BuiltinInfo.Features && llvm::StringRef(BuiltinInfo.Features) != ""; -+} -diff --git a/clang/lib/Basic/TargetInfo.cpp b/clang/lib/Basic/TargetInfo.cpp -index 3a21a19e1f19..795311d5934e 100644 ---- a/clang/lib/Basic/TargetInfo.cpp -+++ b/clang/lib/Basic/TargetInfo.cpp -@@ -377,6 +377,17 @@ void TargetInfo::adjust(LangOptions &Opts) { - HalfFormat = &llvm::APFloat::IEEEhalf(); - FloatFormat = &llvm::APFloat::IEEEsingle(); - LongDoubleFormat = &llvm::APFloat::IEEEquad(); -+ -+ auto &SupportedOCLOpts = getTargetOpts().SupportedOpenCLOptions; -+ -+ SupportedOCLOpts.setOpenCLVersion(Opts); -+ SupportedOCLOpts.adjustFeatures(); -+ -+ if (!Opts.OpenCLCPlusPlus && Opts.OpenCLVersion >= 200) -+ Opts.Blocks = SupportedOCLOpts.isSupported("__opencl_c_device_enqueue"); -+ Opts.OpenCLGenericKeyword = -+ SupportedOCLOpts.isSupported("__opencl_c_generic_address_space"); -+ Opts.OpenCLPipeKeyword = SupportedOCLOpts.isSupported("__opencl_c_pipes"); - } - - if (Opts.LongDoubleSize) { -diff --git a/clang/lib/Basic/Targets.cpp b/clang/lib/Basic/Targets.cpp -index c063f8ca4472..b394be18a234 100644 ---- a/clang/lib/Basic/Targets.cpp -+++ b/clang/lib/Basic/Targets.cpp -@@ -39,7 +39,6 @@ - #include "clang/Basic/Diagnostic.h" - #include "llvm/ADT/StringExtras.h" - #include "llvm/ADT/Triple.h" -- - using namespace clang; - - namespace clang { -diff --git a/clang/lib/CodeGen/CodeGenFunction.cpp b/clang/lib/CodeGen/CodeGenFunction.cpp -index 648e6d9c214a..6d839fa61abc 100644 ---- a/clang/lib/CodeGen/CodeGenFunction.cpp -+++ b/clang/lib/CodeGen/CodeGenFunction.cpp -@@ -2303,11 +2303,11 @@ void CodeGenFunction::checkTargetFeatures(SourceLocation Loc, - std::string MissingFeature; - if (BuiltinID) { - SmallVector<StringRef, 1> ReqFeatures; -- const char *FeatureList = -- CGM.getContext().BuiltinInfo.getRequiredFeatures(BuiltinID); - // Return if the builtin doesn't have any required features. -- if (!FeatureList || StringRef(FeatureList) == "") -+ if (!CGM.getContext().BuiltinInfo.requiresFeatures(BuiltinID)) - return; -+ const char *FeatureList = -+ CGM.getContext().BuiltinInfo.getRequiredFeatures(BuiltinID); - StringRef(FeatureList).split(ReqFeatures, ','); - if (!hasRequiredFeatures(ReqFeatures, CGM, FD, MissingFeature)) - CGM.getDiags().Report(Loc, diag::err_builtin_needs_feature) -diff --git a/clang/lib/Frontend/CompilerInvocation.cpp b/clang/lib/Frontend/CompilerInvocation.cpp -index e98a407ac42f..18fa06bf3c6d 100644 ---- a/clang/lib/Frontend/CompilerInvocation.cpp -+++ b/clang/lib/Frontend/CompilerInvocation.cpp -@@ -2298,6 +2298,8 @@ void CompilerInvocation::setLangDefaults(LangOptions &Opts, InputKind IK, - Opts.OpenCLVersion = 120; - else if (LangStd == LangStandard::lang_opencl20) - Opts.OpenCLVersion = 200; -+ else if (LangStd == LangStandard::lang_opencl30) -+ Opts.OpenCLVersion = 300; - else if (LangStd == LangStandard::lang_openclcpp) - Opts.OpenCLCPlusPlusVersion = 100; - -@@ -2498,14 +2500,15 @@ static void ParseLangArgs(LangOptions &Opts, ArgList &Args, InputKind IK, - // -cl-std only applies for OpenCL language standards. - // Override the -std option in this case. - if (const Arg *A = Args.getLastArg(OPT_cl_std_EQ)) { -- LangStandard::Kind OpenCLLangStd -- = llvm::StringSwitch<LangStandard::Kind>(A->getValue()) -- .Cases("cl", "CL", LangStandard::lang_opencl10) -- .Cases("cl1.1", "CL1.1", LangStandard::lang_opencl11) -- .Cases("cl1.2", "CL1.2", LangStandard::lang_opencl12) -- .Cases("cl2.0", "CL2.0", LangStandard::lang_opencl20) -- .Cases("clc++", "CLC++", LangStandard::lang_openclcpp) -- .Default(LangStandard::lang_unspecified); -+ LangStandard::Kind OpenCLLangStd = -+ llvm::StringSwitch<LangStandard::Kind>(A->getValue()) -+ .Cases("cl", "CL", LangStandard::lang_opencl10) -+ .Cases("cl1.1", "CL1.1", LangStandard::lang_opencl11) -+ .Cases("cl1.2", "CL1.2", LangStandard::lang_opencl12) -+ .Cases("cl2.0", "CL2.0", LangStandard::lang_opencl20) -+ .Cases("cl3.0", "CL3.0", LangStandard::lang_opencl30) -+ .Cases("clc++", "CLC++", LangStandard::lang_openclcpp) -+ .Default(LangStandard::lang_unspecified); - - if (OpenCLLangStd == LangStandard::lang_unspecified) { - Diags.Report(diag::err_drv_invalid_value) -@@ -2787,8 +2790,7 @@ static void ParseLangArgs(LangOptions &Opts, ArgList &Args, InputKind IK, - - Opts.RTTI = Opts.CPlusPlus && !Args.hasArg(OPT_fno_rtti); - Opts.RTTIData = Opts.RTTI && !Args.hasArg(OPT_fno_rtti_data); -- Opts.Blocks = Args.hasArg(OPT_fblocks) || (Opts.OpenCL -- && Opts.OpenCLVersion == 200); -+ Opts.Blocks = Args.hasArg(OPT_fblocks); - Opts.BlocksRuntimeOptional = Args.hasArg(OPT_fblocks_runtime_optional); - Opts.Coroutines = Opts.CPlusPlus2a || Args.hasArg(OPT_fcoroutines_ts); - -diff --git a/clang/lib/Frontend/InitPreprocessor.cpp b/clang/lib/Frontend/InitPreprocessor.cpp -index c273cb96d9b9..aefd208e6cd3 100644 ---- a/clang/lib/Frontend/InitPreprocessor.cpp -+++ b/clang/lib/Frontend/InitPreprocessor.cpp -@@ -445,6 +445,9 @@ static void InitializeStandardPredefinedMacros(const TargetInfo &TI, - case 200: - Builder.defineMacro("__OPENCL_C_VERSION__", "200"); - break; -+ case 300: -+ Builder.defineMacro("__OPENCL_C_VERSION__", "300"); -+ break; - default: - llvm_unreachable("Unsupported OpenCL version"); - } -@@ -453,6 +456,7 @@ static void InitializeStandardPredefinedMacros(const TargetInfo &TI, - Builder.defineMacro("CL_VERSION_1_1", "110"); - Builder.defineMacro("CL_VERSION_1_2", "120"); - Builder.defineMacro("CL_VERSION_2_0", "200"); -+ Builder.defineMacro("CL_VERSION_3_0", "300"); - - if (TI.isLittleEndian()) - Builder.defineMacro("__ENDIAN_LITTLE__"); -@@ -1101,7 +1105,7 @@ static void InitializePredefinedMacros(const TargetInfo &TI, - // OpenCL definitions. - if (LangOpts.OpenCL) { - #define OPENCLEXT(Ext) \ -- if (TI.getSupportedOpenCLOpts().isSupported(#Ext, LangOpts)) \ -+ if (TI.getSupportedOpenCLOpts().isSupported(#Ext)) \ - Builder.defineMacro(#Ext); - #include "clang/Basic/OpenCLExtensions.def" - -diff --git a/clang/lib/Headers/opencl-c-base.h b/clang/lib/Headers/opencl-c-base.h -index 430e07d36f62..2cc688ccc3da 100644 ---- a/clang/lib/Headers/opencl-c-base.h -+++ b/clang/lib/Headers/opencl-c-base.h -@@ -9,6 +9,59 @@ - #ifndef _OPENCL_BASE_H_ - #define _OPENCL_BASE_H_ - -+// Add predefined macros to build headers with standalone executable -+#ifndef CL_VERSION_3_0 -+ #define CL_VERSION_3_0 300 -+#endif -+ -+// Define features for 2.0 for header backward compatibility -+#ifndef __opencl_c_int64 -+ #define __opencl_c_int64 1 -+#endif -+#if __OPENCL_C_VERSION__ != CL_VERSION_3_0 -+ #ifndef __opencl_c_images -+ #define __opencl_c_images 1 -+ #endif -+#endif -+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ == CL_VERSION_2_0) -+#ifndef __opencl_c_pipes -+ #define __opencl_c_pipes 1 -+#endif -+#ifndef __opencl_c_generic_address_space -+ #define __opencl_c_generic_address_space 1 -+#endif -+#ifndef __opencl_c_work_group_collective_functions -+ #define __opencl_c_work_group_collective_functions 1 -+#endif -+#ifndef __opencl_c_atomic_order_acq_rel -+ #define __opencl_c_atomic_order_acq_rel 1 -+#endif -+#ifndef __opencl_c_atomic_order_seq_cst -+ #define __opencl_c_atomic_order_seq_cst 1 -+#endif -+#ifndef __opencl_c_atomic_scope_device -+ #define __opencl_c_atomic_scope_device 1 -+#endif -+#ifndef __opencl_c_atomic_scope_all_devices -+ #define __opencl_c_atomic_scope_all_devices 1 -+#endif -+#ifndef __opencl_c_subgroups -+ #define __opencl_c_subgroups 1 -+#endif -+#ifndef __opencl_c_3d_image_writes -+ #define __opencl_c_3d_image_writes 1 -+#endif -+#ifndef __opencl_c_device_enqueue -+ #define __opencl_c_device_enqueue 1 -+#endif -+#ifndef __opencl_c_read_write_images -+ #define __opencl_c_read_write_images 1 -+#endif -+#ifndef __opencl_c_program_scope_global_variables -+ #define __opencl_c_program_scope_global_variables 1 -+#endif -+#endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ == CL_VERSION_2_0) -+ - // built-in scalar data types: - - /** -@@ -115,7 +168,12 @@ typedef half half4 __attribute__((ext_vector_type(4))); - typedef half half8 __attribute__((ext_vector_type(8))); - typedef half half16 __attribute__((ext_vector_type(16))); - #endif --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+ -+#ifndef __opencl_c_fp64 -+ #define __opencl_c_fp64 1 -+#endif -+ - #if __OPENCL_C_VERSION__ < CL_VERSION_1_2 - #pragma OPENCL EXTENSION cl_khr_fp64 : enable - #endif -@@ -281,9 +339,17 @@ typedef uint cl_mem_fence_flags; - typedef enum memory_scope { - memory_scope_work_item = __OPENCL_MEMORY_SCOPE_WORK_ITEM, - memory_scope_work_group = __OPENCL_MEMORY_SCOPE_WORK_GROUP, -+#ifdef __opencl_c_atomic_scope_device - memory_scope_device = __OPENCL_MEMORY_SCOPE_DEVICE, -+#endif -+#ifdef __opencl_c_atomic_scope_all_devices -+ #if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+ memory_scope_all_devices = __OPENCL_MEMORY_SCOPE_ALL_SVM_DEVICES, -+ #endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) - memory_scope_all_svm_devices = __OPENCL_MEMORY_SCOPE_ALL_SVM_DEVICES, --#if defined(cl_intel_subgroups) || defined(cl_khr_subgroups) -+#endif -+#if defined(cl_intel_subgroups) || defined(cl_khr_subgroups) || \ -+ defined(__opencl_c_subgroups) - memory_scope_sub_group = __OPENCL_MEMORY_SCOPE_SUB_GROUP - #endif - } memory_scope; -@@ -301,13 +367,14 @@ typedef enum memory_scope { - #define ATOMIC_FLAG_INIT 0 - - // enum values aligned with what clang uses in EmitAtomicExpr() --typedef enum memory_order --{ -+typedef enum memory_order { - memory_order_relaxed = __ATOMIC_RELAXED, - memory_order_acquire = __ATOMIC_ACQUIRE, - memory_order_release = __ATOMIC_RELEASE, - memory_order_acq_rel = __ATOMIC_ACQ_REL, -+#ifdef __opencl_c_atomic_order_seq_cst - memory_order_seq_cst = __ATOMIC_SEQ_CST -+#endif //__opencl_c_atomic_order_seq_cst - } memory_order; - - #endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -diff --git a/clang/lib/Headers/opencl-c.h b/clang/lib/Headers/opencl-c.h -index 3210f93cc851..93a946cec5b1 100644 ---- a/clang/lib/Headers/opencl-c.h -+++ b/clang/lib/Headers/opencl-c.h -@@ -35,7 +35,6 @@ - #define __purefn __attribute__((pure)) - #define __cnfn __attribute__((const)) - -- - // OpenCL v1.1/1.2/2.0 s6.2.3 - Explicit conversions - - char __ovld __cnfn convert_char_rte(char); -@@ -4632,7 +4631,7 @@ float16 __ovld __cnfn convert_float16(float16); - - // Conversions with double data type parameters or return value. - --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - char __ovld __cnfn convert_char(double); - char __ovld __cnfn convert_char_rte(double); - char __ovld __cnfn convert_char_rtn(double); -@@ -5452,7 +5451,7 @@ double16 __ovld __cnfn convert_double16_rtz(uchar16); - double16 __ovld __cnfn convert_double16_rtz(uint16); - double16 __ovld __cnfn convert_double16_rtz(ulong16); - double16 __ovld __cnfn convert_double16_rtz(ushort16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - #ifdef cl_khr_fp16 - // Convert half types to non-double types. -@@ -6270,7 +6269,7 @@ half16 __ovld __cnfn convert_half16_rtz(float16); - half16 __ovld __cnfn convert_half16_rtz(half16); - - // Convert half types to double types. --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn convert_double(half); - double __ovld __cnfn convert_double_rte(half); - double __ovld __cnfn convert_double_rtp(half); -@@ -6333,7 +6332,7 @@ half16 __ovld __cnfn convert_half16_rte(double16); - half16 __ovld __cnfn convert_half16_rtp(double16); - half16 __ovld __cnfn convert_half16_rtn(double16); - half16 __ovld __cnfn convert_half16_rtz(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - #endif // cl_khr_fp16 - -@@ -6404,14 +6403,14 @@ half16 __ovld __cnfn convert_half16_rtz(double16); - #define as_float8(x) __builtin_astype((x), float8) - #define as_float16(x) __builtin_astype((x), float16) - --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #define as_double(x) __builtin_astype((x), double) - #define as_double2(x) __builtin_astype((x), double2) - #define as_double3(x) __builtin_astype((x), double3) - #define as_double4(x) __builtin_astype((x), double4) - #define as_double8(x) __builtin_astype((x), double8) - #define as_double16(x) __builtin_astype((x), double16) --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - #ifdef cl_khr_fp16 - #define as_half(x) __builtin_astype((x), half) -@@ -6534,14 +6533,14 @@ float3 __ovld __cnfn acos(float3); - float4 __ovld __cnfn acos(float4); - float8 __ovld __cnfn acos(float8); - float16 __ovld __cnfn acos(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn acos(double); - double2 __ovld __cnfn acos(double2); - double3 __ovld __cnfn acos(double3); - double4 __ovld __cnfn acos(double4); - double8 __ovld __cnfn acos(double8); - double16 __ovld __cnfn acos(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn acos(half); - half2 __ovld __cnfn acos(half2); -@@ -6560,14 +6559,14 @@ float3 __ovld __cnfn acosh(float3); - float4 __ovld __cnfn acosh(float4); - float8 __ovld __cnfn acosh(float8); - float16 __ovld __cnfn acosh(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn acosh(double); - double2 __ovld __cnfn acosh(double2); - double3 __ovld __cnfn acosh(double3); - double4 __ovld __cnfn acosh(double4); - double8 __ovld __cnfn acosh(double8); - double16 __ovld __cnfn acosh(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn acosh(half); - half2 __ovld __cnfn acosh(half2); -@@ -6586,14 +6585,14 @@ float3 __ovld __cnfn acospi(float3 x); - float4 __ovld __cnfn acospi(float4 x); - float8 __ovld __cnfn acospi(float8 x); - float16 __ovld __cnfn acospi(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn acospi(double x); - double2 __ovld __cnfn acospi(double2 x); - double3 __ovld __cnfn acospi(double3 x); - double4 __ovld __cnfn acospi(double4 x); - double8 __ovld __cnfn acospi(double8 x); - double16 __ovld __cnfn acospi(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn acospi(half x); - half2 __ovld __cnfn acospi(half2 x); -@@ -6612,14 +6611,14 @@ float3 __ovld __cnfn asin(float3); - float4 __ovld __cnfn asin(float4); - float8 __ovld __cnfn asin(float8); - float16 __ovld __cnfn asin(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn asin(double); - double2 __ovld __cnfn asin(double2); - double3 __ovld __cnfn asin(double3); - double4 __ovld __cnfn asin(double4); - double8 __ovld __cnfn asin(double8); - double16 __ovld __cnfn asin(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn asin(half); - half2 __ovld __cnfn asin(half2); -@@ -6638,14 +6637,14 @@ float3 __ovld __cnfn asinh(float3); - float4 __ovld __cnfn asinh(float4); - float8 __ovld __cnfn asinh(float8); - float16 __ovld __cnfn asinh(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn asinh(double); - double2 __ovld __cnfn asinh(double2); - double3 __ovld __cnfn asinh(double3); - double4 __ovld __cnfn asinh(double4); - double8 __ovld __cnfn asinh(double8); - double16 __ovld __cnfn asinh(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn asinh(half); - half2 __ovld __cnfn asinh(half2); -@@ -6664,14 +6663,14 @@ float3 __ovld __cnfn asinpi(float3 x); - float4 __ovld __cnfn asinpi(float4 x); - float8 __ovld __cnfn asinpi(float8 x); - float16 __ovld __cnfn asinpi(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn asinpi(double x); - double2 __ovld __cnfn asinpi(double2 x); - double3 __ovld __cnfn asinpi(double3 x); - double4 __ovld __cnfn asinpi(double4 x); - double8 __ovld __cnfn asinpi(double8 x); - double16 __ovld __cnfn asinpi(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn asinpi(half x); - half2 __ovld __cnfn asinpi(half2 x); -@@ -6690,14 +6689,14 @@ float3 __ovld __cnfn atan(float3 y_over_x); - float4 __ovld __cnfn atan(float4 y_over_x); - float8 __ovld __cnfn atan(float8 y_over_x); - float16 __ovld __cnfn atan(float16 y_over_x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn atan(double y_over_x); - double2 __ovld __cnfn atan(double2 y_over_x); - double3 __ovld __cnfn atan(double3 y_over_x); - double4 __ovld __cnfn atan(double4 y_over_x); - double8 __ovld __cnfn atan(double8 y_over_x); - double16 __ovld __cnfn atan(double16 y_over_x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn atan(half y_over_x); - half2 __ovld __cnfn atan(half2 y_over_x); -@@ -6716,14 +6715,14 @@ float3 __ovld __cnfn atan2(float3 y, float3 x); - float4 __ovld __cnfn atan2(float4 y, float4 x); - float8 __ovld __cnfn atan2(float8 y, float8 x); - float16 __ovld __cnfn atan2(float16 y, float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn atan2(double y, double x); - double2 __ovld __cnfn atan2(double2 y, double2 x); - double3 __ovld __cnfn atan2(double3 y, double3 x); - double4 __ovld __cnfn atan2(double4 y, double4 x); - double8 __ovld __cnfn atan2(double8 y, double8 x); - double16 __ovld __cnfn atan2(double16 y, double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn atan2(half y, half x); - half2 __ovld __cnfn atan2(half2 y, half2 x); -@@ -6742,14 +6741,14 @@ float3 __ovld __cnfn atanh(float3); - float4 __ovld __cnfn atanh(float4); - float8 __ovld __cnfn atanh(float8); - float16 __ovld __cnfn atanh(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn atanh(double); - double2 __ovld __cnfn atanh(double2); - double3 __ovld __cnfn atanh(double3); - double4 __ovld __cnfn atanh(double4); - double8 __ovld __cnfn atanh(double8); - double16 __ovld __cnfn atanh(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn atanh(half); - half2 __ovld __cnfn atanh(half2); -@@ -6768,14 +6767,14 @@ float3 __ovld __cnfn atanpi(float3 x); - float4 __ovld __cnfn atanpi(float4 x); - float8 __ovld __cnfn atanpi(float8 x); - float16 __ovld __cnfn atanpi(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn atanpi(double x); - double2 __ovld __cnfn atanpi(double2 x); - double3 __ovld __cnfn atanpi(double3 x); - double4 __ovld __cnfn atanpi(double4 x); - double8 __ovld __cnfn atanpi(double8 x); - double16 __ovld __cnfn atanpi(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn atanpi(half x); - half2 __ovld __cnfn atanpi(half2 x); -@@ -6794,14 +6793,14 @@ float3 __ovld __cnfn atan2pi(float3 y, float3 x); - float4 __ovld __cnfn atan2pi(float4 y, float4 x); - float8 __ovld __cnfn atan2pi(float8 y, float8 x); - float16 __ovld __cnfn atan2pi(float16 y, float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn atan2pi(double y, double x); - double2 __ovld __cnfn atan2pi(double2 y, double2 x); - double3 __ovld __cnfn atan2pi(double3 y, double3 x); - double4 __ovld __cnfn atan2pi(double4 y, double4 x); - double8 __ovld __cnfn atan2pi(double8 y, double8 x); - double16 __ovld __cnfn atan2pi(double16 y, double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn atan2pi(half y, half x); - half2 __ovld __cnfn atan2pi(half2 y, half2 x); -@@ -6820,14 +6819,14 @@ float3 __ovld __cnfn cbrt(float3); - float4 __ovld __cnfn cbrt(float4); - float8 __ovld __cnfn cbrt(float8); - float16 __ovld __cnfn cbrt(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn cbrt(double); - double2 __ovld __cnfn cbrt(double2); - double3 __ovld __cnfn cbrt(double3); - double4 __ovld __cnfn cbrt(double4); - double8 __ovld __cnfn cbrt(double8); - double16 __ovld __cnfn cbrt(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn cbrt(half); - half2 __ovld __cnfn cbrt(half2); -@@ -6847,14 +6846,14 @@ float3 __ovld __cnfn ceil(float3); - float4 __ovld __cnfn ceil(float4); - float8 __ovld __cnfn ceil(float8); - float16 __ovld __cnfn ceil(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn ceil(double); - double2 __ovld __cnfn ceil(double2); - double3 __ovld __cnfn ceil(double3); - double4 __ovld __cnfn ceil(double4); - double8 __ovld __cnfn ceil(double8); - double16 __ovld __cnfn ceil(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn ceil(half); - half2 __ovld __cnfn ceil(half2); -@@ -6873,14 +6872,14 @@ float3 __ovld __cnfn copysign(float3 x, float3 y); - float4 __ovld __cnfn copysign(float4 x, float4 y); - float8 __ovld __cnfn copysign(float8 x, float8 y); - float16 __ovld __cnfn copysign(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn copysign(double x, double y); - double2 __ovld __cnfn copysign(double2 x, double2 y); - double3 __ovld __cnfn copysign(double3 x, double3 y); - double4 __ovld __cnfn copysign(double4 x, double4 y); - double8 __ovld __cnfn copysign(double8 x, double8 y); - double16 __ovld __cnfn copysign(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn copysign(half x, half y); - half2 __ovld __cnfn copysign(half2 x, half2 y); -@@ -6899,14 +6898,14 @@ float3 __ovld __cnfn cos(float3); - float4 __ovld __cnfn cos(float4); - float8 __ovld __cnfn cos(float8); - float16 __ovld __cnfn cos(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn cos(double); - double2 __ovld __cnfn cos(double2); - double3 __ovld __cnfn cos(double3); - double4 __ovld __cnfn cos(double4); - double8 __ovld __cnfn cos(double8); - double16 __ovld __cnfn cos(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn cos(half); - half2 __ovld __cnfn cos(half2); -@@ -6925,14 +6924,14 @@ float3 __ovld __cnfn cosh(float3); - float4 __ovld __cnfn cosh(float4); - float8 __ovld __cnfn cosh(float8); - float16 __ovld __cnfn cosh(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn cosh(double); - double2 __ovld __cnfn cosh(double2); - double3 __ovld __cnfn cosh(double3); - double4 __ovld __cnfn cosh(double4); - double8 __ovld __cnfn cosh(double8); - double16 __ovld __cnfn cosh(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn cosh(half); - half2 __ovld __cnfn cosh(half2); -@@ -6951,14 +6950,14 @@ float3 __ovld __cnfn cospi(float3 x); - float4 __ovld __cnfn cospi(float4 x); - float8 __ovld __cnfn cospi(float8 x); - float16 __ovld __cnfn cospi(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn cospi(double x); - double2 __ovld __cnfn cospi(double2 x); - double3 __ovld __cnfn cospi(double3 x); - double4 __ovld __cnfn cospi(double4 x); - double8 __ovld __cnfn cospi(double8 x); - double16 __ovld __cnfn cospi(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn cospi(half x); - half2 __ovld __cnfn cospi(half2 x); -@@ -6977,14 +6976,14 @@ float3 __ovld __cnfn erfc(float3); - float4 __ovld __cnfn erfc(float4); - float8 __ovld __cnfn erfc(float8); - float16 __ovld __cnfn erfc(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn erfc(double); - double2 __ovld __cnfn erfc(double2); - double3 __ovld __cnfn erfc(double3); - double4 __ovld __cnfn erfc(double4); - double8 __ovld __cnfn erfc(double8); - double16 __ovld __cnfn erfc(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn erfc(half); - half2 __ovld __cnfn erfc(half2); -@@ -7004,14 +7003,14 @@ float3 __ovld __cnfn erf(float3); - float4 __ovld __cnfn erf(float4); - float8 __ovld __cnfn erf(float8); - float16 __ovld __cnfn erf(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn erf(double); - double2 __ovld __cnfn erf(double2); - double3 __ovld __cnfn erf(double3); - double4 __ovld __cnfn erf(double4); - double8 __ovld __cnfn erf(double8); - double16 __ovld __cnfn erf(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn erf(half); - half2 __ovld __cnfn erf(half2); -@@ -7030,14 +7029,14 @@ float3 __ovld __cnfn exp(float3 x); - float4 __ovld __cnfn exp(float4 x); - float8 __ovld __cnfn exp(float8 x); - float16 __ovld __cnfn exp(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn exp(double x); - double2 __ovld __cnfn exp(double2 x); - double3 __ovld __cnfn exp(double3 x); - double4 __ovld __cnfn exp(double4 x); - double8 __ovld __cnfn exp(double8 x); - double16 __ovld __cnfn exp(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn exp(half x); - half2 __ovld __cnfn exp(half2 x); -@@ -7056,14 +7055,14 @@ float3 __ovld __cnfn exp2(float3); - float4 __ovld __cnfn exp2(float4); - float8 __ovld __cnfn exp2(float8); - float16 __ovld __cnfn exp2(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn exp2(double); - double2 __ovld __cnfn exp2(double2); - double3 __ovld __cnfn exp2(double3); - double4 __ovld __cnfn exp2(double4); - double8 __ovld __cnfn exp2(double8); - double16 __ovld __cnfn exp2(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn exp2(half); - half2 __ovld __cnfn exp2(half2); -@@ -7082,14 +7081,14 @@ float3 __ovld __cnfn exp10(float3); - float4 __ovld __cnfn exp10(float4); - float8 __ovld __cnfn exp10(float8); - float16 __ovld __cnfn exp10(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn exp10(double); - double2 __ovld __cnfn exp10(double2); - double3 __ovld __cnfn exp10(double3); - double4 __ovld __cnfn exp10(double4); - double8 __ovld __cnfn exp10(double8); - double16 __ovld __cnfn exp10(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn exp10(half); - half2 __ovld __cnfn exp10(half2); -@@ -7108,14 +7107,14 @@ float3 __ovld __cnfn expm1(float3 x); - float4 __ovld __cnfn expm1(float4 x); - float8 __ovld __cnfn expm1(float8 x); - float16 __ovld __cnfn expm1(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn expm1(double x); - double2 __ovld __cnfn expm1(double2 x); - double3 __ovld __cnfn expm1(double3 x); - double4 __ovld __cnfn expm1(double4 x); - double8 __ovld __cnfn expm1(double8 x); - double16 __ovld __cnfn expm1(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn expm1(half x); - half2 __ovld __cnfn expm1(half2 x); -@@ -7134,14 +7133,14 @@ float3 __ovld __cnfn fabs(float3); - float4 __ovld __cnfn fabs(float4); - float8 __ovld __cnfn fabs(float8); - float16 __ovld __cnfn fabs(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn fabs(double); - double2 __ovld __cnfn fabs(double2); - double3 __ovld __cnfn fabs(double3); - double4 __ovld __cnfn fabs(double4); - double8 __ovld __cnfn fabs(double8); - double16 __ovld __cnfn fabs(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn fabs(half); - half2 __ovld __cnfn fabs(half2); -@@ -7160,14 +7159,14 @@ float3 __ovld __cnfn fdim(float3 x, float3 y); - float4 __ovld __cnfn fdim(float4 x, float4 y); - float8 __ovld __cnfn fdim(float8 x, float8 y); - float16 __ovld __cnfn fdim(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn fdim(double x, double y); - double2 __ovld __cnfn fdim(double2 x, double2 y); - double3 __ovld __cnfn fdim(double3 x, double3 y); - double4 __ovld __cnfn fdim(double4 x, double4 y); - double8 __ovld __cnfn fdim(double8 x, double8 y); - double16 __ovld __cnfn fdim(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn fdim(half x, half y); - half2 __ovld __cnfn fdim(half2 x, half2 y); -@@ -7187,14 +7186,14 @@ float3 __ovld __cnfn floor(float3); - float4 __ovld __cnfn floor(float4); - float8 __ovld __cnfn floor(float8); - float16 __ovld __cnfn floor(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn floor(double); - double2 __ovld __cnfn floor(double2); - double3 __ovld __cnfn floor(double3); - double4 __ovld __cnfn floor(double4); - double8 __ovld __cnfn floor(double8); - double16 __ovld __cnfn floor(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn floor(half); - half2 __ovld __cnfn floor(half2); -@@ -7217,14 +7216,14 @@ float3 __ovld __cnfn fma(float3 a, float3 b, float3 c); - float4 __ovld __cnfn fma(float4 a, float4 b, float4 c); - float8 __ovld __cnfn fma(float8 a, float8 b, float8 c); - float16 __ovld __cnfn fma(float16 a, float16 b, float16 c); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn fma(double a, double b, double c); - double2 __ovld __cnfn fma(double2 a, double2 b, double2 c); - double3 __ovld __cnfn fma(double3 a, double3 b, double3 c); - double4 __ovld __cnfn fma(double4 a, double4 b, double4 c); - double8 __ovld __cnfn fma(double8 a, double8 b, double8 c); - double16 __ovld __cnfn fma(double16 a, double16 b, double16 c); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn fma(half a, half b, half c); - half2 __ovld __cnfn fma(half2 a, half2 b, half2 c); -@@ -7251,7 +7250,7 @@ float3 __ovld __cnfn fmax(float3 x, float y); - float4 __ovld __cnfn fmax(float4 x, float y); - float8 __ovld __cnfn fmax(float8 x, float y); - float16 __ovld __cnfn fmax(float16 x, float y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn fmax(double x, double y); - double2 __ovld __cnfn fmax(double2 x, double2 y); - double3 __ovld __cnfn fmax(double3 x, double3 y); -@@ -7263,7 +7262,7 @@ double3 __ovld __cnfn fmax(double3 x, double y); - double4 __ovld __cnfn fmax(double4 x, double y); - double8 __ovld __cnfn fmax(double8 x, double y); - double16 __ovld __cnfn fmax(double16 x, double y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn fmax(half x, half y); - half2 __ovld __cnfn fmax(half2 x, half2 y); -@@ -7295,7 +7294,7 @@ float3 __ovld __cnfn fmin(float3 x, float y); - float4 __ovld __cnfn fmin(float4 x, float y); - float8 __ovld __cnfn fmin(float8 x, float y); - float16 __ovld __cnfn fmin(float16 x, float y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn fmin(double x, double y); - double2 __ovld __cnfn fmin(double2 x, double2 y); - double3 __ovld __cnfn fmin(double3 x, double3 y); -@@ -7307,7 +7306,7 @@ double3 __ovld __cnfn fmin(double3 x, double y); - double4 __ovld __cnfn fmin(double4 x, double y); - double8 __ovld __cnfn fmin(double8 x, double y); - double16 __ovld __cnfn fmin(double16 x, double y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn fmin(half x, half y); - half2 __ovld __cnfn fmin(half2 x, half2 y); -@@ -7331,14 +7330,14 @@ float3 __ovld __cnfn fmod(float3 x, float3 y); - float4 __ovld __cnfn fmod(float4 x, float4 y); - float8 __ovld __cnfn fmod(float8 x, float8 y); - float16 __ovld __cnfn fmod(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn fmod(double x, double y); - double2 __ovld __cnfn fmod(double2 x, double2 y); - double3 __ovld __cnfn fmod(double3 x, double3 y); - double4 __ovld __cnfn fmod(double4 x, double4 y); - double8 __ovld __cnfn fmod(double8 x, double8 y); - double16 __ovld __cnfn fmod(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn fmod(half x, half y); - half2 __ovld __cnfn fmod(half2 x, half2 y); -@@ -7352,21 +7351,21 @@ half16 __ovld __cnfn fmod(half16 x, half16 y); - * Returns fmin(x - floor (x), 0x1.fffffep-1f ). - * floor(x) is returned in iptr. - */ --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - float __ovld fract(float x, float *iptr); - float2 __ovld fract(float2 x, float2 *iptr); - float3 __ovld fract(float3 x, float3 *iptr); - float4 __ovld fract(float4 x, float4 *iptr); - float8 __ovld fract(float8 x, float8 *iptr); - float16 __ovld fract(float16 x, float16 *iptr); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld fract(double x, double *iptr); - double2 __ovld fract(double2 x, double2 *iptr); - double3 __ovld fract(double3 x, double3 *iptr); - double4 __ovld fract(double4 x, double4 *iptr); - double8 __ovld fract(double8 x, double8 *iptr); - double16 __ovld fract(double16 x, double16 *iptr); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld fract(half x, half *iptr); - half2 __ovld fract(half2 x, half2 *iptr); -@@ -7375,7 +7374,9 @@ half4 __ovld fract(half4 x, half4 *iptr); - half8 __ovld fract(half8 x, half8 *iptr); - half16 __ovld fract(half16 x, half16 *iptr); - #endif //cl_khr_fp16 --#else -+#endif //__opencl_c_generic_address_space -+ -+#if !defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != CL_VERSION_2_0) - float __ovld fract(float x, __global float *iptr); - float2 __ovld fract(float2 x, __global float2 *iptr); - float3 __ovld fract(float3 x, __global float3 *iptr); -@@ -7394,7 +7395,7 @@ float3 __ovld fract(float3 x, __private float3 *iptr); - float4 __ovld fract(float4 x, __private float4 *iptr); - float8 __ovld fract(float8 x, __private float8 *iptr); - float16 __ovld fract(float16 x, __private float16 *iptr); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld fract(double x, __global double *iptr); - double2 __ovld fract(double2 x, __global double2 *iptr); - double3 __ovld fract(double3 x, __global double3 *iptr); -@@ -7413,7 +7414,7 @@ double3 __ovld fract(double3 x, __private double3 *iptr); - double4 __ovld fract(double4 x, __private double4 *iptr); - double8 __ovld fract(double8 x, __private double8 *iptr); - double16 __ovld fract(double16 x, __private double16 *iptr); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld fract(half x, __global half *iptr); - half2 __ovld fract(half2 x, __global half2 *iptr); -@@ -7434,29 +7435,29 @@ half4 __ovld fract(half4 x, __private half4 *iptr); - half8 __ovld fract(half8 x, __private half8 *iptr); - half16 __ovld fract(half16 x, __private half16 *iptr); - #endif //cl_khr_fp16 --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -- -+#endif //! defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != -+ //! CL_VERSION_2_0) - /** - * Extract mantissa and exponent from x. For each - * component the mantissa returned is a float with - * magnitude in the interval [1/2, 1) or 0. Each - * component of x equals mantissa returned * 2^exp. - */ --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - float __ovld frexp(float x, int *exp); - float2 __ovld frexp(float2 x, int2 *exp); - float3 __ovld frexp(float3 x, int3 *exp); - float4 __ovld frexp(float4 x, int4 *exp); - float8 __ovld frexp(float8 x, int8 *exp); - float16 __ovld frexp(float16 x, int16 *exp); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld frexp(double x, int *exp); - double2 __ovld frexp(double2 x, int2 *exp); - double3 __ovld frexp(double3 x, int3 *exp); - double4 __ovld frexp(double4 x, int4 *exp); - double8 __ovld frexp(double8 x, int8 *exp); - double16 __ovld frexp(double16 x, int16 *exp); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld frexp(half x, int *exp); - half2 __ovld frexp(half2 x, int2 *exp); -@@ -7465,7 +7466,9 @@ half4 __ovld frexp(half4 x, int4 *exp); - half8 __ovld frexp(half8 x, int8 *exp); - half16 __ovld frexp(half16 x, int16 *exp); - #endif //cl_khr_fp16 --#else -+#endif //__opencl_c_generic_address_space -+ -+#if !defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != CL_VERSION_2_0) - float __ovld frexp(float x, __global int *exp); - float2 __ovld frexp(float2 x, __global int2 *exp); - float3 __ovld frexp(float3 x, __global int3 *exp); -@@ -7484,7 +7487,7 @@ float3 __ovld frexp(float3 x, __private int3 *exp); - float4 __ovld frexp(float4 x, __private int4 *exp); - float8 __ovld frexp(float8 x, __private int8 *exp); - float16 __ovld frexp(float16 x, __private int16 *exp); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld frexp(double x, __global int *exp); - double2 __ovld frexp(double2 x, __global int2 *exp); - double3 __ovld frexp(double3 x, __global int3 *exp); -@@ -7503,7 +7506,7 @@ double3 __ovld frexp(double3 x, __private int3 *exp); - double4 __ovld frexp(double4 x, __private int4 *exp); - double8 __ovld frexp(double8 x, __private int8 *exp); - double16 __ovld frexp(double16 x, __private int16 *exp); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld frexp(half x, __global int *exp); - half2 __ovld frexp(half2 x, __global int2 *exp); -@@ -7524,7 +7527,8 @@ half4 __ovld frexp(half4 x, __private int4 *exp); - half8 __ovld frexp(half8 x, __private int8 *exp); - half16 __ovld frexp(half16 x, __private int16 *exp); - #endif //cl_khr_fp16 --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#endif //! defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != -+ //! CL_VERSION_2_0) - - /** - * Compute the value of the square root of x^2 + y^2 -@@ -7536,14 +7540,14 @@ float3 __ovld __cnfn hypot(float3 x, float3 y); - float4 __ovld __cnfn hypot(float4 x, float4 y); - float8 __ovld __cnfn hypot(float8 x, float8 y); - float16 __ovld __cnfn hypot(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn hypot(double x, double y); - double2 __ovld __cnfn hypot(double2 x, double2 y); - double3 __ovld __cnfn hypot(double3 x, double3 y); - double4 __ovld __cnfn hypot(double4 x, double4 y); - double8 __ovld __cnfn hypot(double8 x, double8 y); - double16 __ovld __cnfn hypot(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn hypot(half x, half y); - half2 __ovld __cnfn hypot(half2 x, half2 y); -@@ -7562,14 +7566,14 @@ int3 __ovld __cnfn ilogb(float3 x); - int4 __ovld __cnfn ilogb(float4 x); - int8 __ovld __cnfn ilogb(float8 x); - int16 __ovld __cnfn ilogb(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn ilogb(double x); - int2 __ovld __cnfn ilogb(double2 x); - int3 __ovld __cnfn ilogb(double3 x); - int4 __ovld __cnfn ilogb(double4 x); - int8 __ovld __cnfn ilogb(double8 x); - int16 __ovld __cnfn ilogb(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn ilogb(half x); - int2 __ovld __cnfn ilogb(half2 x); -@@ -7593,7 +7597,7 @@ float3 __ovld __cnfn ldexp(float3 x, int n); - float4 __ovld __cnfn ldexp(float4 x, int n); - float8 __ovld __cnfn ldexp(float8 x, int n); - float16 __ovld __cnfn ldexp(float16 x, int n); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn ldexp(double x, int n); - double2 __ovld __cnfn ldexp(double2 x, int2 n); - double3 __ovld __cnfn ldexp(double3 x, int3 n); -@@ -7605,7 +7609,7 @@ double3 __ovld __cnfn ldexp(double3 x, int n); - double4 __ovld __cnfn ldexp(double4 x, int n); - double8 __ovld __cnfn ldexp(double8 x, int n); - double16 __ovld __cnfn ldexp(double16 x, int n); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn ldexp(half x, int n); - half2 __ovld __cnfn ldexp(half2 x, int2 n); -@@ -7632,14 +7636,14 @@ float3 __ovld __cnfn lgamma(float3 x); - float4 __ovld __cnfn lgamma(float4 x); - float8 __ovld __cnfn lgamma(float8 x); - float16 __ovld __cnfn lgamma(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn lgamma(double x); - double2 __ovld __cnfn lgamma(double2 x); - double3 __ovld __cnfn lgamma(double3 x); - double4 __ovld __cnfn lgamma(double4 x); - double8 __ovld __cnfn lgamma(double8 x); - double16 __ovld __cnfn lgamma(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn lgamma(half x); - half2 __ovld __cnfn lgamma(half2 x); -@@ -7649,21 +7653,21 @@ half8 __ovld __cnfn lgamma(half8 x); - half16 __ovld __cnfn lgamma(half16 x); - #endif //cl_khr_fp16 - --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - float __ovld lgamma_r(float x, int *signp); - float2 __ovld lgamma_r(float2 x, int2 *signp); - float3 __ovld lgamma_r(float3 x, int3 *signp); - float4 __ovld lgamma_r(float4 x, int4 *signp); - float8 __ovld lgamma_r(float8 x, int8 *signp); - float16 __ovld lgamma_r(float16 x, int16 *signp); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld lgamma_r(double x, int *signp); - double2 __ovld lgamma_r(double2 x, int2 *signp); - double3 __ovld lgamma_r(double3 x, int3 *signp); - double4 __ovld lgamma_r(double4 x, int4 *signp); - double8 __ovld lgamma_r(double8 x, int8 *signp); - double16 __ovld lgamma_r(double16 x, int16 *signp); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld lgamma_r(half x, int *signp); - half2 __ovld lgamma_r(half2 x, int2 *signp); -@@ -7672,7 +7676,9 @@ half4 __ovld lgamma_r(half4 x, int4 *signp); - half8 __ovld lgamma_r(half8 x, int8 *signp); - half16 __ovld lgamma_r(half16 x, int16 *signp); - #endif //cl_khr_fp16 --#else -+#endif //__opencl_c_generic_address_space -+ -+#if !defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != CL_VERSION_2_0) - float __ovld lgamma_r(float x, __global int *signp); - float2 __ovld lgamma_r(float2 x, __global int2 *signp); - float3 __ovld lgamma_r(float3 x, __global int3 *signp); -@@ -7691,7 +7697,7 @@ float3 __ovld lgamma_r(float3 x, __private int3 *signp); - float4 __ovld lgamma_r(float4 x, __private int4 *signp); - float8 __ovld lgamma_r(float8 x, __private int8 *signp); - float16 __ovld lgamma_r(float16 x, __private int16 *signp); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld lgamma_r(double x, __global int *signp); - double2 __ovld lgamma_r(double2 x, __global int2 *signp); - double3 __ovld lgamma_r(double3 x, __global int3 *signp); -@@ -7710,7 +7716,7 @@ double3 __ovld lgamma_r(double3 x, __private int3 *signp); - double4 __ovld lgamma_r(double4 x, __private int4 *signp); - double8 __ovld lgamma_r(double8 x, __private int8 *signp); - double16 __ovld lgamma_r(double16 x, __private int16 *signp); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld lgamma_r(half x, __global int *signp); - half2 __ovld lgamma_r(half2 x, __global int2 *signp); -@@ -7731,8 +7737,8 @@ half4 __ovld lgamma_r(half4 x, __private int4 *signp); - half8 __ovld lgamma_r(half8 x, __private int8 *signp); - half16 __ovld lgamma_r(half16 x, __private int16 *signp); - #endif //cl_khr_fp16 --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -- -+#endif //! defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != -+ //! CL_VERSION_2_0) - /** - * Compute natural logarithm. - */ -@@ -7742,14 +7748,14 @@ float3 __ovld __cnfn log(float3); - float4 __ovld __cnfn log(float4); - float8 __ovld __cnfn log(float8); - float16 __ovld __cnfn log(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn log(double); - double2 __ovld __cnfn log(double2); - double3 __ovld __cnfn log(double3); - double4 __ovld __cnfn log(double4); - double8 __ovld __cnfn log(double8); - double16 __ovld __cnfn log(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn log(half); - half2 __ovld __cnfn log(half2); -@@ -7760,7 +7766,7 @@ half16 __ovld __cnfn log(half16); - #endif //cl_khr_fp16 - - /** -- * Compute a base 2 logarithm. -+ * Compute a base 2 logarithm - */ - float __ovld __cnfn log2(float); - float2 __ovld __cnfn log2(float2); -@@ -7768,14 +7774,14 @@ float3 __ovld __cnfn log2(float3); - float4 __ovld __cnfn log2(float4); - float8 __ovld __cnfn log2(float8); - float16 __ovld __cnfn log2(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn log2(double); - double2 __ovld __cnfn log2(double2); - double3 __ovld __cnfn log2(double3); - double4 __ovld __cnfn log2(double4); - double8 __ovld __cnfn log2(double8); - double16 __ovld __cnfn log2(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn log2(half); - half2 __ovld __cnfn log2(half2); -@@ -7794,14 +7800,14 @@ float3 __ovld __cnfn log10(float3); - float4 __ovld __cnfn log10(float4); - float8 __ovld __cnfn log10(float8); - float16 __ovld __cnfn log10(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn log10(double); - double2 __ovld __cnfn log10(double2); - double3 __ovld __cnfn log10(double3); - double4 __ovld __cnfn log10(double4); - double8 __ovld __cnfn log10(double8); - double16 __ovld __cnfn log10(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn log10(half); - half2 __ovld __cnfn log10(half2); -@@ -7820,14 +7826,14 @@ float3 __ovld __cnfn log1p(float3 x); - float4 __ovld __cnfn log1p(float4 x); - float8 __ovld __cnfn log1p(float8 x); - float16 __ovld __cnfn log1p(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn log1p(double x); - double2 __ovld __cnfn log1p(double2 x); - double3 __ovld __cnfn log1p(double3 x); - double4 __ovld __cnfn log1p(double4 x); - double8 __ovld __cnfn log1p(double8 x); - double16 __ovld __cnfn log1p(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn log1p(half x); - half2 __ovld __cnfn log1p(half2 x); -@@ -7847,14 +7853,14 @@ float3 __ovld __cnfn logb(float3 x); - float4 __ovld __cnfn logb(float4 x); - float8 __ovld __cnfn logb(float8 x); - float16 __ovld __cnfn logb(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn logb(double x); - double2 __ovld __cnfn logb(double2 x); - double3 __ovld __cnfn logb(double3 x); - double4 __ovld __cnfn logb(double4 x); - double8 __ovld __cnfn logb(double8 x); - double16 __ovld __cnfn logb(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn logb(half x); - half2 __ovld __cnfn logb(half2 x); -@@ -7877,14 +7883,14 @@ float3 __ovld __cnfn mad(float3 a, float3 b, float3 c); - float4 __ovld __cnfn mad(float4 a, float4 b, float4 c); - float8 __ovld __cnfn mad(float8 a, float8 b, float8 c); - float16 __ovld __cnfn mad(float16 a, float16 b, float16 c); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn mad(double a, double b, double c); - double2 __ovld __cnfn mad(double2 a, double2 b, double2 c); - double3 __ovld __cnfn mad(double3 a, double3 b, double3 c); - double4 __ovld __cnfn mad(double4 a, double4 b, double4 c); - double8 __ovld __cnfn mad(double8 a, double8 b, double8 c); - double16 __ovld __cnfn mad(double16 a, double16 b, double16 c); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn mad(half a, half b, half c); - half2 __ovld __cnfn mad(half2 a, half2 b, half2 c); -@@ -7904,14 +7910,14 @@ float3 __ovld __cnfn maxmag(float3 x, float3 y); - float4 __ovld __cnfn maxmag(float4 x, float4 y); - float8 __ovld __cnfn maxmag(float8 x, float8 y); - float16 __ovld __cnfn maxmag(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn maxmag(double x, double y); - double2 __ovld __cnfn maxmag(double2 x, double2 y); - double3 __ovld __cnfn maxmag(double3 x, double3 y); - double4 __ovld __cnfn maxmag(double4 x, double4 y); - double8 __ovld __cnfn maxmag(double8 x, double8 y); - double16 __ovld __cnfn maxmag(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn maxmag(half x, half y); - half2 __ovld __cnfn maxmag(half2 x, half2 y); -@@ -7931,14 +7937,14 @@ float3 __ovld __cnfn minmag(float3 x, float3 y); - float4 __ovld __cnfn minmag(float4 x, float4 y); - float8 __ovld __cnfn minmag(float8 x, float8 y); - float16 __ovld __cnfn minmag(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn minmag(double x, double y); - double2 __ovld __cnfn minmag(double2 x, double2 y); - double3 __ovld __cnfn minmag(double3 x, double3 y); - double4 __ovld __cnfn minmag(double4 x, double4 y); - double8 __ovld __cnfn minmag(double8 x, double8 y); - double16 __ovld __cnfn minmag(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn minmag(half x, half y); - half2 __ovld __cnfn minmag(half2 x, half2 y); -@@ -7955,21 +7961,21 @@ half16 __ovld __cnfn minmag(half16 x, half16 y); - * the argument. It stores the integral part in the object - * pointed to by iptr. - */ --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - float __ovld modf(float x, float *iptr); - float2 __ovld modf(float2 x, float2 *iptr); - float3 __ovld modf(float3 x, float3 *iptr); - float4 __ovld modf(float4 x, float4 *iptr); - float8 __ovld modf(float8 x, float8 *iptr); - float16 __ovld modf(float16 x, float16 *iptr); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld modf(double x, double *iptr); - double2 __ovld modf(double2 x, double2 *iptr); - double3 __ovld modf(double3 x, double3 *iptr); - double4 __ovld modf(double4 x, double4 *iptr); - double8 __ovld modf(double8 x, double8 *iptr); - double16 __ovld modf(double16 x, double16 *iptr); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld modf(half x, half *iptr); - half2 __ovld modf(half2 x, half2 *iptr); -@@ -7978,7 +7984,9 @@ half4 __ovld modf(half4 x, half4 *iptr); - half8 __ovld modf(half8 x, half8 *iptr); - half16 __ovld modf(half16 x, half16 *iptr); - #endif //cl_khr_fp16 --#else -+#endif //__opencl_c_generic_address_space -+ -+#if !defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != CL_VERSION_2_0) - float __ovld modf(float x, __global float *iptr); - float2 __ovld modf(float2 x, __global float2 *iptr); - float3 __ovld modf(float3 x, __global float3 *iptr); -@@ -7997,7 +8005,7 @@ float3 __ovld modf(float3 x, __private float3 *iptr); - float4 __ovld modf(float4 x, __private float4 *iptr); - float8 __ovld modf(float8 x, __private float8 *iptr); - float16 __ovld modf(float16 x, __private float16 *iptr); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld modf(double x, __global double *iptr); - double2 __ovld modf(double2 x, __global double2 *iptr); - double3 __ovld modf(double3 x, __global double3 *iptr); -@@ -8016,7 +8024,7 @@ double3 __ovld modf(double3 x, __private double3 *iptr); - double4 __ovld modf(double4 x, __private double4 *iptr); - double8 __ovld modf(double8 x, __private double8 *iptr); - double16 __ovld modf(double16 x, __private double16 *iptr); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld modf(half x, __global half *iptr); - half2 __ovld modf(half2 x, __global half2 *iptr); -@@ -8037,7 +8045,8 @@ half4 __ovld modf(half4 x, __private half4 *iptr); - half8 __ovld modf(half8 x, __private half8 *iptr); - half16 __ovld modf(half16 x, __private half16 *iptr); - #endif //cl_khr_fp16 --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#endif //! defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != -+ //! CL_VERSION_2_0) - - /** - * Returns a quiet NaN. The nancode may be placed -@@ -8049,14 +8058,14 @@ float3 __ovld __cnfn nan(uint3 nancode); - float4 __ovld __cnfn nan(uint4 nancode); - float8 __ovld __cnfn nan(uint8 nancode); - float16 __ovld __cnfn nan(uint16 nancode); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn nan(ulong nancode); - double2 __ovld __cnfn nan(ulong2 nancode); - double3 __ovld __cnfn nan(ulong3 nancode); - double4 __ovld __cnfn nan(ulong4 nancode); - double8 __ovld __cnfn nan(ulong8 nancode); - double16 __ovld __cnfn nan(ulong16 nancode); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn nan(ushort nancode); - half2 __ovld __cnfn nan(ushort2 nancode); -@@ -8079,14 +8088,14 @@ float3 __ovld __cnfn nextafter(float3 x, float3 y); - float4 __ovld __cnfn nextafter(float4 x, float4 y); - float8 __ovld __cnfn nextafter(float8 x, float8 y); - float16 __ovld __cnfn nextafter(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn nextafter(double x, double y); - double2 __ovld __cnfn nextafter(double2 x, double2 y); - double3 __ovld __cnfn nextafter(double3 x, double3 y); - double4 __ovld __cnfn nextafter(double4 x, double4 y); - double8 __ovld __cnfn nextafter(double8 x, double8 y); - double16 __ovld __cnfn nextafter(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn nextafter(half x, half y); - half2 __ovld __cnfn nextafter(half2 x, half2 y); -@@ -8105,14 +8114,14 @@ float3 __ovld __cnfn pow(float3 x, float3 y); - float4 __ovld __cnfn pow(float4 x, float4 y); - float8 __ovld __cnfn pow(float8 x, float8 y); - float16 __ovld __cnfn pow(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn pow(double x, double y); - double2 __ovld __cnfn pow(double2 x, double2 y); - double3 __ovld __cnfn pow(double3 x, double3 y); - double4 __ovld __cnfn pow(double4 x, double4 y); - double8 __ovld __cnfn pow(double8 x, double8 y); - double16 __ovld __cnfn pow(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn pow(half x, half y); - half2 __ovld __cnfn pow(half2 x, half2 y); -@@ -8131,14 +8140,14 @@ float3 __ovld __cnfn pown(float3 x, int3 y); - float4 __ovld __cnfn pown(float4 x, int4 y); - float8 __ovld __cnfn pown(float8 x, int8 y); - float16 __ovld __cnfn pown(float16 x, int16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn pown(double x, int y); - double2 __ovld __cnfn pown(double2 x, int2 y); - double3 __ovld __cnfn pown(double3 x, int3 y); - double4 __ovld __cnfn pown(double4 x, int4 y); - double8 __ovld __cnfn pown(double8 x, int8 y); - double16 __ovld __cnfn pown(double16 x, int16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn pown(half x, int y); - half2 __ovld __cnfn pown(half2 x, int2 y); -@@ -8157,14 +8166,14 @@ float3 __ovld __cnfn powr(float3 x, float3 y); - float4 __ovld __cnfn powr(float4 x, float4 y); - float8 __ovld __cnfn powr(float8 x, float8 y); - float16 __ovld __cnfn powr(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn powr(double x, double y); - double2 __ovld __cnfn powr(double2 x, double2 y); - double3 __ovld __cnfn powr(double3 x, double3 y); - double4 __ovld __cnfn powr(double4 x, double4 y); - double8 __ovld __cnfn powr(double8 x, double8 y); - double16 __ovld __cnfn powr(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn powr(half x, half y); - half2 __ovld __cnfn powr(half2 x, half2 y); -@@ -8186,14 +8195,14 @@ float3 __ovld __cnfn remainder(float3 x, float3 y); - float4 __ovld __cnfn remainder(float4 x, float4 y); - float8 __ovld __cnfn remainder(float8 x, float8 y); - float16 __ovld __cnfn remainder(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn remainder(double x, double y); - double2 __ovld __cnfn remainder(double2 x, double2 y); - double3 __ovld __cnfn remainder(double3 x, double3 y); - double4 __ovld __cnfn remainder(double4 x, double4 y); - double8 __ovld __cnfn remainder(double8 x, double8 y); - double16 __ovld __cnfn remainder(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn remainder(half x, half y); - half2 __ovld __cnfn remainder(half2 x, half2 y); -@@ -8215,21 +8224,21 @@ half16 __ovld __cnfn remainder(half16 x, half16 y); - * sign as x/y. It stores this signed value in the object - * pointed to by quo. - */ --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - float __ovld remquo(float x, float y, int *quo); - float2 __ovld remquo(float2 x, float2 y, int2 *quo); - float3 __ovld remquo(float3 x, float3 y, int3 *quo); - float4 __ovld remquo(float4 x, float4 y, int4 *quo); - float8 __ovld remquo(float8 x, float8 y, int8 *quo); - float16 __ovld remquo(float16 x, float16 y, int16 *quo); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld remquo(double x, double y, int *quo); - double2 __ovld remquo(double2 x, double2 y, int2 *quo); - double3 __ovld remquo(double3 x, double3 y, int3 *quo); - double4 __ovld remquo(double4 x, double4 y, int4 *quo); - double8 __ovld remquo(double8 x, double8 y, int8 *quo); - double16 __ovld remquo(double16 x, double16 y, int16 *quo); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld remquo(half x, half y, int *quo); - half2 __ovld remquo(half2 x, half2 y, int2 *quo); -@@ -8237,9 +8246,10 @@ half3 __ovld remquo(half3 x, half3 y, int3 *quo); - half4 __ovld remquo(half4 x, half4 y, int4 *quo); - half8 __ovld remquo(half8 x, half8 y, int8 *quo); - half16 __ovld remquo(half16 x, half16 y, int16 *quo); -- - #endif //cl_khr_fp16 --#else -+#endif //__opencl_c_generic_address_space -+ -+#if !defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != CL_VERSION_2_0) - float __ovld remquo(float x, float y, __global int *quo); - float2 __ovld remquo(float2 x, float2 y, __global int2 *quo); - float3 __ovld remquo(float3 x, float3 y, __global int3 *quo); -@@ -8258,7 +8268,7 @@ float3 __ovld remquo(float3 x, float3 y, __private int3 *quo); - float4 __ovld remquo(float4 x, float4 y, __private int4 *quo); - float8 __ovld remquo(float8 x, float8 y, __private int8 *quo); - float16 __ovld remquo(float16 x, float16 y, __private int16 *quo); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld remquo(double x, double y, __global int *quo); - double2 __ovld remquo(double2 x, double2 y, __global int2 *quo); - double3 __ovld remquo(double3 x, double3 y, __global int3 *quo); -@@ -8277,7 +8287,7 @@ double3 __ovld remquo(double3 x, double3 y, __private int3 *quo); - double4 __ovld remquo(double4 x, double4 y, __private int4 *quo); - double8 __ovld remquo(double8 x, double8 y, __private int8 *quo); - double16 __ovld remquo(double16 x, double16 y, __private int16 *quo); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld remquo(half x, half y, __global int *quo); - half2 __ovld remquo(half2 x, half2 y, __global int2 *quo); -@@ -8298,7 +8308,8 @@ half4 __ovld remquo(half4 x, half4 y, __private int4 *quo); - half8 __ovld remquo(half8 x, half8 y, __private int8 *quo); - half16 __ovld remquo(half16 x, half16 y, __private int16 *quo); - #endif //cl_khr_fp16 --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#endif //! defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != -+ //! CL_VERSION_2_0) - /** - * Round to integral value (using round to nearest - * even rounding mode) in floating-point format. -@@ -8311,14 +8322,14 @@ float3 __ovld __cnfn rint(float3); - float4 __ovld __cnfn rint(float4); - float8 __ovld __cnfn rint(float8); - float16 __ovld __cnfn rint(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn rint(double); - double2 __ovld __cnfn rint(double2); - double3 __ovld __cnfn rint(double3); - double4 __ovld __cnfn rint(double4); - double8 __ovld __cnfn rint(double8); - double16 __ovld __cnfn rint(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn rint(half); - half2 __ovld __cnfn rint(half2); -@@ -8337,14 +8348,14 @@ float3 __ovld __cnfn rootn(float3 x, int3 y); - float4 __ovld __cnfn rootn(float4 x, int4 y); - float8 __ovld __cnfn rootn(float8 x, int8 y); - float16 __ovld __cnfn rootn(float16 x, int16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn rootn(double x, int y); - double2 __ovld __cnfn rootn(double2 x, int2 y); - double3 __ovld __cnfn rootn(double3 x, int3 y); - double4 __ovld __cnfn rootn(double4 x, int4 y); - double8 __ovld __cnfn rootn(double8 x, int8 y); - double16 __ovld __cnfn rootn(double16 x, int16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn rootn(half x, int y); - half2 __ovld __cnfn rootn(half2 x, int2 y); -@@ -8365,14 +8376,14 @@ float3 __ovld __cnfn round(float3 x); - float4 __ovld __cnfn round(float4 x); - float8 __ovld __cnfn round(float8 x); - float16 __ovld __cnfn round(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn round(double x); - double2 __ovld __cnfn round(double2 x); - double3 __ovld __cnfn round(double3 x); - double4 __ovld __cnfn round(double4 x); - double8 __ovld __cnfn round(double8 x); - double16 __ovld __cnfn round(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn round(half x); - half2 __ovld __cnfn round(half2 x); -@@ -8391,14 +8402,14 @@ float3 __ovld __cnfn rsqrt(float3); - float4 __ovld __cnfn rsqrt(float4); - float8 __ovld __cnfn rsqrt(float8); - float16 __ovld __cnfn rsqrt(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn rsqrt(double); - double2 __ovld __cnfn rsqrt(double2); - double3 __ovld __cnfn rsqrt(double3); - double4 __ovld __cnfn rsqrt(double4); - double8 __ovld __cnfn rsqrt(double8); - double16 __ovld __cnfn rsqrt(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn rsqrt(half); - half2 __ovld __cnfn rsqrt(half2); -@@ -8417,14 +8428,14 @@ float3 __ovld __cnfn sin(float3); - float4 __ovld __cnfn sin(float4); - float8 __ovld __cnfn sin(float8); - float16 __ovld __cnfn sin(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn sin(double); - double2 __ovld __cnfn sin(double2); - double3 __ovld __cnfn sin(double3); - double4 __ovld __cnfn sin(double4); - double8 __ovld __cnfn sin(double8); - double16 __ovld __cnfn sin(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn sin(half); - half2 __ovld __cnfn sin(half2); -@@ -8439,21 +8450,21 @@ half16 __ovld __cnfn sin(half16); - * is the return value and computed cosine is returned - * in cosval. - */ --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - float __ovld sincos(float x, float *cosval); - float2 __ovld sincos(float2 x, float2 *cosval); - float3 __ovld sincos(float3 x, float3 *cosval); - float4 __ovld sincos(float4 x, float4 *cosval); - float8 __ovld sincos(float8 x, float8 *cosval); - float16 __ovld sincos(float16 x, float16 *cosval); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld sincos(double x, double *cosval); - double2 __ovld sincos(double2 x, double2 *cosval); - double3 __ovld sincos(double3 x, double3 *cosval); - double4 __ovld sincos(double4 x, double4 *cosval); - double8 __ovld sincos(double8 x, double8 *cosval); - double16 __ovld sincos(double16 x, double16 *cosval); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld sincos(half x, half *cosval); - half2 __ovld sincos(half2 x, half2 *cosval); -@@ -8462,7 +8473,9 @@ half4 __ovld sincos(half4 x, half4 *cosval); - half8 __ovld sincos(half8 x, half8 *cosval); - half16 __ovld sincos(half16 x, half16 *cosval); - #endif //cl_khr_fp16 --#else -+#endif //__opencl_c_generic_address_space -+ -+#if !defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != CL_VERSION_2_0) - float __ovld sincos(float x, __global float *cosval); - float2 __ovld sincos(float2 x, __global float2 *cosval); - float3 __ovld sincos(float3 x, __global float3 *cosval); -@@ -8481,7 +8494,7 @@ float3 __ovld sincos(float3 x, __private float3 *cosval); - float4 __ovld sincos(float4 x, __private float4 *cosval); - float8 __ovld sincos(float8 x, __private float8 *cosval); - float16 __ovld sincos(float16 x, __private float16 *cosval); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld sincos(double x, __global double *cosval); - double2 __ovld sincos(double2 x, __global double2 *cosval); - double3 __ovld sincos(double3 x, __global double3 *cosval); -@@ -8500,7 +8513,7 @@ double3 __ovld sincos(double3 x, __private double3 *cosval); - double4 __ovld sincos(double4 x, __private double4 *cosval); - double8 __ovld sincos(double8 x, __private double8 *cosval); - double16 __ovld sincos(double16 x, __private double16 *cosval); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld sincos(half x, __global half *cosval); - half2 __ovld sincos(half2 x, __global half2 *cosval); -@@ -8521,8 +8534,8 @@ half4 __ovld sincos(half4 x, __private half4 *cosval); - half8 __ovld sincos(half8 x, __private half8 *cosval); - half16 __ovld sincos(half16 x, __private half16 *cosval); - #endif //cl_khr_fp16 --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -- -+#endif //! defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != -+ //! CL_VERSION_2_0) - /** - * Compute hyperbolic sine. - */ -@@ -8532,14 +8545,14 @@ float3 __ovld __cnfn sinh(float3); - float4 __ovld __cnfn sinh(float4); - float8 __ovld __cnfn sinh(float8); - float16 __ovld __cnfn sinh(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn sinh(double); - double2 __ovld __cnfn sinh(double2); - double3 __ovld __cnfn sinh(double3); - double4 __ovld __cnfn sinh(double4); - double8 __ovld __cnfn sinh(double8); - double16 __ovld __cnfn sinh(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn sinh(half); - half2 __ovld __cnfn sinh(half2); -@@ -8558,14 +8571,14 @@ float3 __ovld __cnfn sinpi(float3 x); - float4 __ovld __cnfn sinpi(float4 x); - float8 __ovld __cnfn sinpi(float8 x); - float16 __ovld __cnfn sinpi(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn sinpi(double x); - double2 __ovld __cnfn sinpi(double2 x); - double3 __ovld __cnfn sinpi(double3 x); - double4 __ovld __cnfn sinpi(double4 x); - double8 __ovld __cnfn sinpi(double8 x); - double16 __ovld __cnfn sinpi(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn sinpi(half x); - half2 __ovld __cnfn sinpi(half2 x); -@@ -8584,14 +8597,14 @@ float3 __ovld __cnfn sqrt(float3); - float4 __ovld __cnfn sqrt(float4); - float8 __ovld __cnfn sqrt(float8); - float16 __ovld __cnfn sqrt(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn sqrt(double); - double2 __ovld __cnfn sqrt(double2); - double3 __ovld __cnfn sqrt(double3); - double4 __ovld __cnfn sqrt(double4); - double8 __ovld __cnfn sqrt(double8); - double16 __ovld __cnfn sqrt(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn sqrt(half); - half2 __ovld __cnfn sqrt(half2); -@@ -8610,14 +8623,14 @@ float3 __ovld __cnfn tan(float3); - float4 __ovld __cnfn tan(float4); - float8 __ovld __cnfn tan(float8); - float16 __ovld __cnfn tan(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn tan(double); - double2 __ovld __cnfn tan(double2); - double3 __ovld __cnfn tan(double3); - double4 __ovld __cnfn tan(double4); - double8 __ovld __cnfn tan(double8); - double16 __ovld __cnfn tan(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn tan(half); - half2 __ovld __cnfn tan(half2); -@@ -8636,14 +8649,14 @@ float3 __ovld __cnfn tanh(float3); - float4 __ovld __cnfn tanh(float4); - float8 __ovld __cnfn tanh(float8); - float16 __ovld __cnfn tanh(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn tanh(double); - double2 __ovld __cnfn tanh(double2); - double3 __ovld __cnfn tanh(double3); - double4 __ovld __cnfn tanh(double4); - double8 __ovld __cnfn tanh(double8); - double16 __ovld __cnfn tanh(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn tanh(half); - half2 __ovld __cnfn tanh(half2); -@@ -8662,14 +8675,14 @@ float3 __ovld __cnfn tanpi(float3 x); - float4 __ovld __cnfn tanpi(float4 x); - float8 __ovld __cnfn tanpi(float8 x); - float16 __ovld __cnfn tanpi(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn tanpi(double x); - double2 __ovld __cnfn tanpi(double2 x); - double3 __ovld __cnfn tanpi(double3 x); - double4 __ovld __cnfn tanpi(double4 x); - double8 __ovld __cnfn tanpi(double8 x); - double16 __ovld __cnfn tanpi(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn tanpi(half x); - half2 __ovld __cnfn tanpi(half2 x); -@@ -8688,14 +8701,14 @@ float3 __ovld __cnfn tgamma(float3); - float4 __ovld __cnfn tgamma(float4); - float8 __ovld __cnfn tgamma(float8); - float16 __ovld __cnfn tgamma(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn tgamma(double); - double2 __ovld __cnfn tgamma(double2); - double3 __ovld __cnfn tgamma(double3); - double4 __ovld __cnfn tgamma(double4); - double8 __ovld __cnfn tgamma(double8); - double16 __ovld __cnfn tgamma(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn tgamma(half); - half2 __ovld __cnfn tgamma(half2); -@@ -8715,14 +8728,14 @@ float3 __ovld __cnfn trunc(float3); - float4 __ovld __cnfn trunc(float4); - float8 __ovld __cnfn trunc(float8); - float16 __ovld __cnfn trunc(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn trunc(double); - double2 __ovld __cnfn trunc(double2); - double3 __ovld __cnfn trunc(double3); - double4 __ovld __cnfn trunc(double4); - double8 __ovld __cnfn trunc(double8); - double16 __ovld __cnfn trunc(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn trunc(half); - half2 __ovld __cnfn trunc(half2); -@@ -10108,7 +10121,7 @@ float3 __ovld __cnfn clamp(float3 x, float minval, float maxval); - float4 __ovld __cnfn clamp(float4 x, float minval, float maxval); - float8 __ovld __cnfn clamp(float8 x, float minval, float maxval); - float16 __ovld __cnfn clamp(float16 x, float minval, float maxval); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn clamp(double x, double minval, double maxval); - double2 __ovld __cnfn clamp(double2 x, double2 minval, double2 maxval); - double3 __ovld __cnfn clamp(double3 x, double3 minval, double3 maxval); -@@ -10120,7 +10133,7 @@ double3 __ovld __cnfn clamp(double3 x, double minval, double maxval); - double4 __ovld __cnfn clamp(double4 x, double minval, double maxval); - double8 __ovld __cnfn clamp(double8 x, double minval, double maxval); - double16 __ovld __cnfn clamp(double16 x, double minval, double maxval); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn clamp(half x, half minval, half maxval); - half2 __ovld __cnfn clamp(half2 x, half2 minval, half2 maxval); -@@ -10145,14 +10158,14 @@ float3 __ovld __cnfn degrees(float3 radians); - float4 __ovld __cnfn degrees(float4 radians); - float8 __ovld __cnfn degrees(float8 radians); - float16 __ovld __cnfn degrees(float16 radians); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn degrees(double radians); - double2 __ovld __cnfn degrees(double2 radians); - double3 __ovld __cnfn degrees(double3 radians); - double4 __ovld __cnfn degrees(double4 radians); - double8 __ovld __cnfn degrees(double8 radians); - double16 __ovld __cnfn degrees(double16 radians); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn degrees(half radians); - half2 __ovld __cnfn degrees(half2 radians); -@@ -10177,7 +10190,7 @@ float3 __ovld __cnfn max(float3 x, float y); - float4 __ovld __cnfn max(float4 x, float y); - float8 __ovld __cnfn max(float8 x, float y); - float16 __ovld __cnfn max(float16 x, float y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn max(double x, double y); - double2 __ovld __cnfn max(double2 x, double2 y); - double3 __ovld __cnfn max(double3 x, double3 y); -@@ -10189,7 +10202,7 @@ double3 __ovld __cnfn max(double3 x, double y); - double4 __ovld __cnfn max(double4 x, double y); - double8 __ovld __cnfn max(double8 x, double y); - double16 __ovld __cnfn max(double16 x, double y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn max(half x, half y); - half2 __ovld __cnfn max(half2 x, half2 y); -@@ -10219,7 +10232,7 @@ float3 __ovld __cnfn min(float3 x, float y); - float4 __ovld __cnfn min(float4 x, float y); - float8 __ovld __cnfn min(float8 x, float y); - float16 __ovld __cnfn min(float16 x, float y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn min(double x, double y); - double2 __ovld __cnfn min(double2 x, double2 y); - double3 __ovld __cnfn min(double3 x, double3 y); -@@ -10231,7 +10244,7 @@ double3 __ovld __cnfn min(double3 x, double y); - double4 __ovld __cnfn min(double4 x, double y); - double8 __ovld __cnfn min(double8 x, double y); - double16 __ovld __cnfn min(double16 x, double y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn min(half x, half y); - half2 __ovld __cnfn min(half2 x, half2 y); -@@ -10264,7 +10277,7 @@ float3 __ovld __cnfn mix(float3 x, float3 y, float a); - float4 __ovld __cnfn mix(float4 x, float4 y, float a); - float8 __ovld __cnfn mix(float8 x, float8 y, float a); - float16 __ovld __cnfn mix(float16 x, float16 y, float a); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn mix(double x, double y, double a); - double2 __ovld __cnfn mix(double2 x, double2 y, double2 a); - double3 __ovld __cnfn mix(double3 x, double3 y, double3 a); -@@ -10276,7 +10289,7 @@ double3 __ovld __cnfn mix(double3 x, double3 y, double a); - double4 __ovld __cnfn mix(double4 x, double4 y, double a); - double8 __ovld __cnfn mix(double8 x, double8 y, double a); - double16 __ovld __cnfn mix(double16 x, double16 y, double a); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn mix(half x, half y, half a); - half2 __ovld __cnfn mix(half2 x, half2 y, half2 a); -@@ -10301,14 +10314,14 @@ float3 __ovld __cnfn radians(float3 degrees); - float4 __ovld __cnfn radians(float4 degrees); - float8 __ovld __cnfn radians(float8 degrees); - float16 __ovld __cnfn radians(float16 degrees); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn radians(double degrees); - double2 __ovld __cnfn radians(double2 degrees); - double3 __ovld __cnfn radians(double3 degrees); - double4 __ovld __cnfn radians(double4 degrees); - double8 __ovld __cnfn radians(double8 degrees); - double16 __ovld __cnfn radians(double16 degrees); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn radians(half degrees); - half2 __ovld __cnfn radians(half2 degrees); -@@ -10332,7 +10345,7 @@ float3 __ovld __cnfn step(float edge, float3 x); - float4 __ovld __cnfn step(float edge, float4 x); - float8 __ovld __cnfn step(float edge, float8 x); - float16 __ovld __cnfn step(float edge, float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn step(double edge, double x); - double2 __ovld __cnfn step(double2 edge, double2 x); - double3 __ovld __cnfn step(double3 edge, double3 x); -@@ -10344,7 +10357,7 @@ double3 __ovld __cnfn step(double edge, double3 x); - double4 __ovld __cnfn step(double edge, double4 x); - double8 __ovld __cnfn step(double edge, double8 x); - double16 __ovld __cnfn step(double edge, double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn step(half edge, half x); - half2 __ovld __cnfn step(half2 edge, half2 x); -@@ -10383,7 +10396,7 @@ float3 __ovld __cnfn smoothstep(float edge0, float edge1, float3 x); - float4 __ovld __cnfn smoothstep(float edge0, float edge1, float4 x); - float8 __ovld __cnfn smoothstep(float edge0, float edge1, float8 x); - float16 __ovld __cnfn smoothstep(float edge0, float edge1, float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn smoothstep(double edge0, double edge1, double x); - double2 __ovld __cnfn smoothstep(double2 edge0, double2 edge1, double2 x); - double3 __ovld __cnfn smoothstep(double3 edge0, double3 edge1, double3 x); -@@ -10395,7 +10408,7 @@ double3 __ovld __cnfn smoothstep(double edge0, double edge1, double3 x); - double4 __ovld __cnfn smoothstep(double edge0, double edge1, double4 x); - double8 __ovld __cnfn smoothstep(double edge0, double edge1, double8 x); - double16 __ovld __cnfn smoothstep(double edge0, double edge1, double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn smoothstep(half edge0, half edge1, half x); - half2 __ovld __cnfn smoothstep(half2 edge0, half2 edge1, half2 x); -@@ -10420,14 +10433,14 @@ float3 __ovld __cnfn sign(float3 x); - float4 __ovld __cnfn sign(float4 x); - float8 __ovld __cnfn sign(float8 x); - float16 __ovld __cnfn sign(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn sign(double x); - double2 __ovld __cnfn sign(double2 x); - double3 __ovld __cnfn sign(double3 x); - double4 __ovld __cnfn sign(double4 x); - double8 __ovld __cnfn sign(double8 x); - double16 __ovld __cnfn sign(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn sign(half x); - half2 __ovld __cnfn sign(half2 x); -@@ -10445,10 +10458,10 @@ half16 __ovld __cnfn sign(half16 x); - */ - float4 __ovld __cnfn cross(float4 p0, float4 p1); - float3 __ovld __cnfn cross(float3 p0, float3 p1); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double4 __ovld __cnfn cross(double4 p0, double4 p1); - double3 __ovld __cnfn cross(double3 p0, double3 p1); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half4 __ovld __cnfn cross(half4 p0, half4 p1); - half3 __ovld __cnfn cross(half3 p0, half3 p1); -@@ -10461,12 +10474,12 @@ float __ovld __cnfn dot(float p0, float p1); - float __ovld __cnfn dot(float2 p0, float2 p1); - float __ovld __cnfn dot(float3 p0, float3 p1); - float __ovld __cnfn dot(float4 p0, float4 p1); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn dot(double p0, double p1); - double __ovld __cnfn dot(double2 p0, double2 p1); - double __ovld __cnfn dot(double3 p0, double3 p1); - double __ovld __cnfn dot(double4 p0, double4 p1); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn dot(half p0, half p1); - half __ovld __cnfn dot(half2 p0, half2 p1); -@@ -10482,12 +10495,12 @@ float __ovld __cnfn distance(float p0, float p1); - float __ovld __cnfn distance(float2 p0, float2 p1); - float __ovld __cnfn distance(float3 p0, float3 p1); - float __ovld __cnfn distance(float4 p0, float4 p1); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn distance(double p0, double p1); - double __ovld __cnfn distance(double2 p0, double2 p1); - double __ovld __cnfn distance(double3 p0, double3 p1); - double __ovld __cnfn distance(double4 p0, double4 p1); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn distance(half p0, half p1); - half __ovld __cnfn distance(half2 p0, half2 p1); -@@ -10503,12 +10516,12 @@ float __ovld __cnfn length(float p); - float __ovld __cnfn length(float2 p); - float __ovld __cnfn length(float3 p); - float __ovld __cnfn length(float4 p); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn length(double p); - double __ovld __cnfn length(double2 p); - double __ovld __cnfn length(double3 p); - double __ovld __cnfn length(double4 p); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn length(half p); - half __ovld __cnfn length(half2 p); -@@ -10524,12 +10537,12 @@ float __ovld __cnfn normalize(float p); - float2 __ovld __cnfn normalize(float2 p); - float3 __ovld __cnfn normalize(float3 p); - float4 __ovld __cnfn normalize(float4 p); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn normalize(double p); - double2 __ovld __cnfn normalize(double2 p); - double3 __ovld __cnfn normalize(double3 p); - double4 __ovld __cnfn normalize(double4 p); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn normalize(half p); - half2 __ovld __cnfn normalize(half2 p); -@@ -10610,14 +10623,14 @@ int3 __ovld __cnfn isequal(float3 x, float3 y); - int4 __ovld __cnfn isequal(float4 x, float4 y); - int8 __ovld __cnfn isequal(float8 x, float8 y); - int16 __ovld __cnfn isequal(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn isequal(double x, double y); - long2 __ovld __cnfn isequal(double2 x, double2 y); - long3 __ovld __cnfn isequal(double3 x, double3 y); - long4 __ovld __cnfn isequal(double4 x, double4 y); - long8 __ovld __cnfn isequal(double8 x, double8 y); - long16 __ovld __cnfn isequal(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn isequal(half x, half y); - short2 __ovld __cnfn isequal(half2 x, half2 y); -@@ -10636,14 +10649,14 @@ int3 __ovld __cnfn isnotequal(float3 x, float3 y); - int4 __ovld __cnfn isnotequal(float4 x, float4 y); - int8 __ovld __cnfn isnotequal(float8 x, float8 y); - int16 __ovld __cnfn isnotequal(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn isnotequal(double x, double y); - long2 __ovld __cnfn isnotequal(double2 x, double2 y); - long3 __ovld __cnfn isnotequal(double3 x, double3 y); - long4 __ovld __cnfn isnotequal(double4 x, double4 y); - long8 __ovld __cnfn isnotequal(double8 x, double8 y); - long16 __ovld __cnfn isnotequal(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn isnotequal(half x, half y); - short2 __ovld __cnfn isnotequal(half2 x, half2 y); -@@ -10662,14 +10675,14 @@ int3 __ovld __cnfn isgreater(float3 x, float3 y); - int4 __ovld __cnfn isgreater(float4 x, float4 y); - int8 __ovld __cnfn isgreater(float8 x, float8 y); - int16 __ovld __cnfn isgreater(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn isgreater(double x, double y); - long2 __ovld __cnfn isgreater(double2 x, double2 y); - long3 __ovld __cnfn isgreater(double3 x, double3 y); - long4 __ovld __cnfn isgreater(double4 x, double4 y); - long8 __ovld __cnfn isgreater(double8 x, double8 y); - long16 __ovld __cnfn isgreater(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn isgreater(half x, half y); - short2 __ovld __cnfn isgreater(half2 x, half2 y); -@@ -10688,14 +10701,14 @@ int3 __ovld __cnfn isgreaterequal(float3 x, float3 y); - int4 __ovld __cnfn isgreaterequal(float4 x, float4 y); - int8 __ovld __cnfn isgreaterequal(float8 x, float8 y); - int16 __ovld __cnfn isgreaterequal(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn isgreaterequal(double x, double y); - long2 __ovld __cnfn isgreaterequal(double2 x, double2 y); - long3 __ovld __cnfn isgreaterequal(double3 x, double3 y); - long4 __ovld __cnfn isgreaterequal(double4 x, double4 y); - long8 __ovld __cnfn isgreaterequal(double8 x, double8 y); - long16 __ovld __cnfn isgreaterequal(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn isgreaterequal(half x, half y); - short2 __ovld __cnfn isgreaterequal(half2 x, half2 y); -@@ -10714,14 +10727,14 @@ int3 __ovld __cnfn isless(float3 x, float3 y); - int4 __ovld __cnfn isless(float4 x, float4 y); - int8 __ovld __cnfn isless(float8 x, float8 y); - int16 __ovld __cnfn isless(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn isless(double x, double y); - long2 __ovld __cnfn isless(double2 x, double2 y); - long3 __ovld __cnfn isless(double3 x, double3 y); - long4 __ovld __cnfn isless(double4 x, double4 y); - long8 __ovld __cnfn isless(double8 x, double8 y); - long16 __ovld __cnfn isless(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn isless(half x, half y); - short2 __ovld __cnfn isless(half2 x, half2 y); -@@ -10740,14 +10753,14 @@ int3 __ovld __cnfn islessequal(float3 x, float3 y); - int4 __ovld __cnfn islessequal(float4 x, float4 y); - int8 __ovld __cnfn islessequal(float8 x, float8 y); - int16 __ovld __cnfn islessequal(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn islessequal(double x, double y); - long2 __ovld __cnfn islessequal(double2 x, double2 y); - long3 __ovld __cnfn islessequal(double3 x, double3 y); - long4 __ovld __cnfn islessequal(double4 x, double4 y); - long8 __ovld __cnfn islessequal(double8 x, double8 y); - long16 __ovld __cnfn islessequal(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn islessequal(half x, half y); - short2 __ovld __cnfn islessequal(half2 x, half2 y); -@@ -10767,14 +10780,14 @@ int3 __ovld __cnfn islessgreater(float3 x, float3 y); - int4 __ovld __cnfn islessgreater(float4 x, float4 y); - int8 __ovld __cnfn islessgreater(float8 x, float8 y); - int16 __ovld __cnfn islessgreater(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn islessgreater(double x, double y); - long2 __ovld __cnfn islessgreater(double2 x, double2 y); - long3 __ovld __cnfn islessgreater(double3 x, double3 y); - long4 __ovld __cnfn islessgreater(double4 x, double4 y); - long8 __ovld __cnfn islessgreater(double8 x, double8 y); - long16 __ovld __cnfn islessgreater(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn islessgreater(half x, half y); - short2 __ovld __cnfn islessgreater(half2 x, half2 y); -@@ -10793,14 +10806,14 @@ int3 __ovld __cnfn isfinite(float3); - int4 __ovld __cnfn isfinite(float4); - int8 __ovld __cnfn isfinite(float8); - int16 __ovld __cnfn isfinite(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn isfinite(double); - long2 __ovld __cnfn isfinite(double2); - long3 __ovld __cnfn isfinite(double3); - long4 __ovld __cnfn isfinite(double4); - long8 __ovld __cnfn isfinite(double8); - long16 __ovld __cnfn isfinite(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn isfinite(half); - short2 __ovld __cnfn isfinite(half2); -@@ -10819,14 +10832,14 @@ int3 __ovld __cnfn isinf(float3); - int4 __ovld __cnfn isinf(float4); - int8 __ovld __cnfn isinf(float8); - int16 __ovld __cnfn isinf(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn isinf(double); - long2 __ovld __cnfn isinf(double2); - long3 __ovld __cnfn isinf(double3); - long4 __ovld __cnfn isinf(double4); - long8 __ovld __cnfn isinf(double8); - long16 __ovld __cnfn isinf(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn isinf(half); - short2 __ovld __cnfn isinf(half2); -@@ -10845,14 +10858,14 @@ int3 __ovld __cnfn isnan(float3); - int4 __ovld __cnfn isnan(float4); - int8 __ovld __cnfn isnan(float8); - int16 __ovld __cnfn isnan(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn isnan(double); - long2 __ovld __cnfn isnan(double2); - long3 __ovld __cnfn isnan(double3); - long4 __ovld __cnfn isnan(double4); - long8 __ovld __cnfn isnan(double8); - long16 __ovld __cnfn isnan(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn isnan(half); - short2 __ovld __cnfn isnan(half2); -@@ -10871,14 +10884,14 @@ int3 __ovld __cnfn isnormal(float3); - int4 __ovld __cnfn isnormal(float4); - int8 __ovld __cnfn isnormal(float8); - int16 __ovld __cnfn isnormal(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn isnormal(double); - long2 __ovld __cnfn isnormal(double2); - long3 __ovld __cnfn isnormal(double3); - long4 __ovld __cnfn isnormal(double4); - long8 __ovld __cnfn isnormal(double8); - long16 __ovld __cnfn isnormal(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn isnormal(half); - short2 __ovld __cnfn isnormal(half2); -@@ -10899,14 +10912,14 @@ int3 __ovld __cnfn isordered(float3 x, float3 y); - int4 __ovld __cnfn isordered(float4 x, float4 y); - int8 __ovld __cnfn isordered(float8 x, float8 y); - int16 __ovld __cnfn isordered(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn isordered(double x, double y); - long2 __ovld __cnfn isordered(double2 x, double2 y); - long3 __ovld __cnfn isordered(double3 x, double3 y); - long4 __ovld __cnfn isordered(double4 x, double4 y); - long8 __ovld __cnfn isordered(double8 x, double8 y); - long16 __ovld __cnfn isordered(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn isordered(half x, half y); - short2 __ovld __cnfn isordered(half2 x, half2 y); -@@ -10927,14 +10940,14 @@ int3 __ovld __cnfn isunordered(float3 x, float3 y); - int4 __ovld __cnfn isunordered(float4 x, float4 y); - int8 __ovld __cnfn isunordered(float8 x, float8 y); - int16 __ovld __cnfn isunordered(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn isunordered(double x, double y); - long2 __ovld __cnfn isunordered(double2 x, double2 y); - long3 __ovld __cnfn isunordered(double3 x, double3 y); - long4 __ovld __cnfn isunordered(double4 x, double4 y); - long8 __ovld __cnfn isunordered(double8 x, double8 y); - long16 __ovld __cnfn isunordered(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn isunordered(half x, half y); - short2 __ovld __cnfn isunordered(half2 x, half2 y); -@@ -10957,14 +10970,14 @@ int3 __ovld __cnfn signbit(float3); - int4 __ovld __cnfn signbit(float4); - int8 __ovld __cnfn signbit(float8); - int16 __ovld __cnfn signbit(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn signbit(double); - long2 __ovld __cnfn signbit(double2); - long3 __ovld __cnfn signbit(double3); - long4 __ovld __cnfn signbit(double4); - long8 __ovld __cnfn signbit(double8); - long16 __ovld __cnfn signbit(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn signbit(half); - short2 __ovld __cnfn signbit(half2); -@@ -11091,14 +11104,14 @@ float3 __ovld __cnfn bitselect(float3 a, float3 b, float3 c); - float4 __ovld __cnfn bitselect(float4 a, float4 b, float4 c); - float8 __ovld __cnfn bitselect(float8 a, float8 b, float8 c); - float16 __ovld __cnfn bitselect(float16 a, float16 b, float16 c); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn bitselect(double a, double b, double c); - double2 __ovld __cnfn bitselect(double2 a, double2 b, double2 c); - double3 __ovld __cnfn bitselect(double3 a, double3 b, double3 c); - double4 __ovld __cnfn bitselect(double4 a, double4 b, double4 c); - double8 __ovld __cnfn bitselect(double8 a, double8 b, double8 c); - double16 __ovld __cnfn bitselect(double16 a, double16 b, double16 c); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn bitselect(half a, half b, half c); - half2 __ovld __cnfn bitselect(half2 a, half2 b, half2 c); -@@ -11231,7 +11244,7 @@ ulong8 __ovld __cnfn select(ulong8 a, ulong8 b, ulong8 c); - long16 __ovld __cnfn select(long16 a, long16 b, ulong16 c); - ulong16 __ovld __cnfn select(ulong16 a, ulong16 b, ulong16 c); - --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn select(double a, double b, long c); - double2 __ovld __cnfn select(double2 a, double2 b, long2 c); - double3 __ovld __cnfn select(double3 a, double3 b, long3 c); -@@ -11244,7 +11257,7 @@ double3 __ovld __cnfn select(double3 a, double3 b, ulong3 c); - double4 __ovld __cnfn select(double4 a, double4 b, ulong4 c); - double8 __ovld __cnfn select(double8 a, double8 b, ulong8 c); - double16 __ovld __cnfn select(double16 a, double16 b, ulong16 c); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn select(half a, half b, short c); - half2 __ovld __cnfn select(half2 a, half2 b, short2 c); -@@ -11323,13 +11336,13 @@ uint16 __ovld vload16(size_t offset, const __constant uint *p); - long16 __ovld vload16(size_t offset, const __constant long *p); - ulong16 __ovld vload16(size_t offset, const __constant ulong *p); - float16 __ovld vload16(size_t offset, const __constant float *p); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double2 __ovld vload2(size_t offset, const __constant double *p); - double3 __ovld vload3(size_t offset, const __constant double *p); - double4 __ovld vload4(size_t offset, const __constant double *p); - double8 __ovld vload8(size_t offset, const __constant double *p); - double16 __ovld vload16(size_t offset, const __constant double *p); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - #ifdef cl_khr_fp16 - half __ovld vload(size_t offset, const __constant half *p); -@@ -11340,7 +11353,7 @@ half8 __ovld vload8(size_t offset, const __constant half *p); - half16 __ovld vload16(size_t offset, const __constant half *p); - #endif //cl_khr_fp16 - --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - char2 __ovld vload2(size_t offset, const char *p); - uchar2 __ovld vload2(size_t offset, const uchar *p); - short2 __ovld vload2(size_t offset, const short *p); -@@ -11387,13 +11400,13 @@ long16 __ovld vload16(size_t offset, const long *p); - ulong16 __ovld vload16(size_t offset, const ulong *p); - float16 __ovld vload16(size_t offset, const float *p); - --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double2 __ovld vload2(size_t offset, const double *p); - double3 __ovld vload3(size_t offset, const double *p); - double4 __ovld vload4(size_t offset, const double *p); - double8 __ovld vload8(size_t offset, const double *p); - double16 __ovld vload16(size_t offset, const double *p); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - #ifdef cl_khr_fp16 - half __ovld vload(size_t offset, const half *p); -@@ -11403,7 +11416,7 @@ half4 __ovld vload4(size_t offset, const half *p); - half8 __ovld vload8(size_t offset, const half *p); - half16 __ovld vload16(size_t offset, const half *p); - #endif //cl_khr_fp16 --#else -+#endif //__opencl_c_generic_address_space - char2 __ovld vload2(size_t offset, const __global char *p); - uchar2 __ovld vload2(size_t offset, const __global uchar *p); - short2 __ovld vload2(size_t offset, const __global short *p); -@@ -11540,7 +11553,7 @@ long16 __ovld vload16(size_t offset, const __private long *p); - ulong16 __ovld vload16(size_t offset, const __private ulong *p); - float16 __ovld vload16(size_t offset, const __private float *p); - --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double2 __ovld vload2(size_t offset, const __global double *p); - double3 __ovld vload3(size_t offset, const __global double *p); - double4 __ovld vload4(size_t offset, const __global double *p); -@@ -11556,7 +11569,7 @@ double3 __ovld vload3(size_t offset, const __private double *p); - double4 __ovld vload4(size_t offset, const __private double *p); - double8 __ovld vload8(size_t offset, const __private double *p); - double16 __ovld vload16(size_t offset, const __private double *p); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - #ifdef cl_khr_fp16 - half __ovld vload(size_t offset, const __global half *p); -@@ -11578,9 +11591,8 @@ half4 __ovld vload4(size_t offset, const __private half *p); - half8 __ovld vload8(size_t offset, const __private half *p); - half16 __ovld vload16(size_t offset, const __private half *p); - #endif //cl_khr_fp16 --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - void __ovld vstore2(char2 data, size_t offset, char *p); - void __ovld vstore2(uchar2 data, size_t offset, uchar *p); - void __ovld vstore2(short2 data, size_t offset, short *p); -@@ -11626,13 +11638,13 @@ void __ovld vstore16(uint16 data, size_t offset, uint *p); - void __ovld vstore16(long16 data, size_t offset, long *p); - void __ovld vstore16(ulong16 data, size_t offset, ulong *p); - void __ovld vstore16(float16 data, size_t offset, float *p); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - void __ovld vstore2(double2 data, size_t offset, double *p); - void __ovld vstore3(double3 data, size_t offset, double *p); - void __ovld vstore4(double4 data, size_t offset, double *p); - void __ovld vstore8(double8 data, size_t offset, double *p); - void __ovld vstore16(double16 data, size_t offset, double *p); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - void __ovld vstore(half data, size_t offset, half *p); - void __ovld vstore2(half2 data, size_t offset, half *p); -@@ -11641,7 +11653,7 @@ void __ovld vstore4(half4 data, size_t offset, half *p); - void __ovld vstore8(half8 data, size_t offset, half *p); - void __ovld vstore16(half16 data, size_t offset, half *p); - #endif //cl_khr_fp16 --#else -+#endif //__opencl_c_generic_address_space - void __ovld vstore2(char2 data, size_t offset, __global char *p); - void __ovld vstore2(uchar2 data, size_t offset, __global uchar *p); - void __ovld vstore2(short2 data, size_t offset, __global short *p); -@@ -11777,7 +11789,7 @@ void __ovld vstore16(uint16 data, size_t offset, __private uint *p); - void __ovld vstore16(long16 data, size_t offset, __private long *p); - void __ovld vstore16(ulong16 data, size_t offset, __private ulong *p); - void __ovld vstore16(float16 data, size_t offset, __private float *p); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - void __ovld vstore2(double2 data, size_t offset, __global double *p); - void __ovld vstore3(double3 data, size_t offset, __global double *p); - void __ovld vstore4(double4 data, size_t offset, __global double *p); -@@ -11793,7 +11805,7 @@ void __ovld vstore3(double3 data, size_t offset, __private double *p); - void __ovld vstore4(double4 data, size_t offset, __private double *p); - void __ovld vstore8(double8 data, size_t offset, __private double *p); - void __ovld vstore16(double16 data, size_t offset, __private double *p); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - void __ovld vstore(half data, size_t offset, __global half *p); - void __ovld vstore2(half2 data, size_t offset, __global half *p); -@@ -11814,7 +11826,6 @@ void __ovld vstore4(half4 data, size_t offset, __private half *p); - void __ovld vstore8(half8 data, size_t offset, __private half *p); - void __ovld vstore16(half16 data, size_t offset, __private half *p); - #endif //cl_khr_fp16 --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** - * Read sizeof (half) bytes of data from address -@@ -11825,13 +11836,12 @@ void __ovld vstore16(half16 data, size_t offset, __private half *p); - * must be 16-bit aligned. - */ - float __ovld vload_half(size_t offset, const __constant half *p); --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - float __ovld vload_half(size_t offset, const half *p); --#else -+#endif //__opencl_c_generic_address_space - float __ovld vload_half(size_t offset, const __global half *p); - float __ovld vload_half(size_t offset, const __local half *p); - float __ovld vload_half(size_t offset, const __private half *p); --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** - * Read sizeof (halfn) bytes of data from address -@@ -11846,13 +11856,13 @@ float3 __ovld vload_half3(size_t offset, const __constant half *p); - float4 __ovld vload_half4(size_t offset, const __constant half *p); - float8 __ovld vload_half8(size_t offset, const __constant half *p); - float16 __ovld vload_half16(size_t offset, const __constant half *p); --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - float2 __ovld vload_half2(size_t offset, const half *p); - float3 __ovld vload_half3(size_t offset, const half *p); - float4 __ovld vload_half4(size_t offset, const half *p); - float8 __ovld vload_half8(size_t offset, const half *p); - float16 __ovld vload_half16(size_t offset, const half *p); --#else -+#endif //__opencl_c_generic_address_space - float2 __ovld vload_half2(size_t offset, const __global half *p); - float3 __ovld vload_half3(size_t offset, const __global half *p); - float4 __ovld vload_half4(size_t offset, const __global half *p); -@@ -11868,7 +11878,6 @@ float3 __ovld vload_half3(size_t offset, const __private half *p); - float4 __ovld vload_half4(size_t offset, const __private half *p); - float8 __ovld vload_half8(size_t offset, const __private half *p); - float16 __ovld vload_half16(size_t offset, const __private half *p); --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** - * The float value given by data is first -@@ -11881,20 +11890,20 @@ float16 __ovld vload_half16(size_t offset, const __private half *p); - * The default current rounding mode is round to - * nearest even. - */ --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - void __ovld vstore_half(float data, size_t offset, half *p); - void __ovld vstore_half_rte(float data, size_t offset, half *p); - void __ovld vstore_half_rtz(float data, size_t offset, half *p); - void __ovld vstore_half_rtp(float data, size_t offset, half *p); - void __ovld vstore_half_rtn(float data, size_t offset, half *p); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - void __ovld vstore_half(double data, size_t offset, half *p); - void __ovld vstore_half_rte(double data, size_t offset, half *p); - void __ovld vstore_half_rtz(double data, size_t offset, half *p); - void __ovld vstore_half_rtp(double data, size_t offset, half *p); - void __ovld vstore_half_rtn(double data, size_t offset, half *p); --#endif //cl_khr_fp64 --#else -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#endif //__opencl_c_generic_address_space - void __ovld vstore_half(float data, size_t offset, __global half *p); - void __ovld vstore_half_rte(float data, size_t offset, __global half *p); - void __ovld vstore_half_rtz(float data, size_t offset, __global half *p); -@@ -11910,7 +11919,7 @@ void __ovld vstore_half_rte(float data, size_t offset, __private half *p); - void __ovld vstore_half_rtz(float data, size_t offset, __private half *p); - void __ovld vstore_half_rtp(float data, size_t offset, __private half *p); - void __ovld vstore_half_rtn(float data, size_t offset, __private half *p); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - void __ovld vstore_half(double data, size_t offset, __global half *p); - void __ovld vstore_half_rte(double data, size_t offset, __global half *p); - void __ovld vstore_half_rtz(double data, size_t offset, __global half *p); -@@ -11926,8 +11935,7 @@ void __ovld vstore_half_rte(double data, size_t offset, __private half *p); - void __ovld vstore_half_rtz(double data, size_t offset, __private half *p); - void __ovld vstore_half_rtp(double data, size_t offset, __private half *p); - void __ovld vstore_half_rtn(double data, size_t offset, __private half *p); --#endif //cl_khr_fp64 --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - /** - * The floatn value given by data is converted to -@@ -11940,7 +11948,7 @@ void __ovld vstore_half_rtn(double data, size_t offset, __private half *p); - * The default current rounding mode is round to - * nearest even. - */ --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - void __ovld vstore_half2(float2 data, size_t offset, half *p); - void __ovld vstore_half3(float3 data, size_t offset, half *p); - void __ovld vstore_half4(float4 data, size_t offset, half *p); -@@ -11966,7 +11974,7 @@ void __ovld vstore_half3_rtn(float3 data, size_t offset, half *p); - void __ovld vstore_half4_rtn(float4 data, size_t offset, half *p); - void __ovld vstore_half8_rtn(float8 data, size_t offset, half *p); - void __ovld vstore_half16_rtn(float16 data, size_t offset, half *p); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - void __ovld vstore_half2(double2 data, size_t offset, half *p); - void __ovld vstore_half3(double3 data, size_t offset, half *p); - void __ovld vstore_half4(double4 data, size_t offset, half *p); -@@ -11992,8 +12000,8 @@ void __ovld vstore_half3_rtn(double3 data, size_t offset, half *p); - void __ovld vstore_half4_rtn(double4 data, size_t offset, half *p); - void __ovld vstore_half8_rtn(double8 data, size_t offset, half *p); - void __ovld vstore_half16_rtn(double16 data, size_t offset, half *p); --#endif //cl_khr_fp64 --#else -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#endif //__opencl_c_generic_address_space - void __ovld vstore_half2(float2 data, size_t offset, __global half *p); - void __ovld vstore_half3(float3 data, size_t offset, __global half *p); - void __ovld vstore_half4(float4 data, size_t offset, __global half *p); -@@ -12069,7 +12077,7 @@ void __ovld vstore_half3_rtn(float3 data, size_t offset, __private half *p); - void __ovld vstore_half4_rtn(float4 data, size_t offset, __private half *p); - void __ovld vstore_half8_rtn(float8 data, size_t offset, __private half *p); - void __ovld vstore_half16_rtn(float16 data, size_t offset, __private half *p); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - void __ovld vstore_half2(double2 data, size_t offset, __global half *p); - void __ovld vstore_half3(double3 data, size_t offset, __global half *p); - void __ovld vstore_half4(double4 data, size_t offset, __global half *p); -@@ -12145,8 +12153,7 @@ void __ovld vstore_half3_rtn(double3 data, size_t offset, __private half *p); - void __ovld vstore_half4_rtn(double4 data, size_t offset, __private half *p); - void __ovld vstore_half8_rtn(double8 data, size_t offset, __private half *p); - void __ovld vstore_half16_rtn(double16 data, size_t offset, __private half *p); --#endif //cl_khr_fp64 --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - /** - * For n = 1, 2, 4, 8 and 16 read sizeof (halfn) -@@ -12167,14 +12174,14 @@ float3 __ovld vloada_half3(size_t offset, const __constant half *p); - float4 __ovld vloada_half4(size_t offset, const __constant half *p); - float8 __ovld vloada_half8(size_t offset, const __constant half *p); - float16 __ovld vloada_half16(size_t offset, const __constant half *p); --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - float __ovld vloada_half(size_t offset, const half *p); - float2 __ovld vloada_half2(size_t offset, const half *p); - float3 __ovld vloada_half3(size_t offset, const half *p); - float4 __ovld vloada_half4(size_t offset, const half *p); - float8 __ovld vloada_half8(size_t offset, const half *p); - float16 __ovld vloada_half16(size_t offset, const half *p); --#else -+#endif //__opencl_c_generic_address_space - float __ovld vloada_half(size_t offset, const __global half *p); - float2 __ovld vloada_half2(size_t offset, const __global half *p); - float3 __ovld vloada_half3(size_t offset, const __global half *p); -@@ -12193,7 +12200,6 @@ float3 __ovld vloada_half3(size_t offset, const __private half *p); - float4 __ovld vloada_half4(size_t offset, const __private half *p); - float8 __ovld vloada_half8(size_t offset, const __private half *p); - float16 __ovld vloada_half16(size_t offset, const __private half *p); --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** - * The floatn value given by data is converted to -@@ -12211,7 +12217,7 @@ float16 __ovld vloada_half16(size_t offset, const __private half *p); - * mode. The default current rounding mode is - * round to nearest even. - */ --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - void __ovld vstorea_half(float data, size_t offset, half *p); - void __ovld vstorea_half2(float2 data, size_t offset, half *p); - void __ovld vstorea_half3(float3 data, size_t offset, half *p); -@@ -12247,7 +12253,7 @@ void __ovld vstorea_half4_rtn(float4 data, size_t offset, half *p); - void __ovld vstorea_half8_rtn(float8 data, size_t offset, half *p); - void __ovld vstorea_half16_rtn(float16 data, size_t offset, half *p); - --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - void __ovld vstorea_half(double data, size_t offset, half *p); - void __ovld vstorea_half2(double2 data, size_t offset, half *p); - void __ovld vstorea_half3(double3 data, size_t offset, half *p); -@@ -12282,9 +12288,9 @@ void __ovld vstorea_half3_rtn(double3 data, size_t offset, half *p); - void __ovld vstorea_half4_rtn(double4 data, size_t offset, half *p); - void __ovld vstorea_half8_rtn(double8 data, size_t offset, half *p); - void __ovld vstorea_half16_rtn(double16 data, size_t offset, half *p); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#endif //__opencl_c_generic_address_space - --#else - void __ovld vstorea_half(float data, size_t offset, __global half *p); - void __ovld vstorea_half2(float2 data, size_t offset, __global half *p); - void __ovld vstorea_half3(float3 data, size_t offset, __global half *p); -@@ -12390,7 +12396,7 @@ void __ovld vstorea_half4_rtn(float4 data, size_t offset, __private half *p); - void __ovld vstorea_half8_rtn(float8 data, size_t offset, __private half *p); - void __ovld vstorea_half16_rtn(float16 data, size_t offset, __private half *p); - --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - void __ovld vstorea_half(double data, size_t offset, __global half *p); - void __ovld vstorea_half2(double2 data, size_t offset, __global half *p); - void __ovld vstorea_half3(double3 data, size_t offset, __global half *p); -@@ -12495,8 +12501,7 @@ void __ovld vstorea_half3_rtn(double3 data,size_t offset, __private half *p); - void __ovld vstorea_half4_rtn(double4 data,size_t offset, __private half *p); - void __ovld vstorea_half8_rtn(double8 data,size_t offset, __private half *p); - void __ovld vstorea_half16_rtn(double16 data,size_t offset, __private half *p); --#endif //cl_khr_fp64 --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - // OpenCL v1.1 s6.11.8, v1.2 s6.12.8, v2.0 s6.13.8 - Synchronization Functions - -@@ -12580,7 +12585,7 @@ void __ovld write_mem_fence(cl_mem_fence_flags flags); - - // OpenCL v2.0 s6.13.9 - Address Space Qualifier Functions - --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - cl_mem_fence_flags __ovld get_fence(const void *ptr); - cl_mem_fence_flags __ovld get_fence(void *ptr); - -@@ -12591,7 +12596,7 @@ cl_mem_fence_flags __ovld get_fence(void *ptr); - * where gentype is builtin type or user defined type. - */ - --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#endif //__opencl_c_generic_address_space - - // OpenCL v1.1 s6.11.10, v1.2 s6.12.10, v2.0 s6.13.10 - Async Copies from Global to Local Memory, Local to Global Memory, and Prefetch - -@@ -12730,7 +12735,7 @@ event_t __ovld async_work_group_copy(__global uint16 *dst, const __local uint16 - event_t __ovld async_work_group_copy(__global long16 *dst, const __local long16 *src, size_t num_elements, event_t event); - event_t __ovld async_work_group_copy(__global ulong16 *dst, const __local ulong16 *src, size_t num_elements, event_t event); - event_t __ovld async_work_group_copy(__global float16 *dst, const __local float16 *src, size_t num_elements, event_t event); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - event_t __ovld async_work_group_copy(__local double *dst, const __global double *src, size_t num_elements, event_t event); - event_t __ovld async_work_group_copy(__local double2 *dst, const __global double2 *src, size_t num_elements, event_t event); - event_t __ovld async_work_group_copy(__local double3 *dst, const __global double3 *src, size_t num_elements, event_t event); -@@ -12743,7 +12748,7 @@ event_t __ovld async_work_group_copy(__global double3 *dst, const __local double - event_t __ovld async_work_group_copy(__global double4 *dst, const __local double4 *src, size_t num_elements, event_t event); - event_t __ovld async_work_group_copy(__global double8 *dst, const __local double8 *src, size_t num_elements, event_t event); - event_t __ovld async_work_group_copy(__global double16 *dst, const __local double16 *src, size_t num_elements, event_t event); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - event_t __ovld async_work_group_copy(__local half *dst, const __global half *src, size_t num_elements, event_t event); - event_t __ovld async_work_group_copy(__local half2 *dst, const __global half2 *src, size_t num_elements, event_t event); -@@ -12893,7 +12898,7 @@ event_t __ovld async_work_group_strided_copy(__global uint16 *dst, const __local - event_t __ovld async_work_group_strided_copy(__global long16 *dst, const __local long16 *src, size_t num_elements, size_t dst_stride, event_t event); - event_t __ovld async_work_group_strided_copy(__global ulong16 *dst, const __local ulong16 *src, size_t num_elements, size_t dst_stride, event_t event); - event_t __ovld async_work_group_strided_copy(__global float16 *dst, const __local float16 *src, size_t num_elements, size_t dst_stride, event_t event); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - event_t __ovld async_work_group_strided_copy(__local double *dst, const __global double *src, size_t num_elements, size_t src_stride, event_t event); - event_t __ovld async_work_group_strided_copy(__local double2 *dst, const __global double2 *src, size_t num_elements, size_t src_stride, event_t event); - event_t __ovld async_work_group_strided_copy(__local double3 *dst, const __global double3 *src, size_t num_elements, size_t src_stride, event_t event); -@@ -12906,7 +12911,7 @@ event_t __ovld async_work_group_strided_copy(__global double3 *dst, const __loca - event_t __ovld async_work_group_strided_copy(__global double4 *dst, const __local double4 *src, size_t num_elements, size_t dst_stride, event_t event); - event_t __ovld async_work_group_strided_copy(__global double8 *dst, const __local double8 *src, size_t num_elements, size_t dst_stride, event_t event); - event_t __ovld async_work_group_strided_copy(__global double16 *dst, const __local double16 *src, size_t num_elements, size_t dst_stride, event_t event); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - event_t __ovld async_work_group_strided_copy(__local half *dst, const __global half *src, size_t num_elements, size_t src_stride, event_t event); - event_t __ovld async_work_group_strided_copy(__local half2 *dst, const __global half2 *src, size_t num_elements, size_t src_stride, event_t event); -@@ -12996,14 +13001,14 @@ void __ovld prefetch(const __global uint16 *p, size_t num_elements); - void __ovld prefetch(const __global long16 *p, size_t num_elements); - void __ovld prefetch(const __global ulong16 *p, size_t num_elements); - void __ovld prefetch(const __global float16 *p, size_t num_elements); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - void __ovld prefetch(const __global double *p, size_t num_elements); - void __ovld prefetch(const __global double2 *p, size_t num_elements); - void __ovld prefetch(const __global double3 *p, size_t num_elements); - void __ovld prefetch(const __global double4 *p, size_t num_elements); - void __ovld prefetch(const __global double8 *p, size_t num_elements); - void __ovld prefetch(const __global double16 *p, size_t num_elements); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - void __ovld prefetch(const __global half *p, size_t num_elements); - void __ovld prefetch(const __global half2 *p, size_t num_elements); -@@ -13026,9 +13031,11 @@ void __ovld prefetch(const __global half16 *p, size_t num_elements); - * pointed by p. The function returns old. - */ - int __ovld atomic_add(volatile __global int *p, int val); --unsigned int __ovld atomic_add(volatile __global unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_add(volatile __global unsigned int *p, -+ unsigned int val); - int __ovld atomic_add(volatile __local int *p, int val); --unsigned int __ovld atomic_add(volatile __local unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_add(volatile __local unsigned int *p, -+ unsigned int val); - #ifdef __OPENCL_CPP_VERSION__ - int __ovld atomic_add(volatile int *p, int val); - unsigned int __ovld atomic_add(volatile unsigned int *p, unsigned int val); -@@ -13056,9 +13063,11 @@ unsigned long __ovld atom_add(volatile __local unsigned long *p, unsigned long v - * returns old. - */ - int __ovld atomic_sub(volatile __global int *p, int val); --unsigned int __ovld atomic_sub(volatile __global unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_sub(volatile __global unsigned int *p, -+ unsigned int val); - int __ovld atomic_sub(volatile __local int *p, int val); --unsigned int __ovld atomic_sub(volatile __local unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_sub(volatile __local unsigned int *p, -+ unsigned int val); - #ifdef __OPENCL_CPP_VERSION__ - int __ovld atomic_sub(volatile int *p, int val); - unsigned int __ovld atomic_sub(volatile unsigned int *p, unsigned int val); -@@ -13086,9 +13095,11 @@ unsigned long __ovld atom_sub(volatile __local unsigned long *p, unsigned long v - * value. - */ - int __ovld atomic_xchg(volatile __global int *p, int val); --unsigned int __ovld atomic_xchg(volatile __global unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_xchg(volatile __global unsigned int *p, -+ unsigned int val); - int __ovld atomic_xchg(volatile __local int *p, int val); --unsigned int __ovld atomic_xchg(volatile __local unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_xchg(volatile __local unsigned int *p, -+ unsigned int val); - float __ovld atomic_xchg(volatile __global float *p, float val); - float __ovld atomic_xchg(volatile __local float *p, float val); - #ifdef __OPENCL_CPP_VERSION__ -@@ -13183,12 +13194,15 @@ unsigned long __ovld atom_dec(volatile __local unsigned long *p); - * returns old. - */ - int __ovld atomic_cmpxchg(volatile __global int *p, int cmp, int val); --unsigned int __ovld atomic_cmpxchg(volatile __global unsigned int *p, unsigned int cmp, unsigned int val); -+unsigned int __ovld atomic_cmpxchg(volatile __global unsigned int *p, -+ unsigned int cmp, unsigned int val); - int __ovld atomic_cmpxchg(volatile __local int *p, int cmp, int val); --unsigned int __ovld atomic_cmpxchg(volatile __local unsigned int *p, unsigned int cmp, unsigned int val); -+unsigned int __ovld atomic_cmpxchg(volatile __local unsigned int *p, -+ unsigned int cmp, unsigned int val); - #ifdef __OPENCL_CPP_VERSION__ - int __ovld atomic_cmpxchg(volatile int *p, int cmp, int val); --unsigned int __ovld atomic_cmpxchg(volatile unsigned int *p, unsigned int cmp, unsigned int val); -+unsigned int __ovld atomic_cmpxchg(volatile unsigned int *p, unsigned int cmp, -+ unsigned int val); - #endif - - #if defined(cl_khr_global_int32_base_atomics) -@@ -13215,9 +13229,11 @@ unsigned long __ovld atom_cmpxchg(volatile __local unsigned long *p, unsigned lo - * returns old. - */ - int __ovld atomic_min(volatile __global int *p, int val); --unsigned int __ovld atomic_min(volatile __global unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_min(volatile __global unsigned int *p, -+ unsigned int val); - int __ovld atomic_min(volatile __local int *p, int val); --unsigned int __ovld atomic_min(volatile __local unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_min(volatile __local unsigned int *p, -+ unsigned int val); - #ifdef __OPENCL_CPP_VERSION__ - int __ovld atomic_min(volatile int *p, int val); - unsigned int __ovld atomic_min(volatile unsigned int *p, unsigned int val); -@@ -13247,9 +13263,11 @@ unsigned long __ovld atom_min(volatile __local unsigned long *p, unsigned long v - * returns old. - */ - int __ovld atomic_max(volatile __global int *p, int val); --unsigned int __ovld atomic_max(volatile __global unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_max(volatile __global unsigned int *p, -+ unsigned int val); - int __ovld atomic_max(volatile __local int *p, int val); --unsigned int __ovld atomic_max(volatile __local unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_max(volatile __local unsigned int *p, -+ unsigned int val); - #ifdef __OPENCL_CPP_VERSION__ - int __ovld atomic_max(volatile int *p, int val); - unsigned int __ovld atomic_max(volatile unsigned int *p, unsigned int val); -@@ -13278,9 +13296,11 @@ unsigned long __ovld atom_max(volatile __local unsigned long *p, unsigned long v - * pointed by p. The function returns old. - */ - int __ovld atomic_and(volatile __global int *p, int val); --unsigned int __ovld atomic_and(volatile __global unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_and(volatile __global unsigned int *p, -+ unsigned int val); - int __ovld atomic_and(volatile __local int *p, int val); --unsigned int __ovld atomic_and(volatile __local unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_and(volatile __local unsigned int *p, -+ unsigned int val); - #ifdef __OPENCL_CPP_VERSION__ - int __ovld atomic_and(volatile int *p, int val); - unsigned int __ovld atomic_and(volatile unsigned int *p, unsigned int val); -@@ -13309,9 +13329,11 @@ unsigned long __ovld atom_and(volatile __local unsigned long *p, unsigned long v - * pointed by p. The function returns old. - */ - int __ovld atomic_or(volatile __global int *p, int val); --unsigned int __ovld atomic_or(volatile __global unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_or(volatile __global unsigned int *p, -+ unsigned int val); - int __ovld atomic_or(volatile __local int *p, int val); --unsigned int __ovld atomic_or(volatile __local unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_or(volatile __local unsigned int *p, -+ unsigned int val); - #ifdef __OPENCL_CPP_VERSION__ - int __ovld atomic_or(volatile int *p, int val); - unsigned int __ovld atomic_or(volatile unsigned int *p, unsigned int val); -@@ -13340,9 +13362,11 @@ unsigned long __ovld atom_or(volatile __local unsigned long *p, unsigned long va - * pointed by p. The function returns old. - */ - int __ovld atomic_xor(volatile __global int *p, int val); --unsigned int __ovld atomic_xor(volatile __global unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_xor(volatile __global unsigned int *p, -+ unsigned int val); - int __ovld atomic_xor(volatile __local int *p, int val); --unsigned int __ovld atomic_xor(volatile __local unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_xor(volatile __local unsigned int *p, -+ unsigned int val); - #ifdef __OPENCL_CPP_VERSION__ - int __ovld atomic_xor(volatile int *p, int val); - unsigned int __ovld atomic_xor(volatile unsigned int *p, unsigned int val); -@@ -13380,120 +13404,78 @@ unsigned long __ovld atom_xor(volatile __local unsigned long *p, unsigned long v - #endif - - // atomic_init() -+#ifdef __opencl_c_generic_address_space - void __ovld atomic_init(volatile atomic_int *object, int value); - void __ovld atomic_init(volatile atomic_uint *object, uint value); - void __ovld atomic_init(volatile atomic_float *object, float value); - #if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) - void __ovld atomic_init(volatile atomic_long *object, long value); - void __ovld atomic_init(volatile atomic_ulong *object, ulong value); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - void __ovld atomic_init(volatile atomic_double *object, double value); --#endif //cl_khr_fp64 --#endif -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // __opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+void __ovld atomic_init(volatile atomic_int __global *object, int value); -+void __ovld atomic_init(volatile atomic_int __local *object, int value); -+void __ovld atomic_init(volatile atomic_uint __global *object, uint value); -+void __ovld atomic_init(volatile atomic_uint __local *object, uint value); -+void __ovld atomic_init(volatile atomic_float __global *object, float value); -+void __ovld atomic_init(volatile atomic_float __local *object, float value); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+void __ovld atomic_init(volatile atomic_long __global *object, long value); -+void __ovld atomic_init(volatile atomic_long __local *object, long value); -+void __ovld atomic_init(volatile atomic_ulong __global *object, ulong value); -+void __ovld atomic_init(volatile atomic_ulong __local *object, ulong value); -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+void __ovld atomic_init(volatile atomic_double __global *object, double value); -+void __ovld atomic_init(volatile atomic_double __local *object, double value); -+#endif // cl_khr_fp64 -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) - - // atomic_work_item_fence() --void __ovld atomic_work_item_fence(cl_mem_fence_flags flags, memory_order order, memory_scope scope); -+void __ovld atomic_work_item_fence(cl_mem_fence_flags flags, memory_order order, -+ memory_scope scope); - - // atomic_fetch() -- -+#if defined(__opencl_c_atomic_scope_device) && \ -+ defined(__opencl_c_atomic_order_seq_cst) -+#ifdef __opencl_c_generic_address_space - int __ovld atomic_fetch_add(volatile atomic_int *object, int operand); --int __ovld atomic_fetch_add_explicit(volatile atomic_int *object, int operand, memory_order order); --int __ovld atomic_fetch_add_explicit(volatile atomic_int *object, int operand, memory_order order, memory_scope scope); - uint __ovld atomic_fetch_add(volatile atomic_uint *object, uint operand); --uint __ovld atomic_fetch_add_explicit(volatile atomic_uint *object, uint operand, memory_order order); --uint __ovld atomic_fetch_add_explicit(volatile atomic_uint *object, uint operand, memory_order order, memory_scope scope); - int __ovld atomic_fetch_sub(volatile atomic_int *object, int operand); --int __ovld atomic_fetch_sub_explicit(volatile atomic_int *object, int operand, memory_order order); --int __ovld atomic_fetch_sub_explicit(volatile atomic_int *object, int operand, memory_order order, memory_scope scope); - uint __ovld atomic_fetch_sub(volatile atomic_uint *object, uint operand); --uint __ovld atomic_fetch_sub_explicit(volatile atomic_uint *object, uint operand, memory_order order); --uint __ovld atomic_fetch_sub_explicit(volatile atomic_uint *object, uint operand, memory_order order, memory_scope scope); - int __ovld atomic_fetch_or(volatile atomic_int *object, int operand); --int __ovld atomic_fetch_or_explicit(volatile atomic_int *object, int operand, memory_order order); --int __ovld atomic_fetch_or_explicit(volatile atomic_int *object, int operand, memory_order order, memory_scope scope); - uint __ovld atomic_fetch_or(volatile atomic_uint *object, uint operand); --uint __ovld atomic_fetch_or_explicit(volatile atomic_uint *object, uint operand, memory_order order); --uint __ovld atomic_fetch_or_explicit(volatile atomic_uint *object, uint operand, memory_order order, memory_scope scope); - int __ovld atomic_fetch_xor(volatile atomic_int *object, int operand); --int __ovld atomic_fetch_xor_explicit(volatile atomic_int *object, int operand, memory_order order); --int __ovld atomic_fetch_xor_explicit(volatile atomic_int *object, int operand, memory_order order, memory_scope scope); - uint __ovld atomic_fetch_xor(volatile atomic_uint *object, uint operand); --uint __ovld atomic_fetch_xor_explicit(volatile atomic_uint *object, uint operand, memory_order order); --uint __ovld atomic_fetch_xor_explicit(volatile atomic_uint *object, uint operand, memory_order order, memory_scope scope); - int __ovld atomic_fetch_and(volatile atomic_int *object, int operand); --int __ovld atomic_fetch_and_explicit(volatile atomic_int *object, int operand, memory_order order); --int __ovld atomic_fetch_and_explicit(volatile atomic_int *object, int operand, memory_order order, memory_scope scope); - uint __ovld atomic_fetch_and(volatile atomic_uint *object, uint operand); --uint __ovld atomic_fetch_and_explicit(volatile atomic_uint *object, uint operand, memory_order order); --uint __ovld atomic_fetch_and_explicit(volatile atomic_uint *object, uint operand, memory_order order, memory_scope scope); - int __ovld atomic_fetch_min(volatile atomic_int *object, int operand); --int __ovld atomic_fetch_min_explicit(volatile atomic_int *object, int operand, memory_order order); --int __ovld atomic_fetch_min_explicit(volatile atomic_int *object, int operand, memory_order order, memory_scope scope); - uint __ovld atomic_fetch_min(volatile atomic_uint *object, uint operand); --uint __ovld atomic_fetch_min_explicit(volatile atomic_uint *object, uint operand, memory_order order); --uint __ovld atomic_fetch_min_explicit(volatile atomic_uint *object, uint operand, memory_order order, memory_scope scope); --uint __ovld atomic_fetch_min(volatile atomic_uint *object, int operand); --uint __ovld atomic_fetch_min_explicit(volatile atomic_uint *object, int operand, memory_order order); --uint __ovld atomic_fetch_min_explicit(volatile atomic_uint *object, int operand, memory_order order, memory_scope scope); - int __ovld atomic_fetch_max(volatile atomic_int *object, int operand); --int __ovld atomic_fetch_max_explicit(volatile atomic_int *object, int operand, memory_order order); --int __ovld atomic_fetch_max_explicit(volatile atomic_int *object, int operand, memory_order order, memory_scope scope); - uint __ovld atomic_fetch_max(volatile atomic_uint *object, uint operand); --uint __ovld atomic_fetch_max_explicit(volatile atomic_uint *object, uint operand, memory_order order); --uint __ovld atomic_fetch_max_explicit(volatile atomic_uint *object, uint operand, memory_order order, memory_scope scope); --uint __ovld atomic_fetch_max(volatile atomic_uint *object, int operand); --uint __ovld atomic_fetch_max_explicit(volatile atomic_uint *object, int operand, memory_order order); --uint __ovld atomic_fetch_max_explicit(volatile atomic_uint *object, int operand, memory_order order, memory_scope scope); - - #if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) - long __ovld atomic_fetch_add(volatile atomic_long *object, long operand); --long __ovld atomic_fetch_add_explicit(volatile atomic_long *object, long operand, memory_order order); --long __ovld atomic_fetch_add_explicit(volatile atomic_long *object, long operand, memory_order order, memory_scope scope); - ulong __ovld atomic_fetch_add(volatile atomic_ulong *object, ulong operand); --ulong __ovld atomic_fetch_add_explicit(volatile atomic_ulong *object, ulong operand, memory_order order); --ulong __ovld atomic_fetch_add_explicit(volatile atomic_ulong *object, ulong operand, memory_order order, memory_scope scope); - long __ovld atomic_fetch_sub(volatile atomic_long *object, long operand); --long __ovld atomic_fetch_sub_explicit(volatile atomic_long *object, long operand, memory_order order); --long __ovld atomic_fetch_sub_explicit(volatile atomic_long *object, long operand, memory_order order, memory_scope scope); - ulong __ovld atomic_fetch_sub(volatile atomic_ulong *object, ulong operand); --ulong __ovld atomic_fetch_sub_explicit(volatile atomic_ulong *object, ulong operand, memory_order order); --ulong __ovld atomic_fetch_sub_explicit(volatile atomic_ulong *object, ulong operand, memory_order order, memory_scope scope); - long __ovld atomic_fetch_or(volatile atomic_long *object, long operand); --long __ovld atomic_fetch_or_explicit(volatile atomic_long *object, long operand, memory_order order); --long __ovld atomic_fetch_or_explicit(volatile atomic_long *object, long operand, memory_order order, memory_scope scope); - ulong __ovld atomic_fetch_or(volatile atomic_ulong *object, ulong operand); --ulong __ovld atomic_fetch_or_explicit(volatile atomic_ulong *object, ulong operand, memory_order order); --ulong __ovld atomic_fetch_or_explicit(volatile atomic_ulong *object, ulong operand, memory_order order, memory_scope scope); - long __ovld atomic_fetch_xor(volatile atomic_long *object, long operand); --long __ovld atomic_fetch_xor_explicit(volatile atomic_long *object, long operand, memory_order order); --long __ovld atomic_fetch_xor_explicit(volatile atomic_long *object, long operand, memory_order order, memory_scope scope); - ulong __ovld atomic_fetch_xor(volatile atomic_ulong *object, ulong operand); --ulong __ovld atomic_fetch_xor_explicit(volatile atomic_ulong *object, ulong operand, memory_order order); --ulong __ovld atomic_fetch_xor_explicit(volatile atomic_ulong *object, ulong operand, memory_order order, memory_scope scope); - long __ovld atomic_fetch_and(volatile atomic_long *object, long operand); --long __ovld atomic_fetch_and_explicit(volatile atomic_long *object, long operand, memory_order order); --long __ovld atomic_fetch_and_explicit(volatile atomic_long *object, long operand, memory_order order, memory_scope scope); - ulong __ovld atomic_fetch_and(volatile atomic_ulong *object, ulong operand); --ulong __ovld atomic_fetch_and_explicit(volatile atomic_ulong *object, ulong operand, memory_order order); --ulong __ovld atomic_fetch_and_explicit(volatile atomic_ulong *object, ulong operand, memory_order order, memory_scope scope); - long __ovld atomic_fetch_min(volatile atomic_long *object, long operand); --long __ovld atomic_fetch_min_explicit(volatile atomic_long *object, long operand, memory_order order); --long __ovld atomic_fetch_min_explicit(volatile atomic_long *object, long operand, memory_order order, memory_scope scope); - ulong __ovld atomic_fetch_min(volatile atomic_ulong *object, ulong operand); --ulong __ovld atomic_fetch_min_explicit(volatile atomic_ulong *object, ulong operand, memory_order order); --ulong __ovld atomic_fetch_min_explicit(volatile atomic_ulong *object, ulong operand, memory_order order, memory_scope scope); --ulong __ovld atomic_fetch_min(volatile atomic_ulong *object, long operand); --ulong __ovld atomic_fetch_min_explicit(volatile atomic_ulong *object, long operand, memory_order order); --ulong __ovld atomic_fetch_min_explicit(volatile atomic_ulong *object, long operand, memory_order order, memory_scope scope); - long __ovld atomic_fetch_max(volatile atomic_long *object, long operand); --long __ovld atomic_fetch_max_explicit(volatile atomic_long *object, long operand, memory_order order); --long __ovld atomic_fetch_max_explicit(volatile atomic_long *object, long operand, memory_order order, memory_scope scope); - ulong __ovld atomic_fetch_max(volatile atomic_ulong *object, ulong operand); --ulong __ovld atomic_fetch_max_explicit(volatile atomic_ulong *object, ulong operand, memory_order order); --ulong __ovld atomic_fetch_max_explicit(volatile atomic_ulong *object, ulong operand, memory_order order, memory_scope scope); --ulong __ovld atomic_fetch_max(volatile atomic_ulong *object, long operand); --ulong __ovld atomic_fetch_max_explicit(volatile atomic_ulong *object, long operand, memory_order order); --ulong __ovld atomic_fetch_max_explicit(volatile atomic_ulong *object, long operand, memory_order order, memory_scope scope); - #endif //defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) - - // OpenCL v2.0 s6.13.11.7.5: -@@ -13501,196 +13483,2239 @@ ulong __ovld atomic_fetch_max_explicit(volatile atomic_ulong *object, long opera - // or/xor/and/min/max: atomic type argument can be intptr_t/uintptr_t, value type argument can be intptr_t/uintptr_t. - - #if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) --uintptr_t __ovld atomic_fetch_add(volatile atomic_uintptr_t *object, ptrdiff_t operand); --uintptr_t __ovld atomic_fetch_add_explicit(volatile atomic_uintptr_t *object, ptrdiff_t operand, memory_order order); --uintptr_t __ovld atomic_fetch_add_explicit(volatile atomic_uintptr_t *object, ptrdiff_t operand, memory_order order, memory_scope scope); --uintptr_t __ovld atomic_fetch_sub(volatile atomic_uintptr_t *object, ptrdiff_t operand); --uintptr_t __ovld atomic_fetch_sub_explicit(volatile atomic_uintptr_t *object, ptrdiff_t operand, memory_order order); --uintptr_t __ovld atomic_fetch_sub_explicit(volatile atomic_uintptr_t *object, ptrdiff_t operand, memory_order order, memory_scope scope); -- --uintptr_t __ovld atomic_fetch_or(volatile atomic_uintptr_t *object, intptr_t operand); --uintptr_t __ovld atomic_fetch_or_explicit(volatile atomic_uintptr_t *object, intptr_t operand, memory_order order); --uintptr_t __ovld atomic_fetch_or_explicit(volatile atomic_uintptr_t *object, intptr_t operand, memory_order order, memory_scope scope); --uintptr_t __ovld atomic_fetch_xor(volatile atomic_uintptr_t *object, intptr_t operand); --uintptr_t __ovld atomic_fetch_xor_explicit(volatile atomic_uintptr_t *object, intptr_t operand, memory_order order); --uintptr_t __ovld atomic_fetch_xor_explicit(volatile atomic_uintptr_t *object, intptr_t operand, memory_order order, memory_scope scope); --uintptr_t __ovld atomic_fetch_and(volatile atomic_uintptr_t *object, intptr_t operand); --uintptr_t __ovld atomic_fetch_and_explicit(volatile atomic_uintptr_t *object, intptr_t operand, memory_order order); --uintptr_t __ovld atomic_fetch_and_explicit(volatile atomic_uintptr_t *object, intptr_t operand, memory_order order, memory_scope scope); --uintptr_t __ovld atomic_fetch_min(volatile atomic_uintptr_t *object, intptr_t opermax); --uintptr_t __ovld atomic_fetch_min_explicit(volatile atomic_uintptr_t *object, intptr_t opermax, memory_order minder); --uintptr_t __ovld atomic_fetch_min_explicit(volatile atomic_uintptr_t *object, intptr_t opermax, memory_order minder, memory_scope scope); --uintptr_t __ovld atomic_fetch_max(volatile atomic_uintptr_t *object, intptr_t opermax); --uintptr_t __ovld atomic_fetch_max_explicit(volatile atomic_uintptr_t *object, intptr_t opermax, memory_order minder); --uintptr_t __ovld atomic_fetch_max_explicit(volatile atomic_uintptr_t *object, intptr_t opermax, memory_order minder, memory_scope scope); -- --intptr_t __ovld atomic_fetch_or(volatile atomic_intptr_t *object, uintptr_t operand); --intptr_t __ovld atomic_fetch_or_explicit(volatile atomic_intptr_t *object, uintptr_t operand, memory_order order); --intptr_t __ovld atomic_fetch_or_explicit(volatile atomic_intptr_t *object, uintptr_t operand, memory_order order, memory_scope scope); --intptr_t __ovld atomic_fetch_xor(volatile atomic_intptr_t *object, uintptr_t operand); --intptr_t __ovld atomic_fetch_xor_explicit(volatile atomic_intptr_t *object, uintptr_t operand, memory_order order); --intptr_t __ovld atomic_fetch_xor_explicit(volatile atomic_intptr_t *object, uintptr_t operand, memory_order order, memory_scope scope); --intptr_t __ovld atomic_fetch_and(volatile atomic_intptr_t *object, uintptr_t operand); --intptr_t __ovld atomic_fetch_and_explicit(volatile atomic_intptr_t *object, uintptr_t operand, memory_order order); --intptr_t __ovld atomic_fetch_and_explicit(volatile atomic_intptr_t *object, uintptr_t operand, memory_order order, memory_scope scope); --intptr_t __ovld atomic_fetch_min(volatile atomic_intptr_t *object, uintptr_t opermax); --intptr_t __ovld atomic_fetch_min_explicit(volatile atomic_intptr_t *object, uintptr_t opermax, memory_order minder); --intptr_t __ovld atomic_fetch_min_explicit(volatile atomic_intptr_t *object, uintptr_t opermax, memory_order minder, memory_scope scope); --intptr_t __ovld atomic_fetch_max(volatile atomic_intptr_t *object, uintptr_t opermax); --intptr_t __ovld atomic_fetch_max_explicit(volatile atomic_intptr_t *object, uintptr_t opermax, memory_order minder); --intptr_t __ovld atomic_fetch_max_explicit(volatile atomic_intptr_t *object, uintptr_t opermax, memory_order minder, memory_scope scope); -+uintptr_t __ovld atomic_fetch_add(volatile atomic_uintptr_t *object, -+ ptrdiff_t operand); -+uintptr_t __ovld atomic_fetch_sub(volatile atomic_uintptr_t *object, -+ ptrdiff_t operand); -+ -+uintptr_t __ovld atomic_fetch_or(volatile atomic_uintptr_t *object, -+ intptr_t operand); -+uintptr_t __ovld atomic_fetch_xor(volatile atomic_uintptr_t *object, -+ intptr_t operand); -+uintptr_t __ovld atomic_fetch_and(volatile atomic_uintptr_t *object, -+ intptr_t operand); -+uintptr_t __ovld atomic_fetch_min(volatile atomic_uintptr_t *object, -+ intptr_t opermax); -+uintptr_t __ovld atomic_fetch_max(volatile atomic_uintptr_t *object, -+ intptr_t opermax); -+ -+intptr_t __ovld atomic_fetch_or(volatile atomic_intptr_t *object, -+ uintptr_t operand); -+intptr_t __ovld atomic_fetch_xor(volatile atomic_intptr_t *object, -+ uintptr_t operand); -+intptr_t __ovld atomic_fetch_and(volatile atomic_intptr_t *object, -+ uintptr_t operand); -+intptr_t __ovld atomic_fetch_min(volatile atomic_intptr_t *object, -+ uintptr_t opermax); -+intptr_t __ovld atomic_fetch_max(volatile atomic_intptr_t *object, -+ uintptr_t opermax); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // __opencl_c_generic_address_space -+ -+#if(__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+int __ovld atomic_fetch_add(volatile atomic_int __global *object, int operand); -+uint __ovld atomic_fetch_add(volatile atomic_uint __local *object, -+ uint operand); -+int __ovld atomic_fetch_sub(volatile atomic_int __global *object, int operand); -+int __ovld atomic_fetch_sub(volatile atomic_int __local *object, int operand); -+uint __ovld atomic_fetch_sub(volatile atomic_uint __local *object, -+ uint operand); -+uint __ovld atomic_fetch_sub(volatile atomic_uint __global *object, -+ uint operand); -+int __ovld atomic_fetch_or(volatile atomic_int __global *object, int operand); -+uint __ovld atomic_fetch_sub(volatile atomic_uint __local *object, -+ uint operand); -+uint __ovld atomic_fetch_or(volatile atomic_uint __global *object, -+ uint operand); -+uint __ovld atomic_fetch_or(volatile atomic_uint __local *object, uint operand); -+int __ovld atomic_fetch_xor(volatile atomic_int __global *object, int operand); -+int __ovld atomic_fetch_xor(volatile atomic_int __local *object, int operand); -+uint __ovld atomic_fetch_xor(volatile atomic_uint __global *object, -+ uint operand); -+uint __ovld atomic_fetch_xor(volatile atomic_uint __local *object, -+ uint operand); -+int __ovld atomic_fetch_and(volatile atomic_int __global *object, int operand); -+int __ovld atomic_fetch_and(volatile atomic_int __local *object, int operand); -+uint __ovld atomic_fetch_and(volatile atomic_uint __global *object, -+ uint operand); -+uint __ovld atomic_fetch_and(volatile atomic_uint __local *object, -+ uint operand); -+int __ovld atomic_fetch_min(volatile atomic_int __global *object, int operand); -+int __ovld atomic_fetch_min(volatile atomic_int __local *object, int operand); -+uint __ovld atomic_fetch_min(volatile atomic_uint __global *object, -+ uint operand); -+uint __ovld atomic_fetch_min(volatile atomic_uint __local *object, -+ uint operand); -+int __ovld atomic_fetch_max(volatile atomic_int __global *object, int operand); -+int __ovld atomic_fetch_max(volatile atomic_int __local *object, int operand); -+uint __ovld atomic_fetch_max(volatile atomic_uint __global *object, -+ uint operand); -+uint __ovld atomic_fetch_max(volatile atomic_uint __local *object, -+ uint operand); -+ -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+long __ovld atomic_fetch_add(volatile atomic_long __global *object, -+ long operand); -+long __ovld atomic_fetch_add(volatile atomic_long __local *object, -+ long operand); -+ulong __ovld atomic_fetch_add(volatile atomic_ulong __global *object, -+ ulong operand); -+ulong __ovld atomic_fetch_add(volatile atomic_ulong __local *object, -+ ulong operand); -+long __ovld atomic_fetch_sub(volatile atomic_long __global *object, -+ long operand); -+long __ovld atomic_fetch_sub(volatile atomic_long __local *object, -+ long operand); -+ulong __ovld atomic_fetch_sub(volatile atomic_ulong __global *object, -+ ulong operand); -+ulong __ovld atomic_fetch_sub(volatile atomic_ulong __local *object, -+ ulong operand); -+long __ovld atomic_fetch_or(volatile atomic_long __global *object, -+ long operand); -+long __ovld atomic_fetch_or(volatile atomic_long __local *object, long operand); -+ulong __ovld atomic_fetch_or(volatile atomic_ulong __global *object, -+ ulong operand); -+ulong __ovld atomic_fetch_or(volatile atomic_ulong __local *object, -+ ulong operand); -+long __ovld atomic_fetch_xor(volatile atomic_long __global *object, -+ long operand); -+long __ovld atomic_fetch_xor(volatile atomic_long __local *object, -+ long operand); -+ulong __ovld atomic_fetch_xor(volatile atomic_ulong __global *object, -+ ulong operand); -+ulong __ovld atomic_fetch_xor(volatile atomic_ulong __local *object, -+ ulong operand); -+long __ovld atomic_fetch_and(volatile atomic_long __global *object, -+ long operand); -+long __ovld atomic_fetch_and(volatile atomic_long __local *object, -+ long operand); -+ulong __ovld atomic_fetch_and(volatile atomic_ulong __global *object, -+ ulong operand); -+ulong __ovld atomic_fetch_and(volatile atomic_ulong __local *object, -+ ulong operand); -+long __ovld atomic_fetch_min(volatile atomic_long __global *object, -+ long operand); -+long __ovld atomic_fetch_min(volatile atomic_long __local *object, -+ long operand); -+ulong __ovld atomic_fetch_min(volatile atomic_ulong __global *object, -+ ulong operand); -+ulong __ovld atomic_fetch_min(volatile atomic_ulong __local *object, -+ ulong operand); -+long __ovld atomic_fetch_max(volatile atomic_long __global *object, -+ long operand); -+long __ovld atomic_fetch_max(volatile atomic_long __local *object, -+ long operand); -+ulong __ovld atomic_fetch_max(volatile atomic_ulong __global *object, -+ ulong operand); -+ulong __ovld atomic_fetch_max(volatile atomic_ulong __local *object, -+ ulong operand); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+ -+// OpenCL v2.0 s6.13.11.7.5: -+// add/sub: atomic type argument can be uintptr_t/intptr_t, value type argument -+// can be ptrdiff_t. or/xor/and/min/max: atomic type argument can be -+// intptr_t/uintptr_t, value type argument can be intptr_t/uintptr_t. -+ -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+uintptr_t __ovld atomic_fetch_add(volatile atomic_uintptr_t __global *object, -+ ptrdiff_t operand); -+uintptr_t __ovld atomic_fetch_add(volatile atomic_uintptr_t __local *object, -+ ptrdiff_t operand); -+uintptr_t __ovld atomic_fetch_sub(volatile atomic_uintptr_t __global *object, -+ ptrdiff_t operand); -+uintptr_t __ovld atomic_fetch_sub(volatile atomic_uintptr_t __local *object, -+ ptrdiff_t operand); -+ -+uintptr_t __ovld atomic_fetch_or(volatile atomic_uintptr_t __global *object, -+ intptr_t operand); -+uintptr_t __ovld atomic_fetch_or(volatile atomic_uintptr_t __local *object, -+ intptr_t operand); -+uintptr_t __ovld atomic_fetch_xor(volatile atomic_uintptr_t __global *object, -+ intptr_t operand); -+uintptr_t __ovld atomic_fetch_xor(volatile atomic_uintptr_t __local *object, -+ intptr_t operand); -+uintptr_t __ovld atomic_fetch_and(volatile atomic_uintptr_t __global *object, -+ intptr_t operand); -+uintptr_t __ovld atomic_fetch_and(volatile atomic_uintptr_t __local *object, -+ intptr_t operand); -+uintptr_t __ovld atomic_fetch_min(volatile atomic_uintptr_t __global *object, -+ intptr_t opermax); -+uintptr_t __ovld atomic_fetch_min(volatile atomic_uintptr_t __local *object, -+ intptr_t opermax); -+uintptr_t __ovld atomic_fetch_max(volatile atomic_uintptr_t __global *object, -+ intptr_t opermax); -+uintptr_t __ovld atomic_fetch_max(volatile atomic_uintptr_t __local *object, -+ intptr_t opermax); -+ -+intptr_t __ovld atomic_fetch_or(volatile atomic_intptr_t __global *object, -+ uintptr_t operand); -+intptr_t __ovld atomic_fetch_or(volatile atomic_intptr_t __local *object, -+ uintptr_t operand); -+intptr_t __ovld atomic_fetch_xor(volatile atomic_intptr_t __global *object, -+ uintptr_t operand); -+intptr_t __ovld atomic_fetch_xor(volatile atomic_intptr_t __local *object, -+ uintptr_t operand); -+intptr_t __ovld atomic_fetch_and(volatile atomic_intptr_t __global *object, -+ uintptr_t operand); -+intptr_t __ovld atomic_fetch_and(volatile atomic_intptr_t __local *object, -+ uintptr_t operand); -+intptr_t __ovld atomic_fetch_min(volatile atomic_intptr_t __global *object, -+ uintptr_t opermax); -+intptr_t __ovld atomic_fetch_min(volatile atomic_intptr_t __local *object, -+ uintptr_t opermax); -+intptr_t __ovld atomic_fetch_max(volatile atomic_intptr_t __global *object, -+ uintptr_t opermax); -+intptr_t __ovld atomic_fetch_max(volatile atomic_intptr_t __local *object, -+ uintptr_t opermax); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+ -+#endif // defined(__opencl_c_atomic_scope_device) && -+ // defined(__opencl_c_atomic_order_seq_cst) -+ -+#ifdef __opencl_c_generic_address_space -+#ifdef __opencl_c_atomic_scope_device -+int __ovld atomic_fetch_add_explicit(volatile atomic_int *object, int operand, -+ memory_order order); -+uint __ovld atomic_fetch_add_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order); -+int __ovld atomic_fetch_sub_explicit(volatile atomic_int *object, int operand, -+ memory_order order); -+uint __ovld atomic_fetch_sub_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order); -+int __ovld atomic_fetch_or_explicit(volatile atomic_int *object, int operand, -+ memory_order order); -+uint __ovld atomic_fetch_or_explicit(volatile atomic_uint *object, uint operand, -+ memory_order order); -+int __ovld atomic_fetch_xor_explicit(volatile atomic_int *object, int operand, -+ memory_order order); -+uint __ovld atomic_fetch_xor_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order); -+int __ovld atomic_fetch_and_explicit(volatile atomic_int *object, int operand, -+ memory_order order); -+uint __ovld atomic_fetch_and_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order); -+int __ovld atomic_fetch_min_explicit(volatile atomic_int *object, int operand, -+ memory_order order); -+uint __ovld atomic_fetch_min_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order); -+int __ovld atomic_fetch_max_explicit(volatile atomic_int *object, int operand, -+ memory_order order); -+uint __ovld atomic_fetch_max_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+int __ovld atomic_fetch_add_explicit(volatile atomic_int *object, int operand, -+ memory_order order, memory_scope scope); -+uint __ovld atomic_fetch_add_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_sub_explicit(volatile atomic_int *object, int operand, -+ memory_order order, memory_scope scope); -+uint __ovld atomic_fetch_sub_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_or_explicit(volatile atomic_int *object, int operand, -+ memory_order order, memory_scope scope); -+uint __ovld atomic_fetch_or_explicit(volatile atomic_uint *object, uint operand, -+ memory_order order, memory_scope scope); -+int __ovld atomic_fetch_xor_explicit(volatile atomic_int *object, int operand, -+ memory_order order, memory_scope scope); -+uint __ovld atomic_fetch_xor_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_and_explicit(volatile atomic_int *object, int operand, -+ memory_order order, memory_scope scope); -+uint __ovld atomic_fetch_and_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_min_explicit(volatile atomic_int *object, int operand, -+ memory_order order, memory_scope scope); -+uint __ovld atomic_fetch_min_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_max_explicit(volatile atomic_int *object, int operand, -+ memory_order order, memory_scope scope); -+uint __ovld atomic_fetch_max_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+ -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#ifdef __opencl_c_atomic_scope_device -+long __ovld atomic_fetch_add_explicit(volatile atomic_long *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_add_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_sub_explicit(volatile atomic_long *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_sub_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_or_explicit(volatile atomic_long *object, long operand, -+ memory_order order); -+ulong __ovld atomic_fetch_or_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_xor_explicit(volatile atomic_long *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_xor_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_and_explicit(volatile atomic_long *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_and_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_min_explicit(volatile atomic_long *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_min_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_max_explicit(volatile atomic_long *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_max_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+long __ovld atomic_fetch_add_explicit(volatile atomic_long *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_add_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_sub_explicit(volatile atomic_long *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_sub_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_or_explicit(volatile atomic_long *object, long operand, -+ memory_order order, memory_scope scope); -+ulong __ovld atomic_fetch_or_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_xor_explicit(volatile atomic_long *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_xor_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_and_explicit(volatile atomic_long *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_and_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_min_explicit(volatile atomic_long *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_min_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_max_explicit(volatile atomic_long *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_max_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+ -+// OpenCL v2.0 s6.13.11.7.5: -+// add/sub: atomic type argument can be uintptr_t/intptr_t, value type argument -+// can be ptrdiff_t. or/xor/and/min/max: atomic type argument can be -+// intptr_t/uintptr_t, value type argument can be intptr_t/uintptr_t. -+ -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#ifdef __opencl_c_atomic_scope_device -+uintptr_t __ovld atomic_fetch_add_explicit(volatile atomic_uintptr_t *object, -+ ptrdiff_t operand, -+ memory_order order); -+uintptr_t __ovld atomic_fetch_sub_explicit(volatile atomic_uintptr_t *object, -+ ptrdiff_t operand, -+ memory_order order); -+uintptr_t __ovld atomic_fetch_or_explicit(volatile atomic_uintptr_t *object, -+ intptr_t operand, memory_order order); -+uintptr_t __ovld atomic_fetch_xor_explicit(volatile atomic_uintptr_t *object, -+ intptr_t operand, -+ memory_order order); -+uintptr_t __ovld atomic_fetch_and_explicit(volatile atomic_uintptr_t *object, -+ intptr_t operand, -+ memory_order order); -+uintptr_t __ovld atomic_fetch_min_explicit(volatile atomic_uintptr_t *object, -+ intptr_t opermax, -+ memory_order minder); -+uintptr_t __ovld atomic_fetch_max_explicit(volatile atomic_uintptr_t *object, -+ intptr_t opermax, -+ memory_order minder); -+intptr_t __ovld atomic_fetch_or_explicit(volatile atomic_intptr_t *object, -+ uintptr_t operand, memory_order order); -+intptr_t __ovld atomic_fetch_xor_explicit(volatile atomic_intptr_t *object, -+ uintptr_t operand, -+ memory_order order); -+intptr_t __ovld atomic_fetch_and_explicit(volatile atomic_intptr_t *object, -+ uintptr_t operand, -+ memory_order order); -+intptr_t __ovld atomic_fetch_min_explicit(volatile atomic_intptr_t *object, -+ uintptr_t opermax, -+ memory_order minder); -+intptr_t __ovld atomic_fetch_max_explicit(volatile atomic_intptr_t *object, -+ uintptr_t opermax, -+ memory_order minder); -+#endif // __opencl_c_atomic_scope_device -+uintptr_t __ovld atomic_fetch_add_explicit(volatile atomic_uintptr_t *object, -+ ptrdiff_t operand, -+ memory_order order, -+ memory_scope scope); -+uintptr_t __ovld atomic_fetch_sub_explicit(volatile atomic_uintptr_t *object, -+ ptrdiff_t operand, -+ memory_order order, -+ memory_scope scope); -+ -+uintptr_t __ovld atomic_fetch_or_explicit(volatile atomic_uintptr_t *object, -+ intptr_t operand, memory_order order, -+ memory_scope scope); -+uintptr_t __ovld atomic_fetch_xor_explicit(volatile atomic_uintptr_t *object, -+ intptr_t operand, memory_order order, -+ memory_scope scope); -+uintptr_t __ovld atomic_fetch_and_explicit(volatile atomic_uintptr_t *object, -+ intptr_t operand, memory_order order, -+ memory_scope scope); -+uintptr_t __ovld atomic_fetch_min_explicit(volatile atomic_uintptr_t *object, -+ intptr_t opermax, -+ memory_order minder, -+ memory_scope scope); -+uintptr_t __ovld atomic_fetch_max_explicit(volatile atomic_uintptr_t *object, -+ intptr_t opermax, -+ memory_order minder, -+ memory_scope scope); -+ -+intptr_t __ovld atomic_fetch_or_explicit(volatile atomic_intptr_t *object, -+ uintptr_t operand, memory_order order, -+ memory_scope scope); -+intptr_t __ovld atomic_fetch_xor_explicit(volatile atomic_intptr_t *object, -+ uintptr_t operand, memory_order order, -+ memory_scope scope); -+intptr_t __ovld atomic_fetch_and_explicit(volatile atomic_intptr_t *object, -+ uintptr_t operand, memory_order order, -+ memory_scope scope); -+intptr_t __ovld atomic_fetch_min_explicit(volatile atomic_intptr_t *object, -+ uintptr_t opermax, -+ memory_order minder, -+ memory_scope scope); -+intptr_t __ovld atomic_fetch_max_explicit(volatile atomic_intptr_t *object, -+ uintptr_t opermax, -+ memory_order minder, -+ memory_scope scope); - #endif -+#endif // __opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+#ifdef __opencl_c_atomic_scope_device -+int __ovld atomic_fetch_add_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order); -+int __ovld atomic_fetch_add_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order); -+uint __ovld atomic_fetch_add_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order); -+uint __ovld atomic_fetch_add_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order); -+int __ovld atomic_fetch_sub_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order); -+int __ovld atomic_fetch_sub_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order); -+uint __ovld atomic_fetch_sub_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order); -+uint __ovld atomic_fetch_sub_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order); -+int __ovld atomic_fetch_or_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order); -+int __ovld atomic_fetch_or_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order); -+uint __ovld atomic_fetch_or_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order); -+uint __ovld atomic_fetch_or_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order); -+int __ovld atomic_fetch_xor_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order); -+int __ovld atomic_fetch_xor_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order); -+uint __ovld atomic_fetch_xor_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order); -+uint __ovld atomic_fetch_xor_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order); -+int __ovld atomic_fetch_and_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order); -+int __ovld atomic_fetch_and_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order); -+uint __ovld atomic_fetch_and_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order); -+uint __ovld atomic_fetch_and_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order); -+int __ovld atomic_fetch_min_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order); -+int __ovld atomic_fetch_min_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order); -+uint __ovld atomic_fetch_min_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order); -+uint __ovld atomic_fetch_min_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order); -+int __ovld atomic_fetch_max_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order); -+int __ovld atomic_fetch_max_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order); -+uint __ovld atomic_fetch_max_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order); -+uint __ovld atomic_fetch_max_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+int __ovld atomic_fetch_add_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_add_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_add_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_add_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_sub_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_sub_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_sub_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_sub_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_or_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_or_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_or_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_or_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_xor_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_xor_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_xor_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_xor_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_and_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_and_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_and_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_and_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_min_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_min_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_min_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_min_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_max_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_max_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_max_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_max_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+ -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#ifdef __opencl_c_atomic_scope_device -+long __ovld atomic_fetch_add_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order); -+long __ovld atomic_fetch_add_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_add_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order); -+ulong __ovld atomic_fetch_add_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_sub_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order); -+long __ovld atomic_fetch_sub_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_sub_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order); -+ulong __ovld atomic_fetch_sub_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_or_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order); -+long __ovld atomic_fetch_or_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_or_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order); -+ulong __ovld atomic_fetch_or_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_xor_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order); -+long __ovld atomic_fetch_xor_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_xor_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order); -+ulong __ovld atomic_fetch_xor_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_and_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order); -+long __ovld atomic_fetch_and_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_and_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order); -+ulong __ovld atomic_fetch_and_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_min_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order); -+long __ovld atomic_fetch_min_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_min_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order); -+ulong __ovld atomic_fetch_min_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_max_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order); -+long __ovld atomic_fetch_max_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_max_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order); -+ulong __ovld atomic_fetch_max_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+long __ovld atomic_fetch_add_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_add_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_add_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_add_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_sub_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_sub_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_sub_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_sub_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_or_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_or_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_or_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_or_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_xor_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_xor_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_xor_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_xor_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_and_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_and_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_and_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_and_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_min_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_min_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_min_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_min_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_max_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_max_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_max_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_max_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+ -+// OpenCL v2.0 s6.13.11.7.5: -+// add/sub: atomic type argument can be uintptr_t/intptr_t, value type argument -+// can be ptrdiff_t. or/xor/and/min/max: atomic type argument can be -+// intptr_t/uintptr_t, value type argument can be intptr_t/uintptr_t. -+ -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#ifdef __opencl_c_atomic_scope_device -+uintptr_t __ovld -+atomic_fetch_add_explicit(volatile atomic_uintptr_t __global *object, -+ ptrdiff_t operand, memory_order order); -+uintptr_t __ovld -+atomic_fetch_add_explicit(volatile atomic_uintptr_t __local *object, -+ ptrdiff_t operand, memory_order order); -+uintptr_t __ovld -+atomic_fetch_sub_explicit(volatile atomic_uintptr_t __global *object, -+ ptrdiff_t operand, memory_order order); -+uintptr_t __ovld -+atomic_fetch_sub_explicit(volatile atomic_uintptr_t __local *object, -+ ptrdiff_t operand, memory_order order); -+uintptr_t __ovld -+atomic_fetch_or_explicit(volatile atomic_uintptr_t __global *object, -+ intptr_t operand, memory_order order); -+uintptr_t __ovld -+atomic_fetch_or_explicit(volatile atomic_uintptr_t __local *object, -+ intptr_t operand, memory_order order); -+uintptr_t __ovld -+atomic_fetch_xor_explicit(volatile atomic_uintptr_t __global *object, -+ intptr_t operand, memory_order order); -+uintptr_t __ovld -+atomic_fetch_xor_explicit(volatile atomic_uintptr_t __local *object, -+ intptr_t operand, memory_order order); -+uintptr_t __ovld -+atomic_fetch_and_explicit(volatile atomic_uintptr_t __global *object, -+ intptr_t operand, memory_order order); -+uintptr_t __ovld -+atomic_fetch_and_explicit(volatile atomic_uintptr_t __local *object, -+ intptr_t operand, memory_order order); -+uintptr_t __ovld -+atomic_fetch_min_explicit(volatile atomic_uintptr_t __global *object, -+ intptr_t opermax, memory_order minder); -+uintptr_t __ovld -+atomic_fetch_min_explicit(volatile atomic_uintptr_t __local *object, -+ intptr_t opermax, memory_order minder); -+uintptr_t __ovld -+atomic_fetch_max_explicit(volatile atomic_uintptr_t __global *object, -+ intptr_t opermax, memory_order minder); -+uintptr_t __ovld -+atomic_fetch_max_explicit(volatile atomic_uintptr_t __local *object, -+ intptr_t opermax, memory_order minder); -+intptr_t __ovld -+atomic_fetch_or_explicit(volatile atomic_intptr_t __global *object, -+ uintptr_t operand, memory_order order); -+intptr_t __ovld -+atomic_fetch_or_explicit(volatile atomic_intptr_t __local *object, -+ uintptr_t operand, memory_order order); -+intptr_t __ovld -+atomic_fetch_xor_explicit(volatile atomic_intptr_t __global *object, -+ uintptr_t operand, memory_order order); -+intptr_t __ovld -+atomic_fetch_xor_explicit(volatile atomic_intptr_t __local *object, -+ uintptr_t operand, memory_order order); -+intptr_t __ovld -+atomic_fetch_and_explicit(volatile atomic_intptr_t __global *object, -+ uintptr_t operand, memory_order order); -+intptr_t __ovld -+atomic_fetch_and_explicit(volatile atomic_intptr_t __local *object, -+ uintptr_t operand, memory_order order); -+intptr_t __ovld -+atomic_fetch_min_explicit(volatile atomic_intptr_t __global *object, -+ uintptr_t opermax, memory_order minder); -+intptr_t __ovld -+atomic_fetch_min_explicit(volatile atomic_intptr_t __local *object, -+ uintptr_t opermax, memory_order minder); -+intptr_t __ovld -+atomic_fetch_max_explicit(volatile atomic_intptr_t __global *object, -+ uintptr_t opermax, memory_order minder); -+intptr_t __ovld -+atomic_fetch_max_explicit(volatile atomic_intptr_t __local *object, -+ uintptr_t opermax, memory_order minder); -+#endif // __opencl_c_atomic_scope_device -+uintptr_t __ovld atomic_fetch_add_explicit( -+ volatile atomic_uintptr_t __global *object, ptrdiff_t operand, -+ memory_order order, memory_scope scope); -+uintptr_t __ovld atomic_fetch_add_explicit( -+ volatile atomic_uintptr_t __local *object, ptrdiff_t operand, -+ memory_order order, memory_scope scope); -+uintptr_t __ovld atomic_fetch_sub_explicit( -+ volatile atomic_uintptr_t __global *object, ptrdiff_t operand, -+ memory_order order, memory_scope scope); -+uintptr_t __ovld atomic_fetch_sub_explicit( -+ volatile atomic_uintptr_t __local *object, ptrdiff_t operand, -+ memory_order order, memory_scope scope); -+ -+uintptr_t __ovld atomic_fetch_or_explicit( -+ volatile atomic_uintptr_t __global *object, intptr_t operand, -+ memory_order order, memory_scope scope); -+uintptr_t __ovld atomic_fetch_or_explicit( -+ volatile atomic_uintptr_t __local *object, intptr_t operand, -+ memory_order order, memory_scope scope); -+uintptr_t __ovld atomic_fetch_xor_explicit( -+ volatile atomic_uintptr_t __global *object, intptr_t operand, -+ memory_order order, memory_scope scope); -+uintptr_t __ovld atomic_fetch_xor_explicit( -+ volatile atomic_uintptr_t __local *object, intptr_t operand, -+ memory_order order, memory_scope scope); -+uintptr_t __ovld atomic_fetch_and_explicit( -+ volatile atomic_uintptr_t __global *object, intptr_t operand, -+ memory_order order, memory_scope scope); -+uintptr_t __ovld atomic_fetch_and_explicit( -+ volatile atomic_uintptr_t __local *object, intptr_t operand, -+ memory_order order, memory_scope scope); -+uintptr_t __ovld atomic_fetch_min_explicit( -+ volatile atomic_uintptr_t __global *object, intptr_t opermax, -+ memory_order minder, memory_scope scope); -+uintptr_t __ovld atomic_fetch_min_explicit( -+ volatile atomic_uintptr_t __local *object, intptr_t opermax, -+ memory_order minder, memory_scope scope); -+uintptr_t __ovld atomic_fetch_max_explicit( -+ volatile atomic_uintptr_t __global *object, intptr_t opermax, -+ memory_order minder, memory_scope scope); -+uintptr_t __ovld atomic_fetch_max_explicit( -+ volatile atomic_uintptr_t __local *object, intptr_t opermax, -+ memory_order minder, memory_scope scope); -+ -+intptr_t __ovld atomic_fetch_or_explicit( -+ volatile atomic_intptr_t __global *object, uintptr_t operand, -+ memory_order order, memory_scope scope); -+intptr_t __ovld atomic_fetch_or_explicit( -+ volatile atomic_intptr_t __local *object, uintptr_t operand, -+ memory_order order, memory_scope scope); -+intptr_t __ovld atomic_fetch_xor_explicit( -+ volatile atomic_intptr_t __global *object, uintptr_t operand, -+ memory_order order, memory_scope scope); -+intptr_t __ovld atomic_fetch_xor_explicit( -+ volatile atomic_intptr_t __local *object, uintptr_t operand, -+ memory_order order, memory_scope scope); -+intptr_t __ovld atomic_fetch_and_explicit( -+ volatile atomic_intptr_t __global *object, uintptr_t operand, -+ memory_order order, memory_scope scope); -+intptr_t __ovld atomic_fetch_and_explicit( -+ volatile atomic_intptr_t __local *object, uintptr_t operand, -+ memory_order order, memory_scope scope); -+intptr_t __ovld atomic_fetch_min_explicit( -+ volatile atomic_intptr_t __global *object, uintptr_t opermax, -+ memory_order minder, memory_scope scope); -+intptr_t __ovld atomic_fetch_min_explicit( -+ volatile atomic_intptr_t __local *object, uintptr_t opermax, -+ memory_order minder, memory_scope scope); -+intptr_t __ovld atomic_fetch_max_explicit( -+ volatile atomic_intptr_t __global *object, uintptr_t opermax, -+ memory_order minder, memory_scope scope); -+intptr_t __ovld atomic_fetch_max_explicit( -+ volatile atomic_intptr_t __local *object, uintptr_t opermax, -+ memory_order minder, memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) - - // atomic_store() - -+#if defined(__opencl_c_atomic_scope_device) && \ -+ defined(__opencl_c_atomic_order_seq_cst) -+ -+#ifdef __opencl_c_generic_address_space - void __ovld atomic_store(volatile atomic_int *object, int desired); --void __ovld atomic_store_explicit(volatile atomic_int *object, int desired, memory_order order); --void __ovld atomic_store_explicit(volatile atomic_int *object, int desired, memory_order order, memory_scope scope); - void __ovld atomic_store(volatile atomic_uint *object, uint desired); --void __ovld atomic_store_explicit(volatile atomic_uint *object, uint desired, memory_order order); --void __ovld atomic_store_explicit(volatile atomic_uint *object, uint desired, memory_order order, memory_scope scope); - void __ovld atomic_store(volatile atomic_float *object, float desired); --void __ovld atomic_store_explicit(volatile atomic_float *object, float desired, memory_order order); --void __ovld atomic_store_explicit(volatile atomic_float *object, float desired, memory_order order, memory_scope scope); - #if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - void __ovld atomic_store(volatile atomic_double *object, double desired); --void __ovld atomic_store_explicit(volatile atomic_double *object, double desired, memory_order order); --void __ovld atomic_store_explicit(volatile atomic_double *object, double desired, memory_order order, memory_scope scope); --#endif //cl_khr_fp64 -+#endif - void __ovld atomic_store(volatile atomic_long *object, long desired); --void __ovld atomic_store_explicit(volatile atomic_long *object, long desired, memory_order order); --void __ovld atomic_store_explicit(volatile atomic_long *object, long desired, memory_order order, memory_scope scope); - void __ovld atomic_store(volatile atomic_ulong *object, ulong desired); --void __ovld atomic_store_explicit(volatile atomic_ulong *object, ulong desired, memory_order order); --void __ovld atomic_store_explicit(volatile atomic_ulong *object, ulong desired, memory_order order, memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // __opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+void __ovld atomic_store(volatile atomic_int __global *object, int desired); -+void __ovld atomic_store(volatile atomic_int __local *object, int desired); -+void __ovld atomic_store(volatile atomic_uint __global *object, uint desired); -+void __ovld atomic_store(volatile atomic_uint __local *object, uint desired); -+void __ovld atomic_store(volatile atomic_float __global *object, float desired); -+void __ovld atomic_store(volatile atomic_float __local *object, float desired); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+void __ovld atomic_store(volatile atomic_double __global *object, -+ double desired); -+void __ovld atomic_store(volatile atomic_double __local *object, -+ double desired); - #endif -+void __ovld atomic_store(volatile atomic_long __global *object, long desired); -+void __ovld atomic_store(volatile atomic_long __local *object, long desired); -+void __ovld atomic_store(volatile atomic_ulong __global *object, ulong desired); -+void __ovld atomic_store(volatile atomic_ulong __local *object, ulong desired); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+ -+#endif // defined(__opencl_c_atomic_scope_device) && -+ // defined(__opencl_c_atomic_order_seq_cst) -+ -+#ifdef __opencl_c_generic_address_space -+#ifdef __opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_int *object, int desired, -+ memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_uint *object, uint desired, -+ memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_float *object, float desired, -+ memory_order order); -+#endif // __opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_int *object, int desired, -+ memory_order order, memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_uint *object, uint desired, -+ memory_order order, memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_float *object, float desired, -+ memory_order order, memory_scope scope); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_double *object, -+ double desired, memory_order order); -+#endif //__opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_double *object, -+ double desired, memory_order order, -+ memory_scope scope); -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_long *object, long desired, -+ memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_ulong *object, ulong desired, -+ memory_order order); -+#endif //__opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_long *object, long desired, -+ memory_order order, memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_ulong *object, ulong desired, -+ memory_order order, memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // __opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+#ifdef __opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_int __global *object, -+ int desired, memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_int __local *object, -+ int desired, memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_uint __global *object, -+ uint desired, memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_uint __local *object, -+ uint desired, memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_float __global *object, -+ float desired, memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_float __local *object, -+ float desired, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_int __global *object, -+ int desired, memory_order order, -+ memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_int __local *object, -+ int desired, memory_order order, -+ memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_uint __global *object, -+ uint desired, memory_order order, -+ memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_uint __local *object, -+ uint desired, memory_order order, -+ memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_float __global *object, -+ float desired, memory_order order, -+ memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_float __local *object, -+ float desired, memory_order order, -+ memory_scope scope); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_double __global *object, -+ double desired, memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_double __local *object, -+ double desired, memory_order order); -+#endif //__opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_double __global *object, -+ double desired, memory_order order, -+ memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_double __local *object, -+ double desired, memory_order order, -+ memory_scope scope); -+#endif // cl_khr_fp64 -+#ifdef __opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_long __global *object, -+ long desired, memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_long __local *object, -+ long desired, memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_ulong __global *object, -+ ulong desired, memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_ulong __local *object, -+ ulong desired, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_long __global *object, -+ long desired, memory_order order, -+ memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_long __local *object, -+ long desired, memory_order order, -+ memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_ulong __global *object, -+ ulong desired, memory_order order, -+ memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_ulong __local *object, -+ ulong desired, memory_order order, -+ memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) - - // atomic_load() -- -+#ifdef __opencl_c_generic_address_space -+#if defined(__opencl_c_atomic_scope_device) && \ -+ defined(__opencl_c_atomic_order_seq_cst) - int __ovld atomic_load(volatile atomic_int *object); --int __ovld atomic_load_explicit(volatile atomic_int *object, memory_order order); --int __ovld atomic_load_explicit(volatile atomic_int *object, memory_order order, memory_scope scope); - uint __ovld atomic_load(volatile atomic_uint *object); --uint __ovld atomic_load_explicit(volatile atomic_uint *object, memory_order order); --uint __ovld atomic_load_explicit(volatile atomic_uint *object, memory_order order, memory_scope scope); - float __ovld atomic_load(volatile atomic_float *object); --float __ovld atomic_load_explicit(volatile atomic_float *object, memory_order order); --float __ovld atomic_load_explicit(volatile atomic_float *object, memory_order order, memory_scope scope); - #if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld atomic_load(volatile atomic_double *object); --double __ovld atomic_load_explicit(volatile atomic_double *object, memory_order order); --double __ovld atomic_load_explicit(volatile atomic_double *object, memory_order order, memory_scope scope); --#endif //cl_khr_fp64 -+#endif // cl_khr_fp64 - long __ovld atomic_load(volatile atomic_long *object); --long __ovld atomic_load_explicit(volatile atomic_long *object, memory_order order); --long __ovld atomic_load_explicit(volatile atomic_long *object, memory_order order, memory_scope scope); - ulong __ovld atomic_load(volatile atomic_ulong *object); --ulong __ovld atomic_load_explicit(volatile atomic_ulong *object, memory_order order); --ulong __ovld atomic_load_explicit(volatile atomic_ulong *object, memory_order order, memory_scope scope); --#endif -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // defined(__opencl_c_atomic_scope_device) && -+ // defined(__opencl_c_atomic_order_seq_cst) -+#endif //__opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+#if defined(__opencl_c_atomic_scope_device) && \ -+ defined(__opencl_c_atomic_order_seq_cst) -+int __ovld atomic_load(volatile atomic_int __global *object); -+int __ovld atomic_load(volatile atomic_int __local *object); -+uint __ovld atomic_load(volatile atomic_uint __global *object); -+uint __ovld atomic_load(volatile atomic_uint __local *object); -+float __ovld atomic_load(volatile atomic_float __global *object); -+float __ovld atomic_load(volatile atomic_float __local *object); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+double __ovld atomic_load(volatile atomic_double __global *object); -+double __ovld atomic_load(volatile atomic_double __local *object); -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+long __ovld atomic_load(volatile atomic_long __global *object); -+long __ovld atomic_load(volatile atomic_long __local *object); -+ulong __ovld atomic_load(volatile atomic_ulong __global *object); -+ulong __ovld atomic_load(volatile atomic_ulong __local *object); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // defined(__opencl_c_atomic_scope_device) && -+ // defined(__opencl_c_atomic_order_seq_cst) -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+ -+#ifdef __opencl_c_generic_address_space -+#ifdef __opencl_c_atomic_scope_device -+int __ovld atomic_load_explicit(volatile atomic_int *object, -+ memory_order order); -+uint __ovld atomic_load_explicit(volatile atomic_uint *object, -+ memory_order order); -+float __ovld atomic_load_explicit(volatile atomic_float *object, -+ memory_order order); -+#endif // __opencl_c_atomic_scope_device -+int __ovld atomic_load_explicit(volatile atomic_int *object, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_load_explicit(volatile atomic_uint *object, -+ memory_order order, memory_scope scope); -+float __ovld atomic_load_explicit(volatile atomic_float *object, -+ memory_order order, memory_scope scope); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+double __ovld atomic_load_explicit(volatile atomic_double *object, -+ memory_order order); -+#endif // __opencl_c_atomic_scope_device -+double __ovld atomic_load_explicit(volatile atomic_double *object, -+ memory_order order, memory_scope scope); -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+long __ovld atomic_load_explicit(volatile atomic_long *object, -+ memory_order order); -+ulong __ovld atomic_load_explicit(volatile atomic_ulong *object, -+ memory_order order); -+#endif //__opencl_c_atomic_scope_device -+long __ovld atomic_load_explicit(volatile atomic_long *object, -+ memory_order order, memory_scope scope); -+ulong __ovld atomic_load_explicit(volatile atomic_ulong *object, -+ memory_order order, memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // __opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+#ifdef __opencl_c_atomic_scope_device -+int __ovld atomic_load_explicit(volatile atomic_int __global *object, -+ memory_order order); -+int __ovld atomic_load_explicit(volatile atomic_int __local *object, -+ memory_order order); -+uint __ovld atomic_load_explicit(volatile atomic_uint __global *object, -+ memory_order order); -+uint __ovld atomic_load_explicit(volatile atomic_uint __local *object, -+ memory_order order); -+float __ovld atomic_load_explicit(volatile atomic_float __global *object, -+ memory_order order); -+float __ovld atomic_load_explicit(volatile atomic_float __local *object, -+ memory_order order); -+#endif // __opencl_c_atomic_scope_device -+int __ovld atomic_load_explicit(volatile atomic_int __global *object, -+ memory_order order, memory_scope scope); -+int __ovld atomic_load_explicit(volatile atomic_int __local *object, -+ memory_order order, memory_scope scope); -+uint __ovld atomic_load_explicit(volatile atomic_uint __global *object, -+ memory_order order, memory_scope scope); -+uint __ovld atomic_load_explicit(volatile atomic_uint __local *object, -+ memory_order order, memory_scope scope); -+float __ovld atomic_load_explicit(volatile atomic_float __global *object, -+ memory_order order, memory_scope scope); -+float __ovld atomic_load_explicit(volatile atomic_float __local *object, -+ memory_order order, memory_scope scope); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+double __ovld atomic_load_explicit(volatile atomic_double __global *object, -+ memory_order order); -+double __ovld atomic_load_explicit(volatile atomic_double __local *object, -+ memory_order order); -+#endif // __opencl_c_atomic_scope_device -+double __ovld atomic_load_explicit(volatile atomic_double __global *object, -+ memory_order order, memory_scope scope); -+double __ovld atomic_load_explicit(volatile atomic_double __local *object, -+ memory_order order, memory_scope scope); -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+long __ovld atomic_load_explicit(volatile atomic_long __global *object, -+ memory_order order); -+long __ovld atomic_load_explicit(volatile atomic_long __local *object, -+ memory_order order); -+ulong __ovld atomic_load_explicit(volatile atomic_ulong __global *object, -+ memory_order order); -+ulong __ovld atomic_load_explicit(volatile atomic_ulong __local *object, -+ memory_order order); -+#endif // __opencl_c_atomic_scope_device -+long __ovld atomic_load_explicit(volatile atomic_long __global *object, -+ memory_order order, memory_scope scope); -+long __ovld atomic_load_explicit(volatile atomic_long __local *object, -+ memory_order order, memory_scope scope); -+ulong __ovld atomic_load_explicit(volatile atomic_ulong __global *object, -+ memory_order order, memory_scope scope); -+ulong __ovld atomic_load_explicit(volatile atomic_ulong __local *object, -+ memory_order order, memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) - - // atomic_exchange() -- -+#ifdef __opencl_c_generic_address_space -+#if defined(__opencl_c_atomic_scope_device) && \ -+ defined(__opencl_c_atomic_order_seq_cst) - int __ovld atomic_exchange(volatile atomic_int *object, int desired); --int __ovld atomic_exchange_explicit(volatile atomic_int *object, int desired, memory_order order); --int __ovld atomic_exchange_explicit(volatile atomic_int *object, int desired, memory_order order, memory_scope scope); - uint __ovld atomic_exchange(volatile atomic_uint *object, uint desired); --uint __ovld atomic_exchange_explicit(volatile atomic_uint *object, uint desired, memory_order order); --uint __ovld atomic_exchange_explicit(volatile atomic_uint *object, uint desired, memory_order order, memory_scope scope); - float __ovld atomic_exchange(volatile atomic_float *object, float desired); --float __ovld atomic_exchange_explicit(volatile atomic_float *object, float desired, memory_order order); --float __ovld atomic_exchange_explicit(volatile atomic_float *object, float desired, memory_order order, memory_scope scope); - #if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld atomic_exchange(volatile atomic_double *object, double desired); --double __ovld atomic_exchange_explicit(volatile atomic_double *object, double desired, memory_order order); --double __ovld atomic_exchange_explicit(volatile atomic_double *object, double desired, memory_order order, memory_scope scope); --#endif //cl_khr_fp64 -+#endif - long __ovld atomic_exchange(volatile atomic_long *object, long desired); --long __ovld atomic_exchange_explicit(volatile atomic_long *object, long desired, memory_order order); --long __ovld atomic_exchange_explicit(volatile atomic_long *object, long desired, memory_order order, memory_scope scope); - ulong __ovld atomic_exchange(volatile atomic_ulong *object, ulong desired); --ulong __ovld atomic_exchange_explicit(volatile atomic_ulong *object, ulong desired, memory_order order); --ulong __ovld atomic_exchange_explicit(volatile atomic_ulong *object, ulong desired, memory_order order, memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // defined(__opencl_c_atomic_scope_device) && -+ // defined(__opencl_c_atomic_order_seq_cst) -+#endif // __opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+#if defined(__opencl_c_atomic_scope_device) && \ -+ defined(__opencl_c_atomic_order_seq_cst) -+int __ovld atomic_exchange(volatile atomic_int __global *object, int desired); -+int __ovld atomic_exchange(volatile atomic_int __local *object, int desired); -+uint __ovld atomic_exchange(volatile atomic_uint __global *object, -+ uint desired); -+uint __ovld atomic_exchange(volatile atomic_uint __local *object, uint desired); -+float __ovld atomic_exchange(volatile atomic_float __global *object, -+ float desired); -+float __ovld atomic_exchange(volatile atomic_float __local *object, -+ float desired); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+double __ovld atomic_exchange(volatile atomic_double __global *object, -+ double desired); -+double __ovld atomic_exchange(volatile atomic_double __local *object, -+ double desired); - #endif -+long __ovld atomic_exchange(volatile atomic_long __global *object, -+ long desired); -+long __ovld atomic_exchange(volatile atomic_long __local *object, long desired); -+ulong __ovld atomic_exchange(volatile atomic_ulong __global *object, -+ ulong desired); -+ulong __ovld atomic_exchange(volatile atomic_ulong __local *object, -+ ulong desired); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // defined(__opencl_c_atomic_scope_device) && -+ // defined(__opencl_c_atomic_order_seq_cst) -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+ -+#ifdef __opencl_c_generic_address_space -+#ifdef __opencl_c_atomic_scope_device -+int __ovld atomic_exchange_explicit(volatile atomic_int *object, int desired, -+ memory_order order); -+uint __ovld atomic_exchange_explicit(volatile atomic_uint *object, uint desired, -+ memory_order order); -+float __ovld atomic_exchange_explicit(volatile atomic_float *object, -+ float desired, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+int __ovld atomic_exchange_explicit(volatile atomic_int *object, int desired, -+ memory_order order, memory_scope scope); -+uint __ovld atomic_exchange_explicit(volatile atomic_uint *object, uint desired, -+ memory_order order, memory_scope scope); -+float __ovld atomic_exchange_explicit(volatile atomic_float *object, -+ float desired, memory_order order, -+ memory_scope scope); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+double __ovld atomic_exchange_explicit(volatile atomic_double *object, -+ double desired, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+double __ovld atomic_exchange_explicit(volatile atomic_double *object, -+ double desired, memory_order order, -+ memory_scope scope); -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+long __ovld atomic_exchange_explicit(volatile atomic_long *object, long desired, -+ memory_order order); -+ulong __ovld atomic_exchange_explicit(volatile atomic_ulong *object, -+ ulong desired, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+long __ovld atomic_exchange_explicit(volatile atomic_long *object, long desired, -+ memory_order order, memory_scope scope); -+ulong __ovld atomic_exchange_explicit(volatile atomic_ulong *object, -+ ulong desired, memory_order order, -+ memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // __opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+#ifdef __opencl_c_atomic_scope_device -+int __ovld atomic_exchange_explicit(volatile atomic_int __global *object, -+ int desired, memory_order order); -+int __ovld atomic_exchange_explicit(volatile atomic_int __local *object, -+ int desired, memory_order order); -+uint __ovld atomic_exchange_explicit(volatile atomic_uint __global *object, -+ uint desired, memory_order order); -+uint __ovld atomic_exchange_explicit(volatile atomic_uint __local *object, -+ uint desired, memory_order order); -+float __ovld atomic_exchange_explicit(volatile atomic_float __global *object, -+ float desired, memory_order order); -+float __ovld atomic_exchange_explicit(volatile atomic_float __local *object, -+ float desired, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+int __ovld atomic_exchange_explicit(volatile atomic_int __global *object, -+ int desired, memory_order order, -+ memory_scope scope); -+int __ovld atomic_exchange_explicit(volatile atomic_int __local *object, -+ int desired, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_exchange_explicit(volatile atomic_uint __global *object, -+ uint desired, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_exchange_explicit(volatile atomic_uint __local *object, -+ uint desired, memory_order order, -+ memory_scope scope); -+float __ovld atomic_exchange_explicit(volatile atomic_float __global *object, -+ float desired, memory_order order, -+ memory_scope scope); -+float __ovld atomic_exchange_explicit(volatile atomic_float __local *object, -+ float desired, memory_order order, -+ memory_scope scope); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+double __ovld atomic_exchange_explicit(volatile atomic_double __global *object, -+ double desired, memory_order order); -+double __ovld atomic_exchange_explicit(volatile atomic_double __local *object, -+ double desired, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+double __ovld atomic_exchange_explicit(volatile atomic_double __global *object, -+ double desired, memory_order order, -+ memory_scope scope); -+double __ovld atomic_exchange_explicit(volatile atomic_double __local *object, -+ double desired, memory_order order, -+ memory_scope scope); -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+long __ovld atomic_exchange_explicit(volatile atomic_long __global *object, -+ long desired, memory_order order); -+long __ovld atomic_exchange_explicit(volatile atomic_long __local *object, -+ long desired, memory_order order); -+ulong __ovld atomic_exchange_explicit(volatile atomic_ulong __global *object, -+ ulong desired, memory_order order); -+ulong __ovld atomic_exchange_explicit(volatile atomic_ulong __local *object, -+ ulong desired, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+long __ovld atomic_exchange_explicit(volatile atomic_long __global *object, -+ long desired, memory_order order, -+ memory_scope scope); -+long __ovld atomic_exchange_explicit(volatile atomic_long __local *object, -+ long desired, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_exchange_explicit(volatile atomic_ulong __global *object, -+ ulong desired, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_exchange_explicit(volatile atomic_ulong __local *object, -+ ulong desired, memory_order order, -+ memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) - - // atomic_compare_exchange_strong() and atomic_compare_exchange_weak() -- --bool __ovld atomic_compare_exchange_strong(volatile atomic_int *object, int *expected, int desired); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_int *object, int *expected, -- int desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_int *object, int *expected, -- int desired, memory_order success, memory_order failure, memory_scope scope); --bool __ovld atomic_compare_exchange_strong(volatile atomic_uint *object, uint *expected, uint desired); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_uint *object, uint *expected, -- uint desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_uint *object, uint *expected, -- uint desired, memory_order success, memory_order failure, memory_scope scope); --bool __ovld atomic_compare_exchange_weak(volatile atomic_int *object, int *expected, int desired); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_int *object, int *expected, -- int desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_int *object, int *expected, -- int desired, memory_order success, memory_order failure, memory_scope scope); --bool __ovld atomic_compare_exchange_weak(volatile atomic_uint *object, uint *expected, uint desired); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_uint *object, uint *expected, -- uint desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_uint *object, uint *expected, -- uint desired, memory_order success, memory_order failure, memory_scope scope); --bool __ovld atomic_compare_exchange_strong(volatile atomic_float *object, float *expected, float desired); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_float *object, float *expected, -- float desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_float *object, float *expected, -- float desired, memory_order success, memory_order failure, memory_scope scope); --bool __ovld atomic_compare_exchange_weak(volatile atomic_float *object, float *expected, float desired); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_float *object, float *expected, -- float desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_float *object, float *expected, -- float desired, memory_order success, memory_order failure, memory_scope scope); -+#ifdef __opencl_c_generic_address_space -+#if defined(__opencl_c_atomic_scope_device) && \ -+ defined(__opencl_c_atomic_order_seq_cst) -+bool __ovld atomic_compare_exchange_strong(volatile atomic_int *object, -+ int *expected, int desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_uint *object, -+ uint *expected, uint desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_int *object, -+ int *expected, int desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_uint *object, -+ uint *expected, uint desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_float *object, -+ float *expected, float desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_float *object, -+ float *expected, float desired); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+bool __ovld atomic_compare_exchange_strong(volatile atomic_double *object, -+ double *expected, double desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_double *object, -+ double *expected, double desired); -+#endif -+bool __ovld atomic_compare_exchange_strong(volatile atomic_long *object, -+ long *expected, long desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_long *object, -+ long *expected, long desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_ulong *object, -+ ulong *expected, ulong desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_ulong *object, -+ ulong *expected, ulong desired); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // defined(__opencl_c_atomic_scope_device) && -+ // defined(__opencl_c_atomic_order_seq_cst) -+#endif // __opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+#if defined(__opencl_c_atomic_scope_device) && \ -+ defined(__opencl_c_atomic_order_seq_cst) -+bool __ovld atomic_compare_exchange_strong(volatile atomic_int __global *object, -+ int __global *expected, int desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_int __local *object, -+ int __global *expected, int desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_int __global *object, -+ int __local *expected, int desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_int __local *object, -+ int __local *expected, int desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_int __global *object, -+ int __private *expected, -+ int desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_int __local *object, -+ int __private *expected, -+ int desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_uint __global *object, -+ uint __global *expected, uint desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_uint __local *object, -+ uint __global *expected, -+ uint desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_uint __global *object, -+ uint __local *expected, uint desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_uint __local *object, -+ uint __local *expected, -+ uint desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_uint __global *object, -+ uint __private *expected, uint desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_uint __local *object, -+ uint __private *expected, -+ uint desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_int __global *object, -+ int __global *expected, int desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_int __local *object, -+ int __global *expected, int desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_int __global *object, -+ int __local *expected, int desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_int __local *object, -+ int __local *expected, int desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_int __global *object, -+ int __private *expected, int desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_int __local *object, -+ int __private *expected, int desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_uint __global *object, -+ uint __global *expected, uint desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_uint __local *object, -+ uint __global *expected, uint desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_uint __global *object, -+ uint __local *expected, uint desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_uint __local *object, -+ uint __local *expected, uint desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_uint __global *object, -+ uint __private *expected, -+ uint desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_uint __local *object, -+ uint __private *expected, -+ uint desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_float __global *object, -+ float __global *expected, float desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_float __local *object, -+ float __global *expected, float desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_float __global *object, -+ float __local *expected, float desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_float __local *object, -+ float __local *expected, float desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_float __global *object, -+ float __private *expected, float desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_float __local *object, -+ float __private *expected, float desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_float __global *object, -+ float __global *expected, -+ float desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_float __local *object, -+ float __global *expected, -+ float desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_float __global *object, -+ float __local *expected, -+ float desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_float __local *object, -+ float __local *expected, -+ float desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_float __global *object, -+ float __private *expected, -+ float desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_float __local *object, -+ float __private *expected, -+ float desired); - #if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) --#ifdef cl_khr_fp64 --bool __ovld atomic_compare_exchange_strong(volatile atomic_double *object, double *expected, double desired); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_double *object, double *expected, -- double desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_double *object, double *expected, -- double desired, memory_order success, memory_order failure, memory_scope scope); --bool __ovld atomic_compare_exchange_weak(volatile atomic_double *object, double *expected, double desired); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_double *object, double *expected, -- double desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_double *object, double *expected, -- double desired, memory_order success, memory_order failure, memory_scope scope); --#endif //cl_khr_fp64 --bool __ovld atomic_compare_exchange_strong(volatile atomic_long *object, long *expected, long desired); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_long *object, long *expected, -- long desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_long *object, long *expected, -- long desired, memory_order success, memory_order failure, memory_scope scope); --bool __ovld atomic_compare_exchange_weak(volatile atomic_long *object, long *expected, long desired); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_long *object, long *expected, -- long desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_long *object, long *expected, -- long desired, memory_order success, memory_order failure, memory_scope scope); --bool __ovld atomic_compare_exchange_strong(volatile atomic_ulong *object, ulong *expected, ulong desired); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_ulong *object, ulong *expected, -- ulong desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_ulong *object, ulong *expected, -- ulong desired, memory_order success, memory_order failure, memory_scope scope); --bool __ovld atomic_compare_exchange_weak(volatile atomic_ulong *object, ulong *expected, ulong desired); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_ulong *object, ulong *expected, -- ulong desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_ulong *object, ulong *expected, -- ulong desired, memory_order success, memory_order failure, memory_scope scope); -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_double __global *object, -+ double __global *expected, double desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_double __local *object, -+ double __global *expected, double desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_double __global *object, -+ double __local *expected, double desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_double __local *object, -+ double __local *expected, double desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_double __global *object, -+ double __private *expected, double desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_double __local *object, -+ double __private *expected, double desired); -+bool __ovld -+atomic_compare_exchange_weak(volatile atomic_double __global *object, -+ double __global *expected, double desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_double __local *object, -+ double __global *expected, -+ double desired); -+bool __ovld -+atomic_compare_exchange_weak(volatile atomic_double __global *object, -+ double __local *expected, double desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_double __local *object, -+ double __local *expected, -+ double desired); -+bool __ovld -+atomic_compare_exchange_weak(volatile atomic_double __global *object, -+ double __private *expected, double desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_double __local *object, -+ double __private *expected, -+ double desired); - #endif -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_long __global *object, -+ long __global *expected, long desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_long __local *object, -+ long __global *expected, -+ long desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_long __global *object, -+ long __local *expected, long desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_long __local *object, -+ long __local *expected, -+ long desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_long __global *object, -+ long __private *expected, long desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_long __local *object, -+ long __private *expected, -+ long desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_long __global *object, -+ long __global *expected, long desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_long __local *object, -+ long __global *expected, long desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_long __global *object, -+ long __local *expected, long desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_long __local *object, -+ long __local *expected, long desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_long __global *object, -+ long __private *expected, -+ long desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_long __local *object, -+ long __private *expected, -+ long desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_ulong __global *object, -+ ulong __global *expected, ulong desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_ulong __local *object, -+ ulong __global *expected, ulong desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_ulong __global *object, -+ ulong __local *expected, ulong desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_ulong __local *object, -+ ulong __local *expected, ulong desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_ulong __global *object, -+ ulong __private *expected, ulong desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_ulong __local *object, -+ ulong __private *expected, ulong desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_ulong __global *object, -+ ulong __global *expected, -+ ulong desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_ulong __local *object, -+ ulong __global *expected, -+ ulong desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_ulong __global *object, -+ ulong __local *expected, -+ ulong desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_ulong __local *object, -+ ulong __local *expected, -+ ulong desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_ulong __global *object, -+ ulong __private *expected, -+ ulong desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_ulong __local *object, -+ ulong __private *expected, -+ ulong desired); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // defined(__opencl_c_atomic_scope_device) && -+ // defined(__opencl_c_atomic_order_seq_cst) -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+ -+#ifdef __opencl_c_generic_address_space -+#ifdef __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_int *object, -+ int *expected, int desired, -+ memory_order success, -+ memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint *object, uint *expected, uint desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_int *object, -+ int *expected, int desired, -+ memory_order success, -+ memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_uint *object, -+ uint *expected, uint desired, -+ memory_order success, -+ memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float *object, float *expected, float desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_float *object, -+ float *expected, -+ float desired, -+ memory_order success, -+ memory_order failure); -+#endif // __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_int *object, -+ int *expected, int desired, -+ memory_order success, -+ memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint *object, uint *expected, uint desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_int *object, -+ int *expected, int desired, -+ memory_order success, -+ memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_uint *object, -+ uint *expected, uint desired, -+ memory_order success, -+ memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float *object, float *expected, float desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float *object, float *expected, float desired, -+ memory_order success, memory_order failure, memory_scope scope); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double *object, double *expected, double desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double *object, double *expected, double desired, -+ memory_order success, memory_order failure); -+#endif // __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double *object, double *expected, double desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double *object, double *expected, double desired, -+ memory_order success, memory_order failure, memory_scope scope); -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long *object, long *expected, long desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_long *object, -+ long *expected, long desired, -+ memory_order success, -+ memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong *object, ulong *expected, ulong desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_ulong *object, -+ ulong *expected, -+ ulong desired, -+ memory_order success, -+ memory_order failure); -+#endif // __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long *object, long *expected, long desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_long *object, -+ long *expected, long desired, -+ memory_order success, -+ memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong *object, ulong *expected, ulong desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong *object, ulong *expected, ulong desired, -+ memory_order success, memory_order failure, memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // __opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+#ifdef __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __global *object, int __global *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __local *object, int __global *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __global *object, int __local *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __local *object, int __local *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __global *object, int __private *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __local *object, int __private *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __global *object, uint __global *expected, -+ uint desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __local *object, uint __global *expected, uint desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __global *object, uint __local *expected, uint desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __local *object, uint __local *expected, uint desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __global *object, uint __private *expected, -+ uint desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __local *object, uint __private *expected, -+ uint desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __global *object, int __global *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __local *object, int __global *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __global *object, int __local *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __local *object, int __local *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __global *object, int __private *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __local *object, int __private *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __global *object, uint __global *expected, -+ uint desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __local *object, uint __global *expected, uint desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __global *object, uint __local *expected, uint desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __local *object, uint __local *expected, uint desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __global *object, uint __private *expected, -+ uint desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __local *object, uint __private *expected, -+ uint desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __global *object, float __global *expected, -+ float desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __local *object, float __global *expected, -+ float desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __global *object, float __local *expected, -+ float desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __local *object, float __local *expected, -+ float desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __global *object, float __private *expected, -+ float desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __local *object, float __private *expected, -+ float desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __global *object, float __global *expected, -+ float desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __local *object, float __global *expected, -+ float desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __global *object, float __local *expected, -+ float desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __local *object, float __local *expected, -+ float desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __global *object, float __private *expected, -+ float desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __local *object, float __private *expected, -+ float desired, memory_order success, memory_order failure); -+#endif // __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __global *object, int __global *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __local *object, int __global *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __global *object, int __local *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __local *object, int __local *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __global *object, int __private *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __local *object, int __private *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __global *object, uint __global *expected, -+ uint desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __local *object, uint __global *expected, uint desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __global *object, uint __local *expected, uint desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __local *object, uint __local *expected, uint desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __global *object, uint __private *expected, -+ uint desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __local *object, uint __private *expected, -+ uint desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __global *object, int __global *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __local *object, int __global *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __global *object, int __local *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __local *object, int __local *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __global *object, int __private *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __local *object, int __private *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __global *object, uint __global *expected, -+ uint desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __local *object, uint __global *expected, uint desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __global *object, uint __local *expected, uint desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __local *object, uint __local *expected, uint desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __global *object, uint __private *expected, -+ uint desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __local *object, uint __private *expected, -+ uint desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __global *object, float __global *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __local *object, float __global *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __global *object, float __local *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __local *object, float __local *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __global *object, float __private *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __local *object, float __private *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __global *object, float __global *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __local *object, float __global *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __global *object, float __local *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __local *object, float __local *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __global *object, float __private *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __local *object, float __private *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __global *object, double __global *expected, -+ double desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __local *object, double __global *expected, -+ double desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __global *object, double __local *expected, -+ double desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __local *object, double __local *expected, -+ double desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __global *object, double __private *expected, -+ double desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __local *object, double __private *expected, -+ double desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __global *object, double __global *expected, -+ double desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __local *object, double __global *expected, -+ double desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __global *object, double __local *expected, -+ double desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __local *object, double __local *expected, -+ double desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __global *object, double __private *expected, -+ double desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __local *object, double __private *expected, -+ double desired, memory_order success, memory_order failure); -+#endif // __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __global *object, double __global *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __local *object, double __global *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __global *object, double __local *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __local *object, double __local *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __global *object, double __private *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __local *object, double __private *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __global *object, double __global *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __local *object, double __global *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __global *object, double __local *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __local *object, double __local *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __global *object, double __private *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __local *object, double __private *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __global *object, long __global *expected, -+ long desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __local *object, long __global *expected, long desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __global *object, long __local *expected, long desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __local *object, long __local *expected, long desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __global *object, long __private *expected, -+ long desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __local *object, long __private *expected, -+ long desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __global *object, long __global *expected, -+ long desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __local *object, long __global *expected, long desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __global *object, long __local *expected, long desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __local *object, long __local *expected, long desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __global *object, long __private *expected, -+ long desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __local *object, long __private *expected, -+ long desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __global *object, ulong __global *expected, -+ ulong desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __local *object, ulong __global *expected, -+ ulong desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __global *object, ulong __local *expected, -+ ulong desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __local *object, ulong __local *expected, -+ ulong desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __global *object, ulong __private *expected, -+ ulong desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __local *object, ulong __private *expected, -+ ulong desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __global *object, ulong __global *expected, -+ ulong desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __local *object, ulong __global *expected, -+ ulong desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __global *object, ulong __local *expected, -+ ulong desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __local *object, ulong __local *expected, -+ ulong desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __global *object, ulong __private *expected, -+ ulong desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __local *object, ulong __private *expected, -+ ulong desired, memory_order success, memory_order failure); -+#endif // __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __global *object, long __global *expected, -+ long desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __local *object, long __global *expected, long desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __global *object, long __local *expected, long desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __local *object, long __local *expected, long desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __global *object, long __private *expected, -+ long desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __local *object, long __private *expected, -+ long desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __global *object, long __global *expected, -+ long desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __local *object, long __global *expected, long desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __global *object, long __local *expected, long desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __local *object, long __local *expected, long desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __global *object, long __private *expected, -+ long desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __local *object, long __private *expected, -+ long desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __global *object, ulong __global *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __local *object, ulong __global *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __global *object, ulong __local *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __local *object, ulong __local *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __global *object, ulong __private *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __local *object, ulong __private *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __global *object, ulong __global *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __local *object, ulong __global *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __global *object, ulong __local *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __local *object, ulong __local *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __global *object, ulong __private *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __local *object, ulong __private *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) - - // atomic_flag_test_and_set() and atomic_flag_clear() -- -+#if defined(__opencl_c_atomic_scope_device) && \ -+ defined(__opencl_c_atomic_order_seq_cst) -+#ifdef __opencl_c_generic_address_space - bool __ovld atomic_flag_test_and_set(volatile atomic_flag *object); --bool __ovld atomic_flag_test_and_set_explicit(volatile atomic_flag *object, memory_order order); --bool __ovld atomic_flag_test_and_set_explicit(volatile atomic_flag *object, memory_order order, memory_scope scope); - void __ovld atomic_flag_clear(volatile atomic_flag *object); --void __ovld atomic_flag_clear_explicit(volatile atomic_flag *object, memory_order order); --void __ovld atomic_flag_clear_explicit(volatile atomic_flag *object, memory_order order, memory_scope scope); -+#endif // __opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+bool __ovld atomic_flag_test_and_set(volatile atomic_flag __global *object); -+bool __ovld atomic_flag_test_and_set(volatile atomic_flag __local *object); -+void __ovld atomic_flag_clear(volatile atomic_flag __global *object); -+void __ovld atomic_flag_clear(volatile atomic_flag __local *object); -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+#endif // defined(__opencl_c_atomic_scope_device) && -+ // defined(__opencl_c_atomic_order_seq_cst) -+ -+#ifdef __opencl_c_generic_address_space -+#ifdef __opencl_c_atomic_scope_device -+bool __ovld atomic_flag_test_and_set_explicit(volatile atomic_flag *object, -+ memory_order order); -+void __ovld atomic_flag_clear_explicit(volatile atomic_flag *object, -+ memory_order order); -+#endif // __opencl_c_atomic_scope_device -+bool __ovld atomic_flag_test_and_set_explicit(volatile atomic_flag *object, -+ memory_order order, -+ memory_scope scope); -+void __ovld atomic_flag_clear_explicit(volatile atomic_flag *object, -+ memory_order order, memory_scope scope); -+#endif // __opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+#ifdef __opencl_c_atomic_scope_device -+bool __ovld atomic_flag_test_and_set_explicit( -+ volatile atomic_flag __global *object, memory_order order); -+bool __ovld atomic_flag_test_and_set_explicit( -+ volatile atomic_flag __local *object, memory_order order); -+void __ovld atomic_flag_clear_explicit(volatile atomic_flag __global *object, -+ memory_order order); -+void __ovld atomic_flag_clear_explicit(volatile atomic_flag __local *object, -+ memory_order order); -+#endif // __opencl_c_atomic_scope_device -+bool __ovld -+atomic_flag_test_and_set_explicit(volatile atomic_flag __global *object, -+ memory_order order, memory_scope scope); -+bool __ovld -+atomic_flag_test_and_set_explicit(volatile atomic_flag __local *object, -+ memory_order order, memory_scope scope); -+void __ovld atomic_flag_clear_explicit(volatile atomic_flag __global *object, -+ memory_order order, memory_scope scope); -+void __ovld atomic_flag_clear_explicit(volatile atomic_flag __local *object, -+ memory_order order, memory_scope scope); -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) - - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - -@@ -13918,7 +15943,7 @@ float16 __ovld __cnfn shuffle(float4 x, uint16 mask); - float16 __ovld __cnfn shuffle(float8 x, uint16 mask); - float16 __ovld __cnfn shuffle(float16 x, uint16 mask); - --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double2 __ovld __cnfn shuffle(double2 x, ulong2 mask); - double2 __ovld __cnfn shuffle(double4 x, ulong2 mask); - double2 __ovld __cnfn shuffle(double8 x, ulong2 mask); -@@ -13938,7 +15963,7 @@ double16 __ovld __cnfn shuffle(double2 x, ulong16 mask); - double16 __ovld __cnfn shuffle(double4 x, ulong16 mask); - double16 __ovld __cnfn shuffle(double8 x, ulong16 mask); - double16 __ovld __cnfn shuffle(double16 x, ulong16 mask); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - #ifdef cl_khr_fp16 - half2 __ovld __cnfn shuffle(half2 x, ushort2 mask); -@@ -14142,7 +16167,7 @@ float16 __ovld __cnfn shuffle2(float4 x, float4 y, uint16 mask); - float16 __ovld __cnfn shuffle2(float8 x, float8 y, uint16 mask); - float16 __ovld __cnfn shuffle2(float16 x, float16 y, uint16 mask); - --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double2 __ovld __cnfn shuffle2(double2 x, double2 y, ulong2 mask); - double2 __ovld __cnfn shuffle2(double4 x, double4 y, ulong2 mask); - double2 __ovld __cnfn shuffle2(double8 x, double8 y, ulong2 mask); -@@ -14162,7 +16187,7 @@ double16 __ovld __cnfn shuffle2(double2 x, double2 y, ulong16 mask); - double16 __ovld __cnfn shuffle2(double4 x, double4 y, ulong16 mask); - double16 __ovld __cnfn shuffle2(double8 x, double8 y, ulong16 mask); - double16 __ovld __cnfn shuffle2(double16 x, double16 y, ulong16 mask); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - #ifdef cl_khr_fp16 - half2 __ovld __cnfn shuffle2(half2 x, half2 y, ushort2 mask); -@@ -14198,6 +16223,7 @@ int printf(__constant const char* st, ...) __attribute__((format(printf, 1, 2))) - #pragma OPENCL EXTENSION cl_khr_gl_msaa_sharing : enable - #endif //cl_khr_gl_msaa_sharing - -+#ifdef __opencl_c_images - /** - * Use the coordinate (coord.xy) to do an element lookup in - * the 2D image object specified by image. -@@ -14476,6 +16502,7 @@ half4 __purefn __ovld read_imageh(read_only image1d_buffer_t image, int coord); - - // Image read functions for read_write images - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - float4 __purefn __ovld read_imagef(read_write image1d_t image, int coord); - int4 __purefn __ovld read_imagei(read_write image1d_t image, int coord); - uint4 __purefn __ovld read_imageui(read_write image1d_t image, int coord); -@@ -14519,6 +16546,7 @@ float __purefn __ovld read_imagef(read_write image2d_array_msaa_depth_t image, i - #endif //cl_khr_gl_msaa_sharing - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - #ifdef cl_khr_mipmap_image - float4 __purefn __ovld read_imagef(read_write image1d_t image, sampler_t sampler, float coord, float lod); - int4 __purefn __ovld read_imagei(read_write image1d_t image, sampler_t sampler, float coord, float lod); -@@ -14569,6 +16597,7 @@ int4 __purefn __ovld read_imagei(read_write image3d_t image, sampler_t sampler, - uint4 __purefn __ovld read_imageui(read_write image3d_t image, sampler_t sampler, float4 coord, float4 gradientX, float4 gradientY); - - #endif //cl_khr_mipmap_image -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - // Image read functions returning half4 type -@@ -14580,6 +16609,7 @@ half4 __purefn __ovld read_imageh(read_write image1d_array_t image, int2 coord); - half4 __purefn __ovld read_imageh(read_write image2d_array_t image, int4 coord); - half4 __purefn __ovld read_imageh(read_write image1d_buffer_t image, int coord); - #endif //cl_khr_fp16 -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** -@@ -14669,7 +16699,7 @@ void __ovld write_imagef(write_only image1d_array_t image_array, int2 coord, flo - void __ovld write_imagei(write_only image1d_array_t image_array, int2 coord, int4 color); - void __ovld write_imageui(write_only image1d_array_t image_array, int2 coord, uint4 color); - --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - void __ovld write_imagef(write_only image3d_t image, int4 coord, float4 color); - void __ovld write_imagei(write_only image3d_t image, int4 coord, int4 color); - void __ovld write_imageui(write_only image3d_t image, int4 coord, uint4 color); -@@ -14702,7 +16732,7 @@ void __ovld write_imageui(write_only image2d_array_t image_array, int4 coord, in - void __ovld write_imagef(write_only image2d_depth_t image, int2 coord, int lod, float depth); - void __ovld write_imagef(write_only image2d_array_depth_t image, int4 coord, int lod, float depth); - --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - void __ovld write_imagef(write_only image3d_t image, int4 coord, int lod, float4 color); - void __ovld write_imagei(write_only image3d_t image, int4 coord, int lod, int4 color); - void __ovld write_imageui(write_only image3d_t image, int4 coord, int lod, uint4 color); -@@ -14715,7 +16745,7 @@ void __ovld write_imageui(write_only image3d_t image, int4 coord, int lod, uint4 - #ifdef cl_khr_fp16 - void __ovld write_imageh(write_only image1d_t image, int coord, half4 color); - void __ovld write_imageh(write_only image2d_t image, int2 coord, half4 color); --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - void __ovld write_imageh(write_only image3d_t image, int4 coord, half4 color); - #endif - void __ovld write_imageh(write_only image1d_array_t image, int2 coord, half4 color); -@@ -14725,6 +16755,7 @@ void __ovld write_imageh(write_only image1d_buffer_t image, int coord, half4 col - - // Image write functions for read_write images - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - void __ovld write_imagef(read_write image2d_t image, int2 coord, float4 color); - void __ovld write_imagei(read_write image2d_t image, int2 coord, int4 color); - void __ovld write_imageui(read_write image2d_t image, int2 coord, uint4 color); -@@ -14745,7 +16776,7 @@ void __ovld write_imagef(read_write image1d_array_t image_array, int2 coord, flo - void __ovld write_imagei(read_write image1d_array_t image_array, int2 coord, int4 color); - void __ovld write_imageui(read_write image1d_array_t image_array, int2 coord, uint4 color); - --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - void __ovld write_imagef(read_write image3d_t image, int4 coord, float4 color); - void __ovld write_imagei(read_write image3d_t image, int4 coord, int4 color); - void __ovld write_imageui(read_write image3d_t image, int4 coord, uint4 color); -@@ -14777,7 +16808,7 @@ void __ovld write_imageui(read_write image2d_array_t image_array, int4 coord, in - void __ovld write_imagef(read_write image2d_depth_t image, int2 coord, int lod, float color); - void __ovld write_imagef(read_write image2d_array_depth_t image, int4 coord, int lod, float color); - --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - void __ovld write_imagef(read_write image3d_t image, int4 coord, int lod, float4 color); - void __ovld write_imagei(read_write image3d_t image, int4 coord, int lod, int4 color); - void __ovld write_imageui(read_write image3d_t image, int4 coord, int lod, uint4 color); -@@ -14790,13 +16821,14 @@ void __ovld write_imageui(read_write image3d_t image, int4 coord, int lod, uint4 - #ifdef cl_khr_fp16 - void __ovld write_imageh(read_write image1d_t image, int coord, half4 color); - void __ovld write_imageh(read_write image2d_t image, int2 coord, half4 color); --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - void __ovld write_imageh(read_write image3d_t image, int4 coord, half4 color); - #endif - void __ovld write_imageh(read_write image1d_array_t image, int2 coord, half4 color); - void __ovld write_imageh(read_write image2d_array_t image, int4 coord, half4 color); - void __ovld write_imageh(read_write image1d_buffer_t image, int coord, half4 color); - #endif //cl_khr_fp16 -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - // Note: In OpenCL v1.0/1.1/1.2, image argument of image query builtin functions does not have -@@ -14810,7 +16842,7 @@ void __ovld write_imageh(read_write image1d_buffer_t image, int coord, half4 col - int __ovld __cnfn get_image_width(read_only image1d_t image); - int __ovld __cnfn get_image_width(read_only image1d_buffer_t image); - int __ovld __cnfn get_image_width(read_only image2d_t image); --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - int __ovld __cnfn get_image_width(read_only image3d_t image); - #endif - int __ovld __cnfn get_image_width(read_only image1d_array_t image); -@@ -14829,7 +16861,7 @@ int __ovld __cnfn get_image_width(read_only image2d_array_msaa_depth_t image); - int __ovld __cnfn get_image_width(write_only image1d_t image); - int __ovld __cnfn get_image_width(write_only image1d_buffer_t image); - int __ovld __cnfn get_image_width(write_only image2d_t image); --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - int __ovld __cnfn get_image_width(write_only image3d_t image); - #endif - int __ovld __cnfn get_image_width(write_only image1d_array_t image); -@@ -14846,6 +16878,7 @@ int __ovld __cnfn get_image_width(write_only image2d_array_msaa_depth_t image); - #endif //cl_khr_gl_msaa_sharing - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - int __ovld __cnfn get_image_width(read_write image1d_t image); - int __ovld __cnfn get_image_width(read_write image1d_buffer_t image); - int __ovld __cnfn get_image_width(read_write image2d_t image); -@@ -14862,6 +16895,7 @@ int __ovld __cnfn get_image_width(read_write image2d_msaa_depth_t image); - int __ovld __cnfn get_image_width(read_write image2d_array_msaa_t image); - int __ovld __cnfn get_image_width(read_write image2d_array_msaa_depth_t image); - #endif //cl_khr_gl_msaa_sharing -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** -@@ -14882,7 +16916,7 @@ int __ovld __cnfn get_image_height(read_only image2d_array_msaa_depth_t image); - #endif //cl_khr_gl_msaa_sharing - - int __ovld __cnfn get_image_height(write_only image2d_t image); --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - int __ovld __cnfn get_image_height(write_only image3d_t image); - #endif - int __ovld __cnfn get_image_height(write_only image2d_array_t image); -@@ -14898,6 +16932,7 @@ int __ovld __cnfn get_image_height(write_only image2d_array_msaa_depth_t image); - #endif //cl_khr_gl_msaa_sharing - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - int __ovld __cnfn get_image_height(read_write image2d_t image); - int __ovld __cnfn get_image_height(read_write image3d_t image); - int __ovld __cnfn get_image_height(read_write image2d_array_t image); -@@ -14911,6 +16946,7 @@ int __ovld __cnfn get_image_height(read_write image2d_msaa_depth_t image); - int __ovld __cnfn get_image_height(read_write image2d_array_msaa_t image); - int __ovld __cnfn get_image_height(read_write image2d_array_msaa_depth_t image); - #endif //cl_khr_gl_msaa_sharing -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** -@@ -14918,12 +16954,14 @@ int __ovld __cnfn get_image_height(read_write image2d_array_msaa_depth_t image); - */ - int __ovld __cnfn get_image_depth(read_only image3d_t image); - --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - int __ovld __cnfn get_image_depth(write_only image3d_t image); - #endif - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - int __ovld __cnfn get_image_depth(read_write image3d_t image); -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - // OpenCL Extension v2.0 s9.18 - Mipmaps -@@ -14939,13 +16977,15 @@ int __ovld get_image_num_mip_levels(read_only image3d_t image); - - int __ovld get_image_num_mip_levels(write_only image1d_t image); - int __ovld get_image_num_mip_levels(write_only image2d_t image); --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - int __ovld get_image_num_mip_levels(write_only image3d_t image); - #endif - -+#ifdef __opencl_c_read_write_images - int __ovld get_image_num_mip_levels(read_write image1d_t image); - int __ovld get_image_num_mip_levels(read_write image2d_t image); - int __ovld get_image_num_mip_levels(read_write image3d_t image); -+#endif //__opencl_c_read_write_images - - int __ovld get_image_num_mip_levels(read_only image1d_array_t image); - int __ovld get_image_num_mip_levels(read_only image2d_array_t image); -@@ -14957,10 +16997,12 @@ int __ovld get_image_num_mip_levels(write_only image2d_array_t image); - int __ovld get_image_num_mip_levels(write_only image2d_array_depth_t image); - int __ovld get_image_num_mip_levels(write_only image2d_depth_t image); - -+#ifdef __opencl_c_read_write_images - int __ovld get_image_num_mip_levels(read_write image1d_array_t image); - int __ovld get_image_num_mip_levels(read_write image2d_array_t image); - int __ovld get_image_num_mip_levels(read_write image2d_array_depth_t image); - int __ovld get_image_num_mip_levels(read_write image2d_depth_t image); -+#endif //__opencl_c_read_write_images - - #endif //cl_khr_mipmap_image - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -@@ -15004,7 +17046,7 @@ int __ovld __cnfn get_image_channel_data_type(read_only image2d_array_msaa_depth - int __ovld __cnfn get_image_channel_data_type(write_only image1d_t image); - int __ovld __cnfn get_image_channel_data_type(write_only image1d_buffer_t image); - int __ovld __cnfn get_image_channel_data_type(write_only image2d_t image); --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - int __ovld __cnfn get_image_channel_data_type(write_only image3d_t image); - #endif - int __ovld __cnfn get_image_channel_data_type(write_only image1d_array_t image); -@@ -15021,6 +17063,7 @@ int __ovld __cnfn get_image_channel_data_type(write_only image2d_array_msaa_dept - #endif //cl_khr_gl_msaa_sharing - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - int __ovld __cnfn get_image_channel_data_type(read_write image1d_t image); - int __ovld __cnfn get_image_channel_data_type(read_write image1d_buffer_t image); - int __ovld __cnfn get_image_channel_data_type(read_write image2d_t image); -@@ -15037,6 +17080,7 @@ int __ovld __cnfn get_image_channel_data_type(read_write image2d_msaa_depth_t im - int __ovld __cnfn get_image_channel_data_type(read_write image2d_array_msaa_t image); - int __ovld __cnfn get_image_channel_data_type(read_write image2d_array_msaa_depth_t image); - #endif //cl_khr_gl_msaa_sharing -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** -@@ -15076,7 +17120,7 @@ int __ovld __cnfn get_image_channel_order(read_only image2d_array_msaa_depth_t i - int __ovld __cnfn get_image_channel_order(write_only image1d_t image); - int __ovld __cnfn get_image_channel_order(write_only image1d_buffer_t image); - int __ovld __cnfn get_image_channel_order(write_only image2d_t image); --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - int __ovld __cnfn get_image_channel_order(write_only image3d_t image); - #endif - int __ovld __cnfn get_image_channel_order(write_only image1d_array_t image); -@@ -15093,6 +17137,7 @@ int __ovld __cnfn get_image_channel_order(write_only image2d_array_msaa_depth_t - #endif //cl_khr_gl_msaa_sharing - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - int __ovld __cnfn get_image_channel_order(read_write image1d_t image); - int __ovld __cnfn get_image_channel_order(read_write image1d_buffer_t image); - int __ovld __cnfn get_image_channel_order(read_write image2d_t image); -@@ -15109,6 +17154,7 @@ int __ovld __cnfn get_image_channel_order(read_write image2d_msaa_depth_t image) - int __ovld __cnfn get_image_channel_order(read_write image2d_array_msaa_t image); - int __ovld __cnfn get_image_channel_order(read_write image2d_array_msaa_depth_t image); - #endif //cl_khr_gl_msaa_sharing -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** -@@ -15143,6 +17189,7 @@ int2 __ovld __cnfn get_image_dim(write_only image2d_array_msaa_depth_t image); - #endif //cl_khr_gl_msaa_sharing - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - int2 __ovld __cnfn get_image_dim(read_write image2d_t image); - int2 __ovld __cnfn get_image_dim(read_write image2d_array_t image); - #ifdef cl_khr_depth_images -@@ -15155,6 +17202,7 @@ int2 __ovld __cnfn get_image_dim(read_write image2d_msaa_depth_t image); - int2 __ovld __cnfn get_image_dim(read_write image2d_array_msaa_t image); - int2 __ovld __cnfn get_image_dim(read_write image2d_array_msaa_depth_t image); - #endif //cl_khr_gl_msaa_sharing -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** -@@ -15164,11 +17212,13 @@ int2 __ovld __cnfn get_image_dim(read_write image2d_array_msaa_depth_t image); - * component and the w component is 0. - */ - int4 __ovld __cnfn get_image_dim(read_only image3d_t image); --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - int4 __ovld __cnfn get_image_dim(write_only image3d_t image); - #endif - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - int4 __ovld __cnfn get_image_dim(read_write image3d_t image); -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** -@@ -15196,6 +17246,7 @@ size_t __ovld __cnfn get_image_array_size(write_only image2d_array_msaa_depth_t - #endif //cl_khr_gl_msaa_sharing - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - size_t __ovld __cnfn get_image_array_size(read_write image1d_array_t image_array); - size_t __ovld __cnfn get_image_array_size(read_write image2d_array_t image_array); - #ifdef cl_khr_depth_images -@@ -15205,6 +17256,7 @@ size_t __ovld __cnfn get_image_array_size(read_write image2d_array_depth_t image - size_t __ovld __cnfn get_image_array_size(read_write image2d_array_msaa_t image_array); - size_t __ovld __cnfn get_image_array_size(read_write image2d_array_msaa_depth_t image_array); - #endif //cl_khr_gl_msaa_sharing -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** -@@ -15222,16 +17274,21 @@ int __ovld get_image_num_samples(write_only image2d_array_msaa_t image); - int __ovld get_image_num_samples(write_only image2d_array_msaa_depth_t image); - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - int __ovld get_image_num_samples(read_write image2d_msaa_t image); - int __ovld get_image_num_samples(read_write image2d_msaa_depth_t image); - int __ovld get_image_num_samples(read_write image2d_array_msaa_t image); - int __ovld get_image_num_samples(read_write image2d_array_msaa_depth_t image); -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - #endif - -+#endif //__opencl_c_images -+ - // OpenCL v2.0 s6.13.15 - Work-group Functions - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_work_group_collective_functions - int __ovld __conv work_group_all(int predicate); - int __ovld __conv work_group_any(int predicate); - -@@ -15255,11 +17312,11 @@ ulong __ovld __conv work_group_broadcast(ulong a, size_t x, size_t y, size_t z); - float __ovld __conv work_group_broadcast(float a, size_t local_id); - float __ovld __conv work_group_broadcast(float a, size_t x, size_t y); - float __ovld __conv work_group_broadcast(float a, size_t x, size_t y, size_t z); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __conv work_group_broadcast(double a, size_t local_id); - double __ovld __conv work_group_broadcast(double a, size_t x, size_t y); - double __ovld __conv work_group_broadcast(double a, size_t x, size_t y, size_t z); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - #ifdef cl_khr_fp16 - half __ovld __conv work_group_reduce_add(half x); -@@ -15317,7 +17374,7 @@ float __ovld __conv work_group_scan_exclusive_max(float x); - float __ovld __conv work_group_scan_inclusive_add(float x); - float __ovld __conv work_group_scan_inclusive_min(float x); - float __ovld __conv work_group_scan_inclusive_max(float x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __conv work_group_reduce_add(double x); - double __ovld __conv work_group_reduce_min(double x); - double __ovld __conv work_group_reduce_max(double x); -@@ -15327,19 +17384,18 @@ double __ovld __conv work_group_scan_exclusive_max(double x); - double __ovld __conv work_group_scan_inclusive_add(double x); - double __ovld __conv work_group_scan_inclusive_min(double x); - double __ovld __conv work_group_scan_inclusive_max(double x); --#endif //cl_khr_fp64 -- -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#endif //__opencl_c_work_group_collective_functions - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - // OpenCL v2.0 s6.13.16 - Pipe Functions --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_pipes - bool __ovld is_valid_reserve_id(reserve_id_t reserve_id); --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -- -+#endif //__opencl_c_pipes - - // OpenCL v2.0 s6.13.17 - Enqueue Kernels - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -- -+#ifdef __opencl_c_device_enqueue - ndrange_t __ovld ndrange_1D(size_t); - ndrange_t __ovld ndrange_1D(size_t, size_t); - ndrange_t __ovld ndrange_1D(size_t, size_t, size_t); -@@ -15367,11 +17423,13 @@ bool __ovld is_valid_event (clk_event_t event); - void __ovld capture_event_profiling_info(clk_event_t, clk_profiling_info, __global void* value); - - queue_t __ovld get_default_queue(void); -+#endif //__opencl_c_device_enqueue - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - // OpenCL Extension v2.0 s9.17 - Sub-groups - --#if defined(cl_intel_subgroups) || defined(cl_khr_subgroups) -+#if defined(cl_intel_subgroups) || defined(cl_khr_subgroups) || \ -+ defined(__opencl_c_subgroups) - // Shared Sub Group Functions - uint __ovld get_sub_group_size(void); - uint __ovld get_max_sub_group_size(void); -@@ -15457,7 +17515,7 @@ half __ovld __conv sub_group_scan_inclusive_min(half x); - half __ovld __conv sub_group_scan_inclusive_max(half x); - #endif //cl_khr_fp16 - --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __conv sub_group_broadcast(double x, uint sub_group_local_id); - double __ovld __conv sub_group_reduce_add(double x); - double __ovld __conv sub_group_reduce_min(double x); -@@ -15468,7 +17526,7 @@ double __ovld __conv sub_group_scan_exclusive_max(double x); - double __ovld __conv sub_group_scan_inclusive_add(double x); - double __ovld __conv sub_group_scan_inclusive_min(double x); - double __ovld __conv sub_group_scan_inclusive_max(double x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - #endif //cl_khr_subgroups cl_intel_subgroups - -@@ -15570,16 +17628,22 @@ uint16 __ovld __conv intel_sub_group_shuffle_xor( uint16 x, uint c ); - long __ovld __conv intel_sub_group_shuffle_xor( long x, uint c ); - ulong __ovld __conv intel_sub_group_shuffle_xor( ulong x, uint c ); - -+#ifdef __opencl_c_images - uint __ovld __conv intel_sub_group_block_read( read_only image2d_t image, int2 coord ); - uint2 __ovld __conv intel_sub_group_block_read2( read_only image2d_t image, int2 coord ); - uint4 __ovld __conv intel_sub_group_block_read4( read_only image2d_t image, int2 coord ); - uint8 __ovld __conv intel_sub_group_block_read8( read_only image2d_t image, int2 coord ); -+#endif //__opencl_c_images - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_images -+#ifdef __opencl_c_read_write_images - uint __ovld __conv intel_sub_group_block_read(read_write image2d_t image, int2 coord); - uint2 __ovld __conv intel_sub_group_block_read2(read_write image2d_t image, int2 coord); - uint4 __ovld __conv intel_sub_group_block_read4(read_write image2d_t image, int2 coord); - uint8 __ovld __conv intel_sub_group_block_read8(read_write image2d_t image, int2 coord); -+#endif //__opencl_c_read_write_images -+#endif //__opencl_c_images - #endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - uint __ovld __conv intel_sub_group_block_read( const __global uint* p ); -@@ -15587,16 +17651,22 @@ uint2 __ovld __conv intel_sub_group_block_read2( const __global uint* p ); - uint4 __ovld __conv intel_sub_group_block_read4( const __global uint* p ); - uint8 __ovld __conv intel_sub_group_block_read8( const __global uint* p ); - -+#ifdef __opencl_c_images - void __ovld __conv intel_sub_group_block_write(write_only image2d_t image, int2 coord, uint data); - void __ovld __conv intel_sub_group_block_write2(write_only image2d_t image, int2 coord, uint2 data); - void __ovld __conv intel_sub_group_block_write4(write_only image2d_t image, int2 coord, uint4 data); - void __ovld __conv intel_sub_group_block_write8(write_only image2d_t image, int2 coord, uint8 data); -+#endif //__opencl_c_images - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_images -+#ifdef __opencl_c_read_write_images - void __ovld __conv intel_sub_group_block_write(read_write image2d_t image, int2 coord, uint data); - void __ovld __conv intel_sub_group_block_write2(read_write image2d_t image, int2 coord, uint2 data); - void __ovld __conv intel_sub_group_block_write4(read_write image2d_t image, int2 coord, uint4 data); - void __ovld __conv intel_sub_group_block_write8(read_write image2d_t image, int2 coord, uint8 data); -+#endif // __opencl_c_read_write_images -+#endif // __opencl_c_images - #endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - void __ovld __conv intel_sub_group_block_write( __global uint* p, uint data ); -@@ -15611,7 +17681,7 @@ half __ovld __conv intel_sub_group_shuffle_up( half prev, half cur, uint c ); - half __ovld __conv intel_sub_group_shuffle_xor( half x, uint c ); - #endif - --#if defined(cl_khr_fp64) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __conv intel_sub_group_shuffle( double x, uint c ); - double __ovld __conv intel_sub_group_shuffle_down( double prev, double cur, uint c ); - double __ovld __conv intel_sub_group_shuffle_up( double prev, double cur, uint c ); -@@ -15710,16 +17780,22 @@ ushort __ovld __conv intel_sub_group_scan_inclusive_min( ushort x ); - short __ovld __conv intel_sub_group_scan_inclusive_max( short x ); - ushort __ovld __conv intel_sub_group_scan_inclusive_max( ushort x ); - -+#ifdef __opencl_c_images - uint __ovld __conv intel_sub_group_block_read_ui( read_only image2d_t image, int2 byte_coord ); - uint2 __ovld __conv intel_sub_group_block_read_ui2( read_only image2d_t image, int2 byte_coord ); - uint4 __ovld __conv intel_sub_group_block_read_ui4( read_only image2d_t image, int2 byte_coord ); - uint8 __ovld __conv intel_sub_group_block_read_ui8( read_only image2d_t image, int2 byte_coord ); -+#endif //__opencl_c_images - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_images -+#ifdef __opencl_c_read_write_images - uint __ovld __conv intel_sub_group_block_read_ui( read_write image2d_t image, int2 byte_coord ); - uint2 __ovld __conv intel_sub_group_block_read_ui2( read_write image2d_t image, int2 byte_coord ); - uint4 __ovld __conv intel_sub_group_block_read_ui4( read_write image2d_t image, int2 byte_coord ); - uint8 __ovld __conv intel_sub_group_block_read_ui8( read_write image2d_t image, int2 byte_coord ); -+#endif //__opencl_c_read_write_images -+#endif //__opencl_c_images - #endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - uint __ovld __conv intel_sub_group_block_read_ui( const __global uint* p ); -@@ -15727,16 +17803,22 @@ uint2 __ovld __conv intel_sub_group_block_read_ui2( const __global uint* p - uint4 __ovld __conv intel_sub_group_block_read_ui4( const __global uint* p ); - uint8 __ovld __conv intel_sub_group_block_read_ui8( const __global uint* p ); - -+#ifdef __opencl_c_images - void __ovld __conv intel_sub_group_block_write_ui( read_only image2d_t image, int2 byte_coord, uint data ); - void __ovld __conv intel_sub_group_block_write_ui2( read_only image2d_t image, int2 byte_coord, uint2 data ); - void __ovld __conv intel_sub_group_block_write_ui4( read_only image2d_t image, int2 byte_coord, uint4 data ); - void __ovld __conv intel_sub_group_block_write_ui8( read_only image2d_t image, int2 byte_coord, uint8 data ); -+#endif //__opencl_c_images - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_images -+#ifdef __opencl_c_read_write_images - void __ovld __conv intel_sub_group_block_write_ui( read_write image2d_t image, int2 byte_coord, uint data ); - void __ovld __conv intel_sub_group_block_write_ui2( read_write image2d_t image, int2 byte_coord, uint2 data ); - void __ovld __conv intel_sub_group_block_write_ui4( read_write image2d_t image, int2 byte_coord, uint4 data ); - void __ovld __conv intel_sub_group_block_write_ui8( read_write image2d_t image, int2 byte_coord, uint8 data ); -+#endif //__opencl_c_read_write_images -+#endif //__opencl_c_images - #endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - void __ovld __conv intel_sub_group_block_write_ui( __global uint* p, uint data ); -@@ -15744,16 +17826,22 @@ void __ovld __conv intel_sub_group_block_write_ui2( __global uint* p, uint - void __ovld __conv intel_sub_group_block_write_ui4( __global uint* p, uint4 data ); - void __ovld __conv intel_sub_group_block_write_ui8( __global uint* p, uint8 data ); - -+#ifdef __opencl_c_images - ushort __ovld __conv intel_sub_group_block_read_us( read_only image2d_t image, int2 coord ); - ushort2 __ovld __conv intel_sub_group_block_read_us2( read_only image2d_t image, int2 coord ); - ushort4 __ovld __conv intel_sub_group_block_read_us4( read_only image2d_t image, int2 coord ); - ushort8 __ovld __conv intel_sub_group_block_read_us8( read_only image2d_t image, int2 coord ); -+#endif //__opencl_c_images - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_images -+#ifdef __opencl_c_read_write_images - ushort __ovld __conv intel_sub_group_block_read_us(read_write image2d_t image, int2 coord); - ushort2 __ovld __conv intel_sub_group_block_read_us2(read_write image2d_t image, int2 coord); - ushort4 __ovld __conv intel_sub_group_block_read_us4(read_write image2d_t image, int2 coord); - ushort8 __ovld __conv intel_sub_group_block_read_us8(read_write image2d_t image, int2 coord); -+#endif //__opencl_c_read_write_images -+#endif //__opencl_c_images - #endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - ushort __ovld __conv intel_sub_group_block_read_us( const __global ushort* p ); -@@ -15761,16 +17849,22 @@ ushort2 __ovld __conv intel_sub_group_block_read_us2( const __global ushort* - ushort4 __ovld __conv intel_sub_group_block_read_us4( const __global ushort* p ); - ushort8 __ovld __conv intel_sub_group_block_read_us8( const __global ushort* p ); - -+#ifdef __opencl_c_images - void __ovld __conv intel_sub_group_block_write_us(write_only image2d_t image, int2 coord, ushort data); - void __ovld __conv intel_sub_group_block_write_us2(write_only image2d_t image, int2 coord, ushort2 data); - void __ovld __conv intel_sub_group_block_write_us4(write_only image2d_t image, int2 coord, ushort4 data); - void __ovld __conv intel_sub_group_block_write_us8(write_only image2d_t image, int2 coord, ushort8 data); -+#endif //__opencl_c_images - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_images -+#ifdef __opencl_c_read_write_images - void __ovld __conv intel_sub_group_block_write_us(read_write image2d_t image, int2 coord, ushort data); - void __ovld __conv intel_sub_group_block_write_us2(read_write image2d_t image, int2 coord, ushort2 data); - void __ovld __conv intel_sub_group_block_write_us4(read_write image2d_t image, int2 coord, ushort4 data); - void __ovld __conv intel_sub_group_block_write_us8(read_write image2d_t image, int2 coord, ushort8 data); -+#endif //__opencl_c_read_write_images -+#endif //__opencl_c_images - #endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - void __ovld __conv intel_sub_group_block_write_us( __global ushort* p, ushort data ); -@@ -15889,6 +17983,7 @@ short2 __ovld intel_sub_group_avc_ime_adjust_ref_offset( - short2 ref_offset, ushort2 src_coord, ushort2 ref_window_size, - ushort2 image_size); - -+#ifdef __opencl_c_images - intel_sub_group_avc_ime_result_t __ovld - intel_sub_group_avc_ime_evaluate_with_single_reference( - read_only image2d_t src_image, read_only image2d_t ref_image, -@@ -15929,6 +18024,7 @@ intel_sub_group_avc_ime_evaluate_with_dual_reference_streaminout( - read_only image2d_t bwd_ref_image, sampler_t vme_media_sampler, - intel_sub_group_avc_ime_payload_t payload, - intel_sub_group_avc_ime_dual_reference_streamin_t streamin_components); -+#endif //__opencl_c_images - - intel_sub_group_avc_ime_single_reference_streamin_t __ovld - intel_sub_group_avc_ime_get_single_reference_streamin( -@@ -15993,6 +18089,7 @@ intel_sub_group_avc_ref_payload_t __ovld - intel_sub_group_avc_ref_set_bilinear_filter_enable( - intel_sub_group_avc_ref_payload_t payload); - -+#ifdef __opencl_c_images - intel_sub_group_avc_ref_result_t __ovld - intel_sub_group_avc_ref_evaluate_with_single_reference( - read_only image2d_t src_image, read_only image2d_t ref_image, -@@ -16011,6 +18108,7 @@ intel_sub_group_avc_ref_evaluate_with_multi_reference( - read_only image2d_t src_image, uint packed_reference_ids, - uchar packed_reference_field_polarities, sampler_t vme_media_sampler, - intel_sub_group_avc_ref_payload_t payload); -+#endif //__opencl_c_images - - // SIC built-in functions - intel_sub_group_avc_sic_payload_t __ovld -@@ -16061,6 +18159,7 @@ intel_sub_group_avc_sic_set_block_based_raw_skip_sad( - uchar block_based_skip_type, - intel_sub_group_avc_sic_payload_t payload); - -+#ifdef __opencl_c_images - intel_sub_group_avc_sic_result_t __ovld - intel_sub_group_avc_sic_evaluate_ipe( - read_only image2d_t src_image, sampler_t vme_media_sampler, -@@ -16083,6 +18182,7 @@ intel_sub_group_avc_sic_evaluate_with_multi_reference( - read_only image2d_t src_image, uint packed_reference_ids, - uchar packed_reference_field_polarities, sampler_t vme_media_sampler, - intel_sub_group_avc_sic_payload_t payload); -+#endif //__opencl_c_images - - uchar __ovld intel_sub_group_avc_sic_get_ipe_luma_shape( - intel_sub_group_avc_sic_result_t result); -diff --git a/clang/lib/Parse/ParseDecl.cpp b/clang/lib/Parse/ParseDecl.cpp -index 6353e14bc41a..a8c216de6e04 100644 ---- a/clang/lib/Parse/ParseDecl.cpp -+++ b/clang/lib/Parse/ParseDecl.cpp -@@ -3904,8 +3904,7 @@ void Parser::ParseDeclarationSpecifiers(DeclSpec &DS, - isInvalid = DS.SetTypeAltiVecBool(true, Loc, PrevSpec, DiagID, Policy); - break; - case tok::kw_pipe: -- if (!getLangOpts().OpenCL || (getLangOpts().OpenCLVersion < 200 && -- !getLangOpts().OpenCLCPlusPlus)) { -+ if (!getLangOpts().OpenCLPipeKeyword) { - // OpenCL 2.0 defined this keyword. OpenCL 1.2 and earlier should - // support the "pipe" word as identifier. - Tok.getIdentifierInfo()->revertTokenIDToIdentifier(); -@@ -4027,8 +4026,7 @@ void Parser::ParseDeclarationSpecifiers(DeclSpec &DS, - case tok::kw___generic: - // generic address space is introduced only in OpenCL v2.0 - // see OpenCL C Spec v2.0 s6.5.5 -- if (Actions.getLangOpts().OpenCLVersion < 200 && -- !Actions.getLangOpts().OpenCLCPlusPlus) { -+ if (!Actions.getLangOpts().OpenCLGenericKeyword) { - DiagID = diag::err_opencl_unknown_type_specifier; - PrevSpec = Tok.getIdentifierInfo()->getNameStart(); - isInvalid = true; -@@ -5050,8 +5048,7 @@ bool Parser::isDeclarationSpecifier(bool DisambiguatingWithExpression) { - default: return false; - - case tok::kw_pipe: -- return (getLangOpts().OpenCL && getLangOpts().OpenCLVersion >= 200) || -- getLangOpts().OpenCLCPlusPlus; -+ return getLangOpts().OpenCLPipeKeyword; - - case tok::identifier: // foo::bar - // Unfortunate hack to support "Class.factoryMethod" notation. -@@ -5557,8 +5554,7 @@ static bool isPtrOperatorToken(tok::TokenKind Kind, const LangOptions &Lang, - if (Kind == tok::star || Kind == tok::caret) - return true; - -- if (Kind == tok::kw_pipe && -- ((Lang.OpenCL && Lang.OpenCLVersion >= 200) || Lang.OpenCLCPlusPlus)) -+ if (Kind == tok::kw_pipe && Lang.OpenCLPipeKeyword) - return true; - - if (!Lang.CPlusPlus) -diff --git a/clang/lib/Parse/ParsePragma.cpp b/clang/lib/Parse/ParsePragma.cpp -index df411e1928d6..9a40ce888695 100644 ---- a/clang/lib/Parse/ParsePragma.cpp -+++ b/clang/lib/Parse/ParsePragma.cpp -@@ -697,12 +697,14 @@ void Parser::HandlePragmaOpenCLExtension() { - if (Name == "all") { - if (State == Disable) { - Opt.disableAll(); -- Opt.enableSupportedCore(getLangOpts()); -+ Opt.enableSupportedCore(); - } else { - PP.Diag(NameLoc, diag::warn_pragma_expected_predicate) << 1; - } -+ } else if (Opt.isFeature(Name)) { -+ PP.Diag(NameLoc, diag::warn_opencl_pragma_feature_ignore) << Ident; - } else if (State == Begin) { -- if (!Opt.isKnown(Name) || !Opt.isSupported(Name, getLangOpts())) { -+ if (!Opt.isKnown(Name) || !Opt.isSupported(Name)) { - Opt.support(Name); - } - Actions.setCurrentOpenCLExtension(Name); -@@ -712,9 +714,9 @@ void Parser::HandlePragmaOpenCLExtension() { - Actions.setCurrentOpenCLExtension(""); - } else if (!Opt.isKnown(Name)) - PP.Diag(NameLoc, diag::warn_pragma_unknown_extension) << Ident; -- else if (Opt.isSupportedExtension(Name, getLangOpts())) -+ else if (Opt.isSupportedExtension(Name)) - Opt.enable(Name, State == Enable); -- else if (Opt.isSupportedCore(Name, getLangOpts())) -+ else if (Opt.isSupportedCore(Name)) - PP.Diag(NameLoc, diag::warn_pragma_extension_is_core) << Ident; - else - PP.Diag(NameLoc, diag::warn_pragma_unsupported_extension) << Ident; -diff --git a/clang/lib/Sema/OpenCLBuiltins.td b/clang/lib/Sema/OpenCLBuiltins.td -index 9d6bb411eff8..d352d35f1e46 100644 ---- a/clang/lib/Sema/OpenCLBuiltins.td -+++ b/clang/lib/Sema/OpenCLBuiltins.td -@@ -22,11 +22,13 @@ - class Version<int _Version> { - int ID = _Version; - } -+ - def CLAll : Version< 0>; - def CL10 : Version<100>; - def CL11 : Version<110>; - def CL12 : Version<120>; - def CL20 : Version<200>; -+def CL30 : Version<300>; - - // Address spaces - // Pointer types need to be assigned an address space. -@@ -65,6 +67,14 @@ def FuncExtKhrGlMsaaSharing : FunctionExtension<"cl_khr_gl_msaa_sha - // Multiple extensions - def FuncExtKhrMipmapAndWrite3d : FunctionExtension<"cl_khr_mipmap_image cl_khr_3d_image_writes">; - -+// Features -+def FuncExtGenericAddressSpace : FunctionExtension<"__opencl_c_generic_address_space">; -+def FuncExtWorkGroupCollective : FunctionExtension<"__opencl_c_work_group_collective_functions">; -+def FuncExtPipes : FunctionExtension<"__opencl_c_pipes">; -+def FuncExtDeviceSidEenqueue : FunctionExtension<"__opencl_c_device_enqueue">; -+def FuncNonExplicitAtomicFeatures : FunctionExtension<"__opencl_c_atomic_order_seq_cst __opencl_c_atomic_scope_device">; -+def FuncNonExplicitAtomicFeaturesAndGenericAS : FunctionExtension<"__opencl_c_atomic_order_seq_cst __opencl_c_atomic_scope_device __opencl_c_generic_address_space">; -+ - // Qualified Type. These map to ASTContext::QualType. - class QualType<string _Name, bit _IsAbstract=0> { - // Name of the field or function in a clang::ASTContext -@@ -230,13 +240,9 @@ class Builtin<string _Name, list<Type> _Signature, list<bit> _Attributes = Attr. - bit IsConv = _Attributes[2]; - // OpenCL extensions to which the function belongs. - FunctionExtension Extension = FuncExtNone; -- // Version of OpenCL from which the function is available (e.g.: CL10). -- // MinVersion is inclusive. -- Version MinVersion = CL10; -- // Version of OpenCL from which the function is not supported anymore. -- // MaxVersion is exclusive. -+ // List of OpenCL version in which this function available. - // CLAll makes the function available for all versions. -- Version MaxVersion = CLAll; -+ list<Version> Versions = [CLAll]; - } - - //===----------------------------------------------------------------------===// -@@ -398,7 +404,7 @@ foreach name = ["get_global_size", "get_global_id", "get_local_size", - def : Builtin<name, [Size, UInt], Attr.Const>; - } - --let MinVersion = CL20 in { -+let Versions = [CL20, CL30] in { - def : Builtin<"get_enqueued_local_size", [Size, UInt]>; - foreach name = ["get_global_linear_id", "get_local_linear_id"] in { - def : Builtin<name, [Size]>; -@@ -471,7 +477,7 @@ foreach name = ["fma", "mad"] in { - } - - // --- Version dependent --- --let MaxVersion = CL20 in { -+let Versions = [CL10, CL11, CL12, CL30] in { - foreach AS = [GlobalAS, LocalAS, PrivateAS] in { - foreach name = ["fract", "modf", "sincos"] in { - def : Builtin<name, [FGenTypeN, FGenTypeN, PointerType<FGenTypeN, AS>]>; -@@ -488,7 +494,9 @@ let MaxVersion = CL20 in { - } - } - } --let MinVersion = CL20 in { -+ -+let Versions = [CL20, CL30] in { -+ let Extension = FuncExtGenericAddressSpace in { - foreach name = ["fract", "modf", "sincos"] in { - def : Builtin<name, [FGenTypeN, FGenTypeN, PointerType<FGenTypeN, GenericAS>]>; - } -@@ -501,6 +509,7 @@ let MinVersion = CL20 in { - def : Builtin<name, [Type, Type, Type, PointerType<GenTypeIntVecAndScalar, GenericAS>]>; - } - } -+ } - } - - // --- Table 9 --- -@@ -531,7 +540,7 @@ foreach name = ["abs"] in { - foreach name = ["clz", "popcount"] in { - def : Builtin<name, [AIGenTypeN, AIGenTypeN], Attr.Const>; - } --let MinVersion = CL20 in { -+let Versions = [CL20, CL30] in { - foreach name = ["ctz"] in { - def : Builtin<name, [AIGenTypeN, AIGenTypeN]>; - } -@@ -705,7 +714,7 @@ foreach name = ["select"] in { - // --- Table 15 --- - // Variants for OpenCL versions below 2.0, using pointers to the global, local - // and private address spaces. --let MaxVersion = CL20 in { -+let Versions = [CL10, CL11, CL12, CL30] in { - foreach AS = [GlobalAS, LocalAS, PrivateAS] in { - foreach VSize = [2, 3, 4, 8, 16] in { - foreach name = ["vload" # VSize] in { -@@ -748,7 +757,8 @@ let MaxVersion = CL20 in { - } - // Variants for OpenCL versions above 2.0, using pointers to the generic - // address space. --let MinVersion = CL20 in { -+let Versions = [CL20, CL30] in { -+ let Extension = FuncExtGenericAddressSpace in { - foreach VSize = [2, 3, 4, 8, 16] in { - foreach name = ["vload" # VSize] in { - def : Builtin<name, [VectorType<Char, VSize>, Size, PointerType<ConstType<Char>, GenericAS>]>; -@@ -786,6 +796,7 @@ let MinVersion = CL20 in { - } - } - } -+ } - } - // Variants using pointers to the constant address space. - foreach VSize = [2, 3, 4, 8, 16] in { -@@ -812,7 +823,7 @@ foreach VSize = [2, 3, 4, 8, 16] in { - } - } - } --let MaxVersion = CL20 in { -+let Versions = [CL10, CL11, CL12, CL30] in { - foreach AS = [GlobalAS, LocalAS, PrivateAS] in { - def : Builtin<"vload_half", [Float, Size, PointerType<ConstType<Half>, AS>]>; - foreach VSize = [2, 3, 4, 8, 16] in { -@@ -832,7 +843,8 @@ let MaxVersion = CL20 in { - } - } - } --let MinVersion = CL20 in { -+let Versions = [CL20, CL30] in { -+ let Extension = FuncExtGenericAddressSpace in { - foreach AS = [GenericAS] in { - def : Builtin<"vload_half", [Float, Size, PointerType<ConstType<Half>, AS>]>; - foreach VSize = [2, 3, 4, 8, 16] in { -@@ -851,6 +863,7 @@ let MinVersion = CL20 in { - } - } - } -+ } - } - - foreach AS = [ConstantAS] in { -@@ -875,7 +888,9 @@ foreach name = ["async_work_group_strided_copy"] in { - def : Builtin<name, [Event, PointerType<AGenTypeN, GlobalAS>, PointerType<ConstType<AGenTypeN>, LocalAS>, Size, Size, Event]>; - } - foreach name = ["wait_group_events"] in { -+ let Extension = FuncExtGenericAddressSpace in { - def : Builtin<name, [Void, Int, PointerType<Event, GenericAS>]>; -+ } - } - foreach name = ["prefetch"] in { - def : Builtin<name, [Void, PointerType<ConstType<AGenTypeN>, GlobalAS>, Size]>; -@@ -1154,7 +1169,8 @@ foreach aQual = ["WO", "RW"] in { - //-------------------------------------------------------------------- - // OpenCL v2.0 s6.13.15 - Work-group Functions - // --- Table 26 --- --let MinVersion = CL20 in { -+let Versions = [CL20, CL30] in { -+ let Extension = FuncExtWorkGroupCollective in { - foreach name = ["work_group_all", "work_group_any"] in { - def : Builtin<name, [Int, Int], Attr.Convergent>; - } -@@ -1169,11 +1185,12 @@ let MinVersion = CL20 in { - def : Builtin<name # op, [IntLongFloatGenType1, IntLongFloatGenType1], Attr.Convergent>; - } - } -+ } - } - - - // OpenCL v2.0 s9.17.3: Additions to section 6.13.1: Work-Item Functions --let MinVersion = CL20 in { -+let Versions = [CL20] in { - let Extension = FuncExtKhrSubgroups in { - def get_sub_group_size : Builtin<"get_sub_group_size", [UInt]>; - def get_max_sub_group_size : Builtin<"get_max_sub_group_size", [UInt]>; -diff --git a/clang/lib/Sema/Sema.cpp b/clang/lib/Sema/Sema.cpp -index 9cfce5a63b1d..c26f45f62668 100644 ---- a/clang/lib/Sema/Sema.cpp -+++ b/clang/lib/Sema/Sema.cpp -@@ -286,9 +286,10 @@ void Sema::Initialize() { - // Initialize predefined OpenCL types and supported extensions and (optional) - // core features. - if (getLangOpts().OpenCL) { -+ getOpenCLOptions().setOpenCLVersion(getLangOpts()); - getOpenCLOptions().addSupport( - Context.getTargetInfo().getSupportedOpenCLOpts()); -- getOpenCLOptions().enableSupportedCore(getLangOpts()); -+ getOpenCLOptions().enableSupportedCore(); - addImplicitTypedef("sampler_t", Context.OCLSamplerTy); - addImplicitTypedef("event_t", Context.OCLEventTy); - if (getLangOpts().OpenCLCPlusPlus || getLangOpts().OpenCLVersion >= 200) { -@@ -344,12 +345,18 @@ void Sema::Initialize() { - "cl_khr_int64_base_atomics cl_khr_int64_extended_atomics"); - - setOpenCLExtensionForType(AtomicDoubleT, "cl_khr_fp64"); -+ setOpenCLExtensionForType(Context.OCLReserveIDTy, "__opencl_c_pipes"); -+ setOpenCLExtensionForType(Context.OCLClkEventTy, -+ "__opencl_c_device_enqueue"); -+ setOpenCLExtensionForType(Context.OCLQueueTy, -+ "__opencl_c_device_enqueue"); - } - - setOpenCLExtensionForType(Context.DoubleTy, "cl_khr_fp64"); - --#define GENERIC_IMAGE_TYPE_EXT(Type, Id, Ext) \ -- setOpenCLExtensionForType(Context.Id, Ext); -+#define GENERIC_IMAGE_TYPE_EXT(Type, Id, Ext) \ -+ setOpenCLExtensionForType(Context.Id, Ext); \ -+ setOpenCLExtensionForType(Context.Id, "__opencl_c_images"); - #include "clang/Basic/OpenCLImageTypes.def" - #define EXT_OPAQUE_TYPE(ExtType, Id, Ext) \ - addImplicitTypedef(#ExtType, Context.Id##Ty); \ -@@ -2293,6 +2300,27 @@ bool Sema::isOpenCLDisabledDecl(Decl *FD) { - return false; - } - -+template <typename DiagLocT, typename DiagInfoT> -+void Sema::DiagnoseOpenCLRequiresOption(llvm::StringRef OpenCLOptName, -+ DiagLocT DiagLoc, DiagInfoT DiagInfo, -+ unsigned Selector, -+ SourceRange SrcRange) { -+ const auto &LO = getLangOpts(); -+ auto CLVer = LO.OpenCLCPlusPlus ? 200 : LO.OpenCLVersion; -+ // For versions higher that 3.0 diagnosing feature -+ if (CLVer >= 300) { -+ OpenCLOptName = -+ llvm::StringSwitch<llvm::StringRef>(OpenCLOptName) -+ .Case("cl_khr_3d_image_writes", "__opencl_c_3d_image_writes") -+ .Case("cl_khr_subgroups", "__opencl_c_subgroups") -+ .Case("cl_khr_fp64", "__opencl_c_fp64") -+ .Default(OpenCLOptName); -+ } -+ -+ Diag(DiagLoc, diag::err_opencl_requires_extension) -+ << Selector << DiagInfo << OpenCLOptName << SrcRange; -+} -+ - template <typename T, typename DiagLocT, typename DiagInfoT, typename MapT> - bool Sema::checkOpenCLDisabledTypeOrDecl(T D, DiagLocT DiagLoc, - DiagInfoT DiagInfo, MapT &Map, -@@ -2304,8 +2332,7 @@ bool Sema::checkOpenCLDisabledTypeOrDecl(T D, DiagLocT DiagLoc, - bool Disabled = false; - for (auto &I : Loc->second) { - if (I != CurrOpenCLExtension && !getOpenCLOptions().isEnabled(I)) { -- Diag(DiagLoc, diag::err_opencl_requires_extension) << Selector << DiagInfo -- << I << SrcRange; -+ DiagnoseOpenCLRequiresOption(I, DiagLoc, DiagInfo, Selector, SrcRange); - Disabled = true; - } - } -@@ -2341,3 +2368,13 @@ bool Sema::checkOpenCLDisabledDecl(const NamedDecl &D, const Expr &E) { - return checkOpenCLDisabledTypeOrDecl(&D, E.getBeginLoc(), FnName, - OpenCLDeclExtMap, 1, D.getSourceRange()); - } -+ -+bool Sema::checkOpenCLSubgroupExtForCallExpr(CallExpr *Call) { -+ if (!getOpenCLOptions().isEnabled("cl_khr_subgroups")) { -+ DiagnoseOpenCLRequiresOption("cl_khr_subgroups", Call->getBeginLoc(), -+ Call->getDirectCallee(), 1, -+ Call->getSourceRange()); -+ return true; -+ } -+ return false; -+} -diff --git a/clang/lib/Sema/SemaChecking.cpp b/clang/lib/Sema/SemaChecking.cpp -index 74742023d1b3..efa3b6ab0eb6 100644 ---- a/clang/lib/Sema/SemaChecking.cpp -+++ b/clang/lib/Sema/SemaChecking.cpp -@@ -597,20 +597,11 @@ static bool checkOpenCLBlockArgs(Sema &S, Expr *BlockArg) { - return IllegalParams; - } - --static bool checkOpenCLSubgroupExt(Sema &S, CallExpr *Call) { -- if (!S.getOpenCLOptions().isEnabled("cl_khr_subgroups")) { -- S.Diag(Call->getBeginLoc(), diag::err_opencl_requires_extension) -- << 1 << Call->getDirectCallee() << "cl_khr_subgroups"; -- return true; -- } -- return false; --} -- - static bool SemaOpenCLBuiltinNDRangeAndBlock(Sema &S, CallExpr *TheCall) { - if (checkArgCount(S, TheCall, 2)) - return true; - -- if (checkOpenCLSubgroupExt(S, TheCall)) -+ if (S.checkOpenCLSubgroupExtForCallExpr(TheCall)) - return true; - - // First argument is an ndrange_t type. -@@ -1564,7 +1555,7 @@ Sema::CheckBuiltinFunctionCall(FunctionDecl *FDecl, unsigned BuiltinID, - break; - case Builtin::BIsub_group_reserve_read_pipe: - case Builtin::BIsub_group_reserve_write_pipe: -- if (checkOpenCLSubgroupExt(*this, TheCall) || -+ if (checkOpenCLSubgroupExtForCallExpr(TheCall) || - SemaBuiltinReserveRWPipe(*this, TheCall)) - return ExprError(); - break; -@@ -1577,7 +1568,7 @@ Sema::CheckBuiltinFunctionCall(FunctionDecl *FDecl, unsigned BuiltinID, - break; - case Builtin::BIsub_group_commit_read_pipe: - case Builtin::BIsub_group_commit_write_pipe: -- if (checkOpenCLSubgroupExt(*this, TheCall) || -+ if (checkOpenCLSubgroupExtForCallExpr(TheCall) || - SemaBuiltinCommitRWPipe(*this, TheCall)) - return ExprError(); - break; -@@ -4314,6 +4305,20 @@ DiagnoseCStringFormatDirectiveInCFAPI(Sema &S, - } - } - -+bool Sema::isSupportedOpenCLOMemoryOrdering(int64_t Ordering) const { -+ assert(llvm::isValidAtomicOrderingCABI(Ordering)); -+ auto OrderingCABI = (llvm::AtomicOrderingCABI)Ordering; -+ switch (OrderingCABI) { -+ case llvm::AtomicOrderingCABI::acquire: -+ case llvm::AtomicOrderingCABI::release: -+ case llvm::AtomicOrderingCABI::acq_rel: -+ return OpenCLFeatures.isEnabled("__opencl_c_atomic_order_acq_rel"); -+ -+ default: -+ return true; -+ } -+} -+ - /// Determine whether the given type has a non-null nullability annotation. - static bool isNonNullType(ASTContext &ctx, QualType type) { - if (auto nullability = type->getNullability(ctx)) -@@ -5067,10 +5072,17 @@ ExprResult Sema::BuildAtomicExpr(SourceRange CallRange, SourceRange ExprRange, - if (SubExprs.size() >= 2 && Form != Init) { - llvm::APSInt Result(32); - if (SubExprs[1]->isIntegerConstantExpr(Result, Context) && -- !isValidOrderingForOp(Result.getSExtValue(), Op)) -+ !isValidOrderingForOp(Result.getSExtValue(), Op)) { - Diag(SubExprs[1]->getBeginLoc(), - diag::warn_atomic_op_has_invalid_memory_order) - << SubExprs[1]->getSourceRange(); -+ } else if (IsOpenCL && -+ !isSupportedOpenCLOMemoryOrdering(Result.getSExtValue())) { -+ Diag(SubExprs[1]->getBeginLoc(), -+ diag::err_opencl_memory_ordering_require_feat) -+ << SubExprs[1]->getSourceRange(); -+ return ExprError(); -+ } - } - - if (auto ScopeModel = AtomicExpr::getScopeModel(Op)) { -diff --git a/clang/lib/Sema/SemaDecl.cpp b/clang/lib/Sema/SemaDecl.cpp -index 64146f4a912f..c1e629b5dc76 100644 ---- a/clang/lib/Sema/SemaDecl.cpp -+++ b/clang/lib/Sema/SemaDecl.cpp -@@ -6152,7 +6152,9 @@ void Sema::deduceOpenCLAddressSpace(ValueDecl *Decl) { - if (Type->isSamplerT() || Type->isVoidType()) - return; - LangAS ImplAS = LangAS::opencl_private; -- if ((getLangOpts().OpenCLCPlusPlus || getLangOpts().OpenCLVersion >= 200) && -+ if ((getLangOpts().OpenCLCPlusPlus || -+ getOpenCLOptions().isEnabled( -+ "__opencl_c_program_scope_global_variables")) && - Var->hasGlobalStorage()) - ImplAS = LangAS::opencl_global; - // If the original type from a decayed type is an array type and that array -@@ -7682,6 +7684,10 @@ void Sema::CheckVariableDeclarationType(VarDecl *NewVD) { - // OpenCL C v2.0 s6.5.1 - Variables defined at program scope and static - // variables inside a function can also be declared in the global - // address space. -+ // OpenCL C v3.0 s5.5 - For OpenCL C 2.0, or with the -+ // __opencl_c_program_scope_global_variables feature macro, the -+ // address space for a variable at program scope or a static variable -+ // also be __global - // C++ for OpenCL inherits rule from OpenCL C v2.0. - // FIXME: Adding local AS in C++ for OpenCL might make sense. - if (NewVD->isFileVarDecl() || NewVD->isStaticLocal() || -@@ -7689,10 +7695,11 @@ void Sema::CheckVariableDeclarationType(VarDecl *NewVD) { - if (!T->isSamplerT() && - !(T.getAddressSpace() == LangAS::opencl_constant || - (T.getAddressSpace() == LangAS::opencl_global && -- (getLangOpts().OpenCLVersion == 200 || -- getLangOpts().OpenCLCPlusPlus)))) { -+ (OpenCLFeatures.isEnabled( -+ "__opencl_c_program_scope_global_variables"))))) { - int Scope = NewVD->isStaticLocal() | NewVD->hasExternalStorage() << 1; -- if (getLangOpts().OpenCLVersion == 200 || getLangOpts().OpenCLCPlusPlus) -+ if (OpenCLFeatures.isEnabled( -+ "__opencl_c_program_scope_global_variables")) - Diag(NewVD->getLocation(), diag::err_opencl_global_invalid_addr_space) - << Scope << "global or constant"; - else -diff --git a/clang/lib/Sema/SemaDeclAttr.cpp b/clang/lib/Sema/SemaDeclAttr.cpp -index 849bc09063b3..c963b95b131a 100644 ---- a/clang/lib/Sema/SemaDeclAttr.cpp -+++ b/clang/lib/Sema/SemaDeclAttr.cpp -@@ -6362,7 +6362,7 @@ static void handleInternalLinkageAttr(Sema &S, Decl *D, const ParsedAttr &AL) { - } - - static void handleOpenCLNoSVMAttr(Sema &S, Decl *D, const ParsedAttr &AL) { -- if (S.LangOpts.OpenCLVersion != 200) -+ if (S.LangOpts.OpenCLVersion < 200) - S.Diag(AL.getLoc(), diag::err_attribute_requires_opencl_version) - << AL << "2.0" << 0; - else -@@ -6446,6 +6446,13 @@ static void handleOpenCLAccessAttr(Sema &S, Decl *D, const ParsedAttr &AL) { - << AL << PDecl->getType() << DeclTy->isImageType(); - D->setInvalidDecl(true); - return; -+ } else if ((!S.getLangOpts().OpenCLCPlusPlus && -+ S.getLangOpts().OpenCLVersion >= 200) && -+ !S.getOpenCLOptions().isEnabled( -+ "__opencl_c_read_write_images")) { -+ S.Diag(AL.getLoc(), diag::err_opencl_requires_extension) -+ << 0 << PDecl->getType() << "__opencl_c_read_write_images"; -+ return; - } - } - } -diff --git a/clang/lib/Sema/SemaDeclCXX.cpp b/clang/lib/Sema/SemaDeclCXX.cpp -index 831e55046e80..4481a59b4517 100644 ---- a/clang/lib/Sema/SemaDeclCXX.cpp -+++ b/clang/lib/Sema/SemaDeclCXX.cpp -@@ -14906,6 +14906,11 @@ CheckOperatorNewDeleteTypes(Sema &SemaRef, const FunctionDecl *FnDecl, - if (auto *PtrTy = ResultType->getAs<PointerType>()) { - ResultType = RemoveAddressSpaceFromPtr(SemaRef, PtrTy); - } -+ if (CanQual<PointerType> ExpectedPtrTy = -+ ExpectedResultType->getAs<PointerType>()) { -+ ExpectedResultType = SemaRef.Context.getCanonicalType( -+ RemoveAddressSpaceFromPtr(SemaRef, ExpectedPtrTy->getTypePtr())); -+ } - } - - // Check that the result type is what we expect. -@@ -14939,6 +14944,11 @@ CheckOperatorNewDeleteTypes(Sema &SemaRef, const FunctionDecl *FnDecl, - FnDecl->getParamDecl(0)->getType()->getAs<PointerType>()) { - FirstParamType = RemoveAddressSpaceFromPtr(SemaRef, PtrTy); - } -+ if (CanQual<PointerType> ExpectedPtrTy = -+ ExpectedFirstParamType->getAs<PointerType>()) { -+ ExpectedFirstParamType = SemaRef.Context.getCanonicalType( -+ RemoveAddressSpaceFromPtr(SemaRef, ExpectedPtrTy->getTypePtr())); -+ } - } - if (SemaRef.Context.getCanonicalType(FirstParamType).getUnqualifiedType() != - ExpectedFirstParamType) -diff --git a/clang/lib/Sema/SemaLookup.cpp b/clang/lib/Sema/SemaLookup.cpp -index 8d96404a5c27..8f21203b9508 100644 ---- a/clang/lib/Sema/SemaLookup.cpp -+++ b/clang/lib/Sema/SemaLookup.cpp -@@ -771,19 +771,20 @@ static void InsertOCLBuiltinDeclarationsFromTable(Sema &S, LookupResult &LR, - // as argument. Only meaningful for generic types, otherwise equals 1. - unsigned GenTypeMaxCnt; - -+ ASTContext &Context = S.Context; -+ -+ // Ignore this BIF if its version does not match the language options. -+ unsigned OpenCLVersion = Context.getLangOpts().OpenCLVersion; -+ if (Context.getLangOpts().OpenCLCPlusPlus) -+ OpenCLVersion = 200; -+ -+ unsigned short VersionCode = OpenCLOptions::EncodeVersion(OpenCLVersion); -+ - for (unsigned SignatureIndex = 0; SignatureIndex < Len; SignatureIndex++) { - const OpenCLBuiltinStruct &OpenCLBuiltin = - BuiltinTable[FctIndex + SignatureIndex]; -- ASTContext &Context = S.Context; - -- // Ignore this BIF if its version does not match the language options. -- unsigned OpenCLVersion = Context.getLangOpts().OpenCLVersion; -- if (Context.getLangOpts().OpenCLCPlusPlus) -- OpenCLVersion = 200; -- if (OpenCLVersion < OpenCLBuiltin.MinVersion) -- continue; -- if ((OpenCLBuiltin.MaxVersion != 0) && -- (OpenCLVersion >= OpenCLBuiltin.MaxVersion)) -+ if (!(OpenCLBuiltin.AllVersions & VersionCode)) - continue; - - SmallVector<QualType, 1> RetTypes; -diff --git a/clang/lib/Sema/SemaType.cpp b/clang/lib/Sema/SemaType.cpp -index 93ddd047e09b..93211b801f8d 100644 ---- a/clang/lib/Sema/SemaType.cpp -+++ b/clang/lib/Sema/SemaType.cpp -@@ -2016,7 +2016,7 @@ static QualType deduceOpenCLPointeeAddrSpace(Sema &S, QualType PointeeType) { - !PointeeType.hasAddressSpace()) - PointeeType = S.getASTContext().getAddrSpaceQualType( - PointeeType, -- S.getLangOpts().OpenCLCPlusPlus || S.getLangOpts().OpenCLVersion == 200 -+ S.getOpenCLOptions().isEnabled("__opencl_c_generic_address_space") - ? LangAS::opencl_generic - : LangAS::opencl_private); - return PointeeType; -@@ -5160,9 +5160,15 @@ static TypeSourceInfo *GetFullTypeForDeclarator(TypeProcessingState &state, - } - - case DeclaratorChunk::Pipe: { -- T = S.BuildReadPipeType(T, DeclType.Loc); -- processTypeAttrs(state, T, TAL_DeclSpec, -- D.getMutableDeclSpec().getAttributes()); -+ if (S.getOpenCLOptions().isEnabled("__opencl_c_pipes")) { -+ T = S.BuildReadPipeType(T, DeclType.Loc); -+ processTypeAttrs(state, T, TAL_DeclSpec, -+ D.getMutableDeclSpec().getAttributes()); -+ } else { -+ D.setInvalidType(true); -+ T = Context.IntTy; -+ S.Diag(D.getIdentifierLoc(), diag::err_opencl_pipes_require_feat); -+ } - break; - } - } -diff --git a/clang/test/CodeGenOpenCL/addr-space-struct-arg.cl b/clang/test/CodeGenOpenCL/addr-space-struct-arg.cl -index cdbf28bbcad8..0bedff5ef0f3 100644 ---- a/clang/test/CodeGenOpenCL/addr-space-struct-arg.cl -+++ b/clang/test/CodeGenOpenCL/addr-space-struct-arg.cl -@@ -2,7 +2,8 @@ - // RUN: %clang_cc1 %s -emit-llvm -o - -O0 -triple amdgcn | FileCheck -enable-var-scope -check-prefixes=COM,AMDGCN %s - // RUN: %clang_cc1 %s -emit-llvm -o - -cl-std=CL2.0 -O0 -triple amdgcn | FileCheck -enable-var-scope -check-prefixes=COM,AMDGCN,AMDGCN20 %s - // RUN: %clang_cc1 %s -emit-llvm -o - -cl-std=CL1.2 -O0 -triple spir-unknown-unknown-unknown | FileCheck -enable-var-scope -check-prefixes=SPIR %s -- -+// RUN: %clang_cc1 %s -emit-llvm -o - -cl-std=CL3.0 -O0 -triple spir-unknown-unknown-unknown | FileCheck -enable-var-scope -check-prefixes=SPIR %s -+// RUN: %clang_cc1 %s -emit-llvm -o - -cl-std=CL3.0 -cl-ext=__opencl_c_program_scope_global_variables -O0 -triple spir-unknown-unknown-unknown | FileCheck -enable-var-scope -check-prefixes=SPIR %s - typedef int int2 __attribute__((ext_vector_type(2))); - - typedef struct { -@@ -39,7 +40,7 @@ struct LargeStructTwoMember { - int2 y[20]; - }; - --#if __OPENCL_C_VERSION__ >= 200 -+#ifdef __opencl_c_program_scope_global_variables - struct LargeStructOneMember g_s; - #endif - -@@ -98,7 +99,7 @@ void FuncOneLargeMember(struct LargeStructOneMember u) { - // AMDGCN20: %[[r0:.*]] = bitcast %struct.LargeStructOneMember addrspace(5)* %[[byval_temp]] to i8 addrspace(5)* - // AMDGCN20: call void @llvm.memcpy.p5i8.p1i8.i64(i8 addrspace(5)* align 8 %[[r0]], i8 addrspace(1)* align 8 bitcast (%struct.LargeStructOneMember addrspace(1)* @g_s to i8 addrspace(1)*), i64 800, i1 false) - // AMDGCN20: call void @FuncOneLargeMember(%struct.LargeStructOneMember addrspace(5)* byval(%struct.LargeStructOneMember) align 8 %[[byval_temp]]) --#if __OPENCL_C_VERSION__ >= 200 -+#ifdef __opencl_c_program_scope_global_variables - void test_indirect_arg_globl(void) { - FuncOneLargeMember(g_s); - } -diff --git a/clang/test/CodeGenOpenCL/address-spaces-conversions.cl b/clang/test/CodeGenOpenCL/address-spaces-conversions.cl -index 52feccc540bb..9ecffcca5ee9 100644 ---- a/clang/test/CodeGenOpenCL/address-spaces-conversions.cl -+++ b/clang/test/CodeGenOpenCL/address-spaces-conversions.cl -@@ -1,5 +1,7 @@ - // RUN: %clang_cc1 %s -triple x86_64-unknown-linux-gnu -O0 -ffake-address-space-map -cl-std=CL2.0 -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 %s -triple x86_64-unknown-linux-gnu -O0 -ffake-address-space-map -cl-std=CL3.0 -cl-ext=__opencl_c_generic_address_space -emit-llvm -o - | FileCheck %s - // RUN: %clang_cc1 %s -triple x86_64-unknown-linux-gnu -O0 -cl-std=CL2.0 -emit-llvm -o - | FileCheck --check-prefix=CHECK-NOFAKE %s -+// RUN: %clang_cc1 %s -triple x86_64-unknown-linux-gnu -O0 -cl-std=CL3.0 -cl-ext=__opencl_c_generic_address_space -emit-llvm -o - | FileCheck --check-prefix=CHECK-NOFAKE %s - // When -ffake-address-space-map is not used, all addr space mapped to 0 for x86_64. - - // test that we generate address space casts everywhere we need conversions of -diff --git a/clang/test/CodeGenOpenCL/address-spaces-mangling.cl b/clang/test/CodeGenOpenCL/address-spaces-mangling.cl -index 50622f099143..e19ec8451d0d 100644 ---- a/clang/test/CodeGenOpenCL/address-spaces-mangling.cl -+++ b/clang/test/CodeGenOpenCL/address-spaces-mangling.cl -@@ -6,6 +6,7 @@ - // We check that the address spaces are mangled the same in both version of OpenCL - // RUN: %clang_cc1 %s -triple spir-unknown-unknown -cl-std=CL2.0 -emit-llvm -o - | FileCheck -check-prefix=OCL-20 %s - // RUN: %clang_cc1 %s -triple spir-unknown-unknown -cl-std=CL1.2 -emit-llvm -o - | FileCheck -check-prefix=OCL-12 %s -+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -cl-std=CL3.0 -emit-llvm -o - | FileCheck -check-prefix=OCL-12 %s - - // We can't name this f as private is equivalent to default - // no specifier given address space so we get multiple definition -@@ -47,7 +48,7 @@ void f(constant int *arg) { } - // OCL-20-DAG: @_Z1fPU3AS2i - // OCL-12-DAG: @_Z1fPU3AS2i - --#if __OPENCL_C_VERSION__ >= 200 -+#if __OPENCL_C_VERSION__ == 200 - __attribute__((overloadable)) - void f(generic int *arg) { } - // ASMANG20: @_Z1fPU3AS4i -diff --git a/clang/test/CodeGenOpenCL/address-spaces.cl b/clang/test/CodeGenOpenCL/address-spaces.cl -index 3c8fea2a80bc..26a741338b3a 100644 ---- a/clang/test/CodeGenOpenCL/address-spaces.cl -+++ b/clang/test/CodeGenOpenCL/address-spaces.cl -@@ -1,9 +1,13 @@ - // RUN: %clang_cc1 %s -O0 -ffake-address-space-map -emit-llvm -o - | FileCheck %s --check-prefixes=CHECK,SPIR -+// RUN: %clang_cc1 %s -O0 -cl-std=CL3.0 -ffake-address-space-map -emit-llvm -o - | FileCheck %s --check-prefixes=CHECK,SPIR - // RUN: %clang_cc1 %s -O0 -DCL20 -cl-std=CL2.0 -ffake-address-space-map -emit-llvm -o - | FileCheck %s --check-prefixes=CL20,CL20SPIR - // RUN: %clang_cc1 %s -O0 -triple amdgcn-amd-amdhsa -emit-llvm -o - | FileCheck --check-prefixes=CHECK,AMDGCN %s -+// RUN: %clang_cc1 %s -O0 -cl-std=CL3.0 -triple amdgcn-amd-amdhsa -emit-llvm -o - | FileCheck --check-prefixes=CHECK,AMDGCN %s - // RUN: %clang_cc1 %s -O0 -triple amdgcn-amd-amdhsa -DCL20 -cl-std=CL2.0 -emit-llvm -o - | FileCheck %s --check-prefixes=CL20,CL20AMDGCN - // RUN: %clang_cc1 %s -O0 -triple amdgcn-mesa-mesa3d -emit-llvm -o - | FileCheck --check-prefixes=CHECK,AMDGCN %s -+// RUN: %clang_cc1 %s -O0 -cl-std=CL3.0 -triple amdgcn-mesa-mesa3d -emit-llvm -o - | FileCheck --check-prefixes=CHECK,AMDGCN %s - // RUN: %clang_cc1 %s -O0 -triple r600-- -emit-llvm -o - | FileCheck --check-prefixes=CHECK,AMDGCN %s -+// RUN: %clang_cc1 %s -O0 -cl-std=CL3.0 -triple r600-- -emit-llvm -o - | FileCheck --check-prefixes=CHECK,AMDGCN %s - - // SPIR: %struct.S = type { i32, i32, i32* } - // CL20SPIR: %struct.S = type { i32, i32, i32 addrspace(4)* } -diff --git a/clang/test/CodeGenOpenCL/amdgcn-automatic-variable.cl b/clang/test/CodeGenOpenCL/amdgcn-automatic-variable.cl -index 7216cb517495..8d3b30fe8074 100644 ---- a/clang/test/CodeGenOpenCL/amdgcn-automatic-variable.cl -+++ b/clang/test/CodeGenOpenCL/amdgcn-automatic-variable.cl -@@ -1,4 +1,5 @@ - // RUN: %clang_cc1 -O0 -cl-std=CL1.2 -triple amdgcn---amdgizcl -emit-llvm %s -o - | FileCheck -check-prefixes=CHECK,CL12 %s -+// RUN: %clang_cc1 -O0 -cl-std=CL3.0 -triple amdgcn---amdgizcl -emit-llvm %s -o - | FileCheck -check-prefixes=CHECK,CL12 %s - // RUN: %clang_cc1 -O0 -cl-std=CL2.0 -triple amdgcn---amdgizcl -emit-llvm %s -o - | FileCheck -check-prefixes=CHECK,CL20 %s - - // CL12-LABEL: define void @func1(i32 addrspace(5)* %x) -diff --git a/clang/test/CodeGenOpenCL/amdgpu-sizeof-alignof.cl b/clang/test/CodeGenOpenCL/amdgpu-sizeof-alignof.cl -index a5d438933fa4..8073c7756eb6 100644 ---- a/clang/test/CodeGenOpenCL/amdgpu-sizeof-alignof.cl -+++ b/clang/test/CodeGenOpenCL/amdgpu-sizeof-alignof.cl -@@ -4,6 +4,17 @@ - // RUN: %clang_cc1 -triple amdgcn---opencl -cl-std=CL2.0 %s -emit-llvm -o - | FileCheck %s - // RUN: %clang_cc1 -triple amdgcn---amdgizcl -cl-std=CL1.2 %s -emit-llvm -o - | FileCheck %s - // RUN: %clang_cc1 -triple amdgcn---amdgizcl -cl-std=CL2.0 %s -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 -triple r600 -cl-std=CL3.0 %s -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 -triple amdgcn-mesa-mesa3d -cl-std=CL3.0 %s -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 -triple amdgcn---opencl -cl-std=CL3.0 %s -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 -triple amdgcn---amdgizcl -cl-std=CL3.0 %s -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 -triple amdgcn-mesa-mesa3d -cl-ext=__opencl_c_generic_address_space -cl-std=CL3.0 %s -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 -triple amdgcn---opencl -cl-ext=__opencl_c_generic_address_space -cl-std=CL3.0 %s -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 -triple amdgcn---amdgizcl -cl-ext=__opencl_c_generic_address_space -cl-std=CL3.0 %s -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 -triple r600 -cl-ext=__opencl_c_fp64 -cl-std=CL3.0 %s -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 -triple amdgcn-mesa-mesa3d -cl-ext=__opencl_c_fp64 -cl-std=CL3.0 %s -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 -triple amdgcn---opencl -cl-ext=__opencl_c_fp64 -cl-std=CL3.0 %s -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 -triple amdgcn---amdgizcl -cl-ext=__opencl_c_fp64 -cl-std=CL3.0 %s -emit-llvm -o - | FileCheck %s - - #ifdef __AMDGCN__ - #define PTSIZE 8 -@@ -11,7 +22,7 @@ - #define PTSIZE 4 - #endif - --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #pragma OPENCL EXTENSION cl_khr_fp64 : enable - #endif - #ifdef cl_khr_fp16 -@@ -59,8 +70,12 @@ void test() { - check(__alignof__(double) == 8); - #endif - -- check(sizeof(void*) == (__OPENCL_C_VERSION__ >= 200 ? 8 : 4)); -- check(__alignof__(void*) == (__OPENCL_C_VERSION__ >= 200 ? 8 : 4)); -+ check(sizeof(private void*) == 4); -+ check(__alignof__(private void*) == 4); -+#ifdef __opencl_c_generic_address_space -+ check(sizeof(generic void*) == 8); -+ check(__alignof__(generic void*) == 8); -+#endif - check(sizeof(global_ptr_t) == PTSIZE); - check(__alignof__(global_ptr_t) == PTSIZE); - check(sizeof(constant_ptr_t) == PTSIZE); -diff --git a/clang/test/CodeGenOpenCL/arm-integer-dot-product.cl b/clang/test/CodeGenOpenCL/arm-integer-dot-product.cl -index d1ab6aceac5c..70c5bace023b 100644 ---- a/clang/test/CodeGenOpenCL/arm-integer-dot-product.cl -+++ b/clang/test/CodeGenOpenCL/arm-integer-dot-product.cl -@@ -1,4 +1,5 @@ - // RUN: %clang_cc1 %s -triple spir-unknown-unknown -finclude-default-header -cl-std=CL1.2 -emit-llvm -o - -O0 | FileCheck %s -+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -finclude-default-header -cl-std=CL3.0 -emit-llvm -o - -O0 | FileCheck %s - - #pragma OPENCL EXTENSION cl_arm_integer_dot_product_int8 : enable - void test_int8(uchar4 ua, uchar4 ub, char4 sa, char4 sb) { -diff --git a/clang/test/CodeGenOpenCL/cl-uniform-wg-size.cl b/clang/test/CodeGenOpenCL/cl-uniform-wg-size.cl -index 76ace5dca21e..5dc43e222f75 100644 ---- a/clang/test/CodeGenOpenCL/cl-uniform-wg-size.cl -+++ b/clang/test/CodeGenOpenCL/cl-uniform-wg-size.cl -@@ -1,6 +1,8 @@ - // RUN: %clang_cc1 -emit-llvm -O0 -cl-std=CL1.2 -o - %s 2>&1 | FileCheck %s -check-prefixes CHECK,CHECK-UNIFORM - // RUN: %clang_cc1 -emit-llvm -O0 -cl-std=CL2.0 -o - %s 2>&1 | FileCheck %s -check-prefixes CHECK,CHECK-NONUNIFORM - // RUN: %clang_cc1 -emit-llvm -O0 -cl-std=CL2.0 -cl-uniform-work-group-size -o - %s 2>&1 | FileCheck %s -check-prefixes CHECK,CHECK-UNIFORM -+// RUN: %clang_cc1 -emit-llvm -O0 -cl-std=CL3.0 -o - %s 2>&1 | FileCheck %s -check-prefixes CHECK,CHECK-NONUNIFORM -+// RUN: %clang_cc1 -emit-llvm -O0 -cl-std=CL3.0 -cl-uniform-work-group-size -o - %s 2>&1 | FileCheck %s -check-prefixes CHECK,CHECK-UNIFORM - - kernel void ker() {}; - // CHECK: define{{.*}}@ker() #0 -diff --git a/clang/test/CodeGenOpenCL/fpmath.cl b/clang/test/CodeGenOpenCL/fpmath.cl -index 0108d909c94e..b28392739c71 100644 ---- a/clang/test/CodeGenOpenCL/fpmath.cl -+++ b/clang/test/CodeGenOpenCL/fpmath.cl -@@ -2,6 +2,8 @@ - // RUN: %clang_cc1 %s -emit-llvm -o - -triple spir-unknown-unknown -cl-fp32-correctly-rounded-divide-sqrt | FileCheck --check-prefix=CHECK --check-prefix=DIVOPT %s - // RUN: %clang_cc1 %s -emit-llvm -o - -DNOFP64 -cl-std=CL1.2 -triple r600-unknown-unknown -target-cpu r600 -pedantic | FileCheck --check-prefix=CHECK-FLT %s - // RUN: %clang_cc1 %s -emit-llvm -o - -DFP64 -cl-std=CL1.2 -triple spir-unknown-unknown -pedantic | FileCheck --check-prefix=CHECK-DBL %s -+// RUN: %clang_cc1 %s -emit-llvm -o - -DNOFP64 -cl-std=CL3.0 -triple r600-unknown-unknown -target-cpu r600 -pedantic | FileCheck --check-prefix=CHECK-FLT %s -+// RUN: %clang_cc1 %s -emit-llvm -o - -DFP64 -cl-std=CL3.0 -cl-ext=__opencl_c_fp64 -triple spir-unknown-unknown -pedantic | FileCheck --check-prefix=CHECK-DBL %s - - typedef __attribute__(( ext_vector_type(4) )) float float4; - -diff --git a/clang/test/CodeGenOpenCL/generic-address-space-feature.cl b/clang/test/CodeGenOpenCL/generic-address-space-feature.cl -new file mode 100644 -index 000000000000..890860abe4d9 ---- /dev/null -+++ b/clang/test/CodeGenOpenCL/generic-address-space-feature.cl -@@ -0,0 +1,28 @@ -+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -O0 -cl-std=CL1.2 -cl-ext=-cl_khr_fp64 -finclude-default-header -emit-llvm -o - | FileCheck %s --check-prefixes=CL12 -+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -O0 -cl-std=CL1.2 -cl-ext=-cl_khr_fp64 -fdeclare-opencl-builtins -emit-llvm -o - | FileCheck %s --check-prefixes=CL12 -+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -O0 -cl-std=CL2.0 -cl-ext=-cl_khr_fp64 -finclude-default-header -emit-llvm -o - | FileCheck %s --check-prefixes=CL20 -+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -O0 -cl-std=CL2.0 -cl-ext=-cl_khr_fp64 -fdeclare-opencl-builtins -emit-llvm -o - | FileCheck %s --check-prefixes=CL20 -+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -O0 -cl-std=CL3.0 -cl-ext=-cl_khr_fp64 -finclude-default-header -emit-llvm -o - | FileCheck %s --check-prefixes=CL30 -+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -O0 -cl-std=CL3.0 -cl-ext=-cl_khr_fp64,__opencl_c_generic_address_space -finclude-default-header -emit-llvm -o - | FileCheck %s --check-prefixes=CL30-GENERIC -+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -O0 -cl-std=CL3.0 -cl-ext=-cl_khr_fp64 -fdeclare-opencl-builtins -emit-llvm -o - | FileCheck %s --check-prefixes=CL30 -+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -O0 -cl-std=CL3.0 -cl-ext=-cl_khr_fp64,__opencl_c_generic_address_space -fdeclare-opencl-builtins -emit-llvm -o - | FileCheck %s --check-prefixes=CL30-GENERIC -+ -+void test(global float* src1, local float *src2, private float *src3, float *src4, float tmp) { -+ // CL20: %{{.+}} = addrspacecast float addrspace(1)* %{{.+}} to float addrspace(4)* -+ // CL12-NOT: addrspacecast -+ // CL30-NOT: addrspacecast -+ // CL30-GENERIC-NOT: addrspacecast -+ tmp = sincos(tmp, src1); -+ // CL20: %{{.+}} = addrspacecast float addrspace(3)* %{{.+}} to float addrspace(4)* -+ // CL12-NOT: addrspacecast -+ // CL30-NOT: addrspacecast -+ // CL30-GENERIC-NOT: addrspacecast -+ tmp = sincos(tmp, src2); -+ -+ // CL12: %{{.+}} = call {{(spir_func )?}}float [[FUNCNAME:@.*sincos.*]](float %{{.+}}, float* {{.+}}) -+ // CL20: %{{.+}} = call {{(spir_func )?}}float [[FUNCNAME:@.*sincos.*]](float %{{.+}}, float addrspace(4)* {{.+}}) -+ // CL30: %{{.+}} = call {{(spir_func )?}}float [[FUNCNAME:@.*sincos.*]](float %{{.+}}, float* {{.+}}) -+ // CL30-GENERIC: %{{.+}} = call {{(spir_func )?}}float [[FUNCNAME:@.*sincos.*]](float %{{.+}}, float addrspace(4)* {{.+}}) -+ // CHECK: addrspacecast -+ tmp = sincos(tmp, src4); -+} -diff --git a/clang/test/CodeGenOpenCL/intel-subgroups-avc-ext-types.cl b/clang/test/CodeGenOpenCL/intel-subgroups-avc-ext-types.cl -index 515f13f6e768..5aa31ac6f345 100644 ---- a/clang/test/CodeGenOpenCL/intel-subgroups-avc-ext-types.cl -+++ b/clang/test/CodeGenOpenCL/intel-subgroups-avc-ext-types.cl -@@ -1,4 +1,5 @@ - // RUN: %clang_cc1 %s -triple spir-unknown-unknown -cl-std=CL1.2 -cl-ext=+cl_intel_device_side_avc_motion_estimation -emit-llvm -o - -O0 | FileCheck %s -+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -cl-std=CL3.0 -cl-ext=+cl_intel_device_side_avc_motion_estimation -emit-llvm -o - -O0 | FileCheck %s - - // CHECK: %opencl.intel_sub_group_avc_mce_payload_t = type opaque - // CHECK: %opencl.intel_sub_group_avc_ime_payload_t = type opaque -diff --git a/clang/test/CodeGenOpenCL/kernels-have-spir-cc-by-default.cl b/clang/test/CodeGenOpenCL/kernels-have-spir-cc-by-default.cl -index e89237623478..3d6708ac361f 100644 ---- a/clang/test/CodeGenOpenCL/kernels-have-spir-cc-by-default.cl -+++ b/clang/test/CodeGenOpenCL/kernels-have-spir-cc-by-default.cl -@@ -1,5 +1,8 @@ - // RUN: %clang_cc1 %s -cl-std=CL1.2 -emit-llvm -triple x86_64-unknown-unknown -o - | FileCheck %s - // RUN: %clang_cc1 %s -cl-std=CL1.2 -emit-llvm -triple amdgcn-unknown-unknown -o - | FileCheck -check-prefixes=AMDGCN %s -+// RUN: %clang_cc1 %s -cl-std=CL3.0 -cl-ext=__opencl_c_fp64 -emit-llvm -triple x86_64-unknown-unknown -o - | FileCheck %s -+// RUN: %clang_cc1 %s -cl-std=CL3.0 -cl-ext=__opencl_c_fp64 -emit-llvm -triple amdgcn-unknown-unknown -o - | FileCheck -check-prefixes=AMDGCN %s -+ - // Test that the kernels always use the SPIR calling convention - // to have unambiguous mapping of arguments to feasibly implement - // clSetKernelArg(). -diff --git a/clang/test/CodeGenOpenCL/logical-ops.cl b/clang/test/CodeGenOpenCL/logical-ops.cl -index f083a8580ee7..499eab65039b 100644 ---- a/clang/test/CodeGenOpenCL/logical-ops.cl -+++ b/clang/test/CodeGenOpenCL/logical-ops.cl -@@ -1,4 +1,5 @@ - // RUN: %clang_cc1 %s -emit-llvm -o - -cl-std=CL1.2 -O1 -triple x86_64-unknown-linux-gnu | FileCheck %s -+// RUN: %clang_cc1 %s -emit-llvm -o - -cl-std=CL3.0 -cl-ext=__opencl_c_fp64 -O1 -triple x86_64-unknown-linux-gnu | FileCheck %s - // RUN: %clang_cc1 %s -emit-llvm -o - -cl-std=clc++ -O1 -triple x86_64-unknown-linux-gnu | FileCheck %s - - #pragma OPENCL EXTENSION cl_khr_fp64 : enable -diff --git a/clang/test/CodeGenOpenCL/no-half.cl b/clang/test/CodeGenOpenCL/no-half.cl -index aee8f678f01a..46da7fa339e8 100644 ---- a/clang/test/CodeGenOpenCL/no-half.cl -+++ b/clang/test/CodeGenOpenCL/no-half.cl -@@ -1,6 +1,7 @@ - // RUN: %clang_cc1 %s -cl-std=cl2.0 -emit-llvm -o - -triple spir-unknown-unknown | FileCheck %s - // RUN: %clang_cc1 %s -cl-std=cl1.2 -emit-llvm -o - -triple spir-unknown-unknown | FileCheck %s - // RUN: %clang_cc1 %s -cl-std=cl1.1 -emit-llvm -o - -triple spir-unknown-unknown | FileCheck %s -+// RUN: %clang_cc1 %s -cl-std=cl3.0 -cl-ext=__opencl_c_fp64 -emit-llvm -o - -triple spir-unknown-unknown | FileCheck %s - - #pragma OPENCL EXTENSION cl_khr_fp64:enable - -diff --git a/clang/test/CodeGenOpenCL/pipe_builtin.cl b/clang/test/CodeGenOpenCL/pipe_builtin.cl -index 02b9669b7ab1..0bf35c336199 100644 ---- a/clang/test/CodeGenOpenCL/pipe_builtin.cl -+++ b/clang/test/CodeGenOpenCL/pipe_builtin.cl -@@ -1,4 +1,7 @@ - // RUN: %clang_cc1 -triple %itanium_abi_triple -emit-llvm -cl-ext=+cl_khr_subgroups -O0 -cl-std=clc++ -o - %s | FileCheck %s -+// RUN: %clang_cc1 -triple %itanium_abi_triple -emit-llvm -cl-ext=+cl_khr_subgroups -O0 -cl-std=cl2.0 -o - %s | FileCheck %s -+// RUN: %clang_cc1 -triple %itanium_abi_triple -emit-llvm -cl-ext=+cl_khr_subgroups -O0 -cl-std=cl3.0 -cl-ext=__opencl_c_pipes,__opencl_c_subgroups -o - %s | FileCheck %s -+ - // FIXME: Add MS ABI manglings of OpenCL things and remove %itanium_abi_triple - // above to support OpenCL in the MS C++ ABI. - -diff --git a/clang/test/CodeGenOpenCL/pipe_types.cl b/clang/test/CodeGenOpenCL/pipe_types.cl -index ba064c6d7557..b7a523d4f084 100644 ---- a/clang/test/CodeGenOpenCL/pipe_types.cl -+++ b/clang/test/CodeGenOpenCL/pipe_types.cl -@@ -1,4 +1,5 @@ - // RUN: %clang_cc1 -triple x86_64-unknown-linux-gnu -emit-llvm -O0 -cl-std=CL2.0 -o - %s | FileCheck %s -+// RUN: %clang_cc1 -triple x86_64-unknown-linux-gnu -emit-llvm -O0 -cl-std=CL3.0 -cl-ext=__opencl_c_pipes -o - %s | FileCheck %s - - // CHECK: %opencl.pipe_ro_t = type opaque - // CHECK: %opencl.pipe_wo_t = type opaque -diff --git a/clang/test/CodeGenOpenCL/printf.cl b/clang/test/CodeGenOpenCL/printf.cl -index fc139d776db6..0133c5595d81 100644 ---- a/clang/test/CodeGenOpenCL/printf.cl -+++ b/clang/test/CodeGenOpenCL/printf.cl -@@ -1,5 +1,7 @@ - // RUN: %clang_cc1 -cl-std=CL1.2 -cl-ext=-+cl_khr_fp64 -triple spir-unknown-unknown -disable-llvm-passes -emit-llvm -o - %s | FileCheck -check-prefixes=FP64,ALL %s - // RUN: %clang_cc1 -cl-std=CL1.2 -cl-ext=-cl_khr_fp64 -triple spir-unknown-unknown -disable-llvm-passes -emit-llvm -o - %s | FileCheck -check-prefixes=NOFP64,ALL %s -+// RUN: %clang_cc1 -cl-std=CL3.0 -cl-ext=+__opencl_c_fp64 -triple spir-unknown-unknown -disable-llvm-passes -emit-llvm -o - %s | FileCheck -check-prefixes=FP64,ALL %s -+// RUN: %clang_cc1 -cl-std=CL3.0 -triple spir-unknown-unknown -disable-llvm-passes -emit-llvm -o - %s | FileCheck -check-prefixes=NOFP64,ALL %s - - typedef __attribute__((ext_vector_type(2))) float float2; - typedef __attribute__((ext_vector_type(2))) half half2; -diff --git a/clang/test/CodeGenOpenCL/unroll-hint.cl b/clang/test/CodeGenOpenCL/unroll-hint.cl -index 0f84450a1ae6..9347c935869b 100644 ---- a/clang/test/CodeGenOpenCL/unroll-hint.cl -+++ b/clang/test/CodeGenOpenCL/unroll-hint.cl -@@ -1,5 +1,6 @@ - // RUN: %clang_cc1 -emit-llvm -O0 -cl-std=CL2.0 -o - %s | FileCheck %s - // RUN: %clang_cc1 -emit-llvm -O0 -cl-std=CL1.2 -o - %s | FileCheck %s -+// RUN: %clang_cc1 -emit-llvm -O0 -cl-std=CL3.0 -o - %s | FileCheck %s - - /*** for ***/ - void for_count() -diff --git a/clang/test/Driver/autocomplete.c b/clang/test/Driver/autocomplete.c -index 5c0bfb69f9a3..eabdf0b5938d 100644 ---- a/clang/test/Driver/autocomplete.c -+++ b/clang/test/Driver/autocomplete.c -@@ -43,6 +43,8 @@ - // CLSTDALL-NEXT: CL1.2 - // CLSTDALL-NEXT: cl2.0 - // CLSTDALL-NEXT: CL2.0 -+// CLSTDALL-NEXT: cl3.0 -+// CLSTDALL-NEXT: CL3.0 - // CLSTDALL-NEXT: clc++ - // CLSTDALL-NEXT: CLC++ - // RUN: %clang --autocomplete=-fno-sanitize-coverage=,f | FileCheck %s -check-prefix=FNOSANICOVER -diff --git a/clang/test/Driver/opencl.cl b/clang/test/Driver/opencl.cl -index 63b04bc1af41..ffdd4f7d65bb 100644 ---- a/clang/test/Driver/opencl.cl -+++ b/clang/test/Driver/opencl.cl -@@ -2,6 +2,7 @@ - // RUN: %clang -S -### -cl-std=CL1.1 %s 2>&1 | FileCheck --check-prefix=CHECK-CL11 %s - // RUN: %clang -S -### -cl-std=CL1.2 %s 2>&1 | FileCheck --check-prefix=CHECK-CL12 %s - // RUN: %clang -S -### -cl-std=CL2.0 %s 2>&1 | FileCheck --check-prefix=CHECK-CL20 %s -+// RUN: %clang -S -### -cl-std=CL3.0 %s 2>&1 | FileCheck --check-prefix=CHECK-CL30 %s - // RUN: %clang -S -### -cl-std=clc++ %s 2>&1 | FileCheck --check-prefix=CHECK-CLCPP %s - // RUN: %clang -S -### -cl-opt-disable %s 2>&1 | FileCheck --check-prefix=CHECK-OPT-DISABLE %s - // RUN: %clang -S -### -cl-strict-aliasing %s 2>&1 | FileCheck --check-prefix=CHECK-STRICT-ALIASING %s -@@ -22,6 +23,7 @@ - // CHECK-CL11: "-cc1" {{.*}} "-cl-std=CL1.1" - // CHECK-CL12: "-cc1" {{.*}} "-cl-std=CL1.2" - // CHECK-CL20: "-cc1" {{.*}} "-cl-std=CL2.0" -+// CHECK-CL30: "-cc1" {{.*}} "-cl-std=CL3.0" - // CHECK-CLCPP: "-cc1" {{.*}} "-cl-std=clc++" - // CHECK-OPT-DISABLE: "-cc1" {{.*}} "-cl-opt-disable" - // CHECK-STRICT-ALIASING: "-cc1" {{.*}} "-cl-strict-aliasing" -diff --git a/clang/test/Driver/unknown-std.cl b/clang/test/Driver/unknown-std.cl -index 6f371bac13ac..00209fb62556 100644 ---- a/clang/test/Driver/unknown-std.cl -+++ b/clang/test/Driver/unknown-std.cl -@@ -10,6 +10,7 @@ - // CHECK-NEXT: note: use 'cl1.1' for 'OpenCL 1.1' standard - // CHECK-NEXT: note: use 'cl1.2' for 'OpenCL 1.2' standard - // CHECK-NEXT: note: use 'cl2.0' for 'OpenCL 2.0' standard -+// CHECK-NEXT: note: use 'cl3.0' for 'OpenCL 3.0' standard - // CHECK-NEXT: note: use 'clc++' for 'C++ for OpenCL' standard - - // Make sure that no other output is present. -diff --git a/clang/test/Frontend/stdlang.c b/clang/test/Frontend/stdlang.c -index 51484999e37a..eac4632fbdd6 100644 ---- a/clang/test/Frontend/stdlang.c -+++ b/clang/test/Frontend/stdlang.c -@@ -9,6 +9,7 @@ - // RUN: %clang_cc1 -x cl -cl-std=CL1.1 -DOPENCL %s - // RUN: %clang_cc1 -x cl -cl-std=CL1.2 -DOPENCL %s - // RUN: %clang_cc1 -x cl -cl-std=CL2.0 -DOPENCL %s -+// RUN: %clang_cc1 -x cl -cl-std=CL3.0 -DOPENCL %s - // RUN: %clang_cc1 -x cl -cl-std=CLC++ -DOPENCL %s - // RUN: not %clang_cc1 -x cl -std=c99 -DOPENCL %s 2>&1 | FileCheck --check-prefix=CHECK-C99 %s - // RUN: not %clang_cc1 -x cl -cl-std=invalid -DOPENCL %s 2>&1 | FileCheck --check-prefix=CHECK-INVALID %s -diff --git a/clang/test/Headers/opencl-c-header.cl b/clang/test/Headers/opencl-c-header.cl -index 1b151ffdd16a..2716076acdcf 100644 ---- a/clang/test/Headers/opencl-c-header.cl -+++ b/clang/test/Headers/opencl-c-header.cl -@@ -1,6 +1,7 @@ - // RUN: %clang_cc1 -O0 -triple spir-unknown-unknown -internal-isystem ../../lib/Headers -include opencl-c.h -emit-llvm -o - %s -verify | FileCheck %s - // RUN: %clang_cc1 -O0 -triple spir-unknown-unknown -internal-isystem ../../lib/Headers -include opencl-c.h -emit-llvm -o - %s -verify -cl-std=CL1.1 | FileCheck %s - // RUN: %clang_cc1 -O0 -triple spir-unknown-unknown -internal-isystem ../../lib/Headers -include opencl-c.h -emit-llvm -o - %s -verify -cl-std=CL1.2 | FileCheck %s -+// RUN: %clang_cc1 -O0 -triple spir-unknown-unknown -internal-isystem ../../lib/Headers -include opencl-c.h -emit-llvm -o - %s -verify -cl-std=CL3.0 | FileCheck %s - // RUN: %clang_cc1 -O0 -triple spir-unknown-unknown -internal-isystem ../../lib/Headers -include opencl-c.h -emit-llvm -o - %s -verify -cl-std=clc++ | FileCheck %s --check-prefix=CHECK20 - - // Test including the default header as a module. -@@ -39,9 +40,11 @@ - // RUN: rm -rf %t - // RUN: mkdir -p %t - // RUN: %clang_cc1 -triple spir64-unknown-unknown -emit-llvm -o - -cl-std=CL1.2 -finclude-default-header -fmodules -fimplicit-module-maps -fmodules-cache-path=%t -ftime-report %s 2>&1 | FileCheck --check-prefix=CHECK --check-prefix=CHECK-MOD %s -+// RUN: %clang_cc1 -triple spir64-unknown-unknown -emit-llvm -o - -cl-std=CL3.0 -finclude-default-header -fmodules -fimplicit-module-maps -fmodules-cache-path=%t -ftime-report %s 2>&1 | FileCheck --check-prefix=CHECK --check-prefix=CHECK-MOD %s - // RUN: %clang_cc1 -triple amdgcn--amdhsa -O0 -emit-llvm -o - -cl-std=CL2.0 -finclude-default-header -fmodules -fimplicit-module-maps -fmodules-cache-path=%t -ftime-report %s 2>&1 | FileCheck --check-prefix=CHECK20 --check-prefix=CHECK-MOD %s - // RUN: chmod u-w %t - // RUN: %clang_cc1 -triple spir64-unknown-unknown -emit-llvm -o - -cl-std=CL1.2 -finclude-default-header -fmodules -fimplicit-module-maps -fmodules-cache-path=%t -ftime-report %s 2>&1 | FileCheck --check-prefix=CHECK --check-prefix=CHECK-MOD %s -+// RUN: %clang_cc1 -triple spir64-unknown-unknown -emit-llvm -o - -cl-std=CL3.0 -finclude-default-header -fmodules -fimplicit-module-maps -fmodules-cache-path=%t -ftime-report %s 2>&1 | FileCheck --check-prefix=CHECK --check-prefix=CHECK-MOD %s - // RUN: %clang_cc1 -triple amdgcn--amdhsa -O0 -emit-llvm -o - -cl-std=CL2.0 -finclude-default-header -fmodules -fimplicit-module-maps -fmodules-cache-path=%t -ftime-report %s 2>&1 | FileCheck --check-prefix=CHECK20 --check-prefix=CHECK-MOD %s - // RUN: chmod u+w %t - -@@ -67,7 +70,7 @@ char f(char x) { - // from OpenCL 2.0 onwards. - - // CHECK20: _Z12write_imagef14ocl_image3d_wo --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ == CL_VERSION_2_0) - void test_image3dwo(write_only image3d_t img) { - write_imagef(img, (0), (0.0f)); - } -@@ -81,7 +84,7 @@ void test_atomics(__generic volatile unsigned int* a) { - #endif - - // Verify that ATOMIC_VAR_INIT is defined. --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ == CL_VERSION_2_0) - global atomic_int z = ATOMIC_VAR_INIT(99); - #endif //__OPENCL_C_VERSION__ - -diff --git a/clang/test/Index/pipe-size.cl b/clang/test/Index/pipe-size.cl -index 94a1255f0a48..59b76051eda1 100644 ---- a/clang/test/Index/pipe-size.cl -+++ b/clang/test/Index/pipe-size.cl -@@ -2,6 +2,13 @@ - // RUN: %clang_cc1 -x cl -O0 -cl-std=CL2.0 -emit-llvm -triple spir-unknown-unknown %s -o - | FileCheck %s --check-prefix=SPIR - // RUN: %clang_cc1 -x cl -O0 -cl-std=CL2.0 -emit-llvm -triple spir64-unknown-unknown %s -o - | FileCheck %s --check-prefix=SPIR64 - // RUN: %clang_cc1 -x cl -O0 -cl-std=CL2.0 -emit-llvm -triple amdgcn-amd-amdhsa %s -o - | FileCheck %s --check-prefix=AMDGCN -+// RUN: %clang_cc1 -x cl -O0 -cl-std=CL3.0 -cl-ext=__opencl_c_pipes -emit-llvm -triple x86_64-unknown-linux-gnu %s -o - | FileCheck %s --check-prefix=X86 -+// RUN: %clang_cc1 -x cl -O0 -cl-std=CL3.0 -cl-ext=__opencl_c_pipes -emit-llvm -triple spir-unknown-unknown %s -o - | FileCheck %s --check-prefix=SPIR -+// RUN: %clang_cc1 -x cl -O0 -cl-std=CL3.0 -cl-ext=__opencl_c_pipes -emit-llvm -triple spir64-unknown-unknown %s -o - | FileCheck %s --check-prefix=SPIR64 -+// RUN: %clang_cc1 -x cl -O0 -cl-std=CL3.0 -cl-ext=__opencl_c_pipes -emit-llvm -triple amdgcn-amd-amdhsa %s -o - | FileCheck %s --check-prefix=AMDGCN -+ -+ -+ - __kernel void testPipe( pipe int test ) - { - int s = sizeof(test); -diff --git a/clang/test/Preprocessor/predefined-macros.c b/clang/test/Preprocessor/predefined-macros.c -index def105f4c52e..b088a37ba665 100644 ---- a/clang/test/Preprocessor/predefined-macros.c -+++ b/clang/test/Preprocessor/predefined-macros.c -@@ -129,6 +129,8 @@ - // RUN: | FileCheck -match-full-lines %s --check-prefix=CHECK-CL12 - // RUN: %clang_cc1 %s -E -dM -o - -x cl -cl-std=CL2.0 \ - // RUN: | FileCheck -match-full-lines %s --check-prefix=CHECK-CL20 -+// RUN: %clang_cc1 %s -E -dM -o - -x cl -cl-std=CL3.0 \ -+// RUN: | FileCheck -match-full-lines %s --check-prefix=CHECK-CL30 - // RUN: %clang_cc1 %s -E -dM -o - -x cl -cl-fast-relaxed-math \ - // RUN: | FileCheck -match-full-lines %s --check-prefix=CHECK-FRM - // RUN: %clang_cc1 %s -E -dM -o - -x cl -cl-std=clc++ \ -@@ -137,26 +139,37 @@ - // CHECK-CL10: #define CL_VERSION_1_1 110 - // CHECK-CL10: #define CL_VERSION_1_2 120 - // CHECK-CL10: #define CL_VERSION_2_0 200 -+// CHECK-CL10: #define CL_VERSION_3_0 300 - // CHECK-CL10: #define __OPENCL_C_VERSION__ 100 - // CHECK-CL10-NOT: #define __FAST_RELAXED_MATH__ 1 - // CHECK-CL11: #define CL_VERSION_1_0 100 - // CHECK-CL11: #define CL_VERSION_1_1 110 - // CHECK-CL11: #define CL_VERSION_1_2 120 - // CHECK-CL11: #define CL_VERSION_2_0 200 -+// CHECK-CL11: #define CL_VERSION_3_0 300 - // CHECK-CL11: #define __OPENCL_C_VERSION__ 110 - // CHECK-CL11-NOT: #define __FAST_RELAXED_MATH__ 1 - // CHECK-CL12: #define CL_VERSION_1_0 100 - // CHECK-CL12: #define CL_VERSION_1_1 110 - // CHECK-CL12: #define CL_VERSION_1_2 120 - // CHECK-CL12: #define CL_VERSION_2_0 200 -+// CHECK-CL12: #define CL_VERSION_3_0 300 - // CHECK-CL12: #define __OPENCL_C_VERSION__ 120 - // CHECK-CL12-NOT: #define __FAST_RELAXED_MATH__ 1 - // CHECK-CL20: #define CL_VERSION_1_0 100 - // CHECK-CL20: #define CL_VERSION_1_1 110 - // CHECK-CL20: #define CL_VERSION_1_2 120 - // CHECK-CL20: #define CL_VERSION_2_0 200 -+// CHECK-CL20: #define CL_VERSION_3_0 300 - // CHECK-CL20: #define __OPENCL_C_VERSION__ 200 - // CHECK-CL20-NOT: #define __FAST_RELAXED_MATH__ 1 -+// CHECK-CL30: #define CL_VERSION_1_0 100 -+// CHECK-CL30: #define CL_VERSION_1_1 110 -+// CHECK-CL30: #define CL_VERSION_1_2 120 -+// CHECK-CL30: #define CL_VERSION_2_0 200 -+// CHECK-CL30: #define CL_VERSION_3_0 300 -+// CHECK-CL30: #define __OPENCL_C_VERSION__ 300 -+// CHECK-CL30-NOT: #define __FAST_RELAXED_MATH__ 1 - // CHECK-FRM: #define __FAST_RELAXED_MATH__ 1 - // CHECK-CLCPP10: #define __CL_CPP_VERSION_1_0__ 100 - // CHECK-CLCPP10: #define __OPENCL_CPP_VERSION__ 100 -diff --git a/clang/test/Sema/feature-extensions-simult-support.cl b/clang/test/Sema/feature-extensions-simult-support.cl -new file mode 100644 -index 000000000000..0789105002b2 ---- /dev/null -+++ b/clang/test/Sema/feature-extensions-simult-support.cl -@@ -0,0 +1,75 @@ -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL1.2 -cl-ext=-cl_khr_fp64 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL1.2 -cl-ext=+cl_khr_fp64 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL1.2 -cl-ext=-cl_khr_3d_image_writes -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL1.2 -cl-ext=+cl_khr_3d_image_writes -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL1.2 -cl-ext=-cl_khr_subgroups -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL1.2 -cl-ext=+cl_khr_subgroups -+ -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL2.0 -cl-ext=-cl_khr_fp64 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL2.0 -cl-ext=+cl_khr_fp64 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL2.0 -cl-ext=-cl_khr_3d_image_writes -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL2.0 -cl-ext=+cl_khr_3d_image_writes -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL2.0 -cl-ext=-cl_khr_subgroups -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL2.0 -cl-ext=+cl_khr_subgroups -+ -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+__opencl_c_fp64 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+__opencl_c_3d_image_writes -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+__opencl_c_subgroups -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+__opencl_c_pipes -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+__opencl_c_device_enqueue -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+__opencl_c_read_write_images -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+__opencl_c_fp64,-cl_khr_fp64 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+__opencl_c_3d_image_writes,-cl_khr_3d_image_writes -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+__opencl_c_subgroups,-cl_khr_subgroups -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+cl_khr_fp64 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+cl_khr_3d_image_writes -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+cl_khr_subgroups -+ -+// expected-no-diagnostics -+ -+#ifdef cl_khr_fp64 -+ #ifndef __opencl_c_fp64 -+ #error macros were not properly set up -+ #endif -+#endif -+#ifdef __opencl_c_fp64 -+ #ifndef cl_khr_fp64 -+ #error macros were not properly set up -+ #endif -+#endif -+ -+#ifdef cl_khr_3d_image_writes -+ #ifndef __opencl_c_3d_image_writes -+ #error macros were not properly set up -+ #endif -+#endif -+#ifdef __opencl_c_3d_image_writes -+ #ifndef cl_khr_3d_image_writes -+ #error macros were not properly set up -+ #endif -+#endif -+ -+#ifdef cl_khr_subgroups -+ #ifndef __opencl_c_subgroups -+ #error macros were not properly set up -+ #endif -+#endif -+#ifdef __opencl_c_subgroups -+ #ifndef cl_khr_subgroups -+ #error macros were not properly set up -+ #endif -+#endif -+ -+#if defined(__opencl_c_pipes) || defined(__opencl_c_device_enqueue) -+ #ifndef __opencl_c_generic_address_space -+ #error macros were not properly set up -+ #endif -+#endif -+ -+#if defined(__opencl_c_3d_image_writes) || defined(__opencl_c_read_write_images) -+ #ifndef __opencl_c_images -+ #error macros were not properly set up -+ #endif -+#endif -+ -+kernel void test(){} -diff --git a/clang/test/Sema/features-ignore-pragma.cl b/clang/test/Sema/features-ignore-pragma.cl -new file mode 100644 -index 000000000000..046ce5390754 ---- /dev/null -+++ b/clang/test/Sema/features-ignore-pragma.cl -@@ -0,0 +1,24 @@ -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=__opencl_c_fp64 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=__opencl_c_subgroups -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=__opencl_c_3d_image_writes -+ -+#pragma OPENCL EXTENSION __opencl_c_fp64 : enable -+// expected-warning@-1 {{OpenCL feature support can't be controlled via pragma, ignoring}} -+ -+#pragma OPENCL EXTENSION cl_khr_fp64 : enable -+#ifndef __opencl_c_fp64 -+// expected-warning@-2{{unsupported OpenCL extension 'cl_khr_fp64' - ignoring}} -+#endif -+ -+#pragma OPENCL EXTENSION cl_khr_subgroups : enable -+#ifndef __opencl_c_subgroups -+// expected-warning@-2{{unsupported OpenCL extension 'cl_khr_subgroups' - ignoring}} -+#endif -+ -+#pragma OPENCL EXTENSION cl_khr_3d_image_writes : enable -+#ifndef __opencl_c_3d_image_writes -+// expected-warning@-2{{unsupported OpenCL extension 'cl_khr_3d_image_writes' - ignoring}} -+#endif -+ -+kernel void foo() {} -diff --git a/clang/test/Sema/opencl-features-pipes.cl b/clang/test/Sema/opencl-features-pipes.cl -new file mode 100644 -index 000000000000..c0ac778f24a6 ---- /dev/null -+++ b/clang/test/Sema/opencl-features-pipes.cl -@@ -0,0 +1,18 @@ -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL1.1 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL1.2 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL2.0 -DHAS -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=__opencl_c_pipes -DHAS -+// expected-no-diagnostics -+ -+#ifdef HAS -+ #ifndef __opencl_c_pipes -+ #error Feature should be defined -+ #endif -+#else -+ #ifdef __opencl_c_pipes -+ #error Feature should not be defined -+ #endif -+#endif -+ -+kernel void foo() {} -diff --git a/clang/test/Sema/opencl-features.cl b/clang/test/Sema/opencl-features.cl -new file mode 100644 -index 000000000000..aa432f6b60bf ---- /dev/null -+++ b/clang/test/Sema/opencl-features.cl -@@ -0,0 +1,128 @@ -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL2.0 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL2.0 -cl-ext=-__opencl_c_device_enqueue,-__opencl_c_pipes,-__opencl_c_read_write_images -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CLC++ -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -+// expected-no-diagnostics -+ -+#ifndef __opencl_c_int64 -+ #error Feature __opencl_c_int64 shouldn't be defined -+#endif -+ -+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ == CL_VERSION_2_0) -+ #ifndef __opencl_c_3d_image_writes -+ #error Feature __opencl_c_3d_image_writes should be defined -+ #endif -+ -+ #ifndef __opencl_c_atomic_order_acq_rel -+ #error Feature __opencl_c_atomic_order_acq_rel should be defined -+ #endif -+ -+ #ifndef __opencl_c_atomic_order_seq_cst -+ #error Feature __opencl_c_atomic_order_seq_cst should be defined -+ #endif -+ -+ #ifndef __opencl_c_atomic_scope_device -+ #error Feature __opencl_c_atomic_scope_device should be defined -+ #endif -+ -+ #ifndef __opencl_c_atomic_scope_all_devices -+ #error Feature __opencl_c_atomic_scope_all_devices should be defined -+ #endif -+ -+ #ifndef __opencl_c_device_enqueue -+ #error Feature __opencl_c_device_enqueue should be defined -+ #endif -+ -+ #ifndef __opencl_c_generic_address_space -+ #error Feature __opencl_c_generic_address_space should be defined -+ #endif -+ -+ #ifndef __opencl_c_pipes -+ #error Feature __opencl_c_pipes should be defined -+ #endif -+ -+ #ifndef __opencl_c_program_scope_global_variables -+ #error Feature __opencl_c_program_scope_global_variables should be defined -+ #endif -+ -+ #ifndef __opencl_c_read_write_images -+ #error Feature __opencl_c_read_write_images should be defined -+ #endif -+ -+ #ifndef __opencl_c_subgroups -+ #error Feature __opencl_c_subgroups should be defined -+ #endif -+ -+ #ifndef __opencl_c_work_group_collective_functions -+ #error Feature __opencl_c_work_group_collective_functions should be defined -+ #endif -+ -+ #ifndef __opencl_c_fp64 -+ #error Feature __opencl_c_fp64 should be defined -+ #endif -+ -+ #ifndef __opencl_c_images -+ #error Feature __opencl_c_images should be defined -+ #endif -+#endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+ -+ -+#if __OPENCL_C_VERSION__ == CL_VERSION_3_0 -+ #ifdef __opencl_c_3d_image_writes -+ #error Feature __opencl_c_3d_image_writes shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_atomic_order_acq_rel -+ #error Feature __opencl_c_atomic_order_acq_rel shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_atomic_order_seq_cst -+ #error Feature __opencl_c_atomic_order_seq_cst shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_atomic_scope_device -+ #error Feature __opencl_c_atomic_scope_device shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_atomic_scope_all_devices -+ #error Feature __opencl_c_atomic_scope_all_devices shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_device_enqueue -+ #error Feature __opencl_c_device_enqueue shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_generic_address_space -+ #error Feature __opencl_c_generic_address_space shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_pipes -+ #error Feature __opencl_c_pipes shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_program_scope_global_variables -+ #error Feature __opencl_c_program_scope_global_variables shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_read_write_images -+ #error Feature __opencl_c_read_write_images shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_subgroups -+ #error Feature __opencl_c_subgroups shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_work_group_collective_functions -+ #error Feature __opencl_c_work_group_collective_functions shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_fp64 -+ #error Feature __opencl_c_fp64 shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_images -+ #error Feature __opencl_c_images shouldn't be defined -+ #endif -+#endif // __OPENCL_C_VERSION__ == CL_VERSION_3_0 -+ -+kernel void foo() {} -diff --git a/clang/test/Sema/pipe_builtins_feature.cl b/clang/test/Sema/pipe_builtins_feature.cl -new file mode 100644 -index 000000000000..56fa94fc7705 ---- /dev/null -+++ b/clang/test/Sema/pipe_builtins_feature.cl -@@ -0,0 +1,21 @@ -+// RUN: %clang_cc1 -cl-std=CL2.0 -fsyntax-only -verify %s -+// RUN: %clang_cc1 -cl-std=CL3.0 -fsyntax-only -verify %s -+// RUN: %clang_cc1 -cl-std=CL3.0 -cl-ext=__opencl_c_pipes -fsyntax-only -verify %s -+ -+#ifdef __opencl_c_pipes -+ #ifndef __opencl_c_generic_address_space -+ #error Generic address space feature must also be defined -+ #endif -+// CHECK: expected-no-diagnostics -+// check that pragma disable all doesn't touch feature support -+ #pragma OPENCL EXTENSION all : disable -+#endif -+ -+void test(read_only pipe int p, global int *ptr) { -+ reserve_id_t rid; -+} -+ -+#ifndef __opencl_c_pipes -+// expected-error@-5 {{expected parameter declarator}} expected-error@-5 {{expected ')'}} expected-note@-5 {{to match this '('}} -+// expected-error@-5 {{use of type 'reserve_id_t' requires __opencl_c_pipes extension to be enabled}} -+#endif -diff --git a/clang/test/SemaOpenCL/address-spaces-conversions-cl2.0.cl b/clang/test/SemaOpenCL/address-spaces-conversions-cl2.0.cl -index 5efea216346a..09aea27ae6de 100644 ---- a/clang/test/SemaOpenCL/address-spaces-conversions-cl2.0.cl -+++ b/clang/test/SemaOpenCL/address-spaces-conversions-cl2.0.cl -@@ -1,6 +1,9 @@ - // RUN: %clang_cc1 %s -ffake-address-space-map -verify -pedantic -fsyntax-only -DCONSTANT -cl-std=CL2.0 - // RUN: %clang_cc1 %s -ffake-address-space-map -verify -pedantic -fsyntax-only -DGLOBAL -cl-std=CL2.0 - // RUN: %clang_cc1 %s -ffake-address-space-map -verify -pedantic -fsyntax-only -DGENERIC -cl-std=CL2.0 -+// RUN: %clang_cc1 %s -ffake-address-space-map -verify -pedantic -fsyntax-only -DCONSTANT -cl-std=CL3.0 -cl-ext=__opencl_c_generic_address_space -+// RUN: %clang_cc1 %s -ffake-address-space-map -verify -pedantic -fsyntax-only -DGLOBAL -cl-std=CL3.0 -cl-ext=__opencl_c_generic_address_space -+// RUN: %clang_cc1 %s -ffake-address-space-map -verify -pedantic -fsyntax-only -DGENERIC -cl-std=CL3.0 -cl-ext=__opencl_c_generic_address_space - // RUN: %clang_cc1 %s -ffake-address-space-map -verify -pedantic -fsyntax-only -DCONSTANT -cl-std=clc++ - // RUN: %clang_cc1 %s -ffake-address-space-map -verify -pedantic -fsyntax-only -DGLOBAL -cl-std=clc++ - // RUN: %clang_cc1 %s -ffake-address-space-map -verify -pedantic -fsyntax-only -DGENERIC -cl-std=clc++ -diff --git a/clang/test/SemaOpenCL/address-spaces.cl b/clang/test/SemaOpenCL/address-spaces.cl -index 07547ea19680..e367a7c57292 100644 ---- a/clang/test/SemaOpenCL/address-spaces.cl -+++ b/clang/test/SemaOpenCL/address-spaces.cl -@@ -1,5 +1,6 @@ - // RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only - // RUN: %clang_cc1 %s -cl-std=CL2.0 -verify -pedantic -fsyntax-only -+// RUN: %clang_cc1 %s -cl-std=CL3.0 -cl-ext=__opencl_c_generic_address_space -verify -pedantic -fsyntax-only - // RUN: %clang_cc1 %s -cl-std=clc++ -verify -pedantic -fsyntax-only - - __constant int ci = 1; -diff --git a/clang/test/SemaOpenCL/cl20-device-side-enqueue.cl b/clang/test/SemaOpenCL/cl20-device-side-enqueue.cl -index f63e2913c749..727141190a0b 100644 ---- a/clang/test/SemaOpenCL/cl20-device-side-enqueue.cl -+++ b/clang/test/SemaOpenCL/cl20-device-side-enqueue.cl -@@ -2,6 +2,12 @@ - // RUN: %clang_cc1 %s -cl-std=CL2.0 -triple "spir-unknown-unknown" -verify -pedantic -fsyntax-only -DB32 -DQUALS="const volatile" - // RUN: %clang_cc1 %s -cl-std=CL2.0 -triple "spir64-unknown-unknown" -verify -pedantic -fsyntax-only -Wconversion -DWCONV -DQUALS= - // RUN: %clang_cc1 %s -cl-std=CL2.0 -triple "spir64-unknown-unknown" -verify -pedantic -fsyntax-only -Wconversion -DWCONV -DQUALS="const volatile" -+// RUN: %clang_cc1 %s -cl-std=CL3.0 -triple "spir-unknown-unknown" -cl-ext=__opencl_c_device_enqueue,__opencl_c_subgroups -verify -pedantic -fsyntax-only -DB32 -DQUALS= -+// RUN: %clang_cc1 %s -cl-std=CL3.0 -triple "spir-unknown-unknown" -cl-ext=__opencl_c_device_enqueue,__opencl_c_subgroups -verify -pedantic -fsyntax-only -DB32 -DQUALS="const volatile" -+// RUN: %clang_cc1 %s -cl-std=CL3.0 -triple "spir64-unknown-unknown" -cl-ext=__opencl_c_device_enqueue,__opencl_c_subgroups -verify -pedantic -fsyntax-only -Wconversion -DWCONV -DQUALS= -+// RUN: %clang_cc1 %s -cl-std=CL3.0 -triple "spir64-unknown-unknown" -cl-ext=__opencl_c_device_enqueue,__opencl_c_subgroups -verify -pedantic -fsyntax-only -Wconversion -DWCONV -DQUALS="const volatile" -+ -+ - - typedef struct {int a;} ndrange_t; - // Diagnostic tests for different overloads of enqueue_kernel from Table 6.13.17.1 of OpenCL 2.0 Spec. -@@ -235,11 +241,17 @@ kernel void bar(global unsigned int *buf) - kernel void foo1(global unsigned int *buf) - { - ndrange_t n; -- buf[0] = get_kernel_max_sub_group_size_for_ndrange(n, ^(){}); // expected-error {{use of declaration 'get_kernel_max_sub_group_size_for_ndrange' requires cl_khr_subgroups extension to be enabled}} -+ buf[0] = get_kernel_max_sub_group_size_for_ndrange(n, ^(){}); -+#if __OPENCL_C_VERSION__ < 300 -+// expected-error@-2 {{use of declaration 'get_kernel_max_sub_group_size_for_ndrange' requires cl_khr_subgroups extension to be enabled}} -+#endif - } - - kernel void bar1(global unsigned int *buf) - { - ndrange_t n; -- buf[0] = get_kernel_sub_group_count_for_ndrange(n, ^(){}); // expected-error {{use of declaration 'get_kernel_sub_group_count_for_ndrange' requires cl_khr_subgroups extension to be enabled}} -+ buf[0] = get_kernel_sub_group_count_for_ndrange(n, ^(){}); -+#if __OPENCL_C_VERSION__ < 300 -+// expected-error@-2 {{use of declaration 'get_kernel_sub_group_count_for_ndrange' requires cl_khr_subgroups extension to be enabled}} -+#endif - } -diff --git a/clang/test/SemaOpenCL/forget-unsupported-builtins.cl b/clang/test/SemaOpenCL/forget-unsupported-builtins.cl -new file mode 100644 -index 000000000000..14dd03e2c7db ---- /dev/null -+++ b/clang/test/SemaOpenCL/forget-unsupported-builtins.cl -@@ -0,0 +1,22 @@ -+// RUN: %clang_cc1 -cl-std=cl3.0 -fsyntax-only -verify %s -triple spir-unknown-unknown -+// RUN: %clang_cc1 -cl-std=cl3.0 -fsyntax-only -cl-ext=__opencl_c_pipes,__opencl_c_generic_address_space,__opencl_c_device_enqueue -verify %s -triple spir-unknown-unknown -DFEATURES -+ -+#ifndef FEATURES -+ // expected-no-diagnostics -+#else -+ // expected-error@+10 {{cannot redeclare builtin function 'get_pipe_max_packets'}} -+ // expected-note@+9 {{'get_pipe_max_packets' is a builtin with type 'unsigned int ()'}} -+ // expected-error@+9 {{cannot redeclare builtin function 'to_local'}} -+ // expected-note@+8 {{'to_local' is a builtin with type 'void *(void *)'}} -+ // expected-error@+8 {{cannot redeclare builtin function 'to_global'}} -+ // expected-note@+7 {{'to_global' is a builtin with type 'void *(void *)'}} -+ // expected-error@+7 {{cannot redeclare builtin function 'get_kernel_work_group_size'}} -+ // expected-note@+6 {{'get_kernel_work_group_size' is a builtin with type 'unsigned int ()'}} -+#endif -+ -+int get_pipe_max_packets(int); -+int to_local(int); -+int to_global(int); -+int get_kernel_work_group_size(int); -+ -+kernel void test(global int *dst) {} -diff --git a/clang/test/SemaOpenCL/image-features.cl b/clang/test/SemaOpenCL/image-features.cl -new file mode 100644 -index 000000000000..ace6913bb31e ---- /dev/null -+++ b/clang/test/SemaOpenCL/image-features.cl -@@ -0,0 +1,20 @@ -+// RUN: %clang_cc1 -cl-std=cl2.0 -fsyntax-only -verify %s -triple spir-unknown-unknown -+// RUN: %clang_cc1 -cl-std=cl3.0 -cl-ext=__opencl_c_images -fsyntax-only -verify %s -triple spir-unknown-unknown -+// RUN: %clang_cc1 -cl-std=cl3.0 -cl-ext=__opencl_c_3d_image_writes -fsyntax-only -verify %s -triple spir-unknown-unknown -+// RUN: %clang_cc1 -cl-std=cl3.0 -cl-ext=__opencl_c_read_write_images -fsyntax-only -verify %s -triple spir-unknown-unknown -+ -+#if defined(__opencl_c_read_write_images) && defined(__opencl_c_3d_image_writes) -+ // expected-no-diagnostics -+#endif -+ -+__kernel void write_3d_image(__write_only image3d_t i) {} -+ -+#ifndef __opencl_c_3d_image_writes -+ // expected-error@-3 {{use of type '__write_only image3d_t' requires __opencl_c_3d_image_writes extension to be enabled}} -+#endif -+ -+__kernel void read_write_3d_image(__read_write image3d_t i) { } -+ -+#ifndef __opencl_c_read_write_images -+ // expected-error@-3 {{use of type '__read_write image3d_t' requires __opencl_c_read_write_images extension to be enabled}} -+#endif -diff --git a/clang/test/SemaOpenCL/invalid-pipe-builtin-cl2.0.cl b/clang/test/SemaOpenCL/invalid-pipe-builtin-cl2.0.cl -index 36e76621d24a..38b0a04726e3 100644 ---- a/clang/test/SemaOpenCL/invalid-pipe-builtin-cl2.0.cl -+++ b/clang/test/SemaOpenCL/invalid-pipe-builtin-cl2.0.cl -@@ -1,4 +1,5 @@ - // RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL2.0 -cl-ext=+cl_khr_subgroups -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=__opencl_c_subgroups,__opencl_c_pipes - - #pragma OPENCL EXTENSION cl_khr_subgroups : enable - -diff --git a/clang/test/SemaOpenCL/storageclass-cl20.cl b/clang/test/SemaOpenCL/storageclass-cl20.cl -index 581701d2a6a5..469c526ebc30 100644 ---- a/clang/test/SemaOpenCL/storageclass-cl20.cl -+++ b/clang/test/SemaOpenCL/storageclass-cl20.cl -@@ -1,4 +1,5 @@ - // RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL2.0 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=__opencl_c_program_scope_global_variables,__opencl_c_generic_address_space - - int G2 = 0; - global int G3 = 0; -diff --git a/clang/utils/TableGen/ClangOpenCLBuiltinEmitter.cpp b/clang/utils/TableGen/ClangOpenCLBuiltinEmitter.cpp -index 41d33b550680..f50652493e5e 100644 ---- a/clang/utils/TableGen/ClangOpenCLBuiltinEmitter.cpp -+++ b/clang/utils/TableGen/ClangOpenCLBuiltinEmitter.cpp -@@ -56,6 +56,7 @@ - //===----------------------------------------------------------------------===// - - #include "TableGenBackends.h" -+#include "clang/Basic/OpenCLOptions.h" - #include "llvm/ADT/MapVector.h" - #include "llvm/ADT/STLExtras.h" - #include "llvm/ADT/SmallString.h" -@@ -69,6 +70,7 @@ - #include "llvm/TableGen/Record.h" - #include "llvm/TableGen/StringMatcher.h" - #include "llvm/TableGen/TableGenBackend.h" -+#include <numeric> - #include <set> - - using namespace llvm; -@@ -228,6 +230,10 @@ private: - // The function "tan", having the same signatures, would be mapped to the - // same entry (<I1, I2, I3>). - MapVector<BuiltinIndexListTy *, BuiltinTableEntries> SignatureListMap; -+ -+ // Encode all versions -+ unsigned short -+ EncodeBuiltinVersions(std::vector<Record *> BuiltinVersionsRecords) const; - }; - } // namespace - -@@ -338,12 +344,10 @@ struct OpenCLBuiltinStruct { - const bool IsConst; - // Function attribute __attribute__((convergent)) - const bool IsConv; -+ // All opencl versions encoded -+ const unsigned char AllVersions : 5; - // OpenCL extension(s) required for this overload. - const unsigned short Extension; -- // First OpenCL version in which this overload was introduced (e.g. CL20). -- const unsigned short MinVersion; -- // First OpenCL version in which this overload was removed (e.g. CL20). -- const unsigned short MaxVersion; - }; - - )"; -@@ -503,11 +507,9 @@ void BuiltinNameEmitter::EmitBuiltinTable() { - << (Overload.first->getValueAsBit("IsPure")) << ", " - << (Overload.first->getValueAsBit("IsConst")) << ", " - << (Overload.first->getValueAsBit("IsConv")) << ", " -- << FunctionExtensionIndex[ExtName] << ", " -- << Overload.first->getValueAsDef("MinVersion")->getValueAsInt("ID") -- << ", " -- << Overload.first->getValueAsDef("MaxVersion")->getValueAsInt("ID") -- << " },\n"; -+ << EncodeBuiltinVersions( -+ Overload.first->getValueAsListOfDefs("Versions")) -+ << ", " << FunctionExtensionIndex[ExtName] << " },\n"; - Index++; - } - } -@@ -528,10 +530,8 @@ bool BuiltinNameEmitter::CanReuseSignature( - if (Rec->getValueAsBit("IsPure") == Rec2->getValueAsBit("IsPure") && - Rec->getValueAsBit("IsConst") == Rec2->getValueAsBit("IsConst") && - Rec->getValueAsBit("IsConv") == Rec2->getValueAsBit("IsConv") && -- Rec->getValueAsDef("MinVersion")->getValueAsInt("ID") == -- Rec2->getValueAsDef("MinVersion")->getValueAsInt("ID") && -- Rec->getValueAsDef("MaxVersion")->getValueAsInt("ID") == -- Rec2->getValueAsDef("MaxVersion")->getValueAsInt("ID") && -+ EncodeBuiltinVersions(Rec->getValueAsListOfDefs("Versions")) == -+ EncodeBuiltinVersions(Rec2->getValueAsListOfDefs("Versions")) && - Rec->getValueAsDef("Extension")->getName() == - Rec2->getValueAsDef("Extension")->getName()) { - return true; -@@ -806,6 +806,15 @@ static void OCL2Qual(ASTContext &Context, const OpenCLTypeStruct &Ty, - OS << "\n} // OCL2Qual\n"; - } - -+unsigned short BuiltinNameEmitter::EncodeBuiltinVersions( -+ std::vector<Record *> BuiltinVersionsRecords) const { -+ return std::accumulate( -+ BuiltinVersionsRecords.begin(), BuiltinVersionsRecords.end(), -+ (unsigned short)0, [](unsigned short C, Record *R) { -+ return C |= clang::OpenCLOptions::EncodeVersion(R->getValueAsInt("ID")); -+ }); -+} -+ - void clang::EmitClangOpenCLBuiltins(RecordKeeper &Records, raw_ostream &OS) { - BuiltinNameEmitter NameChecker(Records, OS); - NameChecker.Emit(); --- -2.17.1 - diff --git a/dynamic-layers/clang-layer/recipes-devtools/clang/files/llvm10-Remove-repo-name-in-LLVM-IR.patch b/dynamic-layers/clang-layer/recipes-devtools/clang/files/llvm10-Remove-repo-name-in-LLVM-IR.patch deleted file mode 100644 index 232ae063..00000000 --- a/dynamic-layers/clang-layer/recipes-devtools/clang/files/llvm10-Remove-repo-name-in-LLVM-IR.patch +++ /dev/null @@ -1,50 +0,0 @@ -From b53fd86ffdeacb9b13624bdb110fd25e8c35cb92 Mon Sep 17 00:00:00 2001 -From: Feng Zou <feng.zou@intel.com> -Date: Mon, 19 Oct 2020 14:43:38 +0800 -Subject: [PATCH] Remove repo name in LLVM IR - -Upstream-Status: Backport [Taken from opencl-clang patches, https://github.com/intel/opencl-clang/blob/ocl-open-100/patches/llvm/0003-Remove-repo-name-in-LLVM-IR.patch] -Signed-off-by: Feng Zou <feng.zou@intel.com> -Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> - ---- - llvm/cmake/modules/VersionFromVCS.cmake | 23 ++++++++++++----------- - 1 file changed, 12 insertions(+), 11 deletions(-) - -diff --git a/llvm/cmake/modules/VersionFromVCS.cmake b/llvm/cmake/modules/VersionFromVCS.cmake -index 1b6519b4b7c..8fd6b23bb34 100644 ---- a/llvm/cmake/modules/VersionFromVCS.cmake -+++ b/llvm/cmake/modules/VersionFromVCS.cmake -@@ -33,17 +33,18 @@ function(get_source_info path revision repository) - else() - set(remote "origin") - endif() -- execute_process(COMMAND ${GIT_EXECUTABLE} remote get-url ${remote} -- WORKING_DIRECTORY ${path} -- RESULT_VARIABLE git_result -- OUTPUT_VARIABLE git_output -- ERROR_QUIET) -- if(git_result EQUAL 0) -- string(STRIP "${git_output}" git_output) -- set(${repository} ${git_output} PARENT_SCOPE) -- else() -- set(${repository} ${path} PARENT_SCOPE) -- endif() -+ # Do not show repo name in IR -+ # execute_process(COMMAND ${GIT_EXECUTABLE} remote get-url ${remote} -+ # WORKING_DIRECTORY ${path} -+ # RESULT_VARIABLE git_result -+ # OUTPUT_VARIABLE git_output -+ # ERROR_QUIET) -+ # if(git_result EQUAL 0) -+ # string(STRIP "${git_output}" git_output) -+ # set(${repository} ${git_output} PARENT_SCOPE) -+ # else() -+ # set(${repository} ${path} PARENT_SCOPE) -+ # endif() - endif() - endif() - endfunction() --- -2.18.1 - diff --git a/dynamic-layers/clang-layer/recipes-devtools/clang/files/llvm10-skip-building-tests.patch b/dynamic-layers/clang-layer/recipes-devtools/clang/files/llvm10-skip-building-tests.patch deleted file mode 100644 index 8e58ec25..00000000 --- a/dynamic-layers/clang-layer/recipes-devtools/clang/files/llvm10-skip-building-tests.patch +++ /dev/null @@ -1,51 +0,0 @@ -From 455ce9c25df5313f4a6649cc27075bdfbe25af18 Mon Sep 17 00:00:00 2001 -From: Naveen Saini <naveen.kumar.saini@intel.com> -Date: Wed, 21 Aug 2019 14:35:31 +0800 -Subject: [PATCH] llvm-spirv: skip building tests - -Some of these need clang to be built and since we're building this in-tree, -that leads to problems when compiling libcxx, compiler-rt which aren't built -in-tree. - -Instead of using SPIRV_SKIP_CLANG_BUILD to skip clang build and adding this to -all components, disable the building of tests altogether. - -Upstream-Status: Inappropriate - -Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> -Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> ---- - CMakeLists.txt | 10 ---------- - 1 file changed, 10 deletions(-) - -diff --git a/CMakeLists.txt b/CMakeLists.txt -index b718c00..9805140 100644 ---- a/CMakeLists.txt -+++ b/CMakeLists.txt -@@ -24,13 +24,6 @@ if(LLVM_SPIRV_BUILD_EXTERNAL) - set(CMAKE_CXX_STANDARD 14) - set(CMAKE_CXX_STANDARD_REQUIRED ON) - -- if(LLVM_SPIRV_INCLUDE_TESTS) -- set(LLVM_TEST_COMPONENTS -- llvm-as -- llvm-dis -- ) -- endif(LLVM_SPIRV_INCLUDE_TESTS) -- - find_package(LLVM 10.0.0 REQUIRED - COMPONENTS - Analysis -@@ -61,9 +54,6 @@ set(LLVM_SPIRV_INCLUDE_DIRS ${CMAKE_CURRENT_SOURCE_DIR}/include) - - add_subdirectory(lib/SPIRV) - add_subdirectory(tools/llvm-spirv) --if(LLVM_SPIRV_INCLUDE_TESTS) -- add_subdirectory(test) --endif(LLVM_SPIRV_INCLUDE_TESTS) - - install( - FILES --- -2.7.4 - diff --git a/dynamic-layers/clang-layer/recipes-devtools/clang/files/llvm11-OpenCL-3.0-support.patch b/dynamic-layers/clang-layer/recipes-devtools/clang/files/llvm11-OpenCL-3.0-support.patch deleted file mode 100644 index 98545db0..00000000 --- a/dynamic-layers/clang-layer/recipes-devtools/clang/files/llvm11-OpenCL-3.0-support.patch +++ /dev/null @@ -1,8453 +0,0 @@ -From d767afa79d1c8153081eac1ef33e348cadbea5bb Mon Sep 17 00:00:00 2001 -From: Anton Zabaznov <anton.zabaznov@intel.com> -Date: Tue, 22 Sep 2020 19:03:50 +0300 -Subject: [PATCH] OpenCL 3.0 support - -Upstream-Status: Backport [Taken from opencl-clang patches, https://github.com/intel/opencl-clang/blob/ocl-open-110/patches/clang/0001-OpenCL-3.0-support.patch] -Signed-off-by: Anton Zabaznov <anton.zabaznov@intel.com> -Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> - - ---- - clang/include/clang/Basic/Builtins.def | 65 +- - clang/include/clang/Basic/Builtins.h | 13 +- - .../clang/Basic/DiagnosticParseKinds.td | 2 + - .../clang/Basic/DiagnosticSemaKinds.td | 7 + - clang/include/clang/Basic/LangOptions.def | 2 + - clang/include/clang/Basic/LangStandards.def | 4 + - .../include/clang/Basic/OpenCLExtensions.def | 15 + - clang/include/clang/Basic/OpenCLOptions.h | 174 +- - clang/include/clang/Driver/Options.td | 2 +- - clang/include/clang/Sema/Sema.h | 9 + - clang/lib/AST/ASTContext.cpp | 3 +- - clang/lib/Basic/Builtins.cpp | 28 +- - clang/lib/Basic/TargetInfo.cpp | 11 + - clang/lib/Basic/Targets.cpp | 1 - - clang/lib/CodeGen/CodeGenFunction.cpp | 6 +- - clang/lib/Frontend/CompilerInvocation.cpp | 22 +- - clang/lib/Frontend/InitPreprocessor.cpp | 6 +- - clang/lib/Headers/opencl-c-base.h | 75 +- - clang/lib/Headers/opencl-c.h | 3216 ++++++++++++++--- - clang/lib/Parse/ParseDecl.cpp | 12 +- - clang/lib/Parse/ParsePragma.cpp | 10 +- - clang/lib/Sema/OpenCLBuiltins.td | 153 +- - clang/lib/Sema/Sema.cpp | 47 +- - clang/lib/Sema/SemaChecking.cpp | 38 +- - clang/lib/Sema/SemaDecl.cpp | 18 +- - clang/lib/Sema/SemaDeclAttr.cpp | 9 +- - clang/lib/Sema/SemaDeclCXX.cpp | 12 +- - clang/lib/Sema/SemaLookup.cpp | 19 +- - clang/lib/Sema/SemaType.cpp | 14 +- - .../CodeGenOpenCL/addr-space-struct-arg.cl | 7 +- - .../address-spaces-conversions.cl | 2 + - .../CodeGenOpenCL/address-spaces-mangling.cl | 3 +- - clang/test/CodeGenOpenCL/address-spaces.cl | 4 + - .../amdgcn-automatic-variable.cl | 1 + - .../CodeGenOpenCL/amdgpu-sizeof-alignof.cl | 21 +- - .../CodeGenOpenCL/arm-integer-dot-product.cl | 1 + - .../test/CodeGenOpenCL/cl-uniform-wg-size.cl | 2 + - clang/test/CodeGenOpenCL/fpmath.cl | 2 + - .../generic-address-space-feature.cl | 28 + - .../intel-subgroups-avc-ext-types.cl | 1 + - .../kernels-have-spir-cc-by-default.cl | 3 + - clang/test/CodeGenOpenCL/logical-ops.cl | 1 + - clang/test/CodeGenOpenCL/no-half.cl | 1 + - clang/test/CodeGenOpenCL/pipe_builtin.cl | 3 + - clang/test/CodeGenOpenCL/pipe_types.cl | 1 + - clang/test/CodeGenOpenCL/printf.cl | 2 + - clang/test/CodeGenOpenCL/unroll-hint.cl | 1 + - clang/test/Driver/autocomplete.c | 2 + - clang/test/Driver/opencl.cl | 2 + - clang/test/Driver/unknown-std.cl | 1 + - clang/test/Frontend/stdlang.c | 1 + - clang/test/Headers/opencl-c-header.cl | 7 +- - clang/test/Index/pipe-size.cl | 7 + - clang/test/Preprocessor/init-aarch64.c | 12 +- - clang/test/Preprocessor/predefined-macros.c | 13 + - .../Sema/feature-extensions-simult-support.cl | 75 + - clang/test/Sema/features-ignore-pragma.cl | 24 + - clang/test/Sema/opencl-features-pipes.cl | 18 + - clang/test/Sema/opencl-features.cl | 128 + - clang/test/Sema/pipe_builtins_feature.cl | 21 + - .../address-spaces-conversions-cl2.0.cl | 3 + - clang/test/SemaOpenCL/address-spaces.cl | 1 + - .../SemaOpenCL/cl20-device-side-enqueue.cl | 16 +- - .../SemaOpenCL/forget-unsupported-builtins.cl | 22 + - clang/test/SemaOpenCL/image-features.cl | 20 + - .../SemaOpenCL/invalid-pipe-builtin-cl2.0.cl | 1 + - clang/test/SemaOpenCL/storageclass-cl20.cl | 1 + - .../TableGen/ClangOpenCLBuiltinEmitter.cpp | 35 +- - 68 files changed, 3749 insertions(+), 738 deletions(-) - create mode 100644 clang/test/CodeGenOpenCL/generic-address-space-feature.cl - create mode 100644 clang/test/Sema/feature-extensions-simult-support.cl - create mode 100644 clang/test/Sema/features-ignore-pragma.cl - create mode 100644 clang/test/Sema/opencl-features-pipes.cl - create mode 100644 clang/test/Sema/opencl-features.cl - create mode 100644 clang/test/Sema/pipe_builtins_feature.cl - create mode 100644 clang/test/SemaOpenCL/forget-unsupported-builtins.cl - create mode 100644 clang/test/SemaOpenCL/image-features.cl - -diff --git a/clang/include/clang/Basic/Builtins.def b/clang/include/clang/Basic/Builtins.def -index 5463b7dfc18c..e094beab9a6c 100644 ---- a/clang/include/clang/Basic/Builtins.def -+++ b/clang/include/clang/Basic/Builtins.def -@@ -111,6 +111,10 @@ - # define LANGBUILTIN(ID, TYPE, ATTRS, BUILTIN_LANG) BUILTIN(ID, TYPE, ATTRS) - #endif - -+#if defined(BUILTIN) && !defined(OPENCLBUILTIN) -+# define OPENCLBUILTIN(ID, TYPE, ATTRS, BUILTIN_LANG, FEATURE) BUILTIN(ID, TYPE, ATTRS) -+#endif -+ - // Standard libc/libm functions: - BUILTIN(__builtin_atan2 , "ddd" , "Fne") - BUILTIN(__builtin_atan2f, "fff" , "Fne") -@@ -1529,50 +1533,54 @@ BUILTIN(__builtin_coro_param, "bv*v*", "n") - - // OpenCL v2.0 s6.13.16, s9.17.3.5 - Pipe functions. - // We need the generic prototype, since the packet type could be anything. --LANGBUILTIN(read_pipe, "i.", "tn", OCLC20_LANG) --LANGBUILTIN(write_pipe, "i.", "tn", OCLC20_LANG) -+OPENCLBUILTIN(read_pipe, "i.", "tn", OCLC2P_LANG, "__opencl_c_pipes") -+OPENCLBUILTIN(write_pipe, "i.", "tn", OCLC2P_LANG, "__opencl_c_pipes") - --LANGBUILTIN(reserve_read_pipe, "i.", "tn", OCLC20_LANG) --LANGBUILTIN(reserve_write_pipe, "i.", "tn", OCLC20_LANG) -+OPENCLBUILTIN(reserve_read_pipe, "i.", "tn", OCLC2P_LANG, "__opencl_c_pipes") -+OPENCLBUILTIN(reserve_write_pipe, "i.", "tn", OCLC2P_LANG, "__opencl_c_pipes") - --LANGBUILTIN(commit_write_pipe, "v.", "tn", OCLC20_LANG) --LANGBUILTIN(commit_read_pipe, "v.", "tn", OCLC20_LANG) -+OPENCLBUILTIN(commit_write_pipe, "v.", "tn", OCLC2P_LANG, "__opencl_c_pipes") -+OPENCLBUILTIN(commit_read_pipe, "v.", "tn", OCLC2P_LANG, "__opencl_c_pipes") - --LANGBUILTIN(sub_group_reserve_read_pipe, "i.", "tn", OCLC20_LANG) --LANGBUILTIN(sub_group_reserve_write_pipe, "i.", "tn", OCLC20_LANG) -+OPENCLBUILTIN(sub_group_reserve_read_pipe, "i.", "tn", OCLC2P_LANG, "__opencl_c_pipes") -+OPENCLBUILTIN(sub_group_reserve_write_pipe, "i.", "tn", OCLC2P_LANG, "__opencl_c_pipes") - --LANGBUILTIN(sub_group_commit_read_pipe, "v.", "tn", OCLC20_LANG) --LANGBUILTIN(sub_group_commit_write_pipe, "v.", "tn", OCLC20_LANG) -+OPENCLBUILTIN(sub_group_commit_read_pipe, "v.", "tn", OCLC2P_LANG, "__opencl_c_pipes") -+OPENCLBUILTIN(sub_group_commit_write_pipe, "v.", "tn", OCLC2P_LANG, "__opencl_c_pipes") - --LANGBUILTIN(work_group_reserve_read_pipe, "i.", "tn", OCLC20_LANG) --LANGBUILTIN(work_group_reserve_write_pipe, "i.", "tn", OCLC20_LANG) -+OPENCLBUILTIN(work_group_reserve_read_pipe, "i.", "tn", OCLC2P_LANG, "__opencl_c_pipes") -+OPENCLBUILTIN(work_group_reserve_write_pipe, "i.", "tn", OCLC2P_LANG, "__opencl_c_pipes") - --LANGBUILTIN(work_group_commit_read_pipe, "v.", "tn", OCLC20_LANG) --LANGBUILTIN(work_group_commit_write_pipe, "v.", "tn", OCLC20_LANG) -+OPENCLBUILTIN(work_group_commit_read_pipe, "v.", "tn", OCLC2P_LANG, "__opencl_c_pipes") -+OPENCLBUILTIN(work_group_commit_write_pipe, "v.", "tn", OCLC2P_LANG, "__opencl_c_pipes") - --LANGBUILTIN(get_pipe_num_packets, "Ui.", "tn", OCLC20_LANG) --LANGBUILTIN(get_pipe_max_packets, "Ui.", "tn", OCLC20_LANG) -+OPENCLBUILTIN(get_pipe_num_packets, "Ui.", "tn", OCLC2P_LANG, "__opencl_c_pipes") -+OPENCLBUILTIN(get_pipe_max_packets, "Ui.", "tn", OCLC2P_LANG, "__opencl_c_pipes") - - // OpenCL v2.0 s6.13.17 - Enqueue kernel functions. - // Custom builtin check allows to perform special check of passed block arguments. --LANGBUILTIN(enqueue_kernel, "i.", "tn", OCLC20_LANG) --LANGBUILTIN(get_kernel_work_group_size, "Ui.", "tn", OCLC20_LANG) --LANGBUILTIN(get_kernel_preferred_work_group_size_multiple, "Ui.", "tn", OCLC20_LANG) --LANGBUILTIN(get_kernel_max_sub_group_size_for_ndrange, "Ui.", "tn", OCLC20_LANG) --LANGBUILTIN(get_kernel_sub_group_count_for_ndrange, "Ui.", "tn", OCLC20_LANG) -+OPENCLBUILTIN(enqueue_kernel, "i.", "tn", OCLC2P_LANG, -+ "__opencl_c_device_enqueue") -+OPENCLBUILTIN(get_kernel_work_group_size, "Ui.", "tn", OCLC2P_LANG, -+ "__opencl_c_device_enqueue") -+OPENCLBUILTIN(get_kernel_preferred_work_group_size_multiple, "Ui.", "tn", -+ OCLC2P_LANG, "__opencl_c_device_enqueue") -+OPENCLBUILTIN(get_kernel_max_sub_group_size_for_ndrange, "Ui.", "tn", -+ OCLC2P_LANG, "__opencl_c_device_enqueue") -+OPENCLBUILTIN(get_kernel_sub_group_count_for_ndrange, "Ui.", "tn", OCLC2P_LANG, "__opencl_c_device_enqueue") - - // OpenCL v2.0 s6.13.9 - Address space qualifier functions. - // FIXME: Pointer parameters of OpenCL builtins should have their address space - // requirement defined. --LANGBUILTIN(to_global, "v*v*", "tn", OCLC20_LANG) --LANGBUILTIN(to_local, "v*v*", "tn", OCLC20_LANG) --LANGBUILTIN(to_private, "v*v*", "tn", OCLC20_LANG) -+OPENCLBUILTIN(to_global, "v*v*", "tn", OCLC2P_LANG, "__opencl_c_generic_address_space") -+OPENCLBUILTIN(to_local, "v*v*", "tn", OCLC2P_LANG, "__opencl_c_generic_address_space") -+OPENCLBUILTIN(to_private, "v*v*", "tn", OCLC2P_LANG, "__opencl_c_generic_address_space") - - // OpenCL half load/store builtin --LANGBUILTIN(__builtin_store_half, "vdh*", "n", ALL_OCLC_LANGUAGES) --LANGBUILTIN(__builtin_store_halff, "vfh*", "n", ALL_OCLC_LANGUAGES) --LANGBUILTIN(__builtin_load_half, "dhC*", "nc", ALL_OCLC_LANGUAGES) --LANGBUILTIN(__builtin_load_halff, "fhC*", "nc", ALL_OCLC_LANGUAGES) -+OPENCLBUILTIN(__builtin_store_half, "vdh*", "n", ALL_OCLC_LANGUAGES, "") -+OPENCLBUILTIN(__builtin_store_halff, "vfh*", "n", ALL_OCLC_LANGUAGES, "") -+OPENCLBUILTIN(__builtin_load_half, "dhC*", "nc", ALL_OCLC_LANGUAGES, "") -+OPENCLBUILTIN(__builtin_load_halff, "fhC*", "nc", ALL_OCLC_LANGUAGES, "") - - // Builtins for os_log/os_trace - BUILTIN(__builtin_os_log_format_buffer_size, "zcC*.", "p:0:nut") -@@ -1593,3 +1601,4 @@ BUILTIN(__builtin_ms_va_copy, "vc*&c*&", "n") - #undef BUILTIN - #undef LIBBUILTIN - #undef LANGBUILTIN -+#undef OPENCLBUILTIN -diff --git a/clang/include/clang/Basic/Builtins.h b/clang/include/clang/Basic/Builtins.h -index 15bfcf797917..ba3e7853444c 100644 ---- a/clang/include/clang/Basic/Builtins.h -+++ b/clang/include/clang/Basic/Builtins.h -@@ -33,13 +33,13 @@ enum LanguageID { - CXX_LANG = 0x4, // builtin for cplusplus only. - OBJC_LANG = 0x8, // builtin for objective-c and objective-c++ - MS_LANG = 0x10, // builtin requires MS mode. -- OCLC20_LANG = 0x20, // builtin for OpenCL C 2.0 only. -+ OCLC2P_LANG = 0x20, // builtin for OpenCL C 2.0+ versions. - OCLC1X_LANG = 0x40, // builtin for OpenCL C 1.x only. - OMP_LANG = 0x80, // builtin requires OpenMP. - ALL_LANGUAGES = C_LANG | CXX_LANG | OBJC_LANG, // builtin for all languages. - ALL_GNU_LANGUAGES = ALL_LANGUAGES | GNU_LANG, // builtin requires GNU mode. - ALL_MS_LANGUAGES = ALL_LANGUAGES | MS_LANG, // builtin requires MS mode. -- ALL_OCLC_LANGUAGES = OCLC1X_LANG | OCLC20_LANG // builtin for OCLC languages. -+ ALL_OCLC_LANGUAGES = OCLC1X_LANG | OCLC2P_LANG // builtin for OCLC languages. - }; - - namespace Builtin { -@@ -235,6 +235,10 @@ public: - /// for non-builtins. - bool canBeRedeclared(unsigned ID) const; - -+ bool requiresFeatures(unsigned ID) const { -+ return requiresFeatures(getRecord(ID)); -+ } -+ - private: - const Info &getRecord(unsigned ID) const; - -@@ -242,6 +246,11 @@ private: - bool builtinIsSupported(const Builtin::Info &BuiltinInfo, - const LangOptions &LangOpts); - -+ bool OclBuiltinIsSupported(const Builtin::Info &BuiltinInfo, -+ const LangOptions &LangOpts) const; -+ -+ bool requiresFeatures(const Builtin::Info &BuiltinInfo) const; -+ - /// Helper function for isPrintfLike and isScanfLike. - bool isLike(unsigned ID, unsigned &FormatIdx, bool &HasVAListArg, - const char *Fmt) const; -diff --git a/clang/include/clang/Basic/DiagnosticParseKinds.td b/clang/include/clang/Basic/DiagnosticParseKinds.td -index 1038a4119d4c..ba1b1b67a131 100644 ---- a/clang/include/clang/Basic/DiagnosticParseKinds.td -+++ b/clang/include/clang/Basic/DiagnosticParseKinds.td -@@ -1206,6 +1206,8 @@ def warn_pragma_unsupported_extension : Warning< - "unsupported OpenCL extension %0 - ignoring">, InGroup<IgnoredPragmas>; - def warn_pragma_extension_is_core : Warning< - "OpenCL extension %0 is core feature or supported optional core feature - ignoring">, InGroup<DiagGroup<"pedantic-core-features">>, DefaultIgnore; -+def warn_opencl_pragma_feature_ignore : Warning< -+ "OpenCL feature support can't be controlled via pragma, ignoring">, InGroup<IgnoredPragmas>; - - // OpenCL errors. - def err_opencl_taking_function_address_parser : Error< -diff --git a/clang/include/clang/Basic/DiagnosticSemaKinds.td b/clang/include/clang/Basic/DiagnosticSemaKinds.td -index 941f2cafc372..aa9e8b7b011b 100644 ---- a/clang/include/clang/Basic/DiagnosticSemaKinds.td -+++ b/clang/include/clang/Basic/DiagnosticSemaKinds.td -@@ -9741,6 +9741,13 @@ def ext_opencl_ext_vector_type_rgba_selector: ExtWarn< - def err_openclcxx_placement_new : Error< - "use of placement new requires explicit declaration">; - -+def err_opencl_pipes_require_feat : Error< -+ "usage of OpenCL pipes requires feature support">; -+def err_opencl_memory_scope_require_feat : Error< -+ "usage of memory scope requires feature support">; -+def err_opencl_memory_ordering_require_feat : Error< -+ "usage of memory ordering requires feature support">; -+ - // MIG routine annotations. - def warn_mig_server_routine_does_not_return_kern_return_t : Warning< - "'mig_server_routine' attribute only applies to routines that return a kern_return_t">, -diff --git a/clang/include/clang/Basic/LangOptions.def b/clang/include/clang/Basic/LangOptions.def -index 70f68d664bb7..e683f5956975 100644 ---- a/clang/include/clang/Basic/LangOptions.def -+++ b/clang/include/clang/Basic/LangOptions.def -@@ -215,6 +215,8 @@ LANGOPT(OpenCL , 1, 0, "OpenCL") - LANGOPT(OpenCLVersion , 32, 0, "OpenCL C version") - LANGOPT(OpenCLCPlusPlus , 1, 0, "C++ for OpenCL") - LANGOPT(OpenCLCPlusPlusVersion , 32, 0, "C++ for OpenCL version") -+LANGOPT(OpenCLGenericKeyword , 1, 0, "OpenCL generic keyword") -+LANGOPT(OpenCLPipeKeyword , 1, 0, "OpenCL pipe keyword") - LANGOPT(NativeHalfType , 1, 0, "Native half type support") - LANGOPT(NativeHalfArgsAndReturns, 1, 0, "Native half args and returns") - LANGOPT(HalfArgsAndReturns, 1, 0, "half args and returns") -diff --git a/clang/include/clang/Basic/LangStandards.def b/clang/include/clang/Basic/LangStandards.def -index b09568e8b3e8..7b915c312746 100644 ---- a/clang/include/clang/Basic/LangStandards.def -+++ b/clang/include/clang/Basic/LangStandards.def -@@ -167,6 +167,9 @@ LANGSTANDARD(opencl12, "cl1.2", - LANGSTANDARD(opencl20, "cl2.0", - OpenCL, "OpenCL 2.0", - LineComment | C99 | Digraphs | HexFloat | OpenCL) -+LANGSTANDARD(opencl30, "cl3.0", -+ OpenCL, "OpenCL 3.0", -+ LineComment | C99 | Digraphs | HexFloat | OpenCL) - LANGSTANDARD(openclcpp, "clc++", - OpenCL, "C++ for OpenCL", - LineComment | CPlusPlus | CPlusPlus11 | CPlusPlus14 | CPlusPlus17 | -@@ -176,6 +179,7 @@ LANGSTANDARD_ALIAS_DEPR(opencl10, "CL") - LANGSTANDARD_ALIAS_DEPR(opencl11, "CL1.1") - LANGSTANDARD_ALIAS_DEPR(opencl12, "CL1.2") - LANGSTANDARD_ALIAS_DEPR(opencl20, "CL2.0") -+LANGSTANDARD_ALIAS_DEPR(opencl30, "CL3.0") - LANGSTANDARD_ALIAS_DEPR(openclcpp, "CLC++") - - // CUDA -diff --git a/clang/include/clang/Basic/OpenCLExtensions.def b/clang/include/clang/Basic/OpenCLExtensions.def -index 1ae36b32fb0a..d1574164f9b2 100644 ---- a/clang/include/clang/Basic/OpenCLExtensions.def -+++ b/clang/include/clang/Basic/OpenCLExtensions.def -@@ -100,6 +100,21 @@ OPENCLEXT_INTERNAL(cl_intel_subgroups, 120, ~0U) - OPENCLEXT_INTERNAL(cl_intel_subgroups_short, 120, ~0U) - OPENCLEXT_INTERNAL(cl_intel_device_side_avc_motion_estimation, 120, ~0U) - -+OPENCLEXT_INTERNAL(__opencl_c_pipes, 200, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_generic_address_space, 200, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_work_group_collective_functions, 200, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_atomic_order_acq_rel, 200, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_atomic_order_seq_cst, 200, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_atomic_scope_device, 200, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_atomic_scope_all_devices, 200, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_subgroups, 200, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_3d_image_writes, 100, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_device_enqueue, 200, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_read_write_images, 200, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_program_scope_global_variables, 200, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_fp64, 120, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_int64, 100, ~0U) -+OPENCLEXT_INTERNAL(__opencl_c_images, 100, ~0U) - #undef OPENCLEXT_INTERNAL - - #ifdef OPENCLEXT -diff --git a/clang/include/clang/Basic/OpenCLOptions.h b/clang/include/clang/Basic/OpenCLOptions.h -index 15661154eab5..d5800f5611c6 100644 ---- a/clang/include/clang/Basic/OpenCLOptions.h -+++ b/clang/include/clang/Basic/OpenCLOptions.h -@@ -21,6 +21,10 @@ namespace clang { - - /// OpenCL supported extensions and optional core features - class OpenCLOptions { -+ // OpenCL Version -+ unsigned CLVer = 120; -+ bool IsOpenCLCPlusPlus = false; -+ - struct Info { - bool Supported; // Is this option supported - bool Enabled; // Is this option enabled -@@ -31,7 +35,38 @@ class OpenCLOptions { - :Supported(S), Enabled(E), Avail(A), Core(C){} - }; - llvm::StringMap<Info> OptMap; -+ - public: -+ void setOpenCLVersion(const LangOptions &LO) { -+ IsOpenCLCPlusPlus = LO.OpenCLCPlusPlus; -+ CLVer = IsOpenCLCPlusPlus ? 200 : LO.OpenCLVersion; -+ } -+ -+ // Get extension which is semantically equivalent to a given feature -+ // if exists (e.g. __opencl_c_subgroups -> cl_khr_subgroups) -+ llvm::Optional<StringRef> getEquivalentExtension(StringRef Feature) const { -+ return llvm::StringSwitch<llvm::Optional<StringRef>>(Feature) -+ .Case("__opencl_c_3d_image_writes", -+ Optional<StringRef>("cl_khr_3d_image_writes")) -+ .Case("__opencl_c_subgroups", Optional<StringRef>("cl_khr_subgroups")) -+ .Case("__opencl_c_fp64", Optional<StringRef>("cl_khr_fp64")) -+ .Default(Optional<StringRef>()); -+ } -+ -+ // Same as above but for extensions -+ llvm::Optional<StringRef> getEquivalentFeature(StringRef Extension) const { -+ return llvm::StringSwitch<llvm::Optional<StringRef>>(Extension) -+ .Case("cl_khr_3d_image_writes", -+ Optional<StringRef>("__opencl_c_3d_image_writes")) -+ .Case("cl_khr_subgroups", Optional<StringRef>("__opencl_c_subgroups")) -+ .Case("cl_khr_fp64", Optional<StringRef>("__opencl_c_fp64")) -+ .Default(Optional<StringRef>()); -+ } -+ -+ bool isFeature(llvm::StringRef Ext) const { -+ return Ext.startswith("__opencl_c"); -+ } -+ - bool isKnown(llvm::StringRef Ext) const { - return OptMap.find(Ext) != OptMap.end(); - } -@@ -42,32 +77,88 @@ public: - - // Is supported as either an extension or an (optional) core feature for - // OpenCL version \p CLVer. -- bool isSupported(llvm::StringRef Ext, const LangOptions &LO) const { -+ bool isSupported(llvm::StringRef Ext) const { - // In C++ mode all extensions should work at least as in v2.0. -- auto CLVer = LO.OpenCLCPlusPlus ? 200 : LO.OpenCLVersion; - auto I = OptMap.find(Ext)->getValue(); - return I.Supported && I.Avail <= CLVer; - } - - // Is supported (optional) OpenCL core features for OpenCL version \p CLVer. - // For supported extension, return false. -- bool isSupportedCore(llvm::StringRef Ext, const LangOptions &LO) const { -+ bool isSupportedCore(llvm::StringRef Ext) const { - // In C++ mode all extensions should work at least as in v2.0. -- auto CLVer = LO.OpenCLCPlusPlus ? 200 : LO.OpenCLVersion; - auto I = OptMap.find(Ext)->getValue(); - return I.Supported && I.Avail <= CLVer && I.Core != ~0U && CLVer >= I.Core; - } - - // Is supported OpenCL extension for OpenCL version \p CLVer. - // For supported (optional) core feature, return false. -- bool isSupportedExtension(llvm::StringRef Ext, const LangOptions &LO) const { -+ bool isSupportedExtension(llvm::StringRef Ext) const { - // In C++ mode all extensions should work at least as in v2.0. -- auto CLVer = LO.OpenCLCPlusPlus ? 200 : LO.OpenCLVersion; - auto I = OptMap.find(Ext)->getValue(); - return I.Supported && I.Avail <= CLVer && (I.Core == ~0U || CLVer < I.Core); - } - -+ // Support features whose support is directly related to the -+ // specific OpenCL version. For example, OpenCL 2.0 supports -+ // all features that are optional in 3.0 -+ void adjustFeatures() { -+ // Support int64 by default (assume compiling for FULL profile) -+ OptMap["__opencl_c_int64"].Supported = true; -+ -+ if (CLVer >= 300) { -+ // Simultaneously support extension and corresponding feature -+ for (llvm::StringRef F : -+ {"__opencl_c_subgroups", "__opencl_c_3d_image_writes", -+ "__opencl_c_fp64"}) { -+ auto Ext = getEquivalentExtension(F); -+ OptMap[*Ext].Supported = OptMap[F].Supported; -+ } -+ -+ // OpenCL C compilers that define the feature macro __opencl_c_pipes or -+ // or __opencl_c_device_enqueue must also define the -+ // feature macro __opencl_c_generic_address_space. -+ if (OptMap["__opencl_c_pipes"].Supported || -+ OptMap["__opencl_c_device_enqueue"].Supported) -+ OptMap["__opencl_c_generic_address_space"].Supported = true; -+ -+ // OpenCL C compilers that define the feature macro -+ // __opencl_c_3d_image_writes or __opencl_c_read_write_images must also -+ // define the feature macro __opencl_c_images -+ if (OptMap["__opencl_c_3d_image_writes"].Supported || -+ OptMap["__opencl_c_read_write_images"].Supported) -+ OptMap["__opencl_c_images"].Supported = true; -+ -+ // All other features are already supported with options -+ // or in target settings -+ return; -+ } -+ -+ auto FeaturesRange = llvm::make_filter_range( -+ OptMap, [&](llvm::StringMapEntry<Info> &OptVal) { -+ auto Opt = OptVal.getKey(); -+ return isFeature(Opt); -+ }); -+ -+ for (auto &It : FeaturesRange) { -+ auto &Info = It.getValue(); -+ // For OpenCL version less then 3.0 some -+ // features should be supported simulateneously -+ // with specific extension -+ if (Optional<llvm::StringRef> Ext = getEquivalentExtension(It.getKey())) -+ Info.Supported = Info.Enabled = OptMap[*Ext].Supported; -+ else if (Info.Avail <= CLVer) -+ Info.Supported = Info.Enabled = true; -+ } -+ } -+ - void enable(llvm::StringRef Ext, bool V = true) { -+ // Ignore disabling extensions if corresponding features -+ // already supported for OpenCL version higher then 3.0 -+ if (CLVer >= 300) -+ if (Optional<llvm::StringRef> F = getEquivalentFeature(Ext)) -+ if (V != OptMap[*F].Enabled) -+ return; - OptMap[Ext].Enabled = V; - } - -@@ -96,7 +187,7 @@ public: - OptMap[Ext].Supported = V; - } - -- OpenCLOptions(){ -+ OpenCLOptions() { - #define OPENCLEXT_INTERNAL(Ext, AvailVer, CoreVer) \ - OptMap[#Ext].Avail = AvailVer; \ - OptMap[#Ext].Core = CoreVer; -@@ -104,35 +195,86 @@ public: - } - - void addSupport(const OpenCLOptions &Opts) { -+ assert(IsOpenCLCPlusPlus == Opts.IsOpenCLCPlusPlus && CLVer == Opts.CLVer); - for (auto &I:Opts.OptMap) -- if (I.second.Supported) -+ if (I.second.Supported) { - OptMap[I.getKey()].Supported = true; -+ // All features are enabled as they are supported -+ if (isFeature(I.getKey())) -+ OptMap[I.getKey()].Enabled = true; -+ } -+ if (CLVer >= 300) { -+ // Enabling extensions with respect to features -+ for (llvm::StringRef Ext : -+ {"cl_khr_3d_image_writes", "cl_khr_subgroups", "cl_khr_fp64"}) { -+ auto Feature = getEquivalentFeature(Ext); -+ enable(Ext, OptMap[*Feature].Enabled); -+ } -+ } - } - - void copy(const OpenCLOptions &Opts) { -+ CLVer = Opts.CLVer; -+ IsOpenCLCPlusPlus = Opts.IsOpenCLCPlusPlus; - OptMap = Opts.OptMap; - } - - // Turn on or off support of all options. - void supportAll(bool On = true) { -- for (llvm::StringMap<Info>::iterator I = OptMap.begin(), -- E = OptMap.end(); I != E; ++I) -- I->second.Supported = On; -+ for (llvm::StringMap<Info>::iterator I = OptMap.begin(), E = OptMap.end(); -+ I != E; ++I) -+ if (!isFeature(I->getKey())) -+ I->second.Supported = On; - } - - void disableAll() { -- for (llvm::StringMap<Info>::iterator I = OptMap.begin(), -- E = OptMap.end(); I != E; ++I) -- I->second.Enabled = false; -+ for (llvm::StringMap<Info>::iterator I = OptMap.begin(), E = OptMap.end(); -+ I != E; ++I) { -+ auto Ext = I->getKey(); -+ if (!isFeature(Ext)) -+ enable(Ext, false); -+ } - } - -- void enableSupportedCore(LangOptions LO) { -+ void enableSupportedCore() { - for (llvm::StringMap<Info>::iterator I = OptMap.begin(), E = OptMap.end(); - I != E; ++I) -- if (isSupportedCore(I->getKey(), LO)) -+ if (isSupportedCore(I->getKey())) - I->second.Enabled = true; - } - -+ // This enum specifies how OpenCL versions map into values -+ // for encoding. This is used when generating built-ins -+ // from tablegen -+ enum OpenCLVersionsEncodings : unsigned short { -+ OPENCL_C_100_CODE = 0x1, -+ OPENCL_C_110_CODE = 0x2, -+ OPENCL_C_120_CODE = 0x4, -+ OPENCL_C_200_CODE = 0x8, -+ OPENCL_C_300_CODE = 0x10, -+ OPENCL_C_ALL_CODE = 0x1f -+ }; -+ -+ // Encode version into single integer -+ static unsigned short EncodeVersion(unsigned OpenCLVersion) { -+ switch (OpenCLVersion) { -+ default: -+ llvm_unreachable("Unknown OpenCL version"); -+ case 0: -+ return OpenCLVersionsEncodings::OPENCL_C_ALL_CODE; -+ case 100: -+ return OpenCLVersionsEncodings::OPENCL_C_100_CODE; -+ case 110: -+ return OpenCLVersionsEncodings::OPENCL_C_110_CODE; -+ case 120: -+ return OpenCLVersionsEncodings::OPENCL_C_120_CODE; -+ case 200: -+ return OpenCLVersionsEncodings::OPENCL_C_200_CODE; -+ case 300: -+ return OpenCLVersionsEncodings::OPENCL_C_300_CODE; -+ } -+ } -+ - friend class ASTWriter; - friend class ASTReader; - }; -diff --git a/clang/include/clang/Driver/Options.td b/clang/include/clang/Driver/Options.td -index 966cb907b7e2..62536f2470c2 100644 ---- a/clang/include/clang/Driver/Options.td -+++ b/clang/include/clang/Driver/Options.td -@@ -557,7 +557,7 @@ def cl_mad_enable : Flag<["-"], "cl-mad-enable">, Group<opencl_Group>, Flags<[CC - def cl_no_signed_zeros : Flag<["-"], "cl-no-signed-zeros">, Group<opencl_Group>, Flags<[CC1Option]>, - HelpText<"OpenCL only. Allow use of less precise no signed zeros computations in the generated binary.">; - def cl_std_EQ : Joined<["-"], "cl-std=">, Group<opencl_Group>, Flags<[CC1Option]>, -- HelpText<"OpenCL language standard to compile for.">, Values<"cl,CL,cl1.1,CL1.1,cl1.2,CL1.2,cl2.0,CL2.0,clc++,CLC++">; -+ HelpText<"OpenCL language standard to compile for.">, Values<"cl,CL,cl1.1,CL1.1,cl1.2,CL1.2,cl2.0,CL2.0,cl3.0,CL3.0,clc++,CLC++">; - def cl_denorms_are_zero : Flag<["-"], "cl-denorms-are-zero">, Group<opencl_Group>, - HelpText<"OpenCL only. Allow denormals to be flushed to zero.">; - def cl_fp32_correctly_rounded_divide_sqrt : Flag<["-"], "cl-fp32-correctly-rounded-divide-sqrt">, Group<opencl_Group>, Flags<[CC1Option]>, -diff --git a/clang/include/clang/Sema/Sema.h b/clang/include/clang/Sema/Sema.h -index 16a7084f6b08..5920c27229d8 100644 ---- a/clang/include/clang/Sema/Sema.h -+++ b/clang/include/clang/Sema/Sema.h -@@ -9930,6 +9930,10 @@ public: - /// \return true if type is disabled. - bool checkOpenCLDisabledDecl(const NamedDecl &D, const Expr &E); - -+ bool checkOpenCLSubgroupExtForCallExpr(CallExpr *Call); -+ -+ bool isSupportedOpenCLOMemoryOrdering(int64_t Ordering) const; -+ - //===--------------------------------------------------------------------===// - // OpenMP directives and clauses. - // -@@ -11527,6 +11531,11 @@ public: - /// that the user intended an assignment used as condition. - void DiagnoseEqualityWithExtraParens(ParenExpr *ParenE); - -+ template <typename DiagLocT, typename DiagInfoT> -+ void DiagnoseOpenCLRequiresOption(llvm::StringRef OpenCLOptName, -+ DiagLocT DiagLoc, DiagInfoT DiagInfo, -+ unsigned Selector, SourceRange SrcRange); -+ - /// CheckCXXBooleanCondition - Returns true if conversion to bool is invalid. - ExprResult CheckCXXBooleanCondition(Expr *CondExpr, bool IsConstexpr = false); - -diff --git a/clang/lib/AST/ASTContext.cpp b/clang/lib/AST/ASTContext.cpp -index bf51d35d9693..a54ea8475d7d 100644 ---- a/clang/lib/AST/ASTContext.cpp -+++ b/clang/lib/AST/ASTContext.cpp -@@ -1433,7 +1433,8 @@ void ASTContext::InitBuiltinTypes(const TargetInfo &Target, - ObjCSuperType = QualType(); - - // void * type -- if (LangOpts.OpenCLVersion >= 200) { -+ if (Target.getSupportedOpenCLOpts().isEnabled( -+ "__opencl_c_generic_address_space")) { - auto Q = VoidTy.getQualifiers(); - Q.setAddressSpace(LangAS::opencl_generic); - VoidPtrTy = getPointerType(getCanonicalType( -diff --git a/clang/lib/Basic/Builtins.cpp b/clang/lib/Basic/Builtins.cpp -index 0cd89df41b67..7a3067345098 100644 ---- a/clang/lib/Basic/Builtins.cpp -+++ b/clang/lib/Basic/Builtins.cpp -@@ -23,6 +23,8 @@ static const Builtin::Info BuiltinInfo[] = { - { #ID, TYPE, ATTRS, nullptr, ALL_LANGUAGES, nullptr }, - #define LANGBUILTIN(ID, TYPE, ATTRS, LANGS) \ - { #ID, TYPE, ATTRS, nullptr, LANGS, nullptr }, -+#define OPENCLBUILTIN(ID, TYPE, ATTRS, LANGS, FEATURE) \ -+ {#ID, TYPE, ATTRS, nullptr, LANGS, FEATURE}, - #define LIBBUILTIN(ID, TYPE, ATTRS, HEADER, LANGS) \ - { #ID, TYPE, ATTRS, HEADER, LANGS, nullptr }, - #include "clang/Basic/Builtins.def" -@@ -69,16 +71,17 @@ bool Builtin::Context::builtinIsSupported(const Builtin::Info &BuiltinInfo, - bool ObjCUnsupported = !LangOpts.ObjC && BuiltinInfo.Langs == OBJC_LANG; - bool OclC1Unsupported = (LangOpts.OpenCLVersion / 100) != 1 && - (BuiltinInfo.Langs & ALL_OCLC_LANGUAGES ) == OCLC1X_LANG; -- bool OclC2Unsupported = -- (LangOpts.OpenCLVersion != 200 && !LangOpts.OpenCLCPlusPlus) && -- (BuiltinInfo.Langs & ALL_OCLC_LANGUAGES) == OCLC20_LANG; -+ bool OclC2PUnsupported = -+ (BuiltinInfo.Langs & ALL_OCLC_LANGUAGES) == OCLC2P_LANG && -+ ((LangOpts.OpenCLVersion < 200 && !LangOpts.OpenCLCPlusPlus) || -+ !OclBuiltinIsSupported(BuiltinInfo, LangOpts)); - bool OclCUnsupported = !LangOpts.OpenCL && - (BuiltinInfo.Langs & ALL_OCLC_LANGUAGES); - bool OpenMPUnsupported = !LangOpts.OpenMP && BuiltinInfo.Langs == OMP_LANG; - bool CPlusPlusUnsupported = - !LangOpts.CPlusPlus && BuiltinInfo.Langs == CXX_LANG; - return !BuiltinsUnsupported && !MathBuiltinsUnsupported && !OclCUnsupported && -- !OclC1Unsupported && !OclC2Unsupported && !OpenMPUnsupported && -+ !OclC1Unsupported && !OclC2PUnsupported && !OpenMPUnsupported && - !GnuModeUnsupported && !MSModeUnsupported && !ObjCUnsupported && - !CPlusPlusUnsupported; - } -@@ -191,3 +194,20 @@ bool Builtin::Context::canBeRedeclared(unsigned ID) const { - (!hasReferenceArgsOrResult(ID) && - !hasCustomTypechecking(ID)); - } -+ -+bool Builtin::Context::OclBuiltinIsSupported( -+ const Builtin::Info &BuiltinInfo, const LangOptions &LangOpts) const { -+ if (!requiresFeatures(BuiltinInfo)) -+ return true; -+ -+ return llvm::StringSwitch<bool>(BuiltinInfo.Features) -+ .Case("__opencl_c_device_enqueue", LangOpts.Blocks) -+ .Case("__opencl_c_generic_address_space", LangOpts.OpenCLGenericKeyword) -+ .Case("__opencl_c_pipes", LangOpts.OpenCLPipeKeyword) -+ .Default(false); -+} -+ -+bool Builtin::Context::requiresFeatures( -+ const Builtin::Info &BuiltinInfo) const { -+ return BuiltinInfo.Features && llvm::StringRef(BuiltinInfo.Features) != ""; -+} -diff --git a/clang/lib/Basic/TargetInfo.cpp b/clang/lib/Basic/TargetInfo.cpp -index eccdc21d724a..6b4f524d5271 100644 ---- a/clang/lib/Basic/TargetInfo.cpp -+++ b/clang/lib/Basic/TargetInfo.cpp -@@ -387,6 +387,17 @@ void TargetInfo::adjust(LangOptions &Opts) { - HalfFormat = &llvm::APFloat::IEEEhalf(); - FloatFormat = &llvm::APFloat::IEEEsingle(); - LongDoubleFormat = &llvm::APFloat::IEEEquad(); -+ -+ auto &SupportedOCLOpts = getTargetOpts().SupportedOpenCLOptions; -+ -+ SupportedOCLOpts.setOpenCLVersion(Opts); -+ SupportedOCLOpts.adjustFeatures(); -+ -+ if (!Opts.OpenCLCPlusPlus && Opts.OpenCLVersion >= 200) -+ Opts.Blocks = SupportedOCLOpts.isSupported("__opencl_c_device_enqueue"); -+ Opts.OpenCLGenericKeyword = -+ SupportedOCLOpts.isSupported("__opencl_c_generic_address_space"); -+ Opts.OpenCLPipeKeyword = SupportedOCLOpts.isSupported("__opencl_c_pipes"); - } - - if (Opts.DoubleSize) { -diff --git a/clang/lib/Basic/Targets.cpp b/clang/lib/Basic/Targets.cpp -index 965f273892bd..046d7c15f835 100644 ---- a/clang/lib/Basic/Targets.cpp -+++ b/clang/lib/Basic/Targets.cpp -@@ -40,7 +40,6 @@ - #include "clang/Basic/Diagnostic.h" - #include "llvm/ADT/StringExtras.h" - #include "llvm/ADT/Triple.h" -- - using namespace clang; - - namespace clang { -diff --git a/clang/lib/CodeGen/CodeGenFunction.cpp b/clang/lib/CodeGen/CodeGenFunction.cpp -index 8ce488f35dd3..ca4cb0fbf018 100644 ---- a/clang/lib/CodeGen/CodeGenFunction.cpp -+++ b/clang/lib/CodeGen/CodeGenFunction.cpp -@@ -2306,11 +2306,11 @@ void CodeGenFunction::checkTargetFeatures(SourceLocation Loc, - std::string MissingFeature; - if (BuiltinID) { - SmallVector<StringRef, 1> ReqFeatures; -- const char *FeatureList = -- CGM.getContext().BuiltinInfo.getRequiredFeatures(BuiltinID); - // Return if the builtin doesn't have any required features. -- if (!FeatureList || StringRef(FeatureList) == "") -+ if (!CGM.getContext().BuiltinInfo.requiresFeatures(BuiltinID)) - return; -+ const char *FeatureList = -+ CGM.getContext().BuiltinInfo.getRequiredFeatures(BuiltinID); - StringRef(FeatureList).split(ReqFeatures, ','); - if (!hasRequiredFeatures(ReqFeatures, CGM, FD, MissingFeature)) - CGM.getDiags().Report(Loc, diag::err_builtin_needs_feature) -diff --git a/clang/lib/Frontend/CompilerInvocation.cpp b/clang/lib/Frontend/CompilerInvocation.cpp -index 73114c6d76cb..bc54d231b97a 100644 ---- a/clang/lib/Frontend/CompilerInvocation.cpp -+++ b/clang/lib/Frontend/CompilerInvocation.cpp -@@ -2344,6 +2344,8 @@ void CompilerInvocation::setLangDefaults(LangOptions &Opts, InputKind IK, - Opts.OpenCLVersion = 120; - else if (LangStd == LangStandard::lang_opencl20) - Opts.OpenCLVersion = 200; -+ else if (LangStd == LangStandard::lang_opencl30) -+ Opts.OpenCLVersion = 300; - else if (LangStd == LangStandard::lang_openclcpp) - Opts.OpenCLCPlusPlusVersion = 100; - -@@ -2544,14 +2546,15 @@ static void ParseLangArgs(LangOptions &Opts, ArgList &Args, InputKind IK, - // -cl-std only applies for OpenCL language standards. - // Override the -std option in this case. - if (const Arg *A = Args.getLastArg(OPT_cl_std_EQ)) { -- LangStandard::Kind OpenCLLangStd -- = llvm::StringSwitch<LangStandard::Kind>(A->getValue()) -- .Cases("cl", "CL", LangStandard::lang_opencl10) -- .Cases("cl1.1", "CL1.1", LangStandard::lang_opencl11) -- .Cases("cl1.2", "CL1.2", LangStandard::lang_opencl12) -- .Cases("cl2.0", "CL2.0", LangStandard::lang_opencl20) -- .Cases("clc++", "CLC++", LangStandard::lang_openclcpp) -- .Default(LangStandard::lang_unspecified); -+ LangStandard::Kind OpenCLLangStd = -+ llvm::StringSwitch<LangStandard::Kind>(A->getValue()) -+ .Cases("cl", "CL", LangStandard::lang_opencl10) -+ .Cases("cl1.1", "CL1.1", LangStandard::lang_opencl11) -+ .Cases("cl1.2", "CL1.2", LangStandard::lang_opencl12) -+ .Cases("cl2.0", "CL2.0", LangStandard::lang_opencl20) -+ .Cases("cl3.0", "CL3.0", LangStandard::lang_opencl30) -+ .Cases("clc++", "CLC++", LangStandard::lang_openclcpp) -+ .Default(LangStandard::lang_unspecified); - - if (OpenCLLangStd == LangStandard::lang_unspecified) { - Diags.Report(diag::err_drv_invalid_value) -@@ -2852,8 +2855,7 @@ static void ParseLangArgs(LangOptions &Opts, ArgList &Args, InputKind IK, - - Opts.RTTI = Opts.CPlusPlus && !Args.hasArg(OPT_fno_rtti); - Opts.RTTIData = Opts.RTTI && !Args.hasArg(OPT_fno_rtti_data); -- Opts.Blocks = Args.hasArg(OPT_fblocks) || (Opts.OpenCL -- && Opts.OpenCLVersion == 200); -+ Opts.Blocks = Args.hasArg(OPT_fblocks); - Opts.BlocksRuntimeOptional = Args.hasArg(OPT_fblocks_runtime_optional); - Opts.Coroutines = Opts.CPlusPlus20 || Args.hasArg(OPT_fcoroutines_ts); - -diff --git a/clang/lib/Frontend/InitPreprocessor.cpp b/clang/lib/Frontend/InitPreprocessor.cpp -index 6eef1e2376f6..5bb489c11909 100644 ---- a/clang/lib/Frontend/InitPreprocessor.cpp -+++ b/clang/lib/Frontend/InitPreprocessor.cpp -@@ -445,6 +445,9 @@ static void InitializeStandardPredefinedMacros(const TargetInfo &TI, - case 200: - Builder.defineMacro("__OPENCL_C_VERSION__", "200"); - break; -+ case 300: -+ Builder.defineMacro("__OPENCL_C_VERSION__", "300"); -+ break; - default: - llvm_unreachable("Unsupported OpenCL version"); - } -@@ -453,6 +456,7 @@ static void InitializeStandardPredefinedMacros(const TargetInfo &TI, - Builder.defineMacro("CL_VERSION_1_1", "110"); - Builder.defineMacro("CL_VERSION_1_2", "120"); - Builder.defineMacro("CL_VERSION_2_0", "200"); -+ Builder.defineMacro("CL_VERSION_3_0", "300"); - - if (TI.isLittleEndian()) - Builder.defineMacro("__ENDIAN_LITTLE__"); -@@ -1108,7 +1112,7 @@ static void InitializePredefinedMacros(const TargetInfo &TI, - // OpenCL definitions. - if (LangOpts.OpenCL) { - #define OPENCLEXT(Ext) \ -- if (TI.getSupportedOpenCLOpts().isSupported(#Ext, LangOpts)) \ -+ if (TI.getSupportedOpenCLOpts().isSupported(#Ext)) \ - Builder.defineMacro(#Ext); - #include "clang/Basic/OpenCLExtensions.def" - -diff --git a/clang/lib/Headers/opencl-c-base.h b/clang/lib/Headers/opencl-c-base.h -index 430e07d36f62..afa900ab24d9 100644 ---- a/clang/lib/Headers/opencl-c-base.h -+++ b/clang/lib/Headers/opencl-c-base.h -@@ -9,6 +9,59 @@ - #ifndef _OPENCL_BASE_H_ - #define _OPENCL_BASE_H_ - -+// Add predefined macros to build headers with standalone executable -+#ifndef CL_VERSION_3_0 -+ #define CL_VERSION_3_0 300 -+#endif -+ -+// Define features for 2.0 for header backward compatibility -+#ifndef __opencl_c_int64 -+ #define __opencl_c_int64 1 -+#endif -+#if __OPENCL_C_VERSION__ != CL_VERSION_3_0 -+ #ifndef __opencl_c_images -+ #define __opencl_c_images 1 -+ #endif -+#endif -+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ == CL_VERSION_2_0) -+#ifndef __opencl_c_pipes -+ #define __opencl_c_pipes 1 -+#endif -+#ifndef __opencl_c_generic_address_space -+ #define __opencl_c_generic_address_space 1 -+#endif -+#ifndef __opencl_c_work_group_collective_functions -+ #define __opencl_c_work_group_collective_functions 1 -+#endif -+#ifndef __opencl_c_atomic_order_acq_rel -+ #define __opencl_c_atomic_order_acq_rel 1 -+#endif -+#ifndef __opencl_c_atomic_order_seq_cst -+ #define __opencl_c_atomic_order_seq_cst 1 -+#endif -+#ifndef __opencl_c_atomic_scope_device -+ #define __opencl_c_atomic_scope_device 1 -+#endif -+#ifndef __opencl_c_atomic_scope_all_devices -+ #define __opencl_c_atomic_scope_all_devices 1 -+#endif -+#ifndef __opencl_c_subgroups -+ #define __opencl_c_subgroups 1 -+#endif -+#ifndef __opencl_c_3d_image_writes -+ #define __opencl_c_3d_image_writes 1 -+#endif -+#ifndef __opencl_c_device_enqueue -+ #define __opencl_c_device_enqueue 1 -+#endif -+#ifndef __opencl_c_read_write_images -+ #define __opencl_c_read_write_images 1 -+#endif -+#ifndef __opencl_c_program_scope_global_variables -+ #define __opencl_c_program_scope_global_variables 1 -+#endif -+#endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ == CL_VERSION_2_0) -+ - // built-in scalar data types: - - /** -@@ -115,7 +168,12 @@ typedef half half4 __attribute__((ext_vector_type(4))); - typedef half half8 __attribute__((ext_vector_type(8))); - typedef half half16 __attribute__((ext_vector_type(16))); - #endif --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+ -+#ifndef __opencl_c_fp64 -+ #define __opencl_c_fp64 1 -+#endif -+ - #if __OPENCL_C_VERSION__ < CL_VERSION_1_2 - #pragma OPENCL EXTENSION cl_khr_fp64 : enable - #endif -@@ -281,9 +339,17 @@ typedef uint cl_mem_fence_flags; - typedef enum memory_scope { - memory_scope_work_item = __OPENCL_MEMORY_SCOPE_WORK_ITEM, - memory_scope_work_group = __OPENCL_MEMORY_SCOPE_WORK_GROUP, -+#ifdef __opencl_c_atomic_scope_device - memory_scope_device = __OPENCL_MEMORY_SCOPE_DEVICE, -+#endif -+#ifdef __opencl_c_atomic_scope_all_devices -+ #if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+ memory_scope_all_devices = __OPENCL_MEMORY_SCOPE_ALL_SVM_DEVICES, -+ #endif //(__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - memory_scope_all_svm_devices = __OPENCL_MEMORY_SCOPE_ALL_SVM_DEVICES, --#if defined(cl_intel_subgroups) || defined(cl_khr_subgroups) -+#endif // __opencl_c_atomic_scope_all_devices -+#if defined(cl_intel_subgroups) || defined(cl_khr_subgroups) || \ -+ defined(__opencl_c_subgroups) - memory_scope_sub_group = __OPENCL_MEMORY_SCOPE_SUB_GROUP - #endif - } memory_scope; -@@ -301,13 +367,14 @@ typedef enum memory_scope { - #define ATOMIC_FLAG_INIT 0 - - // enum values aligned with what clang uses in EmitAtomicExpr() --typedef enum memory_order --{ -+typedef enum memory_order { - memory_order_relaxed = __ATOMIC_RELAXED, - memory_order_acquire = __ATOMIC_ACQUIRE, - memory_order_release = __ATOMIC_RELEASE, - memory_order_acq_rel = __ATOMIC_ACQ_REL, -+#ifdef __opencl_c_atomic_order_seq_cst - memory_order_seq_cst = __ATOMIC_SEQ_CST -+#endif //__opencl_c_atomic_order_seq_cst - } memory_order; - - #endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -diff --git a/clang/lib/Headers/opencl-c.h b/clang/lib/Headers/opencl-c.h -index 66e18bdd47bb..67d900eb1c3d 100644 ---- a/clang/lib/Headers/opencl-c.h -+++ b/clang/lib/Headers/opencl-c.h -@@ -35,7 +35,6 @@ - #define __purefn __attribute__((pure)) - #define __cnfn __attribute__((const)) - -- - // OpenCL v1.1/1.2/2.0 s6.2.3 - Explicit conversions - - char __ovld __cnfn convert_char_rte(char); -@@ -4632,7 +4631,7 @@ float16 __ovld __cnfn convert_float16(float16); - - // Conversions with double data type parameters or return value. - --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - char __ovld __cnfn convert_char(double); - char __ovld __cnfn convert_char_rte(double); - char __ovld __cnfn convert_char_rtn(double); -@@ -5452,7 +5451,7 @@ double16 __ovld __cnfn convert_double16_rtz(uchar16); - double16 __ovld __cnfn convert_double16_rtz(uint16); - double16 __ovld __cnfn convert_double16_rtz(ulong16); - double16 __ovld __cnfn convert_double16_rtz(ushort16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - #ifdef cl_khr_fp16 - // Convert half types to non-double types. -@@ -6270,7 +6269,7 @@ half16 __ovld __cnfn convert_half16_rtz(float16); - half16 __ovld __cnfn convert_half16_rtz(half16); - - // Convert half types to double types. --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn convert_double(half); - double __ovld __cnfn convert_double_rte(half); - double __ovld __cnfn convert_double_rtp(half); -@@ -6333,7 +6332,7 @@ half16 __ovld __cnfn convert_half16_rte(double16); - half16 __ovld __cnfn convert_half16_rtp(double16); - half16 __ovld __cnfn convert_half16_rtn(double16); - half16 __ovld __cnfn convert_half16_rtz(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - #endif // cl_khr_fp16 - -@@ -6404,14 +6403,14 @@ half16 __ovld __cnfn convert_half16_rtz(double16); - #define as_float8(x) __builtin_astype((x), float8) - #define as_float16(x) __builtin_astype((x), float16) - --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #define as_double(x) __builtin_astype((x), double) - #define as_double2(x) __builtin_astype((x), double2) - #define as_double3(x) __builtin_astype((x), double3) - #define as_double4(x) __builtin_astype((x), double4) - #define as_double8(x) __builtin_astype((x), double8) - #define as_double16(x) __builtin_astype((x), double16) --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - #ifdef cl_khr_fp16 - #define as_half(x) __builtin_astype((x), half) -@@ -6534,14 +6533,14 @@ float3 __ovld __cnfn acos(float3); - float4 __ovld __cnfn acos(float4); - float8 __ovld __cnfn acos(float8); - float16 __ovld __cnfn acos(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn acos(double); - double2 __ovld __cnfn acos(double2); - double3 __ovld __cnfn acos(double3); - double4 __ovld __cnfn acos(double4); - double8 __ovld __cnfn acos(double8); - double16 __ovld __cnfn acos(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn acos(half); - half2 __ovld __cnfn acos(half2); -@@ -6560,14 +6559,14 @@ float3 __ovld __cnfn acosh(float3); - float4 __ovld __cnfn acosh(float4); - float8 __ovld __cnfn acosh(float8); - float16 __ovld __cnfn acosh(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn acosh(double); - double2 __ovld __cnfn acosh(double2); - double3 __ovld __cnfn acosh(double3); - double4 __ovld __cnfn acosh(double4); - double8 __ovld __cnfn acosh(double8); - double16 __ovld __cnfn acosh(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn acosh(half); - half2 __ovld __cnfn acosh(half2); -@@ -6586,14 +6585,14 @@ float3 __ovld __cnfn acospi(float3 x); - float4 __ovld __cnfn acospi(float4 x); - float8 __ovld __cnfn acospi(float8 x); - float16 __ovld __cnfn acospi(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn acospi(double x); - double2 __ovld __cnfn acospi(double2 x); - double3 __ovld __cnfn acospi(double3 x); - double4 __ovld __cnfn acospi(double4 x); - double8 __ovld __cnfn acospi(double8 x); - double16 __ovld __cnfn acospi(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn acospi(half x); - half2 __ovld __cnfn acospi(half2 x); -@@ -6612,14 +6611,14 @@ float3 __ovld __cnfn asin(float3); - float4 __ovld __cnfn asin(float4); - float8 __ovld __cnfn asin(float8); - float16 __ovld __cnfn asin(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn asin(double); - double2 __ovld __cnfn asin(double2); - double3 __ovld __cnfn asin(double3); - double4 __ovld __cnfn asin(double4); - double8 __ovld __cnfn asin(double8); - double16 __ovld __cnfn asin(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn asin(half); - half2 __ovld __cnfn asin(half2); -@@ -6638,14 +6637,14 @@ float3 __ovld __cnfn asinh(float3); - float4 __ovld __cnfn asinh(float4); - float8 __ovld __cnfn asinh(float8); - float16 __ovld __cnfn asinh(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn asinh(double); - double2 __ovld __cnfn asinh(double2); - double3 __ovld __cnfn asinh(double3); - double4 __ovld __cnfn asinh(double4); - double8 __ovld __cnfn asinh(double8); - double16 __ovld __cnfn asinh(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn asinh(half); - half2 __ovld __cnfn asinh(half2); -@@ -6664,14 +6663,14 @@ float3 __ovld __cnfn asinpi(float3 x); - float4 __ovld __cnfn asinpi(float4 x); - float8 __ovld __cnfn asinpi(float8 x); - float16 __ovld __cnfn asinpi(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn asinpi(double x); - double2 __ovld __cnfn asinpi(double2 x); - double3 __ovld __cnfn asinpi(double3 x); - double4 __ovld __cnfn asinpi(double4 x); - double8 __ovld __cnfn asinpi(double8 x); - double16 __ovld __cnfn asinpi(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn asinpi(half x); - half2 __ovld __cnfn asinpi(half2 x); -@@ -6690,14 +6689,14 @@ float3 __ovld __cnfn atan(float3 y_over_x); - float4 __ovld __cnfn atan(float4 y_over_x); - float8 __ovld __cnfn atan(float8 y_over_x); - float16 __ovld __cnfn atan(float16 y_over_x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn atan(double y_over_x); - double2 __ovld __cnfn atan(double2 y_over_x); - double3 __ovld __cnfn atan(double3 y_over_x); - double4 __ovld __cnfn atan(double4 y_over_x); - double8 __ovld __cnfn atan(double8 y_over_x); - double16 __ovld __cnfn atan(double16 y_over_x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn atan(half y_over_x); - half2 __ovld __cnfn atan(half2 y_over_x); -@@ -6716,14 +6715,14 @@ float3 __ovld __cnfn atan2(float3 y, float3 x); - float4 __ovld __cnfn atan2(float4 y, float4 x); - float8 __ovld __cnfn atan2(float8 y, float8 x); - float16 __ovld __cnfn atan2(float16 y, float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn atan2(double y, double x); - double2 __ovld __cnfn atan2(double2 y, double2 x); - double3 __ovld __cnfn atan2(double3 y, double3 x); - double4 __ovld __cnfn atan2(double4 y, double4 x); - double8 __ovld __cnfn atan2(double8 y, double8 x); - double16 __ovld __cnfn atan2(double16 y, double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn atan2(half y, half x); - half2 __ovld __cnfn atan2(half2 y, half2 x); -@@ -6742,14 +6741,14 @@ float3 __ovld __cnfn atanh(float3); - float4 __ovld __cnfn atanh(float4); - float8 __ovld __cnfn atanh(float8); - float16 __ovld __cnfn atanh(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn atanh(double); - double2 __ovld __cnfn atanh(double2); - double3 __ovld __cnfn atanh(double3); - double4 __ovld __cnfn atanh(double4); - double8 __ovld __cnfn atanh(double8); - double16 __ovld __cnfn atanh(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn atanh(half); - half2 __ovld __cnfn atanh(half2); -@@ -6768,14 +6767,14 @@ float3 __ovld __cnfn atanpi(float3 x); - float4 __ovld __cnfn atanpi(float4 x); - float8 __ovld __cnfn atanpi(float8 x); - float16 __ovld __cnfn atanpi(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn atanpi(double x); - double2 __ovld __cnfn atanpi(double2 x); - double3 __ovld __cnfn atanpi(double3 x); - double4 __ovld __cnfn atanpi(double4 x); - double8 __ovld __cnfn atanpi(double8 x); - double16 __ovld __cnfn atanpi(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn atanpi(half x); - half2 __ovld __cnfn atanpi(half2 x); -@@ -6794,14 +6793,14 @@ float3 __ovld __cnfn atan2pi(float3 y, float3 x); - float4 __ovld __cnfn atan2pi(float4 y, float4 x); - float8 __ovld __cnfn atan2pi(float8 y, float8 x); - float16 __ovld __cnfn atan2pi(float16 y, float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn atan2pi(double y, double x); - double2 __ovld __cnfn atan2pi(double2 y, double2 x); - double3 __ovld __cnfn atan2pi(double3 y, double3 x); - double4 __ovld __cnfn atan2pi(double4 y, double4 x); - double8 __ovld __cnfn atan2pi(double8 y, double8 x); - double16 __ovld __cnfn atan2pi(double16 y, double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn atan2pi(half y, half x); - half2 __ovld __cnfn atan2pi(half2 y, half2 x); -@@ -6820,14 +6819,14 @@ float3 __ovld __cnfn cbrt(float3); - float4 __ovld __cnfn cbrt(float4); - float8 __ovld __cnfn cbrt(float8); - float16 __ovld __cnfn cbrt(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn cbrt(double); - double2 __ovld __cnfn cbrt(double2); - double3 __ovld __cnfn cbrt(double3); - double4 __ovld __cnfn cbrt(double4); - double8 __ovld __cnfn cbrt(double8); - double16 __ovld __cnfn cbrt(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn cbrt(half); - half2 __ovld __cnfn cbrt(half2); -@@ -6847,14 +6846,14 @@ float3 __ovld __cnfn ceil(float3); - float4 __ovld __cnfn ceil(float4); - float8 __ovld __cnfn ceil(float8); - float16 __ovld __cnfn ceil(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn ceil(double); - double2 __ovld __cnfn ceil(double2); - double3 __ovld __cnfn ceil(double3); - double4 __ovld __cnfn ceil(double4); - double8 __ovld __cnfn ceil(double8); - double16 __ovld __cnfn ceil(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn ceil(half); - half2 __ovld __cnfn ceil(half2); -@@ -6873,14 +6872,14 @@ float3 __ovld __cnfn copysign(float3 x, float3 y); - float4 __ovld __cnfn copysign(float4 x, float4 y); - float8 __ovld __cnfn copysign(float8 x, float8 y); - float16 __ovld __cnfn copysign(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn copysign(double x, double y); - double2 __ovld __cnfn copysign(double2 x, double2 y); - double3 __ovld __cnfn copysign(double3 x, double3 y); - double4 __ovld __cnfn copysign(double4 x, double4 y); - double8 __ovld __cnfn copysign(double8 x, double8 y); - double16 __ovld __cnfn copysign(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn copysign(half x, half y); - half2 __ovld __cnfn copysign(half2 x, half2 y); -@@ -6899,14 +6898,14 @@ float3 __ovld __cnfn cos(float3); - float4 __ovld __cnfn cos(float4); - float8 __ovld __cnfn cos(float8); - float16 __ovld __cnfn cos(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn cos(double); - double2 __ovld __cnfn cos(double2); - double3 __ovld __cnfn cos(double3); - double4 __ovld __cnfn cos(double4); - double8 __ovld __cnfn cos(double8); - double16 __ovld __cnfn cos(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn cos(half); - half2 __ovld __cnfn cos(half2); -@@ -6925,14 +6924,14 @@ float3 __ovld __cnfn cosh(float3); - float4 __ovld __cnfn cosh(float4); - float8 __ovld __cnfn cosh(float8); - float16 __ovld __cnfn cosh(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn cosh(double); - double2 __ovld __cnfn cosh(double2); - double3 __ovld __cnfn cosh(double3); - double4 __ovld __cnfn cosh(double4); - double8 __ovld __cnfn cosh(double8); - double16 __ovld __cnfn cosh(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn cosh(half); - half2 __ovld __cnfn cosh(half2); -@@ -6951,14 +6950,14 @@ float3 __ovld __cnfn cospi(float3 x); - float4 __ovld __cnfn cospi(float4 x); - float8 __ovld __cnfn cospi(float8 x); - float16 __ovld __cnfn cospi(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn cospi(double x); - double2 __ovld __cnfn cospi(double2 x); - double3 __ovld __cnfn cospi(double3 x); - double4 __ovld __cnfn cospi(double4 x); - double8 __ovld __cnfn cospi(double8 x); - double16 __ovld __cnfn cospi(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn cospi(half x); - half2 __ovld __cnfn cospi(half2 x); -@@ -6977,14 +6976,14 @@ float3 __ovld __cnfn erfc(float3); - float4 __ovld __cnfn erfc(float4); - float8 __ovld __cnfn erfc(float8); - float16 __ovld __cnfn erfc(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn erfc(double); - double2 __ovld __cnfn erfc(double2); - double3 __ovld __cnfn erfc(double3); - double4 __ovld __cnfn erfc(double4); - double8 __ovld __cnfn erfc(double8); - double16 __ovld __cnfn erfc(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn erfc(half); - half2 __ovld __cnfn erfc(half2); -@@ -7004,14 +7003,14 @@ float3 __ovld __cnfn erf(float3); - float4 __ovld __cnfn erf(float4); - float8 __ovld __cnfn erf(float8); - float16 __ovld __cnfn erf(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn erf(double); - double2 __ovld __cnfn erf(double2); - double3 __ovld __cnfn erf(double3); - double4 __ovld __cnfn erf(double4); - double8 __ovld __cnfn erf(double8); - double16 __ovld __cnfn erf(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn erf(half); - half2 __ovld __cnfn erf(half2); -@@ -7030,14 +7029,14 @@ float3 __ovld __cnfn exp(float3 x); - float4 __ovld __cnfn exp(float4 x); - float8 __ovld __cnfn exp(float8 x); - float16 __ovld __cnfn exp(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn exp(double x); - double2 __ovld __cnfn exp(double2 x); - double3 __ovld __cnfn exp(double3 x); - double4 __ovld __cnfn exp(double4 x); - double8 __ovld __cnfn exp(double8 x); - double16 __ovld __cnfn exp(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn exp(half x); - half2 __ovld __cnfn exp(half2 x); -@@ -7056,14 +7055,14 @@ float3 __ovld __cnfn exp2(float3); - float4 __ovld __cnfn exp2(float4); - float8 __ovld __cnfn exp2(float8); - float16 __ovld __cnfn exp2(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn exp2(double); - double2 __ovld __cnfn exp2(double2); - double3 __ovld __cnfn exp2(double3); - double4 __ovld __cnfn exp2(double4); - double8 __ovld __cnfn exp2(double8); - double16 __ovld __cnfn exp2(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn exp2(half); - half2 __ovld __cnfn exp2(half2); -@@ -7082,14 +7081,14 @@ float3 __ovld __cnfn exp10(float3); - float4 __ovld __cnfn exp10(float4); - float8 __ovld __cnfn exp10(float8); - float16 __ovld __cnfn exp10(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn exp10(double); - double2 __ovld __cnfn exp10(double2); - double3 __ovld __cnfn exp10(double3); - double4 __ovld __cnfn exp10(double4); - double8 __ovld __cnfn exp10(double8); - double16 __ovld __cnfn exp10(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn exp10(half); - half2 __ovld __cnfn exp10(half2); -@@ -7108,14 +7107,14 @@ float3 __ovld __cnfn expm1(float3 x); - float4 __ovld __cnfn expm1(float4 x); - float8 __ovld __cnfn expm1(float8 x); - float16 __ovld __cnfn expm1(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn expm1(double x); - double2 __ovld __cnfn expm1(double2 x); - double3 __ovld __cnfn expm1(double3 x); - double4 __ovld __cnfn expm1(double4 x); - double8 __ovld __cnfn expm1(double8 x); - double16 __ovld __cnfn expm1(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn expm1(half x); - half2 __ovld __cnfn expm1(half2 x); -@@ -7134,14 +7133,14 @@ float3 __ovld __cnfn fabs(float3); - float4 __ovld __cnfn fabs(float4); - float8 __ovld __cnfn fabs(float8); - float16 __ovld __cnfn fabs(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn fabs(double); - double2 __ovld __cnfn fabs(double2); - double3 __ovld __cnfn fabs(double3); - double4 __ovld __cnfn fabs(double4); - double8 __ovld __cnfn fabs(double8); - double16 __ovld __cnfn fabs(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn fabs(half); - half2 __ovld __cnfn fabs(half2); -@@ -7160,14 +7159,14 @@ float3 __ovld __cnfn fdim(float3 x, float3 y); - float4 __ovld __cnfn fdim(float4 x, float4 y); - float8 __ovld __cnfn fdim(float8 x, float8 y); - float16 __ovld __cnfn fdim(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn fdim(double x, double y); - double2 __ovld __cnfn fdim(double2 x, double2 y); - double3 __ovld __cnfn fdim(double3 x, double3 y); - double4 __ovld __cnfn fdim(double4 x, double4 y); - double8 __ovld __cnfn fdim(double8 x, double8 y); - double16 __ovld __cnfn fdim(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn fdim(half x, half y); - half2 __ovld __cnfn fdim(half2 x, half2 y); -@@ -7187,14 +7186,14 @@ float3 __ovld __cnfn floor(float3); - float4 __ovld __cnfn floor(float4); - float8 __ovld __cnfn floor(float8); - float16 __ovld __cnfn floor(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn floor(double); - double2 __ovld __cnfn floor(double2); - double3 __ovld __cnfn floor(double3); - double4 __ovld __cnfn floor(double4); - double8 __ovld __cnfn floor(double8); - double16 __ovld __cnfn floor(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn floor(half); - half2 __ovld __cnfn floor(half2); -@@ -7217,14 +7216,14 @@ float3 __ovld __cnfn fma(float3 a, float3 b, float3 c); - float4 __ovld __cnfn fma(float4 a, float4 b, float4 c); - float8 __ovld __cnfn fma(float8 a, float8 b, float8 c); - float16 __ovld __cnfn fma(float16 a, float16 b, float16 c); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn fma(double a, double b, double c); - double2 __ovld __cnfn fma(double2 a, double2 b, double2 c); - double3 __ovld __cnfn fma(double3 a, double3 b, double3 c); - double4 __ovld __cnfn fma(double4 a, double4 b, double4 c); - double8 __ovld __cnfn fma(double8 a, double8 b, double8 c); - double16 __ovld __cnfn fma(double16 a, double16 b, double16 c); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn fma(half a, half b, half c); - half2 __ovld __cnfn fma(half2 a, half2 b, half2 c); -@@ -7251,7 +7250,7 @@ float3 __ovld __cnfn fmax(float3 x, float y); - float4 __ovld __cnfn fmax(float4 x, float y); - float8 __ovld __cnfn fmax(float8 x, float y); - float16 __ovld __cnfn fmax(float16 x, float y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn fmax(double x, double y); - double2 __ovld __cnfn fmax(double2 x, double2 y); - double3 __ovld __cnfn fmax(double3 x, double3 y); -@@ -7263,7 +7262,7 @@ double3 __ovld __cnfn fmax(double3 x, double y); - double4 __ovld __cnfn fmax(double4 x, double y); - double8 __ovld __cnfn fmax(double8 x, double y); - double16 __ovld __cnfn fmax(double16 x, double y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn fmax(half x, half y); - half2 __ovld __cnfn fmax(half2 x, half2 y); -@@ -7295,7 +7294,7 @@ float3 __ovld __cnfn fmin(float3 x, float y); - float4 __ovld __cnfn fmin(float4 x, float y); - float8 __ovld __cnfn fmin(float8 x, float y); - float16 __ovld __cnfn fmin(float16 x, float y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn fmin(double x, double y); - double2 __ovld __cnfn fmin(double2 x, double2 y); - double3 __ovld __cnfn fmin(double3 x, double3 y); -@@ -7307,7 +7306,7 @@ double3 __ovld __cnfn fmin(double3 x, double y); - double4 __ovld __cnfn fmin(double4 x, double y); - double8 __ovld __cnfn fmin(double8 x, double y); - double16 __ovld __cnfn fmin(double16 x, double y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn fmin(half x, half y); - half2 __ovld __cnfn fmin(half2 x, half2 y); -@@ -7331,14 +7330,14 @@ float3 __ovld __cnfn fmod(float3 x, float3 y); - float4 __ovld __cnfn fmod(float4 x, float4 y); - float8 __ovld __cnfn fmod(float8 x, float8 y); - float16 __ovld __cnfn fmod(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn fmod(double x, double y); - double2 __ovld __cnfn fmod(double2 x, double2 y); - double3 __ovld __cnfn fmod(double3 x, double3 y); - double4 __ovld __cnfn fmod(double4 x, double4 y); - double8 __ovld __cnfn fmod(double8 x, double8 y); - double16 __ovld __cnfn fmod(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn fmod(half x, half y); - half2 __ovld __cnfn fmod(half2 x, half2 y); -@@ -7352,21 +7351,21 @@ half16 __ovld __cnfn fmod(half16 x, half16 y); - * Returns fmin(x - floor (x), 0x1.fffffep-1f ). - * floor(x) is returned in iptr. - */ --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - float __ovld fract(float x, float *iptr); - float2 __ovld fract(float2 x, float2 *iptr); - float3 __ovld fract(float3 x, float3 *iptr); - float4 __ovld fract(float4 x, float4 *iptr); - float8 __ovld fract(float8 x, float8 *iptr); - float16 __ovld fract(float16 x, float16 *iptr); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld fract(double x, double *iptr); - double2 __ovld fract(double2 x, double2 *iptr); - double3 __ovld fract(double3 x, double3 *iptr); - double4 __ovld fract(double4 x, double4 *iptr); - double8 __ovld fract(double8 x, double8 *iptr); - double16 __ovld fract(double16 x, double16 *iptr); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld fract(half x, half *iptr); - half2 __ovld fract(half2 x, half2 *iptr); -@@ -7375,7 +7374,9 @@ half4 __ovld fract(half4 x, half4 *iptr); - half8 __ovld fract(half8 x, half8 *iptr); - half16 __ovld fract(half16 x, half16 *iptr); - #endif //cl_khr_fp16 --#else -+#endif //__opencl_c_generic_address_space -+ -+#if !defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != CL_VERSION_2_0) - float __ovld fract(float x, __global float *iptr); - float2 __ovld fract(float2 x, __global float2 *iptr); - float3 __ovld fract(float3 x, __global float3 *iptr); -@@ -7394,7 +7395,7 @@ float3 __ovld fract(float3 x, __private float3 *iptr); - float4 __ovld fract(float4 x, __private float4 *iptr); - float8 __ovld fract(float8 x, __private float8 *iptr); - float16 __ovld fract(float16 x, __private float16 *iptr); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld fract(double x, __global double *iptr); - double2 __ovld fract(double2 x, __global double2 *iptr); - double3 __ovld fract(double3 x, __global double3 *iptr); -@@ -7413,7 +7414,7 @@ double3 __ovld fract(double3 x, __private double3 *iptr); - double4 __ovld fract(double4 x, __private double4 *iptr); - double8 __ovld fract(double8 x, __private double8 *iptr); - double16 __ovld fract(double16 x, __private double16 *iptr); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld fract(half x, __global half *iptr); - half2 __ovld fract(half2 x, __global half2 *iptr); -@@ -7434,29 +7435,29 @@ half4 __ovld fract(half4 x, __private half4 *iptr); - half8 __ovld fract(half8 x, __private half8 *iptr); - half16 __ovld fract(half16 x, __private half16 *iptr); - #endif //cl_khr_fp16 --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -- -+#endif //! defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != -+ //! CL_VERSION_2_0) - /** - * Extract mantissa and exponent from x. For each - * component the mantissa returned is a float with - * magnitude in the interval [1/2, 1) or 0. Each - * component of x equals mantissa returned * 2^exp. - */ --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - float __ovld frexp(float x, int *exp); - float2 __ovld frexp(float2 x, int2 *exp); - float3 __ovld frexp(float3 x, int3 *exp); - float4 __ovld frexp(float4 x, int4 *exp); - float8 __ovld frexp(float8 x, int8 *exp); - float16 __ovld frexp(float16 x, int16 *exp); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld frexp(double x, int *exp); - double2 __ovld frexp(double2 x, int2 *exp); - double3 __ovld frexp(double3 x, int3 *exp); - double4 __ovld frexp(double4 x, int4 *exp); - double8 __ovld frexp(double8 x, int8 *exp); - double16 __ovld frexp(double16 x, int16 *exp); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld frexp(half x, int *exp); - half2 __ovld frexp(half2 x, int2 *exp); -@@ -7465,7 +7466,9 @@ half4 __ovld frexp(half4 x, int4 *exp); - half8 __ovld frexp(half8 x, int8 *exp); - half16 __ovld frexp(half16 x, int16 *exp); - #endif //cl_khr_fp16 --#else -+#endif //__opencl_c_generic_address_space -+ -+#if !defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != CL_VERSION_2_0) - float __ovld frexp(float x, __global int *exp); - float2 __ovld frexp(float2 x, __global int2 *exp); - float3 __ovld frexp(float3 x, __global int3 *exp); -@@ -7484,7 +7487,7 @@ float3 __ovld frexp(float3 x, __private int3 *exp); - float4 __ovld frexp(float4 x, __private int4 *exp); - float8 __ovld frexp(float8 x, __private int8 *exp); - float16 __ovld frexp(float16 x, __private int16 *exp); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld frexp(double x, __global int *exp); - double2 __ovld frexp(double2 x, __global int2 *exp); - double3 __ovld frexp(double3 x, __global int3 *exp); -@@ -7503,7 +7506,7 @@ double3 __ovld frexp(double3 x, __private int3 *exp); - double4 __ovld frexp(double4 x, __private int4 *exp); - double8 __ovld frexp(double8 x, __private int8 *exp); - double16 __ovld frexp(double16 x, __private int16 *exp); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld frexp(half x, __global int *exp); - half2 __ovld frexp(half2 x, __global int2 *exp); -@@ -7524,7 +7527,8 @@ half4 __ovld frexp(half4 x, __private int4 *exp); - half8 __ovld frexp(half8 x, __private int8 *exp); - half16 __ovld frexp(half16 x, __private int16 *exp); - #endif //cl_khr_fp16 --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#endif //! defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != -+ //! CL_VERSION_2_0) - - /** - * Compute the value of the square root of x^2 + y^2 -@@ -7536,14 +7540,14 @@ float3 __ovld __cnfn hypot(float3 x, float3 y); - float4 __ovld __cnfn hypot(float4 x, float4 y); - float8 __ovld __cnfn hypot(float8 x, float8 y); - float16 __ovld __cnfn hypot(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn hypot(double x, double y); - double2 __ovld __cnfn hypot(double2 x, double2 y); - double3 __ovld __cnfn hypot(double3 x, double3 y); - double4 __ovld __cnfn hypot(double4 x, double4 y); - double8 __ovld __cnfn hypot(double8 x, double8 y); - double16 __ovld __cnfn hypot(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn hypot(half x, half y); - half2 __ovld __cnfn hypot(half2 x, half2 y); -@@ -7562,14 +7566,14 @@ int3 __ovld __cnfn ilogb(float3 x); - int4 __ovld __cnfn ilogb(float4 x); - int8 __ovld __cnfn ilogb(float8 x); - int16 __ovld __cnfn ilogb(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn ilogb(double x); - int2 __ovld __cnfn ilogb(double2 x); - int3 __ovld __cnfn ilogb(double3 x); - int4 __ovld __cnfn ilogb(double4 x); - int8 __ovld __cnfn ilogb(double8 x); - int16 __ovld __cnfn ilogb(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn ilogb(half x); - int2 __ovld __cnfn ilogb(half2 x); -@@ -7593,7 +7597,7 @@ float3 __ovld __cnfn ldexp(float3 x, int n); - float4 __ovld __cnfn ldexp(float4 x, int n); - float8 __ovld __cnfn ldexp(float8 x, int n); - float16 __ovld __cnfn ldexp(float16 x, int n); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn ldexp(double x, int n); - double2 __ovld __cnfn ldexp(double2 x, int2 n); - double3 __ovld __cnfn ldexp(double3 x, int3 n); -@@ -7605,7 +7609,7 @@ double3 __ovld __cnfn ldexp(double3 x, int n); - double4 __ovld __cnfn ldexp(double4 x, int n); - double8 __ovld __cnfn ldexp(double8 x, int n); - double16 __ovld __cnfn ldexp(double16 x, int n); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn ldexp(half x, int n); - half2 __ovld __cnfn ldexp(half2 x, int2 n); -@@ -7632,14 +7636,14 @@ float3 __ovld __cnfn lgamma(float3 x); - float4 __ovld __cnfn lgamma(float4 x); - float8 __ovld __cnfn lgamma(float8 x); - float16 __ovld __cnfn lgamma(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn lgamma(double x); - double2 __ovld __cnfn lgamma(double2 x); - double3 __ovld __cnfn lgamma(double3 x); - double4 __ovld __cnfn lgamma(double4 x); - double8 __ovld __cnfn lgamma(double8 x); - double16 __ovld __cnfn lgamma(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn lgamma(half x); - half2 __ovld __cnfn lgamma(half2 x); -@@ -7649,21 +7653,21 @@ half8 __ovld __cnfn lgamma(half8 x); - half16 __ovld __cnfn lgamma(half16 x); - #endif //cl_khr_fp16 - --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - float __ovld lgamma_r(float x, int *signp); - float2 __ovld lgamma_r(float2 x, int2 *signp); - float3 __ovld lgamma_r(float3 x, int3 *signp); - float4 __ovld lgamma_r(float4 x, int4 *signp); - float8 __ovld lgamma_r(float8 x, int8 *signp); - float16 __ovld lgamma_r(float16 x, int16 *signp); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld lgamma_r(double x, int *signp); - double2 __ovld lgamma_r(double2 x, int2 *signp); - double3 __ovld lgamma_r(double3 x, int3 *signp); - double4 __ovld lgamma_r(double4 x, int4 *signp); - double8 __ovld lgamma_r(double8 x, int8 *signp); - double16 __ovld lgamma_r(double16 x, int16 *signp); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld lgamma_r(half x, int *signp); - half2 __ovld lgamma_r(half2 x, int2 *signp); -@@ -7672,7 +7676,9 @@ half4 __ovld lgamma_r(half4 x, int4 *signp); - half8 __ovld lgamma_r(half8 x, int8 *signp); - half16 __ovld lgamma_r(half16 x, int16 *signp); - #endif //cl_khr_fp16 --#else -+#endif //__opencl_c_generic_address_space -+ -+#if !defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != CL_VERSION_2_0) - float __ovld lgamma_r(float x, __global int *signp); - float2 __ovld lgamma_r(float2 x, __global int2 *signp); - float3 __ovld lgamma_r(float3 x, __global int3 *signp); -@@ -7691,7 +7697,7 @@ float3 __ovld lgamma_r(float3 x, __private int3 *signp); - float4 __ovld lgamma_r(float4 x, __private int4 *signp); - float8 __ovld lgamma_r(float8 x, __private int8 *signp); - float16 __ovld lgamma_r(float16 x, __private int16 *signp); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld lgamma_r(double x, __global int *signp); - double2 __ovld lgamma_r(double2 x, __global int2 *signp); - double3 __ovld lgamma_r(double3 x, __global int3 *signp); -@@ -7710,7 +7716,7 @@ double3 __ovld lgamma_r(double3 x, __private int3 *signp); - double4 __ovld lgamma_r(double4 x, __private int4 *signp); - double8 __ovld lgamma_r(double8 x, __private int8 *signp); - double16 __ovld lgamma_r(double16 x, __private int16 *signp); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld lgamma_r(half x, __global int *signp); - half2 __ovld lgamma_r(half2 x, __global int2 *signp); -@@ -7731,8 +7737,8 @@ half4 __ovld lgamma_r(half4 x, __private int4 *signp); - half8 __ovld lgamma_r(half8 x, __private int8 *signp); - half16 __ovld lgamma_r(half16 x, __private int16 *signp); - #endif //cl_khr_fp16 --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -- -+#endif //! defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != -+ //! CL_VERSION_2_0) - /** - * Compute natural logarithm. - */ -@@ -7742,14 +7748,14 @@ float3 __ovld __cnfn log(float3); - float4 __ovld __cnfn log(float4); - float8 __ovld __cnfn log(float8); - float16 __ovld __cnfn log(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn log(double); - double2 __ovld __cnfn log(double2); - double3 __ovld __cnfn log(double3); - double4 __ovld __cnfn log(double4); - double8 __ovld __cnfn log(double8); - double16 __ovld __cnfn log(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn log(half); - half2 __ovld __cnfn log(half2); -@@ -7760,7 +7766,7 @@ half16 __ovld __cnfn log(half16); - #endif //cl_khr_fp16 - - /** -- * Compute a base 2 logarithm. -+ * Compute a base 2 logarithm - */ - float __ovld __cnfn log2(float); - float2 __ovld __cnfn log2(float2); -@@ -7768,14 +7774,14 @@ float3 __ovld __cnfn log2(float3); - float4 __ovld __cnfn log2(float4); - float8 __ovld __cnfn log2(float8); - float16 __ovld __cnfn log2(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn log2(double); - double2 __ovld __cnfn log2(double2); - double3 __ovld __cnfn log2(double3); - double4 __ovld __cnfn log2(double4); - double8 __ovld __cnfn log2(double8); - double16 __ovld __cnfn log2(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn log2(half); - half2 __ovld __cnfn log2(half2); -@@ -7794,14 +7800,14 @@ float3 __ovld __cnfn log10(float3); - float4 __ovld __cnfn log10(float4); - float8 __ovld __cnfn log10(float8); - float16 __ovld __cnfn log10(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn log10(double); - double2 __ovld __cnfn log10(double2); - double3 __ovld __cnfn log10(double3); - double4 __ovld __cnfn log10(double4); - double8 __ovld __cnfn log10(double8); - double16 __ovld __cnfn log10(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn log10(half); - half2 __ovld __cnfn log10(half2); -@@ -7820,14 +7826,14 @@ float3 __ovld __cnfn log1p(float3 x); - float4 __ovld __cnfn log1p(float4 x); - float8 __ovld __cnfn log1p(float8 x); - float16 __ovld __cnfn log1p(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn log1p(double x); - double2 __ovld __cnfn log1p(double2 x); - double3 __ovld __cnfn log1p(double3 x); - double4 __ovld __cnfn log1p(double4 x); - double8 __ovld __cnfn log1p(double8 x); - double16 __ovld __cnfn log1p(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn log1p(half x); - half2 __ovld __cnfn log1p(half2 x); -@@ -7847,14 +7853,14 @@ float3 __ovld __cnfn logb(float3 x); - float4 __ovld __cnfn logb(float4 x); - float8 __ovld __cnfn logb(float8 x); - float16 __ovld __cnfn logb(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn logb(double x); - double2 __ovld __cnfn logb(double2 x); - double3 __ovld __cnfn logb(double3 x); - double4 __ovld __cnfn logb(double4 x); - double8 __ovld __cnfn logb(double8 x); - double16 __ovld __cnfn logb(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn logb(half x); - half2 __ovld __cnfn logb(half2 x); -@@ -7877,14 +7883,14 @@ float3 __ovld __cnfn mad(float3 a, float3 b, float3 c); - float4 __ovld __cnfn mad(float4 a, float4 b, float4 c); - float8 __ovld __cnfn mad(float8 a, float8 b, float8 c); - float16 __ovld __cnfn mad(float16 a, float16 b, float16 c); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn mad(double a, double b, double c); - double2 __ovld __cnfn mad(double2 a, double2 b, double2 c); - double3 __ovld __cnfn mad(double3 a, double3 b, double3 c); - double4 __ovld __cnfn mad(double4 a, double4 b, double4 c); - double8 __ovld __cnfn mad(double8 a, double8 b, double8 c); - double16 __ovld __cnfn mad(double16 a, double16 b, double16 c); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn mad(half a, half b, half c); - half2 __ovld __cnfn mad(half2 a, half2 b, half2 c); -@@ -7904,14 +7910,14 @@ float3 __ovld __cnfn maxmag(float3 x, float3 y); - float4 __ovld __cnfn maxmag(float4 x, float4 y); - float8 __ovld __cnfn maxmag(float8 x, float8 y); - float16 __ovld __cnfn maxmag(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn maxmag(double x, double y); - double2 __ovld __cnfn maxmag(double2 x, double2 y); - double3 __ovld __cnfn maxmag(double3 x, double3 y); - double4 __ovld __cnfn maxmag(double4 x, double4 y); - double8 __ovld __cnfn maxmag(double8 x, double8 y); - double16 __ovld __cnfn maxmag(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn maxmag(half x, half y); - half2 __ovld __cnfn maxmag(half2 x, half2 y); -@@ -7931,14 +7937,14 @@ float3 __ovld __cnfn minmag(float3 x, float3 y); - float4 __ovld __cnfn minmag(float4 x, float4 y); - float8 __ovld __cnfn minmag(float8 x, float8 y); - float16 __ovld __cnfn minmag(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn minmag(double x, double y); - double2 __ovld __cnfn minmag(double2 x, double2 y); - double3 __ovld __cnfn minmag(double3 x, double3 y); - double4 __ovld __cnfn minmag(double4 x, double4 y); - double8 __ovld __cnfn minmag(double8 x, double8 y); - double16 __ovld __cnfn minmag(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn minmag(half x, half y); - half2 __ovld __cnfn minmag(half2 x, half2 y); -@@ -7955,21 +7961,21 @@ half16 __ovld __cnfn minmag(half16 x, half16 y); - * the argument. It stores the integral part in the object - * pointed to by iptr. - */ --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - float __ovld modf(float x, float *iptr); - float2 __ovld modf(float2 x, float2 *iptr); - float3 __ovld modf(float3 x, float3 *iptr); - float4 __ovld modf(float4 x, float4 *iptr); - float8 __ovld modf(float8 x, float8 *iptr); - float16 __ovld modf(float16 x, float16 *iptr); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld modf(double x, double *iptr); - double2 __ovld modf(double2 x, double2 *iptr); - double3 __ovld modf(double3 x, double3 *iptr); - double4 __ovld modf(double4 x, double4 *iptr); - double8 __ovld modf(double8 x, double8 *iptr); - double16 __ovld modf(double16 x, double16 *iptr); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld modf(half x, half *iptr); - half2 __ovld modf(half2 x, half2 *iptr); -@@ -7978,7 +7984,9 @@ half4 __ovld modf(half4 x, half4 *iptr); - half8 __ovld modf(half8 x, half8 *iptr); - half16 __ovld modf(half16 x, half16 *iptr); - #endif //cl_khr_fp16 --#else -+#endif //__opencl_c_generic_address_space -+ -+#if !defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != CL_VERSION_2_0) - float __ovld modf(float x, __global float *iptr); - float2 __ovld modf(float2 x, __global float2 *iptr); - float3 __ovld modf(float3 x, __global float3 *iptr); -@@ -7997,7 +8005,7 @@ float3 __ovld modf(float3 x, __private float3 *iptr); - float4 __ovld modf(float4 x, __private float4 *iptr); - float8 __ovld modf(float8 x, __private float8 *iptr); - float16 __ovld modf(float16 x, __private float16 *iptr); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld modf(double x, __global double *iptr); - double2 __ovld modf(double2 x, __global double2 *iptr); - double3 __ovld modf(double3 x, __global double3 *iptr); -@@ -8016,7 +8024,7 @@ double3 __ovld modf(double3 x, __private double3 *iptr); - double4 __ovld modf(double4 x, __private double4 *iptr); - double8 __ovld modf(double8 x, __private double8 *iptr); - double16 __ovld modf(double16 x, __private double16 *iptr); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld modf(half x, __global half *iptr); - half2 __ovld modf(half2 x, __global half2 *iptr); -@@ -8037,7 +8045,8 @@ half4 __ovld modf(half4 x, __private half4 *iptr); - half8 __ovld modf(half8 x, __private half8 *iptr); - half16 __ovld modf(half16 x, __private half16 *iptr); - #endif //cl_khr_fp16 --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#endif //! defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != -+ //! CL_VERSION_2_0) - - /** - * Returns a quiet NaN. The nancode may be placed -@@ -8049,14 +8058,14 @@ float3 __ovld __cnfn nan(uint3 nancode); - float4 __ovld __cnfn nan(uint4 nancode); - float8 __ovld __cnfn nan(uint8 nancode); - float16 __ovld __cnfn nan(uint16 nancode); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn nan(ulong nancode); - double2 __ovld __cnfn nan(ulong2 nancode); - double3 __ovld __cnfn nan(ulong3 nancode); - double4 __ovld __cnfn nan(ulong4 nancode); - double8 __ovld __cnfn nan(ulong8 nancode); - double16 __ovld __cnfn nan(ulong16 nancode); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn nan(ushort nancode); - half2 __ovld __cnfn nan(ushort2 nancode); -@@ -8079,14 +8088,14 @@ float3 __ovld __cnfn nextafter(float3 x, float3 y); - float4 __ovld __cnfn nextafter(float4 x, float4 y); - float8 __ovld __cnfn nextafter(float8 x, float8 y); - float16 __ovld __cnfn nextafter(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn nextafter(double x, double y); - double2 __ovld __cnfn nextafter(double2 x, double2 y); - double3 __ovld __cnfn nextafter(double3 x, double3 y); - double4 __ovld __cnfn nextafter(double4 x, double4 y); - double8 __ovld __cnfn nextafter(double8 x, double8 y); - double16 __ovld __cnfn nextafter(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn nextafter(half x, half y); - half2 __ovld __cnfn nextafter(half2 x, half2 y); -@@ -8105,14 +8114,14 @@ float3 __ovld __cnfn pow(float3 x, float3 y); - float4 __ovld __cnfn pow(float4 x, float4 y); - float8 __ovld __cnfn pow(float8 x, float8 y); - float16 __ovld __cnfn pow(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn pow(double x, double y); - double2 __ovld __cnfn pow(double2 x, double2 y); - double3 __ovld __cnfn pow(double3 x, double3 y); - double4 __ovld __cnfn pow(double4 x, double4 y); - double8 __ovld __cnfn pow(double8 x, double8 y); - double16 __ovld __cnfn pow(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn pow(half x, half y); - half2 __ovld __cnfn pow(half2 x, half2 y); -@@ -8131,14 +8140,14 @@ float3 __ovld __cnfn pown(float3 x, int3 y); - float4 __ovld __cnfn pown(float4 x, int4 y); - float8 __ovld __cnfn pown(float8 x, int8 y); - float16 __ovld __cnfn pown(float16 x, int16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn pown(double x, int y); - double2 __ovld __cnfn pown(double2 x, int2 y); - double3 __ovld __cnfn pown(double3 x, int3 y); - double4 __ovld __cnfn pown(double4 x, int4 y); - double8 __ovld __cnfn pown(double8 x, int8 y); - double16 __ovld __cnfn pown(double16 x, int16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn pown(half x, int y); - half2 __ovld __cnfn pown(half2 x, int2 y); -@@ -8157,14 +8166,14 @@ float3 __ovld __cnfn powr(float3 x, float3 y); - float4 __ovld __cnfn powr(float4 x, float4 y); - float8 __ovld __cnfn powr(float8 x, float8 y); - float16 __ovld __cnfn powr(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn powr(double x, double y); - double2 __ovld __cnfn powr(double2 x, double2 y); - double3 __ovld __cnfn powr(double3 x, double3 y); - double4 __ovld __cnfn powr(double4 x, double4 y); - double8 __ovld __cnfn powr(double8 x, double8 y); - double16 __ovld __cnfn powr(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn powr(half x, half y); - half2 __ovld __cnfn powr(half2 x, half2 y); -@@ -8186,14 +8195,14 @@ float3 __ovld __cnfn remainder(float3 x, float3 y); - float4 __ovld __cnfn remainder(float4 x, float4 y); - float8 __ovld __cnfn remainder(float8 x, float8 y); - float16 __ovld __cnfn remainder(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn remainder(double x, double y); - double2 __ovld __cnfn remainder(double2 x, double2 y); - double3 __ovld __cnfn remainder(double3 x, double3 y); - double4 __ovld __cnfn remainder(double4 x, double4 y); - double8 __ovld __cnfn remainder(double8 x, double8 y); - double16 __ovld __cnfn remainder(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn remainder(half x, half y); - half2 __ovld __cnfn remainder(half2 x, half2 y); -@@ -8215,21 +8224,21 @@ half16 __ovld __cnfn remainder(half16 x, half16 y); - * sign as x/y. It stores this signed value in the object - * pointed to by quo. - */ --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - float __ovld remquo(float x, float y, int *quo); - float2 __ovld remquo(float2 x, float2 y, int2 *quo); - float3 __ovld remquo(float3 x, float3 y, int3 *quo); - float4 __ovld remquo(float4 x, float4 y, int4 *quo); - float8 __ovld remquo(float8 x, float8 y, int8 *quo); - float16 __ovld remquo(float16 x, float16 y, int16 *quo); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld remquo(double x, double y, int *quo); - double2 __ovld remquo(double2 x, double2 y, int2 *quo); - double3 __ovld remquo(double3 x, double3 y, int3 *quo); - double4 __ovld remquo(double4 x, double4 y, int4 *quo); - double8 __ovld remquo(double8 x, double8 y, int8 *quo); - double16 __ovld remquo(double16 x, double16 y, int16 *quo); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld remquo(half x, half y, int *quo); - half2 __ovld remquo(half2 x, half2 y, int2 *quo); -@@ -8237,9 +8246,10 @@ half3 __ovld remquo(half3 x, half3 y, int3 *quo); - half4 __ovld remquo(half4 x, half4 y, int4 *quo); - half8 __ovld remquo(half8 x, half8 y, int8 *quo); - half16 __ovld remquo(half16 x, half16 y, int16 *quo); -- - #endif //cl_khr_fp16 --#else -+#endif //__opencl_c_generic_address_space -+ -+#if !defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != CL_VERSION_2_0) - float __ovld remquo(float x, float y, __global int *quo); - float2 __ovld remquo(float2 x, float2 y, __global int2 *quo); - float3 __ovld remquo(float3 x, float3 y, __global int3 *quo); -@@ -8258,7 +8268,7 @@ float3 __ovld remquo(float3 x, float3 y, __private int3 *quo); - float4 __ovld remquo(float4 x, float4 y, __private int4 *quo); - float8 __ovld remquo(float8 x, float8 y, __private int8 *quo); - float16 __ovld remquo(float16 x, float16 y, __private int16 *quo); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld remquo(double x, double y, __global int *quo); - double2 __ovld remquo(double2 x, double2 y, __global int2 *quo); - double3 __ovld remquo(double3 x, double3 y, __global int3 *quo); -@@ -8277,7 +8287,7 @@ double3 __ovld remquo(double3 x, double3 y, __private int3 *quo); - double4 __ovld remquo(double4 x, double4 y, __private int4 *quo); - double8 __ovld remquo(double8 x, double8 y, __private int8 *quo); - double16 __ovld remquo(double16 x, double16 y, __private int16 *quo); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld remquo(half x, half y, __global int *quo); - half2 __ovld remquo(half2 x, half2 y, __global int2 *quo); -@@ -8298,7 +8308,8 @@ half4 __ovld remquo(half4 x, half4 y, __private int4 *quo); - half8 __ovld remquo(half8 x, half8 y, __private int8 *quo); - half16 __ovld remquo(half16 x, half16 y, __private int16 *quo); - #endif //cl_khr_fp16 --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#endif //! defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != -+ //! CL_VERSION_2_0) - /** - * Round to integral value (using round to nearest - * even rounding mode) in floating-point format. -@@ -8311,14 +8322,14 @@ float3 __ovld __cnfn rint(float3); - float4 __ovld __cnfn rint(float4); - float8 __ovld __cnfn rint(float8); - float16 __ovld __cnfn rint(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn rint(double); - double2 __ovld __cnfn rint(double2); - double3 __ovld __cnfn rint(double3); - double4 __ovld __cnfn rint(double4); - double8 __ovld __cnfn rint(double8); - double16 __ovld __cnfn rint(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn rint(half); - half2 __ovld __cnfn rint(half2); -@@ -8337,14 +8348,14 @@ float3 __ovld __cnfn rootn(float3 x, int3 y); - float4 __ovld __cnfn rootn(float4 x, int4 y); - float8 __ovld __cnfn rootn(float8 x, int8 y); - float16 __ovld __cnfn rootn(float16 x, int16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn rootn(double x, int y); - double2 __ovld __cnfn rootn(double2 x, int2 y); - double3 __ovld __cnfn rootn(double3 x, int3 y); - double4 __ovld __cnfn rootn(double4 x, int4 y); - double8 __ovld __cnfn rootn(double8 x, int8 y); - double16 __ovld __cnfn rootn(double16 x, int16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn rootn(half x, int y); - half2 __ovld __cnfn rootn(half2 x, int2 y); -@@ -8365,14 +8376,14 @@ float3 __ovld __cnfn round(float3 x); - float4 __ovld __cnfn round(float4 x); - float8 __ovld __cnfn round(float8 x); - float16 __ovld __cnfn round(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn round(double x); - double2 __ovld __cnfn round(double2 x); - double3 __ovld __cnfn round(double3 x); - double4 __ovld __cnfn round(double4 x); - double8 __ovld __cnfn round(double8 x); - double16 __ovld __cnfn round(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn round(half x); - half2 __ovld __cnfn round(half2 x); -@@ -8391,14 +8402,14 @@ float3 __ovld __cnfn rsqrt(float3); - float4 __ovld __cnfn rsqrt(float4); - float8 __ovld __cnfn rsqrt(float8); - float16 __ovld __cnfn rsqrt(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn rsqrt(double); - double2 __ovld __cnfn rsqrt(double2); - double3 __ovld __cnfn rsqrt(double3); - double4 __ovld __cnfn rsqrt(double4); - double8 __ovld __cnfn rsqrt(double8); - double16 __ovld __cnfn rsqrt(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn rsqrt(half); - half2 __ovld __cnfn rsqrt(half2); -@@ -8417,14 +8428,14 @@ float3 __ovld __cnfn sin(float3); - float4 __ovld __cnfn sin(float4); - float8 __ovld __cnfn sin(float8); - float16 __ovld __cnfn sin(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn sin(double); - double2 __ovld __cnfn sin(double2); - double3 __ovld __cnfn sin(double3); - double4 __ovld __cnfn sin(double4); - double8 __ovld __cnfn sin(double8); - double16 __ovld __cnfn sin(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn sin(half); - half2 __ovld __cnfn sin(half2); -@@ -8439,21 +8450,21 @@ half16 __ovld __cnfn sin(half16); - * is the return value and computed cosine is returned - * in cosval. - */ --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - float __ovld sincos(float x, float *cosval); - float2 __ovld sincos(float2 x, float2 *cosval); - float3 __ovld sincos(float3 x, float3 *cosval); - float4 __ovld sincos(float4 x, float4 *cosval); - float8 __ovld sincos(float8 x, float8 *cosval); - float16 __ovld sincos(float16 x, float16 *cosval); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld sincos(double x, double *cosval); - double2 __ovld sincos(double2 x, double2 *cosval); - double3 __ovld sincos(double3 x, double3 *cosval); - double4 __ovld sincos(double4 x, double4 *cosval); - double8 __ovld sincos(double8 x, double8 *cosval); - double16 __ovld sincos(double16 x, double16 *cosval); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld sincos(half x, half *cosval); - half2 __ovld sincos(half2 x, half2 *cosval); -@@ -8462,7 +8473,9 @@ half4 __ovld sincos(half4 x, half4 *cosval); - half8 __ovld sincos(half8 x, half8 *cosval); - half16 __ovld sincos(half16 x, half16 *cosval); - #endif //cl_khr_fp16 --#else -+#endif //__opencl_c_generic_address_space -+ -+#if !defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != CL_VERSION_2_0) - float __ovld sincos(float x, __global float *cosval); - float2 __ovld sincos(float2 x, __global float2 *cosval); - float3 __ovld sincos(float3 x, __global float3 *cosval); -@@ -8481,7 +8494,7 @@ float3 __ovld sincos(float3 x, __private float3 *cosval); - float4 __ovld sincos(float4 x, __private float4 *cosval); - float8 __ovld sincos(float8 x, __private float8 *cosval); - float16 __ovld sincos(float16 x, __private float16 *cosval); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld sincos(double x, __global double *cosval); - double2 __ovld sincos(double2 x, __global double2 *cosval); - double3 __ovld sincos(double3 x, __global double3 *cosval); -@@ -8500,7 +8513,7 @@ double3 __ovld sincos(double3 x, __private double3 *cosval); - double4 __ovld sincos(double4 x, __private double4 *cosval); - double8 __ovld sincos(double8 x, __private double8 *cosval); - double16 __ovld sincos(double16 x, __private double16 *cosval); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld sincos(half x, __global half *cosval); - half2 __ovld sincos(half2 x, __global half2 *cosval); -@@ -8521,8 +8534,8 @@ half4 __ovld sincos(half4 x, __private half4 *cosval); - half8 __ovld sincos(half8 x, __private half8 *cosval); - half16 __ovld sincos(half16 x, __private half16 *cosval); - #endif //cl_khr_fp16 --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -- -+#endif //! defined(__OPENCL_CPP_VERSION__) && (__OPENCL_C_VERSION__ != -+ //! CL_VERSION_2_0) - /** - * Compute hyperbolic sine. - */ -@@ -8532,14 +8545,14 @@ float3 __ovld __cnfn sinh(float3); - float4 __ovld __cnfn sinh(float4); - float8 __ovld __cnfn sinh(float8); - float16 __ovld __cnfn sinh(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn sinh(double); - double2 __ovld __cnfn sinh(double2); - double3 __ovld __cnfn sinh(double3); - double4 __ovld __cnfn sinh(double4); - double8 __ovld __cnfn sinh(double8); - double16 __ovld __cnfn sinh(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn sinh(half); - half2 __ovld __cnfn sinh(half2); -@@ -8558,14 +8571,14 @@ float3 __ovld __cnfn sinpi(float3 x); - float4 __ovld __cnfn sinpi(float4 x); - float8 __ovld __cnfn sinpi(float8 x); - float16 __ovld __cnfn sinpi(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn sinpi(double x); - double2 __ovld __cnfn sinpi(double2 x); - double3 __ovld __cnfn sinpi(double3 x); - double4 __ovld __cnfn sinpi(double4 x); - double8 __ovld __cnfn sinpi(double8 x); - double16 __ovld __cnfn sinpi(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn sinpi(half x); - half2 __ovld __cnfn sinpi(half2 x); -@@ -8584,14 +8597,14 @@ float3 __ovld __cnfn sqrt(float3); - float4 __ovld __cnfn sqrt(float4); - float8 __ovld __cnfn sqrt(float8); - float16 __ovld __cnfn sqrt(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn sqrt(double); - double2 __ovld __cnfn sqrt(double2); - double3 __ovld __cnfn sqrt(double3); - double4 __ovld __cnfn sqrt(double4); - double8 __ovld __cnfn sqrt(double8); - double16 __ovld __cnfn sqrt(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn sqrt(half); - half2 __ovld __cnfn sqrt(half2); -@@ -8610,14 +8623,14 @@ float3 __ovld __cnfn tan(float3); - float4 __ovld __cnfn tan(float4); - float8 __ovld __cnfn tan(float8); - float16 __ovld __cnfn tan(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn tan(double); - double2 __ovld __cnfn tan(double2); - double3 __ovld __cnfn tan(double3); - double4 __ovld __cnfn tan(double4); - double8 __ovld __cnfn tan(double8); - double16 __ovld __cnfn tan(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn tan(half); - half2 __ovld __cnfn tan(half2); -@@ -8636,14 +8649,14 @@ float3 __ovld __cnfn tanh(float3); - float4 __ovld __cnfn tanh(float4); - float8 __ovld __cnfn tanh(float8); - float16 __ovld __cnfn tanh(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn tanh(double); - double2 __ovld __cnfn tanh(double2); - double3 __ovld __cnfn tanh(double3); - double4 __ovld __cnfn tanh(double4); - double8 __ovld __cnfn tanh(double8); - double16 __ovld __cnfn tanh(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn tanh(half); - half2 __ovld __cnfn tanh(half2); -@@ -8662,14 +8675,14 @@ float3 __ovld __cnfn tanpi(float3 x); - float4 __ovld __cnfn tanpi(float4 x); - float8 __ovld __cnfn tanpi(float8 x); - float16 __ovld __cnfn tanpi(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn tanpi(double x); - double2 __ovld __cnfn tanpi(double2 x); - double3 __ovld __cnfn tanpi(double3 x); - double4 __ovld __cnfn tanpi(double4 x); - double8 __ovld __cnfn tanpi(double8 x); - double16 __ovld __cnfn tanpi(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn tanpi(half x); - half2 __ovld __cnfn tanpi(half2 x); -@@ -8688,14 +8701,14 @@ float3 __ovld __cnfn tgamma(float3); - float4 __ovld __cnfn tgamma(float4); - float8 __ovld __cnfn tgamma(float8); - float16 __ovld __cnfn tgamma(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn tgamma(double); - double2 __ovld __cnfn tgamma(double2); - double3 __ovld __cnfn tgamma(double3); - double4 __ovld __cnfn tgamma(double4); - double8 __ovld __cnfn tgamma(double8); - double16 __ovld __cnfn tgamma(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn tgamma(half); - half2 __ovld __cnfn tgamma(half2); -@@ -8715,14 +8728,14 @@ float3 __ovld __cnfn trunc(float3); - float4 __ovld __cnfn trunc(float4); - float8 __ovld __cnfn trunc(float8); - float16 __ovld __cnfn trunc(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn trunc(double); - double2 __ovld __cnfn trunc(double2); - double3 __ovld __cnfn trunc(double3); - double4 __ovld __cnfn trunc(double4); - double8 __ovld __cnfn trunc(double8); - double16 __ovld __cnfn trunc(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn trunc(half); - half2 __ovld __cnfn trunc(half2); -@@ -10108,7 +10121,7 @@ float3 __ovld __cnfn clamp(float3 x, float minval, float maxval); - float4 __ovld __cnfn clamp(float4 x, float minval, float maxval); - float8 __ovld __cnfn clamp(float8 x, float minval, float maxval); - float16 __ovld __cnfn clamp(float16 x, float minval, float maxval); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn clamp(double x, double minval, double maxval); - double2 __ovld __cnfn clamp(double2 x, double2 minval, double2 maxval); - double3 __ovld __cnfn clamp(double3 x, double3 minval, double3 maxval); -@@ -10120,7 +10133,7 @@ double3 __ovld __cnfn clamp(double3 x, double minval, double maxval); - double4 __ovld __cnfn clamp(double4 x, double minval, double maxval); - double8 __ovld __cnfn clamp(double8 x, double minval, double maxval); - double16 __ovld __cnfn clamp(double16 x, double minval, double maxval); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn clamp(half x, half minval, half maxval); - half2 __ovld __cnfn clamp(half2 x, half2 minval, half2 maxval); -@@ -10145,14 +10158,14 @@ float3 __ovld __cnfn degrees(float3 radians); - float4 __ovld __cnfn degrees(float4 radians); - float8 __ovld __cnfn degrees(float8 radians); - float16 __ovld __cnfn degrees(float16 radians); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn degrees(double radians); - double2 __ovld __cnfn degrees(double2 radians); - double3 __ovld __cnfn degrees(double3 radians); - double4 __ovld __cnfn degrees(double4 radians); - double8 __ovld __cnfn degrees(double8 radians); - double16 __ovld __cnfn degrees(double16 radians); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn degrees(half radians); - half2 __ovld __cnfn degrees(half2 radians); -@@ -10177,7 +10190,7 @@ float3 __ovld __cnfn max(float3 x, float y); - float4 __ovld __cnfn max(float4 x, float y); - float8 __ovld __cnfn max(float8 x, float y); - float16 __ovld __cnfn max(float16 x, float y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn max(double x, double y); - double2 __ovld __cnfn max(double2 x, double2 y); - double3 __ovld __cnfn max(double3 x, double3 y); -@@ -10189,7 +10202,7 @@ double3 __ovld __cnfn max(double3 x, double y); - double4 __ovld __cnfn max(double4 x, double y); - double8 __ovld __cnfn max(double8 x, double y); - double16 __ovld __cnfn max(double16 x, double y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn max(half x, half y); - half2 __ovld __cnfn max(half2 x, half2 y); -@@ -10219,7 +10232,7 @@ float3 __ovld __cnfn min(float3 x, float y); - float4 __ovld __cnfn min(float4 x, float y); - float8 __ovld __cnfn min(float8 x, float y); - float16 __ovld __cnfn min(float16 x, float y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn min(double x, double y); - double2 __ovld __cnfn min(double2 x, double2 y); - double3 __ovld __cnfn min(double3 x, double3 y); -@@ -10231,7 +10244,7 @@ double3 __ovld __cnfn min(double3 x, double y); - double4 __ovld __cnfn min(double4 x, double y); - double8 __ovld __cnfn min(double8 x, double y); - double16 __ovld __cnfn min(double16 x, double y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn min(half x, half y); - half2 __ovld __cnfn min(half2 x, half2 y); -@@ -10264,7 +10277,7 @@ float3 __ovld __cnfn mix(float3 x, float3 y, float a); - float4 __ovld __cnfn mix(float4 x, float4 y, float a); - float8 __ovld __cnfn mix(float8 x, float8 y, float a); - float16 __ovld __cnfn mix(float16 x, float16 y, float a); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn mix(double x, double y, double a); - double2 __ovld __cnfn mix(double2 x, double2 y, double2 a); - double3 __ovld __cnfn mix(double3 x, double3 y, double3 a); -@@ -10276,7 +10289,7 @@ double3 __ovld __cnfn mix(double3 x, double3 y, double a); - double4 __ovld __cnfn mix(double4 x, double4 y, double a); - double8 __ovld __cnfn mix(double8 x, double8 y, double a); - double16 __ovld __cnfn mix(double16 x, double16 y, double a); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn mix(half x, half y, half a); - half2 __ovld __cnfn mix(half2 x, half2 y, half2 a); -@@ -10301,14 +10314,14 @@ float3 __ovld __cnfn radians(float3 degrees); - float4 __ovld __cnfn radians(float4 degrees); - float8 __ovld __cnfn radians(float8 degrees); - float16 __ovld __cnfn radians(float16 degrees); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn radians(double degrees); - double2 __ovld __cnfn radians(double2 degrees); - double3 __ovld __cnfn radians(double3 degrees); - double4 __ovld __cnfn radians(double4 degrees); - double8 __ovld __cnfn radians(double8 degrees); - double16 __ovld __cnfn radians(double16 degrees); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn radians(half degrees); - half2 __ovld __cnfn radians(half2 degrees); -@@ -10332,7 +10345,7 @@ float3 __ovld __cnfn step(float edge, float3 x); - float4 __ovld __cnfn step(float edge, float4 x); - float8 __ovld __cnfn step(float edge, float8 x); - float16 __ovld __cnfn step(float edge, float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn step(double edge, double x); - double2 __ovld __cnfn step(double2 edge, double2 x); - double3 __ovld __cnfn step(double3 edge, double3 x); -@@ -10344,7 +10357,7 @@ double3 __ovld __cnfn step(double edge, double3 x); - double4 __ovld __cnfn step(double edge, double4 x); - double8 __ovld __cnfn step(double edge, double8 x); - double16 __ovld __cnfn step(double edge, double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn step(half edge, half x); - half2 __ovld __cnfn step(half2 edge, half2 x); -@@ -10383,7 +10396,7 @@ float3 __ovld __cnfn smoothstep(float edge0, float edge1, float3 x); - float4 __ovld __cnfn smoothstep(float edge0, float edge1, float4 x); - float8 __ovld __cnfn smoothstep(float edge0, float edge1, float8 x); - float16 __ovld __cnfn smoothstep(float edge0, float edge1, float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn smoothstep(double edge0, double edge1, double x); - double2 __ovld __cnfn smoothstep(double2 edge0, double2 edge1, double2 x); - double3 __ovld __cnfn smoothstep(double3 edge0, double3 edge1, double3 x); -@@ -10395,7 +10408,7 @@ double3 __ovld __cnfn smoothstep(double edge0, double edge1, double3 x); - double4 __ovld __cnfn smoothstep(double edge0, double edge1, double4 x); - double8 __ovld __cnfn smoothstep(double edge0, double edge1, double8 x); - double16 __ovld __cnfn smoothstep(double edge0, double edge1, double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn smoothstep(half edge0, half edge1, half x); - half2 __ovld __cnfn smoothstep(half2 edge0, half2 edge1, half2 x); -@@ -10420,14 +10433,14 @@ float3 __ovld __cnfn sign(float3 x); - float4 __ovld __cnfn sign(float4 x); - float8 __ovld __cnfn sign(float8 x); - float16 __ovld __cnfn sign(float16 x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn sign(double x); - double2 __ovld __cnfn sign(double2 x); - double3 __ovld __cnfn sign(double3 x); - double4 __ovld __cnfn sign(double4 x); - double8 __ovld __cnfn sign(double8 x); - double16 __ovld __cnfn sign(double16 x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn sign(half x); - half2 __ovld __cnfn sign(half2 x); -@@ -10445,10 +10458,10 @@ half16 __ovld __cnfn sign(half16 x); - */ - float4 __ovld __cnfn cross(float4 p0, float4 p1); - float3 __ovld __cnfn cross(float3 p0, float3 p1); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double4 __ovld __cnfn cross(double4 p0, double4 p1); - double3 __ovld __cnfn cross(double3 p0, double3 p1); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half4 __ovld __cnfn cross(half4 p0, half4 p1); - half3 __ovld __cnfn cross(half3 p0, half3 p1); -@@ -10461,12 +10474,12 @@ float __ovld __cnfn dot(float p0, float p1); - float __ovld __cnfn dot(float2 p0, float2 p1); - float __ovld __cnfn dot(float3 p0, float3 p1); - float __ovld __cnfn dot(float4 p0, float4 p1); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn dot(double p0, double p1); - double __ovld __cnfn dot(double2 p0, double2 p1); - double __ovld __cnfn dot(double3 p0, double3 p1); - double __ovld __cnfn dot(double4 p0, double4 p1); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn dot(half p0, half p1); - half __ovld __cnfn dot(half2 p0, half2 p1); -@@ -10482,12 +10495,12 @@ float __ovld __cnfn distance(float p0, float p1); - float __ovld __cnfn distance(float2 p0, float2 p1); - float __ovld __cnfn distance(float3 p0, float3 p1); - float __ovld __cnfn distance(float4 p0, float4 p1); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn distance(double p0, double p1); - double __ovld __cnfn distance(double2 p0, double2 p1); - double __ovld __cnfn distance(double3 p0, double3 p1); - double __ovld __cnfn distance(double4 p0, double4 p1); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn distance(half p0, half p1); - half __ovld __cnfn distance(half2 p0, half2 p1); -@@ -10503,12 +10516,12 @@ float __ovld __cnfn length(float p); - float __ovld __cnfn length(float2 p); - float __ovld __cnfn length(float3 p); - float __ovld __cnfn length(float4 p); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn length(double p); - double __ovld __cnfn length(double2 p); - double __ovld __cnfn length(double3 p); - double __ovld __cnfn length(double4 p); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn length(half p); - half __ovld __cnfn length(half2 p); -@@ -10524,12 +10537,12 @@ float __ovld __cnfn normalize(float p); - float2 __ovld __cnfn normalize(float2 p); - float3 __ovld __cnfn normalize(float3 p); - float4 __ovld __cnfn normalize(float4 p); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn normalize(double p); - double2 __ovld __cnfn normalize(double2 p); - double3 __ovld __cnfn normalize(double3 p); - double4 __ovld __cnfn normalize(double4 p); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn normalize(half p); - half2 __ovld __cnfn normalize(half2 p); -@@ -10610,14 +10623,14 @@ int3 __ovld __cnfn isequal(float3 x, float3 y); - int4 __ovld __cnfn isequal(float4 x, float4 y); - int8 __ovld __cnfn isequal(float8 x, float8 y); - int16 __ovld __cnfn isequal(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn isequal(double x, double y); - long2 __ovld __cnfn isequal(double2 x, double2 y); - long3 __ovld __cnfn isequal(double3 x, double3 y); - long4 __ovld __cnfn isequal(double4 x, double4 y); - long8 __ovld __cnfn isequal(double8 x, double8 y); - long16 __ovld __cnfn isequal(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn isequal(half x, half y); - short2 __ovld __cnfn isequal(half2 x, half2 y); -@@ -10636,14 +10649,14 @@ int3 __ovld __cnfn isnotequal(float3 x, float3 y); - int4 __ovld __cnfn isnotequal(float4 x, float4 y); - int8 __ovld __cnfn isnotequal(float8 x, float8 y); - int16 __ovld __cnfn isnotequal(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn isnotequal(double x, double y); - long2 __ovld __cnfn isnotequal(double2 x, double2 y); - long3 __ovld __cnfn isnotequal(double3 x, double3 y); - long4 __ovld __cnfn isnotequal(double4 x, double4 y); - long8 __ovld __cnfn isnotequal(double8 x, double8 y); - long16 __ovld __cnfn isnotequal(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn isnotequal(half x, half y); - short2 __ovld __cnfn isnotequal(half2 x, half2 y); -@@ -10662,14 +10675,14 @@ int3 __ovld __cnfn isgreater(float3 x, float3 y); - int4 __ovld __cnfn isgreater(float4 x, float4 y); - int8 __ovld __cnfn isgreater(float8 x, float8 y); - int16 __ovld __cnfn isgreater(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn isgreater(double x, double y); - long2 __ovld __cnfn isgreater(double2 x, double2 y); - long3 __ovld __cnfn isgreater(double3 x, double3 y); - long4 __ovld __cnfn isgreater(double4 x, double4 y); - long8 __ovld __cnfn isgreater(double8 x, double8 y); - long16 __ovld __cnfn isgreater(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn isgreater(half x, half y); - short2 __ovld __cnfn isgreater(half2 x, half2 y); -@@ -10688,14 +10701,14 @@ int3 __ovld __cnfn isgreaterequal(float3 x, float3 y); - int4 __ovld __cnfn isgreaterequal(float4 x, float4 y); - int8 __ovld __cnfn isgreaterequal(float8 x, float8 y); - int16 __ovld __cnfn isgreaterequal(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn isgreaterequal(double x, double y); - long2 __ovld __cnfn isgreaterequal(double2 x, double2 y); - long3 __ovld __cnfn isgreaterequal(double3 x, double3 y); - long4 __ovld __cnfn isgreaterequal(double4 x, double4 y); - long8 __ovld __cnfn isgreaterequal(double8 x, double8 y); - long16 __ovld __cnfn isgreaterequal(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn isgreaterequal(half x, half y); - short2 __ovld __cnfn isgreaterequal(half2 x, half2 y); -@@ -10714,14 +10727,14 @@ int3 __ovld __cnfn isless(float3 x, float3 y); - int4 __ovld __cnfn isless(float4 x, float4 y); - int8 __ovld __cnfn isless(float8 x, float8 y); - int16 __ovld __cnfn isless(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn isless(double x, double y); - long2 __ovld __cnfn isless(double2 x, double2 y); - long3 __ovld __cnfn isless(double3 x, double3 y); - long4 __ovld __cnfn isless(double4 x, double4 y); - long8 __ovld __cnfn isless(double8 x, double8 y); - long16 __ovld __cnfn isless(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn isless(half x, half y); - short2 __ovld __cnfn isless(half2 x, half2 y); -@@ -10740,14 +10753,14 @@ int3 __ovld __cnfn islessequal(float3 x, float3 y); - int4 __ovld __cnfn islessequal(float4 x, float4 y); - int8 __ovld __cnfn islessequal(float8 x, float8 y); - int16 __ovld __cnfn islessequal(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn islessequal(double x, double y); - long2 __ovld __cnfn islessequal(double2 x, double2 y); - long3 __ovld __cnfn islessequal(double3 x, double3 y); - long4 __ovld __cnfn islessequal(double4 x, double4 y); - long8 __ovld __cnfn islessequal(double8 x, double8 y); - long16 __ovld __cnfn islessequal(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn islessequal(half x, half y); - short2 __ovld __cnfn islessequal(half2 x, half2 y); -@@ -10767,14 +10780,14 @@ int3 __ovld __cnfn islessgreater(float3 x, float3 y); - int4 __ovld __cnfn islessgreater(float4 x, float4 y); - int8 __ovld __cnfn islessgreater(float8 x, float8 y); - int16 __ovld __cnfn islessgreater(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn islessgreater(double x, double y); - long2 __ovld __cnfn islessgreater(double2 x, double2 y); - long3 __ovld __cnfn islessgreater(double3 x, double3 y); - long4 __ovld __cnfn islessgreater(double4 x, double4 y); - long8 __ovld __cnfn islessgreater(double8 x, double8 y); - long16 __ovld __cnfn islessgreater(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn islessgreater(half x, half y); - short2 __ovld __cnfn islessgreater(half2 x, half2 y); -@@ -10793,14 +10806,14 @@ int3 __ovld __cnfn isfinite(float3); - int4 __ovld __cnfn isfinite(float4); - int8 __ovld __cnfn isfinite(float8); - int16 __ovld __cnfn isfinite(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn isfinite(double); - long2 __ovld __cnfn isfinite(double2); - long3 __ovld __cnfn isfinite(double3); - long4 __ovld __cnfn isfinite(double4); - long8 __ovld __cnfn isfinite(double8); - long16 __ovld __cnfn isfinite(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn isfinite(half); - short2 __ovld __cnfn isfinite(half2); -@@ -10819,14 +10832,14 @@ int3 __ovld __cnfn isinf(float3); - int4 __ovld __cnfn isinf(float4); - int8 __ovld __cnfn isinf(float8); - int16 __ovld __cnfn isinf(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn isinf(double); - long2 __ovld __cnfn isinf(double2); - long3 __ovld __cnfn isinf(double3); - long4 __ovld __cnfn isinf(double4); - long8 __ovld __cnfn isinf(double8); - long16 __ovld __cnfn isinf(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn isinf(half); - short2 __ovld __cnfn isinf(half2); -@@ -10845,14 +10858,14 @@ int3 __ovld __cnfn isnan(float3); - int4 __ovld __cnfn isnan(float4); - int8 __ovld __cnfn isnan(float8); - int16 __ovld __cnfn isnan(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn isnan(double); - long2 __ovld __cnfn isnan(double2); - long3 __ovld __cnfn isnan(double3); - long4 __ovld __cnfn isnan(double4); - long8 __ovld __cnfn isnan(double8); - long16 __ovld __cnfn isnan(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn isnan(half); - short2 __ovld __cnfn isnan(half2); -@@ -10871,14 +10884,14 @@ int3 __ovld __cnfn isnormal(float3); - int4 __ovld __cnfn isnormal(float4); - int8 __ovld __cnfn isnormal(float8); - int16 __ovld __cnfn isnormal(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn isnormal(double); - long2 __ovld __cnfn isnormal(double2); - long3 __ovld __cnfn isnormal(double3); - long4 __ovld __cnfn isnormal(double4); - long8 __ovld __cnfn isnormal(double8); - long16 __ovld __cnfn isnormal(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn isnormal(half); - short2 __ovld __cnfn isnormal(half2); -@@ -10899,14 +10912,14 @@ int3 __ovld __cnfn isordered(float3 x, float3 y); - int4 __ovld __cnfn isordered(float4 x, float4 y); - int8 __ovld __cnfn isordered(float8 x, float8 y); - int16 __ovld __cnfn isordered(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn isordered(double x, double y); - long2 __ovld __cnfn isordered(double2 x, double2 y); - long3 __ovld __cnfn isordered(double3 x, double3 y); - long4 __ovld __cnfn isordered(double4 x, double4 y); - long8 __ovld __cnfn isordered(double8 x, double8 y); - long16 __ovld __cnfn isordered(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn isordered(half x, half y); - short2 __ovld __cnfn isordered(half2 x, half2 y); -@@ -10927,14 +10940,14 @@ int3 __ovld __cnfn isunordered(float3 x, float3 y); - int4 __ovld __cnfn isunordered(float4 x, float4 y); - int8 __ovld __cnfn isunordered(float8 x, float8 y); - int16 __ovld __cnfn isunordered(float16 x, float16 y); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn isunordered(double x, double y); - long2 __ovld __cnfn isunordered(double2 x, double2 y); - long3 __ovld __cnfn isunordered(double3 x, double3 y); - long4 __ovld __cnfn isunordered(double4 x, double4 y); - long8 __ovld __cnfn isunordered(double8 x, double8 y); - long16 __ovld __cnfn isunordered(double16 x, double16 y); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn isunordered(half x, half y); - short2 __ovld __cnfn isunordered(half2 x, half2 y); -@@ -10957,14 +10970,14 @@ int3 __ovld __cnfn signbit(float3); - int4 __ovld __cnfn signbit(float4); - int8 __ovld __cnfn signbit(float8); - int16 __ovld __cnfn signbit(float16); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - int __ovld __cnfn signbit(double); - long2 __ovld __cnfn signbit(double2); - long3 __ovld __cnfn signbit(double3); - long4 __ovld __cnfn signbit(double4); - long8 __ovld __cnfn signbit(double8); - long16 __ovld __cnfn signbit(double16); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - int __ovld __cnfn signbit(half); - short2 __ovld __cnfn signbit(half2); -@@ -11091,14 +11104,14 @@ float3 __ovld __cnfn bitselect(float3 a, float3 b, float3 c); - float4 __ovld __cnfn bitselect(float4 a, float4 b, float4 c); - float8 __ovld __cnfn bitselect(float8 a, float8 b, float8 c); - float16 __ovld __cnfn bitselect(float16 a, float16 b, float16 c); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn bitselect(double a, double b, double c); - double2 __ovld __cnfn bitselect(double2 a, double2 b, double2 c); - double3 __ovld __cnfn bitselect(double3 a, double3 b, double3 c); - double4 __ovld __cnfn bitselect(double4 a, double4 b, double4 c); - double8 __ovld __cnfn bitselect(double8 a, double8 b, double8 c); - double16 __ovld __cnfn bitselect(double16 a, double16 b, double16 c); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn bitselect(half a, half b, half c); - half2 __ovld __cnfn bitselect(half2 a, half2 b, half2 c); -@@ -11231,7 +11244,7 @@ ulong8 __ovld __cnfn select(ulong8 a, ulong8 b, ulong8 c); - long16 __ovld __cnfn select(long16 a, long16 b, ulong16 c); - ulong16 __ovld __cnfn select(ulong16 a, ulong16 b, ulong16 c); - --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __cnfn select(double a, double b, long c); - double2 __ovld __cnfn select(double2 a, double2 b, long2 c); - double3 __ovld __cnfn select(double3 a, double3 b, long3 c); -@@ -11244,7 +11257,7 @@ double3 __ovld __cnfn select(double3 a, double3 b, ulong3 c); - double4 __ovld __cnfn select(double4 a, double4 b, ulong4 c); - double8 __ovld __cnfn select(double8 a, double8 b, ulong8 c); - double16 __ovld __cnfn select(double16 a, double16 b, ulong16 c); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - half __ovld __cnfn select(half a, half b, short c); - half2 __ovld __cnfn select(half2 a, half2 b, short2 c); -@@ -11323,13 +11336,13 @@ uint16 __ovld vload16(size_t offset, const __constant uint *p); - long16 __ovld vload16(size_t offset, const __constant long *p); - ulong16 __ovld vload16(size_t offset, const __constant ulong *p); - float16 __ovld vload16(size_t offset, const __constant float *p); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double2 __ovld vload2(size_t offset, const __constant double *p); - double3 __ovld vload3(size_t offset, const __constant double *p); - double4 __ovld vload4(size_t offset, const __constant double *p); - double8 __ovld vload8(size_t offset, const __constant double *p); - double16 __ovld vload16(size_t offset, const __constant double *p); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - #ifdef cl_khr_fp16 - half __ovld vload(size_t offset, const __constant half *p); -@@ -11340,7 +11353,7 @@ half8 __ovld vload8(size_t offset, const __constant half *p); - half16 __ovld vload16(size_t offset, const __constant half *p); - #endif //cl_khr_fp16 - --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - char2 __ovld vload2(size_t offset, const char *p); - uchar2 __ovld vload2(size_t offset, const uchar *p); - short2 __ovld vload2(size_t offset, const short *p); -@@ -11387,13 +11400,13 @@ long16 __ovld vload16(size_t offset, const long *p); - ulong16 __ovld vload16(size_t offset, const ulong *p); - float16 __ovld vload16(size_t offset, const float *p); - --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double2 __ovld vload2(size_t offset, const double *p); - double3 __ovld vload3(size_t offset, const double *p); - double4 __ovld vload4(size_t offset, const double *p); - double8 __ovld vload8(size_t offset, const double *p); - double16 __ovld vload16(size_t offset, const double *p); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - #ifdef cl_khr_fp16 - half __ovld vload(size_t offset, const half *p); -@@ -11403,7 +11416,7 @@ half4 __ovld vload4(size_t offset, const half *p); - half8 __ovld vload8(size_t offset, const half *p); - half16 __ovld vload16(size_t offset, const half *p); - #endif //cl_khr_fp16 --#else -+#endif //__opencl_c_generic_address_space - char2 __ovld vload2(size_t offset, const __global char *p); - uchar2 __ovld vload2(size_t offset, const __global uchar *p); - short2 __ovld vload2(size_t offset, const __global short *p); -@@ -11540,7 +11553,7 @@ long16 __ovld vload16(size_t offset, const __private long *p); - ulong16 __ovld vload16(size_t offset, const __private ulong *p); - float16 __ovld vload16(size_t offset, const __private float *p); - --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double2 __ovld vload2(size_t offset, const __global double *p); - double3 __ovld vload3(size_t offset, const __global double *p); - double4 __ovld vload4(size_t offset, const __global double *p); -@@ -11556,7 +11569,7 @@ double3 __ovld vload3(size_t offset, const __private double *p); - double4 __ovld vload4(size_t offset, const __private double *p); - double8 __ovld vload8(size_t offset, const __private double *p); - double16 __ovld vload16(size_t offset, const __private double *p); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - #ifdef cl_khr_fp16 - half __ovld vload(size_t offset, const __global half *p); -@@ -11578,9 +11591,8 @@ half4 __ovld vload4(size_t offset, const __private half *p); - half8 __ovld vload8(size_t offset, const __private half *p); - half16 __ovld vload16(size_t offset, const __private half *p); - #endif //cl_khr_fp16 --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - void __ovld vstore2(char2 data, size_t offset, char *p); - void __ovld vstore2(uchar2 data, size_t offset, uchar *p); - void __ovld vstore2(short2 data, size_t offset, short *p); -@@ -11626,13 +11638,13 @@ void __ovld vstore16(uint16 data, size_t offset, uint *p); - void __ovld vstore16(long16 data, size_t offset, long *p); - void __ovld vstore16(ulong16 data, size_t offset, ulong *p); - void __ovld vstore16(float16 data, size_t offset, float *p); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - void __ovld vstore2(double2 data, size_t offset, double *p); - void __ovld vstore3(double3 data, size_t offset, double *p); - void __ovld vstore4(double4 data, size_t offset, double *p); - void __ovld vstore8(double8 data, size_t offset, double *p); - void __ovld vstore16(double16 data, size_t offset, double *p); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - void __ovld vstore(half data, size_t offset, half *p); - void __ovld vstore2(half2 data, size_t offset, half *p); -@@ -11641,7 +11653,7 @@ void __ovld vstore4(half4 data, size_t offset, half *p); - void __ovld vstore8(half8 data, size_t offset, half *p); - void __ovld vstore16(half16 data, size_t offset, half *p); - #endif //cl_khr_fp16 --#else -+#endif //__opencl_c_generic_address_space - void __ovld vstore2(char2 data, size_t offset, __global char *p); - void __ovld vstore2(uchar2 data, size_t offset, __global uchar *p); - void __ovld vstore2(short2 data, size_t offset, __global short *p); -@@ -11777,7 +11789,7 @@ void __ovld vstore16(uint16 data, size_t offset, __private uint *p); - void __ovld vstore16(long16 data, size_t offset, __private long *p); - void __ovld vstore16(ulong16 data, size_t offset, __private ulong *p); - void __ovld vstore16(float16 data, size_t offset, __private float *p); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - void __ovld vstore2(double2 data, size_t offset, __global double *p); - void __ovld vstore3(double3 data, size_t offset, __global double *p); - void __ovld vstore4(double4 data, size_t offset, __global double *p); -@@ -11793,7 +11805,7 @@ void __ovld vstore3(double3 data, size_t offset, __private double *p); - void __ovld vstore4(double4 data, size_t offset, __private double *p); - void __ovld vstore8(double8 data, size_t offset, __private double *p); - void __ovld vstore16(double16 data, size_t offset, __private double *p); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - void __ovld vstore(half data, size_t offset, __global half *p); - void __ovld vstore2(half2 data, size_t offset, __global half *p); -@@ -11814,7 +11826,6 @@ void __ovld vstore4(half4 data, size_t offset, __private half *p); - void __ovld vstore8(half8 data, size_t offset, __private half *p); - void __ovld vstore16(half16 data, size_t offset, __private half *p); - #endif //cl_khr_fp16 --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** - * Read sizeof (half) bytes of data from address -@@ -11825,13 +11836,12 @@ void __ovld vstore16(half16 data, size_t offset, __private half *p); - * must be 16-bit aligned. - */ - float __ovld vload_half(size_t offset, const __constant half *p); --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - float __ovld vload_half(size_t offset, const half *p); --#else -+#endif //__opencl_c_generic_address_space - float __ovld vload_half(size_t offset, const __global half *p); - float __ovld vload_half(size_t offset, const __local half *p); - float __ovld vload_half(size_t offset, const __private half *p); --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** - * Read sizeof (halfn) bytes of data from address -@@ -11846,13 +11856,13 @@ float3 __ovld vload_half3(size_t offset, const __constant half *p); - float4 __ovld vload_half4(size_t offset, const __constant half *p); - float8 __ovld vload_half8(size_t offset, const __constant half *p); - float16 __ovld vload_half16(size_t offset, const __constant half *p); --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - float2 __ovld vload_half2(size_t offset, const half *p); - float3 __ovld vload_half3(size_t offset, const half *p); - float4 __ovld vload_half4(size_t offset, const half *p); - float8 __ovld vload_half8(size_t offset, const half *p); - float16 __ovld vload_half16(size_t offset, const half *p); --#else -+#endif //__opencl_c_generic_address_space - float2 __ovld vload_half2(size_t offset, const __global half *p); - float3 __ovld vload_half3(size_t offset, const __global half *p); - float4 __ovld vload_half4(size_t offset, const __global half *p); -@@ -11868,7 +11878,6 @@ float3 __ovld vload_half3(size_t offset, const __private half *p); - float4 __ovld vload_half4(size_t offset, const __private half *p); - float8 __ovld vload_half8(size_t offset, const __private half *p); - float16 __ovld vload_half16(size_t offset, const __private half *p); --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** - * The float value given by data is first -@@ -11881,20 +11890,20 @@ float16 __ovld vload_half16(size_t offset, const __private half *p); - * The default current rounding mode is round to - * nearest even. - */ --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - void __ovld vstore_half(float data, size_t offset, half *p); - void __ovld vstore_half_rte(float data, size_t offset, half *p); - void __ovld vstore_half_rtz(float data, size_t offset, half *p); - void __ovld vstore_half_rtp(float data, size_t offset, half *p); - void __ovld vstore_half_rtn(float data, size_t offset, half *p); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - void __ovld vstore_half(double data, size_t offset, half *p); - void __ovld vstore_half_rte(double data, size_t offset, half *p); - void __ovld vstore_half_rtz(double data, size_t offset, half *p); - void __ovld vstore_half_rtp(double data, size_t offset, half *p); - void __ovld vstore_half_rtn(double data, size_t offset, half *p); --#endif //cl_khr_fp64 --#else -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#endif //__opencl_c_generic_address_space - void __ovld vstore_half(float data, size_t offset, __global half *p); - void __ovld vstore_half_rte(float data, size_t offset, __global half *p); - void __ovld vstore_half_rtz(float data, size_t offset, __global half *p); -@@ -11910,7 +11919,7 @@ void __ovld vstore_half_rte(float data, size_t offset, __private half *p); - void __ovld vstore_half_rtz(float data, size_t offset, __private half *p); - void __ovld vstore_half_rtp(float data, size_t offset, __private half *p); - void __ovld vstore_half_rtn(float data, size_t offset, __private half *p); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - void __ovld vstore_half(double data, size_t offset, __global half *p); - void __ovld vstore_half_rte(double data, size_t offset, __global half *p); - void __ovld vstore_half_rtz(double data, size_t offset, __global half *p); -@@ -11926,8 +11935,7 @@ void __ovld vstore_half_rte(double data, size_t offset, __private half *p); - void __ovld vstore_half_rtz(double data, size_t offset, __private half *p); - void __ovld vstore_half_rtp(double data, size_t offset, __private half *p); - void __ovld vstore_half_rtn(double data, size_t offset, __private half *p); --#endif //cl_khr_fp64 --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - /** - * The floatn value given by data is converted to -@@ -11940,7 +11948,7 @@ void __ovld vstore_half_rtn(double data, size_t offset, __private half *p); - * The default current rounding mode is round to - * nearest even. - */ --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - void __ovld vstore_half2(float2 data, size_t offset, half *p); - void __ovld vstore_half3(float3 data, size_t offset, half *p); - void __ovld vstore_half4(float4 data, size_t offset, half *p); -@@ -11966,7 +11974,7 @@ void __ovld vstore_half3_rtn(float3 data, size_t offset, half *p); - void __ovld vstore_half4_rtn(float4 data, size_t offset, half *p); - void __ovld vstore_half8_rtn(float8 data, size_t offset, half *p); - void __ovld vstore_half16_rtn(float16 data, size_t offset, half *p); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - void __ovld vstore_half2(double2 data, size_t offset, half *p); - void __ovld vstore_half3(double3 data, size_t offset, half *p); - void __ovld vstore_half4(double4 data, size_t offset, half *p); -@@ -11992,8 +12000,8 @@ void __ovld vstore_half3_rtn(double3 data, size_t offset, half *p); - void __ovld vstore_half4_rtn(double4 data, size_t offset, half *p); - void __ovld vstore_half8_rtn(double8 data, size_t offset, half *p); - void __ovld vstore_half16_rtn(double16 data, size_t offset, half *p); --#endif //cl_khr_fp64 --#else -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#endif //__opencl_c_generic_address_space - void __ovld vstore_half2(float2 data, size_t offset, __global half *p); - void __ovld vstore_half3(float3 data, size_t offset, __global half *p); - void __ovld vstore_half4(float4 data, size_t offset, __global half *p); -@@ -12069,7 +12077,7 @@ void __ovld vstore_half3_rtn(float3 data, size_t offset, __private half *p); - void __ovld vstore_half4_rtn(float4 data, size_t offset, __private half *p); - void __ovld vstore_half8_rtn(float8 data, size_t offset, __private half *p); - void __ovld vstore_half16_rtn(float16 data, size_t offset, __private half *p); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - void __ovld vstore_half2(double2 data, size_t offset, __global half *p); - void __ovld vstore_half3(double3 data, size_t offset, __global half *p); - void __ovld vstore_half4(double4 data, size_t offset, __global half *p); -@@ -12145,8 +12153,7 @@ void __ovld vstore_half3_rtn(double3 data, size_t offset, __private half *p); - void __ovld vstore_half4_rtn(double4 data, size_t offset, __private half *p); - void __ovld vstore_half8_rtn(double8 data, size_t offset, __private half *p); - void __ovld vstore_half16_rtn(double16 data, size_t offset, __private half *p); --#endif //cl_khr_fp64 --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - /** - * For n = 1, 2, 4, 8 and 16 read sizeof (halfn) -@@ -12167,14 +12174,14 @@ float3 __ovld vloada_half3(size_t offset, const __constant half *p); - float4 __ovld vloada_half4(size_t offset, const __constant half *p); - float8 __ovld vloada_half8(size_t offset, const __constant half *p); - float16 __ovld vloada_half16(size_t offset, const __constant half *p); --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - float __ovld vloada_half(size_t offset, const half *p); - float2 __ovld vloada_half2(size_t offset, const half *p); - float3 __ovld vloada_half3(size_t offset, const half *p); - float4 __ovld vloada_half4(size_t offset, const half *p); - float8 __ovld vloada_half8(size_t offset, const half *p); - float16 __ovld vloada_half16(size_t offset, const half *p); --#else -+#endif //__opencl_c_generic_address_space - float __ovld vloada_half(size_t offset, const __global half *p); - float2 __ovld vloada_half2(size_t offset, const __global half *p); - float3 __ovld vloada_half3(size_t offset, const __global half *p); -@@ -12193,7 +12200,6 @@ float3 __ovld vloada_half3(size_t offset, const __private half *p); - float4 __ovld vloada_half4(size_t offset, const __private half *p); - float8 __ovld vloada_half8(size_t offset, const __private half *p); - float16 __ovld vloada_half16(size_t offset, const __private half *p); --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** - * The floatn value given by data is converted to -@@ -12211,7 +12217,7 @@ float16 __ovld vloada_half16(size_t offset, const __private half *p); - * mode. The default current rounding mode is - * round to nearest even. - */ --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - void __ovld vstorea_half(float data, size_t offset, half *p); - void __ovld vstorea_half2(float2 data, size_t offset, half *p); - void __ovld vstorea_half3(float3 data, size_t offset, half *p); -@@ -12247,7 +12253,7 @@ void __ovld vstorea_half4_rtn(float4 data, size_t offset, half *p); - void __ovld vstorea_half8_rtn(float8 data, size_t offset, half *p); - void __ovld vstorea_half16_rtn(float16 data, size_t offset, half *p); - --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - void __ovld vstorea_half(double data, size_t offset, half *p); - void __ovld vstorea_half2(double2 data, size_t offset, half *p); - void __ovld vstorea_half3(double3 data, size_t offset, half *p); -@@ -12282,9 +12288,9 @@ void __ovld vstorea_half3_rtn(double3 data, size_t offset, half *p); - void __ovld vstorea_half4_rtn(double4 data, size_t offset, half *p); - void __ovld vstorea_half8_rtn(double8 data, size_t offset, half *p); - void __ovld vstorea_half16_rtn(double16 data, size_t offset, half *p); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#endif //__opencl_c_generic_address_space - --#else - void __ovld vstorea_half(float data, size_t offset, __global half *p); - void __ovld vstorea_half2(float2 data, size_t offset, __global half *p); - void __ovld vstorea_half3(float3 data, size_t offset, __global half *p); -@@ -12390,7 +12396,7 @@ void __ovld vstorea_half4_rtn(float4 data, size_t offset, __private half *p); - void __ovld vstorea_half8_rtn(float8 data, size_t offset, __private half *p); - void __ovld vstorea_half16_rtn(float16 data, size_t offset, __private half *p); - --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - void __ovld vstorea_half(double data, size_t offset, __global half *p); - void __ovld vstorea_half2(double2 data, size_t offset, __global half *p); - void __ovld vstorea_half3(double3 data, size_t offset, __global half *p); -@@ -12495,8 +12501,7 @@ void __ovld vstorea_half3_rtn(double3 data,size_t offset, __private half *p); - void __ovld vstorea_half4_rtn(double4 data,size_t offset, __private half *p); - void __ovld vstorea_half8_rtn(double8 data,size_t offset, __private half *p); - void __ovld vstorea_half16_rtn(double16 data,size_t offset, __private half *p); --#endif //cl_khr_fp64 --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - // OpenCL v1.1 s6.11.8, v1.2 s6.12.8, v2.0 s6.13.8 - Synchronization Functions - -@@ -12580,7 +12585,7 @@ void __ovld write_mem_fence(cl_mem_fence_flags flags); - - // OpenCL v2.0 s6.13.9 - Address Space Qualifier Functions - --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_generic_address_space - cl_mem_fence_flags __ovld get_fence(const void *ptr); - cl_mem_fence_flags __ovld get_fence(void *ptr); - -@@ -12591,7 +12596,7 @@ cl_mem_fence_flags __ovld get_fence(void *ptr); - * where gentype is builtin type or user defined type. - */ - --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#endif //__opencl_c_generic_address_space - - // OpenCL v1.1 s6.11.10, v1.2 s6.12.10, v2.0 s6.13.10 - Async Copies from Global to Local Memory, Local to Global Memory, and Prefetch - -@@ -12730,7 +12735,7 @@ event_t __ovld async_work_group_copy(__global uint16 *dst, const __local uint16 - event_t __ovld async_work_group_copy(__global long16 *dst, const __local long16 *src, size_t num_elements, event_t event); - event_t __ovld async_work_group_copy(__global ulong16 *dst, const __local ulong16 *src, size_t num_elements, event_t event); - event_t __ovld async_work_group_copy(__global float16 *dst, const __local float16 *src, size_t num_elements, event_t event); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - event_t __ovld async_work_group_copy(__local double *dst, const __global double *src, size_t num_elements, event_t event); - event_t __ovld async_work_group_copy(__local double2 *dst, const __global double2 *src, size_t num_elements, event_t event); - event_t __ovld async_work_group_copy(__local double3 *dst, const __global double3 *src, size_t num_elements, event_t event); -@@ -12743,7 +12748,7 @@ event_t __ovld async_work_group_copy(__global double3 *dst, const __local double - event_t __ovld async_work_group_copy(__global double4 *dst, const __local double4 *src, size_t num_elements, event_t event); - event_t __ovld async_work_group_copy(__global double8 *dst, const __local double8 *src, size_t num_elements, event_t event); - event_t __ovld async_work_group_copy(__global double16 *dst, const __local double16 *src, size_t num_elements, event_t event); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - event_t __ovld async_work_group_copy(__local half *dst, const __global half *src, size_t num_elements, event_t event); - event_t __ovld async_work_group_copy(__local half2 *dst, const __global half2 *src, size_t num_elements, event_t event); -@@ -12893,7 +12898,7 @@ event_t __ovld async_work_group_strided_copy(__global uint16 *dst, const __local - event_t __ovld async_work_group_strided_copy(__global long16 *dst, const __local long16 *src, size_t num_elements, size_t dst_stride, event_t event); - event_t __ovld async_work_group_strided_copy(__global ulong16 *dst, const __local ulong16 *src, size_t num_elements, size_t dst_stride, event_t event); - event_t __ovld async_work_group_strided_copy(__global float16 *dst, const __local float16 *src, size_t num_elements, size_t dst_stride, event_t event); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - event_t __ovld async_work_group_strided_copy(__local double *dst, const __global double *src, size_t num_elements, size_t src_stride, event_t event); - event_t __ovld async_work_group_strided_copy(__local double2 *dst, const __global double2 *src, size_t num_elements, size_t src_stride, event_t event); - event_t __ovld async_work_group_strided_copy(__local double3 *dst, const __global double3 *src, size_t num_elements, size_t src_stride, event_t event); -@@ -12906,7 +12911,7 @@ event_t __ovld async_work_group_strided_copy(__global double3 *dst, const __loca - event_t __ovld async_work_group_strided_copy(__global double4 *dst, const __local double4 *src, size_t num_elements, size_t dst_stride, event_t event); - event_t __ovld async_work_group_strided_copy(__global double8 *dst, const __local double8 *src, size_t num_elements, size_t dst_stride, event_t event); - event_t __ovld async_work_group_strided_copy(__global double16 *dst, const __local double16 *src, size_t num_elements, size_t dst_stride, event_t event); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - event_t __ovld async_work_group_strided_copy(__local half *dst, const __global half *src, size_t num_elements, size_t src_stride, event_t event); - event_t __ovld async_work_group_strided_copy(__local half2 *dst, const __global half2 *src, size_t num_elements, size_t src_stride, event_t event); -@@ -12996,14 +13001,14 @@ void __ovld prefetch(const __global uint16 *p, size_t num_elements); - void __ovld prefetch(const __global long16 *p, size_t num_elements); - void __ovld prefetch(const __global ulong16 *p, size_t num_elements); - void __ovld prefetch(const __global float16 *p, size_t num_elements); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - void __ovld prefetch(const __global double *p, size_t num_elements); - void __ovld prefetch(const __global double2 *p, size_t num_elements); - void __ovld prefetch(const __global double3 *p, size_t num_elements); - void __ovld prefetch(const __global double4 *p, size_t num_elements); - void __ovld prefetch(const __global double8 *p, size_t num_elements); - void __ovld prefetch(const __global double16 *p, size_t num_elements); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #ifdef cl_khr_fp16 - void __ovld prefetch(const __global half *p, size_t num_elements); - void __ovld prefetch(const __global half2 *p, size_t num_elements); -@@ -13026,9 +13031,11 @@ void __ovld prefetch(const __global half16 *p, size_t num_elements); - * pointed by p. The function returns old. - */ - int __ovld atomic_add(volatile __global int *p, int val); --unsigned int __ovld atomic_add(volatile __global unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_add(volatile __global unsigned int *p, -+ unsigned int val); - int __ovld atomic_add(volatile __local int *p, int val); --unsigned int __ovld atomic_add(volatile __local unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_add(volatile __local unsigned int *p, -+ unsigned int val); - #ifdef __OPENCL_CPP_VERSION__ - int __ovld atomic_add(volatile int *p, int val); - unsigned int __ovld atomic_add(volatile unsigned int *p, unsigned int val); -@@ -13056,9 +13063,11 @@ unsigned long __ovld atom_add(volatile __local unsigned long *p, unsigned long v - * returns old. - */ - int __ovld atomic_sub(volatile __global int *p, int val); --unsigned int __ovld atomic_sub(volatile __global unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_sub(volatile __global unsigned int *p, -+ unsigned int val); - int __ovld atomic_sub(volatile __local int *p, int val); --unsigned int __ovld atomic_sub(volatile __local unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_sub(volatile __local unsigned int *p, -+ unsigned int val); - #ifdef __OPENCL_CPP_VERSION__ - int __ovld atomic_sub(volatile int *p, int val); - unsigned int __ovld atomic_sub(volatile unsigned int *p, unsigned int val); -@@ -13086,9 +13095,11 @@ unsigned long __ovld atom_sub(volatile __local unsigned long *p, unsigned long v - * value. - */ - int __ovld atomic_xchg(volatile __global int *p, int val); --unsigned int __ovld atomic_xchg(volatile __global unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_xchg(volatile __global unsigned int *p, -+ unsigned int val); - int __ovld atomic_xchg(volatile __local int *p, int val); --unsigned int __ovld atomic_xchg(volatile __local unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_xchg(volatile __local unsigned int *p, -+ unsigned int val); - float __ovld atomic_xchg(volatile __global float *p, float val); - float __ovld atomic_xchg(volatile __local float *p, float val); - #ifdef __OPENCL_CPP_VERSION__ -@@ -13183,12 +13194,15 @@ unsigned long __ovld atom_dec(volatile __local unsigned long *p); - * returns old. - */ - int __ovld atomic_cmpxchg(volatile __global int *p, int cmp, int val); --unsigned int __ovld atomic_cmpxchg(volatile __global unsigned int *p, unsigned int cmp, unsigned int val); -+unsigned int __ovld atomic_cmpxchg(volatile __global unsigned int *p, -+ unsigned int cmp, unsigned int val); - int __ovld atomic_cmpxchg(volatile __local int *p, int cmp, int val); --unsigned int __ovld atomic_cmpxchg(volatile __local unsigned int *p, unsigned int cmp, unsigned int val); -+unsigned int __ovld atomic_cmpxchg(volatile __local unsigned int *p, -+ unsigned int cmp, unsigned int val); - #ifdef __OPENCL_CPP_VERSION__ - int __ovld atomic_cmpxchg(volatile int *p, int cmp, int val); --unsigned int __ovld atomic_cmpxchg(volatile unsigned int *p, unsigned int cmp, unsigned int val); -+unsigned int __ovld atomic_cmpxchg(volatile unsigned int *p, unsigned int cmp, -+ unsigned int val); - #endif - - #if defined(cl_khr_global_int32_base_atomics) -@@ -13215,9 +13229,11 @@ unsigned long __ovld atom_cmpxchg(volatile __local unsigned long *p, unsigned lo - * returns old. - */ - int __ovld atomic_min(volatile __global int *p, int val); --unsigned int __ovld atomic_min(volatile __global unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_min(volatile __global unsigned int *p, -+ unsigned int val); - int __ovld atomic_min(volatile __local int *p, int val); --unsigned int __ovld atomic_min(volatile __local unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_min(volatile __local unsigned int *p, -+ unsigned int val); - #ifdef __OPENCL_CPP_VERSION__ - int __ovld atomic_min(volatile int *p, int val); - unsigned int __ovld atomic_min(volatile unsigned int *p, unsigned int val); -@@ -13247,9 +13263,11 @@ unsigned long __ovld atom_min(volatile __local unsigned long *p, unsigned long v - * returns old. - */ - int __ovld atomic_max(volatile __global int *p, int val); --unsigned int __ovld atomic_max(volatile __global unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_max(volatile __global unsigned int *p, -+ unsigned int val); - int __ovld atomic_max(volatile __local int *p, int val); --unsigned int __ovld atomic_max(volatile __local unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_max(volatile __local unsigned int *p, -+ unsigned int val); - #ifdef __OPENCL_CPP_VERSION__ - int __ovld atomic_max(volatile int *p, int val); - unsigned int __ovld atomic_max(volatile unsigned int *p, unsigned int val); -@@ -13278,9 +13296,11 @@ unsigned long __ovld atom_max(volatile __local unsigned long *p, unsigned long v - * pointed by p. The function returns old. - */ - int __ovld atomic_and(volatile __global int *p, int val); --unsigned int __ovld atomic_and(volatile __global unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_and(volatile __global unsigned int *p, -+ unsigned int val); - int __ovld atomic_and(volatile __local int *p, int val); --unsigned int __ovld atomic_and(volatile __local unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_and(volatile __local unsigned int *p, -+ unsigned int val); - #ifdef __OPENCL_CPP_VERSION__ - int __ovld atomic_and(volatile int *p, int val); - unsigned int __ovld atomic_and(volatile unsigned int *p, unsigned int val); -@@ -13309,9 +13329,11 @@ unsigned long __ovld atom_and(volatile __local unsigned long *p, unsigned long v - * pointed by p. The function returns old. - */ - int __ovld atomic_or(volatile __global int *p, int val); --unsigned int __ovld atomic_or(volatile __global unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_or(volatile __global unsigned int *p, -+ unsigned int val); - int __ovld atomic_or(volatile __local int *p, int val); --unsigned int __ovld atomic_or(volatile __local unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_or(volatile __local unsigned int *p, -+ unsigned int val); - #ifdef __OPENCL_CPP_VERSION__ - int __ovld atomic_or(volatile int *p, int val); - unsigned int __ovld atomic_or(volatile unsigned int *p, unsigned int val); -@@ -13340,9 +13362,11 @@ unsigned long __ovld atom_or(volatile __local unsigned long *p, unsigned long va - * pointed by p. The function returns old. - */ - int __ovld atomic_xor(volatile __global int *p, int val); --unsigned int __ovld atomic_xor(volatile __global unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_xor(volatile __global unsigned int *p, -+ unsigned int val); - int __ovld atomic_xor(volatile __local int *p, int val); --unsigned int __ovld atomic_xor(volatile __local unsigned int *p, unsigned int val); -+unsigned int __ovld atomic_xor(volatile __local unsigned int *p, -+ unsigned int val); - #ifdef __OPENCL_CPP_VERSION__ - int __ovld atomic_xor(volatile int *p, int val); - unsigned int __ovld atomic_xor(volatile unsigned int *p, unsigned int val); -@@ -13380,108 +13404,78 @@ unsigned long __ovld atom_xor(volatile __local unsigned long *p, unsigned long v - #endif - - // atomic_init() -+#ifdef __opencl_c_generic_address_space - void __ovld atomic_init(volatile atomic_int *object, int value); - void __ovld atomic_init(volatile atomic_uint *object, uint value); - void __ovld atomic_init(volatile atomic_float *object, float value); - #if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) - void __ovld atomic_init(volatile atomic_long *object, long value); - void __ovld atomic_init(volatile atomic_ulong *object, ulong value); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - void __ovld atomic_init(volatile atomic_double *object, double value); --#endif //cl_khr_fp64 --#endif -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // __opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+void __ovld atomic_init(volatile atomic_int __global *object, int value); -+void __ovld atomic_init(volatile atomic_int __local *object, int value); -+void __ovld atomic_init(volatile atomic_uint __global *object, uint value); -+void __ovld atomic_init(volatile atomic_uint __local *object, uint value); -+void __ovld atomic_init(volatile atomic_float __global *object, float value); -+void __ovld atomic_init(volatile atomic_float __local *object, float value); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+void __ovld atomic_init(volatile atomic_long __global *object, long value); -+void __ovld atomic_init(volatile atomic_long __local *object, long value); -+void __ovld atomic_init(volatile atomic_ulong __global *object, ulong value); -+void __ovld atomic_init(volatile atomic_ulong __local *object, ulong value); -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+void __ovld atomic_init(volatile atomic_double __global *object, double value); -+void __ovld atomic_init(volatile atomic_double __local *object, double value); -+#endif // cl_khr_fp64 -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) - - // atomic_work_item_fence() --void __ovld atomic_work_item_fence(cl_mem_fence_flags flags, memory_order order, memory_scope scope); -+void __ovld atomic_work_item_fence(cl_mem_fence_flags flags, memory_order order, -+ memory_scope scope); - - // atomic_fetch() -- -+#if defined(__opencl_c_atomic_scope_device) && \ -+ defined(__opencl_c_atomic_order_seq_cst) -+#ifdef __opencl_c_generic_address_space - int __ovld atomic_fetch_add(volatile atomic_int *object, int operand); --int __ovld atomic_fetch_add_explicit(volatile atomic_int *object, int operand, memory_order order); --int __ovld atomic_fetch_add_explicit(volatile atomic_int *object, int operand, memory_order order, memory_scope scope); - uint __ovld atomic_fetch_add(volatile atomic_uint *object, uint operand); --uint __ovld atomic_fetch_add_explicit(volatile atomic_uint *object, uint operand, memory_order order); --uint __ovld atomic_fetch_add_explicit(volatile atomic_uint *object, uint operand, memory_order order, memory_scope scope); - int __ovld atomic_fetch_sub(volatile atomic_int *object, int operand); --int __ovld atomic_fetch_sub_explicit(volatile atomic_int *object, int operand, memory_order order); --int __ovld atomic_fetch_sub_explicit(volatile atomic_int *object, int operand, memory_order order, memory_scope scope); - uint __ovld atomic_fetch_sub(volatile atomic_uint *object, uint operand); --uint __ovld atomic_fetch_sub_explicit(volatile atomic_uint *object, uint operand, memory_order order); --uint __ovld atomic_fetch_sub_explicit(volatile atomic_uint *object, uint operand, memory_order order, memory_scope scope); - int __ovld atomic_fetch_or(volatile atomic_int *object, int operand); --int __ovld atomic_fetch_or_explicit(volatile atomic_int *object, int operand, memory_order order); --int __ovld atomic_fetch_or_explicit(volatile atomic_int *object, int operand, memory_order order, memory_scope scope); - uint __ovld atomic_fetch_or(volatile atomic_uint *object, uint operand); --uint __ovld atomic_fetch_or_explicit(volatile atomic_uint *object, uint operand, memory_order order); --uint __ovld atomic_fetch_or_explicit(volatile atomic_uint *object, uint operand, memory_order order, memory_scope scope); - int __ovld atomic_fetch_xor(volatile atomic_int *object, int operand); --int __ovld atomic_fetch_xor_explicit(volatile atomic_int *object, int operand, memory_order order); --int __ovld atomic_fetch_xor_explicit(volatile atomic_int *object, int operand, memory_order order, memory_scope scope); - uint __ovld atomic_fetch_xor(volatile atomic_uint *object, uint operand); --uint __ovld atomic_fetch_xor_explicit(volatile atomic_uint *object, uint operand, memory_order order); --uint __ovld atomic_fetch_xor_explicit(volatile atomic_uint *object, uint operand, memory_order order, memory_scope scope); - int __ovld atomic_fetch_and(volatile atomic_int *object, int operand); --int __ovld atomic_fetch_and_explicit(volatile atomic_int *object, int operand, memory_order order); --int __ovld atomic_fetch_and_explicit(volatile atomic_int *object, int operand, memory_order order, memory_scope scope); - uint __ovld atomic_fetch_and(volatile atomic_uint *object, uint operand); --uint __ovld atomic_fetch_and_explicit(volatile atomic_uint *object, uint operand, memory_order order); --uint __ovld atomic_fetch_and_explicit(volatile atomic_uint *object, uint operand, memory_order order, memory_scope scope); - int __ovld atomic_fetch_min(volatile atomic_int *object, int operand); --int __ovld atomic_fetch_min_explicit(volatile atomic_int *object, int operand, memory_order order); --int __ovld atomic_fetch_min_explicit(volatile atomic_int *object, int operand, memory_order order, memory_scope scope); - uint __ovld atomic_fetch_min(volatile atomic_uint *object, uint operand); --uint __ovld atomic_fetch_min_explicit(volatile atomic_uint *object, uint operand, memory_order order); --uint __ovld atomic_fetch_min_explicit(volatile atomic_uint *object, uint operand, memory_order order, memory_scope scope); - int __ovld atomic_fetch_max(volatile atomic_int *object, int operand); --int __ovld atomic_fetch_max_explicit(volatile atomic_int *object, int operand, memory_order order); --int __ovld atomic_fetch_max_explicit(volatile atomic_int *object, int operand, memory_order order, memory_scope scope); - uint __ovld atomic_fetch_max(volatile atomic_uint *object, uint operand); --uint __ovld atomic_fetch_max_explicit(volatile atomic_uint *object, uint operand, memory_order order); --uint __ovld atomic_fetch_max_explicit(volatile atomic_uint *object, uint operand, memory_order order, memory_scope scope); - - #if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) - long __ovld atomic_fetch_add(volatile atomic_long *object, long operand); --long __ovld atomic_fetch_add_explicit(volatile atomic_long *object, long operand, memory_order order); --long __ovld atomic_fetch_add_explicit(volatile atomic_long *object, long operand, memory_order order, memory_scope scope); - ulong __ovld atomic_fetch_add(volatile atomic_ulong *object, ulong operand); --ulong __ovld atomic_fetch_add_explicit(volatile atomic_ulong *object, ulong operand, memory_order order); --ulong __ovld atomic_fetch_add_explicit(volatile atomic_ulong *object, ulong operand, memory_order order, memory_scope scope); - long __ovld atomic_fetch_sub(volatile atomic_long *object, long operand); --long __ovld atomic_fetch_sub_explicit(volatile atomic_long *object, long operand, memory_order order); --long __ovld atomic_fetch_sub_explicit(volatile atomic_long *object, long operand, memory_order order, memory_scope scope); - ulong __ovld atomic_fetch_sub(volatile atomic_ulong *object, ulong operand); --ulong __ovld atomic_fetch_sub_explicit(volatile atomic_ulong *object, ulong operand, memory_order order); --ulong __ovld atomic_fetch_sub_explicit(volatile atomic_ulong *object, ulong operand, memory_order order, memory_scope scope); - long __ovld atomic_fetch_or(volatile atomic_long *object, long operand); --long __ovld atomic_fetch_or_explicit(volatile atomic_long *object, long operand, memory_order order); --long __ovld atomic_fetch_or_explicit(volatile atomic_long *object, long operand, memory_order order, memory_scope scope); - ulong __ovld atomic_fetch_or(volatile atomic_ulong *object, ulong operand); --ulong __ovld atomic_fetch_or_explicit(volatile atomic_ulong *object, ulong operand, memory_order order); --ulong __ovld atomic_fetch_or_explicit(volatile atomic_ulong *object, ulong operand, memory_order order, memory_scope scope); - long __ovld atomic_fetch_xor(volatile atomic_long *object, long operand); --long __ovld atomic_fetch_xor_explicit(volatile atomic_long *object, long operand, memory_order order); --long __ovld atomic_fetch_xor_explicit(volatile atomic_long *object, long operand, memory_order order, memory_scope scope); - ulong __ovld atomic_fetch_xor(volatile atomic_ulong *object, ulong operand); --ulong __ovld atomic_fetch_xor_explicit(volatile atomic_ulong *object, ulong operand, memory_order order); --ulong __ovld atomic_fetch_xor_explicit(volatile atomic_ulong *object, ulong operand, memory_order order, memory_scope scope); - long __ovld atomic_fetch_and(volatile atomic_long *object, long operand); --long __ovld atomic_fetch_and_explicit(volatile atomic_long *object, long operand, memory_order order); --long __ovld atomic_fetch_and_explicit(volatile atomic_long *object, long operand, memory_order order, memory_scope scope); - ulong __ovld atomic_fetch_and(volatile atomic_ulong *object, ulong operand); --ulong __ovld atomic_fetch_and_explicit(volatile atomic_ulong *object, ulong operand, memory_order order); --ulong __ovld atomic_fetch_and_explicit(volatile atomic_ulong *object, ulong operand, memory_order order, memory_scope scope); - long __ovld atomic_fetch_min(volatile atomic_long *object, long operand); --long __ovld atomic_fetch_min_explicit(volatile atomic_long *object, long operand, memory_order order); --long __ovld atomic_fetch_min_explicit(volatile atomic_long *object, long operand, memory_order order, memory_scope scope); - ulong __ovld atomic_fetch_min(volatile atomic_ulong *object, ulong operand); --ulong __ovld atomic_fetch_min_explicit(volatile atomic_ulong *object, ulong operand, memory_order order); --ulong __ovld atomic_fetch_min_explicit(volatile atomic_ulong *object, ulong operand, memory_order order, memory_scope scope); - long __ovld atomic_fetch_max(volatile atomic_long *object, long operand); --long __ovld atomic_fetch_max_explicit(volatile atomic_long *object, long operand, memory_order order); --long __ovld atomic_fetch_max_explicit(volatile atomic_long *object, long operand, memory_order order, memory_scope scope); - ulong __ovld atomic_fetch_max(volatile atomic_ulong *object, ulong operand); --ulong __ovld atomic_fetch_max_explicit(volatile atomic_ulong *object, ulong operand, memory_order order); --ulong __ovld atomic_fetch_max_explicit(volatile atomic_ulong *object, ulong operand, memory_order order, memory_scope scope); - #endif //defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) - - // OpenCL v2.0 s6.13.11.7.5: -@@ -13489,196 +13483,2239 @@ ulong __ovld atomic_fetch_max_explicit(volatile atomic_ulong *object, ulong oper - // or/xor/and/min/max: atomic type argument can be intptr_t/uintptr_t, value type argument can be intptr_t/uintptr_t. - - #if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) --uintptr_t __ovld atomic_fetch_add(volatile atomic_uintptr_t *object, ptrdiff_t operand); --uintptr_t __ovld atomic_fetch_add_explicit(volatile atomic_uintptr_t *object, ptrdiff_t operand, memory_order order); --uintptr_t __ovld atomic_fetch_add_explicit(volatile atomic_uintptr_t *object, ptrdiff_t operand, memory_order order, memory_scope scope); --uintptr_t __ovld atomic_fetch_sub(volatile atomic_uintptr_t *object, ptrdiff_t operand); --uintptr_t __ovld atomic_fetch_sub_explicit(volatile atomic_uintptr_t *object, ptrdiff_t operand, memory_order order); --uintptr_t __ovld atomic_fetch_sub_explicit(volatile atomic_uintptr_t *object, ptrdiff_t operand, memory_order order, memory_scope scope); -- --uintptr_t __ovld atomic_fetch_or(volatile atomic_uintptr_t *object, intptr_t operand); --uintptr_t __ovld atomic_fetch_or_explicit(volatile atomic_uintptr_t *object, intptr_t operand, memory_order order); --uintptr_t __ovld atomic_fetch_or_explicit(volatile atomic_uintptr_t *object, intptr_t operand, memory_order order, memory_scope scope); --uintptr_t __ovld atomic_fetch_xor(volatile atomic_uintptr_t *object, intptr_t operand); --uintptr_t __ovld atomic_fetch_xor_explicit(volatile atomic_uintptr_t *object, intptr_t operand, memory_order order); --uintptr_t __ovld atomic_fetch_xor_explicit(volatile atomic_uintptr_t *object, intptr_t operand, memory_order order, memory_scope scope); --uintptr_t __ovld atomic_fetch_and(volatile atomic_uintptr_t *object, intptr_t operand); --uintptr_t __ovld atomic_fetch_and_explicit(volatile atomic_uintptr_t *object, intptr_t operand, memory_order order); --uintptr_t __ovld atomic_fetch_and_explicit(volatile atomic_uintptr_t *object, intptr_t operand, memory_order order, memory_scope scope); --uintptr_t __ovld atomic_fetch_min(volatile atomic_uintptr_t *object, intptr_t opermax); --uintptr_t __ovld atomic_fetch_min_explicit(volatile atomic_uintptr_t *object, intptr_t opermax, memory_order minder); --uintptr_t __ovld atomic_fetch_min_explicit(volatile atomic_uintptr_t *object, intptr_t opermax, memory_order minder, memory_scope scope); --uintptr_t __ovld atomic_fetch_max(volatile atomic_uintptr_t *object, intptr_t opermax); --uintptr_t __ovld atomic_fetch_max_explicit(volatile atomic_uintptr_t *object, intptr_t opermax, memory_order minder); --uintptr_t __ovld atomic_fetch_max_explicit(volatile atomic_uintptr_t *object, intptr_t opermax, memory_order minder, memory_scope scope); -- --intptr_t __ovld atomic_fetch_or(volatile atomic_intptr_t *object, uintptr_t operand); --intptr_t __ovld atomic_fetch_or_explicit(volatile atomic_intptr_t *object, uintptr_t operand, memory_order order); --intptr_t __ovld atomic_fetch_or_explicit(volatile atomic_intptr_t *object, uintptr_t operand, memory_order order, memory_scope scope); --intptr_t __ovld atomic_fetch_xor(volatile atomic_intptr_t *object, uintptr_t operand); --intptr_t __ovld atomic_fetch_xor_explicit(volatile atomic_intptr_t *object, uintptr_t operand, memory_order order); --intptr_t __ovld atomic_fetch_xor_explicit(volatile atomic_intptr_t *object, uintptr_t operand, memory_order order, memory_scope scope); --intptr_t __ovld atomic_fetch_and(volatile atomic_intptr_t *object, uintptr_t operand); --intptr_t __ovld atomic_fetch_and_explicit(volatile atomic_intptr_t *object, uintptr_t operand, memory_order order); --intptr_t __ovld atomic_fetch_and_explicit(volatile atomic_intptr_t *object, uintptr_t operand, memory_order order, memory_scope scope); --intptr_t __ovld atomic_fetch_min(volatile atomic_intptr_t *object, uintptr_t opermax); --intptr_t __ovld atomic_fetch_min_explicit(volatile atomic_intptr_t *object, uintptr_t opermax, memory_order minder); --intptr_t __ovld atomic_fetch_min_explicit(volatile atomic_intptr_t *object, uintptr_t opermax, memory_order minder, memory_scope scope); --intptr_t __ovld atomic_fetch_max(volatile atomic_intptr_t *object, uintptr_t opermax); --intptr_t __ovld atomic_fetch_max_explicit(volatile atomic_intptr_t *object, uintptr_t opermax, memory_order minder); --intptr_t __ovld atomic_fetch_max_explicit(volatile atomic_intptr_t *object, uintptr_t opermax, memory_order minder, memory_scope scope); -+uintptr_t __ovld atomic_fetch_add(volatile atomic_uintptr_t *object, -+ ptrdiff_t operand); -+uintptr_t __ovld atomic_fetch_sub(volatile atomic_uintptr_t *object, -+ ptrdiff_t operand); -+ -+uintptr_t __ovld atomic_fetch_or(volatile atomic_uintptr_t *object, -+ intptr_t operand); -+uintptr_t __ovld atomic_fetch_xor(volatile atomic_uintptr_t *object, -+ intptr_t operand); -+uintptr_t __ovld atomic_fetch_and(volatile atomic_uintptr_t *object, -+ intptr_t operand); -+uintptr_t __ovld atomic_fetch_min(volatile atomic_uintptr_t *object, -+ intptr_t opermax); -+uintptr_t __ovld atomic_fetch_max(volatile atomic_uintptr_t *object, -+ intptr_t opermax); -+ -+intptr_t __ovld atomic_fetch_or(volatile atomic_intptr_t *object, -+ uintptr_t operand); -+intptr_t __ovld atomic_fetch_xor(volatile atomic_intptr_t *object, -+ uintptr_t operand); -+intptr_t __ovld atomic_fetch_and(volatile atomic_intptr_t *object, -+ uintptr_t operand); -+intptr_t __ovld atomic_fetch_min(volatile atomic_intptr_t *object, -+ uintptr_t opermax); -+intptr_t __ovld atomic_fetch_max(volatile atomic_intptr_t *object, -+ uintptr_t opermax); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // __opencl_c_generic_address_space -+ -+#if(__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+int __ovld atomic_fetch_add(volatile atomic_int __global *object, int operand); -+uint __ovld atomic_fetch_add(volatile atomic_uint __local *object, -+ uint operand); -+int __ovld atomic_fetch_sub(volatile atomic_int __global *object, int operand); -+int __ovld atomic_fetch_sub(volatile atomic_int __local *object, int operand); -+uint __ovld atomic_fetch_sub(volatile atomic_uint __local *object, -+ uint operand); -+uint __ovld atomic_fetch_sub(volatile atomic_uint __global *object, -+ uint operand); -+int __ovld atomic_fetch_or(volatile atomic_int __global *object, int operand); -+uint __ovld atomic_fetch_sub(volatile atomic_uint __local *object, -+ uint operand); -+uint __ovld atomic_fetch_or(volatile atomic_uint __global *object, -+ uint operand); -+uint __ovld atomic_fetch_or(volatile atomic_uint __local *object, uint operand); -+int __ovld atomic_fetch_xor(volatile atomic_int __global *object, int operand); -+int __ovld atomic_fetch_xor(volatile atomic_int __local *object, int operand); -+uint __ovld atomic_fetch_xor(volatile atomic_uint __global *object, -+ uint operand); -+uint __ovld atomic_fetch_xor(volatile atomic_uint __local *object, -+ uint operand); -+int __ovld atomic_fetch_and(volatile atomic_int __global *object, int operand); -+int __ovld atomic_fetch_and(volatile atomic_int __local *object, int operand); -+uint __ovld atomic_fetch_and(volatile atomic_uint __global *object, -+ uint operand); -+uint __ovld atomic_fetch_and(volatile atomic_uint __local *object, -+ uint operand); -+int __ovld atomic_fetch_min(volatile atomic_int __global *object, int operand); -+int __ovld atomic_fetch_min(volatile atomic_int __local *object, int operand); -+uint __ovld atomic_fetch_min(volatile atomic_uint __global *object, -+ uint operand); -+uint __ovld atomic_fetch_min(volatile atomic_uint __local *object, -+ uint operand); -+int __ovld atomic_fetch_max(volatile atomic_int __global *object, int operand); -+int __ovld atomic_fetch_max(volatile atomic_int __local *object, int operand); -+uint __ovld atomic_fetch_max(volatile atomic_uint __global *object, -+ uint operand); -+uint __ovld atomic_fetch_max(volatile atomic_uint __local *object, -+ uint operand); -+ -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+long __ovld atomic_fetch_add(volatile atomic_long __global *object, -+ long operand); -+long __ovld atomic_fetch_add(volatile atomic_long __local *object, -+ long operand); -+ulong __ovld atomic_fetch_add(volatile atomic_ulong __global *object, -+ ulong operand); -+ulong __ovld atomic_fetch_add(volatile atomic_ulong __local *object, -+ ulong operand); -+long __ovld atomic_fetch_sub(volatile atomic_long __global *object, -+ long operand); -+long __ovld atomic_fetch_sub(volatile atomic_long __local *object, -+ long operand); -+ulong __ovld atomic_fetch_sub(volatile atomic_ulong __global *object, -+ ulong operand); -+ulong __ovld atomic_fetch_sub(volatile atomic_ulong __local *object, -+ ulong operand); -+long __ovld atomic_fetch_or(volatile atomic_long __global *object, -+ long operand); -+long __ovld atomic_fetch_or(volatile atomic_long __local *object, long operand); -+ulong __ovld atomic_fetch_or(volatile atomic_ulong __global *object, -+ ulong operand); -+ulong __ovld atomic_fetch_or(volatile atomic_ulong __local *object, -+ ulong operand); -+long __ovld atomic_fetch_xor(volatile atomic_long __global *object, -+ long operand); -+long __ovld atomic_fetch_xor(volatile atomic_long __local *object, -+ long operand); -+ulong __ovld atomic_fetch_xor(volatile atomic_ulong __global *object, -+ ulong operand); -+ulong __ovld atomic_fetch_xor(volatile atomic_ulong __local *object, -+ ulong operand); -+long __ovld atomic_fetch_and(volatile atomic_long __global *object, -+ long operand); -+long __ovld atomic_fetch_and(volatile atomic_long __local *object, -+ long operand); -+ulong __ovld atomic_fetch_and(volatile atomic_ulong __global *object, -+ ulong operand); -+ulong __ovld atomic_fetch_and(volatile atomic_ulong __local *object, -+ ulong operand); -+long __ovld atomic_fetch_min(volatile atomic_long __global *object, -+ long operand); -+long __ovld atomic_fetch_min(volatile atomic_long __local *object, -+ long operand); -+ulong __ovld atomic_fetch_min(volatile atomic_ulong __global *object, -+ ulong operand); -+ulong __ovld atomic_fetch_min(volatile atomic_ulong __local *object, -+ ulong operand); -+long __ovld atomic_fetch_max(volatile atomic_long __global *object, -+ long operand); -+long __ovld atomic_fetch_max(volatile atomic_long __local *object, -+ long operand); -+ulong __ovld atomic_fetch_max(volatile atomic_ulong __global *object, -+ ulong operand); -+ulong __ovld atomic_fetch_max(volatile atomic_ulong __local *object, -+ ulong operand); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+ -+// OpenCL v2.0 s6.13.11.7.5: -+// add/sub: atomic type argument can be uintptr_t/intptr_t, value type argument -+// can be ptrdiff_t. or/xor/and/min/max: atomic type argument can be -+// intptr_t/uintptr_t, value type argument can be intptr_t/uintptr_t. -+ -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+uintptr_t __ovld atomic_fetch_add(volatile atomic_uintptr_t __global *object, -+ ptrdiff_t operand); -+uintptr_t __ovld atomic_fetch_add(volatile atomic_uintptr_t __local *object, -+ ptrdiff_t operand); -+uintptr_t __ovld atomic_fetch_sub(volatile atomic_uintptr_t __global *object, -+ ptrdiff_t operand); -+uintptr_t __ovld atomic_fetch_sub(volatile atomic_uintptr_t __local *object, -+ ptrdiff_t operand); -+ -+uintptr_t __ovld atomic_fetch_or(volatile atomic_uintptr_t __global *object, -+ intptr_t operand); -+uintptr_t __ovld atomic_fetch_or(volatile atomic_uintptr_t __local *object, -+ intptr_t operand); -+uintptr_t __ovld atomic_fetch_xor(volatile atomic_uintptr_t __global *object, -+ intptr_t operand); -+uintptr_t __ovld atomic_fetch_xor(volatile atomic_uintptr_t __local *object, -+ intptr_t operand); -+uintptr_t __ovld atomic_fetch_and(volatile atomic_uintptr_t __global *object, -+ intptr_t operand); -+uintptr_t __ovld atomic_fetch_and(volatile atomic_uintptr_t __local *object, -+ intptr_t operand); -+uintptr_t __ovld atomic_fetch_min(volatile atomic_uintptr_t __global *object, -+ intptr_t opermax); -+uintptr_t __ovld atomic_fetch_min(volatile atomic_uintptr_t __local *object, -+ intptr_t opermax); -+uintptr_t __ovld atomic_fetch_max(volatile atomic_uintptr_t __global *object, -+ intptr_t opermax); -+uintptr_t __ovld atomic_fetch_max(volatile atomic_uintptr_t __local *object, -+ intptr_t opermax); -+ -+intptr_t __ovld atomic_fetch_or(volatile atomic_intptr_t __global *object, -+ uintptr_t operand); -+intptr_t __ovld atomic_fetch_or(volatile atomic_intptr_t __local *object, -+ uintptr_t operand); -+intptr_t __ovld atomic_fetch_xor(volatile atomic_intptr_t __global *object, -+ uintptr_t operand); -+intptr_t __ovld atomic_fetch_xor(volatile atomic_intptr_t __local *object, -+ uintptr_t operand); -+intptr_t __ovld atomic_fetch_and(volatile atomic_intptr_t __global *object, -+ uintptr_t operand); -+intptr_t __ovld atomic_fetch_and(volatile atomic_intptr_t __local *object, -+ uintptr_t operand); -+intptr_t __ovld atomic_fetch_min(volatile atomic_intptr_t __global *object, -+ uintptr_t opermax); -+intptr_t __ovld atomic_fetch_min(volatile atomic_intptr_t __local *object, -+ uintptr_t opermax); -+intptr_t __ovld atomic_fetch_max(volatile atomic_intptr_t __global *object, -+ uintptr_t opermax); -+intptr_t __ovld atomic_fetch_max(volatile atomic_intptr_t __local *object, -+ uintptr_t opermax); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+ -+#endif // defined(__opencl_c_atomic_scope_device) && -+ // defined(__opencl_c_atomic_order_seq_cst) -+ -+#ifdef __opencl_c_generic_address_space -+#ifdef __opencl_c_atomic_scope_device -+int __ovld atomic_fetch_add_explicit(volatile atomic_int *object, int operand, -+ memory_order order); -+uint __ovld atomic_fetch_add_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order); -+int __ovld atomic_fetch_sub_explicit(volatile atomic_int *object, int operand, -+ memory_order order); -+uint __ovld atomic_fetch_sub_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order); -+int __ovld atomic_fetch_or_explicit(volatile atomic_int *object, int operand, -+ memory_order order); -+uint __ovld atomic_fetch_or_explicit(volatile atomic_uint *object, uint operand, -+ memory_order order); -+int __ovld atomic_fetch_xor_explicit(volatile atomic_int *object, int operand, -+ memory_order order); -+uint __ovld atomic_fetch_xor_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order); -+int __ovld atomic_fetch_and_explicit(volatile atomic_int *object, int operand, -+ memory_order order); -+uint __ovld atomic_fetch_and_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order); -+int __ovld atomic_fetch_min_explicit(volatile atomic_int *object, int operand, -+ memory_order order); -+uint __ovld atomic_fetch_min_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order); -+int __ovld atomic_fetch_max_explicit(volatile atomic_int *object, int operand, -+ memory_order order); -+uint __ovld atomic_fetch_max_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+int __ovld atomic_fetch_add_explicit(volatile atomic_int *object, int operand, -+ memory_order order, memory_scope scope); -+uint __ovld atomic_fetch_add_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_sub_explicit(volatile atomic_int *object, int operand, -+ memory_order order, memory_scope scope); -+uint __ovld atomic_fetch_sub_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_or_explicit(volatile atomic_int *object, int operand, -+ memory_order order, memory_scope scope); -+uint __ovld atomic_fetch_or_explicit(volatile atomic_uint *object, uint operand, -+ memory_order order, memory_scope scope); -+int __ovld atomic_fetch_xor_explicit(volatile atomic_int *object, int operand, -+ memory_order order, memory_scope scope); -+uint __ovld atomic_fetch_xor_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_and_explicit(volatile atomic_int *object, int operand, -+ memory_order order, memory_scope scope); -+uint __ovld atomic_fetch_and_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_min_explicit(volatile atomic_int *object, int operand, -+ memory_order order, memory_scope scope); -+uint __ovld atomic_fetch_min_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_max_explicit(volatile atomic_int *object, int operand, -+ memory_order order, memory_scope scope); -+uint __ovld atomic_fetch_max_explicit(volatile atomic_uint *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+ -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#ifdef __opencl_c_atomic_scope_device -+long __ovld atomic_fetch_add_explicit(volatile atomic_long *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_add_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_sub_explicit(volatile atomic_long *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_sub_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_or_explicit(volatile atomic_long *object, long operand, -+ memory_order order); -+ulong __ovld atomic_fetch_or_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_xor_explicit(volatile atomic_long *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_xor_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_and_explicit(volatile atomic_long *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_and_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_min_explicit(volatile atomic_long *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_min_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_max_explicit(volatile atomic_long *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_max_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+long __ovld atomic_fetch_add_explicit(volatile atomic_long *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_add_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_sub_explicit(volatile atomic_long *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_sub_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_or_explicit(volatile atomic_long *object, long operand, -+ memory_order order, memory_scope scope); -+ulong __ovld atomic_fetch_or_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_xor_explicit(volatile atomic_long *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_xor_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_and_explicit(volatile atomic_long *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_and_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_min_explicit(volatile atomic_long *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_min_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_max_explicit(volatile atomic_long *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_max_explicit(volatile atomic_ulong *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+ -+// OpenCL v2.0 s6.13.11.7.5: -+// add/sub: atomic type argument can be uintptr_t/intptr_t, value type argument -+// can be ptrdiff_t. or/xor/and/min/max: atomic type argument can be -+// intptr_t/uintptr_t, value type argument can be intptr_t/uintptr_t. -+ -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#ifdef __opencl_c_atomic_scope_device -+uintptr_t __ovld atomic_fetch_add_explicit(volatile atomic_uintptr_t *object, -+ ptrdiff_t operand, -+ memory_order order); -+uintptr_t __ovld atomic_fetch_sub_explicit(volatile atomic_uintptr_t *object, -+ ptrdiff_t operand, -+ memory_order order); -+uintptr_t __ovld atomic_fetch_or_explicit(volatile atomic_uintptr_t *object, -+ intptr_t operand, memory_order order); -+uintptr_t __ovld atomic_fetch_xor_explicit(volatile atomic_uintptr_t *object, -+ intptr_t operand, -+ memory_order order); -+uintptr_t __ovld atomic_fetch_and_explicit(volatile atomic_uintptr_t *object, -+ intptr_t operand, -+ memory_order order); -+uintptr_t __ovld atomic_fetch_min_explicit(volatile atomic_uintptr_t *object, -+ intptr_t opermax, -+ memory_order minder); -+uintptr_t __ovld atomic_fetch_max_explicit(volatile atomic_uintptr_t *object, -+ intptr_t opermax, -+ memory_order minder); -+intptr_t __ovld atomic_fetch_or_explicit(volatile atomic_intptr_t *object, -+ uintptr_t operand, memory_order order); -+intptr_t __ovld atomic_fetch_xor_explicit(volatile atomic_intptr_t *object, -+ uintptr_t operand, -+ memory_order order); -+intptr_t __ovld atomic_fetch_and_explicit(volatile atomic_intptr_t *object, -+ uintptr_t operand, -+ memory_order order); -+intptr_t __ovld atomic_fetch_min_explicit(volatile atomic_intptr_t *object, -+ uintptr_t opermax, -+ memory_order minder); -+intptr_t __ovld atomic_fetch_max_explicit(volatile atomic_intptr_t *object, -+ uintptr_t opermax, -+ memory_order minder); -+#endif // __opencl_c_atomic_scope_device -+uintptr_t __ovld atomic_fetch_add_explicit(volatile atomic_uintptr_t *object, -+ ptrdiff_t operand, -+ memory_order order, -+ memory_scope scope); -+uintptr_t __ovld atomic_fetch_sub_explicit(volatile atomic_uintptr_t *object, -+ ptrdiff_t operand, -+ memory_order order, -+ memory_scope scope); -+ -+uintptr_t __ovld atomic_fetch_or_explicit(volatile atomic_uintptr_t *object, -+ intptr_t operand, memory_order order, -+ memory_scope scope); -+uintptr_t __ovld atomic_fetch_xor_explicit(volatile atomic_uintptr_t *object, -+ intptr_t operand, memory_order order, -+ memory_scope scope); -+uintptr_t __ovld atomic_fetch_and_explicit(volatile atomic_uintptr_t *object, -+ intptr_t operand, memory_order order, -+ memory_scope scope); -+uintptr_t __ovld atomic_fetch_min_explicit(volatile atomic_uintptr_t *object, -+ intptr_t opermax, -+ memory_order minder, -+ memory_scope scope); -+uintptr_t __ovld atomic_fetch_max_explicit(volatile atomic_uintptr_t *object, -+ intptr_t opermax, -+ memory_order minder, -+ memory_scope scope); -+ -+intptr_t __ovld atomic_fetch_or_explicit(volatile atomic_intptr_t *object, -+ uintptr_t operand, memory_order order, -+ memory_scope scope); -+intptr_t __ovld atomic_fetch_xor_explicit(volatile atomic_intptr_t *object, -+ uintptr_t operand, memory_order order, -+ memory_scope scope); -+intptr_t __ovld atomic_fetch_and_explicit(volatile atomic_intptr_t *object, -+ uintptr_t operand, memory_order order, -+ memory_scope scope); -+intptr_t __ovld atomic_fetch_min_explicit(volatile atomic_intptr_t *object, -+ uintptr_t opermax, -+ memory_order minder, -+ memory_scope scope); -+intptr_t __ovld atomic_fetch_max_explicit(volatile atomic_intptr_t *object, -+ uintptr_t opermax, -+ memory_order minder, -+ memory_scope scope); - #endif -+#endif // __opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+#ifdef __opencl_c_atomic_scope_device -+int __ovld atomic_fetch_add_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order); -+int __ovld atomic_fetch_add_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order); -+uint __ovld atomic_fetch_add_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order); -+uint __ovld atomic_fetch_add_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order); -+int __ovld atomic_fetch_sub_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order); -+int __ovld atomic_fetch_sub_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order); -+uint __ovld atomic_fetch_sub_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order); -+uint __ovld atomic_fetch_sub_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order); -+int __ovld atomic_fetch_or_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order); -+int __ovld atomic_fetch_or_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order); -+uint __ovld atomic_fetch_or_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order); -+uint __ovld atomic_fetch_or_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order); -+int __ovld atomic_fetch_xor_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order); -+int __ovld atomic_fetch_xor_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order); -+uint __ovld atomic_fetch_xor_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order); -+uint __ovld atomic_fetch_xor_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order); -+int __ovld atomic_fetch_and_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order); -+int __ovld atomic_fetch_and_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order); -+uint __ovld atomic_fetch_and_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order); -+uint __ovld atomic_fetch_and_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order); -+int __ovld atomic_fetch_min_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order); -+int __ovld atomic_fetch_min_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order); -+uint __ovld atomic_fetch_min_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order); -+uint __ovld atomic_fetch_min_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order); -+int __ovld atomic_fetch_max_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order); -+int __ovld atomic_fetch_max_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order); -+uint __ovld atomic_fetch_max_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order); -+uint __ovld atomic_fetch_max_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+int __ovld atomic_fetch_add_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_add_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_add_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_add_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_sub_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_sub_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_sub_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_sub_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_or_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_or_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_or_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_or_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_xor_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_xor_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_xor_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_xor_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_and_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_and_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_and_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_and_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_min_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_min_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_min_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_min_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_max_explicit(volatile atomic_int __global *object, -+ int operand, memory_order order, -+ memory_scope scope); -+int __ovld atomic_fetch_max_explicit(volatile atomic_int __local *object, -+ int operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_max_explicit(volatile atomic_uint __global *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_fetch_max_explicit(volatile atomic_uint __local *object, -+ uint operand, memory_order order, -+ memory_scope scope); -+ -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#ifdef __opencl_c_atomic_scope_device -+long __ovld atomic_fetch_add_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order); -+long __ovld atomic_fetch_add_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_add_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order); -+ulong __ovld atomic_fetch_add_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_sub_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order); -+long __ovld atomic_fetch_sub_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_sub_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order); -+ulong __ovld atomic_fetch_sub_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_or_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order); -+long __ovld atomic_fetch_or_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_or_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order); -+ulong __ovld atomic_fetch_or_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_xor_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order); -+long __ovld atomic_fetch_xor_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_xor_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order); -+ulong __ovld atomic_fetch_xor_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_and_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order); -+long __ovld atomic_fetch_and_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_and_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order); -+ulong __ovld atomic_fetch_and_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_min_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order); -+long __ovld atomic_fetch_min_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_min_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order); -+ulong __ovld atomic_fetch_min_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order); -+long __ovld atomic_fetch_max_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order); -+long __ovld atomic_fetch_max_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order); -+ulong __ovld atomic_fetch_max_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order); -+ulong __ovld atomic_fetch_max_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+long __ovld atomic_fetch_add_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_add_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_add_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_add_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_sub_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_sub_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_sub_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_sub_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_or_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_or_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_or_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_or_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_xor_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_xor_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_xor_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_xor_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_and_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_and_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_and_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_and_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_min_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_min_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_min_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_min_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_max_explicit(volatile atomic_long __global *object, -+ long operand, memory_order order, -+ memory_scope scope); -+long __ovld atomic_fetch_max_explicit(volatile atomic_long __local *object, -+ long operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_max_explicit(volatile atomic_ulong __global *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_fetch_max_explicit(volatile atomic_ulong __local *object, -+ ulong operand, memory_order order, -+ memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+ -+// OpenCL v2.0 s6.13.11.7.5: -+// add/sub: atomic type argument can be uintptr_t/intptr_t, value type argument -+// can be ptrdiff_t. or/xor/and/min/max: atomic type argument can be -+// intptr_t/uintptr_t, value type argument can be intptr_t/uintptr_t. -+ -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#ifdef __opencl_c_atomic_scope_device -+uintptr_t __ovld -+atomic_fetch_add_explicit(volatile atomic_uintptr_t __global *object, -+ ptrdiff_t operand, memory_order order); -+uintptr_t __ovld -+atomic_fetch_add_explicit(volatile atomic_uintptr_t __local *object, -+ ptrdiff_t operand, memory_order order); -+uintptr_t __ovld -+atomic_fetch_sub_explicit(volatile atomic_uintptr_t __global *object, -+ ptrdiff_t operand, memory_order order); -+uintptr_t __ovld -+atomic_fetch_sub_explicit(volatile atomic_uintptr_t __local *object, -+ ptrdiff_t operand, memory_order order); -+uintptr_t __ovld -+atomic_fetch_or_explicit(volatile atomic_uintptr_t __global *object, -+ intptr_t operand, memory_order order); -+uintptr_t __ovld -+atomic_fetch_or_explicit(volatile atomic_uintptr_t __local *object, -+ intptr_t operand, memory_order order); -+uintptr_t __ovld -+atomic_fetch_xor_explicit(volatile atomic_uintptr_t __global *object, -+ intptr_t operand, memory_order order); -+uintptr_t __ovld -+atomic_fetch_xor_explicit(volatile atomic_uintptr_t __local *object, -+ intptr_t operand, memory_order order); -+uintptr_t __ovld -+atomic_fetch_and_explicit(volatile atomic_uintptr_t __global *object, -+ intptr_t operand, memory_order order); -+uintptr_t __ovld -+atomic_fetch_and_explicit(volatile atomic_uintptr_t __local *object, -+ intptr_t operand, memory_order order); -+uintptr_t __ovld -+atomic_fetch_min_explicit(volatile atomic_uintptr_t __global *object, -+ intptr_t opermax, memory_order minder); -+uintptr_t __ovld -+atomic_fetch_min_explicit(volatile atomic_uintptr_t __local *object, -+ intptr_t opermax, memory_order minder); -+uintptr_t __ovld -+atomic_fetch_max_explicit(volatile atomic_uintptr_t __global *object, -+ intptr_t opermax, memory_order minder); -+uintptr_t __ovld -+atomic_fetch_max_explicit(volatile atomic_uintptr_t __local *object, -+ intptr_t opermax, memory_order minder); -+intptr_t __ovld -+atomic_fetch_or_explicit(volatile atomic_intptr_t __global *object, -+ uintptr_t operand, memory_order order); -+intptr_t __ovld -+atomic_fetch_or_explicit(volatile atomic_intptr_t __local *object, -+ uintptr_t operand, memory_order order); -+intptr_t __ovld -+atomic_fetch_xor_explicit(volatile atomic_intptr_t __global *object, -+ uintptr_t operand, memory_order order); -+intptr_t __ovld -+atomic_fetch_xor_explicit(volatile atomic_intptr_t __local *object, -+ uintptr_t operand, memory_order order); -+intptr_t __ovld -+atomic_fetch_and_explicit(volatile atomic_intptr_t __global *object, -+ uintptr_t operand, memory_order order); -+intptr_t __ovld -+atomic_fetch_and_explicit(volatile atomic_intptr_t __local *object, -+ uintptr_t operand, memory_order order); -+intptr_t __ovld -+atomic_fetch_min_explicit(volatile atomic_intptr_t __global *object, -+ uintptr_t opermax, memory_order minder); -+intptr_t __ovld -+atomic_fetch_min_explicit(volatile atomic_intptr_t __local *object, -+ uintptr_t opermax, memory_order minder); -+intptr_t __ovld -+atomic_fetch_max_explicit(volatile atomic_intptr_t __global *object, -+ uintptr_t opermax, memory_order minder); -+intptr_t __ovld -+atomic_fetch_max_explicit(volatile atomic_intptr_t __local *object, -+ uintptr_t opermax, memory_order minder); -+#endif // __opencl_c_atomic_scope_device -+uintptr_t __ovld atomic_fetch_add_explicit( -+ volatile atomic_uintptr_t __global *object, ptrdiff_t operand, -+ memory_order order, memory_scope scope); -+uintptr_t __ovld atomic_fetch_add_explicit( -+ volatile atomic_uintptr_t __local *object, ptrdiff_t operand, -+ memory_order order, memory_scope scope); -+uintptr_t __ovld atomic_fetch_sub_explicit( -+ volatile atomic_uintptr_t __global *object, ptrdiff_t operand, -+ memory_order order, memory_scope scope); -+uintptr_t __ovld atomic_fetch_sub_explicit( -+ volatile atomic_uintptr_t __local *object, ptrdiff_t operand, -+ memory_order order, memory_scope scope); -+ -+uintptr_t __ovld atomic_fetch_or_explicit( -+ volatile atomic_uintptr_t __global *object, intptr_t operand, -+ memory_order order, memory_scope scope); -+uintptr_t __ovld atomic_fetch_or_explicit( -+ volatile atomic_uintptr_t __local *object, intptr_t operand, -+ memory_order order, memory_scope scope); -+uintptr_t __ovld atomic_fetch_xor_explicit( -+ volatile atomic_uintptr_t __global *object, intptr_t operand, -+ memory_order order, memory_scope scope); -+uintptr_t __ovld atomic_fetch_xor_explicit( -+ volatile atomic_uintptr_t __local *object, intptr_t operand, -+ memory_order order, memory_scope scope); -+uintptr_t __ovld atomic_fetch_and_explicit( -+ volatile atomic_uintptr_t __global *object, intptr_t operand, -+ memory_order order, memory_scope scope); -+uintptr_t __ovld atomic_fetch_and_explicit( -+ volatile atomic_uintptr_t __local *object, intptr_t operand, -+ memory_order order, memory_scope scope); -+uintptr_t __ovld atomic_fetch_min_explicit( -+ volatile atomic_uintptr_t __global *object, intptr_t opermax, -+ memory_order minder, memory_scope scope); -+uintptr_t __ovld atomic_fetch_min_explicit( -+ volatile atomic_uintptr_t __local *object, intptr_t opermax, -+ memory_order minder, memory_scope scope); -+uintptr_t __ovld atomic_fetch_max_explicit( -+ volatile atomic_uintptr_t __global *object, intptr_t opermax, -+ memory_order minder, memory_scope scope); -+uintptr_t __ovld atomic_fetch_max_explicit( -+ volatile atomic_uintptr_t __local *object, intptr_t opermax, -+ memory_order minder, memory_scope scope); -+ -+intptr_t __ovld atomic_fetch_or_explicit( -+ volatile atomic_intptr_t __global *object, uintptr_t operand, -+ memory_order order, memory_scope scope); -+intptr_t __ovld atomic_fetch_or_explicit( -+ volatile atomic_intptr_t __local *object, uintptr_t operand, -+ memory_order order, memory_scope scope); -+intptr_t __ovld atomic_fetch_xor_explicit( -+ volatile atomic_intptr_t __global *object, uintptr_t operand, -+ memory_order order, memory_scope scope); -+intptr_t __ovld atomic_fetch_xor_explicit( -+ volatile atomic_intptr_t __local *object, uintptr_t operand, -+ memory_order order, memory_scope scope); -+intptr_t __ovld atomic_fetch_and_explicit( -+ volatile atomic_intptr_t __global *object, uintptr_t operand, -+ memory_order order, memory_scope scope); -+intptr_t __ovld atomic_fetch_and_explicit( -+ volatile atomic_intptr_t __local *object, uintptr_t operand, -+ memory_order order, memory_scope scope); -+intptr_t __ovld atomic_fetch_min_explicit( -+ volatile atomic_intptr_t __global *object, uintptr_t opermax, -+ memory_order minder, memory_scope scope); -+intptr_t __ovld atomic_fetch_min_explicit( -+ volatile atomic_intptr_t __local *object, uintptr_t opermax, -+ memory_order minder, memory_scope scope); -+intptr_t __ovld atomic_fetch_max_explicit( -+ volatile atomic_intptr_t __global *object, uintptr_t opermax, -+ memory_order minder, memory_scope scope); -+intptr_t __ovld atomic_fetch_max_explicit( -+ volatile atomic_intptr_t __local *object, uintptr_t opermax, -+ memory_order minder, memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) - - // atomic_store() - -+#if defined(__opencl_c_atomic_scope_device) && \ -+ defined(__opencl_c_atomic_order_seq_cst) -+ -+#ifdef __opencl_c_generic_address_space - void __ovld atomic_store(volatile atomic_int *object, int desired); --void __ovld atomic_store_explicit(volatile atomic_int *object, int desired, memory_order order); --void __ovld atomic_store_explicit(volatile atomic_int *object, int desired, memory_order order, memory_scope scope); - void __ovld atomic_store(volatile atomic_uint *object, uint desired); --void __ovld atomic_store_explicit(volatile atomic_uint *object, uint desired, memory_order order); --void __ovld atomic_store_explicit(volatile atomic_uint *object, uint desired, memory_order order, memory_scope scope); - void __ovld atomic_store(volatile atomic_float *object, float desired); --void __ovld atomic_store_explicit(volatile atomic_float *object, float desired, memory_order order); --void __ovld atomic_store_explicit(volatile atomic_float *object, float desired, memory_order order, memory_scope scope); - #if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - void __ovld atomic_store(volatile atomic_double *object, double desired); --void __ovld atomic_store_explicit(volatile atomic_double *object, double desired, memory_order order); --void __ovld atomic_store_explicit(volatile atomic_double *object, double desired, memory_order order, memory_scope scope); --#endif //cl_khr_fp64 -+#endif - void __ovld atomic_store(volatile atomic_long *object, long desired); --void __ovld atomic_store_explicit(volatile atomic_long *object, long desired, memory_order order); --void __ovld atomic_store_explicit(volatile atomic_long *object, long desired, memory_order order, memory_scope scope); - void __ovld atomic_store(volatile atomic_ulong *object, ulong desired); --void __ovld atomic_store_explicit(volatile atomic_ulong *object, ulong desired, memory_order order); --void __ovld atomic_store_explicit(volatile atomic_ulong *object, ulong desired, memory_order order, memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // __opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+void __ovld atomic_store(volatile atomic_int __global *object, int desired); -+void __ovld atomic_store(volatile atomic_int __local *object, int desired); -+void __ovld atomic_store(volatile atomic_uint __global *object, uint desired); -+void __ovld atomic_store(volatile atomic_uint __local *object, uint desired); -+void __ovld atomic_store(volatile atomic_float __global *object, float desired); -+void __ovld atomic_store(volatile atomic_float __local *object, float desired); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+void __ovld atomic_store(volatile atomic_double __global *object, -+ double desired); -+void __ovld atomic_store(volatile atomic_double __local *object, -+ double desired); - #endif -+void __ovld atomic_store(volatile atomic_long __global *object, long desired); -+void __ovld atomic_store(volatile atomic_long __local *object, long desired); -+void __ovld atomic_store(volatile atomic_ulong __global *object, ulong desired); -+void __ovld atomic_store(volatile atomic_ulong __local *object, ulong desired); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+ -+#endif // defined(__opencl_c_atomic_scope_device) && -+ // defined(__opencl_c_atomic_order_seq_cst) -+ -+#ifdef __opencl_c_generic_address_space -+#ifdef __opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_int *object, int desired, -+ memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_uint *object, uint desired, -+ memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_float *object, float desired, -+ memory_order order); -+#endif // __opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_int *object, int desired, -+ memory_order order, memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_uint *object, uint desired, -+ memory_order order, memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_float *object, float desired, -+ memory_order order, memory_scope scope); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_double *object, -+ double desired, memory_order order); -+#endif //__opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_double *object, -+ double desired, memory_order order, -+ memory_scope scope); -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_long *object, long desired, -+ memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_ulong *object, ulong desired, -+ memory_order order); -+#endif //__opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_long *object, long desired, -+ memory_order order, memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_ulong *object, ulong desired, -+ memory_order order, memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // __opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+#ifdef __opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_int __global *object, -+ int desired, memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_int __local *object, -+ int desired, memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_uint __global *object, -+ uint desired, memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_uint __local *object, -+ uint desired, memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_float __global *object, -+ float desired, memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_float __local *object, -+ float desired, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_int __global *object, -+ int desired, memory_order order, -+ memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_int __local *object, -+ int desired, memory_order order, -+ memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_uint __global *object, -+ uint desired, memory_order order, -+ memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_uint __local *object, -+ uint desired, memory_order order, -+ memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_float __global *object, -+ float desired, memory_order order, -+ memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_float __local *object, -+ float desired, memory_order order, -+ memory_scope scope); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_double __global *object, -+ double desired, memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_double __local *object, -+ double desired, memory_order order); -+#endif //__opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_double __global *object, -+ double desired, memory_order order, -+ memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_double __local *object, -+ double desired, memory_order order, -+ memory_scope scope); -+#endif // cl_khr_fp64 -+#ifdef __opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_long __global *object, -+ long desired, memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_long __local *object, -+ long desired, memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_ulong __global *object, -+ ulong desired, memory_order order); -+void __ovld atomic_store_explicit(volatile atomic_ulong __local *object, -+ ulong desired, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+void __ovld atomic_store_explicit(volatile atomic_long __global *object, -+ long desired, memory_order order, -+ memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_long __local *object, -+ long desired, memory_order order, -+ memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_ulong __global *object, -+ ulong desired, memory_order order, -+ memory_scope scope); -+void __ovld atomic_store_explicit(volatile atomic_ulong __local *object, -+ ulong desired, memory_order order, -+ memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) - - // atomic_load() -- -+#ifdef __opencl_c_generic_address_space -+#if defined(__opencl_c_atomic_scope_device) && \ -+ defined(__opencl_c_atomic_order_seq_cst) - int __ovld atomic_load(volatile atomic_int *object); --int __ovld atomic_load_explicit(volatile atomic_int *object, memory_order order); --int __ovld atomic_load_explicit(volatile atomic_int *object, memory_order order, memory_scope scope); - uint __ovld atomic_load(volatile atomic_uint *object); --uint __ovld atomic_load_explicit(volatile atomic_uint *object, memory_order order); --uint __ovld atomic_load_explicit(volatile atomic_uint *object, memory_order order, memory_scope scope); - float __ovld atomic_load(volatile atomic_float *object); --float __ovld atomic_load_explicit(volatile atomic_float *object, memory_order order); --float __ovld atomic_load_explicit(volatile atomic_float *object, memory_order order, memory_scope scope); - #if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld atomic_load(volatile atomic_double *object); --double __ovld atomic_load_explicit(volatile atomic_double *object, memory_order order); --double __ovld atomic_load_explicit(volatile atomic_double *object, memory_order order, memory_scope scope); --#endif //cl_khr_fp64 -+#endif // cl_khr_fp64 - long __ovld atomic_load(volatile atomic_long *object); --long __ovld atomic_load_explicit(volatile atomic_long *object, memory_order order); --long __ovld atomic_load_explicit(volatile atomic_long *object, memory_order order, memory_scope scope); - ulong __ovld atomic_load(volatile atomic_ulong *object); --ulong __ovld atomic_load_explicit(volatile atomic_ulong *object, memory_order order); --ulong __ovld atomic_load_explicit(volatile atomic_ulong *object, memory_order order, memory_scope scope); --#endif -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // defined(__opencl_c_atomic_scope_device) && -+ // defined(__opencl_c_atomic_order_seq_cst) -+#endif //__opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+#if defined(__opencl_c_atomic_scope_device) && \ -+ defined(__opencl_c_atomic_order_seq_cst) -+int __ovld atomic_load(volatile atomic_int __global *object); -+int __ovld atomic_load(volatile atomic_int __local *object); -+uint __ovld atomic_load(volatile atomic_uint __global *object); -+uint __ovld atomic_load(volatile atomic_uint __local *object); -+float __ovld atomic_load(volatile atomic_float __global *object); -+float __ovld atomic_load(volatile atomic_float __local *object); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+double __ovld atomic_load(volatile atomic_double __global *object); -+double __ovld atomic_load(volatile atomic_double __local *object); -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+long __ovld atomic_load(volatile atomic_long __global *object); -+long __ovld atomic_load(volatile atomic_long __local *object); -+ulong __ovld atomic_load(volatile atomic_ulong __global *object); -+ulong __ovld atomic_load(volatile atomic_ulong __local *object); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // defined(__opencl_c_atomic_scope_device) && -+ // defined(__opencl_c_atomic_order_seq_cst) -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+ -+#ifdef __opencl_c_generic_address_space -+#ifdef __opencl_c_atomic_scope_device -+int __ovld atomic_load_explicit(volatile atomic_int *object, -+ memory_order order); -+uint __ovld atomic_load_explicit(volatile atomic_uint *object, -+ memory_order order); -+float __ovld atomic_load_explicit(volatile atomic_float *object, -+ memory_order order); -+#endif // __opencl_c_atomic_scope_device -+int __ovld atomic_load_explicit(volatile atomic_int *object, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_load_explicit(volatile atomic_uint *object, -+ memory_order order, memory_scope scope); -+float __ovld atomic_load_explicit(volatile atomic_float *object, -+ memory_order order, memory_scope scope); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+double __ovld atomic_load_explicit(volatile atomic_double *object, -+ memory_order order); -+#endif // __opencl_c_atomic_scope_device -+double __ovld atomic_load_explicit(volatile atomic_double *object, -+ memory_order order, memory_scope scope); -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+long __ovld atomic_load_explicit(volatile atomic_long *object, -+ memory_order order); -+ulong __ovld atomic_load_explicit(volatile atomic_ulong *object, -+ memory_order order); -+#endif //__opencl_c_atomic_scope_device -+long __ovld atomic_load_explicit(volatile atomic_long *object, -+ memory_order order, memory_scope scope); -+ulong __ovld atomic_load_explicit(volatile atomic_ulong *object, -+ memory_order order, memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // __opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+#ifdef __opencl_c_atomic_scope_device -+int __ovld atomic_load_explicit(volatile atomic_int __global *object, -+ memory_order order); -+int __ovld atomic_load_explicit(volatile atomic_int __local *object, -+ memory_order order); -+uint __ovld atomic_load_explicit(volatile atomic_uint __global *object, -+ memory_order order); -+uint __ovld atomic_load_explicit(volatile atomic_uint __local *object, -+ memory_order order); -+float __ovld atomic_load_explicit(volatile atomic_float __global *object, -+ memory_order order); -+float __ovld atomic_load_explicit(volatile atomic_float __local *object, -+ memory_order order); -+#endif // __opencl_c_atomic_scope_device -+int __ovld atomic_load_explicit(volatile atomic_int __global *object, -+ memory_order order, memory_scope scope); -+int __ovld atomic_load_explicit(volatile atomic_int __local *object, -+ memory_order order, memory_scope scope); -+uint __ovld atomic_load_explicit(volatile atomic_uint __global *object, -+ memory_order order, memory_scope scope); -+uint __ovld atomic_load_explicit(volatile atomic_uint __local *object, -+ memory_order order, memory_scope scope); -+float __ovld atomic_load_explicit(volatile atomic_float __global *object, -+ memory_order order, memory_scope scope); -+float __ovld atomic_load_explicit(volatile atomic_float __local *object, -+ memory_order order, memory_scope scope); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+double __ovld atomic_load_explicit(volatile atomic_double __global *object, -+ memory_order order); -+double __ovld atomic_load_explicit(volatile atomic_double __local *object, -+ memory_order order); -+#endif // __opencl_c_atomic_scope_device -+double __ovld atomic_load_explicit(volatile atomic_double __global *object, -+ memory_order order, memory_scope scope); -+double __ovld atomic_load_explicit(volatile atomic_double __local *object, -+ memory_order order, memory_scope scope); -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+long __ovld atomic_load_explicit(volatile atomic_long __global *object, -+ memory_order order); -+long __ovld atomic_load_explicit(volatile atomic_long __local *object, -+ memory_order order); -+ulong __ovld atomic_load_explicit(volatile atomic_ulong __global *object, -+ memory_order order); -+ulong __ovld atomic_load_explicit(volatile atomic_ulong __local *object, -+ memory_order order); -+#endif // __opencl_c_atomic_scope_device -+long __ovld atomic_load_explicit(volatile atomic_long __global *object, -+ memory_order order, memory_scope scope); -+long __ovld atomic_load_explicit(volatile atomic_long __local *object, -+ memory_order order, memory_scope scope); -+ulong __ovld atomic_load_explicit(volatile atomic_ulong __global *object, -+ memory_order order, memory_scope scope); -+ulong __ovld atomic_load_explicit(volatile atomic_ulong __local *object, -+ memory_order order, memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) - - // atomic_exchange() -- -+#ifdef __opencl_c_generic_address_space -+#if defined(__opencl_c_atomic_scope_device) && \ -+ defined(__opencl_c_atomic_order_seq_cst) - int __ovld atomic_exchange(volatile atomic_int *object, int desired); --int __ovld atomic_exchange_explicit(volatile atomic_int *object, int desired, memory_order order); --int __ovld atomic_exchange_explicit(volatile atomic_int *object, int desired, memory_order order, memory_scope scope); - uint __ovld atomic_exchange(volatile atomic_uint *object, uint desired); --uint __ovld atomic_exchange_explicit(volatile atomic_uint *object, uint desired, memory_order order); --uint __ovld atomic_exchange_explicit(volatile atomic_uint *object, uint desired, memory_order order, memory_scope scope); - float __ovld atomic_exchange(volatile atomic_float *object, float desired); --float __ovld atomic_exchange_explicit(volatile atomic_float *object, float desired, memory_order order); --float __ovld atomic_exchange_explicit(volatile atomic_float *object, float desired, memory_order order, memory_scope scope); - #if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld atomic_exchange(volatile atomic_double *object, double desired); --double __ovld atomic_exchange_explicit(volatile atomic_double *object, double desired, memory_order order); --double __ovld atomic_exchange_explicit(volatile atomic_double *object, double desired, memory_order order, memory_scope scope); --#endif //cl_khr_fp64 -+#endif - long __ovld atomic_exchange(volatile atomic_long *object, long desired); --long __ovld atomic_exchange_explicit(volatile atomic_long *object, long desired, memory_order order); --long __ovld atomic_exchange_explicit(volatile atomic_long *object, long desired, memory_order order, memory_scope scope); - ulong __ovld atomic_exchange(volatile atomic_ulong *object, ulong desired); --ulong __ovld atomic_exchange_explicit(volatile atomic_ulong *object, ulong desired, memory_order order); --ulong __ovld atomic_exchange_explicit(volatile atomic_ulong *object, ulong desired, memory_order order, memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // defined(__opencl_c_atomic_scope_device) && -+ // defined(__opencl_c_atomic_order_seq_cst) -+#endif // __opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+#if defined(__opencl_c_atomic_scope_device) && \ -+ defined(__opencl_c_atomic_order_seq_cst) -+int __ovld atomic_exchange(volatile atomic_int __global *object, int desired); -+int __ovld atomic_exchange(volatile atomic_int __local *object, int desired); -+uint __ovld atomic_exchange(volatile atomic_uint __global *object, -+ uint desired); -+uint __ovld atomic_exchange(volatile atomic_uint __local *object, uint desired); -+float __ovld atomic_exchange(volatile atomic_float __global *object, -+ float desired); -+float __ovld atomic_exchange(volatile atomic_float __local *object, -+ float desired); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+double __ovld atomic_exchange(volatile atomic_double __global *object, -+ double desired); -+double __ovld atomic_exchange(volatile atomic_double __local *object, -+ double desired); - #endif -+long __ovld atomic_exchange(volatile atomic_long __global *object, -+ long desired); -+long __ovld atomic_exchange(volatile atomic_long __local *object, long desired); -+ulong __ovld atomic_exchange(volatile atomic_ulong __global *object, -+ ulong desired); -+ulong __ovld atomic_exchange(volatile atomic_ulong __local *object, -+ ulong desired); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // defined(__opencl_c_atomic_scope_device) && -+ // defined(__opencl_c_atomic_order_seq_cst) -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+ -+#ifdef __opencl_c_generic_address_space -+#ifdef __opencl_c_atomic_scope_device -+int __ovld atomic_exchange_explicit(volatile atomic_int *object, int desired, -+ memory_order order); -+uint __ovld atomic_exchange_explicit(volatile atomic_uint *object, uint desired, -+ memory_order order); -+float __ovld atomic_exchange_explicit(volatile atomic_float *object, -+ float desired, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+int __ovld atomic_exchange_explicit(volatile atomic_int *object, int desired, -+ memory_order order, memory_scope scope); -+uint __ovld atomic_exchange_explicit(volatile atomic_uint *object, uint desired, -+ memory_order order, memory_scope scope); -+float __ovld atomic_exchange_explicit(volatile atomic_float *object, -+ float desired, memory_order order, -+ memory_scope scope); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+double __ovld atomic_exchange_explicit(volatile atomic_double *object, -+ double desired, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+double __ovld atomic_exchange_explicit(volatile atomic_double *object, -+ double desired, memory_order order, -+ memory_scope scope); -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+long __ovld atomic_exchange_explicit(volatile atomic_long *object, long desired, -+ memory_order order); -+ulong __ovld atomic_exchange_explicit(volatile atomic_ulong *object, -+ ulong desired, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+long __ovld atomic_exchange_explicit(volatile atomic_long *object, long desired, -+ memory_order order, memory_scope scope); -+ulong __ovld atomic_exchange_explicit(volatile atomic_ulong *object, -+ ulong desired, memory_order order, -+ memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // __opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+#ifdef __opencl_c_atomic_scope_device -+int __ovld atomic_exchange_explicit(volatile atomic_int __global *object, -+ int desired, memory_order order); -+int __ovld atomic_exchange_explicit(volatile atomic_int __local *object, -+ int desired, memory_order order); -+uint __ovld atomic_exchange_explicit(volatile atomic_uint __global *object, -+ uint desired, memory_order order); -+uint __ovld atomic_exchange_explicit(volatile atomic_uint __local *object, -+ uint desired, memory_order order); -+float __ovld atomic_exchange_explicit(volatile atomic_float __global *object, -+ float desired, memory_order order); -+float __ovld atomic_exchange_explicit(volatile atomic_float __local *object, -+ float desired, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+int __ovld atomic_exchange_explicit(volatile atomic_int __global *object, -+ int desired, memory_order order, -+ memory_scope scope); -+int __ovld atomic_exchange_explicit(volatile atomic_int __local *object, -+ int desired, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_exchange_explicit(volatile atomic_uint __global *object, -+ uint desired, memory_order order, -+ memory_scope scope); -+uint __ovld atomic_exchange_explicit(volatile atomic_uint __local *object, -+ uint desired, memory_order order, -+ memory_scope scope); -+float __ovld atomic_exchange_explicit(volatile atomic_float __global *object, -+ float desired, memory_order order, -+ memory_scope scope); -+float __ovld atomic_exchange_explicit(volatile atomic_float __local *object, -+ float desired, memory_order order, -+ memory_scope scope); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+double __ovld atomic_exchange_explicit(volatile atomic_double __global *object, -+ double desired, memory_order order); -+double __ovld atomic_exchange_explicit(volatile atomic_double __local *object, -+ double desired, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+double __ovld atomic_exchange_explicit(volatile atomic_double __global *object, -+ double desired, memory_order order, -+ memory_scope scope); -+double __ovld atomic_exchange_explicit(volatile atomic_double __local *object, -+ double desired, memory_order order, -+ memory_scope scope); -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+long __ovld atomic_exchange_explicit(volatile atomic_long __global *object, -+ long desired, memory_order order); -+long __ovld atomic_exchange_explicit(volatile atomic_long __local *object, -+ long desired, memory_order order); -+ulong __ovld atomic_exchange_explicit(volatile atomic_ulong __global *object, -+ ulong desired, memory_order order); -+ulong __ovld atomic_exchange_explicit(volatile atomic_ulong __local *object, -+ ulong desired, memory_order order); -+#endif // __opencl_c_atomic_scope_device -+long __ovld atomic_exchange_explicit(volatile atomic_long __global *object, -+ long desired, memory_order order, -+ memory_scope scope); -+long __ovld atomic_exchange_explicit(volatile atomic_long __local *object, -+ long desired, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_exchange_explicit(volatile atomic_ulong __global *object, -+ ulong desired, memory_order order, -+ memory_scope scope); -+ulong __ovld atomic_exchange_explicit(volatile atomic_ulong __local *object, -+ ulong desired, memory_order order, -+ memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) - - // atomic_compare_exchange_strong() and atomic_compare_exchange_weak() -- --bool __ovld atomic_compare_exchange_strong(volatile atomic_int *object, int *expected, int desired); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_int *object, int *expected, -- int desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_int *object, int *expected, -- int desired, memory_order success, memory_order failure, memory_scope scope); --bool __ovld atomic_compare_exchange_strong(volatile atomic_uint *object, uint *expected, uint desired); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_uint *object, uint *expected, -- uint desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_uint *object, uint *expected, -- uint desired, memory_order success, memory_order failure, memory_scope scope); --bool __ovld atomic_compare_exchange_weak(volatile atomic_int *object, int *expected, int desired); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_int *object, int *expected, -- int desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_int *object, int *expected, -- int desired, memory_order success, memory_order failure, memory_scope scope); --bool __ovld atomic_compare_exchange_weak(volatile atomic_uint *object, uint *expected, uint desired); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_uint *object, uint *expected, -- uint desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_uint *object, uint *expected, -- uint desired, memory_order success, memory_order failure, memory_scope scope); --bool __ovld atomic_compare_exchange_strong(volatile atomic_float *object, float *expected, float desired); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_float *object, float *expected, -- float desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_float *object, float *expected, -- float desired, memory_order success, memory_order failure, memory_scope scope); --bool __ovld atomic_compare_exchange_weak(volatile atomic_float *object, float *expected, float desired); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_float *object, float *expected, -- float desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_float *object, float *expected, -- float desired, memory_order success, memory_order failure, memory_scope scope); -+#ifdef __opencl_c_generic_address_space -+#if defined(__opencl_c_atomic_scope_device) && \ -+ defined(__opencl_c_atomic_order_seq_cst) -+bool __ovld atomic_compare_exchange_strong(volatile atomic_int *object, -+ int *expected, int desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_uint *object, -+ uint *expected, uint desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_int *object, -+ int *expected, int desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_uint *object, -+ uint *expected, uint desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_float *object, -+ float *expected, float desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_float *object, -+ float *expected, float desired); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+bool __ovld atomic_compare_exchange_strong(volatile atomic_double *object, -+ double *expected, double desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_double *object, -+ double *expected, double desired); -+#endif -+bool __ovld atomic_compare_exchange_strong(volatile atomic_long *object, -+ long *expected, long desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_long *object, -+ long *expected, long desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_ulong *object, -+ ulong *expected, ulong desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_ulong *object, -+ ulong *expected, ulong desired); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // defined(__opencl_c_atomic_scope_device) && -+ // defined(__opencl_c_atomic_order_seq_cst) -+#endif // __opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+#if defined(__opencl_c_atomic_scope_device) && \ -+ defined(__opencl_c_atomic_order_seq_cst) -+bool __ovld atomic_compare_exchange_strong(volatile atomic_int __global *object, -+ int __global *expected, int desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_int __local *object, -+ int __global *expected, int desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_int __global *object, -+ int __local *expected, int desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_int __local *object, -+ int __local *expected, int desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_int __global *object, -+ int __private *expected, -+ int desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_int __local *object, -+ int __private *expected, -+ int desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_uint __global *object, -+ uint __global *expected, uint desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_uint __local *object, -+ uint __global *expected, -+ uint desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_uint __global *object, -+ uint __local *expected, uint desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_uint __local *object, -+ uint __local *expected, -+ uint desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_uint __global *object, -+ uint __private *expected, uint desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_uint __local *object, -+ uint __private *expected, -+ uint desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_int __global *object, -+ int __global *expected, int desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_int __local *object, -+ int __global *expected, int desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_int __global *object, -+ int __local *expected, int desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_int __local *object, -+ int __local *expected, int desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_int __global *object, -+ int __private *expected, int desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_int __local *object, -+ int __private *expected, int desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_uint __global *object, -+ uint __global *expected, uint desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_uint __local *object, -+ uint __global *expected, uint desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_uint __global *object, -+ uint __local *expected, uint desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_uint __local *object, -+ uint __local *expected, uint desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_uint __global *object, -+ uint __private *expected, -+ uint desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_uint __local *object, -+ uint __private *expected, -+ uint desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_float __global *object, -+ float __global *expected, float desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_float __local *object, -+ float __global *expected, float desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_float __global *object, -+ float __local *expected, float desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_float __local *object, -+ float __local *expected, float desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_float __global *object, -+ float __private *expected, float desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_float __local *object, -+ float __private *expected, float desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_float __global *object, -+ float __global *expected, -+ float desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_float __local *object, -+ float __global *expected, -+ float desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_float __global *object, -+ float __local *expected, -+ float desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_float __local *object, -+ float __local *expected, -+ float desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_float __global *object, -+ float __private *expected, -+ float desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_float __local *object, -+ float __private *expected, -+ float desired); - #if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) --#ifdef cl_khr_fp64 --bool __ovld atomic_compare_exchange_strong(volatile atomic_double *object, double *expected, double desired); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_double *object, double *expected, -- double desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_double *object, double *expected, -- double desired, memory_order success, memory_order failure, memory_scope scope); --bool __ovld atomic_compare_exchange_weak(volatile atomic_double *object, double *expected, double desired); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_double *object, double *expected, -- double desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_double *object, double *expected, -- double desired, memory_order success, memory_order failure, memory_scope scope); --#endif //cl_khr_fp64 --bool __ovld atomic_compare_exchange_strong(volatile atomic_long *object, long *expected, long desired); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_long *object, long *expected, -- long desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_long *object, long *expected, -- long desired, memory_order success, memory_order failure, memory_scope scope); --bool __ovld atomic_compare_exchange_weak(volatile atomic_long *object, long *expected, long desired); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_long *object, long *expected, -- long desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_long *object, long *expected, -- long desired, memory_order success, memory_order failure, memory_scope scope); --bool __ovld atomic_compare_exchange_strong(volatile atomic_ulong *object, ulong *expected, ulong desired); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_ulong *object, ulong *expected, -- ulong desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_ulong *object, ulong *expected, -- ulong desired, memory_order success, memory_order failure, memory_scope scope); --bool __ovld atomic_compare_exchange_weak(volatile atomic_ulong *object, ulong *expected, ulong desired); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_ulong *object, ulong *expected, -- ulong desired, memory_order success, memory_order failure); --bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_ulong *object, ulong *expected, -- ulong desired, memory_order success, memory_order failure, memory_scope scope); -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_double __global *object, -+ double __global *expected, double desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_double __local *object, -+ double __global *expected, double desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_double __global *object, -+ double __local *expected, double desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_double __local *object, -+ double __local *expected, double desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_double __global *object, -+ double __private *expected, double desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_double __local *object, -+ double __private *expected, double desired); -+bool __ovld -+atomic_compare_exchange_weak(volatile atomic_double __global *object, -+ double __global *expected, double desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_double __local *object, -+ double __global *expected, -+ double desired); -+bool __ovld -+atomic_compare_exchange_weak(volatile atomic_double __global *object, -+ double __local *expected, double desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_double __local *object, -+ double __local *expected, -+ double desired); -+bool __ovld -+atomic_compare_exchange_weak(volatile atomic_double __global *object, -+ double __private *expected, double desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_double __local *object, -+ double __private *expected, -+ double desired); - #endif -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_long __global *object, -+ long __global *expected, long desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_long __local *object, -+ long __global *expected, -+ long desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_long __global *object, -+ long __local *expected, long desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_long __local *object, -+ long __local *expected, -+ long desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_long __global *object, -+ long __private *expected, long desired); -+bool __ovld atomic_compare_exchange_strong(volatile atomic_long __local *object, -+ long __private *expected, -+ long desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_long __global *object, -+ long __global *expected, long desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_long __local *object, -+ long __global *expected, long desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_long __global *object, -+ long __local *expected, long desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_long __local *object, -+ long __local *expected, long desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_long __global *object, -+ long __private *expected, -+ long desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_long __local *object, -+ long __private *expected, -+ long desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_ulong __global *object, -+ ulong __global *expected, ulong desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_ulong __local *object, -+ ulong __global *expected, ulong desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_ulong __global *object, -+ ulong __local *expected, ulong desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_ulong __local *object, -+ ulong __local *expected, ulong desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_ulong __global *object, -+ ulong __private *expected, ulong desired); -+bool __ovld -+atomic_compare_exchange_strong(volatile atomic_ulong __local *object, -+ ulong __private *expected, ulong desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_ulong __global *object, -+ ulong __global *expected, -+ ulong desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_ulong __local *object, -+ ulong __global *expected, -+ ulong desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_ulong __global *object, -+ ulong __local *expected, -+ ulong desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_ulong __local *object, -+ ulong __local *expected, -+ ulong desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_ulong __global *object, -+ ulong __private *expected, -+ ulong desired); -+bool __ovld atomic_compare_exchange_weak(volatile atomic_ulong __local *object, -+ ulong __private *expected, -+ ulong desired); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // defined(__opencl_c_atomic_scope_device) && -+ // defined(__opencl_c_atomic_order_seq_cst) -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+ -+#ifdef __opencl_c_generic_address_space -+#ifdef __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_int *object, -+ int *expected, int desired, -+ memory_order success, -+ memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint *object, uint *expected, uint desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_int *object, -+ int *expected, int desired, -+ memory_order success, -+ memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_uint *object, -+ uint *expected, uint desired, -+ memory_order success, -+ memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float *object, float *expected, float desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_float *object, -+ float *expected, -+ float desired, -+ memory_order success, -+ memory_order failure); -+#endif // __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit(volatile atomic_int *object, -+ int *expected, int desired, -+ memory_order success, -+ memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint *object, uint *expected, uint desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_int *object, -+ int *expected, int desired, -+ memory_order success, -+ memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_uint *object, -+ uint *expected, uint desired, -+ memory_order success, -+ memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float *object, float *expected, float desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float *object, float *expected, float desired, -+ memory_order success, memory_order failure, memory_scope scope); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double *object, double *expected, double desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double *object, double *expected, double desired, -+ memory_order success, memory_order failure); -+#endif // __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double *object, double *expected, double desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double *object, double *expected, double desired, -+ memory_order success, memory_order failure, memory_scope scope); -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long *object, long *expected, long desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_long *object, -+ long *expected, long desired, -+ memory_order success, -+ memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong *object, ulong *expected, ulong desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_ulong *object, -+ ulong *expected, -+ ulong desired, -+ memory_order success, -+ memory_order failure); -+#endif // __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long *object, long *expected, long desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit(volatile atomic_long *object, -+ long *expected, long desired, -+ memory_order success, -+ memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong *object, ulong *expected, ulong desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong *object, ulong *expected, ulong desired, -+ memory_order success, memory_order failure, memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // __opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+#ifdef __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __global *object, int __global *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __local *object, int __global *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __global *object, int __local *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __local *object, int __local *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __global *object, int __private *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __local *object, int __private *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __global *object, uint __global *expected, -+ uint desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __local *object, uint __global *expected, uint desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __global *object, uint __local *expected, uint desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __local *object, uint __local *expected, uint desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __global *object, uint __private *expected, -+ uint desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __local *object, uint __private *expected, -+ uint desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __global *object, int __global *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __local *object, int __global *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __global *object, int __local *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __local *object, int __local *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __global *object, int __private *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __local *object, int __private *expected, int desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __global *object, uint __global *expected, -+ uint desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __local *object, uint __global *expected, uint desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __global *object, uint __local *expected, uint desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __local *object, uint __local *expected, uint desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __global *object, uint __private *expected, -+ uint desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __local *object, uint __private *expected, -+ uint desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __global *object, float __global *expected, -+ float desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __local *object, float __global *expected, -+ float desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __global *object, float __local *expected, -+ float desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __local *object, float __local *expected, -+ float desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __global *object, float __private *expected, -+ float desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __local *object, float __private *expected, -+ float desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __global *object, float __global *expected, -+ float desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __local *object, float __global *expected, -+ float desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __global *object, float __local *expected, -+ float desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __local *object, float __local *expected, -+ float desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __global *object, float __private *expected, -+ float desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __local *object, float __private *expected, -+ float desired, memory_order success, memory_order failure); -+#endif // __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __global *object, int __global *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __local *object, int __global *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __global *object, int __local *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __local *object, int __local *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __global *object, int __private *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_int __local *object, int __private *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __global *object, uint __global *expected, -+ uint desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __local *object, uint __global *expected, uint desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __global *object, uint __local *expected, uint desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __local *object, uint __local *expected, uint desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __global *object, uint __private *expected, -+ uint desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_uint __local *object, uint __private *expected, -+ uint desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __global *object, int __global *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __local *object, int __global *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __global *object, int __local *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __local *object, int __local *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __global *object, int __private *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_int __local *object, int __private *expected, int desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __global *object, uint __global *expected, -+ uint desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __local *object, uint __global *expected, uint desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __global *object, uint __local *expected, uint desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __local *object, uint __local *expected, uint desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __global *object, uint __private *expected, -+ uint desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_uint __local *object, uint __private *expected, -+ uint desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __global *object, float __global *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __local *object, float __global *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __global *object, float __local *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __local *object, float __local *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __global *object, float __private *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_float __local *object, float __private *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __global *object, float __global *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __local *object, float __global *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __global *object, float __local *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __local *object, float __local *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __global *object, float __private *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_float __local *object, float __private *expected, -+ float desired, memory_order success, memory_order failure, -+ memory_scope scope); -+#if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __global *object, double __global *expected, -+ double desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __local *object, double __global *expected, -+ double desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __global *object, double __local *expected, -+ double desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __local *object, double __local *expected, -+ double desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __global *object, double __private *expected, -+ double desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __local *object, double __private *expected, -+ double desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __global *object, double __global *expected, -+ double desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __local *object, double __global *expected, -+ double desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __global *object, double __local *expected, -+ double desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __local *object, double __local *expected, -+ double desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __global *object, double __private *expected, -+ double desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __local *object, double __private *expected, -+ double desired, memory_order success, memory_order failure); -+#endif // __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __global *object, double __global *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __local *object, double __global *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __global *object, double __local *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __local *object, double __local *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __global *object, double __private *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_double __local *object, double __private *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __global *object, double __global *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __local *object, double __global *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __global *object, double __local *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __local *object, double __local *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __global *object, double __private *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_double __local *object, double __private *expected, -+ double desired, memory_order success, memory_order failure, -+ memory_scope scope); -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#ifdef __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __global *object, long __global *expected, -+ long desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __local *object, long __global *expected, long desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __global *object, long __local *expected, long desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __local *object, long __local *expected, long desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __global *object, long __private *expected, -+ long desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __local *object, long __private *expected, -+ long desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __global *object, long __global *expected, -+ long desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __local *object, long __global *expected, long desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __global *object, long __local *expected, long desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __local *object, long __local *expected, long desired, -+ memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __global *object, long __private *expected, -+ long desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __local *object, long __private *expected, -+ long desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __global *object, ulong __global *expected, -+ ulong desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __local *object, ulong __global *expected, -+ ulong desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __global *object, ulong __local *expected, -+ ulong desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __local *object, ulong __local *expected, -+ ulong desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __global *object, ulong __private *expected, -+ ulong desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __local *object, ulong __private *expected, -+ ulong desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __global *object, ulong __global *expected, -+ ulong desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __local *object, ulong __global *expected, -+ ulong desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __global *object, ulong __local *expected, -+ ulong desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __local *object, ulong __local *expected, -+ ulong desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __global *object, ulong __private *expected, -+ ulong desired, memory_order success, memory_order failure); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __local *object, ulong __private *expected, -+ ulong desired, memory_order success, memory_order failure); -+#endif // __opencl_c_atomic_scope_device -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __global *object, long __global *expected, -+ long desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __local *object, long __global *expected, long desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __global *object, long __local *expected, long desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __local *object, long __local *expected, long desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __global *object, long __private *expected, -+ long desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_long __local *object, long __private *expected, -+ long desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __global *object, long __global *expected, -+ long desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __local *object, long __global *expected, long desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __global *object, long __local *expected, long desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __local *object, long __local *expected, long desired, -+ memory_order success, memory_order failure, memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __global *object, long __private *expected, -+ long desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_long __local *object, long __private *expected, -+ long desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __global *object, ulong __global *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __local *object, ulong __global *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __global *object, ulong __local *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __local *object, ulong __local *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __global *object, ulong __private *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_strong_explicit( -+ volatile atomic_ulong __local *object, ulong __private *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __global *object, ulong __global *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __local *object, ulong __global *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __global *object, ulong __local *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __local *object, ulong __local *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __global *object, ulong __private *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+bool __ovld atomic_compare_exchange_weak_explicit( -+ volatile atomic_ulong __local *object, ulong __private *expected, -+ ulong desired, memory_order success, memory_order failure, -+ memory_scope scope); -+#endif // defined(cl_khr_int64_base_atomics) && -+ // defined(cl_khr_int64_extended_atomics) -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) - - // atomic_flag_test_and_set() and atomic_flag_clear() -- -+#if defined(__opencl_c_atomic_scope_device) && \ -+ defined(__opencl_c_atomic_order_seq_cst) -+#ifdef __opencl_c_generic_address_space - bool __ovld atomic_flag_test_and_set(volatile atomic_flag *object); --bool __ovld atomic_flag_test_and_set_explicit(volatile atomic_flag *object, memory_order order); --bool __ovld atomic_flag_test_and_set_explicit(volatile atomic_flag *object, memory_order order, memory_scope scope); - void __ovld atomic_flag_clear(volatile atomic_flag *object); --void __ovld atomic_flag_clear_explicit(volatile atomic_flag *object, memory_order order); --void __ovld atomic_flag_clear_explicit(volatile atomic_flag *object, memory_order order, memory_scope scope); -+#endif // __opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+bool __ovld atomic_flag_test_and_set(volatile atomic_flag __global *object); -+bool __ovld atomic_flag_test_and_set(volatile atomic_flag __local *object); -+void __ovld atomic_flag_clear(volatile atomic_flag __global *object); -+void __ovld atomic_flag_clear(volatile atomic_flag __local *object); -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+#endif // defined(__opencl_c_atomic_scope_device) && -+ // defined(__opencl_c_atomic_order_seq_cst) -+ -+#ifdef __opencl_c_generic_address_space -+#ifdef __opencl_c_atomic_scope_device -+bool __ovld atomic_flag_test_and_set_explicit(volatile atomic_flag *object, -+ memory_order order); -+void __ovld atomic_flag_clear_explicit(volatile atomic_flag *object, -+ memory_order order); -+#endif // __opencl_c_atomic_scope_device -+bool __ovld atomic_flag_test_and_set_explicit(volatile atomic_flag *object, -+ memory_order order, -+ memory_scope scope); -+void __ovld atomic_flag_clear_explicit(volatile atomic_flag *object, -+ memory_order order, memory_scope scope); -+#endif // __opencl_c_generic_address_space -+ -+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) -+#ifdef __opencl_c_atomic_scope_device -+bool __ovld atomic_flag_test_and_set_explicit( -+ volatile atomic_flag __global *object, memory_order order); -+bool __ovld atomic_flag_test_and_set_explicit( -+ volatile atomic_flag __local *object, memory_order order); -+void __ovld atomic_flag_clear_explicit(volatile atomic_flag __global *object, -+ memory_order order); -+void __ovld atomic_flag_clear_explicit(volatile atomic_flag __local *object, -+ memory_order order); -+#endif // __opencl_c_atomic_scope_device -+bool __ovld -+atomic_flag_test_and_set_explicit(volatile atomic_flag __global *object, -+ memory_order order, memory_scope scope); -+bool __ovld -+atomic_flag_test_and_set_explicit(volatile atomic_flag __local *object, -+ memory_order order, memory_scope scope); -+void __ovld atomic_flag_clear_explicit(volatile atomic_flag __global *object, -+ memory_order order, memory_scope scope); -+void __ovld atomic_flag_clear_explicit(volatile atomic_flag __local *object, -+ memory_order order, memory_scope scope); -+#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_3_0) - - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - -@@ -13906,7 +15943,7 @@ float16 __ovld __cnfn shuffle(float4 x, uint16 mask); - float16 __ovld __cnfn shuffle(float8 x, uint16 mask); - float16 __ovld __cnfn shuffle(float16 x, uint16 mask); - --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double2 __ovld __cnfn shuffle(double2 x, ulong2 mask); - double2 __ovld __cnfn shuffle(double4 x, ulong2 mask); - double2 __ovld __cnfn shuffle(double8 x, ulong2 mask); -@@ -13926,7 +15963,7 @@ double16 __ovld __cnfn shuffle(double2 x, ulong16 mask); - double16 __ovld __cnfn shuffle(double4 x, ulong16 mask); - double16 __ovld __cnfn shuffle(double8 x, ulong16 mask); - double16 __ovld __cnfn shuffle(double16 x, ulong16 mask); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - #ifdef cl_khr_fp16 - half2 __ovld __cnfn shuffle(half2 x, ushort2 mask); -@@ -14130,7 +16167,7 @@ float16 __ovld __cnfn shuffle2(float4 x, float4 y, uint16 mask); - float16 __ovld __cnfn shuffle2(float8 x, float8 y, uint16 mask); - float16 __ovld __cnfn shuffle2(float16 x, float16 y, uint16 mask); - --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double2 __ovld __cnfn shuffle2(double2 x, double2 y, ulong2 mask); - double2 __ovld __cnfn shuffle2(double4 x, double4 y, ulong2 mask); - double2 __ovld __cnfn shuffle2(double8 x, double8 y, ulong2 mask); -@@ -14150,7 +16187,7 @@ double16 __ovld __cnfn shuffle2(double2 x, double2 y, ulong16 mask); - double16 __ovld __cnfn shuffle2(double4 x, double4 y, ulong16 mask); - double16 __ovld __cnfn shuffle2(double8 x, double8 y, ulong16 mask); - double16 __ovld __cnfn shuffle2(double16 x, double16 y, ulong16 mask); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - #ifdef cl_khr_fp16 - half2 __ovld __cnfn shuffle2(half2 x, half2 y, ushort2 mask); -@@ -14186,6 +16223,7 @@ int printf(__constant const char* st, ...) __attribute__((format(printf, 1, 2))) - #pragma OPENCL EXTENSION cl_khr_gl_msaa_sharing : enable - #endif //cl_khr_gl_msaa_sharing - -+#ifdef __opencl_c_images - /** - * Use the coordinate (coord.xy) to do an element lookup in - * the 2D image object specified by image. -@@ -14464,6 +16502,7 @@ half4 __purefn __ovld read_imageh(read_only image1d_buffer_t image, int coord); - - // Image read functions for read_write images - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - float4 __purefn __ovld read_imagef(read_write image1d_t image, int coord); - int4 __purefn __ovld read_imagei(read_write image1d_t image, int coord); - uint4 __purefn __ovld read_imageui(read_write image1d_t image, int coord); -@@ -14507,6 +16546,7 @@ float __purefn __ovld read_imagef(read_write image2d_array_msaa_depth_t image, i - #endif //cl_khr_gl_msaa_sharing - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - #ifdef cl_khr_mipmap_image - float4 __purefn __ovld read_imagef(read_write image1d_t image, sampler_t sampler, float coord, float lod); - int4 __purefn __ovld read_imagei(read_write image1d_t image, sampler_t sampler, float coord, float lod); -@@ -14557,6 +16597,7 @@ int4 __purefn __ovld read_imagei(read_write image3d_t image, sampler_t sampler, - uint4 __purefn __ovld read_imageui(read_write image3d_t image, sampler_t sampler, float4 coord, float4 gradientX, float4 gradientY); - - #endif //cl_khr_mipmap_image -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - // Image read functions returning half4 type -@@ -14568,6 +16609,7 @@ half4 __purefn __ovld read_imageh(read_write image1d_array_t image, int2 coord); - half4 __purefn __ovld read_imageh(read_write image2d_array_t image, int4 coord); - half4 __purefn __ovld read_imageh(read_write image1d_buffer_t image, int coord); - #endif //cl_khr_fp16 -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** -@@ -14657,7 +16699,7 @@ void __ovld write_imagef(write_only image1d_array_t image_array, int2 coord, flo - void __ovld write_imagei(write_only image1d_array_t image_array, int2 coord, int4 color); - void __ovld write_imageui(write_only image1d_array_t image_array, int2 coord, uint4 color); - --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - void __ovld write_imagef(write_only image3d_t image, int4 coord, float4 color); - void __ovld write_imagei(write_only image3d_t image, int4 coord, int4 color); - void __ovld write_imageui(write_only image3d_t image, int4 coord, uint4 color); -@@ -14690,7 +16732,7 @@ void __ovld write_imageui(write_only image2d_array_t image_array, int4 coord, in - void __ovld write_imagef(write_only image2d_depth_t image, int2 coord, int lod, float depth); - void __ovld write_imagef(write_only image2d_array_depth_t image, int4 coord, int lod, float depth); - --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - void __ovld write_imagef(write_only image3d_t image, int4 coord, int lod, float4 color); - void __ovld write_imagei(write_only image3d_t image, int4 coord, int lod, int4 color); - void __ovld write_imageui(write_only image3d_t image, int4 coord, int lod, uint4 color); -@@ -14703,7 +16745,7 @@ void __ovld write_imageui(write_only image3d_t image, int4 coord, int lod, uint4 - #ifdef cl_khr_fp16 - void __ovld write_imageh(write_only image1d_t image, int coord, half4 color); - void __ovld write_imageh(write_only image2d_t image, int2 coord, half4 color); --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - void __ovld write_imageh(write_only image3d_t image, int4 coord, half4 color); - #endif - void __ovld write_imageh(write_only image1d_array_t image, int2 coord, half4 color); -@@ -14713,6 +16755,7 @@ void __ovld write_imageh(write_only image1d_buffer_t image, int coord, half4 col - - // Image write functions for read_write images - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - void __ovld write_imagef(read_write image2d_t image, int2 coord, float4 color); - void __ovld write_imagei(read_write image2d_t image, int2 coord, int4 color); - void __ovld write_imageui(read_write image2d_t image, int2 coord, uint4 color); -@@ -14733,7 +16776,7 @@ void __ovld write_imagef(read_write image1d_array_t image_array, int2 coord, flo - void __ovld write_imagei(read_write image1d_array_t image_array, int2 coord, int4 color); - void __ovld write_imageui(read_write image1d_array_t image_array, int2 coord, uint4 color); - --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - void __ovld write_imagef(read_write image3d_t image, int4 coord, float4 color); - void __ovld write_imagei(read_write image3d_t image, int4 coord, int4 color); - void __ovld write_imageui(read_write image3d_t image, int4 coord, uint4 color); -@@ -14765,7 +16808,7 @@ void __ovld write_imageui(read_write image2d_array_t image_array, int4 coord, in - void __ovld write_imagef(read_write image2d_depth_t image, int2 coord, int lod, float color); - void __ovld write_imagef(read_write image2d_array_depth_t image, int4 coord, int lod, float color); - --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - void __ovld write_imagef(read_write image3d_t image, int4 coord, int lod, float4 color); - void __ovld write_imagei(read_write image3d_t image, int4 coord, int lod, int4 color); - void __ovld write_imageui(read_write image3d_t image, int4 coord, int lod, uint4 color); -@@ -14778,13 +16821,14 @@ void __ovld write_imageui(read_write image3d_t image, int4 coord, int lod, uint4 - #ifdef cl_khr_fp16 - void __ovld write_imageh(read_write image1d_t image, int coord, half4 color); - void __ovld write_imageh(read_write image2d_t image, int2 coord, half4 color); --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - void __ovld write_imageh(read_write image3d_t image, int4 coord, half4 color); - #endif - void __ovld write_imageh(read_write image1d_array_t image, int2 coord, half4 color); - void __ovld write_imageh(read_write image2d_array_t image, int4 coord, half4 color); - void __ovld write_imageh(read_write image1d_buffer_t image, int coord, half4 color); - #endif //cl_khr_fp16 -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - // Note: In OpenCL v1.0/1.1/1.2, image argument of image query builtin functions does not have -@@ -14798,7 +16842,7 @@ void __ovld write_imageh(read_write image1d_buffer_t image, int coord, half4 col - int __ovld __cnfn get_image_width(read_only image1d_t image); - int __ovld __cnfn get_image_width(read_only image1d_buffer_t image); - int __ovld __cnfn get_image_width(read_only image2d_t image); --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - int __ovld __cnfn get_image_width(read_only image3d_t image); - #endif - int __ovld __cnfn get_image_width(read_only image1d_array_t image); -@@ -14817,7 +16861,7 @@ int __ovld __cnfn get_image_width(read_only image2d_array_msaa_depth_t image); - int __ovld __cnfn get_image_width(write_only image1d_t image); - int __ovld __cnfn get_image_width(write_only image1d_buffer_t image); - int __ovld __cnfn get_image_width(write_only image2d_t image); --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - int __ovld __cnfn get_image_width(write_only image3d_t image); - #endif - int __ovld __cnfn get_image_width(write_only image1d_array_t image); -@@ -14834,6 +16878,7 @@ int __ovld __cnfn get_image_width(write_only image2d_array_msaa_depth_t image); - #endif //cl_khr_gl_msaa_sharing - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - int __ovld __cnfn get_image_width(read_write image1d_t image); - int __ovld __cnfn get_image_width(read_write image1d_buffer_t image); - int __ovld __cnfn get_image_width(read_write image2d_t image); -@@ -14850,6 +16895,7 @@ int __ovld __cnfn get_image_width(read_write image2d_msaa_depth_t image); - int __ovld __cnfn get_image_width(read_write image2d_array_msaa_t image); - int __ovld __cnfn get_image_width(read_write image2d_array_msaa_depth_t image); - #endif //cl_khr_gl_msaa_sharing -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** -@@ -14870,7 +16916,7 @@ int __ovld __cnfn get_image_height(read_only image2d_array_msaa_depth_t image); - #endif //cl_khr_gl_msaa_sharing - - int __ovld __cnfn get_image_height(write_only image2d_t image); --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - int __ovld __cnfn get_image_height(write_only image3d_t image); - #endif - int __ovld __cnfn get_image_height(write_only image2d_array_t image); -@@ -14886,6 +16932,7 @@ int __ovld __cnfn get_image_height(write_only image2d_array_msaa_depth_t image); - #endif //cl_khr_gl_msaa_sharing - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - int __ovld __cnfn get_image_height(read_write image2d_t image); - int __ovld __cnfn get_image_height(read_write image3d_t image); - int __ovld __cnfn get_image_height(read_write image2d_array_t image); -@@ -14899,6 +16946,7 @@ int __ovld __cnfn get_image_height(read_write image2d_msaa_depth_t image); - int __ovld __cnfn get_image_height(read_write image2d_array_msaa_t image); - int __ovld __cnfn get_image_height(read_write image2d_array_msaa_depth_t image); - #endif //cl_khr_gl_msaa_sharing -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** -@@ -14906,12 +16954,14 @@ int __ovld __cnfn get_image_height(read_write image2d_array_msaa_depth_t image); - */ - int __ovld __cnfn get_image_depth(read_only image3d_t image); - --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - int __ovld __cnfn get_image_depth(write_only image3d_t image); - #endif - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - int __ovld __cnfn get_image_depth(read_write image3d_t image); -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - // OpenCL Extension v2.0 s9.18 - Mipmaps -@@ -14927,13 +16977,15 @@ int __ovld get_image_num_mip_levels(read_only image3d_t image); - - int __ovld get_image_num_mip_levels(write_only image1d_t image); - int __ovld get_image_num_mip_levels(write_only image2d_t image); --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - int __ovld get_image_num_mip_levels(write_only image3d_t image); - #endif - -+#ifdef __opencl_c_read_write_images - int __ovld get_image_num_mip_levels(read_write image1d_t image); - int __ovld get_image_num_mip_levels(read_write image2d_t image); - int __ovld get_image_num_mip_levels(read_write image3d_t image); -+#endif //__opencl_c_read_write_images - - int __ovld get_image_num_mip_levels(read_only image1d_array_t image); - int __ovld get_image_num_mip_levels(read_only image2d_array_t image); -@@ -14945,10 +16997,12 @@ int __ovld get_image_num_mip_levels(write_only image2d_array_t image); - int __ovld get_image_num_mip_levels(write_only image2d_array_depth_t image); - int __ovld get_image_num_mip_levels(write_only image2d_depth_t image); - -+#ifdef __opencl_c_read_write_images - int __ovld get_image_num_mip_levels(read_write image1d_array_t image); - int __ovld get_image_num_mip_levels(read_write image2d_array_t image); - int __ovld get_image_num_mip_levels(read_write image2d_array_depth_t image); - int __ovld get_image_num_mip_levels(read_write image2d_depth_t image); -+#endif //__opencl_c_read_write_images - - #endif //cl_khr_mipmap_image - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -@@ -14992,7 +17046,7 @@ int __ovld __cnfn get_image_channel_data_type(read_only image2d_array_msaa_depth - int __ovld __cnfn get_image_channel_data_type(write_only image1d_t image); - int __ovld __cnfn get_image_channel_data_type(write_only image1d_buffer_t image); - int __ovld __cnfn get_image_channel_data_type(write_only image2d_t image); --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - int __ovld __cnfn get_image_channel_data_type(write_only image3d_t image); - #endif - int __ovld __cnfn get_image_channel_data_type(write_only image1d_array_t image); -@@ -15009,6 +17063,7 @@ int __ovld __cnfn get_image_channel_data_type(write_only image2d_array_msaa_dept - #endif //cl_khr_gl_msaa_sharing - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - int __ovld __cnfn get_image_channel_data_type(read_write image1d_t image); - int __ovld __cnfn get_image_channel_data_type(read_write image1d_buffer_t image); - int __ovld __cnfn get_image_channel_data_type(read_write image2d_t image); -@@ -15025,6 +17080,7 @@ int __ovld __cnfn get_image_channel_data_type(read_write image2d_msaa_depth_t im - int __ovld __cnfn get_image_channel_data_type(read_write image2d_array_msaa_t image); - int __ovld __cnfn get_image_channel_data_type(read_write image2d_array_msaa_depth_t image); - #endif //cl_khr_gl_msaa_sharing -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** -@@ -15064,7 +17120,7 @@ int __ovld __cnfn get_image_channel_order(read_only image2d_array_msaa_depth_t i - int __ovld __cnfn get_image_channel_order(write_only image1d_t image); - int __ovld __cnfn get_image_channel_order(write_only image1d_buffer_t image); - int __ovld __cnfn get_image_channel_order(write_only image2d_t image); --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - int __ovld __cnfn get_image_channel_order(write_only image3d_t image); - #endif - int __ovld __cnfn get_image_channel_order(write_only image1d_array_t image); -@@ -15081,6 +17137,7 @@ int __ovld __cnfn get_image_channel_order(write_only image2d_array_msaa_depth_t - #endif //cl_khr_gl_msaa_sharing - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - int __ovld __cnfn get_image_channel_order(read_write image1d_t image); - int __ovld __cnfn get_image_channel_order(read_write image1d_buffer_t image); - int __ovld __cnfn get_image_channel_order(read_write image2d_t image); -@@ -15097,6 +17154,7 @@ int __ovld __cnfn get_image_channel_order(read_write image2d_msaa_depth_t image) - int __ovld __cnfn get_image_channel_order(read_write image2d_array_msaa_t image); - int __ovld __cnfn get_image_channel_order(read_write image2d_array_msaa_depth_t image); - #endif //cl_khr_gl_msaa_sharing -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** -@@ -15131,6 +17189,7 @@ int2 __ovld __cnfn get_image_dim(write_only image2d_array_msaa_depth_t image); - #endif //cl_khr_gl_msaa_sharing - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - int2 __ovld __cnfn get_image_dim(read_write image2d_t image); - int2 __ovld __cnfn get_image_dim(read_write image2d_array_t image); - #ifdef cl_khr_depth_images -@@ -15143,6 +17202,7 @@ int2 __ovld __cnfn get_image_dim(read_write image2d_msaa_depth_t image); - int2 __ovld __cnfn get_image_dim(read_write image2d_array_msaa_t image); - int2 __ovld __cnfn get_image_dim(read_write image2d_array_msaa_depth_t image); - #endif //cl_khr_gl_msaa_sharing -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** -@@ -15152,11 +17212,13 @@ int2 __ovld __cnfn get_image_dim(read_write image2d_array_msaa_depth_t image); - * component and the w component is 0. - */ - int4 __ovld __cnfn get_image_dim(read_only image3d_t image); --#ifdef cl_khr_3d_image_writes -+#if defined(cl_khr_3d_image_writes) || defined(__opencl_c_3d_image_writes) - int4 __ovld __cnfn get_image_dim(write_only image3d_t image); - #endif - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - int4 __ovld __cnfn get_image_dim(read_write image3d_t image); -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** -@@ -15184,6 +17246,7 @@ size_t __ovld __cnfn get_image_array_size(write_only image2d_array_msaa_depth_t - #endif //cl_khr_gl_msaa_sharing - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - size_t __ovld __cnfn get_image_array_size(read_write image1d_array_t image_array); - size_t __ovld __cnfn get_image_array_size(read_write image2d_array_t image_array); - #ifdef cl_khr_depth_images -@@ -15193,6 +17256,7 @@ size_t __ovld __cnfn get_image_array_size(read_write image2d_array_depth_t image - size_t __ovld __cnfn get_image_array_size(read_write image2d_array_msaa_t image_array); - size_t __ovld __cnfn get_image_array_size(read_write image2d_array_msaa_depth_t image_array); - #endif //cl_khr_gl_msaa_sharing -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - /** -@@ -15210,16 +17274,21 @@ int __ovld get_image_num_samples(write_only image2d_array_msaa_t image); - int __ovld get_image_num_samples(write_only image2d_array_msaa_depth_t image); - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_read_write_images - int __ovld get_image_num_samples(read_write image2d_msaa_t image); - int __ovld get_image_num_samples(read_write image2d_msaa_depth_t image); - int __ovld get_image_num_samples(read_write image2d_array_msaa_t image); - int __ovld get_image_num_samples(read_write image2d_array_msaa_depth_t image); -+#endif //__opencl_c_read_write_images - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - #endif - -+#endif //__opencl_c_images -+ - // OpenCL v2.0 s6.13.15 - Work-group Functions - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_work_group_collective_functions - int __ovld __conv work_group_all(int predicate); - int __ovld __conv work_group_any(int predicate); - -@@ -15243,11 +17312,11 @@ ulong __ovld __conv work_group_broadcast(ulong a, size_t x, size_t y, size_t z); - float __ovld __conv work_group_broadcast(float a, size_t local_id); - float __ovld __conv work_group_broadcast(float a, size_t x, size_t y); - float __ovld __conv work_group_broadcast(float a, size_t x, size_t y, size_t z); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __conv work_group_broadcast(double a, size_t local_id); - double __ovld __conv work_group_broadcast(double a, size_t x, size_t y); - double __ovld __conv work_group_broadcast(double a, size_t x, size_t y, size_t z); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - #ifdef cl_khr_fp16 - half __ovld __conv work_group_reduce_add(half x); -@@ -15305,7 +17374,7 @@ float __ovld __conv work_group_scan_exclusive_max(float x); - float __ovld __conv work_group_scan_inclusive_add(float x); - float __ovld __conv work_group_scan_inclusive_min(float x); - float __ovld __conv work_group_scan_inclusive_max(float x); --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __conv work_group_reduce_add(double x); - double __ovld __conv work_group_reduce_min(double x); - double __ovld __conv work_group_reduce_max(double x); -@@ -15315,19 +17384,18 @@ double __ovld __conv work_group_scan_exclusive_max(double x); - double __ovld __conv work_group_scan_inclusive_add(double x); - double __ovld __conv work_group_scan_inclusive_min(double x); - double __ovld __conv work_group_scan_inclusive_max(double x); --#endif //cl_khr_fp64 -- -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) -+#endif //__opencl_c_work_group_collective_functions - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - // OpenCL v2.0 s6.13.16 - Pipe Functions --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_pipes - bool __ovld is_valid_reserve_id(reserve_id_t reserve_id); --#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -- -+#endif //__opencl_c_pipes - - // OpenCL v2.0 s6.13.17 - Enqueue Kernels - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -- -+#ifdef __opencl_c_device_enqueue - ndrange_t __ovld ndrange_1D(size_t); - ndrange_t __ovld ndrange_1D(size_t, size_t); - ndrange_t __ovld ndrange_1D(size_t, size_t, size_t); -@@ -15355,11 +17423,13 @@ bool __ovld is_valid_event (clk_event_t event); - void __ovld capture_event_profiling_info(clk_event_t, clk_profiling_info, __global void* value); - - queue_t __ovld get_default_queue(void); -+#endif //__opencl_c_device_enqueue - #endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - // OpenCL Extension v2.0 s9.17 - Sub-groups - --#if defined(cl_intel_subgroups) || defined(cl_khr_subgroups) -+#if defined(cl_intel_subgroups) || defined(cl_khr_subgroups) || \ -+ defined(__opencl_c_subgroups) - // Shared Sub Group Functions - uint __ovld get_sub_group_size(void); - uint __ovld get_max_sub_group_size(void); -@@ -15445,7 +17515,7 @@ half __ovld __conv sub_group_scan_inclusive_min(half x); - half __ovld __conv sub_group_scan_inclusive_max(half x); - #endif //cl_khr_fp16 - --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __conv sub_group_broadcast(double x, uint sub_group_local_id); - double __ovld __conv sub_group_reduce_add(double x); - double __ovld __conv sub_group_reduce_min(double x); -@@ -15456,7 +17526,7 @@ double __ovld __conv sub_group_scan_exclusive_max(double x); - double __ovld __conv sub_group_scan_inclusive_add(double x); - double __ovld __conv sub_group_scan_inclusive_min(double x); - double __ovld __conv sub_group_scan_inclusive_max(double x); --#endif //cl_khr_fp64 -+#endif // defined(cl_khr_fp64) || defined(__opencl_c_fp64) - - #endif //cl_khr_subgroups cl_intel_subgroups - -@@ -16226,16 +18296,22 @@ uint16 __ovld __conv intel_sub_group_shuffle_xor( uint16 x, uint c ); - long __ovld __conv intel_sub_group_shuffle_xor( long x, uint c ); - ulong __ovld __conv intel_sub_group_shuffle_xor( ulong x, uint c ); - -+#ifdef __opencl_c_images - uint __ovld __conv intel_sub_group_block_read( read_only image2d_t image, int2 coord ); - uint2 __ovld __conv intel_sub_group_block_read2( read_only image2d_t image, int2 coord ); - uint4 __ovld __conv intel_sub_group_block_read4( read_only image2d_t image, int2 coord ); - uint8 __ovld __conv intel_sub_group_block_read8( read_only image2d_t image, int2 coord ); -+#endif //__opencl_c_images - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_images -+#ifdef __opencl_c_read_write_images - uint __ovld __conv intel_sub_group_block_read(read_write image2d_t image, int2 coord); - uint2 __ovld __conv intel_sub_group_block_read2(read_write image2d_t image, int2 coord); - uint4 __ovld __conv intel_sub_group_block_read4(read_write image2d_t image, int2 coord); - uint8 __ovld __conv intel_sub_group_block_read8(read_write image2d_t image, int2 coord); -+#endif //__opencl_c_read_write_images -+#endif //__opencl_c_images - #endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - uint __ovld __conv intel_sub_group_block_read( const __global uint* p ); -@@ -16243,16 +18319,22 @@ uint2 __ovld __conv intel_sub_group_block_read2( const __global uint* p ); - uint4 __ovld __conv intel_sub_group_block_read4( const __global uint* p ); - uint8 __ovld __conv intel_sub_group_block_read8( const __global uint* p ); - -+#ifdef __opencl_c_images - void __ovld __conv intel_sub_group_block_write(write_only image2d_t image, int2 coord, uint data); - void __ovld __conv intel_sub_group_block_write2(write_only image2d_t image, int2 coord, uint2 data); - void __ovld __conv intel_sub_group_block_write4(write_only image2d_t image, int2 coord, uint4 data); - void __ovld __conv intel_sub_group_block_write8(write_only image2d_t image, int2 coord, uint8 data); -+#endif //__opencl_c_images - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_images -+#ifdef __opencl_c_read_write_images - void __ovld __conv intel_sub_group_block_write(read_write image2d_t image, int2 coord, uint data); - void __ovld __conv intel_sub_group_block_write2(read_write image2d_t image, int2 coord, uint2 data); - void __ovld __conv intel_sub_group_block_write4(read_write image2d_t image, int2 coord, uint4 data); - void __ovld __conv intel_sub_group_block_write8(read_write image2d_t image, int2 coord, uint8 data); -+#endif // __opencl_c_read_write_images -+#endif // __opencl_c_images - #endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - void __ovld __conv intel_sub_group_block_write( __global uint* p, uint data ); -@@ -16267,7 +18349,7 @@ half __ovld __conv intel_sub_group_shuffle_up( half prev, half cur, uint c ); - half __ovld __conv intel_sub_group_shuffle_xor( half x, uint c ); - #endif - --#if defined(cl_khr_fp64) -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - double __ovld __conv intel_sub_group_shuffle( double x, uint c ); - double __ovld __conv intel_sub_group_shuffle_down( double prev, double cur, uint c ); - double __ovld __conv intel_sub_group_shuffle_up( double prev, double cur, uint c ); -@@ -16366,16 +18448,22 @@ ushort __ovld __conv intel_sub_group_scan_inclusive_min( ushort x ); - short __ovld __conv intel_sub_group_scan_inclusive_max( short x ); - ushort __ovld __conv intel_sub_group_scan_inclusive_max( ushort x ); - -+#ifdef __opencl_c_images - uint __ovld __conv intel_sub_group_block_read_ui( read_only image2d_t image, int2 byte_coord ); - uint2 __ovld __conv intel_sub_group_block_read_ui2( read_only image2d_t image, int2 byte_coord ); - uint4 __ovld __conv intel_sub_group_block_read_ui4( read_only image2d_t image, int2 byte_coord ); - uint8 __ovld __conv intel_sub_group_block_read_ui8( read_only image2d_t image, int2 byte_coord ); -+#endif //__opencl_c_images - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_images -+#ifdef __opencl_c_read_write_images - uint __ovld __conv intel_sub_group_block_read_ui( read_write image2d_t image, int2 byte_coord ); - uint2 __ovld __conv intel_sub_group_block_read_ui2( read_write image2d_t image, int2 byte_coord ); - uint4 __ovld __conv intel_sub_group_block_read_ui4( read_write image2d_t image, int2 byte_coord ); - uint8 __ovld __conv intel_sub_group_block_read_ui8( read_write image2d_t image, int2 byte_coord ); -+#endif //__opencl_c_read_write_images -+#endif //__opencl_c_images - #endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - uint __ovld __conv intel_sub_group_block_read_ui( const __global uint* p ); -@@ -16383,16 +18471,22 @@ uint2 __ovld __conv intel_sub_group_block_read_ui2( const __global uint* p - uint4 __ovld __conv intel_sub_group_block_read_ui4( const __global uint* p ); - uint8 __ovld __conv intel_sub_group_block_read_ui8( const __global uint* p ); - -+#ifdef __opencl_c_images - void __ovld __conv intel_sub_group_block_write_ui( read_only image2d_t image, int2 byte_coord, uint data ); - void __ovld __conv intel_sub_group_block_write_ui2( read_only image2d_t image, int2 byte_coord, uint2 data ); - void __ovld __conv intel_sub_group_block_write_ui4( read_only image2d_t image, int2 byte_coord, uint4 data ); - void __ovld __conv intel_sub_group_block_write_ui8( read_only image2d_t image, int2 byte_coord, uint8 data ); -+#endif //__opencl_c_images - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_images -+#ifdef __opencl_c_read_write_images - void __ovld __conv intel_sub_group_block_write_ui( read_write image2d_t image, int2 byte_coord, uint data ); - void __ovld __conv intel_sub_group_block_write_ui2( read_write image2d_t image, int2 byte_coord, uint2 data ); - void __ovld __conv intel_sub_group_block_write_ui4( read_write image2d_t image, int2 byte_coord, uint4 data ); - void __ovld __conv intel_sub_group_block_write_ui8( read_write image2d_t image, int2 byte_coord, uint8 data ); -+#endif //__opencl_c_read_write_images -+#endif //__opencl_c_images - #endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - void __ovld __conv intel_sub_group_block_write_ui( __global uint* p, uint data ); -@@ -16400,16 +18494,22 @@ void __ovld __conv intel_sub_group_block_write_ui2( __global uint* p, uint - void __ovld __conv intel_sub_group_block_write_ui4( __global uint* p, uint4 data ); - void __ovld __conv intel_sub_group_block_write_ui8( __global uint* p, uint8 data ); - -+#ifdef __opencl_c_images - ushort __ovld __conv intel_sub_group_block_read_us( read_only image2d_t image, int2 coord ); - ushort2 __ovld __conv intel_sub_group_block_read_us2( read_only image2d_t image, int2 coord ); - ushort4 __ovld __conv intel_sub_group_block_read_us4( read_only image2d_t image, int2 coord ); - ushort8 __ovld __conv intel_sub_group_block_read_us8( read_only image2d_t image, int2 coord ); -+#endif //__opencl_c_images - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_images -+#ifdef __opencl_c_read_write_images - ushort __ovld __conv intel_sub_group_block_read_us(read_write image2d_t image, int2 coord); - ushort2 __ovld __conv intel_sub_group_block_read_us2(read_write image2d_t image, int2 coord); - ushort4 __ovld __conv intel_sub_group_block_read_us4(read_write image2d_t image, int2 coord); - ushort8 __ovld __conv intel_sub_group_block_read_us8(read_write image2d_t image, int2 coord); -+#endif //__opencl_c_read_write_images -+#endif //__opencl_c_images - #endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - ushort __ovld __conv intel_sub_group_block_read_us( const __global ushort* p ); -@@ -16417,16 +18517,22 @@ ushort2 __ovld __conv intel_sub_group_block_read_us2( const __global ushort* - ushort4 __ovld __conv intel_sub_group_block_read_us4( const __global ushort* p ); - ushort8 __ovld __conv intel_sub_group_block_read_us8( const __global ushort* p ); - -+#ifdef __opencl_c_images - void __ovld __conv intel_sub_group_block_write_us(write_only image2d_t image, int2 coord, ushort data); - void __ovld __conv intel_sub_group_block_write_us2(write_only image2d_t image, int2 coord, ushort2 data); - void __ovld __conv intel_sub_group_block_write_us4(write_only image2d_t image, int2 coord, ushort4 data); - void __ovld __conv intel_sub_group_block_write_us8(write_only image2d_t image, int2 coord, ushort8 data); -+#endif //__opencl_c_images - - #if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#ifdef __opencl_c_images -+#ifdef __opencl_c_read_write_images - void __ovld __conv intel_sub_group_block_write_us(read_write image2d_t image, int2 coord, ushort data); - void __ovld __conv intel_sub_group_block_write_us2(read_write image2d_t image, int2 coord, ushort2 data); - void __ovld __conv intel_sub_group_block_write_us4(read_write image2d_t image, int2 coord, ushort4 data); - void __ovld __conv intel_sub_group_block_write_us8(read_write image2d_t image, int2 coord, ushort8 data); -+#endif //__opencl_c_read_write_images -+#endif //__opencl_c_images - #endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) - - void __ovld __conv intel_sub_group_block_write_us( __global ushort* p, ushort data ); -@@ -16545,6 +18651,7 @@ short2 __ovld intel_sub_group_avc_ime_adjust_ref_offset( - short2 ref_offset, ushort2 src_coord, ushort2 ref_window_size, - ushort2 image_size); - -+#ifdef __opencl_c_images - intel_sub_group_avc_ime_result_t __ovld - intel_sub_group_avc_ime_evaluate_with_single_reference( - read_only image2d_t src_image, read_only image2d_t ref_image, -@@ -16585,6 +18692,7 @@ intel_sub_group_avc_ime_evaluate_with_dual_reference_streaminout( - read_only image2d_t bwd_ref_image, sampler_t vme_media_sampler, - intel_sub_group_avc_ime_payload_t payload, - intel_sub_group_avc_ime_dual_reference_streamin_t streamin_components); -+#endif //__opencl_c_images - - intel_sub_group_avc_ime_single_reference_streamin_t __ovld - intel_sub_group_avc_ime_get_single_reference_streamin( -@@ -16649,6 +18757,7 @@ intel_sub_group_avc_ref_payload_t __ovld - intel_sub_group_avc_ref_set_bilinear_filter_enable( - intel_sub_group_avc_ref_payload_t payload); - -+#ifdef __opencl_c_images - intel_sub_group_avc_ref_result_t __ovld - intel_sub_group_avc_ref_evaluate_with_single_reference( - read_only image2d_t src_image, read_only image2d_t ref_image, -@@ -16667,6 +18776,7 @@ intel_sub_group_avc_ref_evaluate_with_multi_reference( - read_only image2d_t src_image, uint packed_reference_ids, - uchar packed_reference_field_polarities, sampler_t vme_media_sampler, - intel_sub_group_avc_ref_payload_t payload); -+#endif //__opencl_c_images - - // SIC built-in functions - intel_sub_group_avc_sic_payload_t __ovld -@@ -16717,6 +18827,7 @@ intel_sub_group_avc_sic_set_block_based_raw_skip_sad( - uchar block_based_skip_type, - intel_sub_group_avc_sic_payload_t payload); - -+#ifdef __opencl_c_images - intel_sub_group_avc_sic_result_t __ovld - intel_sub_group_avc_sic_evaluate_ipe( - read_only image2d_t src_image, sampler_t vme_media_sampler, -@@ -16739,6 +18850,7 @@ intel_sub_group_avc_sic_evaluate_with_multi_reference( - read_only image2d_t src_image, uint packed_reference_ids, - uchar packed_reference_field_polarities, sampler_t vme_media_sampler, - intel_sub_group_avc_sic_payload_t payload); -+#endif //__opencl_c_images - - uchar __ovld intel_sub_group_avc_sic_get_ipe_luma_shape( - intel_sub_group_avc_sic_result_t result); -diff --git a/clang/lib/Parse/ParseDecl.cpp b/clang/lib/Parse/ParseDecl.cpp -index c87d240a8206..8c8f03bba0f3 100644 ---- a/clang/lib/Parse/ParseDecl.cpp -+++ b/clang/lib/Parse/ParseDecl.cpp -@@ -3799,8 +3799,7 @@ void Parser::ParseDeclarationSpecifiers(DeclSpec &DS, - isInvalid = DS.SetTypeAltiVecBool(true, Loc, PrevSpec, DiagID, Policy); - break; - case tok::kw_pipe: -- if (!getLangOpts().OpenCL || (getLangOpts().OpenCLVersion < 200 && -- !getLangOpts().OpenCLCPlusPlus)) { -+ if (!getLangOpts().OpenCLPipeKeyword) { - // OpenCL 2.0 defined this keyword. OpenCL 1.2 and earlier should - // support the "pipe" word as identifier. - Tok.getIdentifierInfo()->revertTokenIDToIdentifier(); -@@ -3922,8 +3921,7 @@ void Parser::ParseDeclarationSpecifiers(DeclSpec &DS, - case tok::kw___generic: - // generic address space is introduced only in OpenCL v2.0 - // see OpenCL C Spec v2.0 s6.5.5 -- if (Actions.getLangOpts().OpenCLVersion < 200 && -- !Actions.getLangOpts().OpenCLCPlusPlus) { -+ if (!Actions.getLangOpts().OpenCLGenericKeyword) { - DiagID = diag::err_opencl_unknown_type_specifier; - PrevSpec = Tok.getIdentifierInfo()->getNameStart(); - isInvalid = true; -@@ -4973,8 +4971,7 @@ bool Parser::isDeclarationSpecifier(bool DisambiguatingWithExpression) { - default: return false; - - case tok::kw_pipe: -- return (getLangOpts().OpenCL && getLangOpts().OpenCLVersion >= 200) || -- getLangOpts().OpenCLCPlusPlus; -+ return getLangOpts().OpenCLPipeKeyword; - - case tok::identifier: // foo::bar - // Unfortunate hack to support "Class.factoryMethod" notation. -@@ -5499,8 +5496,7 @@ static bool isPtrOperatorToken(tok::TokenKind Kind, const LangOptions &Lang, - if (Kind == tok::star || Kind == tok::caret) - return true; - -- if (Kind == tok::kw_pipe && -- ((Lang.OpenCL && Lang.OpenCLVersion >= 200) || Lang.OpenCLCPlusPlus)) -+ if (Kind == tok::kw_pipe && Lang.OpenCLPipeKeyword) - return true; - - if (!Lang.CPlusPlus) -diff --git a/clang/lib/Parse/ParsePragma.cpp b/clang/lib/Parse/ParsePragma.cpp -index 6402b31d00b2..65f283ad556b 100644 ---- a/clang/lib/Parse/ParsePragma.cpp -+++ b/clang/lib/Parse/ParsePragma.cpp -@@ -749,12 +749,14 @@ void Parser::HandlePragmaOpenCLExtension() { - if (Name == "all") { - if (State == Disable) { - Opt.disableAll(); -- Opt.enableSupportedCore(getLangOpts()); -+ Opt.enableSupportedCore(); - } else { - PP.Diag(NameLoc, diag::warn_pragma_expected_predicate) << 1; - } -+ } else if (Opt.isFeature(Name)) { -+ PP.Diag(NameLoc, diag::warn_opencl_pragma_feature_ignore) << Ident; - } else if (State == Begin) { -- if (!Opt.isKnown(Name) || !Opt.isSupported(Name, getLangOpts())) { -+ if (!Opt.isKnown(Name) || !Opt.isSupported(Name)) { - Opt.support(Name); - } - Actions.setCurrentOpenCLExtension(Name); -@@ -764,9 +766,9 @@ void Parser::HandlePragmaOpenCLExtension() { - Actions.setCurrentOpenCLExtension(""); - } else if (!Opt.isKnown(Name)) - PP.Diag(NameLoc, diag::warn_pragma_unknown_extension) << Ident; -- else if (Opt.isSupportedExtension(Name, getLangOpts())) -+ else if (Opt.isSupportedExtension(Name)) - Opt.enable(Name, State == Enable); -- else if (Opt.isSupportedCore(Name, getLangOpts())) -+ else if (Opt.isSupportedCore(Name)) - PP.Diag(NameLoc, diag::warn_pragma_extension_is_core) << Ident; - else - PP.Diag(NameLoc, diag::warn_pragma_unsupported_extension) << Ident; -diff --git a/clang/lib/Sema/OpenCLBuiltins.td b/clang/lib/Sema/OpenCLBuiltins.td -index 745363a6b43f..6a1fce567090 100644 ---- a/clang/lib/Sema/OpenCLBuiltins.td -+++ b/clang/lib/Sema/OpenCLBuiltins.td -@@ -22,11 +22,13 @@ - class Version<int _Version> { - int ID = _Version; - } -+ - def CLAll : Version< 0>; - def CL10 : Version<100>; - def CL11 : Version<110>; - def CL12 : Version<120>; - def CL20 : Version<200>; -+def CL30 : Version<300>; - - // Address spaces - // Pointer types need to be assigned an address space. -@@ -72,6 +74,14 @@ def ArmIntegerDotProductAccumulateInt8 : FunctionExtension<"cl_arm_integ - def ArmIntegerDotProductAccumulateInt16 : FunctionExtension<"cl_arm_integer_dot_product_accumulate_int16">; - def ArmIntegerDotProductAccumulateSaturateInt8 : FunctionExtension<"cl_arm_integer_dot_product_accumulate_saturate_int8">; - -+// Features -+def FuncExtGenericAddressSpace : FunctionExtension<"__opencl_c_generic_address_space">; -+def FuncExtWorkGroupCollective : FunctionExtension<"__opencl_c_work_group_collective_functions">; -+def FuncExtPipes : FunctionExtension<"__opencl_c_pipes">; -+def FuncExtDeviceSidEenqueue : FunctionExtension<"__opencl_c_device_enqueue">; -+def FuncNonExplicitAtomicFeatures : FunctionExtension<"__opencl_c_atomic_order_seq_cst __opencl_c_atomic_scope_device">; -+def FuncNonExplicitAtomicFeaturesAndGenericAS : FunctionExtension<"__opencl_c_atomic_order_seq_cst __opencl_c_atomic_scope_device __opencl_c_generic_address_space">; -+ - // Qualified Type. These map to ASTContext::QualType. - class QualType<string _Name, bit _IsAbstract=0> { - // Name of the field or function in a clang::ASTContext -@@ -236,13 +246,9 @@ class Builtin<string _Name, list<Type> _Signature, list<bit> _Attributes = Attr. - bit IsConv = _Attributes[2]; - // OpenCL extensions to which the function belongs. - FunctionExtension Extension = FuncExtNone; -- // Version of OpenCL from which the function is available (e.g.: CL10). -- // MinVersion is inclusive. -- Version MinVersion = CL10; -- // Version of OpenCL from which the function is not supported anymore. -- // MaxVersion is exclusive. -+ // List of OpenCL version in which this function available. - // CLAll makes the function available for all versions. -- Version MaxVersion = CLAll; -+ list<Version> Versions = [CLAll]; - } - - //===----------------------------------------------------------------------===// -@@ -421,7 +427,7 @@ foreach name = ["get_global_size", "get_global_id", "get_local_size", - def : Builtin<name, [Size, UInt], Attr.Const>; - } - --let MinVersion = CL20 in { -+let Versions = [CL20, CL30] in { - def : Builtin<"get_enqueued_local_size", [Size, UInt]>; - foreach name = ["get_global_linear_id", "get_local_linear_id"] in { - def : Builtin<name, [Size]>; -@@ -494,7 +500,7 @@ foreach name = ["fma", "mad"] in { - } - - // --- Version dependent --- --let MaxVersion = CL20 in { -+let Versions = [CL10, CL11, CL12, CL30] in { - foreach AS = [GlobalAS, LocalAS, PrivateAS] in { - foreach name = ["fract", "modf", "sincos"] in { - def : Builtin<name, [FGenTypeN, FGenTypeN, PointerType<FGenTypeN, AS>]>; -@@ -511,7 +517,9 @@ let MaxVersion = CL20 in { - } - } - } --let MinVersion = CL20 in { -+ -+let Versions = [CL20, CL30] in { -+ let Extension = FuncExtGenericAddressSpace in { - foreach name = ["fract", "modf", "sincos"] in { - def : Builtin<name, [FGenTypeN, FGenTypeN, PointerType<FGenTypeN, GenericAS>]>; - } -@@ -524,6 +532,7 @@ let MinVersion = CL20 in { - def : Builtin<name, [Type, Type, Type, PointerType<GenTypeIntVecAndScalar, GenericAS>]>; - } - } -+ } - } - - // --- Table 9 --- -@@ -554,7 +563,7 @@ foreach name = ["abs"] in { - foreach name = ["clz", "popcount"] in { - def : Builtin<name, [AIGenTypeN, AIGenTypeN], Attr.Const>; - } --let MinVersion = CL20 in { -+let Versions = [CL20, CL30] in { - foreach name = ["ctz"] in { - def : Builtin<name, [AIGenTypeN, AIGenTypeN]>; - } -@@ -728,7 +737,7 @@ foreach name = ["select"] in { - // --- Table 15 --- - // Variants for OpenCL versions below 2.0, using pointers to the global, local - // and private address spaces. --let MaxVersion = CL20 in { -+let Versions = [CL10, CL11, CL12, CL30] in { - foreach AS = [GlobalAS, LocalAS, PrivateAS] in { - foreach VSize = [2, 3, 4, 8, 16] in { - foreach name = ["vload" # VSize] in { -@@ -771,7 +780,8 @@ let MaxVersion = CL20 in { - } - // Variants for OpenCL versions above 2.0, using pointers to the generic - // address space. --let MinVersion = CL20 in { -+let Versions = [CL20, CL30] in { -+ let Extension = FuncExtGenericAddressSpace in { - foreach VSize = [2, 3, 4, 8, 16] in { - foreach name = ["vload" # VSize] in { - def : Builtin<name, [VectorType<Char, VSize>, Size, PointerType<ConstType<Char>, GenericAS>]>; -@@ -809,6 +819,7 @@ let MinVersion = CL20 in { - } - } - } -+ } - } - // Variants using pointers to the constant address space. - foreach VSize = [2, 3, 4, 8, 16] in { -@@ -829,7 +840,7 @@ foreach VSize = [2, 3, 4, 8, 16] in { - def : Builtin<name, [VectorType<Float, VSize>, Size, PointerType<ConstType<Half>, ConstantAS>]>; - } - } --let MaxVersion = CL20 in { -+let Versions = [CL10, CL11, CL12, CL30] in { - foreach AS = [GlobalAS, LocalAS, PrivateAS] in { - def : Builtin<"vload_half", [Float, Size, PointerType<ConstType<Half>, AS>]>; - def : Builtin<"vloada_half", [Float, Size, PointerType<ConstType<Half>, AS>]>; -@@ -852,7 +863,8 @@ let MaxVersion = CL20 in { - } - } - } --let MinVersion = CL20 in { -+let Versions = [CL20, CL30] in { -+ let Extension = FuncExtGenericAddressSpace in { - foreach AS = [GenericAS] in { - def : Builtin<"vload_half", [Float, Size, PointerType<ConstType<Half>, AS>]>; - def : Builtin<"vloada_half", [Float, Size, PointerType<ConstType<Half>, AS>]>; -@@ -874,6 +886,7 @@ let MinVersion = CL20 in { - } - } - } -+ } - } - - foreach AS = [ConstantAS] in { -@@ -899,7 +912,9 @@ foreach name = ["async_work_group_strided_copy"] in { - def : Builtin<name, [Event, PointerType<AGenTypeN, GlobalAS>, PointerType<ConstType<AGenTypeN>, LocalAS>, Size, Size, Event]>; - } - foreach name = ["wait_group_events"] in { -+ let Extension = FuncExtGenericAddressSpace in { - def : Builtin<name, [Void, Int, PointerType<Event, GenericAS>]>; -+ } - } - foreach name = ["prefetch"] in { - def : Builtin<name, [Void, PointerType<ConstType<AGenTypeN>, GlobalAS>, Size]>; -@@ -1000,13 +1015,25 @@ foreach AS = [GlobalAS, LocalAS] in { - } - } - } -+ - // OpenCL v2.0 s6.13.11 - Atomic Functions. --let MinVersion = CL20 in { -+ -+let Versions = [CL20, CL30] in { -+ let Extension = FuncExtGenericAddressSpace in { -+ foreach TypePair = [[AtomicInt, Int], [AtomicUInt, UInt], -+ [AtomicLong, Long], [AtomicULong, ULong], -+ [AtomicFloat, Float], [AtomicDouble, Double]] in { -+ def : Builtin<"atomic_init", -+ [Void, PointerType<VolatileType<TypePair[0]>, GenericAS>, TypePair[1]]>; -+ } -+ } -+} -+ -+let Versions = [CL20, CL30] in { -+ let Extension = FuncNonExplicitAtomicFeaturesAndGenericAS in { - foreach TypePair = [[AtomicInt, Int], [AtomicUInt, UInt], - [AtomicLong, Long], [AtomicULong, ULong], - [AtomicFloat, Float], [AtomicDouble, Double]] in { -- def : Builtin<"atomic_init", -- [Void, PointerType<VolatileType<TypePair[0]>, GenericAS>, TypePair[1]]>; - def : Builtin<"atomic_store", - [Void, PointerType<VolatileType<TypePair[0]>, GenericAS>, TypePair[1]]>; - def : Builtin<"atomic_load", -@@ -1038,6 +1065,59 @@ let MinVersion = CL20 in { - [TypePair[1], PointerType<VolatileType<TypePair[0]>, GenericAS>, TypePair[2]]>; - } - } -+ } -+} -+ -+let Versions = [CL30] in { -+ foreach AS = [GlobalAS, LocalAS] in { -+ foreach TypePair = [[AtomicInt, Int], [AtomicUInt, UInt], -+ [AtomicLong, Long], [AtomicULong, ULong], -+ [AtomicFloat, Float], [AtomicDouble, Double]] in { -+ def : Builtin<"atomic_init", -+ [Void, PointerType<VolatileType<TypePair[0]>, AS>, TypePair[1]]>; -+ } -+ } -+} -+ -+let Versions = [CL30] in { -+ let Extension = FuncNonExplicitAtomicFeatures in { -+ foreach AS = [GlobalAS, LocalAS] in { -+ foreach TypePair = [[AtomicInt, Int], [AtomicUInt, UInt], -+ [AtomicLong, Long], [AtomicULong, ULong], -+ [AtomicFloat, Float], [AtomicDouble, Double]] in { -+ def : Builtin<"atomic_store", -+ [Void, PointerType<VolatileType<TypePair[0]>, AS>, TypePair[1]]>; -+ def : Builtin<"atomic_load", -+ [TypePair[1], PointerType<VolatileType<TypePair[0]>, AS>]>; -+ def : Builtin<"atomic_exchange", -+ [TypePair[1], PointerType<VolatileType<TypePair[0]>, AS>, TypePair[1]]>; -+ foreach Variant = ["weak", "strong"] in { -+ def : Builtin<"atomic_compare_exchange_" # Variant, -+ [Bool, PointerType<VolatileType<TypePair[0]>, AS>, -+ PointerType<TypePair[1], AS>, TypePair[1]]>; -+ } -+ } -+ -+ foreach TypePair = [[AtomicInt, Int, Int], [AtomicUInt, UInt, UInt], -+ [AtomicLong, Long, Long], [AtomicULong, ULong, ULong], -+ [AtomicIntPtr, IntPtr, PtrDiff], -+ [AtomicUIntPtr, UIntPtr, PtrDiff]] in { -+ foreach ModOp = ["add", "sub"] in { -+ def : Builtin<"atomic_fetch_" # ModOp, -+ [TypePair[1], PointerType<VolatileType<TypePair[0]>, AS>, TypePair[2]]>; -+ } -+ } -+ foreach TypePair = [[AtomicInt, Int, Int], [AtomicUInt, UInt, UInt], -+ [AtomicLong, Long, Long], [AtomicULong, ULong, ULong], -+ [AtomicIntPtr, IntPtr, IntPtr], -+ [AtomicUIntPtr, UIntPtr, UIntPtr]] in { -+ foreach ModOp = ["or", "xor", "and", "min", "max"] in { -+ def : Builtin<"atomic_fetch_" # ModOp, -+ [TypePair[1], PointerType<VolatileType<TypePair[0]>, AS>, TypePair[2]]>; -+ } -+ } -+ } -+ } - } - - //-------------------------------------------------------------------- -@@ -1217,7 +1297,8 @@ foreach aQual = ["WO", "RW"] in { - //-------------------------------------------------------------------- - // OpenCL v2.0 s6.13.15 - Work-group Functions - // --- Table 26 --- --let MinVersion = CL20 in { -+let Versions = [CL20, CL30] in { -+ let Extension = FuncExtWorkGroupCollective in { - foreach name = ["work_group_all", "work_group_any"] in { - def : Builtin<name, [Int, Int], Attr.Convergent>; - } -@@ -1232,6 +1313,7 @@ let MinVersion = CL20 in { - def : Builtin<name # op, [IntLongFloatGenType1, IntLongFloatGenType1], Attr.Convergent>; - } - } -+ } - } - - -@@ -1241,8 +1323,10 @@ let MinVersion = CL20 in { - // Defined in Builtins.def - - // --- Table 28 --- --// Builtins taking pipe arguments are defined in Builtins.def --def : Builtin<"is_valid_reserve_id", [Bool, ReserveId]>; -+let Extension = FuncExtPipes in { -+ // Builtins taking pipe arguments are defined in Builtins.def -+ def : Builtin<"is_valid_reserve_id", [Bool, ReserveId]>; -+} - - // --- Table 29 --- - // Defined in Builtins.def -@@ -1257,21 +1341,22 @@ def : Builtin<"is_valid_reserve_id", [Bool, ReserveId]>; - // Defined in Builtins.def - - // --- Table 33 --- --def : Builtin<"enqueue_marker", -- [Int, Queue, UInt, PointerType<ConstType<ClkEvent>, GenericAS>, PointerType<ClkEvent, GenericAS>]>; -+let Extension = FuncExtDeviceSidEenqueue in { -+ def : Builtin<"enqueue_marker", -+ [Int, Queue, UInt, PointerType<ConstType<ClkEvent>, GenericAS>, PointerType<ClkEvent, GenericAS>]>; - --// --- Table 34 --- --def : Builtin<"retain_event", [Void, ClkEvent]>; --def : Builtin<"release_event", [Void, ClkEvent]>; --def : Builtin<"create_user_event", [ClkEvent]>; --def : Builtin<"is_valid_event", [Bool, ClkEvent]>; --def : Builtin<"set_user_event_status", [Void, ClkEvent, Int]>; --// TODO: capture_event_profiling_info -- --// --- Table 35 --- --def : Builtin<"get_default_queue", [Queue]>; --// TODO: ndrange functions -+ // --- Table 34 --- -+ def : Builtin<"retain_event", [Void, ClkEvent]>; -+ def : Builtin<"release_event", [Void, ClkEvent]>; -+ def : Builtin<"create_user_event", [ClkEvent]>; -+ def : Builtin<"is_valid_event", [Bool, ClkEvent]>; -+ def : Builtin<"set_user_event_status", [Void, ClkEvent, Int]>; -+ // TODO: capture_event_profiling_info - -+ // --- Table 35 --- -+ def : Builtin<"get_default_queue", [Queue]>; -+ // TODO: ndrange functions -+} - - //-------------------------------------------------------------------- - // End of the builtin functions defined in the OpenCL C specification. -@@ -1456,7 +1541,7 @@ let Extension = FuncExtKhrSubgroups in { - "get_sub_group_local_id"] in { - def : Builtin<name, [UInt]>; - } -- let MinVersion = CL20 in { -+ let Versions = [CL20] in { - foreach name = ["get_enqueued_num_sub_groups"] in { - def : Builtin<name, [UInt]>; - } -diff --git a/clang/lib/Sema/Sema.cpp b/clang/lib/Sema/Sema.cpp -index 2f2b52106f3d..5092a4691b9b 100644 ---- a/clang/lib/Sema/Sema.cpp -+++ b/clang/lib/Sema/Sema.cpp -@@ -291,9 +291,10 @@ void Sema::Initialize() { - // Initialize predefined OpenCL types and supported extensions and (optional) - // core features. - if (getLangOpts().OpenCL) { -+ getOpenCLOptions().setOpenCLVersion(getLangOpts()); - getOpenCLOptions().addSupport( - Context.getTargetInfo().getSupportedOpenCLOpts()); -- getOpenCLOptions().enableSupportedCore(getLangOpts()); -+ getOpenCLOptions().enableSupportedCore(); - addImplicitTypedef("sampler_t", Context.OCLSamplerTy); - addImplicitTypedef("event_t", Context.OCLEventTy); - if (getLangOpts().OpenCLCPlusPlus || getLangOpts().OpenCLVersion >= 200) { -@@ -349,12 +350,18 @@ void Sema::Initialize() { - "cl_khr_int64_base_atomics cl_khr_int64_extended_atomics"); - - setOpenCLExtensionForType(AtomicDoubleT, "cl_khr_fp64"); -+ setOpenCLExtensionForType(Context.OCLReserveIDTy, "__opencl_c_pipes"); -+ setOpenCLExtensionForType(Context.OCLClkEventTy, -+ "__opencl_c_device_enqueue"); -+ setOpenCLExtensionForType(Context.OCLQueueTy, -+ "__opencl_c_device_enqueue"); - } - - setOpenCLExtensionForType(Context.DoubleTy, "cl_khr_fp64"); - --#define GENERIC_IMAGE_TYPE_EXT(Type, Id, Ext) \ -- setOpenCLExtensionForType(Context.Id, Ext); -+#define GENERIC_IMAGE_TYPE_EXT(Type, Id, Ext) \ -+ setOpenCLExtensionForType(Context.Id, Ext); \ -+ setOpenCLExtensionForType(Context.Id, "__opencl_c_images"); - #include "clang/Basic/OpenCLImageTypes.def" - #define EXT_OPAQUE_TYPE(ExtType, Id, Ext) \ - addImplicitTypedef(#ExtType, Context.Id##Ty); \ -@@ -2423,6 +2430,27 @@ bool Sema::isOpenCLDisabledDecl(Decl *FD) { - return false; - } - -+template <typename DiagLocT, typename DiagInfoT> -+void Sema::DiagnoseOpenCLRequiresOption(llvm::StringRef OpenCLOptName, -+ DiagLocT DiagLoc, DiagInfoT DiagInfo, -+ unsigned Selector, -+ SourceRange SrcRange) { -+ const auto &LO = getLangOpts(); -+ auto CLVer = LO.OpenCLCPlusPlus ? 200 : LO.OpenCLVersion; -+ // For versions higher that 3.0 diagnosing feature -+ if (CLVer >= 300) { -+ OpenCLOptName = -+ llvm::StringSwitch<llvm::StringRef>(OpenCLOptName) -+ .Case("cl_khr_3d_image_writes", "__opencl_c_3d_image_writes") -+ .Case("cl_khr_subgroups", "__opencl_c_subgroups") -+ .Case("cl_khr_fp64", "__opencl_c_fp64") -+ .Default(OpenCLOptName); -+ } -+ -+ Diag(DiagLoc, diag::err_opencl_requires_extension) -+ << Selector << DiagInfo << OpenCLOptName << SrcRange; -+} -+ - template <typename T, typename DiagLocT, typename DiagInfoT, typename MapT> - bool Sema::checkOpenCLDisabledTypeOrDecl(T D, DiagLocT DiagLoc, - DiagInfoT DiagInfo, MapT &Map, -@@ -2434,8 +2462,7 @@ bool Sema::checkOpenCLDisabledTypeOrDecl(T D, DiagLocT DiagLoc, - bool Disabled = false; - for (auto &I : Loc->second) { - if (I != CurrOpenCLExtension && !getOpenCLOptions().isEnabled(I)) { -- Diag(DiagLoc, diag::err_opencl_requires_extension) << Selector << DiagInfo -- << I << SrcRange; -+ DiagnoseOpenCLRequiresOption(I, DiagLoc, DiagInfo, Selector, SrcRange); - Disabled = true; - } - } -@@ -2471,3 +2498,13 @@ bool Sema::checkOpenCLDisabledDecl(const NamedDecl &D, const Expr &E) { - return checkOpenCLDisabledTypeOrDecl(&D, E.getBeginLoc(), FnName, - OpenCLDeclExtMap, 1, D.getSourceRange()); - } -+ -+bool Sema::checkOpenCLSubgroupExtForCallExpr(CallExpr *Call) { -+ if (!getOpenCLOptions().isEnabled("cl_khr_subgroups")) { -+ DiagnoseOpenCLRequiresOption("cl_khr_subgroups", Call->getBeginLoc(), -+ Call->getDirectCallee(), 1, -+ Call->getSourceRange()); -+ return true; -+ } -+ return false; -+} -diff --git a/clang/lib/Sema/SemaChecking.cpp b/clang/lib/Sema/SemaChecking.cpp -index 1bf04d9cb4f2..9dd63db0f420 100644 ---- a/clang/lib/Sema/SemaChecking.cpp -+++ b/clang/lib/Sema/SemaChecking.cpp -@@ -835,20 +835,11 @@ static bool checkOpenCLBlockArgs(Sema &S, Expr *BlockArg) { - return IllegalParams; - } - --static bool checkOpenCLSubgroupExt(Sema &S, CallExpr *Call) { -- if (!S.getOpenCLOptions().isEnabled("cl_khr_subgroups")) { -- S.Diag(Call->getBeginLoc(), diag::err_opencl_requires_extension) -- << 1 << Call->getDirectCallee() << "cl_khr_subgroups"; -- return true; -- } -- return false; --} -- - static bool SemaOpenCLBuiltinNDRangeAndBlock(Sema &S, CallExpr *TheCall) { - if (checkArgCount(S, TheCall, 2)) - return true; - -- if (checkOpenCLSubgroupExt(S, TheCall)) -+ if (S.checkOpenCLSubgroupExtForCallExpr(TheCall)) - return true; - - // First argument is an ndrange_t type. -@@ -1883,7 +1874,7 @@ Sema::CheckBuiltinFunctionCall(FunctionDecl *FDecl, unsigned BuiltinID, - break; - case Builtin::BIsub_group_reserve_read_pipe: - case Builtin::BIsub_group_reserve_write_pipe: -- if (checkOpenCLSubgroupExt(*this, TheCall) || -+ if (checkOpenCLSubgroupExtForCallExpr(TheCall) || - SemaBuiltinReserveRWPipe(*this, TheCall)) - return ExprError(); - break; -@@ -1896,7 +1887,7 @@ Sema::CheckBuiltinFunctionCall(FunctionDecl *FDecl, unsigned BuiltinID, - break; - case Builtin::BIsub_group_commit_read_pipe: - case Builtin::BIsub_group_commit_write_pipe: -- if (checkOpenCLSubgroupExt(*this, TheCall) || -+ if (checkOpenCLSubgroupExtForCallExpr(TheCall) || - SemaBuiltinCommitRWPipe(*this, TheCall)) - return ExprError(); - break; -@@ -4149,6 +4140,20 @@ DiagnoseCStringFormatDirectiveInCFAPI(Sema &S, - } - } - -+bool Sema::isSupportedOpenCLOMemoryOrdering(int64_t Ordering) const { -+ assert(llvm::isValidAtomicOrderingCABI(Ordering)); -+ auto OrderingCABI = (llvm::AtomicOrderingCABI)Ordering; -+ switch (OrderingCABI) { -+ case llvm::AtomicOrderingCABI::acquire: -+ case llvm::AtomicOrderingCABI::release: -+ case llvm::AtomicOrderingCABI::acq_rel: -+ return OpenCLFeatures.isEnabled("__opencl_c_atomic_order_acq_rel"); -+ -+ default: -+ return true; -+ } -+} -+ - /// Determine whether the given type has a non-null nullability annotation. - static bool isNonNullType(ASTContext &ctx, QualType type) { - if (auto nullability = type->getNullability(ctx)) -@@ -4920,10 +4925,17 @@ ExprResult Sema::BuildAtomicExpr(SourceRange CallRange, SourceRange ExprRange, - if (SubExprs.size() >= 2 && Form != Init) { - llvm::APSInt Result(32); - if (SubExprs[1]->isIntegerConstantExpr(Result, Context) && -- !isValidOrderingForOp(Result.getSExtValue(), Op)) -+ !isValidOrderingForOp(Result.getSExtValue(), Op)) { - Diag(SubExprs[1]->getBeginLoc(), - diag::warn_atomic_op_has_invalid_memory_order) - << SubExprs[1]->getSourceRange(); -+ } else if (IsOpenCL && -+ !isSupportedOpenCLOMemoryOrdering(Result.getSExtValue())) { -+ Diag(SubExprs[1]->getBeginLoc(), -+ diag::err_opencl_memory_ordering_require_feat) -+ << SubExprs[1]->getSourceRange(); -+ return ExprError(); -+ } - } - - if (auto ScopeModel = AtomicExpr::getScopeModel(Op)) { -diff --git a/clang/lib/Sema/SemaDecl.cpp b/clang/lib/Sema/SemaDecl.cpp -index 5b0417fa8859..2dee60763d57 100644 ---- a/clang/lib/Sema/SemaDecl.cpp -+++ b/clang/lib/Sema/SemaDecl.cpp -@@ -6290,7 +6290,9 @@ void Sema::deduceOpenCLAddressSpace(ValueDecl *Decl) { - if (Type->isSamplerT() || Type->isVoidType()) - return; - LangAS ImplAS = LangAS::opencl_private; -- if ((getLangOpts().OpenCLCPlusPlus || getLangOpts().OpenCLVersion >= 200) && -+ if ((getLangOpts().OpenCLCPlusPlus || -+ getOpenCLOptions().isEnabled( -+ "__opencl_c_program_scope_global_variables")) && - Var->hasGlobalStorage()) - ImplAS = LangAS::opencl_global; - // If the original type from a decayed type is an array type and that array -@@ -7849,18 +7851,22 @@ void Sema::CheckVariableDeclarationType(VarDecl *NewVD) { - // OpenCL C v2.0 s6.5.1 - Variables defined at program scope and static - // variables inside a function can also be declared in the global - // address space. -+ // OpenCL C v3.0 s5.5 - For OpenCL C 2.0, or with the -+ // __opencl_c_program_scope_global_variables feature macro, the -+ // address space for a variable at program scope or a static variable -+ // also be __global - // C++ for OpenCL inherits rule from OpenCL C v2.0. - // FIXME: Adding local AS in C++ for OpenCL might make sense. - if (NewVD->isFileVarDecl() || NewVD->isStaticLocal() || - NewVD->hasExternalStorage()) { -- if (!T->isSamplerT() && -- !T->isDependentType() && -+ if (!T->isSamplerT() && !T->isDependentType() && - !(T.getAddressSpace() == LangAS::opencl_constant || - (T.getAddressSpace() == LangAS::opencl_global && -- (getLangOpts().OpenCLVersion == 200 || -- getLangOpts().OpenCLCPlusPlus)))) { -+ (OpenCLFeatures.isEnabled( -+ "__opencl_c_program_scope_global_variables"))))) { - int Scope = NewVD->isStaticLocal() | NewVD->hasExternalStorage() << 1; -- if (getLangOpts().OpenCLVersion == 200 || getLangOpts().OpenCLCPlusPlus) -+ if (OpenCLFeatures.isEnabled( -+ "__opencl_c_program_scope_global_variables")) - Diag(NewVD->getLocation(), diag::err_opencl_global_invalid_addr_space) - << Scope << "global or constant"; - else -diff --git a/clang/lib/Sema/SemaDeclAttr.cpp b/clang/lib/Sema/SemaDeclAttr.cpp -index a9a2a19b4797..73b35eae176a 100644 ---- a/clang/lib/Sema/SemaDeclAttr.cpp -+++ b/clang/lib/Sema/SemaDeclAttr.cpp -@@ -6488,7 +6488,7 @@ static void handleInternalLinkageAttr(Sema &S, Decl *D, const ParsedAttr &AL) { - } - - static void handleOpenCLNoSVMAttr(Sema &S, Decl *D, const ParsedAttr &AL) { -- if (S.LangOpts.OpenCLVersion != 200) -+ if (S.LangOpts.OpenCLVersion < 200) - S.Diag(AL.getLoc(), diag::err_attribute_requires_opencl_version) - << AL << "2.0" << 0; - else -@@ -6572,6 +6572,13 @@ static void handleOpenCLAccessAttr(Sema &S, Decl *D, const ParsedAttr &AL) { - << AL << PDecl->getType() << DeclTy->isImageType(); - D->setInvalidDecl(true); - return; -+ } else if ((!S.getLangOpts().OpenCLCPlusPlus && -+ S.getLangOpts().OpenCLVersion >= 200) && -+ !S.getOpenCLOptions().isEnabled( -+ "__opencl_c_read_write_images")) { -+ S.Diag(AL.getLoc(), diag::err_opencl_requires_extension) -+ << 0 << PDecl->getType() << "__opencl_c_read_write_images"; -+ return; - } - } - } -diff --git a/clang/lib/Sema/SemaDeclCXX.cpp b/clang/lib/Sema/SemaDeclCXX.cpp -index 22bf35dbd0cb..b2081bb6b407 100644 ---- a/clang/lib/Sema/SemaDeclCXX.cpp -+++ b/clang/lib/Sema/SemaDeclCXX.cpp -@@ -38,9 +38,9 @@ - #include "clang/Sema/ScopeInfo.h" - #include "clang/Sema/SemaInternal.h" - #include "clang/Sema/Template.h" -+#include "llvm/ADT/STLExtras.h" - #include "llvm/ADT/ScopeExit.h" - #include "llvm/ADT/SmallString.h" --#include "llvm/ADT/STLExtras.h" - #include "llvm/ADT/StringExtras.h" - #include <map> - #include <set> -@@ -15228,6 +15228,11 @@ CheckOperatorNewDeleteTypes(Sema &SemaRef, const FunctionDecl *FnDecl, - if (auto *PtrTy = ResultType->getAs<PointerType>()) { - ResultType = RemoveAddressSpaceFromPtr(SemaRef, PtrTy); - } -+ if (CanQual<PointerType> ExpectedPtrTy = -+ ExpectedResultType->getAs<PointerType>()) { -+ ExpectedResultType = SemaRef.Context.getCanonicalType( -+ RemoveAddressSpaceFromPtr(SemaRef, ExpectedPtrTy->getTypePtr())); -+ } - } - - // Check that the result type is what we expect. -@@ -15261,6 +15266,11 @@ CheckOperatorNewDeleteTypes(Sema &SemaRef, const FunctionDecl *FnDecl, - FnDecl->getParamDecl(0)->getType()->getAs<PointerType>()) { - FirstParamType = RemoveAddressSpaceFromPtr(SemaRef, PtrTy); - } -+ if (CanQual<PointerType> ExpectedPtrTy = -+ ExpectedFirstParamType->getAs<PointerType>()) { -+ ExpectedFirstParamType = SemaRef.Context.getCanonicalType( -+ RemoveAddressSpaceFromPtr(SemaRef, ExpectedPtrTy->getTypePtr())); -+ } - } - - // Check that the first parameter type is what we expect. -diff --git a/clang/lib/Sema/SemaLookup.cpp b/clang/lib/Sema/SemaLookup.cpp -index bcbecd545398..40eaebbc9495 100644 ---- a/clang/lib/Sema/SemaLookup.cpp -+++ b/clang/lib/Sema/SemaLookup.cpp -@@ -775,19 +775,20 @@ static void InsertOCLBuiltinDeclarationsFromTable(Sema &S, LookupResult &LR, - // as argument. Only meaningful for generic types, otherwise equals 1. - unsigned GenTypeMaxCnt; - -+ ASTContext &Context = S.Context; -+ -+ // Ignore this BIF if its version does not match the language options. -+ unsigned OpenCLVersion = Context.getLangOpts().OpenCLVersion; -+ if (Context.getLangOpts().OpenCLCPlusPlus) -+ OpenCLVersion = 200; -+ -+ unsigned short VersionCode = OpenCLOptions::EncodeVersion(OpenCLVersion); -+ - for (unsigned SignatureIndex = 0; SignatureIndex < Len; SignatureIndex++) { - const OpenCLBuiltinStruct &OpenCLBuiltin = - BuiltinTable[FctIndex + SignatureIndex]; -- ASTContext &Context = S.Context; - -- // Ignore this BIF if its version does not match the language options. -- unsigned OpenCLVersion = Context.getLangOpts().OpenCLVersion; -- if (Context.getLangOpts().OpenCLCPlusPlus) -- OpenCLVersion = 200; -- if (OpenCLVersion < OpenCLBuiltin.MinVersion) -- continue; -- if ((OpenCLBuiltin.MaxVersion != 0) && -- (OpenCLVersion >= OpenCLBuiltin.MaxVersion)) -+ if (!(OpenCLBuiltin.AllVersions & VersionCode)) - continue; - - SmallVector<QualType, 1> RetTypes; -diff --git a/clang/lib/Sema/SemaType.cpp b/clang/lib/Sema/SemaType.cpp -index cc151a048b98..f380a5656a57 100644 ---- a/clang/lib/Sema/SemaType.cpp -+++ b/clang/lib/Sema/SemaType.cpp -@@ -2042,7 +2042,7 @@ static QualType deduceOpenCLPointeeAddrSpace(Sema &S, QualType PointeeType) { - !PointeeType.hasAddressSpace()) - PointeeType = S.getASTContext().getAddrSpaceQualType( - PointeeType, -- S.getLangOpts().OpenCLCPlusPlus || S.getLangOpts().OpenCLVersion == 200 -+ S.getOpenCLOptions().isEnabled("__opencl_c_generic_address_space") - ? LangAS::opencl_generic - : LangAS::opencl_private); - return PointeeType; -@@ -5316,9 +5316,15 @@ static TypeSourceInfo *GetFullTypeForDeclarator(TypeProcessingState &state, - } - - case DeclaratorChunk::Pipe: { -- T = S.BuildReadPipeType(T, DeclType.Loc); -- processTypeAttrs(state, T, TAL_DeclSpec, -- D.getMutableDeclSpec().getAttributes()); -+ if (S.getOpenCLOptions().isEnabled("__opencl_c_pipes")) { -+ T = S.BuildReadPipeType(T, DeclType.Loc); -+ processTypeAttrs(state, T, TAL_DeclSpec, -+ D.getMutableDeclSpec().getAttributes()); -+ } else { -+ D.setInvalidType(true); -+ T = Context.IntTy; -+ S.Diag(D.getIdentifierLoc(), diag::err_opencl_pipes_require_feat); -+ } - break; - } - } -diff --git a/clang/test/CodeGenOpenCL/addr-space-struct-arg.cl b/clang/test/CodeGenOpenCL/addr-space-struct-arg.cl -index e1f3f6fe1419..b43d9df5935c 100644 ---- a/clang/test/CodeGenOpenCL/addr-space-struct-arg.cl -+++ b/clang/test/CodeGenOpenCL/addr-space-struct-arg.cl -@@ -2,7 +2,8 @@ - // RUN: %clang_cc1 %s -emit-llvm -o - -O0 -triple amdgcn | FileCheck -enable-var-scope -check-prefixes=ALL,AMDGCN %s - // RUN: %clang_cc1 %s -emit-llvm -o - -cl-std=CL2.0 -O0 -triple amdgcn | FileCheck -enable-var-scope -check-prefixes=ALL,AMDGCN,AMDGCN20 %s - // RUN: %clang_cc1 %s -emit-llvm -o - -cl-std=CL1.2 -O0 -triple spir-unknown-unknown-unknown | FileCheck -enable-var-scope -check-prefixes=SPIR %s -- -+// RUN: %clang_cc1 %s -emit-llvm -o - -cl-std=CL3.0 -O0 -triple spir-unknown-unknown-unknown | FileCheck -enable-var-scope -check-prefixes=SPIR %s -+// RUN: %clang_cc1 %s -emit-llvm -o - -cl-std=CL3.0 -cl-ext=__opencl_c_program_scope_global_variables -O0 -triple spir-unknown-unknown-unknown | FileCheck -enable-var-scope -check-prefixes=SPIR %s - typedef int int2 __attribute__((ext_vector_type(2))); - - typedef struct { -@@ -39,7 +40,7 @@ struct LargeStructTwoMember { - int2 y[20]; - }; - --#if __OPENCL_C_VERSION__ >= 200 -+#ifdef __opencl_c_program_scope_global_variables - struct LargeStructOneMember g_s; - #endif - -@@ -98,7 +99,7 @@ void FuncOneLargeMember(struct LargeStructOneMember u) { - // AMDGCN20: %[[r0:.*]] = bitcast %struct.LargeStructOneMember addrspace(5)* %[[byval_temp]] to i8 addrspace(5)* - // AMDGCN20: call void @llvm.memcpy.p5i8.p1i8.i64(i8 addrspace(5)* align 8 %[[r0]], i8 addrspace(1)* align 8 bitcast (%struct.LargeStructOneMember addrspace(1)* @g_s to i8 addrspace(1)*), i64 800, i1 false) - // AMDGCN20: call void @FuncOneLargeMember(%struct.LargeStructOneMember addrspace(5)* byval(%struct.LargeStructOneMember) align 8 %[[byval_temp]]) --#if __OPENCL_C_VERSION__ >= 200 -+#ifdef __opencl_c_program_scope_global_variables - void test_indirect_arg_globl(void) { - FuncOneLargeMember(g_s); - } -diff --git a/clang/test/CodeGenOpenCL/address-spaces-conversions.cl b/clang/test/CodeGenOpenCL/address-spaces-conversions.cl -index 52feccc540bb..9ecffcca5ee9 100644 ---- a/clang/test/CodeGenOpenCL/address-spaces-conversions.cl -+++ b/clang/test/CodeGenOpenCL/address-spaces-conversions.cl -@@ -1,5 +1,7 @@ - // RUN: %clang_cc1 %s -triple x86_64-unknown-linux-gnu -O0 -ffake-address-space-map -cl-std=CL2.0 -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 %s -triple x86_64-unknown-linux-gnu -O0 -ffake-address-space-map -cl-std=CL3.0 -cl-ext=__opencl_c_generic_address_space -emit-llvm -o - | FileCheck %s - // RUN: %clang_cc1 %s -triple x86_64-unknown-linux-gnu -O0 -cl-std=CL2.0 -emit-llvm -o - | FileCheck --check-prefix=CHECK-NOFAKE %s -+// RUN: %clang_cc1 %s -triple x86_64-unknown-linux-gnu -O0 -cl-std=CL3.0 -cl-ext=__opencl_c_generic_address_space -emit-llvm -o - | FileCheck --check-prefix=CHECK-NOFAKE %s - // When -ffake-address-space-map is not used, all addr space mapped to 0 for x86_64. - - // test that we generate address space casts everywhere we need conversions of -diff --git a/clang/test/CodeGenOpenCL/address-spaces-mangling.cl b/clang/test/CodeGenOpenCL/address-spaces-mangling.cl -index 50622f099143..e19ec8451d0d 100644 ---- a/clang/test/CodeGenOpenCL/address-spaces-mangling.cl -+++ b/clang/test/CodeGenOpenCL/address-spaces-mangling.cl -@@ -6,6 +6,7 @@ - // We check that the address spaces are mangled the same in both version of OpenCL - // RUN: %clang_cc1 %s -triple spir-unknown-unknown -cl-std=CL2.0 -emit-llvm -o - | FileCheck -check-prefix=OCL-20 %s - // RUN: %clang_cc1 %s -triple spir-unknown-unknown -cl-std=CL1.2 -emit-llvm -o - | FileCheck -check-prefix=OCL-12 %s -+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -cl-std=CL3.0 -emit-llvm -o - | FileCheck -check-prefix=OCL-12 %s - - // We can't name this f as private is equivalent to default - // no specifier given address space so we get multiple definition -@@ -47,7 +48,7 @@ void f(constant int *arg) { } - // OCL-20-DAG: @_Z1fPU3AS2i - // OCL-12-DAG: @_Z1fPU3AS2i - --#if __OPENCL_C_VERSION__ >= 200 -+#if __OPENCL_C_VERSION__ == 200 - __attribute__((overloadable)) - void f(generic int *arg) { } - // ASMANG20: @_Z1fPU3AS4i -diff --git a/clang/test/CodeGenOpenCL/address-spaces.cl b/clang/test/CodeGenOpenCL/address-spaces.cl -index 144466a690b3..b150e1080b53 100644 ---- a/clang/test/CodeGenOpenCL/address-spaces.cl -+++ b/clang/test/CodeGenOpenCL/address-spaces.cl -@@ -1,9 +1,13 @@ - // RUN: %clang_cc1 %s -O0 -ffake-address-space-map -emit-llvm -o - | FileCheck %s --check-prefixes=CHECK,SPIR -+// RUN: %clang_cc1 %s -O0 -cl-std=CL3.0 -ffake-address-space-map -emit-llvm -o - | FileCheck %s --check-prefixes=CHECK,SPIR - // RUN: %clang_cc1 %s -O0 -DCL20 -cl-std=CL2.0 -ffake-address-space-map -emit-llvm -o - | FileCheck %s --check-prefixes=CL20,CL20SPIR - // RUN: %clang_cc1 %s -O0 -triple amdgcn-amd-amdhsa -emit-llvm -o - | FileCheck --check-prefixes=CHECK,AMDGCN %s -+// RUN: %clang_cc1 %s -O0 -cl-std=CL3.0 -triple amdgcn-amd-amdhsa -emit-llvm -o - | FileCheck --check-prefixes=CHECK,AMDGCN %s - // RUN: %clang_cc1 %s -O0 -triple amdgcn-amd-amdhsa -DCL20 -cl-std=CL2.0 -emit-llvm -o - | FileCheck %s --check-prefixes=CL20,CL20AMDGCN - // RUN: %clang_cc1 %s -O0 -triple amdgcn-mesa-mesa3d -emit-llvm -o - | FileCheck --check-prefixes=CHECK,AMDGCN %s -+// RUN: %clang_cc1 %s -O0 -cl-std=CL3.0 -triple amdgcn-mesa-mesa3d -emit-llvm -o - | FileCheck --check-prefixes=CHECK,AMDGCN %s - // RUN: %clang_cc1 %s -O0 -triple r600-- -emit-llvm -o - | FileCheck --check-prefixes=CHECK,AMDGCN %s -+// RUN: %clang_cc1 %s -O0 -cl-std=CL3.0 -triple r600-- -emit-llvm -o - | FileCheck --check-prefixes=CHECK,AMDGCN %s - - // SPIR: %struct.S = type { i32, i32, i32* } - // CL20SPIR: %struct.S = type { i32, i32, i32 addrspace(4)* } -diff --git a/clang/test/CodeGenOpenCL/amdgcn-automatic-variable.cl b/clang/test/CodeGenOpenCL/amdgcn-automatic-variable.cl -index 7216cb517495..8d3b30fe8074 100644 ---- a/clang/test/CodeGenOpenCL/amdgcn-automatic-variable.cl -+++ b/clang/test/CodeGenOpenCL/amdgcn-automatic-variable.cl -@@ -1,4 +1,5 @@ - // RUN: %clang_cc1 -O0 -cl-std=CL1.2 -triple amdgcn---amdgizcl -emit-llvm %s -o - | FileCheck -check-prefixes=CHECK,CL12 %s -+// RUN: %clang_cc1 -O0 -cl-std=CL3.0 -triple amdgcn---amdgizcl -emit-llvm %s -o - | FileCheck -check-prefixes=CHECK,CL12 %s - // RUN: %clang_cc1 -O0 -cl-std=CL2.0 -triple amdgcn---amdgizcl -emit-llvm %s -o - | FileCheck -check-prefixes=CHECK,CL20 %s - - // CL12-LABEL: define void @func1(i32 addrspace(5)* %x) -diff --git a/clang/test/CodeGenOpenCL/amdgpu-sizeof-alignof.cl b/clang/test/CodeGenOpenCL/amdgpu-sizeof-alignof.cl -index a5d438933fa4..8073c7756eb6 100644 ---- a/clang/test/CodeGenOpenCL/amdgpu-sizeof-alignof.cl -+++ b/clang/test/CodeGenOpenCL/amdgpu-sizeof-alignof.cl -@@ -4,6 +4,17 @@ - // RUN: %clang_cc1 -triple amdgcn---opencl -cl-std=CL2.0 %s -emit-llvm -o - | FileCheck %s - // RUN: %clang_cc1 -triple amdgcn---amdgizcl -cl-std=CL1.2 %s -emit-llvm -o - | FileCheck %s - // RUN: %clang_cc1 -triple amdgcn---amdgizcl -cl-std=CL2.0 %s -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 -triple r600 -cl-std=CL3.0 %s -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 -triple amdgcn-mesa-mesa3d -cl-std=CL3.0 %s -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 -triple amdgcn---opencl -cl-std=CL3.0 %s -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 -triple amdgcn---amdgizcl -cl-std=CL3.0 %s -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 -triple amdgcn-mesa-mesa3d -cl-ext=__opencl_c_generic_address_space -cl-std=CL3.0 %s -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 -triple amdgcn---opencl -cl-ext=__opencl_c_generic_address_space -cl-std=CL3.0 %s -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 -triple amdgcn---amdgizcl -cl-ext=__opencl_c_generic_address_space -cl-std=CL3.0 %s -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 -triple r600 -cl-ext=__opencl_c_fp64 -cl-std=CL3.0 %s -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 -triple amdgcn-mesa-mesa3d -cl-ext=__opencl_c_fp64 -cl-std=CL3.0 %s -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 -triple amdgcn---opencl -cl-ext=__opencl_c_fp64 -cl-std=CL3.0 %s -emit-llvm -o - | FileCheck %s -+// RUN: %clang_cc1 -triple amdgcn---amdgizcl -cl-ext=__opencl_c_fp64 -cl-std=CL3.0 %s -emit-llvm -o - | FileCheck %s - - #ifdef __AMDGCN__ - #define PTSIZE 8 -@@ -11,7 +22,7 @@ - #define PTSIZE 4 - #endif - --#ifdef cl_khr_fp64 -+#if defined(cl_khr_fp64) || defined(__opencl_c_fp64) - #pragma OPENCL EXTENSION cl_khr_fp64 : enable - #endif - #ifdef cl_khr_fp16 -@@ -59,8 +70,12 @@ void test() { - check(__alignof__(double) == 8); - #endif - -- check(sizeof(void*) == (__OPENCL_C_VERSION__ >= 200 ? 8 : 4)); -- check(__alignof__(void*) == (__OPENCL_C_VERSION__ >= 200 ? 8 : 4)); -+ check(sizeof(private void*) == 4); -+ check(__alignof__(private void*) == 4); -+#ifdef __opencl_c_generic_address_space -+ check(sizeof(generic void*) == 8); -+ check(__alignof__(generic void*) == 8); -+#endif - check(sizeof(global_ptr_t) == PTSIZE); - check(__alignof__(global_ptr_t) == PTSIZE); - check(sizeof(constant_ptr_t) == PTSIZE); -diff --git a/clang/test/CodeGenOpenCL/arm-integer-dot-product.cl b/clang/test/CodeGenOpenCL/arm-integer-dot-product.cl -index d1ab6aceac5c..70c5bace023b 100644 ---- a/clang/test/CodeGenOpenCL/arm-integer-dot-product.cl -+++ b/clang/test/CodeGenOpenCL/arm-integer-dot-product.cl -@@ -1,4 +1,5 @@ - // RUN: %clang_cc1 %s -triple spir-unknown-unknown -finclude-default-header -cl-std=CL1.2 -emit-llvm -o - -O0 | FileCheck %s -+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -finclude-default-header -cl-std=CL3.0 -emit-llvm -o - -O0 | FileCheck %s - - #pragma OPENCL EXTENSION cl_arm_integer_dot_product_int8 : enable - void test_int8(uchar4 ua, uchar4 ub, char4 sa, char4 sb) { -diff --git a/clang/test/CodeGenOpenCL/cl-uniform-wg-size.cl b/clang/test/CodeGenOpenCL/cl-uniform-wg-size.cl -index 76ace5dca21e..5dc43e222f75 100644 ---- a/clang/test/CodeGenOpenCL/cl-uniform-wg-size.cl -+++ b/clang/test/CodeGenOpenCL/cl-uniform-wg-size.cl -@@ -1,6 +1,8 @@ - // RUN: %clang_cc1 -emit-llvm -O0 -cl-std=CL1.2 -o - %s 2>&1 | FileCheck %s -check-prefixes CHECK,CHECK-UNIFORM - // RUN: %clang_cc1 -emit-llvm -O0 -cl-std=CL2.0 -o - %s 2>&1 | FileCheck %s -check-prefixes CHECK,CHECK-NONUNIFORM - // RUN: %clang_cc1 -emit-llvm -O0 -cl-std=CL2.0 -cl-uniform-work-group-size -o - %s 2>&1 | FileCheck %s -check-prefixes CHECK,CHECK-UNIFORM -+// RUN: %clang_cc1 -emit-llvm -O0 -cl-std=CL3.0 -o - %s 2>&1 | FileCheck %s -check-prefixes CHECK,CHECK-NONUNIFORM -+// RUN: %clang_cc1 -emit-llvm -O0 -cl-std=CL3.0 -cl-uniform-work-group-size -o - %s 2>&1 | FileCheck %s -check-prefixes CHECK,CHECK-UNIFORM - - kernel void ker() {}; - // CHECK: define{{.*}}@ker() #0 -diff --git a/clang/test/CodeGenOpenCL/fpmath.cl b/clang/test/CodeGenOpenCL/fpmath.cl -index 0108d909c94e..b28392739c71 100644 ---- a/clang/test/CodeGenOpenCL/fpmath.cl -+++ b/clang/test/CodeGenOpenCL/fpmath.cl -@@ -2,6 +2,8 @@ - // RUN: %clang_cc1 %s -emit-llvm -o - -triple spir-unknown-unknown -cl-fp32-correctly-rounded-divide-sqrt | FileCheck --check-prefix=CHECK --check-prefix=DIVOPT %s - // RUN: %clang_cc1 %s -emit-llvm -o - -DNOFP64 -cl-std=CL1.2 -triple r600-unknown-unknown -target-cpu r600 -pedantic | FileCheck --check-prefix=CHECK-FLT %s - // RUN: %clang_cc1 %s -emit-llvm -o - -DFP64 -cl-std=CL1.2 -triple spir-unknown-unknown -pedantic | FileCheck --check-prefix=CHECK-DBL %s -+// RUN: %clang_cc1 %s -emit-llvm -o - -DNOFP64 -cl-std=CL3.0 -triple r600-unknown-unknown -target-cpu r600 -pedantic | FileCheck --check-prefix=CHECK-FLT %s -+// RUN: %clang_cc1 %s -emit-llvm -o - -DFP64 -cl-std=CL3.0 -cl-ext=__opencl_c_fp64 -triple spir-unknown-unknown -pedantic | FileCheck --check-prefix=CHECK-DBL %s - - typedef __attribute__(( ext_vector_type(4) )) float float4; - -diff --git a/clang/test/CodeGenOpenCL/generic-address-space-feature.cl b/clang/test/CodeGenOpenCL/generic-address-space-feature.cl -new file mode 100644 -index 000000000000..890860abe4d9 ---- /dev/null -+++ b/clang/test/CodeGenOpenCL/generic-address-space-feature.cl -@@ -0,0 +1,28 @@ -+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -O0 -cl-std=CL1.2 -cl-ext=-cl_khr_fp64 -finclude-default-header -emit-llvm -o - | FileCheck %s --check-prefixes=CL12 -+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -O0 -cl-std=CL1.2 -cl-ext=-cl_khr_fp64 -fdeclare-opencl-builtins -emit-llvm -o - | FileCheck %s --check-prefixes=CL12 -+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -O0 -cl-std=CL2.0 -cl-ext=-cl_khr_fp64 -finclude-default-header -emit-llvm -o - | FileCheck %s --check-prefixes=CL20 -+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -O0 -cl-std=CL2.0 -cl-ext=-cl_khr_fp64 -fdeclare-opencl-builtins -emit-llvm -o - | FileCheck %s --check-prefixes=CL20 -+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -O0 -cl-std=CL3.0 -cl-ext=-cl_khr_fp64 -finclude-default-header -emit-llvm -o - | FileCheck %s --check-prefixes=CL30 -+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -O0 -cl-std=CL3.0 -cl-ext=-cl_khr_fp64,__opencl_c_generic_address_space -finclude-default-header -emit-llvm -o - | FileCheck %s --check-prefixes=CL30-GENERIC -+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -O0 -cl-std=CL3.0 -cl-ext=-cl_khr_fp64 -fdeclare-opencl-builtins -emit-llvm -o - | FileCheck %s --check-prefixes=CL30 -+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -O0 -cl-std=CL3.0 -cl-ext=-cl_khr_fp64,__opencl_c_generic_address_space -fdeclare-opencl-builtins -emit-llvm -o - | FileCheck %s --check-prefixes=CL30-GENERIC -+ -+void test(global float* src1, local float *src2, private float *src3, float *src4, float tmp) { -+ // CL20: %{{.+}} = addrspacecast float addrspace(1)* %{{.+}} to float addrspace(4)* -+ // CL12-NOT: addrspacecast -+ // CL30-NOT: addrspacecast -+ // CL30-GENERIC-NOT: addrspacecast -+ tmp = sincos(tmp, src1); -+ // CL20: %{{.+}} = addrspacecast float addrspace(3)* %{{.+}} to float addrspace(4)* -+ // CL12-NOT: addrspacecast -+ // CL30-NOT: addrspacecast -+ // CL30-GENERIC-NOT: addrspacecast -+ tmp = sincos(tmp, src2); -+ -+ // CL12: %{{.+}} = call {{(spir_func )?}}float [[FUNCNAME:@.*sincos.*]](float %{{.+}}, float* {{.+}}) -+ // CL20: %{{.+}} = call {{(spir_func )?}}float [[FUNCNAME:@.*sincos.*]](float %{{.+}}, float addrspace(4)* {{.+}}) -+ // CL30: %{{.+}} = call {{(spir_func )?}}float [[FUNCNAME:@.*sincos.*]](float %{{.+}}, float* {{.+}}) -+ // CL30-GENERIC: %{{.+}} = call {{(spir_func )?}}float [[FUNCNAME:@.*sincos.*]](float %{{.+}}, float addrspace(4)* {{.+}}) -+ // CHECK: addrspacecast -+ tmp = sincos(tmp, src4); -+} -diff --git a/clang/test/CodeGenOpenCL/intel-subgroups-avc-ext-types.cl b/clang/test/CodeGenOpenCL/intel-subgroups-avc-ext-types.cl -index 515f13f6e768..5aa31ac6f345 100644 ---- a/clang/test/CodeGenOpenCL/intel-subgroups-avc-ext-types.cl -+++ b/clang/test/CodeGenOpenCL/intel-subgroups-avc-ext-types.cl -@@ -1,4 +1,5 @@ - // RUN: %clang_cc1 %s -triple spir-unknown-unknown -cl-std=CL1.2 -cl-ext=+cl_intel_device_side_avc_motion_estimation -emit-llvm -o - -O0 | FileCheck %s -+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -cl-std=CL3.0 -cl-ext=+cl_intel_device_side_avc_motion_estimation -emit-llvm -o - -O0 | FileCheck %s - - // CHECK: %opencl.intel_sub_group_avc_mce_payload_t = type opaque - // CHECK: %opencl.intel_sub_group_avc_ime_payload_t = type opaque -diff --git a/clang/test/CodeGenOpenCL/kernels-have-spir-cc-by-default.cl b/clang/test/CodeGenOpenCL/kernels-have-spir-cc-by-default.cl -index e89237623478..3d6708ac361f 100644 ---- a/clang/test/CodeGenOpenCL/kernels-have-spir-cc-by-default.cl -+++ b/clang/test/CodeGenOpenCL/kernels-have-spir-cc-by-default.cl -@@ -1,5 +1,8 @@ - // RUN: %clang_cc1 %s -cl-std=CL1.2 -emit-llvm -triple x86_64-unknown-unknown -o - | FileCheck %s - // RUN: %clang_cc1 %s -cl-std=CL1.2 -emit-llvm -triple amdgcn-unknown-unknown -o - | FileCheck -check-prefixes=AMDGCN %s -+// RUN: %clang_cc1 %s -cl-std=CL3.0 -cl-ext=__opencl_c_fp64 -emit-llvm -triple x86_64-unknown-unknown -o - | FileCheck %s -+// RUN: %clang_cc1 %s -cl-std=CL3.0 -cl-ext=__opencl_c_fp64 -emit-llvm -triple amdgcn-unknown-unknown -o - | FileCheck -check-prefixes=AMDGCN %s -+ - // Test that the kernels always use the SPIR calling convention - // to have unambiguous mapping of arguments to feasibly implement - // clSetKernelArg(). -diff --git a/clang/test/CodeGenOpenCL/logical-ops.cl b/clang/test/CodeGenOpenCL/logical-ops.cl -index f083a8580ee7..499eab65039b 100644 ---- a/clang/test/CodeGenOpenCL/logical-ops.cl -+++ b/clang/test/CodeGenOpenCL/logical-ops.cl -@@ -1,4 +1,5 @@ - // RUN: %clang_cc1 %s -emit-llvm -o - -cl-std=CL1.2 -O1 -triple x86_64-unknown-linux-gnu | FileCheck %s -+// RUN: %clang_cc1 %s -emit-llvm -o - -cl-std=CL3.0 -cl-ext=__opencl_c_fp64 -O1 -triple x86_64-unknown-linux-gnu | FileCheck %s - // RUN: %clang_cc1 %s -emit-llvm -o - -cl-std=clc++ -O1 -triple x86_64-unknown-linux-gnu | FileCheck %s - - #pragma OPENCL EXTENSION cl_khr_fp64 : enable -diff --git a/clang/test/CodeGenOpenCL/no-half.cl b/clang/test/CodeGenOpenCL/no-half.cl -index aee8f678f01a..46da7fa339e8 100644 ---- a/clang/test/CodeGenOpenCL/no-half.cl -+++ b/clang/test/CodeGenOpenCL/no-half.cl -@@ -1,6 +1,7 @@ - // RUN: %clang_cc1 %s -cl-std=cl2.0 -emit-llvm -o - -triple spir-unknown-unknown | FileCheck %s - // RUN: %clang_cc1 %s -cl-std=cl1.2 -emit-llvm -o - -triple spir-unknown-unknown | FileCheck %s - // RUN: %clang_cc1 %s -cl-std=cl1.1 -emit-llvm -o - -triple spir-unknown-unknown | FileCheck %s -+// RUN: %clang_cc1 %s -cl-std=cl3.0 -cl-ext=__opencl_c_fp64 -emit-llvm -o - -triple spir-unknown-unknown | FileCheck %s - - #pragma OPENCL EXTENSION cl_khr_fp64:enable - -diff --git a/clang/test/CodeGenOpenCL/pipe_builtin.cl b/clang/test/CodeGenOpenCL/pipe_builtin.cl -index 02b9669b7ab1..0bf35c336199 100644 ---- a/clang/test/CodeGenOpenCL/pipe_builtin.cl -+++ b/clang/test/CodeGenOpenCL/pipe_builtin.cl -@@ -1,4 +1,7 @@ - // RUN: %clang_cc1 -triple %itanium_abi_triple -emit-llvm -cl-ext=+cl_khr_subgroups -O0 -cl-std=clc++ -o - %s | FileCheck %s -+// RUN: %clang_cc1 -triple %itanium_abi_triple -emit-llvm -cl-ext=+cl_khr_subgroups -O0 -cl-std=cl2.0 -o - %s | FileCheck %s -+// RUN: %clang_cc1 -triple %itanium_abi_triple -emit-llvm -cl-ext=+cl_khr_subgroups -O0 -cl-std=cl3.0 -cl-ext=__opencl_c_pipes,__opencl_c_subgroups -o - %s | FileCheck %s -+ - // FIXME: Add MS ABI manglings of OpenCL things and remove %itanium_abi_triple - // above to support OpenCL in the MS C++ ABI. - -diff --git a/clang/test/CodeGenOpenCL/pipe_types.cl b/clang/test/CodeGenOpenCL/pipe_types.cl -index ba064c6d7557..b7a523d4f084 100644 ---- a/clang/test/CodeGenOpenCL/pipe_types.cl -+++ b/clang/test/CodeGenOpenCL/pipe_types.cl -@@ -1,4 +1,5 @@ - // RUN: %clang_cc1 -triple x86_64-unknown-linux-gnu -emit-llvm -O0 -cl-std=CL2.0 -o - %s | FileCheck %s -+// RUN: %clang_cc1 -triple x86_64-unknown-linux-gnu -emit-llvm -O0 -cl-std=CL3.0 -cl-ext=__opencl_c_pipes -o - %s | FileCheck %s - - // CHECK: %opencl.pipe_ro_t = type opaque - // CHECK: %opencl.pipe_wo_t = type opaque -diff --git a/clang/test/CodeGenOpenCL/printf.cl b/clang/test/CodeGenOpenCL/printf.cl -index fc139d776db6..0133c5595d81 100644 ---- a/clang/test/CodeGenOpenCL/printf.cl -+++ b/clang/test/CodeGenOpenCL/printf.cl -@@ -1,5 +1,7 @@ - // RUN: %clang_cc1 -cl-std=CL1.2 -cl-ext=-+cl_khr_fp64 -triple spir-unknown-unknown -disable-llvm-passes -emit-llvm -o - %s | FileCheck -check-prefixes=FP64,ALL %s - // RUN: %clang_cc1 -cl-std=CL1.2 -cl-ext=-cl_khr_fp64 -triple spir-unknown-unknown -disable-llvm-passes -emit-llvm -o - %s | FileCheck -check-prefixes=NOFP64,ALL %s -+// RUN: %clang_cc1 -cl-std=CL3.0 -cl-ext=+__opencl_c_fp64 -triple spir-unknown-unknown -disable-llvm-passes -emit-llvm -o - %s | FileCheck -check-prefixes=FP64,ALL %s -+// RUN: %clang_cc1 -cl-std=CL3.0 -triple spir-unknown-unknown -disable-llvm-passes -emit-llvm -o - %s | FileCheck -check-prefixes=NOFP64,ALL %s - - typedef __attribute__((ext_vector_type(2))) float float2; - typedef __attribute__((ext_vector_type(2))) half half2; -diff --git a/clang/test/CodeGenOpenCL/unroll-hint.cl b/clang/test/CodeGenOpenCL/unroll-hint.cl -index 0f84450a1ae6..9347c935869b 100644 ---- a/clang/test/CodeGenOpenCL/unroll-hint.cl -+++ b/clang/test/CodeGenOpenCL/unroll-hint.cl -@@ -1,5 +1,6 @@ - // RUN: %clang_cc1 -emit-llvm -O0 -cl-std=CL2.0 -o - %s | FileCheck %s - // RUN: %clang_cc1 -emit-llvm -O0 -cl-std=CL1.2 -o - %s | FileCheck %s -+// RUN: %clang_cc1 -emit-llvm -O0 -cl-std=CL3.0 -o - %s | FileCheck %s - - /*** for ***/ - void for_count() -diff --git a/clang/test/Driver/autocomplete.c b/clang/test/Driver/autocomplete.c -index 18b31320ea80..a6e7be887c8c 100644 ---- a/clang/test/Driver/autocomplete.c -+++ b/clang/test/Driver/autocomplete.c -@@ -43,6 +43,8 @@ - // CLSTDALL-NEXT: CL1.2 - // CLSTDALL-NEXT: cl2.0 - // CLSTDALL-NEXT: CL2.0 -+// CLSTDALL-NEXT: cl3.0 -+// CLSTDALL-NEXT: CL3.0 - // CLSTDALL-NEXT: clc++ - // CLSTDALL-NEXT: CLC++ - // RUN: %clang --autocomplete=-fno-sanitize-coverage=,f | FileCheck %s -check-prefix=FNOSANICOVER -diff --git a/clang/test/Driver/opencl.cl b/clang/test/Driver/opencl.cl -index 05588f2c8b81..cc0a9143ab37 100644 ---- a/clang/test/Driver/opencl.cl -+++ b/clang/test/Driver/opencl.cl -@@ -2,6 +2,7 @@ - // RUN: %clang -S -### -cl-std=CL1.1 %s 2>&1 | FileCheck --check-prefix=CHECK-CL11 %s - // RUN: %clang -S -### -cl-std=CL1.2 %s 2>&1 | FileCheck --check-prefix=CHECK-CL12 %s - // RUN: %clang -S -### -cl-std=CL2.0 %s 2>&1 | FileCheck --check-prefix=CHECK-CL20 %s -+// RUN: %clang -S -### -cl-std=CL3.0 %s 2>&1 | FileCheck --check-prefix=CHECK-CL30 %s - // RUN: %clang -S -### -cl-std=clc++ %s 2>&1 | FileCheck --check-prefix=CHECK-CLCPP %s - // RUN: %clang -S -### -cl-opt-disable %s 2>&1 | FileCheck --check-prefix=CHECK-OPT-DISABLE %s - // RUN: %clang -S -### -cl-strict-aliasing %s 2>&1 | FileCheck --check-prefix=CHECK-STRICT-ALIASING %s -@@ -22,6 +23,7 @@ - // CHECK-CL11: "-cc1" {{.*}} "-cl-std=CL1.1" - // CHECK-CL12: "-cc1" {{.*}} "-cl-std=CL1.2" - // CHECK-CL20: "-cc1" {{.*}} "-cl-std=CL2.0" -+// CHECK-CL30: "-cc1" {{.*}} "-cl-std=CL3.0" - // CHECK-CLCPP: "-cc1" {{.*}} "-cl-std=clc++" - // CHECK-OPT-DISABLE: "-cc1" {{.*}} "-cl-opt-disable" - // CHECK-STRICT-ALIASING: "-cc1" {{.*}} "-cl-strict-aliasing" -diff --git a/clang/test/Driver/unknown-std.cl b/clang/test/Driver/unknown-std.cl -index 6f371bac13ac..00209fb62556 100644 ---- a/clang/test/Driver/unknown-std.cl -+++ b/clang/test/Driver/unknown-std.cl -@@ -10,6 +10,7 @@ - // CHECK-NEXT: note: use 'cl1.1' for 'OpenCL 1.1' standard - // CHECK-NEXT: note: use 'cl1.2' for 'OpenCL 1.2' standard - // CHECK-NEXT: note: use 'cl2.0' for 'OpenCL 2.0' standard -+// CHECK-NEXT: note: use 'cl3.0' for 'OpenCL 3.0' standard - // CHECK-NEXT: note: use 'clc++' for 'C++ for OpenCL' standard - - // Make sure that no other output is present. -diff --git a/clang/test/Frontend/stdlang.c b/clang/test/Frontend/stdlang.c -index 51484999e37a..eac4632fbdd6 100644 ---- a/clang/test/Frontend/stdlang.c -+++ b/clang/test/Frontend/stdlang.c -@@ -9,6 +9,7 @@ - // RUN: %clang_cc1 -x cl -cl-std=CL1.1 -DOPENCL %s - // RUN: %clang_cc1 -x cl -cl-std=CL1.2 -DOPENCL %s - // RUN: %clang_cc1 -x cl -cl-std=CL2.0 -DOPENCL %s -+// RUN: %clang_cc1 -x cl -cl-std=CL3.0 -DOPENCL %s - // RUN: %clang_cc1 -x cl -cl-std=CLC++ -DOPENCL %s - // RUN: not %clang_cc1 -x cl -std=c99 -DOPENCL %s 2>&1 | FileCheck --check-prefix=CHECK-C99 %s - // RUN: not %clang_cc1 -x cl -cl-std=invalid -DOPENCL %s 2>&1 | FileCheck --check-prefix=CHECK-INVALID %s -diff --git a/clang/test/Headers/opencl-c-header.cl b/clang/test/Headers/opencl-c-header.cl -index 1b151ffdd16a..2716076acdcf 100644 ---- a/clang/test/Headers/opencl-c-header.cl -+++ b/clang/test/Headers/opencl-c-header.cl -@@ -1,6 +1,7 @@ - // RUN: %clang_cc1 -O0 -triple spir-unknown-unknown -internal-isystem ../../lib/Headers -include opencl-c.h -emit-llvm -o - %s -verify | FileCheck %s - // RUN: %clang_cc1 -O0 -triple spir-unknown-unknown -internal-isystem ../../lib/Headers -include opencl-c.h -emit-llvm -o - %s -verify -cl-std=CL1.1 | FileCheck %s - // RUN: %clang_cc1 -O0 -triple spir-unknown-unknown -internal-isystem ../../lib/Headers -include opencl-c.h -emit-llvm -o - %s -verify -cl-std=CL1.2 | FileCheck %s -+// RUN: %clang_cc1 -O0 -triple spir-unknown-unknown -internal-isystem ../../lib/Headers -include opencl-c.h -emit-llvm -o - %s -verify -cl-std=CL3.0 | FileCheck %s - // RUN: %clang_cc1 -O0 -triple spir-unknown-unknown -internal-isystem ../../lib/Headers -include opencl-c.h -emit-llvm -o - %s -verify -cl-std=clc++ | FileCheck %s --check-prefix=CHECK20 - - // Test including the default header as a module. -@@ -39,9 +40,11 @@ - // RUN: rm -rf %t - // RUN: mkdir -p %t - // RUN: %clang_cc1 -triple spir64-unknown-unknown -emit-llvm -o - -cl-std=CL1.2 -finclude-default-header -fmodules -fimplicit-module-maps -fmodules-cache-path=%t -ftime-report %s 2>&1 | FileCheck --check-prefix=CHECK --check-prefix=CHECK-MOD %s -+// RUN: %clang_cc1 -triple spir64-unknown-unknown -emit-llvm -o - -cl-std=CL3.0 -finclude-default-header -fmodules -fimplicit-module-maps -fmodules-cache-path=%t -ftime-report %s 2>&1 | FileCheck --check-prefix=CHECK --check-prefix=CHECK-MOD %s - // RUN: %clang_cc1 -triple amdgcn--amdhsa -O0 -emit-llvm -o - -cl-std=CL2.0 -finclude-default-header -fmodules -fimplicit-module-maps -fmodules-cache-path=%t -ftime-report %s 2>&1 | FileCheck --check-prefix=CHECK20 --check-prefix=CHECK-MOD %s - // RUN: chmod u-w %t - // RUN: %clang_cc1 -triple spir64-unknown-unknown -emit-llvm -o - -cl-std=CL1.2 -finclude-default-header -fmodules -fimplicit-module-maps -fmodules-cache-path=%t -ftime-report %s 2>&1 | FileCheck --check-prefix=CHECK --check-prefix=CHECK-MOD %s -+// RUN: %clang_cc1 -triple spir64-unknown-unknown -emit-llvm -o - -cl-std=CL3.0 -finclude-default-header -fmodules -fimplicit-module-maps -fmodules-cache-path=%t -ftime-report %s 2>&1 | FileCheck --check-prefix=CHECK --check-prefix=CHECK-MOD %s - // RUN: %clang_cc1 -triple amdgcn--amdhsa -O0 -emit-llvm -o - -cl-std=CL2.0 -finclude-default-header -fmodules -fimplicit-module-maps -fmodules-cache-path=%t -ftime-report %s 2>&1 | FileCheck --check-prefix=CHECK20 --check-prefix=CHECK-MOD %s - // RUN: chmod u+w %t - -@@ -67,7 +70,7 @@ char f(char x) { - // from OpenCL 2.0 onwards. - - // CHECK20: _Z12write_imagef14ocl_image3d_wo --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ == CL_VERSION_2_0) - void test_image3dwo(write_only image3d_t img) { - write_imagef(img, (0), (0.0f)); - } -@@ -81,7 +84,7 @@ void test_atomics(__generic volatile unsigned int* a) { - #endif - - // Verify that ATOMIC_VAR_INIT is defined. --#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ == CL_VERSION_2_0) - global atomic_int z = ATOMIC_VAR_INIT(99); - #endif //__OPENCL_C_VERSION__ - -diff --git a/clang/test/Index/pipe-size.cl b/clang/test/Index/pipe-size.cl -index 94a1255f0a48..59b76051eda1 100644 ---- a/clang/test/Index/pipe-size.cl -+++ b/clang/test/Index/pipe-size.cl -@@ -2,6 +2,13 @@ - // RUN: %clang_cc1 -x cl -O0 -cl-std=CL2.0 -emit-llvm -triple spir-unknown-unknown %s -o - | FileCheck %s --check-prefix=SPIR - // RUN: %clang_cc1 -x cl -O0 -cl-std=CL2.0 -emit-llvm -triple spir64-unknown-unknown %s -o - | FileCheck %s --check-prefix=SPIR64 - // RUN: %clang_cc1 -x cl -O0 -cl-std=CL2.0 -emit-llvm -triple amdgcn-amd-amdhsa %s -o - | FileCheck %s --check-prefix=AMDGCN -+// RUN: %clang_cc1 -x cl -O0 -cl-std=CL3.0 -cl-ext=__opencl_c_pipes -emit-llvm -triple x86_64-unknown-linux-gnu %s -o - | FileCheck %s --check-prefix=X86 -+// RUN: %clang_cc1 -x cl -O0 -cl-std=CL3.0 -cl-ext=__opencl_c_pipes -emit-llvm -triple spir-unknown-unknown %s -o - | FileCheck %s --check-prefix=SPIR -+// RUN: %clang_cc1 -x cl -O0 -cl-std=CL3.0 -cl-ext=__opencl_c_pipes -emit-llvm -triple spir64-unknown-unknown %s -o - | FileCheck %s --check-prefix=SPIR64 -+// RUN: %clang_cc1 -x cl -O0 -cl-std=CL3.0 -cl-ext=__opencl_c_pipes -emit-llvm -triple amdgcn-amd-amdhsa %s -o - | FileCheck %s --check-prefix=AMDGCN -+ -+ -+ - __kernel void testPipe( pipe int test ) - { - int s = sizeof(test); -diff --git a/clang/test/Preprocessor/init-aarch64.c b/clang/test/Preprocessor/init-aarch64.c -index df2a6128989b..cbb0995b48b0 100644 ---- a/clang/test/Preprocessor/init-aarch64.c -+++ b/clang/test/Preprocessor/init-aarch64.c -@@ -109,12 +109,12 @@ - // AARCH64-NEXT: #define __GCC_HAVE_SYNC_COMPARE_AND_SWAP_8 1 - // AARCH64_CXX-NEXT: #define __GLIBCXX_BITSIZE_INT_N_0 128 - // AARCH64_CXX-NEXT: #define __GLIBCXX_TYPE_INT_N_0 __int128 --// AARCH64-NEXT: #define __INT16_C_SUFFIX__ -+// AARCH64-NEXT: #define __INT16_C_SUFFIX__ - // AARCH64-NEXT: #define __INT16_FMTd__ "hd" - // AARCH64-NEXT: #define __INT16_FMTi__ "hi" - // AARCH64-NEXT: #define __INT16_MAX__ 32767 - // AARCH64-NEXT: #define __INT16_TYPE__ short --// AARCH64-NEXT: #define __INT32_C_SUFFIX__ -+// AARCH64-NEXT: #define __INT32_C_SUFFIX__ - // AARCH64-NEXT: #define __INT32_FMTd__ "d" - // AARCH64-NEXT: #define __INT32_FMTi__ "i" - // AARCH64-NEXT: #define __INT32_MAX__ 2147483647 -@@ -124,7 +124,7 @@ - // AARCH64-NEXT: #define __INT64_FMTi__ "li" - // AARCH64-NEXT: #define __INT64_MAX__ 9223372036854775807L - // AARCH64-NEXT: #define __INT64_TYPE__ long int --// AARCH64-NEXT: #define __INT8_C_SUFFIX__ -+// AARCH64-NEXT: #define __INT8_C_SUFFIX__ - // AARCH64-NEXT: #define __INT8_FMTd__ "hhd" - // AARCH64-NEXT: #define __INT8_FMTi__ "hhi" - // AARCH64-NEXT: #define __INT8_MAX__ 127 -@@ -238,7 +238,7 @@ - // AARCH64-NEXT: #define __STDC_UTF_32__ 1 - // AARCH64_C: #define __STDC_VERSION__ 201710L - // AARCH64-NEXT: #define __STDC__ 1 --// AARCH64-NEXT: #define __UINT16_C_SUFFIX__ -+// AARCH64-NEXT: #define __UINT16_C_SUFFIX__ - // AARCH64-NEXT: #define __UINT16_FMTX__ "hX" - // AARCH64-NEXT: #define __UINT16_FMTo__ "ho" - // AARCH64-NEXT: #define __UINT16_FMTu__ "hu" -@@ -259,7 +259,7 @@ - // AARCH64-NEXT: #define __UINT64_FMTx__ "lx" - // AARCH64-NEXT: #define __UINT64_MAX__ 18446744073709551615UL - // AARCH64-NEXT: #define __UINT64_TYPE__ long unsigned int --// AARCH64-NEXT: #define __UINT8_C_SUFFIX__ -+// AARCH64-NEXT: #define __UINT8_C_SUFFIX__ - // AARCH64-NEXT: #define __UINT8_FMTX__ "hhX" - // AARCH64-NEXT: #define __UINT8_FMTo__ "hho" - // AARCH64-NEXT: #define __UINT8_FMTu__ "hhu" -@@ -329,7 +329,7 @@ - // AARCH64-NEXT: #define __UINT_LEAST8_FMTx__ "hhx" - // AARCH64-NEXT: #define __UINT_LEAST8_MAX__ 255 - // AARCH64-NEXT: #define __UINT_LEAST8_TYPE__ unsigned char --// AARCH64-NEXT: #define __USER_LABEL_PREFIX__ -+// AARCH64-NEXT: #define __USER_LABEL_PREFIX__ - // AARCH64-NEXT: #define __VERSION__ "{{.*}}" - // AARCH64-NEXT: #define __WCHAR_MAX__ 4294967295U - // AARCH64-NEXT: #define __WCHAR_TYPE__ unsigned int -diff --git a/clang/test/Preprocessor/predefined-macros.c b/clang/test/Preprocessor/predefined-macros.c -index 083f0e539d88..6c80517ec4d4 100644 ---- a/clang/test/Preprocessor/predefined-macros.c -+++ b/clang/test/Preprocessor/predefined-macros.c -@@ -129,6 +129,8 @@ - // RUN: | FileCheck -match-full-lines %s --check-prefix=CHECK-CL12 - // RUN: %clang_cc1 %s -E -dM -o - -x cl -cl-std=CL2.0 \ - // RUN: | FileCheck -match-full-lines %s --check-prefix=CHECK-CL20 -+// RUN: %clang_cc1 %s -E -dM -o - -x cl -cl-std=CL3.0 \ -+// RUN: | FileCheck -match-full-lines %s --check-prefix=CHECK-CL30 - // RUN: %clang_cc1 %s -E -dM -o - -x cl -cl-fast-relaxed-math \ - // RUN: | FileCheck -match-full-lines %s --check-prefix=CHECK-FRM - // RUN: %clang_cc1 %s -E -dM -o - -x cl -cl-std=clc++ \ -@@ -137,26 +139,37 @@ - // CHECK-CL10: #define CL_VERSION_1_1 110 - // CHECK-CL10: #define CL_VERSION_1_2 120 - // CHECK-CL10: #define CL_VERSION_2_0 200 -+// CHECK-CL10: #define CL_VERSION_3_0 300 - // CHECK-CL10: #define __OPENCL_C_VERSION__ 100 - // CHECK-CL10-NOT: #define __FAST_RELAXED_MATH__ 1 - // CHECK-CL11: #define CL_VERSION_1_0 100 - // CHECK-CL11: #define CL_VERSION_1_1 110 - // CHECK-CL11: #define CL_VERSION_1_2 120 - // CHECK-CL11: #define CL_VERSION_2_0 200 -+// CHECK-CL11: #define CL_VERSION_3_0 300 - // CHECK-CL11: #define __OPENCL_C_VERSION__ 110 - // CHECK-CL11-NOT: #define __FAST_RELAXED_MATH__ 1 - // CHECK-CL12: #define CL_VERSION_1_0 100 - // CHECK-CL12: #define CL_VERSION_1_1 110 - // CHECK-CL12: #define CL_VERSION_1_2 120 - // CHECK-CL12: #define CL_VERSION_2_0 200 -+// CHECK-CL12: #define CL_VERSION_3_0 300 - // CHECK-CL12: #define __OPENCL_C_VERSION__ 120 - // CHECK-CL12-NOT: #define __FAST_RELAXED_MATH__ 1 - // CHECK-CL20: #define CL_VERSION_1_0 100 - // CHECK-CL20: #define CL_VERSION_1_1 110 - // CHECK-CL20: #define CL_VERSION_1_2 120 - // CHECK-CL20: #define CL_VERSION_2_0 200 -+// CHECK-CL20: #define CL_VERSION_3_0 300 - // CHECK-CL20: #define __OPENCL_C_VERSION__ 200 - // CHECK-CL20-NOT: #define __FAST_RELAXED_MATH__ 1 -+// CHECK-CL30: #define CL_VERSION_1_0 100 -+// CHECK-CL30: #define CL_VERSION_1_1 110 -+// CHECK-CL30: #define CL_VERSION_1_2 120 -+// CHECK-CL30: #define CL_VERSION_2_0 200 -+// CHECK-CL30: #define CL_VERSION_3_0 300 -+// CHECK-CL30: #define __OPENCL_C_VERSION__ 300 -+// CHECK-CL30-NOT: #define __FAST_RELAXED_MATH__ 1 - // CHECK-FRM: #define __FAST_RELAXED_MATH__ 1 - // CHECK-CLCPP10: #define __CL_CPP_VERSION_1_0__ 100 - // CHECK-CLCPP10: #define __OPENCL_CPP_VERSION__ 100 -diff --git a/clang/test/Sema/feature-extensions-simult-support.cl b/clang/test/Sema/feature-extensions-simult-support.cl -new file mode 100644 -index 000000000000..0789105002b2 ---- /dev/null -+++ b/clang/test/Sema/feature-extensions-simult-support.cl -@@ -0,0 +1,75 @@ -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL1.2 -cl-ext=-cl_khr_fp64 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL1.2 -cl-ext=+cl_khr_fp64 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL1.2 -cl-ext=-cl_khr_3d_image_writes -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL1.2 -cl-ext=+cl_khr_3d_image_writes -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL1.2 -cl-ext=-cl_khr_subgroups -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL1.2 -cl-ext=+cl_khr_subgroups -+ -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL2.0 -cl-ext=-cl_khr_fp64 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL2.0 -cl-ext=+cl_khr_fp64 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL2.0 -cl-ext=-cl_khr_3d_image_writes -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL2.0 -cl-ext=+cl_khr_3d_image_writes -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL2.0 -cl-ext=-cl_khr_subgroups -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL2.0 -cl-ext=+cl_khr_subgroups -+ -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+__opencl_c_fp64 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+__opencl_c_3d_image_writes -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+__opencl_c_subgroups -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+__opencl_c_pipes -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+__opencl_c_device_enqueue -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+__opencl_c_read_write_images -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+__opencl_c_fp64,-cl_khr_fp64 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+__opencl_c_3d_image_writes,-cl_khr_3d_image_writes -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+__opencl_c_subgroups,-cl_khr_subgroups -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+cl_khr_fp64 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+cl_khr_3d_image_writes -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=+cl_khr_subgroups -+ -+// expected-no-diagnostics -+ -+#ifdef cl_khr_fp64 -+ #ifndef __opencl_c_fp64 -+ #error macros were not properly set up -+ #endif -+#endif -+#ifdef __opencl_c_fp64 -+ #ifndef cl_khr_fp64 -+ #error macros were not properly set up -+ #endif -+#endif -+ -+#ifdef cl_khr_3d_image_writes -+ #ifndef __opencl_c_3d_image_writes -+ #error macros were not properly set up -+ #endif -+#endif -+#ifdef __opencl_c_3d_image_writes -+ #ifndef cl_khr_3d_image_writes -+ #error macros were not properly set up -+ #endif -+#endif -+ -+#ifdef cl_khr_subgroups -+ #ifndef __opencl_c_subgroups -+ #error macros were not properly set up -+ #endif -+#endif -+#ifdef __opencl_c_subgroups -+ #ifndef cl_khr_subgroups -+ #error macros were not properly set up -+ #endif -+#endif -+ -+#if defined(__opencl_c_pipes) || defined(__opencl_c_device_enqueue) -+ #ifndef __opencl_c_generic_address_space -+ #error macros were not properly set up -+ #endif -+#endif -+ -+#if defined(__opencl_c_3d_image_writes) || defined(__opencl_c_read_write_images) -+ #ifndef __opencl_c_images -+ #error macros were not properly set up -+ #endif -+#endif -+ -+kernel void test(){} -diff --git a/clang/test/Sema/features-ignore-pragma.cl b/clang/test/Sema/features-ignore-pragma.cl -new file mode 100644 -index 000000000000..046ce5390754 ---- /dev/null -+++ b/clang/test/Sema/features-ignore-pragma.cl -@@ -0,0 +1,24 @@ -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=__opencl_c_fp64 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=__opencl_c_subgroups -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=__opencl_c_3d_image_writes -+ -+#pragma OPENCL EXTENSION __opencl_c_fp64 : enable -+// expected-warning@-1 {{OpenCL feature support can't be controlled via pragma, ignoring}} -+ -+#pragma OPENCL EXTENSION cl_khr_fp64 : enable -+#ifndef __opencl_c_fp64 -+// expected-warning@-2{{unsupported OpenCL extension 'cl_khr_fp64' - ignoring}} -+#endif -+ -+#pragma OPENCL EXTENSION cl_khr_subgroups : enable -+#ifndef __opencl_c_subgroups -+// expected-warning@-2{{unsupported OpenCL extension 'cl_khr_subgroups' - ignoring}} -+#endif -+ -+#pragma OPENCL EXTENSION cl_khr_3d_image_writes : enable -+#ifndef __opencl_c_3d_image_writes -+// expected-warning@-2{{unsupported OpenCL extension 'cl_khr_3d_image_writes' - ignoring}} -+#endif -+ -+kernel void foo() {} -diff --git a/clang/test/Sema/opencl-features-pipes.cl b/clang/test/Sema/opencl-features-pipes.cl -new file mode 100644 -index 000000000000..c0ac778f24a6 ---- /dev/null -+++ b/clang/test/Sema/opencl-features-pipes.cl -@@ -0,0 +1,18 @@ -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL1.1 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL1.2 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL2.0 -DHAS -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=__opencl_c_pipes -DHAS -+// expected-no-diagnostics -+ -+#ifdef HAS -+ #ifndef __opencl_c_pipes -+ #error Feature should be defined -+ #endif -+#else -+ #ifdef __opencl_c_pipes -+ #error Feature should not be defined -+ #endif -+#endif -+ -+kernel void foo() {} -diff --git a/clang/test/Sema/opencl-features.cl b/clang/test/Sema/opencl-features.cl -new file mode 100644 -index 000000000000..aa432f6b60bf ---- /dev/null -+++ b/clang/test/Sema/opencl-features.cl -@@ -0,0 +1,128 @@ -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL2.0 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL2.0 -cl-ext=-__opencl_c_device_enqueue,-__opencl_c_pipes,-__opencl_c_read_write_images -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CLC++ -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -+// expected-no-diagnostics -+ -+#ifndef __opencl_c_int64 -+ #error Feature __opencl_c_int64 shouldn't be defined -+#endif -+ -+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ == CL_VERSION_2_0) -+ #ifndef __opencl_c_3d_image_writes -+ #error Feature __opencl_c_3d_image_writes should be defined -+ #endif -+ -+ #ifndef __opencl_c_atomic_order_acq_rel -+ #error Feature __opencl_c_atomic_order_acq_rel should be defined -+ #endif -+ -+ #ifndef __opencl_c_atomic_order_seq_cst -+ #error Feature __opencl_c_atomic_order_seq_cst should be defined -+ #endif -+ -+ #ifndef __opencl_c_atomic_scope_device -+ #error Feature __opencl_c_atomic_scope_device should be defined -+ #endif -+ -+ #ifndef __opencl_c_atomic_scope_all_devices -+ #error Feature __opencl_c_atomic_scope_all_devices should be defined -+ #endif -+ -+ #ifndef __opencl_c_device_enqueue -+ #error Feature __opencl_c_device_enqueue should be defined -+ #endif -+ -+ #ifndef __opencl_c_generic_address_space -+ #error Feature __opencl_c_generic_address_space should be defined -+ #endif -+ -+ #ifndef __opencl_c_pipes -+ #error Feature __opencl_c_pipes should be defined -+ #endif -+ -+ #ifndef __opencl_c_program_scope_global_variables -+ #error Feature __opencl_c_program_scope_global_variables should be defined -+ #endif -+ -+ #ifndef __opencl_c_read_write_images -+ #error Feature __opencl_c_read_write_images should be defined -+ #endif -+ -+ #ifndef __opencl_c_subgroups -+ #error Feature __opencl_c_subgroups should be defined -+ #endif -+ -+ #ifndef __opencl_c_work_group_collective_functions -+ #error Feature __opencl_c_work_group_collective_functions should be defined -+ #endif -+ -+ #ifndef __opencl_c_fp64 -+ #error Feature __opencl_c_fp64 should be defined -+ #endif -+ -+ #ifndef __opencl_c_images -+ #error Feature __opencl_c_images should be defined -+ #endif -+#endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0) -+ -+ -+#if __OPENCL_C_VERSION__ == CL_VERSION_3_0 -+ #ifdef __opencl_c_3d_image_writes -+ #error Feature __opencl_c_3d_image_writes shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_atomic_order_acq_rel -+ #error Feature __opencl_c_atomic_order_acq_rel shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_atomic_order_seq_cst -+ #error Feature __opencl_c_atomic_order_seq_cst shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_atomic_scope_device -+ #error Feature __opencl_c_atomic_scope_device shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_atomic_scope_all_devices -+ #error Feature __opencl_c_atomic_scope_all_devices shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_device_enqueue -+ #error Feature __opencl_c_device_enqueue shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_generic_address_space -+ #error Feature __opencl_c_generic_address_space shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_pipes -+ #error Feature __opencl_c_pipes shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_program_scope_global_variables -+ #error Feature __opencl_c_program_scope_global_variables shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_read_write_images -+ #error Feature __opencl_c_read_write_images shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_subgroups -+ #error Feature __opencl_c_subgroups shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_work_group_collective_functions -+ #error Feature __opencl_c_work_group_collective_functions shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_fp64 -+ #error Feature __opencl_c_fp64 shouldn't be defined -+ #endif -+ -+ #ifdef __opencl_c_images -+ #error Feature __opencl_c_images shouldn't be defined -+ #endif -+#endif // __OPENCL_C_VERSION__ == CL_VERSION_3_0 -+ -+kernel void foo() {} -diff --git a/clang/test/Sema/pipe_builtins_feature.cl b/clang/test/Sema/pipe_builtins_feature.cl -new file mode 100644 -index 000000000000..56fa94fc7705 ---- /dev/null -+++ b/clang/test/Sema/pipe_builtins_feature.cl -@@ -0,0 +1,21 @@ -+// RUN: %clang_cc1 -cl-std=CL2.0 -fsyntax-only -verify %s -+// RUN: %clang_cc1 -cl-std=CL3.0 -fsyntax-only -verify %s -+// RUN: %clang_cc1 -cl-std=CL3.0 -cl-ext=__opencl_c_pipes -fsyntax-only -verify %s -+ -+#ifdef __opencl_c_pipes -+ #ifndef __opencl_c_generic_address_space -+ #error Generic address space feature must also be defined -+ #endif -+// CHECK: expected-no-diagnostics -+// check that pragma disable all doesn't touch feature support -+ #pragma OPENCL EXTENSION all : disable -+#endif -+ -+void test(read_only pipe int p, global int *ptr) { -+ reserve_id_t rid; -+} -+ -+#ifndef __opencl_c_pipes -+// expected-error@-5 {{expected parameter declarator}} expected-error@-5 {{expected ')'}} expected-note@-5 {{to match this '('}} -+// expected-error@-5 {{use of type 'reserve_id_t' requires __opencl_c_pipes extension to be enabled}} -+#endif -diff --git a/clang/test/SemaOpenCL/address-spaces-conversions-cl2.0.cl b/clang/test/SemaOpenCL/address-spaces-conversions-cl2.0.cl -index a5a838241347..dbdc5cffd073 100644 ---- a/clang/test/SemaOpenCL/address-spaces-conversions-cl2.0.cl -+++ b/clang/test/SemaOpenCL/address-spaces-conversions-cl2.0.cl -@@ -1,6 +1,9 @@ - // RUN: %clang_cc1 %s -ffake-address-space-map -verify -pedantic -fsyntax-only -DCONSTANT -cl-std=CL2.0 - // RUN: %clang_cc1 %s -ffake-address-space-map -verify -pedantic -fsyntax-only -DGLOBAL -cl-std=CL2.0 - // RUN: %clang_cc1 %s -ffake-address-space-map -verify -pedantic -fsyntax-only -DGENERIC -cl-std=CL2.0 -+// RUN: %clang_cc1 %s -ffake-address-space-map -verify -pedantic -fsyntax-only -DCONSTANT -cl-std=CL3.0 -cl-ext=__opencl_c_generic_address_space -+// RUN: %clang_cc1 %s -ffake-address-space-map -verify -pedantic -fsyntax-only -DGLOBAL -cl-std=CL3.0 -cl-ext=__opencl_c_generic_address_space -+// RUN: %clang_cc1 %s -ffake-address-space-map -verify -pedantic -fsyntax-only -DGENERIC -cl-std=CL3.0 -cl-ext=__opencl_c_generic_address_space - // RUN: %clang_cc1 %s -ffake-address-space-map -verify -pedantic -fsyntax-only -DCONSTANT -cl-std=clc++ - // RUN: %clang_cc1 %s -ffake-address-space-map -verify -pedantic -fsyntax-only -DGLOBAL -cl-std=clc++ - // RUN: %clang_cc1 %s -ffake-address-space-map -verify -pedantic -fsyntax-only -DGENERIC -cl-std=clc++ -diff --git a/clang/test/SemaOpenCL/address-spaces.cl b/clang/test/SemaOpenCL/address-spaces.cl -index e9825fd1682a..563f1e649a82 100644 ---- a/clang/test/SemaOpenCL/address-spaces.cl -+++ b/clang/test/SemaOpenCL/address-spaces.cl -@@ -1,5 +1,6 @@ - // RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only - // RUN: %clang_cc1 %s -cl-std=CL2.0 -verify -pedantic -fsyntax-only -+// RUN: %clang_cc1 %s -cl-std=CL3.0 -cl-ext=__opencl_c_generic_address_space -verify -pedantic -fsyntax-only - // RUN: %clang_cc1 %s -cl-std=clc++ -verify -pedantic -fsyntax-only - - __constant int ci = 1; -diff --git a/clang/test/SemaOpenCL/cl20-device-side-enqueue.cl b/clang/test/SemaOpenCL/cl20-device-side-enqueue.cl -index f63e2913c749..727141190a0b 100644 ---- a/clang/test/SemaOpenCL/cl20-device-side-enqueue.cl -+++ b/clang/test/SemaOpenCL/cl20-device-side-enqueue.cl -@@ -2,6 +2,12 @@ - // RUN: %clang_cc1 %s -cl-std=CL2.0 -triple "spir-unknown-unknown" -verify -pedantic -fsyntax-only -DB32 -DQUALS="const volatile" - // RUN: %clang_cc1 %s -cl-std=CL2.0 -triple "spir64-unknown-unknown" -verify -pedantic -fsyntax-only -Wconversion -DWCONV -DQUALS= - // RUN: %clang_cc1 %s -cl-std=CL2.0 -triple "spir64-unknown-unknown" -verify -pedantic -fsyntax-only -Wconversion -DWCONV -DQUALS="const volatile" -+// RUN: %clang_cc1 %s -cl-std=CL3.0 -triple "spir-unknown-unknown" -cl-ext=__opencl_c_device_enqueue,__opencl_c_subgroups -verify -pedantic -fsyntax-only -DB32 -DQUALS= -+// RUN: %clang_cc1 %s -cl-std=CL3.0 -triple "spir-unknown-unknown" -cl-ext=__opencl_c_device_enqueue,__opencl_c_subgroups -verify -pedantic -fsyntax-only -DB32 -DQUALS="const volatile" -+// RUN: %clang_cc1 %s -cl-std=CL3.0 -triple "spir64-unknown-unknown" -cl-ext=__opencl_c_device_enqueue,__opencl_c_subgroups -verify -pedantic -fsyntax-only -Wconversion -DWCONV -DQUALS= -+// RUN: %clang_cc1 %s -cl-std=CL3.0 -triple "spir64-unknown-unknown" -cl-ext=__opencl_c_device_enqueue,__opencl_c_subgroups -verify -pedantic -fsyntax-only -Wconversion -DWCONV -DQUALS="const volatile" -+ -+ - - typedef struct {int a;} ndrange_t; - // Diagnostic tests for different overloads of enqueue_kernel from Table 6.13.17.1 of OpenCL 2.0 Spec. -@@ -235,11 +241,17 @@ kernel void bar(global unsigned int *buf) - kernel void foo1(global unsigned int *buf) - { - ndrange_t n; -- buf[0] = get_kernel_max_sub_group_size_for_ndrange(n, ^(){}); // expected-error {{use of declaration 'get_kernel_max_sub_group_size_for_ndrange' requires cl_khr_subgroups extension to be enabled}} -+ buf[0] = get_kernel_max_sub_group_size_for_ndrange(n, ^(){}); -+#if __OPENCL_C_VERSION__ < 300 -+// expected-error@-2 {{use of declaration 'get_kernel_max_sub_group_size_for_ndrange' requires cl_khr_subgroups extension to be enabled}} -+#endif - } - - kernel void bar1(global unsigned int *buf) - { - ndrange_t n; -- buf[0] = get_kernel_sub_group_count_for_ndrange(n, ^(){}); // expected-error {{use of declaration 'get_kernel_sub_group_count_for_ndrange' requires cl_khr_subgroups extension to be enabled}} -+ buf[0] = get_kernel_sub_group_count_for_ndrange(n, ^(){}); -+#if __OPENCL_C_VERSION__ < 300 -+// expected-error@-2 {{use of declaration 'get_kernel_sub_group_count_for_ndrange' requires cl_khr_subgroups extension to be enabled}} -+#endif - } -diff --git a/clang/test/SemaOpenCL/forget-unsupported-builtins.cl b/clang/test/SemaOpenCL/forget-unsupported-builtins.cl -new file mode 100644 -index 000000000000..14dd03e2c7db ---- /dev/null -+++ b/clang/test/SemaOpenCL/forget-unsupported-builtins.cl -@@ -0,0 +1,22 @@ -+// RUN: %clang_cc1 -cl-std=cl3.0 -fsyntax-only -verify %s -triple spir-unknown-unknown -+// RUN: %clang_cc1 -cl-std=cl3.0 -fsyntax-only -cl-ext=__opencl_c_pipes,__opencl_c_generic_address_space,__opencl_c_device_enqueue -verify %s -triple spir-unknown-unknown -DFEATURES -+ -+#ifndef FEATURES -+ // expected-no-diagnostics -+#else -+ // expected-error@+10 {{cannot redeclare builtin function 'get_pipe_max_packets'}} -+ // expected-note@+9 {{'get_pipe_max_packets' is a builtin with type 'unsigned int ()'}} -+ // expected-error@+9 {{cannot redeclare builtin function 'to_local'}} -+ // expected-note@+8 {{'to_local' is a builtin with type 'void *(void *)'}} -+ // expected-error@+8 {{cannot redeclare builtin function 'to_global'}} -+ // expected-note@+7 {{'to_global' is a builtin with type 'void *(void *)'}} -+ // expected-error@+7 {{cannot redeclare builtin function 'get_kernel_work_group_size'}} -+ // expected-note@+6 {{'get_kernel_work_group_size' is a builtin with type 'unsigned int ()'}} -+#endif -+ -+int get_pipe_max_packets(int); -+int to_local(int); -+int to_global(int); -+int get_kernel_work_group_size(int); -+ -+kernel void test(global int *dst) {} -diff --git a/clang/test/SemaOpenCL/image-features.cl b/clang/test/SemaOpenCL/image-features.cl -new file mode 100644 -index 000000000000..ace6913bb31e ---- /dev/null -+++ b/clang/test/SemaOpenCL/image-features.cl -@@ -0,0 +1,20 @@ -+// RUN: %clang_cc1 -cl-std=cl2.0 -fsyntax-only -verify %s -triple spir-unknown-unknown -+// RUN: %clang_cc1 -cl-std=cl3.0 -cl-ext=__opencl_c_images -fsyntax-only -verify %s -triple spir-unknown-unknown -+// RUN: %clang_cc1 -cl-std=cl3.0 -cl-ext=__opencl_c_3d_image_writes -fsyntax-only -verify %s -triple spir-unknown-unknown -+// RUN: %clang_cc1 -cl-std=cl3.0 -cl-ext=__opencl_c_read_write_images -fsyntax-only -verify %s -triple spir-unknown-unknown -+ -+#if defined(__opencl_c_read_write_images) && defined(__opencl_c_3d_image_writes) -+ // expected-no-diagnostics -+#endif -+ -+__kernel void write_3d_image(__write_only image3d_t i) {} -+ -+#ifndef __opencl_c_3d_image_writes -+ // expected-error@-3 {{use of type '__write_only image3d_t' requires __opencl_c_3d_image_writes extension to be enabled}} -+#endif -+ -+__kernel void read_write_3d_image(__read_write image3d_t i) { } -+ -+#ifndef __opencl_c_read_write_images -+ // expected-error@-3 {{use of type '__read_write image3d_t' requires __opencl_c_read_write_images extension to be enabled}} -+#endif -diff --git a/clang/test/SemaOpenCL/invalid-pipe-builtin-cl2.0.cl b/clang/test/SemaOpenCL/invalid-pipe-builtin-cl2.0.cl -index 36e76621d24a..38b0a04726e3 100644 ---- a/clang/test/SemaOpenCL/invalid-pipe-builtin-cl2.0.cl -+++ b/clang/test/SemaOpenCL/invalid-pipe-builtin-cl2.0.cl -@@ -1,4 +1,5 @@ - // RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL2.0 -cl-ext=+cl_khr_subgroups -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=__opencl_c_subgroups,__opencl_c_pipes - - #pragma OPENCL EXTENSION cl_khr_subgroups : enable - -diff --git a/clang/test/SemaOpenCL/storageclass-cl20.cl b/clang/test/SemaOpenCL/storageclass-cl20.cl -index 581701d2a6a5..469c526ebc30 100644 ---- a/clang/test/SemaOpenCL/storageclass-cl20.cl -+++ b/clang/test/SemaOpenCL/storageclass-cl20.cl -@@ -1,4 +1,5 @@ - // RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL2.0 -+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -cl-std=CL3.0 -cl-ext=__opencl_c_program_scope_global_variables,__opencl_c_generic_address_space - - int G2 = 0; - global int G3 = 0; -diff --git a/clang/utils/TableGen/ClangOpenCLBuiltinEmitter.cpp b/clang/utils/TableGen/ClangOpenCLBuiltinEmitter.cpp -index 7c63cf51ecfa..785ff93aaf85 100644 ---- a/clang/utils/TableGen/ClangOpenCLBuiltinEmitter.cpp -+++ b/clang/utils/TableGen/ClangOpenCLBuiltinEmitter.cpp -@@ -56,6 +56,7 @@ - //===----------------------------------------------------------------------===// - - #include "TableGenBackends.h" -+#include "clang/Basic/OpenCLOptions.h" - #include "llvm/ADT/MapVector.h" - #include "llvm/ADT/STLExtras.h" - #include "llvm/ADT/SmallString.h" -@@ -69,6 +70,7 @@ - #include "llvm/TableGen/Record.h" - #include "llvm/TableGen/StringMatcher.h" - #include "llvm/TableGen/TableGenBackend.h" -+#include <numeric> - #include <set> - - using namespace llvm; -@@ -228,6 +230,10 @@ private: - // The function "tan", having the same signatures, would be mapped to the - // same entry (<I1, I2, I3>). - MapVector<BuiltinIndexListTy *, BuiltinTableEntries> SignatureListMap; -+ -+ // Encode all versions -+ unsigned short -+ EncodeBuiltinVersions(std::vector<Record *> BuiltinVersionsRecords) const; - }; - } // namespace - -@@ -338,12 +344,10 @@ struct OpenCLBuiltinStruct { - const bool IsConst : 1; - // Function attribute __attribute__((convergent)) - const bool IsConv : 1; -+ // All opencl versions encoded -+ const unsigned char AllVersions : 5; - // OpenCL extension(s) required for this overload. - const unsigned short Extension; -- // First OpenCL version in which this overload was introduced (e.g. CL20). -- const unsigned short MinVersion; -- // First OpenCL version in which this overload was removed (e.g. CL20). -- const unsigned short MaxVersion; - }; - - )"; -@@ -510,11 +514,9 @@ void BuiltinNameEmitter::EmitBuiltinTable() { - << (Overload.first->getValueAsBit("IsPure")) << ", " - << (Overload.first->getValueAsBit("IsConst")) << ", " - << (Overload.first->getValueAsBit("IsConv")) << ", " -- << FunctionExtensionIndex[ExtName] << ", " -- << Overload.first->getValueAsDef("MinVersion")->getValueAsInt("ID") -- << ", " -- << Overload.first->getValueAsDef("MaxVersion")->getValueAsInt("ID") -- << " },\n"; -+ << EncodeBuiltinVersions( -+ Overload.first->getValueAsListOfDefs("Versions")) -+ << ", " << FunctionExtensionIndex[ExtName] << " },\n"; - Index++; - } - } -@@ -535,10 +537,8 @@ bool BuiltinNameEmitter::CanReuseSignature( - if (Rec->getValueAsBit("IsPure") == Rec2->getValueAsBit("IsPure") && - Rec->getValueAsBit("IsConst") == Rec2->getValueAsBit("IsConst") && - Rec->getValueAsBit("IsConv") == Rec2->getValueAsBit("IsConv") && -- Rec->getValueAsDef("MinVersion")->getValueAsInt("ID") == -- Rec2->getValueAsDef("MinVersion")->getValueAsInt("ID") && -- Rec->getValueAsDef("MaxVersion")->getValueAsInt("ID") == -- Rec2->getValueAsDef("MaxVersion")->getValueAsInt("ID") && -+ EncodeBuiltinVersions(Rec->getValueAsListOfDefs("Versions")) == -+ EncodeBuiltinVersions(Rec2->getValueAsListOfDefs("Versions")) && - Rec->getValueAsDef("Extension")->getName() == - Rec2->getValueAsDef("Extension")->getName()) { - return true; -@@ -814,6 +814,15 @@ static void OCL2Qual(ASTContext &Context, const OpenCLTypeStruct &Ty, - OS << "\n} // OCL2Qual\n"; - } - -+unsigned short BuiltinNameEmitter::EncodeBuiltinVersions( -+ std::vector<Record *> BuiltinVersionsRecords) const { -+ return std::accumulate( -+ BuiltinVersionsRecords.begin(), BuiltinVersionsRecords.end(), -+ (unsigned short)0, [](unsigned short C, Record *R) { -+ return C |= clang::OpenCLOptions::EncodeVersion(R->getValueAsInt("ID")); -+ }); -+} -+ - void clang::EmitClangOpenCLBuiltins(RecordKeeper &Records, raw_ostream &OS) { - BuiltinNameEmitter NameChecker(Records, OS); - NameChecker.Emit(); --- -2.17.1 - diff --git a/dynamic-layers/clang-layer/recipes-devtools/clang/files/llvm11-Remove-repo-name-in-LLVM-IR.patch b/dynamic-layers/clang-layer/recipes-devtools/clang/files/llvm11-Remove-repo-name-in-LLVM-IR.patch deleted file mode 100644 index 09089432..00000000 --- a/dynamic-layers/clang-layer/recipes-devtools/clang/files/llvm11-Remove-repo-name-in-LLVM-IR.patch +++ /dev/null @@ -1,50 +0,0 @@ -From ff0a6da84b94c16c4519c649f1f7bed3cdf89bbb Mon Sep 17 00:00:00 2001 -From: Feng Zou <feng.zou@intel.com> -Date: Tue, 20 Oct 2020 11:29:04 +0800 -Subject: [PATCH] Remove repo name in LLVM IR - -Upstream-Status: Backport [Taken from opencl-clang patches, https://github.com/intel/opencl-clang/blob/ocl-open-110/patches/llvm/0002-Remove-repo-name-in-LLVM-IR.patch] -Signed-off-by: Feng Zou <feng.zou@intel.com> -Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> - ---- - llvm/cmake/modules/VersionFromVCS.cmake | 23 ++++++++++++----------- - 1 file changed, 12 insertions(+), 11 deletions(-) - -diff --git a/llvm/cmake/modules/VersionFromVCS.cmake b/llvm/cmake/modules/VersionFromVCS.cmake -index 18edbeabe3e..2d965263478 100644 ---- a/llvm/cmake/modules/VersionFromVCS.cmake -+++ b/llvm/cmake/modules/VersionFromVCS.cmake -@@ -33,17 +33,18 @@ function(get_source_info path revision repository) - else() - set(remote "origin") - endif() -- execute_process(COMMAND ${GIT_EXECUTABLE} remote get-url ${remote} -- WORKING_DIRECTORY ${path} -- RESULT_VARIABLE git_result -- OUTPUT_VARIABLE git_output -- ERROR_QUIET) -- if(git_result EQUAL 0) -- string(STRIP "${git_output}" git_output) -- set(${repository} ${git_output} PARENT_SCOPE) -- else() -- set(${repository} ${path} PARENT_SCOPE) -- endif() -+ # Do not show repo name in IR -+ # execute_process(COMMAND ${GIT_EXECUTABLE} remote get-url ${remote} -+ # WORKING_DIRECTORY ${path} -+ # RESULT_VARIABLE git_result -+ # OUTPUT_VARIABLE git_output -+ # ERROR_QUIET) -+ # if(git_result EQUAL 0) -+ # string(STRIP "${git_output}" git_output) -+ # set(${repository} ${git_output} PARENT_SCOPE) -+ # else() -+ # set(${repository} ${path} PARENT_SCOPE) -+ # endif() - endif() - else() - message(WARNING "Git not found. Version cannot be determined.") --- -2.18.1 - diff --git a/dynamic-layers/clang-layer/recipes-devtools/clang/files/llvm11-skip-building-tests.patch b/dynamic-layers/clang-layer/recipes-devtools/clang/files/llvm11-skip-building-tests.patch deleted file mode 100644 index 011c09ee..00000000 --- a/dynamic-layers/clang-layer/recipes-devtools/clang/files/llvm11-skip-building-tests.patch +++ /dev/null @@ -1,51 +0,0 @@ -From d362652617c5e840089273df0c6623a9745c92a2 Mon Sep 17 00:00:00 2001 -From: Naveen Saini <naveen.kumar.saini@intel.com> -Date: Wed, 21 Aug 2019 14:35:31 +0800 -Subject: [PATCH] llvm-spirv: skip building tests - -Some of these need clang to be built and since we're building this in-tree, -that leads to problems when compiling libcxx, compiler-rt which aren't built -in-tree. - -Instead of using SPIRV_SKIP_CLANG_BUILD to skip clang build and adding this to -all components, disable the building of tests altogether. - -Upstream-Status: Inappropriate - -Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> -Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> ---- - CMakeLists.txt | 10 ---------- - 1 file changed, 10 deletions(-) - -diff --git a/CMakeLists.txt b/CMakeLists.txt -index ecebb4cb..578ca602 100644 ---- a/CMakeLists.txt -+++ b/CMakeLists.txt -@@ -25,13 +25,6 @@ if(LLVM_SPIRV_BUILD_EXTERNAL) - set(CMAKE_CXX_STANDARD 14) - set(CMAKE_CXX_STANDARD_REQUIRED ON) - -- if(LLVM_SPIRV_INCLUDE_TESTS) -- set(LLVM_TEST_COMPONENTS -- llvm-as -- llvm-dis -- ) -- endif(LLVM_SPIRV_INCLUDE_TESTS) -- - find_package(LLVM ${BASE_LLVM_VERSION} REQUIRED - COMPONENTS - Analysis -@@ -62,9 +55,6 @@ set(LLVM_SPIRV_INCLUDE_DIRS ${CMAKE_CURRENT_SOURCE_DIR}/include) - - add_subdirectory(lib/SPIRV) - add_subdirectory(tools/llvm-spirv) --if(LLVM_SPIRV_INCLUDE_TESTS) -- add_subdirectory(test) --endif(LLVM_SPIRV_INCLUDE_TESTS) - - install( - FILES --- -2.26.2 - diff --git a/dynamic-layers/clang-layer/recipes-devtools/clang/llvm-project-source.bbappend b/dynamic-layers/clang-layer/recipes-devtools/clang/llvm-project-source.bbappend deleted file mode 100644 index 4e2bc535..00000000 --- a/dynamic-layers/clang-layer/recipes-devtools/clang/llvm-project-source.bbappend +++ /dev/null @@ -1,35 +0,0 @@ -FILESEXTRAPATHS_prepend_intel-x86-common := "${THISDIR}/files:" - -SPIRV_BRANCH = "${@bb.utils.contains('LLVMVERSION', '10.0.1', 'llvm_release_100', 'llvm_release_110', d)}" - -SPIRV10_SRCREV = "576abae62cecd171992017a4a786e3831221ab8d" -SPIRV11_SRCREV = "2a8c1e6c9778deaa720a23e08c293006dc5d56fd" - -SPIRV_SRCREV = "${@bb.utils.contains('LLVMVERSION', '10.0.1', '${SPIRV10_SRCREV}', '${SPIRV11_SRCREV}', d)}" - -SRC_URI_LLVM10 = " \ - file://llvm10-skip-building-tests.patch;patchdir=llvm/projects/llvm-spirv \ - file://fix-shared-libs.patch;patchdir=llvm/projects/llvm-spirv \ - file://BasicBlockUtils-Add-metadata-fixing-in-SplitBlockPre.patch;patchdir=llvm \ - file://IndVarSimplify-Do-not-use-SCEV-expander-for-IVCount-.patch;patchdir=llvm \ - file://llvm10-OpenCL-3.0-support.patch \ - file://0002-Add-cl_khr_extended_subgroup-extensions.patch \ - file://0001-Memory-leak-fix-for-Managed-Static-Mutex.patch \ - file://llvm10-Remove-repo-name-in-LLVM-IR.patch \ - file://0001-Fix-debug-info-of-work-item-builtin-translation-745.patch;patchdir=llvm/projects/llvm-spirv \ - " - -SRC_URI_LLVM11 = " \ - file://llvm11-skip-building-tests.patch;patchdir=llvm/projects/llvm-spirv \ - file://llvm11-OpenCL-3.0-support.patch \ - file://0001-Memory-leak-fix-for-Managed-Static-Mutex.patch \ - file://llvm11-Remove-repo-name-in-LLVM-IR.patch \ - " - -SPIRV_LLVM_SRC_URI = "git://github.com/KhronosGroup/SPIRV-LLVM-Translator.git;protocol=https;branch=${SPIRV_BRANCH};destsuffix=git/llvm/projects/llvm-spirv;name=spirv" - -SPIRV_LLVM_PATCHES = "${@bb.utils.contains('LLVMVERSION', '10.0.1', '${SRC_URI_LLVM10}', '${SRC_URI_LLVM11}', d)}" - - -SRC_URI_append_intel-x86-common = "${@bb.utils.contains('LLVMVERSION', '12.0.0', '', ' ${SPIRV_LLVM_SRC_URI} ${SPIRV_LLVM_PATCHES} ', d)}" -SRCREV_spirv = "${@bb.utils.contains('LLVMVERSION', '12.0.0', '', '${SPIRV_SRCREV}', d)}" diff --git a/dynamic-layers/clang-layer/recipes-opencl/compute-runtime/intel-compute-runtime/allow-to-find-cpp-generation-tool.patch b/dynamic-layers/clang-layer/recipes-opencl/compute-runtime/intel-compute-runtime/allow-to-find-cpp-generation-tool.patch index 631e0ce6..265fcfa2 100644 --- a/dynamic-layers/clang-layer/recipes-opencl/compute-runtime/intel-compute-runtime/allow-to-find-cpp-generation-tool.patch +++ b/dynamic-layers/clang-layer/recipes-opencl/compute-runtime/intel-compute-runtime/allow-to-find-cpp-generation-tool.patch @@ -1,44 +1,51 @@ -From 32851d269ac0c027730168c04d9ce32d0eb50905 Mon Sep 17 00:00:00 2001 -From: Naveen Saini <naveen.kumar.saini@intel.com> -Date: Wed, 27 May 2020 11:12:09 +0000 -Subject: [PATCH] intel-compute-runtime: allow to find cpp_generation_tool - -Build not able to locate cpp_generation_tool. +From a6361d635e5f3046853883f3ac06fb175116933c Mon Sep 17 00:00:00 2001 +From: Dongwon Kim <dongwon.kim@intel.com> +Date: Sat, 21 Aug 2021 16:09:39 -0700 +Subject: [PATCH] Build not able to locate cpp_generation_tool. Upstream-Status: Inappropriate [oe specific] + Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> +Signed-off-by: Dongwon Kim <dongwon.kim@intel.com> --- - shared/source/built_ins/kernels/CMakeLists.txt | 8 ++++---- - 1 file changed, 4 insertions(+), 4 deletions(-) + shared/source/built_ins/kernels/CMakeLists.txt | 10 +++++----- + 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/shared/source/built_ins/kernels/CMakeLists.txt b/shared/source/built_ins/kernels/CMakeLists.txt -index 1b1225077..2e8148c25 100644 +index ed85a37c52..f7c9e79137 100644 --- a/shared/source/built_ins/kernels/CMakeLists.txt +++ b/shared/source/built_ins/kernels/CMakeLists.txt -@@ -93,9 +93,9 @@ function(compile_builtin gen_type platform_type builtin bits builtin_options mod - ) - add_custom_command( - OUTPUT ${OUTPUT_FILE_CPP} -- COMMAND $<TARGET_FILE:cpp_generate_tool> --file ${OUTPUTPATH_BASE}.gen --output ${OUTPUT_FILE_CPP} --array ${mode}_${BASENAME} --platform ${family_name_with_type} -+ COMMAND cpp_generate_tool --file ${OUTPUTPATH_BASE}.gen --output ${OUTPUT_FILE_CPP} --array ${mode}_${BASENAME} --platform ${family_name_with_type} - WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR} -- DEPENDS ${OUTPUTPATH_BASE}.gen $<TARGET_FILE:cpp_generate_tool> -+ DEPENDS ${OUTPUTPATH_BASE}.gen cpp_generate_tool - ) - endfunction() - -@@ -136,9 +136,9 @@ function(generate_cpp_spirv builtin) - ) - add_custom_command( - OUTPUT ${OUTPUT_FILE_CPP} -- COMMAND $<TARGET_FILE:cpp_generate_tool> --file ${GENERATED_SPV_INPUT} --output ${OUTPUT_FILE_CPP} --array ${BASENAME} -+ COMMAND cpp_generate_tool --file ${GENERATED_SPV_INPUT} --output ${OUTPUT_FILE_CPP} --array ${BASENAME} - WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR} -- DEPENDS ${GENERATED_SPV_INPUT} $<TARGET_FILE:cpp_generate_tool> -+ DEPENDS ${GENERATED_SPV_INPUT} cpp_generate_tool +@@ -107,9 +107,9 @@ function(compile_builtin core_type platform_it builtin bits builtin_options mode + ) + add_custom_command( + OUTPUT ${OUTPUT_FILE_CPP} +- COMMAND $<TARGET_FILE:cpp_generate_tool> --file ${BINARY_OUTPUT}.bin --output ${OUTPUT_FILE_CPP} --array ${mode}_${BASENAME} --device ${RELEASE_FILENAME} ++ COMMAND cpp_generate_tool --file ${BINARY_OUTPUT}.bin --output ${OUTPUT_FILE_CPP} --array ${mode}_${BASENAME} --device ${RELEASE_FILENAME} + WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR} +- DEPENDS ${OUTPUT_FILES_BINARIES} $<TARGET_FILE:cpp_generate_tool> ++ DEPENDS ${OUTPUT_FILES_BINARIES} cpp_generate_tool + ) + list(APPEND BUILTINS_COMMANDS "${OUTPUT_FILE_CPP}") + else() +@@ -159,9 +159,9 @@ function(generate_cpp_spirv builtin) + ) + add_custom_command( + OUTPUT ${OUTPUT_FILE_CPP} +- COMMAND $<TARGET_FILE:cpp_generate_tool> --file ${GENERATED_SPV_INPUT} --output ${OUTPUT_FILE_CPP} --array ${BASENAME} ++ COMMAND cpp_generate_tool --file ${GENERATED_SPV_INPUT} --output ${OUTPUT_FILE_CPP} --array ${BASENAME} + WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR} +- DEPENDS ${GENERATED_SPV_INPUT} $<TARGET_FILE:cpp_generate_tool> ++ DEPENDS ${GENERATED_SPV_INPUT} cpp_generate_tool + ) + set(OUTPUT_LIST_CPP_FILES ${OUTPUT_LIST_CPP_FILES} ${OUTPUT_FILE_CPP} PARENT_SCOPE) + else() +@@ -277,4 +277,4 @@ if(NOT "${OUTPUT_LIST_CPP_FILES}" STREQUAL "") ) - endfunction() + endif() +-apply_macro_for_each_core_type("SUPPORTED") +\ No newline at end of file ++apply_macro_for_each_core_type("SUPPORTED") -- -2.28.0 +2.43.2 diff --git a/dynamic-layers/clang-layer/recipes-opencl/compute-runtime/intel-compute-runtime/disable-werror.patch b/dynamic-layers/clang-layer/recipes-opencl/compute-runtime/intel-compute-runtime/disable-werror.patch new file mode 100644 index 00000000..20d9b847 --- /dev/null +++ b/dynamic-layers/clang-layer/recipes-opencl/compute-runtime/intel-compute-runtime/disable-werror.patch @@ -0,0 +1,16 @@ +Upstream-Status: Inappropriate +Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> + +diff --git a/CMakeLists.txt b/CMakeLists.txt +index d52e089778..bc0cf35014 100644 +--- a/CMakeLists.txt ++++ b/CMakeLists.txt +@@ -727,7 +727,7 @@ if(NOT MSVC) + set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wno-noexcept-type") # Added for gtest + endif() + endif() +- set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Werror -Werror=vla") ++ set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Werror=vla") + + if(USE_SANITIZE_UB) + check_cxx_compiler_flag(-fsanitize=undefined COMPILER_SUPPORTS_UNDEFINED_BEHAVIOR_SANITIZER) diff --git a/dynamic-layers/clang-layer/recipes-opencl/compute-runtime/intel-compute-runtime/external-ocloc.patch b/dynamic-layers/clang-layer/recipes-opencl/compute-runtime/intel-compute-runtime/external-ocloc.patch new file mode 100644 index 00000000..5f93b7b6 --- /dev/null +++ b/dynamic-layers/clang-layer/recipes-opencl/compute-runtime/intel-compute-runtime/external-ocloc.patch @@ -0,0 +1,40 @@ +From 1f58c22992ddea4167b01b44448528de427f50d5 Mon Sep 17 00:00:00 2001 +From: Dongwon Kim <dongwon.kim@intel.com> +Date: Wed, 2 Mar 2022 15:52:45 -0800 +Subject: [PATCH] external ocloc + +Upstream-Status: Inappropriate + +Signed-off-by: Dongwon Kim <dongwon.kim@intel.com> +--- + cmake/ocloc_cmd_prefix.cmake | 14 ++++++++------ + 1 file changed, 8 insertions(+), 6 deletions(-) + +diff --git a/cmake/ocloc_cmd_prefix.cmake b/cmake/ocloc_cmd_prefix.cmake +index 2b44330831..03067c9df0 100644 +--- a/cmake/ocloc_cmd_prefix.cmake ++++ b/cmake/ocloc_cmd_prefix.cmake +@@ -4,12 +4,14 @@ + # SPDX-License-Identifier: MIT + # + +-if(WIN32) +- set(ocloc_cmd_prefix ocloc) +-else() +- if(DEFINED NEO__IGC_LIBRARY_PATH) +- set(ocloc_cmd_prefix ${CMAKE_COMMAND} -E env "LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:${NEO__IGC_LIBRARY_PATH}:$<TARGET_FILE_DIR:ocloc_lib>" $<TARGET_FILE:ocloc>) ++if(NOT DEFINED ocloc_cmd_prefix) ++ if(WIN32) ++ set(ocloc_cmd_prefix ocloc) + else() +- set(ocloc_cmd_prefix ${CMAKE_COMMAND} -E env "LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:$<TARGET_FILE_DIR:ocloc_lib>" $<TARGET_FILE:ocloc>) ++ if(DEFINED NEO__IGC_LIBRARY_PATH) ++ set(ocloc_cmd_prefix LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:${NEO__IGC_LIBRARY_PATH}:$<TARGET_FILE_DIR:ocloc_lib> $<TARGET_FILE:ocloc>) ++ else() ++ set(ocloc_cmd_prefix LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:$<TARGET_FILE_DIR:ocloc_lib> $<TARGET_FILE:ocloc>) ++ endif() + endif() + endif() +-- +2.37.3 + diff --git a/dynamic-layers/clang-layer/recipes-opencl/compute-runtime/intel-compute-runtime_21.04.18912.bb b/dynamic-layers/clang-layer/recipes-opencl/compute-runtime/intel-compute-runtime_21.04.18912.bb deleted file mode 100644 index d0078f3a..00000000 --- a/dynamic-layers/clang-layer/recipes-opencl/compute-runtime/intel-compute-runtime_21.04.18912.bb +++ /dev/null @@ -1,56 +0,0 @@ -SUMMARY = "The Intel(R) Graphics Compute Runtime for OpenCL(TM)" -DESCRIPTION = "The Intel(R) Graphics Compute Runtime for OpenCL(TM) \ -is an open source project to converge Intel's development efforts \ -on OpenCL(TM) compute stacks supporting the GEN graphics hardware \ -architecture." - -LICENSE = "MIT" -LIC_FILES_CHKSUM = "file://LICENSE;md5=ae27f47fd6755510247c19e547e4c804 \ - file://third_party/opencl_headers/LICENSE;md5=dcefc90f4c3c689ec0c2489064e7273b" - -SRC_URI = "git://github.com/intel/compute-runtime.git;protocol=https \ - " -SRC_URI_append_class-target = " file://allow-to-find-cpp-generation-tool.patch" - -SRCREV = "5a05d7643c4d8cc14e281b686c78101a7a01e8e1" - -S = "${WORKDIR}/git" - -DEPENDS += " intel-graphics-compiler gmmlib" -DEPENDS_append_class-target = " intel-compute-runtime-native libva" - -RDEPENDS_${PN} += " intel-graphics-compiler gmmlib" - -inherit cmake pkgconfig - -COMPATIBLE_HOST = '(x86_64).*-linux' -COMPATIBLE_HOST_libc-musl = "null" - -EXTRA_OECMAKE = " \ - -DIGC_DIR=${STAGING_INCDIR}/igc \ - -DBUILD_TYPE=Release \ - -DSKIP_UNIT_TESTS=1 \ - -DCCACHE_ALLOWED=FALSE \ - " -EXTRA_OECMAKE_append_class-target = " \ - -Dcloc_cmd_prefix=ocloc \ - " - -PACKAGECONFIG ??= "" -PACKAGECONFIG[levelzero] = "-DBUILD_WITH_L0=ON, -DBUILD_WITH_L0=OFF, level-zero" - -do_install_append_class-native() { - install -d ${D}${bindir} - install ${B}/bin/cpp_generate_tool ${D}${bindir}/ -} - -FILES_${PN} += " \ - ${libdir}/intel-opencl/libigdrcl.so \ - ${libdir}/libocloc.so \ - " - -FILES_${PN}-dev = "${includedir}" - -BBCLASSEXTEND = "native nativesdk" - -UPSTREAM_CHECK_GITTAGREGEX = "(?P<pver>\d+(\.\d+)+)" diff --git a/dynamic-layers/clang-layer/recipes-opencl/compute-runtime/intel-compute-runtime_24.05.28454.6.bb b/dynamic-layers/clang-layer/recipes-opencl/compute-runtime/intel-compute-runtime_24.05.28454.6.bb new file mode 100644 index 00000000..60125171 --- /dev/null +++ b/dynamic-layers/clang-layer/recipes-opencl/compute-runtime/intel-compute-runtime_24.05.28454.6.bb @@ -0,0 +1,64 @@ +SUMMARY = "The Intel(R) Graphics Compute Runtime for OpenCL(TM)" +DESCRIPTION = "The Intel(R) Graphics Compute Runtime for OpenCL(TM) \ +is an open source project to converge Intel's development efforts \ +on OpenCL(TM) compute stacks supporting the GEN graphics hardware \ +architecture." + +LICENSE = "MIT & Apache-2.0" +LIC_FILES_CHKSUM = "file://LICENSE.md;md5=eca6ec6997e18db166db7109cdbe611c \ + file://third_party/opencl_headers/LICENSE;md5=3b83ef96387f14655fc854ddc3c6bd57" + +SRC_URI = "git://github.com/intel/compute-runtime.git;protocol=https;branch=releases/24.05 \ + file://disable-werror.patch \ + file://allow-to-find-cpp-generation-tool.patch \ + file://external-ocloc.patch \ + " + +SRCREV = "23a9725e62c8181193c90e5e5207507a1f136587" + +S = "${WORKDIR}/git" + +DEPENDS += " intel-graphics-compiler gmmlib libva qemu-native" + +RDEPENDS:${PN} += " intel-graphics-compiler gmmlib" + +inherit cmake pkgconfig qemu + +COMPATIBLE_HOST = '(x86_64).*-linux' +COMPATIBLE_HOST:libc-musl = "null" + +EXTRA_OECMAKE = " \ + -DIGC_DIR=${STAGING_INCDIR}/igc \ + -DBUILD_TYPE=Release \ + -DSKIP_UNIT_TESTS=1 \ + -DCCACHE_ALLOWED=FALSE \ + -DNEO_DISABLE_LD_LLD=ON \ + -DNEO_DISABLE_LD_GOLD=ON \ + " + +EXTRA_OECMAKE:append:class-target = " \ + -Docloc_cmd_prefix=ocloc \ + -DCMAKE_CROSSCOMPILING_EMULATOR=${WORKDIR}/qemuwrapper \ + " + +PACKAGECONFIG ??= "" +PACKAGECONFIG[levelzero] = "-DBUILD_WITH_L0=ON, -DBUILD_WITH_L0=OFF, level-zero" + +do_configure:prepend:class-target () { + # Write out a qemu wrapper that will be used by cmake. + qemu_binary="${@qemu_wrapper_cmdline(d, d.getVar('STAGING_DIR_HOST'), [d.expand('${B}/bin'),d.expand('${STAGING_DIR_HOST}${libdir}'),d.expand('${STAGING_DIR_HOST}${base_libdir}')])}" + cat > ${WORKDIR}/qemuwrapper << EOF +#!/bin/sh +$qemu_binary "\$@" +EOF + chmod +x ${WORKDIR}/qemuwrapper +} + +FILES:${PN} += " \ + ${libdir}/intel-opencl/libigdrcl.so \ + ${libdir}/libocloc.so \ + " + +FILES:${PN}-dev = "${includedir}" + +UPSTREAM_CHECK_GITTAGREGEX = "(?P<pver>\d+(\.\d+)+)" diff --git a/dynamic-layers/clang-layer/recipes-opencl/igc/files/0001-BiF-CMakeLists.txt-remove-opt-from-DEPENDS.patch b/dynamic-layers/clang-layer/recipes-opencl/igc/files/0001-BiF-CMakeLists.txt-remove-opt-from-DEPENDS.patch new file mode 100644 index 00000000..377081fd --- /dev/null +++ b/dynamic-layers/clang-layer/recipes-opencl/igc/files/0001-BiF-CMakeLists.txt-remove-opt-from-DEPENDS.patch @@ -0,0 +1,35 @@ +From 3d71670f8ad5b54d434c2f5f71713bb1d5433ae4 Mon Sep 17 00:00:00 2001 +From: Anuj Mittal <anuj.mittal@intel.com> +Date: Tue, 12 Oct 2021 23:46:42 +0800 +Subject: [PATCH] BiF/CMakeLists.txt: remove opt from DEPENDS + +Otherwise it starts failing with: + +| ninja: error: 'IGC/VectorCompiler/lib/BiF/opt', needed by 'IGC/VectorCompiler/lib/BiF/VCBiFPrintfOCL32.opt.bc', missing and no known rule to make it + +We don't need to explicitly make sure opt is built when +using prebuilt binaries. + +Upstream-Status: Inappropriate + +Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> +--- + IGC/VectorCompiler/lib/BiF/cmake/Functions.cmake | 2 +- + 1 file changed, 1 insertion(+), 1 deletion(-) + +diff --git a/IGC/VectorCompiler/lib/BiF/cmake/Functions.cmake b/IGC/VectorCompiler/lib/BiF/cmake/Functions.cmake +index d20d7f887..882e09fea 100644 +--- a/IGC/VectorCompiler/lib/BiF/cmake/Functions.cmake ++++ b/IGC/VectorCompiler/lib/BiF/cmake/Functions.cmake +@@ -109,7 +109,7 @@ function(vc_build_bif TARGET RES_FILE CMCL_SRC_PATH BIF_NAME PTR_BIT_SIZE) + COMMENT "vc_build_bif: Translating CMCL builtins: ${BIF_CLANG_BC_NAME_FINAL} -> ${BIF_OPT_BC_NAME}" + COMMAND CMCLTranslatorTool -o ${BIF_CMCL_BC_NAME} ${BIF_CLANG_BC_NAME_FINAL} + COMMAND ${LLVM_OPT_EXE} ${IGC_LLVM_DEPENDENT_OPT_FLAGS} --O2 -o ${BIF_OPT_BC_NAME} ${BIF_CMCL_BC_NAME} +- DEPENDS CMCLTranslatorTool ${LLVM_OPT_EXE} ${BIF_CLANG_BC_PATH_FINAL} ++ DEPENDS CMCLTranslatorTool ${BIF_CLANG_BC_PATH_FINAL} + BYPRODUCTS ${BIF_OPT_BC_PATH} + SOURCES ${CMCL_SRC_PATH}) + set(${RES_FILE} ${BIF_OPT_BC_NAME} PARENT_SCOPE) +-- +2.43.2 + diff --git a/dynamic-layers/clang-layer/recipes-opencl/igc/files/0001-Fix-build-with-LLVM-12.patch b/dynamic-layers/clang-layer/recipes-opencl/igc/files/0001-Fix-build-with-LLVM-12.patch deleted file mode 100644 index 1d2c5767..00000000 --- a/dynamic-layers/clang-layer/recipes-opencl/igc/files/0001-Fix-build-with-LLVM-12.patch +++ /dev/null @@ -1,2433 +0,0 @@ -From 869cdb784aa062bd08bd26465bf0b5651e8f112e Mon Sep 17 00:00:00 2001 -From: =?UTF-8?q?Zolt=C3=A1n=20B=C3=B6sz=C3=B6rm=C3=A9nyi?= <zboszor@pr.hu> -Date: Thu, 25 Feb 2021 10:39:27 +0100 -Subject: [PATCH 1/3] Fix build with LLVM 12 -MIME-Version: 1.0 -Content-Type: text/plain; charset=UTF-8 -Content-Transfer-Encoding: 8bit - -Upstream-Status: Pending - -Signed-off-by: Zoltán Böszörményi <zboszor@gmail.com> -Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> ---- - .../LegalizeFunctionSignatures.cpp | 2 +- - IGC/AdaptorOCL/SPIRV/SPIRVReader.cpp | 26 +++++++++----- - IGC/AdaptorOCL/SPIRV/SPIRVUtil.cpp | 6 +++- - IGC/Compiler/CISACodeGen/AdvMemOpt.cpp | 4 +++ - IGC/Compiler/CISACodeGen/CShader.cpp | 14 ++++---- - .../CISACodeGen/ConstantCoalescing.cpp | 14 ++++---- - IGC/Compiler/CISACodeGen/DeSSA.cpp | 3 +- - IGC/Compiler/CISACodeGen/EmitVISAPass.cpp | 35 +++++++++--------- - IGC/Compiler/CISACodeGen/Emu64OpsPass.cpp | 4 +-- - IGC/Compiler/CISACodeGen/GenIRLowering.cpp | 4 +-- - .../CISACodeGen/GenSimplification.cpp | 5 +-- - .../CISACodeGen/GeometryShaderLowering.cpp | 3 +- - IGC/Compiler/CISACodeGen/LdShrink.cpp | 10 ++++-- - .../CISACodeGen/LowerGEPForPrivMem.cpp | 8 ++--- - IGC/Compiler/CISACodeGen/MemOpt.cpp | 8 ++--- - IGC/Compiler/CISACodeGen/MemOpt2.cpp | 5 +-- - .../CISACodeGen/OpenCLKernelCodeGen.cpp | 2 +- - IGC/Compiler/CISACodeGen/PatternMatchPass.cpp | 4 +-- - IGC/Compiler/CISACodeGen/PushAnalysis.cpp | 4 +-- - .../CISACodeGen/RegisterEstimator.cpp | 3 +- - .../CISACodeGen/RegisterPressureEstimate.hpp | 3 +- - IGC/Compiler/CISACodeGen/SLMConstProp.cpp | 3 +- - .../CISACodeGen/ScalarizerCodeGen.cpp | 4 +-- - IGC/Compiler/CISACodeGen/ShaderCodeGen.cpp | 4 +++ - .../CISACodeGen/Simd32Profitability.cpp | 11 ++++-- - IGC/Compiler/CISACodeGen/SimplifyConstant.cpp | 12 +++---- - IGC/Compiler/CISACodeGen/TypeDemote.cpp | 3 +- - .../CISACodeGen/VariableReuseAnalysis.cpp | 5 +-- - .../CISACodeGen/VariableReuseAnalysis.hpp | 3 +- - IGC/Compiler/CISACodeGen/VectorPreProcess.cpp | 24 ++++++------- - IGC/Compiler/CISACodeGen/VectorProcess.cpp | 4 +-- - IGC/Compiler/CISACodeGen/helper.cpp | 10 +++--- - IGC/Compiler/ConvertMSAAPayloadTo16Bit.cpp | 15 ++++++-- - IGC/Compiler/CustomSafeOptPass.cpp | 26 +++++++------- - IGC/Compiler/DebugInfo/ScalarVISAModule.cpp | 3 +- - IGC/Compiler/GenTTI.cpp | 8 ++++- - IGC/Compiler/GenUpdateCB.cpp | 3 +- - IGC/Compiler/HandleFRemInstructions.cpp | 3 +- - IGC/Compiler/HandleLoadStoreInstructions.cpp | 4 +-- - IGC/Compiler/LegalizationPass.cpp | 28 +++++++-------- - IGC/Compiler/Legalizer/InstPromoter.cpp | 11 ++++-- - IGC/Compiler/Legalizer/InstScalarizer.cpp | 6 ++-- - .../Legalizer/PeepholeTypeLegalizer.cpp | 6 ++-- - IGC/Compiler/Legalizer/TypeLegalizer.cpp | 3 +- - IGC/Compiler/Legalizer/TypeLegalizer.h | 3 +- - IGC/Compiler/LowPrecisionOptPass.cpp | 2 +- - .../AddressSpaceAliasAnalysis.cpp | 9 ++--- - .../AggregateArguments/AggregateArguments.cpp | 3 +- - .../CorrectlyRoundedDivSqrt.cpp | 3 +- - .../DeviceEnqueueFuncs/TransformBlocks.cpp | 6 +++- - .../Optimizer/OpenCLPasses/KernelArgs.cpp | 3 +- - .../OpenCLPrintf/OpenCLPrintfResolution.cpp | 4 +-- - .../PrivateMemory/PrivateMemoryResolution.cpp | 11 ++++-- - .../ReplaceUnsupportedIntrinsics.cpp | 4 +-- - .../OpenCLPasses/WIFuncs/WIFuncResolution.cpp | 4 +++ - .../Optimizer/PreCompiledFuncImport.cpp | 4 +-- - IGC/Compiler/Optimizer/Scalarizer.cpp | 36 +++++++++---------- - IGC/Compiler/Optimizer/ValueTracker.cpp | 2 +- - IGC/Compiler/VerificationPass.cpp | 2 +- - IGC/DebugInfo/DebugInfoUtils.hpp | 4 +++ - IGC/DebugInfo/DwarfDebug.cpp | 8 +++++ - IGC/GenISAIntrinsics/GenIntrinsics.cpp | 6 ++-- - .../BuiltinsFrontendDefinitions.hpp | 2 +- - .../include/llvmWrapper/Support/TypeSize.h | 8 ++++- - .../include/llvmWrapper/Transforms/Scalar.h | 4 +-- - IGC/common/igc_resourceDimTypes.h | 5 +++ - 66 files changed, 303 insertions(+), 196 deletions(-) - -diff --git a/IGC/AdaptorCommon/LegalizeFunctionSignatures.cpp b/IGC/AdaptorCommon/LegalizeFunctionSignatures.cpp -index 4046b4a6..8ed36089 100644 ---- a/IGC/AdaptorCommon/LegalizeFunctionSignatures.cpp -+++ b/IGC/AdaptorCommon/LegalizeFunctionSignatures.cpp -@@ -156,7 +156,7 @@ inline Type* LegalizedIntVectorType(Module& M, const Type* const oldTy) - else if (size <= 64) newSize = 64; - else IGC_ASSERT_MESSAGE(0, "Currently don't support upscaling int sizes > 64 bits"); - -- return IGCLLVM::FixedVectorType::get(IntegerType::get(M.getContext(), newSize), (unsigned)cast<VectorType>(oldTy)->getNumElements()); -+ return IGCLLVM::FixedVectorType::get(IntegerType::get(M.getContext(), newSize), (unsigned)cast<IGCLLVM::FixedVectorType>(oldTy)->getNumElements()); - } - - void LegalizeFunctionSignatures::FixFunctionSignatures() -diff --git a/IGC/AdaptorOCL/SPIRV/SPIRVReader.cpp b/IGC/AdaptorOCL/SPIRV/SPIRVReader.cpp -index 41d1f2ea..725a1512 100644 ---- a/IGC/AdaptorOCL/SPIRV/SPIRVReader.cpp -+++ b/IGC/AdaptorOCL/SPIRV/SPIRVReader.cpp -@@ -1576,7 +1576,11 @@ void SPIRVToLLVMDbgTran::transDbgInfo(SPIRVValue *SV, Value *V) { - Line->getColumn(), scope, iat); - - if(scope && !isa<DIFile>(scope)) -+#if LLVM_VERSION_MAJOR >= 12 -+ I->setDebugLoc(DILocation::get(scope->getContext(), Line->getLine(), Line->getColumn(), -+#else - I->setDebugLoc(DebugLoc::get(Line->getLine(), Line->getColumn(), -+#endif - scope, iat)); - } - } -@@ -1925,7 +1929,11 @@ SPIRVToLLVM::transType(SPIRVType *T) { - auto name = isSubgroupAvcINTELTypeOpCode(OC) ? - OCLSubgroupINTELTypeOpCodeMap::rmap(OC) : - BuiltinOpaqueGenericTypeOpCodeMap::rmap(OC); -+#if LLVM_VERSION_MAJOR >= 12 -+ auto *pST = llvm::StructType::getTypeByName(M->getContext(), name); -+#else - auto *pST = M->getTypeByName(name); -+#endif - pST = pST ? pST : StructType::create(*Context, name); - - return mapType(T, PointerType::get(pST, getOCLOpaqueTypeAddrSpace(OC))); -@@ -2403,7 +2411,7 @@ Value *SPIRVToLLVM::promoteBool(Value *pVal, BasicBlock *BB) - - auto *PromoType = isa<VectorType>(pVal->getType()) ? - cast<Type>(IGCLLVM::FixedVectorType::get(Type::getInt8Ty(pVal->getContext()), -- (unsigned)cast<VectorType>(pVal->getType())->getNumElements())) : -+ (unsigned)cast<IGCLLVM::FixedVectorType>(pVal->getType())->getNumElements())) : - Type::getInt8Ty(pVal->getContext()); - - if (auto *C = dyn_cast<Constant>(pVal)) -@@ -2445,7 +2453,7 @@ Value *SPIRVToLLVM::truncBool(Value *pVal, BasicBlock *BB) - - auto *TruncType = isa<VectorType>(pVal->getType()) ? - cast<Type>(IGCLLVM::FixedVectorType::get(Type::getInt1Ty(pVal->getContext()), -- (unsigned)cast<VectorType>(pVal->getType())->getNumElements())) : -+ (unsigned)cast<IGCLLVM::FixedVectorType>(pVal->getType())->getNumElements())) : - Type::getInt1Ty(pVal->getContext()); - - if (auto *C = dyn_cast<Constant>(pVal)) -@@ -2491,7 +2499,7 @@ Type *SPIRVToLLVM::truncBoolType(SPIRVType *SPVType, Type *LLType) - - return isa<VectorType>(LLType) ? - cast<Type>(IGCLLVM::FixedVectorType::get(Type::getInt1Ty(LLType->getContext()), -- (unsigned)cast<VectorType>(LLType)->getNumElements())) : -+ (unsigned)cast<IGCLLVM::FixedVectorType>(LLType)->getNumElements())) : - Type::getInt1Ty(LLType->getContext()); - } - -@@ -2648,7 +2656,7 @@ SPIRVToLLVM::transValueWithoutDecoration(SPIRVValue *BV, Function *F, - { - if(CV[i]->getType()->isVectorTy()) - { -- for(uint32_t j = 0; j < cast<VectorType>(CV[i]->getType())->getNumElements(); j++) -+ for(uint32_t j = 0; j < cast<IGCLLVM::FixedVectorType>(CV[i]->getType())->getNumElements(); j++) - { - Value *v = ExtractElementInst::Create( CV[i],ConstantInt::get( *Context,APInt( 32,j ) ),BCC->getName(),BB ); - elm1 = CreateCompositeConstruct( elm1,v,pos++ ); -@@ -3336,7 +3344,7 @@ SPIRVToLLVM::transValueWithoutDecoration(SPIRVValue *BV, Function *F, - auto Vector = transValue(BI->getOperand(0), F, BB); - auto Scalar = transValue(BI->getOperand(1), F, BB); - -- auto VecType = cast<VectorType>(Vector->getType()); -+ auto VecType = cast<IGCLLVM::FixedVectorType>(Vector->getType()); - auto Undef = UndefValue::get(VecType); - - auto ScalarVec = InsertElementInst::Create(Undef, Scalar, -@@ -3361,7 +3369,7 @@ SPIRVToLLVM::transValueWithoutDecoration(SPIRVValue *BV, Function *F, - a->getType()->getScalarSizeInBits() - 1); - auto *ShiftOp = isa<VectorType>(a->getType()) ? - ConstantVector::getSplat( -- IGCLLVM::getElementCount((unsigned)cast<VectorType>(a->getType())->getNumElements()), ShiftAmt) : -+ IGCLLVM::getElementCount((unsigned)cast<IGCLLVM::FixedVectorType>(a->getType())->getNumElements()), ShiftAmt) : - ShiftAmt; - - // OCL C: -@@ -3705,15 +3713,15 @@ SPIRVToLLVM::transSPIRVBuiltinFromInst(SPIRVInstruction *BI, BasicBlock *BB) { - "", - BB); - } -- else if (cast<VectorType>(coordType)->getNumElements() != 4) -+ else if (cast<IGCLLVM::FixedVectorType>(coordType)->getNumElements() != 4) - { - Value *undef = UndefValue::get(coordType); - - SmallVector<Constant*, 4> shuffleIdx; -- for (unsigned i = 0; i < cast<VectorType>(coordType)->getNumElements(); i++) -+ for (unsigned i = 0; i < cast<IGCLLVM::FixedVectorType>(coordType)->getNumElements(); i++) - shuffleIdx.push_back(ConstantInt::get(Type::getInt32Ty(*Context), i)); - -- for (uint64_t i = (unsigned)cast<VectorType>(coordType)->getNumElements(); i < 4; i++) -+ for (uint64_t i = (unsigned)cast<IGCLLVM::FixedVectorType>(coordType)->getNumElements(); i < 4; i++) - shuffleIdx.push_back(ConstantInt::get(Type::getInt32Ty(*Context), 0)); - - imageCoordinateWiden = new ShuffleVectorInst( -diff --git a/IGC/AdaptorOCL/SPIRV/SPIRVUtil.cpp b/IGC/AdaptorOCL/SPIRV/SPIRVUtil.cpp -index 57821556..91b4623c 100644 ---- a/IGC/AdaptorOCL/SPIRV/SPIRVUtil.cpp -+++ b/IGC/AdaptorOCL/SPIRV/SPIRVUtil.cpp -@@ -93,7 +93,11 @@ saveLLVMModule(Module *M, const std::string &OutputFile) { - PointerType* - getOrCreateOpaquePtrType(Module *M, const std::string &Name, - unsigned AddrSpace) { -+#if LLVM_VERSION_MAJOR >= 12 -+ auto OpaqueType = llvm::StructType::getTypeByName(M->getContext(), Name); -+#else - auto OpaqueType = M->getTypeByName(Name); -+#endif - if (!OpaqueType) - OpaqueType = StructType::create(M->getContext(), Name); - return PointerType::get(OpaqueType, AddrSpace); -@@ -159,7 +163,7 @@ std::string recursive_mangle(const Type* pType) - return "i" + utostr(pType->getIntegerBitWidth()); - case IGCLLVM::VectorTyID: - { -- unsigned vecLen = (unsigned)cast<VectorType>(pType)->getNumElements(); -+ unsigned vecLen = (unsigned)cast<IGCLLVM::FixedVectorType>(pType)->getNumElements(); - Type* pEltType = cast<VectorType>(pType)->getElementType(); - return "v" + utostr(vecLen) + recursive_mangle(pEltType); - } -diff --git a/IGC/Compiler/CISACodeGen/AdvMemOpt.cpp b/IGC/Compiler/CISACodeGen/AdvMemOpt.cpp -index 81acc6ce..fc45a510 100644 ---- a/IGC/Compiler/CISACodeGen/AdvMemOpt.cpp -+++ b/IGC/Compiler/CISACodeGen/AdvMemOpt.cpp -@@ -134,7 +134,11 @@ bool AdvMemOpt::runOnFunction(Function& F) { - for (auto I = LI->begin(), E = LI->end(); I != E; ++I) - for (auto DFI = df_begin(*I), DFE = df_end(*I); DFI != DFE; ++DFI) { - Loop* L = *DFI; -+#if LLVM_VERSION_MAJOR >= 12 -+ if (L->isInnermost()) -+#else - if (L->empty()) -+#endif - InnermostLoops.push_back(L); - } - -diff --git a/IGC/Compiler/CISACodeGen/CShader.cpp b/IGC/Compiler/CISACodeGen/CShader.cpp -index e3cb8153..ebc99615 100644 ---- a/IGC/Compiler/CISACodeGen/CShader.cpp -+++ b/IGC/Compiler/CISACodeGen/CShader.cpp -@@ -430,7 +430,7 @@ void CShader::CreateAliasVars() - continue; - - Type* Ty = V->getType(); -- VectorType* VTy = dyn_cast<VectorType>(Ty); -+ IGCLLVM::FixedVectorType* VTy = dyn_cast<IGCLLVM::FixedVectorType>(Ty); - Type* BTy = VTy ? VTy->getElementType() : Ty; - int nelts = (VTy ? (int)VTy->getNumElements() : 1); - -@@ -961,7 +961,7 @@ bool CShader::InsideDivergentCF(llvm::Instruction* inst) - uint CShader::GetNbVectorElementAndMask(llvm::Value* val, uint32_t& mask) - { - llvm::Type* type = val->getType(); -- uint nbElement = int_cast<uint>(cast<VectorType>(type)->getNumElements()); -+ uint nbElement = int_cast<uint>(cast<IGCLLVM::FixedVectorType>(type)->getNumElements()); - mask = 0; - // we don't process vector bigger than 31 elements as the mask has only 32bits - // If we want to support longer vectors we need to extend the mask size -@@ -1166,7 +1166,7 @@ uint32_t CShader::GetExtractMask(llvm::Value* vecVal) - { - return it->second; - } -- const unsigned int numChannels = vecVal->getType()->isVectorTy() ? (unsigned)cast<VectorType>(vecVal->getType())->getNumElements() : 1; -+ const unsigned int numChannels = vecVal->getType()->isVectorTy() ? (unsigned)cast<IGCLLVM::FixedVectorType>(vecVal->getType())->getNumElements() : 1; - IGC_ASSERT_MESSAGE(numChannels <= 32, "Mask has 32 bits maximally!"); - return (1ULL << numChannels) - 1; - } -@@ -1174,7 +1174,7 @@ uint32_t CShader::GetExtractMask(llvm::Value* vecVal) - uint16_t CShader::AdjustExtractIndex(llvm::Value* vecVal, uint16_t index) - { - uint16_t result = index; -- if (cast<VectorType>(vecVal->getType())->getNumElements() < 32) -+ if (cast<IGCLLVM::FixedVectorType>(vecVal->getType())->getNumElements() < 32) - { - uint32_t mask = GetExtractMask(vecVal); - for (uint i = 0; i < index; ++i) -@@ -1591,7 +1591,7 @@ auto sizeToSIMDMode = [](uint32_t size) - - CVariable* CShader::GetConstant(llvm::Constant* C, CVariable* dstVar) - { -- llvm::VectorType* VTy = llvm::dyn_cast<llvm::VectorType>(C->getType()); -+ IGCLLVM::FixedVectorType* VTy = llvm::dyn_cast<IGCLLVM::FixedVectorType>(C->getType()); - if (C && VTy) - { // Vector constant - llvm::Type* eTy = VTy->getElementType(); -@@ -1816,7 +1816,7 @@ uint32_t CShader::GetNumElts(llvm::Type* type, bool isUniform) - { - IGC_ASSERT(type->getContainedType(0)->isIntegerTy() || type->getContainedType(0)->isFloatingPointTy()); - -- auto VT = cast<VectorType>(type); -+ auto VT = cast<IGCLLVM::FixedVectorType>(type); - numElts *= (uint16_t)VT->getNumElements(); - } - return numElts; -@@ -2516,7 +2516,7 @@ CVariable* CShader::GetSymbol(llvm::Value* value, bool fromConstantPool) - if (isVecType) - { - // Map the entire vector value to the CVar -- unsigned numElements = (unsigned)cast<VectorType>(value->getType())->getNumElements(); -+ unsigned numElements = (unsigned)cast<IGCLLVM::FixedVectorType>(value->getType())->getNumElements(); - var = GetNewVariable(numElements, ISA_TYPE_UQ, (GetContext()->platform.getGRFSize() == 64) ? EALIGN_32WORD : EALIGN_HWORD, true, 1, valName); - symbolMapping.insert(std::pair<llvm::Value*, CVariable*>(value, var)); - -diff --git a/IGC/Compiler/CISACodeGen/ConstantCoalescing.cpp b/IGC/Compiler/CISACodeGen/ConstantCoalescing.cpp -index 1efe116f..47b7c6c0 100644 ---- a/IGC/Compiler/CISACodeGen/ConstantCoalescing.cpp -+++ b/IGC/Compiler/CISACodeGen/ConstantCoalescing.cpp -@@ -224,7 +224,7 @@ static void checkInsertExtractMatch(InsertElementInst* insertInst, Value* base, - - static bool canReplaceInsert(InsertElementInst* insertElt) - { -- VectorType* VTy = cast<VectorType>(insertElt->getOperand(0)->getType()); -+ IGCLLVM::FixedVectorType* VTy = cast<IGCLLVM::FixedVectorType>(insertElt->getOperand(0)->getType()); - ConstantInt* index = dyn_cast<ConstantInt>(insertElt->getOperand(2)); - if (!index || index->getZExtValue() != VTy->getNumElements() - 1) - { -@@ -312,7 +312,7 @@ void ConstantCoalescing::VectorizePrep(llvm::BasicBlock* bb) - { - if (load->getType()->isVectorTy() && wiAns->isUniform(load)) - { -- srcNElts = (uint32_t)cast<VectorType>(load->getType())->getNumElements(); -+ srcNElts = (uint32_t)cast<IGCLLVM::FixedVectorType>(load->getType())->getNumElements(); - DenseMap<uint64_t, Instruction*> extractElementMap; - - for (auto iter = load->user_begin(); iter != load->user_end(); iter++) -@@ -396,7 +396,7 @@ bool ConstantCoalescing::isProfitableLoad( - (isa<LoadInst>(I) && wiAns->isUniform(I)) ? - 16 : 4; - -- if (cast<VectorType>(LoadTy)->getNumElements() > MaxVectorInput) -+ if (cast<IGCLLVM::FixedVectorType>(LoadTy)->getNumElements() > MaxVectorInput) - return false; - - MaxEltPlus = CheckVectorElementUses(I); -@@ -1787,7 +1787,7 @@ void ConstantCoalescing::AdjustChunk(BufChunk* cov_chunk, uint start_adj, uint s - WIAnalysis::WIDependancy loadDep = wiAns->whichDepend(cov_chunk->chunkIO); - irBuilder->SetInsertPoint(cov_chunk->chunkIO->getNextNode()); - Value* vec = UndefValue::get(originalType); -- for (unsigned i = 0; i < cast<VectorType>(originalType)->getNumElements(); i++) -+ for (unsigned i = 0; i < cast<IGCLLVM::FixedVectorType>(originalType)->getNumElements(); i++) - { - Value* channel = irBuilder->CreateExtractElement( - cov_chunk->chunkIO, irBuilder->getInt32(i + start_adj)); -@@ -1851,7 +1851,7 @@ void ConstantCoalescing::MoveExtracts(BufChunk* cov_chunk, Instruction* load, ui - irBuilder->SetInsertPoint(load->getNextNode()); - Type* vecType = load->getType(); - Value* vec = UndefValue::get(vecType); -- for (unsigned i = 0; i < cast<VectorType>(vecType)->getNumElements(); i++) -+ for (unsigned i = 0; i < cast<IGCLLVM::FixedVectorType>(vecType)->getNumElements(); i++) - { - Value* channel = irBuilder->CreateExtractElement( - cov_chunk->chunkIO, irBuilder->getInt32(i + start_adj)); -@@ -1915,7 +1915,7 @@ void ConstantCoalescing::EnlargeChunk(BufChunk* cov_chunk, uint size_adj) - WIAnalysis::WIDependancy loadDep = wiAns->whichDepend(cov_chunk->chunkIO); - irBuilder->SetInsertPoint(cov_chunk->chunkIO->getNextNode()); - Value* vec = UndefValue::get(originalType); -- for (unsigned i = 0; i < cast<VectorType>(originalType)->getNumElements(); i++) -+ for (unsigned i = 0; i < cast<IGCLLVM::FixedVectorType>(originalType)->getNumElements(); i++) - { - Value* channel = irBuilder->CreateExtractElement( - cov_chunk->chunkIO, irBuilder->getInt32(i)); -@@ -2343,7 +2343,7 @@ void ConstantCoalescing::ReplaceLoadWithSamplerLoad( - if (dstTy->isVectorTy()) - { - result = UndefValue::get(dstTy); -- for (uint i = 0; i < cast<VectorType>(dstTy)->getNumElements(); i++) -+ for (uint i = 0; i < cast<IGCLLVM::FixedVectorType>(dstTy)->getNumElements(); i++) - { - Value* tmpData = ExtractFromSamplerData(cast<VectorType>(dstTy)->getElementType(), i); - result = irBuilder->CreateInsertElement(result, tmpData, irBuilder->getInt32(i)); -diff --git a/IGC/Compiler/CISACodeGen/DeSSA.cpp b/IGC/Compiler/CISACodeGen/DeSSA.cpp -index f540dc11..6fab0246 100644 ---- a/IGC/Compiler/CISACodeGen/DeSSA.cpp -+++ b/IGC/Compiler/CISACodeGen/DeSSA.cpp -@@ -83,6 +83,7 @@ IN THE SOFTWARE. - #include "Compiler/IGCPassSupport.h" - #include "common/LLVMWarningsPush.hpp" - #include <llvm/IR/InstIterator.h> -+#include <llvmWrapper/IR/DerivedTypes.h> - #include "common/LLVMWarningsPop.hpp" - #include <algorithm> - #include "Probe/Assertion.h" -@@ -1284,7 +1285,7 @@ int DeSSA::checkInsertElementAlias( - // - // If found, return the actual vector size; - // otherwise, return 0. -- VectorType* VTy = cast<VectorType>(IEI->getType()); -+ IGCLLVM::FixedVectorType* VTy = cast<IGCLLVM::FixedVectorType>(IEI->getType()); - IGC_ASSERT(nullptr != VTy); - int nelts = (int)VTy->getNumElements(); - AllIEIs.resize(nelts, nullptr); -diff --git a/IGC/Compiler/CISACodeGen/EmitVISAPass.cpp b/IGC/Compiler/CISACodeGen/EmitVISAPass.cpp -index 3ab1cc5f..4502ef62 100644 ---- a/IGC/Compiler/CISACodeGen/EmitVISAPass.cpp -+++ b/IGC/Compiler/CISACodeGen/EmitVISAPass.cpp -@@ -53,6 +53,7 @@ IN THE SOFTWARE. - #include "Compiler/IGCPassSupport.h" - #include "common/LLVMWarningsPush.hpp" - #include "llvmWrapper/IR/Instructions.h" -+#include "llvmWrapper/IR/DerivedTypes.h" - #include "llvm/Support/Path.h" - #include "llvmWrapper/IR/Intrinsics.h" - #include "common/LLVMWarningsPop.hpp" -@@ -1165,7 +1166,7 @@ bool EmitPass::canRelocatePhiMov( - if (dst != src) - { - int numElt = 1; -- if (VectorType * vTy = dyn_cast<VectorType>(PN->getType())) -+ if (IGCLLVM::FixedVectorType * vTy = dyn_cast<IGCLLVM::FixedVectorType>(PN->getType())) - { - numElt = int_cast<int>(vTy->getNumElements()); - } -@@ -1302,7 +1303,7 @@ void EmitPass::MovPhiSources(llvm::BasicBlock* aBB) - phiSrcDstList.push_back(phiInfo); - - int numElt = 0; -- if (VectorType * vTy = dyn_cast<VectorType>(PN->getType())) -+ if (IGCLLVM::FixedVectorType * vTy = dyn_cast<IGCLLVM::FixedVectorType>(PN->getType())) - { - numElt = int_cast<int>(vTy->getNumElements()); - } -@@ -5375,7 +5376,7 @@ void EmitPass::emitLegacySimdBlockWrite(llvm::Instruction* inst, llvm::Value* pt - bool useA64 = isA64Ptr(ptrType, m_currShader->GetContext()); - - Type* Ty = dataPtr->getType(); -- VectorType* VTy = dyn_cast<VectorType>(Ty); -+ IGCLLVM::FixedVectorType* VTy = dyn_cast<IGCLLVM::FixedVectorType>(Ty); - uint32_t nbElements = VTy ? int_cast<uint32_t>(VTy->getNumElements()) : 1; - - uint32_t typeSizeInBytes = Ty->getScalarSizeInBits() / 8; -@@ -5621,7 +5622,7 @@ void EmitPass::emitLegacySimdBlockRead(llvm::Instruction* inst, llvm::Value* ptr - bool useA64 = isA64Ptr(ptrType, m_currShader->GetContext()); - - Type* Ty = inst->getType(); -- VectorType* VTy = dyn_cast<VectorType>(Ty); -+ IGCLLVM::FixedVectorType* VTy = dyn_cast<IGCLLVM::FixedVectorType>(Ty); - uint32_t nbElements = VTy ? int_cast<uint32_t>(VTy->getNumElements()) : 1; - - uint32_t typeSizeInBytes = Ty->getScalarSizeInBits() / 8; -@@ -5913,7 +5914,7 @@ void EmitPass::emitMediaBlockIO(const llvm::GenIntrinsicInst* inst, bool isRead) - }; - - uint nElts = isa<VectorType>(pDataType) ? -- (uint)cast<VectorType>(pDataType)->getNumElements() : -+ (uint)cast<IGCLLVM::FixedVectorType>(pDataType)->getNumElements() : - 1; - - // Now, do the copies. -@@ -6003,7 +6004,7 @@ void EmitPass::emitSimdMediaBlockRead(llvm::Instruction* inst) - uint32_t nbElements = 1; - if (inst->getType()->isVectorTy()) - { -- nbElements = (uint32_t)cast<VectorType>(inst->getType())->getNumElements(); -+ nbElements = (uint32_t)cast<IGCLLVM::FixedVectorType>(inst->getType())->getNumElements(); - } - IGC_ASSERT_MESSAGE(nbElements <= 8, "InValid Vector Size"); - -@@ -6247,7 +6248,7 @@ void EmitPass::emitSimdMediaBlockWrite(llvm::Instruction* inst) - uint32_t nbElements = 1; - if (dataPtr->getType()->isVectorTy()) - { -- nbElements = (uint32_t)cast<VectorType>(dataPtr->getType())->getNumElements(); -+ nbElements = (uint32_t)cast<IGCLLVM::FixedVectorType>(dataPtr->getType())->getNumElements(); - } - IGC_ASSERT_MESSAGE(nbElements <= 8, "InValid Vector Size"); - -@@ -8993,8 +8994,8 @@ void EmitPass::emitBitCast(llvm::BitCastInst* btCst) - { - Type* srcType = btCst->getOperand(0)->getType(); - Type* dstType = btCst->getType(); -- unsigned int numSrcElement = srcType->isVectorTy() ? (unsigned)cast<VectorType>(srcType)->getNumElements() : 1; -- unsigned int numDstElement = dstType->isVectorTy() ? (unsigned)cast<VectorType>(dstType)->getNumElements() : 1; -+ unsigned int numSrcElement = srcType->isVectorTy() ? (unsigned)cast<IGCLLVM::FixedVectorType>(srcType)->getNumElements() : 1; -+ unsigned int numDstElement = dstType->isVectorTy() ? (unsigned)cast<IGCLLVM::FixedVectorType>(dstType)->getNumElements() : 1; - - if (srcType->isPointerTy()) - { -@@ -9468,7 +9469,7 @@ void EmitPass::emitLoad3DInner(LdRawIntrinsic* inst, ResourceDescriptor& resourc - { - IGC_ASSERT_MESSAGE(predDefSurface != ESURFACE_STATELESS, "scratch cannot be uniform"); - Type* loadType = inst->getType(); -- uint numElement = loadType->isVectorTy() ? (uint)cast<VectorType>(loadType)->getNumElements() : 1; -+ uint numElement = loadType->isVectorTy() ? (uint)cast<IGCLLVM::FixedVectorType>(loadType)->getNumElements() : 1; - if (predDefSurface == ESURFACE_SLM) - { - IGC_ASSERT(numElement <= 4); -@@ -10795,7 +10796,7 @@ void EmitPass::emitInsert(llvm::Instruction* inst) - pVecVar = GetSymbol(pVec); - if (pVecVar != pInstVar) - { -- emitVectorCopy(pInstVar, pVecVar, int_cast<unsigned>(dyn_cast<VectorType>(pVecType)->getNumElements())); -+ emitVectorCopy(pInstVar, pVecVar, int_cast<unsigned>(dyn_cast<IGCLLVM::FixedVectorType>(pVecType)->getNumElements())); - } - } - } -@@ -14596,7 +14597,7 @@ bool EmitPass::isUniformStoreOCL(llvm::StoreInst* SI) - - Value* storeVal = SI->getValueOperand(); - Type* Ty = storeVal->getType(); -- VectorType* VTy = dyn_cast<VectorType>(Ty); -+ IGCLLVM::FixedVectorType* VTy = dyn_cast<IGCLLVM::FixedVectorType>(Ty); - uint32_t elts = VTy ? int_cast<uint32_t>(VTy->getNumElements()) : 1; - Type* eltTy = VTy ? VTy->getElementType() : Ty; - -@@ -14626,7 +14627,7 @@ void EmitPass::emitVectorBitCast(llvm::BitCastInst* BCI) - if (srcTy->isVectorTy()) - { - srcEltTy = cast<VectorType>(srcTy)->getElementType(); -- srcNElts = (uint32_t)cast<VectorType>(srcTy)->getNumElements(); -+ srcNElts = (uint32_t)cast<IGCLLVM::FixedVectorType>(srcTy)->getNumElements(); - } - else - { -@@ -14636,7 +14637,7 @@ void EmitPass::emitVectorBitCast(llvm::BitCastInst* BCI) - if (dstTy->isVectorTy()) - { - dstEltTy = cast<VectorType>(dstTy)->getElementType(); -- dstNElts = (uint32_t)cast<VectorType>(dstTy)->getNumElements(); -+ dstNElts = (uint32_t)cast<IGCLLVM::FixedVectorType>(dstTy)->getNumElements(); - } - else - { -@@ -15215,7 +15216,7 @@ void EmitPass::emitVectorLoad(LoadInst* inst, Value* offset, ConstantInt* immOff - } - - Type* Ty = inst->getType(); -- VectorType* VTy = dyn_cast<VectorType>(Ty); -+ IGCLLVM::FixedVectorType* VTy = dyn_cast<IGCLLVM::FixedVectorType>(Ty); - Type* eltTy = VTy ? VTy->getElementType() : Ty; - uint32_t eltBytes = GetScalarTypeSizeInRegister(eltTy); - IGC_ASSERT_MESSAGE((eltBytes == 1) || (eltBytes == 2) || (eltBytes == 4) || (eltBytes == 8), -@@ -15667,7 +15668,7 @@ void EmitPass::emitVectorStore(StoreInst* inst, Value* offset, ConstantInt* immO - - Value* storedVal = inst->getValueOperand(); - Type* Ty = storedVal->getType(); -- VectorType* VTy = dyn_cast<VectorType>(Ty); -+ IGCLLVM::FixedVectorType* VTy = dyn_cast<IGCLLVM::FixedVectorType>(Ty); - Type* eltTy = VTy ? VTy->getElementType() : Ty; - uint32_t eltBytes = GetScalarTypeSizeInRegister(eltTy); - -@@ -16471,7 +16472,7 @@ void EmitPass::emitCopyAll(CVariable* Dst, CVariable* Src, llvm::Type* Ty) - } - else if (Ty->isVectorTy()) - { -- unsigned NElts = (unsigned)cast<VectorType>(Ty)->getNumElements(); -+ unsigned NElts = (unsigned)cast<IGCLLVM::FixedVectorType>(Ty)->getNumElements(); - emitVectorCopy(Dst, Src, NElts); - } - else -diff --git a/IGC/Compiler/CISACodeGen/Emu64OpsPass.cpp b/IGC/Compiler/CISACodeGen/Emu64OpsPass.cpp -index 28248f30..767b477a 100644 ---- a/IGC/Compiler/CISACodeGen/Emu64OpsPass.cpp -+++ b/IGC/Compiler/CISACodeGen/Emu64OpsPass.cpp -@@ -1961,7 +1961,7 @@ bool InstExpander::visitExtractElement(ExtractElementInst& EEI) { - // later. - - Value* V = EEI.getVectorOperand(); -- unsigned NumElts = (unsigned)cast<VectorType>(V->getType())->getNumElements(); -+ unsigned NumElts = (unsigned)cast<IGCLLVM::FixedVectorType>(V->getType())->getNumElements(); - V = IRB->CreateBitCast(V, Emu->getV2Int32Ty(NumElts)); - // Re-calculate indices to Lo and Hi parts. - Value* Idx = EEI.getIndexOperand(); -@@ -1998,7 +1998,7 @@ bool InstExpander::visitInsertElement(InsertElementInst& IEI) { - - // Create the emulated vector. - Value* NewVal = IEI.getOperand(0); -- unsigned NumElts = (unsigned)cast<VectorType>(NewVal->getType())->getNumElements(); -+ unsigned NumElts = (unsigned)cast<IGCLLVM::FixedVectorType>(NewVal->getType())->getNumElements(); - NewVal = IRB->CreateBitCast(NewVal, Emu->getV2Int32Ty(NumElts)); - // Re-calculate indices to Lo and Hi parts. - Value* Idx = IEI.getOperand(2); -diff --git a/IGC/Compiler/CISACodeGen/GenIRLowering.cpp b/IGC/Compiler/CISACodeGen/GenIRLowering.cpp -index c74a2ac6..78e8a4aa 100644 ---- a/IGC/Compiler/CISACodeGen/GenIRLowering.cpp -+++ b/IGC/Compiler/CISACodeGen/GenIRLowering.cpp -@@ -453,7 +453,7 @@ Value* GEPLowering::getSExtOrTrunc(Value* Val, Type* NewTy) const { - - IGC_ASSERT_MESSAGE(OldTy->isIntOrIntVectorTy(), "Index should be Integer or vector of Integer!"); - -- if (auto OldVecTy = dyn_cast<VectorType>(OldTy)) { -+ if (auto OldVecTy = dyn_cast<IGCLLVM::FixedVectorType>(OldTy)) { - OldWidth = (unsigned)OldVecTy->getNumElements() * OldVecTy->getElementType()->getIntegerBitWidth(); - NewWidth = (unsigned)OldVecTy->getNumElements() * NewTy->getIntegerBitWidth(); - } -@@ -805,7 +805,7 @@ bool GEPLowering::lowerGetElementPtrInst(GetElementPtrInst* GEP) const - } - else - { -- if (auto NewIdxVT = dyn_cast<VectorType>(NewIdx->getType())) { -+ if (auto NewIdxVT = dyn_cast<IGCLLVM::FixedVectorType>(NewIdx->getType())) { - Value* result = llvm::UndefValue::get(FixedVectorType::get(PtrMathTy, (unsigned)NewIdxVT->getNumElements())); - for (uint32_t j = 0; j < (uint32_t)NewIdxVT->getNumElements(); j++) { - result = Builder->CreateInsertElement(result, PointerValue, Builder->getInt32(j)); -diff --git a/IGC/Compiler/CISACodeGen/GenSimplification.cpp b/IGC/Compiler/CISACodeGen/GenSimplification.cpp -index 8c38f229..f432061e 100644 ---- a/IGC/Compiler/CISACodeGen/GenSimplification.cpp -+++ b/IGC/Compiler/CISACodeGen/GenSimplification.cpp -@@ -35,6 +35,7 @@ IN THE SOFTWARE. - #include <llvm/IR/InstVisitor.h> - #include <llvm/IR/Instruction.h> - #include <llvm/Support/raw_ostream.h> -+#include <llvmWrapper/IR/DerivedTypes.h> - #include "common/LLVMWarningsPop.hpp" - #include "Probe/Assertion.h" - -@@ -164,7 +165,7 @@ bool GenSimplification::simplifyVectorPHINodeCase2(PHINode& PN) const { - - Type* Ty = PN.getType(); - Type* EltTy = Ty->getScalarType(); -- unsigned NumElts = (unsigned)cast<VectorType>(Ty)->getNumElements(); -+ unsigned NumElts = (unsigned)cast<IGCLLVM::FixedVectorType>(Ty)->getNumElements(); - - SmallVector<Value*, 8> Lanes; - SmallVector<SmallVector<Value*, 8>, 4> Values; -@@ -238,7 +239,7 @@ void GenSimplification::visitPHINode(PHINode& PN) { - void GenSimplification::visitExtractElement(ExtractElementInst& EEI) { - // Skip non-2-element vector. - Value* Vec = EEI.getVectorOperand(); -- VectorType* VTy = cast<VectorType>(Vec->getType()); -+ IGCLLVM::FixedVectorType* VTy = cast<IGCLLVM::FixedVectorType>(Vec->getType()); - if (VTy->getNumElements() != 2) - return; - -diff --git a/IGC/Compiler/CISACodeGen/GeometryShaderLowering.cpp b/IGC/Compiler/CISACodeGen/GeometryShaderLowering.cpp -index e00e8469..a60b9892 100644 ---- a/IGC/Compiler/CISACodeGen/GeometryShaderLowering.cpp -+++ b/IGC/Compiler/CISACodeGen/GeometryShaderLowering.cpp -@@ -33,6 +33,7 @@ IN THE SOFTWARE. - #include <llvm/IR/PassManager.h> - #include <llvm/IR/IRBuilder.h> - #include <llvm/IR/Verifier.h> -+#include <llvmWrapper/IR/DerivedTypes.h> - #include "common/LLVMWarningsPop.hpp" - #include "Compiler/InitializePasses.h" - #include "Probe/Assertion.h" -@@ -507,7 +508,7 @@ void GeometryShaderLowering::AddURBRead( - { - Value* vec = UndefValue::get(inst->getType()); - IRBuilder<> builder(inst); -- for (unsigned int i = 0; i < cast<VectorType>(inst->getType())->getNumElements(); i++) -+ for (unsigned int i = 0; i < cast<IGCLLVM::FixedVectorType>(inst->getType())->getNumElements(); i++) - { - Value* vecElement = builder.CreateExtractElement(urbRead, builder.getInt32(i)); - vec = builder.CreateInsertElement(vec, vecElement, builder.getInt32(i)); -diff --git a/IGC/Compiler/CISACodeGen/LdShrink.cpp b/IGC/Compiler/CISACodeGen/LdShrink.cpp -index d89fbd0e..2df06df2 100644 ---- a/IGC/Compiler/CISACodeGen/LdShrink.cpp -+++ b/IGC/Compiler/CISACodeGen/LdShrink.cpp -@@ -30,6 +30,7 @@ IN THE SOFTWARE. - #include <llvm/Support/Debug.h> - #include <llvm/Support/MathExtras.h> - #include <llvm/Support/raw_ostream.h> -+#include <llvmWrapper/IR/DerivedTypes.h> - #include "common/LLVMWarningsPop.hpp" - #include "Compiler/CISACodeGen/ShaderCodeGen.hpp" - #include "Compiler/IGCPassSupport.h" -@@ -79,7 +80,7 @@ IGC_INITIALIZE_PASS_BEGIN(LdShrink, PASS_FLAG, PASS_DESC, PASS_CFG_ONLY, PASS_AN - IGC_INITIALIZE_PASS_END(LdShrink, PASS_FLAG, PASS_DESC, PASS_CFG_ONLY, PASS_ANALYSIS) - - unsigned LdShrink::getExtractIndexMask(LoadInst* LI) const { -- VectorType* VTy = dyn_cast<VectorType>(LI->getType()); -+ IGCLLVM::FixedVectorType* VTy = dyn_cast<IGCLLVM::FixedVectorType>(LI->getType()); - // Skip non-vector loads. - if (!VTy) - return 0; -@@ -91,9 +92,12 @@ unsigned LdShrink::getExtractIndexMask(LoadInst* LI) const { - Type* Ty = VTy->getScalarType(); - // Skip non-BYTE addressable data types. So far, check integer types - // only. -- if (IntegerType * ITy = dyn_cast<IntegerType>(Ty)) -- if (!ITy->isPowerOf2ByteWidth()) -+ if (IntegerType * ITy = dyn_cast<IntegerType>(Ty)) { -+ // Unroll isPowerOf2ByteWidth, it was removed in LLVM 12. -+ unsigned BitWidth = ITy->getBitWidth(); -+ if (!((BitWidth > 7) && isPowerOf2_32(BitWidth))) - return 0; -+ } - - unsigned Mask = 0; // Maxmimally 32 elements. - -diff --git a/IGC/Compiler/CISACodeGen/LowerGEPForPrivMem.cpp b/IGC/Compiler/CISACodeGen/LowerGEPForPrivMem.cpp -index c12b20d5..5245117d 100644 ---- a/IGC/Compiler/CISACodeGen/LowerGEPForPrivMem.cpp -+++ b/IGC/Compiler/CISACodeGen/LowerGEPForPrivMem.cpp -@@ -634,7 +634,7 @@ void TransposeHelper::handleGEPInst( - } - else - { -- arr_sz = (unsigned)cast<VectorType>(T)->getNumElements(); -+ arr_sz = (unsigned)cast<IGCLLVM::FixedVectorType>(T)->getNumElements(); - } - T = cast<VectorType>(T)->getElementType(); - } -@@ -656,7 +656,7 @@ void TransposeHelper::handleGEPInst( - } - else if (T->isVectorTy()) - { -- arr_sz = (unsigned)cast<VectorType>(T)->getNumElements(); -+ arr_sz = (unsigned)cast<IGCLLVM::FixedVectorType>(T)->getNumElements(); - T = cast<VectorType>(T)->getElementType(); - } - else -@@ -716,7 +716,7 @@ void TransposeHelperPromote::handleLoadInst( - IRBuilder<> IRB(pLoad); - IGC_ASSERT(nullptr != pLoad->getType()); - unsigned N = pLoad->getType()->isVectorTy() -- ? (unsigned)cast<VectorType>(pLoad->getType())->getNumElements() -+ ? (unsigned)cast<IGCLLVM::FixedVectorType>(pLoad->getType())->getNumElements() - : 1; - Value* Val = loadEltsFromVecAlloca(N, pVecAlloca, pScalarizedIdx, IRB, pLoad->getType()->getScalarType()); - pLoad->replaceAllUsesWith(Val); -@@ -748,7 +748,7 @@ void TransposeHelperPromote::handleStoreInst( - // %v1 = extractelement <2 x float> %v, i32 1 - // %w1 = insertelement <32 x float> %w0, float %v1, i32 %idx+1 - // store <32 x float> %w1, <32 x float>* %ptr1 -- for (unsigned i = 0, e = (unsigned)cast<VectorType>(pStoreVal->getType())->getNumElements(); i < e; ++i) -+ for (unsigned i = 0, e = (unsigned)cast<IGCLLVM::FixedVectorType>(pStoreVal->getType())->getNumElements(); i < e; ++i) - { - Value* VectorIdx = ConstantInt::get(pScalarizedIdx->getType(), i); - auto Val = IRB.CreateExtractElement(pStoreVal, VectorIdx); -diff --git a/IGC/Compiler/CISACodeGen/MemOpt.cpp b/IGC/Compiler/CISACodeGen/MemOpt.cpp -index e4279cca..aa788b11 100644 ---- a/IGC/Compiler/CISACodeGen/MemOpt.cpp -+++ b/IGC/Compiler/CISACodeGen/MemOpt.cpp -@@ -125,7 +125,7 @@ namespace { - MemRefListTy& MemRefs, TrivialMemRefListTy& ToOpt); - - unsigned getNumElements(Type* Ty) const { -- return Ty->isVectorTy() ? (unsigned)cast<VectorType>(Ty)->getNumElements() : 1; -+ return Ty->isVectorTy() ? (unsigned)cast<IGCLLVM::FixedVectorType>(Ty)->getNumElements() : 1; - } - - MemoryLocation getLocation(Instruction* I) const { -@@ -797,7 +797,7 @@ bool MemOpt::mergeLoad(LoadInst* LeadingLoad, - Pos = unsigned((std::get<1>(I) - FirstOffset) / LdScalarSize); - - if (Ty->isVectorTy()) { -- if (Pos + cast<VectorType>(Ty)->getNumElements() > NumElts) { -+ if (Pos + cast<IGCLLVM::FixedVectorType>(Ty)->getNumElements() > NumElts) { - // This implies we're trying to extract an element from our new load - // with an index > the size of the new load. If this happens, - // we'll generate correct code if it does since we don't remove the -@@ -805,7 +805,7 @@ bool MemOpt::mergeLoad(LoadInst* LeadingLoad, - continue; - } - Value* Val = UndefValue::get(Ty); -- for (unsigned i = 0, e = (unsigned)cast<VectorType>(Ty)->getNumElements(); i != e; ++i) { -+ for (unsigned i = 0, e = (unsigned)cast<IGCLLVM::FixedVectorType>(Ty)->getNumElements(); i != e; ++i) { - Value* Ex = Builder.CreateExtractElement(NewLoad, Builder.getInt32(Pos + i)); - Ex = createBitOrPointerCast(Ex, ScalarTy, Builder); - Val = Builder.CreateInsertElement(Val, Ex, Builder.getInt32(i)); -@@ -1083,7 +1083,7 @@ bool MemOpt::mergeStore(StoreInst* LeadingStore, - IGC_ASSERT(hasSameSize(ScalarTy, LeadingStoreScalarType)); - - if (Ty->isVectorTy()) { -- for (unsigned i = 0, e = (unsigned)cast<VectorType>(Ty)->getNumElements(); i != e; ++i) { -+ for (unsigned i = 0, e = (unsigned)cast<IGCLLVM::FixedVectorType>(Ty)->getNumElements(); i != e; ++i) { - Value* Ex = Builder.CreateExtractElement(Val, Builder.getInt32(i)); - Ex = createBitOrPointerCast(Ex, LeadingStoreScalarType, Builder); - NewStoreVal = Builder.CreateInsertElement(NewStoreVal, Ex, -diff --git a/IGC/Compiler/CISACodeGen/MemOpt2.cpp b/IGC/Compiler/CISACodeGen/MemOpt2.cpp -index 4b2628d2..20115c73 100644 ---- a/IGC/Compiler/CISACodeGen/MemOpt2.cpp -+++ b/IGC/Compiler/CISACodeGen/MemOpt2.cpp -@@ -32,6 +32,7 @@ IN THE SOFTWARE. - #include <llvm/Support/Debug.h> - #include <llvm/Support/raw_ostream.h> - #include <llvm/Transforms/Utils/Local.h> -+#include <llvmWrapper/IR/DerivedTypes.h> - #include "common/LLVMWarningsPop.hpp" - #include "GenISAIntrinsics/GenIntrinsics.h" - -@@ -124,7 +125,7 @@ namespace { - if (!Ty->isSingleValueType()) - return UINT_MAX; - // Simply return 1 so far for scalar types. -- VectorType* VecTy = dyn_cast<VectorType>(Ty); -+ IGCLLVM::FixedVectorType* VecTy = dyn_cast<IGCLLVM::FixedVectorType>(Ty); - if (!VecTy) - return 1; - // Check how that vector is used. -@@ -153,7 +154,7 @@ namespace { - return UINT_MAX; - unsigned EltByte = (Ty->getScalarSizeInBits() + 7) / 8; - // Simply return 1 so far for scalar types. -- VectorType* VecTy = dyn_cast<VectorType>(Ty); -+ IGCLLVM::FixedVectorType* VecTy = dyn_cast<IGCLLVM::FixedVectorType>(Ty); - if (!VecTy) - return EltByte; - // Check how that vector is used. -diff --git a/IGC/Compiler/CISACodeGen/OpenCLKernelCodeGen.cpp b/IGC/Compiler/CISACodeGen/OpenCLKernelCodeGen.cpp -index 5baff64f..032d127a 100644 ---- a/IGC/Compiler/CISACodeGen/OpenCLKernelCodeGen.cpp -+++ b/IGC/Compiler/CISACodeGen/OpenCLKernelCodeGen.cpp -@@ -410,7 +410,7 @@ namespace IGC - unsigned int numElements = 1; - if (baseType->isVectorTy()) - { -- numElements = (unsigned)cast<VectorType>(baseType)->getNumElements(); -+ numElements = (unsigned)cast<IGCLLVM::FixedVectorType>(baseType)->getNumElements(); - baseType = cast<VectorType>(baseType)->getElementType(); - } - -diff --git a/IGC/Compiler/CISACodeGen/PatternMatchPass.cpp b/IGC/Compiler/CISACodeGen/PatternMatchPass.cpp -index 356f1141..52396304 100644 ---- a/IGC/Compiler/CISACodeGen/PatternMatchPass.cpp -+++ b/IGC/Compiler/CISACodeGen/PatternMatchPass.cpp -@@ -2916,8 +2916,8 @@ namespace IGC - llvm::Type* srcTy = bTInst->getOperand(0)->getType(); - llvm::Type* dstTy = bTInst->getType(); - -- srcNElts = (srcTy->isVectorTy()) ? (uint32_t)cast<VectorType>(srcTy)->getNumElements() : 1; -- dstNElts = (dstTy->isVectorTy()) ? (uint32_t)cast<VectorType>(dstTy)->getNumElements() : 1; -+ srcNElts = (srcTy->isVectorTy()) ? (uint32_t)cast<IGCLLVM::FixedVectorType>(srcTy)->getNumElements() : 1; -+ dstNElts = (dstTy->isVectorTy()) ? (uint32_t)cast<IGCLLVM::FixedVectorType>(dstTy)->getNumElements() : 1; - - if (srcNElts < dstNElts && srcTy->getScalarSizeInBits() < 64) - { -diff --git a/IGC/Compiler/CISACodeGen/PushAnalysis.cpp b/IGC/Compiler/CISACodeGen/PushAnalysis.cpp -index ba8c2770..8da82db0 100644 ---- a/IGC/Compiler/CISACodeGen/PushAnalysis.cpp -+++ b/IGC/Compiler/CISACodeGen/PushAnalysis.cpp -@@ -899,7 +899,7 @@ namespace IGC - - if (pTypeToPush->isVectorTy()) - { -- num_elms = (unsigned)cast<VectorType>(pTypeToPush)->getNumElements(); -+ num_elms = (unsigned)cast<IGCLLVM::FixedVectorType>(pTypeToPush)->getNumElements(); - pTypeToPush = cast<VectorType>(pTypeToPush)->getElementType(); - llvm::Type* pVecTy = IGCLLVM::FixedVectorType::get(pTypeToPush, num_elms); - pReplacedInst = llvm::UndefValue::get(pVecTy); -@@ -1102,7 +1102,7 @@ namespace IGC - } - - unsigned num_elms = -- inst->getType()->isVectorTy() ? (unsigned)cast<VectorType>(inst->getType())->getNumElements() : 1; -+ inst->getType()->isVectorTy() ? (unsigned)cast<IGCLLVM::FixedVectorType>(inst->getType())->getNumElements() : 1; - llvm::Type* pTypeToPush = inst->getType(); - llvm::Value* replaceVector = nullptr; - unsigned int numberChannelReplaced = 0; -diff --git a/IGC/Compiler/CISACodeGen/RegisterEstimator.cpp b/IGC/Compiler/CISACodeGen/RegisterEstimator.cpp -index 95b937e7..1c6544a4 100644 ---- a/IGC/Compiler/CISACodeGen/RegisterEstimator.cpp -+++ b/IGC/Compiler/CISACodeGen/RegisterEstimator.cpp -@@ -41,6 +41,7 @@ IN THE SOFTWARE. - #include <llvm/IR/IRBuilder.h> - #include <llvm/IR/InstIterator.h> - #include <llvm/Support/MathExtras.h> -+#include <llvmWrapper/IR/DerivedTypes.h> - #include "common/LLVMWarningsPop.hpp" - #include "Probe/Assertion.h" - -@@ -146,7 +147,7 @@ RegUse RegisterEstimator::estimateNumOfRegs(Value* V) const - Type* Ty = V->getType(); - if (!Ty->isVoidTy()) - { -- VectorType* VTy = dyn_cast<VectorType>(Ty); -+ IGCLLVM::FixedVectorType* VTy = dyn_cast<IGCLLVM::FixedVectorType>(Ty); - Type* eltTy = VTy ? VTy->getElementType() : Ty; - uint32_t nelts = VTy ? int_cast<uint32_t>(VTy->getNumElements()) : 1; - uint32_t eltBits = (uint32_t)m_DL->getTypeSizeInBits(eltTy); -diff --git a/IGC/Compiler/CISACodeGen/RegisterPressureEstimate.hpp b/IGC/Compiler/CISACodeGen/RegisterPressureEstimate.hpp -index beac138d..ceeafa48 100644 ---- a/IGC/Compiler/CISACodeGen/RegisterPressureEstimate.hpp -+++ b/IGC/Compiler/CISACodeGen/RegisterPressureEstimate.hpp -@@ -42,6 +42,7 @@ See LRCENSE.TXT for details. - #include "llvm/Pass.h" - #include <llvm/IR/InstVisitor.h> - #include "llvm/Analysis/LoopInfo.h" -+#include <llvmWrapper/IR/DerivedTypes.h> - #include "common/LLVMWarningsPop.hpp" - #include "Compiler/IGCPassSupport.h" - #include "Compiler/CISACodeGen/WIAnalysis.hpp" -@@ -240,7 +241,7 @@ namespace IGC - auto Ty = V->getType(); - if (Ty->isVoidTy()) - return 0; -- auto VTy = llvm::dyn_cast<llvm::VectorType>(Ty); -+ auto VTy = llvm::dyn_cast<IGCLLVM::FixedVectorType>(Ty); - auto eltTy = VTy ? VTy->getElementType() : Ty; - uint32_t nelts = VTy ? int_cast<uint32_t>(VTy->getNumElements()) : 1; - uint32_t eltBits = (uint32_t)m_DL->getTypeSizeInBits(eltTy); -diff --git a/IGC/Compiler/CISACodeGen/SLMConstProp.cpp b/IGC/Compiler/CISACodeGen/SLMConstProp.cpp -index 962083f4..26a7b8be 100644 ---- a/IGC/Compiler/CISACodeGen/SLMConstProp.cpp -+++ b/IGC/Compiler/CISACodeGen/SLMConstProp.cpp -@@ -39,6 +39,7 @@ IN THE SOFTWARE. - #include <llvm/IR/InstIterator.h> - #include <llvm/Support/Debug.h> - #include <llvm/Support/MathExtras.h> -+#include "llvmWrapper/IR/DerivedTypes.h" - #include "common/LLVMWarningsPop.hpp" - #include <vector> - #include "Probe/Assertion.h" -@@ -629,7 +630,7 @@ bool SLMConstProp::isEqual(Constant* C0, Constant* C1) - - bool SLMConstProp::isFloatType(Type* Ty) - { -- if (VectorType * vTy = dyn_cast<VectorType>(Ty)) -+ if (IGCLLVM::FixedVectorType * vTy = dyn_cast<IGCLLVM::FixedVectorType>(Ty)) - { - if (vTy->getNumElements() > 1) - { -diff --git a/IGC/Compiler/CISACodeGen/ScalarizerCodeGen.cpp b/IGC/Compiler/CISACodeGen/ScalarizerCodeGen.cpp -index ce816416..32ab79c1 100644 ---- a/IGC/Compiler/CISACodeGen/ScalarizerCodeGen.cpp -+++ b/IGC/Compiler/CISACodeGen/ScalarizerCodeGen.cpp -@@ -66,7 +66,7 @@ void ScalarizerCodeGen::visitBinaryOperator(llvm::BinaryOperator& I) - { - bool isNewTypeVector = false; - -- VectorType* instType = cast<VectorType>(I.getType()); -+ IGCLLVM::FixedVectorType* instType = cast<IGCLLVM::FixedVectorType>(I.getType()); - unsigned numElements = int_cast<unsigned>(instType->getNumElements()); - unsigned scalarSize = instType->getScalarSizeInBits(); - unsigned newScalarBits = numElements * scalarSize; -@@ -111,7 +111,7 @@ void ScalarizerCodeGen::visitBinaryOperator(llvm::BinaryOperator& I) - } - else - { -- VectorType* newVecType = cast<VectorType>(newType); -+ IGCLLVM::FixedVectorType* newVecType = cast<IGCLLVM::FixedVectorType>(newType); - unsigned newVecTypeNumEle = int_cast<unsigned>(newVecType->getNumElements()); - Value* ieLogicOp = UndefValue::get(newType); - for (unsigned i = 0; i < newVecTypeNumEle; i++) -diff --git a/IGC/Compiler/CISACodeGen/ShaderCodeGen.cpp b/IGC/Compiler/CISACodeGen/ShaderCodeGen.cpp -index d7baf806..e6fbdf23 100644 ---- a/IGC/Compiler/CISACodeGen/ShaderCodeGen.cpp -+++ b/IGC/Compiler/CISACodeGen/ShaderCodeGen.cpp -@@ -589,12 +589,14 @@ static void AddLegalizationPasses(CodeGenContext& ctx, IGCPassManager& mpm, PSSi - { - mpm.add(createPruneUnusedArgumentsPass()); - -+#if LLVM_VERSION_MAJOR < 12 - if (IGC_GET_FLAG_VALUE(FunctionControl) == FLAG_FCALL_DEFAULT) - { - // Don't run IPConstantProp when debugging function calls, to avoid folding function arg/ret constants - mpm.add(createIPConstantPropagationPass()); - } - mpm.add(createConstantPropagationPass()); -+#endif - mpm.add(createDeadCodeEliminationPass()); - mpm.add(createCFGSimplificationPass()); - } -@@ -1475,8 +1477,10 @@ void OptimizeIR(CodeGenContext* const pContext) - // possible which potentially allows late stage code sinking of - // those calls by the instruction combiner. - mpm.add(createPostOrderFunctionAttrsLegacyPass()); -+#if LLVM_VERSION_MAJOR < 12 - mpm.add(createConstantPropagationPass()); - mpm.add(createIPConstantPropagationPass()); -+#endif - } - - // enable this only when Pooled EU is not supported -diff --git a/IGC/Compiler/CISACodeGen/Simd32Profitability.cpp b/IGC/Compiler/CISACodeGen/Simd32Profitability.cpp -index 87ba2751..c1f4a419 100644 ---- a/IGC/Compiler/CISACodeGen/Simd32Profitability.cpp -+++ b/IGC/Compiler/CISACodeGen/Simd32Profitability.cpp -@@ -30,6 +30,7 @@ IN THE SOFTWARE. - #include <llvmWrapper/IR/DerivedTypes.h> - #include <llvm/IR/InstIterator.h> - #include <llvm/IR/Operator.h> -+#include <llvmWrapper/IR/DerivedTypes.h> - #include "common/LLVMWarningsPop.hpp" - #include "GenISAIntrinsics/GenIntrinsics.h" - #include "GenISAIntrinsics/GenIntrinsicInst.h" -@@ -577,7 +578,7 @@ static bool isPayloadHeader(Value* V) { - Argument* Arg = dyn_cast<Argument>(V); - if (!Arg || !Arg->hasName()) - return false; -- VectorType* VTy = dyn_cast<VectorType>(Arg->getType()); -+ IGCLLVM::FixedVectorType* VTy = dyn_cast<IGCLLVM::FixedVectorType>(Arg->getType()); - if (!VTy || VTy->getNumElements() != 8 || - !VTy->getElementType()->isIntegerTy(32)) - return false; -@@ -588,7 +589,7 @@ static bool isR0(Value* V) { - Argument* Arg = dyn_cast<Argument>(V); - if (!Arg || !Arg->hasName()) - return false; -- VectorType* VTy = dyn_cast<VectorType>(Arg->getType()); -+ IGCLLVM::FixedVectorType* VTy = dyn_cast<IGCLLVM::FixedVectorType>(Arg->getType()); - if (!VTy || VTy->getNumElements() != 8 || - !VTy->getElementType()->isIntegerTy(32)) - return false; -@@ -599,7 +600,7 @@ static bool isEnqueuedLocalSize(Value* V) { - Argument* Arg = dyn_cast<Argument>(V); - if (!Arg || !Arg->hasName()) - return false; -- VectorType* VTy = dyn_cast<VectorType>(Arg->getType()); -+ IGCLLVM::FixedVectorType* VTy = dyn_cast<IGCLLVM::FixedVectorType>(Arg->getType()); - if (!VTy || VTy->getNumElements() != 3 || - !VTy->getElementType()->isIntegerTy(32)) - return false; -@@ -994,7 +995,11 @@ static bool hasLongStridedLdStInLoop(Function* F, LoopInfo* LI, WIAnalysis* WI) - // Collect innermost simple loop. - for (auto I = LI->begin(), E = LI->end(); I != E; ++I) { - auto L = *I; -+#if LLVM_VERSION_MAJOR >= 12 -+ if (!L->isInnermost()) -+#else - if (!L->empty()) -+#endif - continue; - if (L->getNumBlocks() != 2) - continue; -diff --git a/IGC/Compiler/CISACodeGen/SimplifyConstant.cpp b/IGC/Compiler/CISACodeGen/SimplifyConstant.cpp -index 73f472b5..0069c4b1 100644 ---- a/IGC/Compiler/CISACodeGen/SimplifyConstant.cpp -+++ b/IGC/Compiler/CISACodeGen/SimplifyConstant.cpp -@@ -352,7 +352,7 @@ static unsigned getLegalVectorSize(unsigned N) { - // Check vector size. We may demote the data type if all values can fit into - // smaller data type. - // --static bool checkSize(GlobalVariable* GV, VectorType*& DataType, -+static bool checkSize(GlobalVariable* GV, IGCLLVM::FixedVectorType*& DataType, - bool& IsSigned) { - Constant* Init = GV->getInitializer(); - IGC_ASSERT(isa<ArrayType>(Init->getType())); -@@ -360,7 +360,7 @@ static bool checkSize(GlobalVariable* GV, VectorType*& DataType, - unsigned N = (unsigned)ArrayTy->getArrayNumElements(); - Type* BaseTy = ArrayTy->getArrayElementType(); - unsigned VectorSize = 1; -- if (auto VT = dyn_cast<VectorType>(BaseTy)) { -+ if (auto VT = dyn_cast<IGCLLVM::FixedVectorType>(BaseTy)) { - BaseTy = VT->getElementType(); - VectorSize = int_cast<unsigned>(VT->getNumElements()); - N *= VectorSize; -@@ -483,7 +483,7 @@ static Constant* getConstantVal(Type* VEltTy, Constant* V, bool IsSigned) { - return ConstantInt::get(VEltTy, IVal, IsSigned); - } - --static void promote(GlobalVariable* GV, VectorType* AllocaType, bool IsSigned, -+static void promote(GlobalVariable* GV, IGCLLVM::FixedVectorType* AllocaType, bool IsSigned, - Function* F) { - // Build the constant vector from constant array. - unsigned VS = int_cast<unsigned>(AllocaType->getNumElements()); -@@ -507,7 +507,7 @@ static void promote(GlobalVariable* GV, VectorType* AllocaType, bool IsSigned, - Constant* const Elt = CA->getAggregateElement(i); - IGC_ASSERT_MESSAGE(nullptr != Elt, "Null AggregateElement"); - if (auto EltTy = dyn_cast<VectorType>(Elt->getType())) { -- unsigned VectorSize = (unsigned)cast<VectorType>(EltTy)->getNumElements(); -+ unsigned VectorSize = (unsigned)cast<IGCLLVM::FixedVectorType>(EltTy)->getNumElements(); - for (unsigned j = 0; j < VectorSize; ++j) { - Constant* V = Elt->getAggregateElement(j); - Vals[i * VectorSize + j] = getConstantVal(VEltTy, V, IsSigned); -@@ -549,7 +549,7 @@ static void promote(GlobalVariable* GV, VectorType* AllocaType, bool IsSigned, - unsigned N = 1; - Value* Offset = Index; - if (Ty->isVectorTy()) { -- N = (unsigned)cast<VectorType>(Ty)->getNumElements(); -+ N = (unsigned)cast<IGCLLVM::FixedVectorType>(Ty)->getNumElements(); - Offset = Builder.CreateMul(Offset, ConstantInt::get(Offset->getType(), N)); - } - Value* Val = extractNElts(N, VectorData, Offset, Builder); -@@ -684,7 +684,7 @@ bool PromoteConstant::runOnFunction(Function& F) { - - // If possible demote the data into smaller type. Uses of value will be - // promoted back with ZExt or SExt. -- VectorType* AllocaType = nullptr; -+ IGCLLVM::FixedVectorType* AllocaType = nullptr; - bool IsSigned = false; - if (!checkSize(GV, AllocaType, IsSigned)) - continue; -diff --git a/IGC/Compiler/CISACodeGen/TypeDemote.cpp b/IGC/Compiler/CISACodeGen/TypeDemote.cpp -index eb7a7b64..b04a0f6e 100644 ---- a/IGC/Compiler/CISACodeGen/TypeDemote.cpp -+++ b/IGC/Compiler/CISACodeGen/TypeDemote.cpp -@@ -32,6 +32,7 @@ IN THE SOFTWARE. - #include <llvm/ADT/PostOrderIterator.h> - #include <llvm/IR/IRBuilder.h> - #include <llvm/Pass.h> -+#include <llvmWrapper/IR/DerivedTypes.h> - #include "common/LLVMWarningsPop.hpp" - #include "GenISAIntrinsics/GenIntrinsics.h" - #include "Probe/Assertion.h" -@@ -288,7 +289,7 @@ bool TypeDemote::demoteOnBasicBlock(BasicBlock* BB) const { - CastInst* CI = dyn_cast<CastInst>(Index); - if (CI && (CI->getOpcode() == Instruction::ZExt || - CI->getOpcode() == Instruction::SExt)) { -- unsigned VS = (unsigned)cast<VectorType>(EEI->getVectorOperandType())->getNumElements(); -+ unsigned VS = (unsigned)cast<IGCLLVM::FixedVectorType>(EEI->getVectorOperandType())->getNumElements(); - unsigned N = (unsigned int)CI->getSrcTy()->getPrimitiveSizeInBits(); - unsigned Bound = (N < 32) ? (1U << N) : UINT32_MAX; - if (VS <= Bound) { -diff --git a/IGC/Compiler/CISACodeGen/VariableReuseAnalysis.cpp b/IGC/Compiler/CISACodeGen/VariableReuseAnalysis.cpp -index 1cac7b12..12c4734e 100644 ---- a/IGC/Compiler/CISACodeGen/VariableReuseAnalysis.cpp -+++ b/IGC/Compiler/CISACodeGen/VariableReuseAnalysis.cpp -@@ -28,6 +28,7 @@ IN THE SOFTWARE. - #include "Compiler/CodeGenPublic.h" - #include "common/LLVMWarningsPush.hpp" - #include <llvm/Support/Debug.h> -+#include "llvmWrapper/IR/DerivedTypes.h" - #include "common/LLVMWarningsPop.hpp" - #include <algorithm> - #include "Probe/Assertion.h" -@@ -40,13 +41,13 @@ namespace - // If V is scalar, return 1. - // if V is vector, return the number of elements. - inline int getNumElts(Value* V) { -- VectorType* VTy = dyn_cast<VectorType>(V->getType()); -+ IGCLLVM::FixedVectorType* VTy = dyn_cast<IGCLLVM::FixedVectorType>(V->getType()); - return VTy ? (int)VTy->getNumElements() : 1; - } - - inline int getTypeSizeInBits(Type* Ty) { - int scalarBits = Ty->getScalarSizeInBits(); -- VectorType* VTy = dyn_cast<VectorType>(Ty); -+ IGCLLVM::FixedVectorType* VTy = dyn_cast<IGCLLVM::FixedVectorType>(Ty); - return scalarBits * (VTy ? (int)VTy->getNumElements() : 1); - } - } -diff --git a/IGC/Compiler/CISACodeGen/VariableReuseAnalysis.hpp b/IGC/Compiler/CISACodeGen/VariableReuseAnalysis.hpp -index 56dac17e..42081dca 100644 ---- a/IGC/Compiler/CISACodeGen/VariableReuseAnalysis.hpp -+++ b/IGC/Compiler/CISACodeGen/VariableReuseAnalysis.hpp -@@ -37,6 +37,7 @@ IN THE SOFTWARE. - #include <llvm/IR/InstIterator.h> - #include <llvm/IR/InstVisitor.h> - #include "llvm/Pass.h" -+#include "llvmWrapper/IR/DerivedTypes.h" - #include "llvm/Support/raw_ostream.h" - #include "common/LLVMWarningsPop.hpp" - #include "Compiler/CISACodeGen/RegisterEstimator.hpp" -@@ -79,7 +80,7 @@ namespace IGC { - SSubVecDesc(llvm::Value* V) - : Aliaser(V), BaseVector(V), StartElementOffset(0) - { -- llvm::VectorType* VTy = llvm::dyn_cast<llvm::VectorType>(V->getType()); -+ IGCLLVM::FixedVectorType* VTy = llvm::dyn_cast<IGCLLVM::FixedVectorType>(V->getType()); - NumElts = VTy ? (short)VTy->getNumElements() : 1; - } - -diff --git a/IGC/Compiler/CISACodeGen/VectorPreProcess.cpp b/IGC/Compiler/CISACodeGen/VectorPreProcess.cpp -index 4c2b0644..4cd6a2db 100644 ---- a/IGC/Compiler/CISACodeGen/VectorPreProcess.cpp -+++ b/IGC/Compiler/CISACodeGen/VectorPreProcess.cpp -@@ -500,7 +500,7 @@ void VectorPreProcess::replaceAllVectorUsesWithScalars(Instruction* VI, ValVecto - I = VI->getParent()->getFirstNonPHI(); - } - IRBuilder<> Builder(I); -- VectorType* VTy = cast<VectorType>(VI->getType()); -+ IGCLLVM::FixedVectorType* VTy = cast<IGCLLVM::FixedVectorType>(VI->getType()); - Value* newVec = UndefValue::get(VTy); - for (uint32_t i = 0, e = int_cast<uint32_t>(VTy->getNumElements()); i < e; ++i) - { -@@ -619,7 +619,7 @@ bool VectorPreProcess::splitStore( - { - Instruction* SI = ASI.getInst(); - Value* StoredVal = ASI.getValueOperand(); -- VectorType* VTy = cast<VectorType>(StoredVal->getType()); -+ IGCLLVM::FixedVectorType* VTy = cast<IGCLLVM::FixedVectorType>(StoredVal->getType()); - Type* ETy = VTy->getElementType(); - uint32_t nelts = int_cast<uint32_t>(VTy->getNumElements()); - -@@ -705,7 +705,7 @@ bool VectorPreProcess::splitStore( - { - Type* Ty1 = splitInfo[i].first; - uint32_t len1 = splitInfo[i].second; -- VectorType* VTy1 = dyn_cast<VectorType>(Ty1); -+ IGCLLVM::FixedVectorType* VTy1 = dyn_cast<IGCLLVM::FixedVectorType>(Ty1); - for (uint32_t j = 0; j < len1; ++j) - { - Value* subVec; -@@ -742,7 +742,7 @@ bool VectorPreProcess::splitStore( - { - Type* Ty1 = splitInfo[i].first; - uint32_t len1 = splitInfo[i].second; -- VectorType* VTy1 = dyn_cast<VectorType>(Ty1); -+ IGCLLVM::FixedVectorType* VTy1 = dyn_cast<IGCLLVM::FixedVectorType>(Ty1); - for (uint32_t j = 0; j < len1; ++j) - { - uint32_t vAlign = (uint32_t)MinAlign(Align, (uint32_t)eOffset * EBytes); -@@ -786,7 +786,7 @@ bool VectorPreProcess::splitLoad( - { - Instruction* LI = ALI.getInst(); - bool isLdRaw = isa<LdRawIntrinsic>(LI); -- VectorType* VTy = cast<VectorType>(LI->getType()); -+ IGCLLVM::FixedVectorType* VTy = cast<IGCLLVM::FixedVectorType>(LI->getType()); - Type* ETy = VTy->getElementType(); - uint32_t nelts = int_cast<uint32_t>(VTy->getNumElements()); - -@@ -848,7 +848,7 @@ bool VectorPreProcess::splitLoad( - { - Type* Ty1 = splitInfo[i].first; - uint32_t len1 = splitInfo[i].second; -- VectorType* VTy1 = dyn_cast<VectorType>(Ty1); -+ IGCLLVM::FixedVectorType* VTy1 = dyn_cast<IGCLLVM::FixedVectorType>(Ty1); - for (uint32_t j = 0; j < len1; ++j) - { - uint32_t vAlign = (uint32_t)MinAlign(Align, eOffset * EBytes); -@@ -916,7 +916,7 @@ bool VectorPreProcess::splitLoadStore( - Optional<AbstractStoreInst> ASI = AbstractStoreInst::get(Inst); - IGC_ASSERT_MESSAGE((ALI || ASI), "Inst should be either load or store"); - Type* Ty = ALI ? ALI->getInst()->getType() : ASI->getValueOperand()->getType(); -- VectorType* VTy = dyn_cast<VectorType>(Ty); -+ IGCLLVM::FixedVectorType* VTy = dyn_cast<IGCLLVM::FixedVectorType>(Ty); - if (!VTy) - { - return false; -@@ -1147,7 +1147,7 @@ void VectorPreProcess::getOrGenScalarValues( - { - availBeforeInst = nullptr; - -- VectorType* VTy = cast<VectorType>(VecVal->getType()); -+ IGCLLVM::FixedVectorType* VTy = cast<IGCLLVM::FixedVectorType>(VecVal->getType()); - if (!VTy) - { - scalars[0] = VecVal; -@@ -1298,7 +1298,7 @@ Instruction* VectorPreProcess::simplifyLoadStore(Instruction* Inst) - // - // TODO: further optimize this load into a message with channel masks - // for cases in which use indices are sparse like {0, 2}. -- unsigned N = (unsigned)cast<VectorType>(Inst->getType())->getNumElements(); -+ unsigned N = (unsigned)cast<IGCLLVM::FixedVectorType>(Inst->getType())->getNumElements(); - if (N == MaxIndex + 1) - return Inst; - -@@ -1355,7 +1355,7 @@ Instruction* VectorPreProcess::simplifyLoadStore(Instruction* Inst) - if (NBits < 32) - return Inst; - -- unsigned N = (unsigned)cast<VectorType>(Val->getType())->getNumElements(); -+ unsigned N = (unsigned)cast<IGCLLVM::FixedVectorType>(Val->getType())->getNumElements(); - if (auto CV = dyn_cast<ConstantVector>(Val)) - { - unsigned MaxIndex = 0; -@@ -1524,7 +1524,7 @@ bool VectorPreProcess::runOnFunction(Function& F) - for (uint32_t j = 0; j < svals.size(); ++j) - { - Type* Ty1 = svals[j]->getType(); -- VectorType* VTy1 = dyn_cast<VectorType>(Ty1); -+ IGCLLVM::FixedVectorType* VTy1 = dyn_cast<IGCLLVM::FixedVectorType>(Ty1); - if (VTy1) { - for (uint32_t k = 0; k < VTy1->getNumElements(); ++k) - { -@@ -1573,7 +1573,7 @@ bool VectorPreProcess::runOnFunction(Function& F) - // If this is a 3-element vector load, remove it - // from m_Vector3List as well. - if (isAbstractLoadInst(tInst) && tInst->getType()->isVectorTy() && -- cast<VectorType>(tInst->getType())->getNumElements() == 3) -+ cast<IGCLLVM::FixedVectorType>(tInst->getType())->getNumElements() == 3) - { - InstWorkVector::iterator - tI = m_Vector3List.begin(), -diff --git a/IGC/Compiler/CISACodeGen/VectorProcess.cpp b/IGC/Compiler/CISACodeGen/VectorProcess.cpp -index 9f39cdc0..cfe45321 100644 ---- a/IGC/Compiler/CISACodeGen/VectorProcess.cpp -+++ b/IGC/Compiler/CISACodeGen/VectorProcess.cpp -@@ -240,7 +240,7 @@ bool VectorProcess::reLayoutLoadStore(Instruction* Inst) - IGC_ASSERT(nullptr != Ptr); - IGC_ASSERT(nullptr != Ty); - -- VectorType* const VTy = dyn_cast<VectorType>(Ty); -+ IGCLLVM::FixedVectorType* const VTy = dyn_cast<IGCLLVM::FixedVectorType>(Ty); - - // Treat a scalar as 1-element vector - uint32_t nelts = VTy ? int_cast<uint32_t>(VTy->getNumElements()) : 1; -@@ -654,7 +654,7 @@ void VectorMessage::getInfo(Type* Ty, uint32_t Align, bool useA32, - VectorType* VTy = dyn_cast<VectorType>(Ty); - Type* eTy = VTy ? cast<VectorType>(VTy)->getElementType() : Ty; - unsigned eltSize = m_emitter->GetScalarTypeSizeInRegister(eTy); -- unsigned nElts = VTy ? (unsigned)cast<VectorType>(VTy)->getNumElements() : 1; -+ unsigned nElts = VTy ? (unsigned)cast<IGCLLVM::FixedVectorType>(VTy)->getNumElements() : 1; - // total bytes - const unsigned TBytes = nElts * eltSize; - -diff --git a/IGC/Compiler/CISACodeGen/helper.cpp b/IGC/Compiler/CISACodeGen/helper.cpp -index 784affaa..e5507283 100644 ---- a/IGC/Compiler/CISACodeGen/helper.cpp -+++ b/IGC/Compiler/CISACodeGen/helper.cpp -@@ -1634,7 +1634,7 @@ namespace IGC - { - instList[i] = builder.CreateExtractElement(val, static_cast<uint64_t>(0)); - size_t iOld = i; -- for (unsigned j = 1; j < cast<VectorType>(val->getType())->getNumElements(); j++) -+ for (unsigned j = 1; j < cast<IGCLLVM::FixedVectorType>(val->getType())->getNumElements(); j++) - { - instList.insert(instList.begin()+ iOld +j, builder.CreateExtractElement(val, j)); - i++; -@@ -1667,7 +1667,7 @@ namespace IGC - } - break; - case IGCLLVM::VectorTyID: -- num = (unsigned)cast<VectorType>(type)->getNumElements(); -+ num = (unsigned)cast<IGCLLVM::FixedVectorType>(type)->getNumElements(); - for (unsigned i = 0; i < num; i++) - { - ScalarizeAggregateMembers(builder, builder.CreateExtractElement(val, i), instList); -@@ -1707,7 +1707,7 @@ namespace IGC - } - break; - case IGCLLVM::VectorTyID: -- num = (unsigned)cast<VectorType>(type)->getNumElements(); -+ num = (unsigned)cast<IGCLLVM::FixedVectorType>(type)->getNumElements(); - for (unsigned i = 0; i < num; i++) - { - indices.push_back(builder.getInt32(i)); -@@ -1972,8 +1972,8 @@ namespace IGC - return false; - } - -- VectorType* dVTy = dyn_cast<VectorType>(dTy); -- VectorType* sVTy = dyn_cast<VectorType>(sTy); -+ IGCLLVM::FixedVectorType* dVTy = dyn_cast<IGCLLVM::FixedVectorType>(dTy); -+ IGCLLVM::FixedVectorType* sVTy = dyn_cast<IGCLLVM::FixedVectorType>(sTy); - int d_nelts = dVTy ? (int)dVTy->getNumElements() : 1; - int s_nelts = sVTy ? (int)sVTy->getNumElements() : 1; - if (d_nelts != s_nelts) { -diff --git a/IGC/Compiler/ConvertMSAAPayloadTo16Bit.cpp b/IGC/Compiler/ConvertMSAAPayloadTo16Bit.cpp -index cd7fc66e..adf992cb 100644 ---- a/IGC/Compiler/ConvertMSAAPayloadTo16Bit.cpp -+++ b/IGC/Compiler/ConvertMSAAPayloadTo16Bit.cpp -@@ -153,15 +153,24 @@ void ConvertMSAAPayloadTo16Bit::visitCallInst(CallInst& I) - // In OGL there are uses of ldmcs other then ldms, using vec4float type. - // Fix them to use newly created 16bit ldmcs. - if (ldmcs->getType()->isVectorTy() && -+#if LLVM_VERSION_MAJOR >= 12 -+ ldmcs->getType()->getScalarType()->isFloatTy()) -+#else - ldmcs->getType()->getVectorElementType()->isFloatTy()) -+#endif - { - m_builder->SetInsertPoint(ldmcs); - -+#if LLVM_VERSION_MAJOR >= 12 -+ uint ldmcsNumOfElements = cast<IGCLLVM::FixedVectorType>(ldmcs->getType())->getNumElements(); -+ uint new_mcs_callNumOfElements = cast<IGCLLVM::FixedVectorType>(new_mcs_call->getType())->getNumElements(); -+#else - uint ldmcsNumOfElements = ldmcs->getType()->getVectorNumElements(); - uint new_mcs_callNumOfElements = new_mcs_call->getType()->getVectorNumElements(); -+#endif - - // vec of 16bit ints to vec of 32bit ints -- Type* new_mcs_callVecType = VectorType::get(m_builder->getInt32Ty(), new_mcs_callNumOfElements); -+ Type* new_mcs_callVecType = IGCLLVM::FixedVectorType::get(m_builder->getInt32Ty(), new_mcs_callNumOfElements); - Value* ldmcsExtendedToInt32 = m_builder->CreateSExt(new_mcs_call, new_mcs_callVecType); - - // if new ldmcs has fewer elements than ldmcs, extend vector -@@ -175,7 +184,7 @@ void ConvertMSAAPayloadTo16Bit::visitCallInst(CallInst& I) - } - auto* pMask = ConstantDataVector::get(I.getContext(), maskVals); - -- newLdmcsSizedVector = m_builder->CreateShuffleVector(ldmcsExtendedToInt32, UndefValue::get(VectorType::get(m_builder->getInt32Ty(), ldmcsNumOfElements)), pMask); -+ newLdmcsSizedVector = m_builder->CreateShuffleVector(ldmcsExtendedToInt32, UndefValue::get(IGCLLVM::FixedVectorType::get(m_builder->getInt32Ty(), ldmcsNumOfElements)), pMask); - } - else - { -@@ -183,7 +192,7 @@ void ConvertMSAAPayloadTo16Bit::visitCallInst(CallInst& I) - } - IGC_ASSERT(newLdmcsSizedVector); - -- Type* ldmcsFloatVecType = VectorType::get(m_builder->getFloatTy(), ldmcsNumOfElements); -+ Type* ldmcsFloatVecType = IGCLLVM::FixedVectorType::get(m_builder->getFloatTy(), ldmcsNumOfElements); - Value* ldmcsBitcastedToFloat = m_builder->CreateBitCast(ldmcsExtendedToInt32, ldmcsFloatVecType); - ldmcs->replaceAllUsesWith(ldmcsBitcastedToFloat); - } -diff --git a/IGC/Compiler/CustomSafeOptPass.cpp b/IGC/Compiler/CustomSafeOptPass.cpp -index 707bbfbe..72859a91 100644 ---- a/IGC/Compiler/CustomSafeOptPass.cpp -+++ b/IGC/Compiler/CustomSafeOptPass.cpp -@@ -1372,7 +1372,7 @@ void IGC::CustomSafeOptPass::visitSampleBptr(llvm::SampleIntrinsic* sampleInst) - bool CustomSafeOptPass::isIdentityMatrix(ExtractElementInst& I) - { - bool found = false; -- auto extractType = cast<VectorType>(I.getVectorOperandType()); -+ auto extractType = cast<IGCLLVM::FixedVectorType>(I.getVectorOperandType()); - auto extractTypeVecSize = (uint32_t)extractType->getNumElements(); - if (extractTypeVecSize == 20 || - extractTypeVecSize == 16) -@@ -1601,7 +1601,7 @@ void CustomSafeOptPass::visitExtractElementInst(ExtractElementInst& I) - int elOffset = (int)(bitShift / eltSize); - elOffset = rightShift ? elOffset : -elOffset; - unsigned int newIndex = (unsigned int)((int)cstIndex->getZExtValue() + elOffset); -- if (newIndex < cast<VectorType>(vecType)->getNumElements()) -+ if (newIndex < cast<IGCLLVM::FixedVectorType>(vecType)->getNumElements()) - { - IRBuilder<> builder(&I); - Value* newBitCast = builder.CreateBitCast(binOp->getOperand(0), vecType); -@@ -2001,7 +2001,7 @@ void GenSpecificPattern::createBitcastExtractInsertPattern(BinaryOperator& I, Va - else if (auto IEIInst = dyn_cast<InsertElementInst>(Op)) - { - auto opType = IEIInst->getType(); -- if (opType->isVectorTy() && cast<VectorType>(opType)->getElementType()->isIntegerTy(32) && cast<VectorType>(opType)->getNumElements() == 2) -+ if (opType->isVectorTy() && cast<VectorType>(opType)->getElementType()->isIntegerTy(32) && cast<IGCLLVM::FixedVectorType>(opType)->getNumElements() == 2) - { - elem = IEIInst->getOperand(1); - } -@@ -2064,7 +2064,7 @@ void GenSpecificPattern::visitBinaryOperator(BinaryOperator& I) - else if (match(&I, pattern2) && AndOp2->getType()->isIntegerTy(64)) - { - ConstantVector* cVec = dyn_cast<ConstantVector>(VecOp); -- VectorType* vector_type = dyn_cast<VectorType>(VecOp->getType()); -+ IGCLLVM::FixedVectorType* vector_type = dyn_cast<IGCLLVM::FixedVectorType>(VecOp->getType()); - if (cVec && vector_type && - isa<ConstantInt>(cVec->getOperand(0)) && - cast<ConstantInt>(cVec->getOperand(0))->isZero() && -@@ -2210,7 +2210,7 @@ void GenSpecificPattern::visitBinaryOperator(BinaryOperator& I) - BitCastInst* opBC = cast<BitCastInst>(op); - - auto opType = opBC->getType(); -- if (!(opType->isVectorTy() && cast<VectorType>(opType)->getElementType()->isIntegerTy(32) && cast<VectorType>(opType)->getNumElements() == 2)) -+ if (!(opType->isVectorTy() && cast<VectorType>(opType)->getElementType()->isIntegerTy(32) && cast<IGCLLVM::FixedVectorType>(opType)->getNumElements() == 2)) - return nullptr; - - if (opBC->getSrcTy()->isDoubleTy()) -@@ -2630,8 +2630,8 @@ void GenSpecificPattern::visitBitCastInst(BitCastInst& I) - if (zExtInst->getOperand(0)->getType()->isIntegerTy(32) && - isa<InsertElementInst>(bitCastInst->getOperand(0)) && - bitCastInst->getOperand(0)->getType()->isVectorTy() && -- cast<VectorType>(bitCastInst->getOperand(0)->getType())->getElementType()->isIntegerTy(32) && -- cast<VectorType>(bitCastInst->getOperand(0)->getType())->getNumElements() == 2) -+ cast<IGCLLVM::FixedVectorType>(bitCastInst->getOperand(0)->getType())->getElementType()->isIntegerTy(32) && -+ cast<IGCLLVM::FixedVectorType>(bitCastInst->getOperand(0)->getType())->getNumElements() == 2) - { - InsertElementInst* insertElementInst = cast<InsertElementInst>(bitCastInst->getOperand(0)); - -@@ -2731,7 +2731,7 @@ void GenSpecificPattern::visitFNeg(llvm::UnaryOperator& I) - } - else - { -- uint32_t vectorSize = cast<VectorType>(I.getType())->getNumElements(); -+ uint32_t vectorSize = cast<IGCLLVM::FixedVectorType>(I.getType())->getNumElements(); - fsub = llvm::UndefValue::get(I.getType()); - - for (uint32_t i = 0; i < vectorSize; ++i) -@@ -2845,7 +2845,7 @@ Constant* IGCConstProp::replaceShaderConstant(LoadInst* inst) - if (inst->getType()->isVectorTy()) - { - Type* srcEltTy = cast<VectorType>(inst->getType())->getElementType(); -- uint32_t srcNElts = (uint32_t)cast<VectorType>(inst->getType())->getNumElements(); -+ uint32_t srcNElts = (uint32_t)cast<IGCLLVM::FixedVectorType>(inst->getType())->getNumElements(); - uint32_t eltSize_in_bytes = (unsigned int)srcEltTy->getPrimitiveSizeInBits() / 8; - IRBuilder<> builder(inst); - Value* vectorValue = UndefValue::get(inst->getType()); -@@ -3122,7 +3122,7 @@ Constant* IGCConstProp::ConstantFoldCmpInst(CmpInst* CI) - { - bool AllTrue = true, AllFalse = true; - auto VecOpnd = cast<Constant>(EEI->getVectorOperand()); -- unsigned N = (unsigned)cast<VectorType>(VecOpnd->getType())->getNumElements(); -+ unsigned N = (unsigned)cast<IGCLLVM::FixedVectorType>(VecOpnd->getType())->getNumElements(); - for (unsigned i = 0; i < N; ++i) - { - Constant* const Opnd = VecOpnd->getAggregateElement(i); -@@ -3918,8 +3918,8 @@ namespace IGC - BitCastInst* BC = dyn_cast<BitCastInst>(&*BI++); - if (!BC) continue; - // Skip non-element-wise bitcast. -- VectorType* DstVTy = dyn_cast<VectorType>(BC->getType()); -- VectorType* SrcVTy = dyn_cast<VectorType>(BC->getOperand(0)->getType()); -+ IGCLLVM::FixedVectorType* DstVTy = dyn_cast<IGCLLVM::FixedVectorType>(BC->getType()); -+ IGCLLVM::FixedVectorType* SrcVTy = dyn_cast<IGCLLVM::FixedVectorType>(BC->getOperand(0)->getType()); - if (!DstVTy || !SrcVTy || DstVTy->getNumElements() != SrcVTy->getNumElements()) - continue; - // Skip if it's not used only all extractelement. -@@ -4689,7 +4689,7 @@ void SplitIndirectEEtoSel::visitExtractElementInst(llvm::ExtractElementInst& I) - { - using namespace llvm::PatternMatch; - -- VectorType* vecTy = I.getVectorOperandType(); -+ IGCLLVM::FixedVectorType* vecTy = dyn_cast<IGCLLVM::FixedVectorType>(I.getVectorOperandType()); - uint64_t num = vecTy->getNumElements(); - Type* eleType = vecTy->getElementType(); - -diff --git a/IGC/Compiler/DebugInfo/ScalarVISAModule.cpp b/IGC/Compiler/DebugInfo/ScalarVISAModule.cpp -index 885ad65a..923f296e 100644 ---- a/IGC/Compiler/DebugInfo/ScalarVISAModule.cpp -+++ b/IGC/Compiler/DebugInfo/ScalarVISAModule.cpp -@@ -32,6 +32,7 @@ IN THE SOFTWARE. - #include "DebugInfo/DebugInfoUtils.hpp" - - #include "common/LLVMWarningsPush.hpp" -+#include "llvmWrapper/IR/DerivedTypes.h" - #include "llvm/IR/Function.h" - #include "llvm/IR/Module.h" - #include "common/LLVMWarningsPop.hpp" -@@ -581,7 +582,7 @@ ScalarVisaModule::GetVariableLocation(const llvm::Instruction* pInst) const - - if (pType->isVectorTy()) - { -- vectorNumElements = (unsigned)cast<VectorType>(pType)->getNumElements(); -+ vectorNumElements = (unsigned)cast<IGCLLVM::FixedVectorType>(pType)->getNumElements(); - } - else if (!pVar->IsUniform()) - { -diff --git a/IGC/Compiler/GenTTI.cpp b/IGC/Compiler/GenTTI.cpp -index 34b21c73..9e4d2f26 100644 ---- a/IGC/Compiler/GenTTI.cpp -+++ b/IGC/Compiler/GenTTI.cpp -@@ -216,7 +216,13 @@ namespace llvm { - - // Skip non-simple loop. - if (L->getNumBlocks() != 1) { -- if (IGC_IS_FLAG_ENABLED(EnableAdvRuntimeUnroll) && L->empty()) { -+ if (IGC_IS_FLAG_ENABLED(EnableAdvRuntimeUnroll) && -+#if LLVM_VERSION_MAJOR >= 12 -+ L->isInnermost() -+#else -+ L->empty() -+#endif -+ ) { - auto countNonPHI = [](BasicBlock* BB) { - unsigned Total = BB->size(); - unsigned PHIs = 0; -diff --git a/IGC/Compiler/GenUpdateCB.cpp b/IGC/Compiler/GenUpdateCB.cpp -index 89b414b6..9034066c 100644 ---- a/IGC/Compiler/GenUpdateCB.cpp -+++ b/IGC/Compiler/GenUpdateCB.cpp -@@ -24,6 +24,7 @@ IN THE SOFTWARE. - - #include "common/LLVMWarningsPush.hpp" - #include "llvmWrapper/Bitcode/BitcodeWriter.h" -+#include "llvmWrapper/IR/DerivedTypes.h" - #include <llvm/Support/ScaledNumber.h> - #include <llvm/Bitcode/BitcodeReader.h> - #include "llvm/IR/DebugInfo.h" -@@ -384,7 +385,7 @@ bool GenUpdateCB::runOnFunction(Function& F) - isResInfo(dyn_cast<GenIntrinsicInst>(inst, - GenISAIntrinsic::GenISA_resinfoptr), texId, lod, isUAV)) - { -- unsigned nelems = (unsigned)cast<VectorType>(inst->getType())->getNumElements(); -+ unsigned nelems = (unsigned)cast<IGCLLVM::FixedVectorType>(inst->getType())->getNumElements(); - SmallVector< SmallVector<ExtractElementInst*, 1>, 4> extracts(nelems); - if (VectorUsedByConstExtractOnly(inst, extracts)) - { -diff --git a/IGC/Compiler/HandleFRemInstructions.cpp b/IGC/Compiler/HandleFRemInstructions.cpp -index 94995a56..4523f3f0 100644 ---- a/IGC/Compiler/HandleFRemInstructions.cpp -+++ b/IGC/Compiler/HandleFRemInstructions.cpp -@@ -28,6 +28,7 @@ IN THE SOFTWARE. - #include <llvm/IR/Instructions.h> - #include <llvm/IR/InstIterator.h> - #include <llvm/Transforms/Utils/BasicBlockUtils.h> -+#include "llvmWrapper/IR/DerivedTypes.h" - #include "common/LLVMWarningsPop.hpp" - #include "Compiler/IGCPassSupport.h" - #include "Probe/Assertion.h" -@@ -77,7 +78,7 @@ void HandleFRemInstructions::visitFRem(llvm::BinaryOperator& I) - - if (ValType->isVectorTy()) - { -- auto VecCount = cast<VectorType>(ValType)->getNumElements(); -+ auto VecCount = cast<IGCLLVM::FixedVectorType>(ValType)->getNumElements(); - if (VecCount == 2 || VecCount == 3 || VecCount == 4 || VecCount == 8 || VecCount == 16) - { - VecStr = "v" + std::to_string(VecCount); -diff --git a/IGC/Compiler/HandleLoadStoreInstructions.cpp b/IGC/Compiler/HandleLoadStoreInstructions.cpp -index d18eec73..c5535701 100644 ---- a/IGC/Compiler/HandleLoadStoreInstructions.cpp -+++ b/IGC/Compiler/HandleLoadStoreInstructions.cpp -@@ -77,7 +77,7 @@ void HandleLoadStoreInstructions::visitLoadInst(llvm::LoadInst& I) - - if (I.getType()->isVectorTy()) - { -- numVectorElements = (uint32_t)cast<VectorType>(I.getType())->getNumElements(); -+ numVectorElements = (uint32_t)cast<IGCLLVM::FixedVectorType>(I.getType())->getNumElements(); - doubleDstType = IGCLLVM::FixedVectorType::get(builder.getDoubleTy(), numVectorElements); - } - uint as = ptrv->getType()->getPointerAddressSpace(); -@@ -163,7 +163,7 @@ void HandleLoadStoreInstructions::visitStoreInst(llvm::StoreInst& I) - - if (I.getValueOperand()->getType()->isVectorTy()) - { -- numVectorElements = (uint32_t)cast<VectorType>(I.getValueOperand()->getType())->getNumElements(); -+ numVectorElements = (uint32_t)cast<IGCLLVM::FixedVectorType>(I.getValueOperand()->getType())->getNumElements(); - } - - -diff --git a/IGC/Compiler/LegalizationPass.cpp b/IGC/Compiler/LegalizationPass.cpp -index b651350b..9524a495 100644 ---- a/IGC/Compiler/LegalizationPass.cpp -+++ b/IGC/Compiler/LegalizationPass.cpp -@@ -670,7 +670,7 @@ void Legalization::visitBitCastInst(llvm::BitCastInst& I) - if (!isa<TruncInst>(pZ->getOperand(0))) - return; - -- auto* pVecTy = cast<VectorType>(pZ->getDestTy()); -+ auto* pVecTy = cast<IGCLLVM::FixedVectorType>(pZ->getDestTy()); - if (pVecTy->getNumElements() != 3) - return; - -@@ -765,7 +765,7 @@ void Legalization::visitSelectInst(SelectInst& I) - } - else if (I.getType()->isVectorTy()) - { -- unsigned int vecSize = (unsigned)cast<VectorType>(I.getType())->getNumElements(); -+ unsigned int vecSize = (unsigned)cast<IGCLLVM::FixedVectorType>(I.getType())->getNumElements(); - Value* newVec = UndefValue::get(I.getType()); - m_builder->SetInsertPoint(&I); - for (unsigned int i = 0; i < vecSize; i++) -@@ -1147,7 +1147,7 @@ void Legalization::visitStoreInst(StoreInst& I) - if (ConstantDataVector * vec = dyn_cast<ConstantDataVector>(I.getOperand(0))) - { - Value* newVec = UndefValue::get(vec->getType()); -- unsigned int nbElement = (unsigned)cast<VectorType>(vec->getType())->getNumElements(); -+ unsigned int nbElement = (unsigned)cast<IGCLLVM::FixedVectorType>(vec->getType())->getNumElements(); - for (unsigned int i = 0; i < nbElement; i++) - { - Constant* cst = vec->getElementAsConstant(i); -@@ -1168,7 +1168,7 @@ void Legalization::visitStoreInst(StoreInst& I) - else if (ConstantVector * vec = dyn_cast<ConstantVector>(I.getOperand(0))) - { - Value* newVec = UndefValue::get(vec->getType()); -- unsigned int nbElement = (unsigned)cast<VectorType>(vec->getType())->getNumElements(); -+ unsigned int nbElement = (unsigned)cast<IGCLLVM::FixedVectorType>(vec->getType())->getNumElements(); - for (unsigned int i = 0; i < nbElement; i++) - { - Constant* cst = vec->getOperand(i); -@@ -1189,7 +1189,7 @@ void Legalization::visitStoreInst(StoreInst& I) - else if (ConstantAggregateZero * vec = dyn_cast<ConstantAggregateZero>(I.getOperand(0))) - { - Value* newVec = UndefValue::get(vec->getType()); -- unsigned int nbElement = (unsigned)cast<VectorType>(vec->getType())->getNumElements(); -+ unsigned int nbElement = (unsigned)cast<IGCLLVM::FixedVectorType>(vec->getType())->getNumElements(); - for (unsigned int i = 0; i < nbElement; i++) - { - Constant* cst = vec->getElementValue(i); -@@ -1340,7 +1340,7 @@ void Legalization::visitInsertElementInst(InsertElementInst& I) - if (ConstantDataVector * vec = dyn_cast<ConstantDataVector>(I.getOperand(0))) - { - Value* newVec = UndefValue::get(vec->getType()); -- unsigned int nbElement = (unsigned)cast<VectorType>(vec->getType())->getNumElements(); -+ unsigned int nbElement = (unsigned)cast<IGCLLVM::FixedVectorType>(vec->getType())->getNumElements(); - for (unsigned int i = 0; i < nbElement; i++) - { - Constant* cst = vec->getElementAsConstant(i); -@@ -1360,7 +1360,7 @@ void Legalization::visitInsertElementInst(InsertElementInst& I) - else if (ConstantVector * vec = dyn_cast<ConstantVector>(I.getOperand(0))) - { - Value* newVec = UndefValue::get(I.getType()); -- unsigned int nbElement = (unsigned)cast<VectorType>(vec->getType())->getNumElements(); -+ unsigned int nbElement = (unsigned)cast<IGCLLVM::FixedVectorType>(vec->getType())->getNumElements(); - for (unsigned int i = 0; i < nbElement; i++) - { - Constant* cst = vec->getOperand(i); -@@ -1380,7 +1380,7 @@ void Legalization::visitInsertElementInst(InsertElementInst& I) - else if (ConstantAggregateZero * vec = dyn_cast<ConstantAggregateZero>(I.getOperand(0))) - { - Value* newVec = UndefValue::get(I.getType()); -- unsigned int nbElement = (unsigned)cast<VectorType>(vec->getType())->getNumElements(); -+ unsigned int nbElement = (unsigned)cast<IGCLLVM::FixedVectorType>(vec->getType())->getNumElements(); - for (unsigned int i = 0; i < nbElement; i++) - { - Constant* cst = vec->getElementValue(i); -@@ -1397,7 +1397,7 @@ void Legalization::visitInsertElementInst(InsertElementInst& I) - else if (I.getOperand(1)->getType()->isIntegerTy(1)) - { - // This promotes i1 insertelement to i32 -- unsigned int nbElement = (unsigned)cast<VectorType>(I.getOperand(0)->getType())->getNumElements(); -+ unsigned int nbElement = (unsigned)cast<IGCLLVM::FixedVectorType>(I.getOperand(0)->getType())->getNumElements(); - Value* newVec = UndefValue::get(IGCLLVM::FixedVectorType::get(m_builder->getInt32Ty(), nbElement)); - PromoteInsertElement(&I, newVec); - } -@@ -1410,7 +1410,7 @@ void Legalization::visitShuffleVectorInst(ShuffleVectorInst& I) - // If the original vector is a constant, just use the scalar constant, - // otherwise extract from the original vector. - -- VectorType* resType = cast<VectorType>(I.getType()); -+ IGCLLVM::FixedVectorType* resType = cast<IGCLLVM::FixedVectorType>(I.getType()); - Value* newVec = UndefValue::get(resType); - Value* src0 = I.getOperand(0); - Value* src1 = I.getOperand(1); -@@ -1421,7 +1421,7 @@ void Legalization::visitShuffleVectorInst(ShuffleVectorInst& I) - Constant* mask = I.getShuffleMaskForBitcode(); - #endif - // The two inputs are guaranteed to be of the same type -- VectorType* inType = cast<VectorType>(src0->getType()); -+ IGCLLVM::FixedVectorType* inType = cast<IGCLLVM::FixedVectorType>(src0->getType()); - int inCount = int_cast<int>(inType->getNumElements()); - int inBase = 2; // 2 means using undef - // if inType == resType, use src0/src1 as the input -@@ -1637,7 +1637,7 @@ Type* Legalization::LegalAllocaType(Type* type) const - #endif - legalType = IGCLLVM::FixedVectorType::get( - LegalAllocaType(cast<VectorType>(type)->getElementType()), -- (unsigned)cast<VectorType>(type)->getNumElements()); -+ (unsigned)cast<IGCLLVM::FixedVectorType>(type)->getNumElements()); - break; - case Type::StructTyID: - return LegalStructAllocaType(type); -@@ -1937,7 +1937,7 @@ void Legalization::visitIntrinsicInst(llvm::IntrinsicInst& I) - Value* newValue = nullptr; - if (srcType->isVectorTy()) - { -- auto sourceVT = cast<VectorType>(srcType); -+ auto sourceVT = cast<IGCLLVM::FixedVectorType>(srcType); - const unsigned int numElements = (uint32_t)sourceVT->getNumElements(); - Value* dstVec = UndefValue::get(srcType); - for (unsigned int i = 0; i < numElements; ++i) -@@ -2081,7 +2081,7 @@ void Legalization::visitTruncInst(llvm::TruncInst& I) { - } - - Src = BC->getOperand(0); -- VectorType* VTy = dyn_cast<VectorType>(Src->getType()); -+ IGCLLVM::FixedVectorType* VTy = dyn_cast<IGCLLVM::FixedVectorType>(Src->getType()); - // Bail out if it's not bitcasted from <3 x i16> - if (!VTy || VTy->getNumElements() != 3 || !VTy->getElementType()->isIntegerTy(16)) - return; -diff --git a/IGC/Compiler/Legalizer/InstPromoter.cpp b/IGC/Compiler/Legalizer/InstPromoter.cpp -index bba1d96d..8fadf89f 100644 ---- a/IGC/Compiler/Legalizer/InstPromoter.cpp -+++ b/IGC/Compiler/Legalizer/InstPromoter.cpp -@@ -25,6 +25,7 @@ IN THE SOFTWARE. - #define DEBUG_TYPE "type-legalizer" - #include "TypeLegalizer.h" - #include "InstPromoter.h" -+#include "llvmWrapper/IR/DerivedTypes.h" - #include "common/LLVMWarningsPush.hpp" - #include "llvm/Support/Debug.h" - #include "llvm/Support/MathExtras.h" -@@ -394,10 +395,16 @@ bool InstPromoter::visitBitCastInst(BitCastInst& I) { - unsigned N = - Val->getType()->getScalarSizeInBits() / DestTy->getScalarSizeInBits(); - Value* BC = -- IRB->CreateBitCast(Val, VectorType::get(DestTy->getScalarType(), N)); -+ IRB->CreateBitCast(Val, IGCLLVM::FixedVectorType::get(DestTy->getScalarType(), N)); - - std::vector<Constant*> Vals; -- for (unsigned i = 0; i < DestTy->getVectorNumElements(); i++) -+ for (unsigned i = 0; -+#if LLVM_VERSION_MAJOR >= 12 -+ i < cast<IGCLLVM::FixedVectorType>(DestTy)->getNumElements(); -+#else -+ i < DestTy->getVectorNumElements(); -+#endif -+ i++) - Vals.push_back(IRB->getInt32(i)); - - Value* Mask = ConstantVector::get(Vals); -diff --git a/IGC/Compiler/Legalizer/InstScalarizer.cpp b/IGC/Compiler/Legalizer/InstScalarizer.cpp -index 3902f740..c8e18cc5 100644 ---- a/IGC/Compiler/Legalizer/InstScalarizer.cpp -+++ b/IGC/Compiler/Legalizer/InstScalarizer.cpp -@@ -130,7 +130,7 @@ bool InstScalarizer::visitLoadInst(LoadInst& I) { - // otherwise, it's broken. - IGC_ASSERT(TL->getTypeSizeInBits(EltTy) == TL->getTypeStoreSizeInBits(EltTy)); - -- unsigned NumElts = (unsigned)cast<VectorType>(OrigTy)->getNumElements(); -+ unsigned NumElts = (unsigned)cast<IGCLLVM::FixedVectorType>(OrigTy)->getNumElements(); - unsigned Elt = 0; - - Type* NewPtrTy = PointerType::get(EltTy, AS); -@@ -268,7 +268,7 @@ bool InstScalarizer::visitStoreInst(StoreInst& I) { - // otherwise, it's broken. - IGC_ASSERT(TL->getTypeSizeInBits(EltTy) == TL->getTypeStoreSizeInBits(EltTy)); - -- unsigned NumElts = (unsigned)cast<VectorType>(OrigTy)->getNumElements(); -+ unsigned NumElts = (unsigned)cast<IGCLLVM::FixedVectorType>(OrigTy)->getNumElements(); - unsigned Elt = 0; - - Type* NewPtrTy = PointerType::get(EltTy, AS); -@@ -498,7 +498,7 @@ bool InstScalarizer::visitInsertElementInst(InsertElementInst& I) { - IGC_ASSERT(EltSeq->size()); - IGC_ASSERT(VecSeqCopy.size() % EltSeq->size() == 0); - -- unsigned NumElts = (unsigned)cast<VectorType>(I.getOperand(0)->getType())->getNumElements(); -+ unsigned NumElts = (unsigned)cast<IGCLLVM::FixedVectorType>(I.getOperand(0)->getType())->getNumElements(); - unsigned i = 0; - for (unsigned Elt = 0; Elt != NumElts; ++Elt) { - if (Elt == Idx) { -diff --git a/IGC/Compiler/Legalizer/PeepholeTypeLegalizer.cpp b/IGC/Compiler/Legalizer/PeepholeTypeLegalizer.cpp -index 28f8ce84..778132c4 100644 ---- a/IGC/Compiler/Legalizer/PeepholeTypeLegalizer.cpp -+++ b/IGC/Compiler/Legalizer/PeepholeTypeLegalizer.cpp -@@ -162,7 +162,7 @@ void PeepholeTypeLegalizer::legalizePhiInstruction(Instruction& I) - - if (quotient > 1) - { -- unsigned numElements = I.getType()->isVectorTy() ? (unsigned)cast<VectorType>(I.getType())->getNumElements() : 1; -+ unsigned numElements = I.getType()->isVectorTy() ? (unsigned)cast<IGCLLVM::FixedVectorType>(I.getType())->getNumElements() : 1; - Type* newType = IGCLLVM::FixedVectorType::get(Type::getIntNTy(I.getContext(), promoteToInt), quotient * numElements); - - PHINode* newPhi = m_builder->CreatePHI(newType, oldPhi->getNumIncomingValues()); -@@ -223,7 +223,7 @@ void PeepholeTypeLegalizer::legalizeExtractElement(Instruction& I) - unsigned elementWidth = extract->getType()->getScalarSizeInBits(); - if (!isLegalInteger(elementWidth) && extract->getType()->isIntOrIntVectorTy()) - { -- unsigned numElements = (unsigned)cast<VectorType>(extract->getOperand(0)->getType())->getNumElements(); -+ unsigned numElements = (unsigned)cast<IGCLLVM::FixedVectorType>(extract->getOperand(0)->getType())->getNumElements(); - unsigned quotient, promoteToInt; - promoteInt(elementWidth, quotient, promoteToInt, DL->getLargestLegalIntTypeSizeInBits()); - -@@ -850,7 +850,7 @@ void PeepholeTypeLegalizer::cleanupZExtInst(Instruction& I) { - } - - unsigned ipElmtSize = prevInst->getOperand(0)->getType()->getScalarSizeInBits(); -- unsigned ipVecSize = (unsigned)cast<VectorType>(prevInst->getOperand(0)->getType())->getNumElements(); -+ unsigned ipVecSize = (unsigned)cast<IGCLLVM::FixedVectorType>(prevInst->getOperand(0)->getType())->getNumElements(); - unsigned convFactor = promoteToInt / ipElmtSize; - - Value* vecRes = UndefValue::get(IGCLLVM::FixedVectorType::get(llvm::Type::getIntNTy(I.getContext(), promoteToInt), quotient)); -diff --git a/IGC/Compiler/Legalizer/TypeLegalizer.cpp b/IGC/Compiler/Legalizer/TypeLegalizer.cpp -index f8f71d71..a46f42bd 100644 ---- a/IGC/Compiler/Legalizer/TypeLegalizer.cpp -+++ b/IGC/Compiler/Legalizer/TypeLegalizer.cpp -@@ -32,6 +32,7 @@ IN THE SOFTWARE. - #include "InstScalarizer.h" - #include "InstElementizer.h" - #include "common/LLVMWarningsPush.hpp" -+#include "llvmWrapper/IR/DerivedTypes.h" - #include "llvm/ADT/PostOrderIterator.h" - #include "llvm/IR/CFG.h" - #include "llvm/Support/Debug.h" -@@ -290,7 +291,7 @@ TypeSeq* TypeLegalizer::getScalarizedTypeSeq(Type* Ty) { - } - - Type* EltTy = cast<VectorType>(Ty)->getElementType(); -- for (unsigned i = 0, e = (unsigned)cast<VectorType>(Ty)->getNumElements(); i != e; ++i) -+ for (unsigned i = 0, e = (unsigned)cast<IGCLLVM::FixedVectorType>(Ty)->getNumElements(); i != e; ++i) - TMI->second.push_back(EltTy); - - return &TMI->second; -diff --git a/IGC/Compiler/Legalizer/TypeLegalizer.h b/IGC/Compiler/Legalizer/TypeLegalizer.h -index f3ce264c..a6910d08 100644 ---- a/IGC/Compiler/Legalizer/TypeLegalizer.h -+++ b/IGC/Compiler/Legalizer/TypeLegalizer.h -@@ -28,6 +28,7 @@ IN THE SOFTWARE. - #include "llvmWrapper/IR/Instructions.h" - #include "llvmWrapper/Analysis/InlineCost.h" - #include "llvmWrapper/IR/InstrTypes.h" -+#include "llvmWrapper/IR/DerivedTypes.h" - #include "llvmWrapper/Support/Alignment.h" - #include "llvm/ADT/ArrayRef.h" - #include "llvm/ADT/DenseMap.h" -@@ -287,7 +288,7 @@ namespace IGC { - if (!Ty->isVectorTy()) - return false; - -- unsigned NumElts = (unsigned)cast<VectorType>(Ty)->getNumElements(); -+ unsigned NumElts = (unsigned)cast<IGCLLVM::FixedVectorType>(Ty)->getNumElements(); - Type* EltTy = cast<VectorType>(Ty)->getElementType(); - const auto& ProfitLengths = getProfitLoadVectorLength(EltTy); - -diff --git a/IGC/Compiler/LowPrecisionOptPass.cpp b/IGC/Compiler/LowPrecisionOptPass.cpp -index 64912719..1d1db949 100644 ---- a/IGC/Compiler/LowPrecisionOptPass.cpp -+++ b/IGC/Compiler/LowPrecisionOptPass.cpp -@@ -252,7 +252,7 @@ bool LowPrecisionOpt::propagateSamplerType(llvm::GenIntrinsicInst& I) - return false; - } - -- VectorType* oldTy = cast<VectorType>(I.getType()); -+ IGCLLVM::FixedVectorType* oldTy = cast<IGCLLVM::FixedVectorType>(I.getType()); - llvm::SmallVector<llvm::Type*, 4> overloadTys; - auto retTy = IGCLLVM::FixedVectorType::get(newDstType, int_cast<unsigned int>(oldTy->getNumElements())); - overloadTys.push_back(retTy); -diff --git a/IGC/Compiler/Optimizer/OpenCLPasses/AddressSpaceAliasAnalysis/AddressSpaceAliasAnalysis.cpp b/IGC/Compiler/Optimizer/OpenCLPasses/AddressSpaceAliasAnalysis/AddressSpaceAliasAnalysis.cpp -index 393b4b0a..e9c07b34 100644 ---- a/IGC/Compiler/Optimizer/OpenCLPasses/AddressSpaceAliasAnalysis/AddressSpaceAliasAnalysis.cpp -+++ b/IGC/Compiler/Optimizer/OpenCLPasses/AddressSpaceAliasAnalysis/AddressSpaceAliasAnalysis.cpp -@@ -23,6 +23,8 @@ IN THE SOFTWARE. - ============================= end_copyright_notice ===========================*/ - - #include "llvm/Config/llvm-config.h" -+#include "llvmWrapper/IR/DerivedTypes.h" -+#include "llvmWrapper/Analysis/TargetLibraryInfo.h" - #include "Compiler/Optimizer/OpenCLPasses/AddressSpaceAliasAnalysis/AddressSpaceAliasAnalysis.h" - #include "Compiler/CodeGenPublic.h" - #include "Compiler/IGCPassSupport.h" -@@ -178,12 +180,11 @@ namespace { - bool doInitialization(Module& M) override { - if(M.size() > 0) - { -- Result.reset(new AddressSpaceAAResult( -- getAnalysis<TargetLibraryInfoWrapperPass>().getTLI( - #if LLVM_VERSION_MAJOR >= 10 -- *M.begin() -+ Function &F = *M.begin(); - #endif -- ), -+ Result.reset(new AddressSpaceAAResult( -+ getAnalysis<TargetLibraryInfoWrapperPass>().getTLI(), - *getAnalysis<CodeGenContextWrapper>().getCodeGenContext())); - } - return false; -diff --git a/IGC/Compiler/Optimizer/OpenCLPasses/AggregateArguments/AggregateArguments.cpp b/IGC/Compiler/Optimizer/OpenCLPasses/AggregateArguments/AggregateArguments.cpp -index 67535cfc..3cdd2975 100644 ---- a/IGC/Compiler/Optimizer/OpenCLPasses/AggregateArguments/AggregateArguments.cpp -+++ b/IGC/Compiler/Optimizer/OpenCLPasses/AggregateArguments/AggregateArguments.cpp -@@ -24,6 +24,7 @@ IN THE SOFTWARE. - - #include "Compiler/Optimizer/OpenCLPasses/AggregateArguments/AggregateArguments.hpp" - #include "Compiler/IGCPassSupport.h" -+#include "llvmWrapper/IR/DerivedTypes.h" - #include "common/LLVMWarningsPush.hpp" - #include "llvm/IR/Function.h" - #include "common/LLVMWarningsPop.hpp" -@@ -130,7 +131,7 @@ static uint64_t getNumElements(Type* type) - { - return arrayType->getNumElements(); - } -- if (VectorType * vectorType = dyn_cast<VectorType>(type)) -+ if (IGCLLVM::FixedVectorType * vectorType = dyn_cast<IGCLLVM::FixedVectorType>(type)) - { - return vectorType->getNumElements(); - } -diff --git a/IGC/Compiler/Optimizer/OpenCLPasses/CorrectlyRoundedDivSqrt/CorrectlyRoundedDivSqrt.cpp b/IGC/Compiler/Optimizer/OpenCLPasses/CorrectlyRoundedDivSqrt/CorrectlyRoundedDivSqrt.cpp -index 2a9d7499..79ed0948 100644 ---- a/IGC/Compiler/Optimizer/OpenCLPasses/CorrectlyRoundedDivSqrt/CorrectlyRoundedDivSqrt.cpp -+++ b/IGC/Compiler/Optimizer/OpenCLPasses/CorrectlyRoundedDivSqrt/CorrectlyRoundedDivSqrt.cpp -@@ -27,6 +27,7 @@ IN THE SOFTWARE. - #include "Compiler/IGCPassSupport.h" - #include "GenISAIntrinsics/GenIntrinsicInst.h" - -+#include "llvmWrapper/IR/DerivedTypes.h" - #include "common/LLVMWarningsPush.hpp" - #include <llvm/IR/Function.h> - #include <llvm/IR/Instructions.h> -@@ -133,7 +134,7 @@ Value* CorrectlyRoundedDivSqrt::emitIEEEDivide(BinaryOperator* I, Value* Op0, Va - } - else - { -- auto vType = dyn_cast<VectorType>(Ty); -+ auto vType = dyn_cast<IGCLLVM::FixedVectorType>(Ty); - unsigned VecLen = (uint32_t)vType->getNumElements(); - Divide = UndefValue::get(Ty); - for (unsigned i = 0; i < VecLen; i++) -diff --git a/IGC/Compiler/Optimizer/OpenCLPasses/DeviceEnqueueFuncs/TransformBlocks.cpp b/IGC/Compiler/Optimizer/OpenCLPasses/DeviceEnqueueFuncs/TransformBlocks.cpp -index c5ff3ac0..119520ed 100644 ---- a/IGC/Compiler/Optimizer/OpenCLPasses/DeviceEnqueueFuncs/TransformBlocks.cpp -+++ b/IGC/Compiler/Optimizer/OpenCLPasses/DeviceEnqueueFuncs/TransformBlocks.cpp -@@ -952,7 +952,11 @@ namespace //Anonymous - { - auto ndrangeStructName = "struct.ndrange_t"; - auto module = _deviceExecCall->getModule(); -+#if LLVM_VERSION_MAJOR >= 12 -+ auto ndrangeTy = llvm::StructType::getTypeByName(module->getContext(), ndrangeStructName); -+#else - auto ndrangeTy = module->getTypeByName(ndrangeStructName); -+#endif - if (ndrangeTy == nullptr) - { - //create struct type -@@ -1697,7 +1701,7 @@ namespace //Anonymous - // this generates <element_type><num_elements> string. Ie for char2 element_type is char and num_elements is 2 - // that is done by callin BaseTypeName on vector element type, this recursive call has only a depth of one since - // there are no compound vectors in OpenCL. -- auto vType = llvm::dyn_cast<VectorType>(type); -+ auto vType = llvm::dyn_cast<IGCLLVM::FixedVectorType>(type); - return BaseTypeName(type->getContainedType(0), os) << vType->getNumElements(); - } - default: -diff --git a/IGC/Compiler/Optimizer/OpenCLPasses/KernelArgs.cpp b/IGC/Compiler/Optimizer/OpenCLPasses/KernelArgs.cpp -index 6448463d..7df71079 100644 ---- a/IGC/Compiler/Optimizer/OpenCLPasses/KernelArgs.cpp -+++ b/IGC/Compiler/Optimizer/OpenCLPasses/KernelArgs.cpp -@@ -24,6 +24,7 @@ IN THE SOFTWARE. - - #include "Compiler/Optimizer/OpenCLPasses/KernelArgs.hpp" - #include "AdaptorCommon/ImplicitArgs.hpp" -+#include "llvmWrapper/IR/DerivedTypes.h" - #include "common/LLVMWarningsPush.hpp" - #include <llvm/IR/Argument.h> - #include <llvm/IR/DataLayout.h> -@@ -388,7 +389,7 @@ unsigned int KernelArg::calcAssociatedArgNo(const ImplicitArg& implicitArg, cons - - unsigned int KernelArg::getNumComponents() const - { -- if (VectorType * vecType = dyn_cast<VectorType>(m_arg->getType())) -+ if (IGCLLVM::FixedVectorType * vecType = dyn_cast<IGCLLVM::FixedVectorType>(m_arg->getType())) - { - // Vector - return int_cast<unsigned int>(vecType->getNumElements()); -diff --git a/IGC/Compiler/Optimizer/OpenCLPasses/OpenCLPrintf/OpenCLPrintfResolution.cpp b/IGC/Compiler/Optimizer/OpenCLPasses/OpenCLPrintf/OpenCLPrintfResolution.cpp -index c7414b8e..74964510 100644 ---- a/IGC/Compiler/Optimizer/OpenCLPasses/OpenCLPrintf/OpenCLPrintfResolution.cpp -+++ b/IGC/Compiler/Optimizer/OpenCLPasses/OpenCLPrintf/OpenCLPrintfResolution.cpp -@@ -719,7 +719,7 @@ Value* OpenCLPrintfResolution::fixupPrintfArg(CallInst& printfCall, Value* arg, - } - - Type* newType = Type::getFloatTy(arg->getContext()); -- if (auto argVT = dyn_cast<VectorType>(arg->getType())) -+ if (auto argVT = dyn_cast<IGCLLVM::FixedVectorType>(arg->getType())) - { - newType = IGCLLVM::FixedVectorType::get(newType, (unsigned)argVT->getNumElements()); - } -@@ -749,7 +749,7 @@ void OpenCLPrintfResolution::preprocessPrintfArgs(CallInst& printfCall) - IGC::SHADER_PRINTF_TYPE argDataType = getPrintfArgDataType(arg); - arg = fixupPrintfArg(printfCall, arg, argDataType); - uint vecSize = 0; -- if (auto argVType = dyn_cast<VectorType>(argType)) { -+ if (auto argVType = dyn_cast<IGCLLVM::FixedVectorType>(argType)) { - vecSize = (uint)argVType->getNumElements(); - } - m_argDescriptors.push_back(SPrintfArgDescriptor(argDataType, arg, vecSize)); -diff --git a/IGC/Compiler/Optimizer/OpenCLPasses/PrivateMemory/PrivateMemoryResolution.cpp b/IGC/Compiler/Optimizer/OpenCLPasses/PrivateMemory/PrivateMemoryResolution.cpp -index 2ed5a2b6..07f85f4c 100644 ---- a/IGC/Compiler/Optimizer/OpenCLPasses/PrivateMemory/PrivateMemoryResolution.cpp -+++ b/IGC/Compiler/Optimizer/OpenCLPasses/PrivateMemory/PrivateMemoryResolution.cpp -@@ -29,6 +29,7 @@ IN THE SOFTWARE. - #include "Compiler/IGCPassSupport.h" - #include "Compiler/CISACodeGen/GenCodeGenModule.h" - #include "Compiler/CISACodeGen/LowerGEPForPrivMem.hpp" -+#include "llvmWrapper/IR/DerivedTypes.h" - #include "common/LLVMWarningsPush.hpp" - #include "llvm/Transforms/Utils/Local.h" - #include "llvm/IR/DataLayout.h" -@@ -604,7 +605,7 @@ public: - Type* scalarptrTy = PointerType::get(scalarType, pLoad->getPointerAddressSpace()); - IGC_ASSERT(scalarType->getPrimitiveSizeInBits() / 8 == elementSize); - Value* vec = UndefValue::get(pLoad->getType()); -- auto pLoadVT = cast<VectorType>(pLoad->getType()); -+ auto pLoadVT = cast<IGCLLVM::FixedVectorType>(pLoad->getType()); - for (unsigned i = 0, e = (unsigned)pLoadVT->getNumElements(); i < e; ++i) - { - Value* ptr = IRB.CreateIntToPtr(address, scalarptrTy); -@@ -643,7 +644,7 @@ public: - IGC_ASSERT(scalarType->getPrimitiveSizeInBits() / 8 == elementSize); - Value* vec = pStore->getValueOperand(); - -- unsigned vecNumElts = (unsigned)cast<VectorType>(vec->getType())->getNumElements(); -+ unsigned vecNumElts = (unsigned)cast<IGCLLVM::FixedVectorType>(vec->getType())->getNumElements(); - for (unsigned i = 0; i < vecNumElts; ++i) - { - Value* ptr = IRB.CreateIntToPtr(address, scalarptrTy); -@@ -707,7 +708,7 @@ bool PrivateMemoryResolution::testTransposedMemory(const Type* pTmpType, const T - } - else if(pTmpType->isVectorTy()) - { -- auto pTmpVType = cast<VectorType>(pTmpType); -+ auto pTmpVType = cast<IGCLLVM::FixedVectorType>(pTmpType); - tmpAllocaSize *= pTmpVType->getNumElements(); - pTmpType = pTmpType->getContainedType(0); - ok = (nullptr != pTmpType); -@@ -815,7 +816,11 @@ bool PrivateMemoryResolution::resolveAllocaInstructions(bool privateOnStack) - // Construct an empty DebugLoc. - IF_DEBUG_INFO(DebugLoc entryDebugLoc); - // Assign with the function location if available. -+#if LLVM_VERSION_MAJOR >= 12 -+ IF_DEBUG_INFO_IF(DISubprogram *subprogram = m_currFunction->getSubprogram(), entryDebugLoc = DILocation::get(subprogram->getContext(), subprogram->getLine(), 0, subprogram);); -+#else - IF_DEBUG_INFO_IF(DISubprogram *subprogram = m_currFunction->getSubprogram(), entryDebugLoc = DebugLoc::get(subprogram->getLine(), 0, subprogram);); -+#endif - IF_DEBUG_INFO(entryBuilder.SetCurrentDebugLocation(entryDebugLoc)); - - if (privateOnStack) -diff --git a/IGC/Compiler/Optimizer/OpenCLPasses/ReplaceUnsupportedIntrinsics/ReplaceUnsupportedIntrinsics.cpp b/IGC/Compiler/Optimizer/OpenCLPasses/ReplaceUnsupportedIntrinsics/ReplaceUnsupportedIntrinsics.cpp -index c8414efa..68491d48 100644 ---- a/IGC/Compiler/Optimizer/OpenCLPasses/ReplaceUnsupportedIntrinsics/ReplaceUnsupportedIntrinsics.cpp -+++ b/IGC/Compiler/Optimizer/OpenCLPasses/ReplaceUnsupportedIntrinsics/ReplaceUnsupportedIntrinsics.cpp -@@ -254,7 +254,7 @@ Instruction* ReplaceUnsupportedIntrinsics::insertLoop(Instruction* Loc, Value* L - Value* ReplaceUnsupportedIntrinsics::replicateScalar( - Value* ScalarVal, Type* Ty, Instruction* InsertBefore) - { -- VectorType* VTy = dyn_cast<VectorType>(Ty); -+ IGCLLVM::FixedVectorType* VTy = dyn_cast<IGCLLVM::FixedVectorType>(Ty); - Type* ETy = VTy ? VTy->getElementType() : Ty; - uint32_t sBits = (unsigned int)ScalarVal->getType()->getPrimitiveSizeInBits(); - uint32_t nBits = (unsigned int)ETy->getPrimitiveSizeInBits(); -@@ -868,7 +868,7 @@ void ReplaceUnsupportedIntrinsics::replaceFunnelShift(IntrinsicInst* I) { - } - - Value* numBits = Builder.getIntN(sizeInBits, sizeInBits); -- if (auto IVT = dyn_cast<VectorType>(I->getType())) { -+ if (auto IVT = dyn_cast<IGCLLVM::FixedVectorType>(I->getType())) { - numBits = ConstantVector::getSplat(IGCLLVM::getElementCount((uint32_t)IVT->getNumElements()), cast<Constant>(numBits)); - } - auto shiftModulo = Builder.CreateURem(I->getArgOperand(2), numBits); -diff --git a/IGC/Compiler/Optimizer/OpenCLPasses/WIFuncs/WIFuncResolution.cpp b/IGC/Compiler/Optimizer/OpenCLPasses/WIFuncs/WIFuncResolution.cpp -index 7188cc50..535d6268 100644 ---- a/IGC/Compiler/Optimizer/OpenCLPasses/WIFuncs/WIFuncResolution.cpp -+++ b/IGC/Compiler/Optimizer/OpenCLPasses/WIFuncs/WIFuncResolution.cpp -@@ -303,7 +303,11 @@ static Value* BuildLoadInst(CallInst& CI, unsigned int Offset, Type* DataType) - auto Size = ElemByteSize; - if (DataType->isVectorTy()) - { -+#if LLVM_VERSION_MAJOR >= 12 -+ Size *= cast<IGCLLVM::FixedVectorType>(DataType)->getNumElements(); -+#else - Size *= DataType->getVectorNumElements(); -+#endif - } - unsigned int AlignedOffset = (Offset / ElemByteSize) * ElemByteSize; - unsigned int LoadByteSize = (Offset == AlignedOffset) ? Size : Size * 2; -diff --git a/IGC/Compiler/Optimizer/PreCompiledFuncImport.cpp b/IGC/Compiler/Optimizer/PreCompiledFuncImport.cpp -index 1fda21d5..63a22864 100644 ---- a/IGC/Compiler/Optimizer/PreCompiledFuncImport.cpp -+++ b/IGC/Compiler/Optimizer/PreCompiledFuncImport.cpp -@@ -340,7 +340,7 @@ bool PreCompiledFuncImport::preProcessDouble() - } - else - { -- uint32_t vectorSize = cast<VectorType>(Inst->getType())->getNumElements(); -+ uint32_t vectorSize = cast<IGCLLVM::FixedVectorType>(Inst->getType())->getNumElements(); - fsub = llvm::UndefValue::get(Inst->getType()); - - for (uint32_t i = 0; i < vectorSize; ++i) -@@ -936,7 +936,7 @@ void PreCompiledFuncImport::processDivide(BinaryOperator& inst, EmulatedFunction - - Type* argumentType = inst.getOperand(0)->getType(); - -- if (auto argumentVType = dyn_cast<VectorType>(argumentType)) -+ if (auto argumentVType = dyn_cast<IGCLLVM::FixedVectorType>(argumentType)) - { - numElements = (unsigned)argumentVType->getNumElements(); - } -diff --git a/IGC/Compiler/Optimizer/Scalarizer.cpp b/IGC/Compiler/Optimizer/Scalarizer.cpp -index 4f59eaf9..a4e73a6d 100644 ---- a/IGC/Compiler/Optimizer/Scalarizer.cpp -+++ b/IGC/Compiler/Optimizer/Scalarizer.cpp -@@ -289,7 +289,7 @@ void ScalarizeFunction::scalarizeInstruction(BinaryOperator* BI) - { - V_PRINT(scalarizer, "\t\tBinary instruction\n"); - IGC_ASSERT_MESSAGE(BI, "instruction type dynamic cast failed"); -- VectorType* instType = dyn_cast<VectorType>(BI->getType()); -+ IGCLLVM::FixedVectorType* instType = dyn_cast<IGCLLVM::FixedVectorType>(BI->getType()); - // Only need handling for vector binary ops - if (!instType) return; - -@@ -349,7 +349,7 @@ void ScalarizeFunction::scalarizeInstruction(CmpInst* CI) - { - V_PRINT(scalarizer, "\t\tCompare instruction\n"); - IGC_ASSERT_MESSAGE(CI, "instruction type dynamic cast failed"); -- VectorType* instType = dyn_cast<VectorType>(CI->getType()); -+ IGCLLVM::FixedVectorType* instType = dyn_cast<IGCLLVM::FixedVectorType>(CI->getType()); - // Only need handling for vector compares - if (!instType) return; - -@@ -397,13 +397,13 @@ void ScalarizeFunction::scalarizeInstruction(CastInst* CI) - { - V_PRINT(scalarizer, "\t\tCast instruction\n"); - IGC_ASSERT_MESSAGE(CI, "instruction type dynamic cast failed"); -- VectorType* instType = dyn_cast<VectorType>(CI->getType()); -+ IGCLLVM::FixedVectorType* instType = dyn_cast<IGCLLVM::FixedVectorType>(CI->getType()); - - // For BitCast - we only scalarize if src and dst types have same vector length - if (isa<BitCastInst>(CI)) - { - if (!instType) return recoverNonScalarizableInst(CI); -- VectorType* srcType = dyn_cast<VectorType>(CI->getOperand(0)->getType()); -+ IGCLLVM::FixedVectorType* srcType = dyn_cast<IGCLLVM::FixedVectorType>(CI->getOperand(0)->getType()); - if (!srcType || (instType->getNumElements() != srcType->getNumElements())) - { - return recoverNonScalarizableInst(CI); -@@ -458,7 +458,7 @@ void ScalarizeFunction::scalarizeInstruction(PHINode* PI) - { - V_PRINT(scalarizer, "\t\tPHI instruction\n"); - IGC_ASSERT_MESSAGE(PI, "instruction type dynamic cast failed"); -- VectorType* instType = dyn_cast<VectorType>(PI->getType()); -+ IGCLLVM::FixedVectorType* instType = dyn_cast<IGCLLVM::FixedVectorType>(PI->getType()); - // Only need handling for vector PHI - if (!instType) return; - -@@ -569,7 +569,7 @@ void ScalarizeFunction::scalarizeInstruction(SelectInst* SI) - { - V_PRINT(scalarizer, "\t\tSelect instruction\n"); - IGC_ASSERT_MESSAGE(SI, "instruction type dynamic cast failed"); -- VectorType* instType = dyn_cast<VectorType>(SI->getType()); -+ IGCLLVM::FixedVectorType* instType = dyn_cast<IGCLLVM::FixedVectorType>(SI->getType()); - // Only need handling for vector select - if (!instType) return; - -@@ -651,7 +651,7 @@ void ScalarizeFunction::scalarizeInstruction(ExtractElementInst* EI) - - // Connect the "extracted" value to all its consumers - uint64_t scalarIndex = cast<ConstantInt>(scalarIndexVal)->getZExtValue(); -- auto valueVType = cast<VectorType>(vectorValue->getType()); -+ auto valueVType = cast<IGCLLVM::FixedVectorType>(vectorValue->getType()); - if (static_cast<unsigned int>(scalarIndex) < (unsigned)valueVType->getNumElements()) - { - IGC_ASSERT_MESSAGE(NULL != operand[static_cast<unsigned int>(scalarIndex)], "SCM error"); -@@ -715,7 +715,7 @@ void ScalarizeFunction::scalarizeInstruction(InsertElementInst* II) - if (isa<UndefValue>(sourceVectorValue)) - { - // Scalarize the undef value (generate a scalar undef) -- VectorType* inputVectorType = dyn_cast<VectorType>(sourceVectorValue->getType()); -+ IGCLLVM::FixedVectorType* inputVectorType = dyn_cast<IGCLLVM::FixedVectorType>(sourceVectorValue->getType()); - IGC_ASSERT_MESSAGE(inputVectorType, "expected vector argument"); - - UndefValue* undefVal = UndefValue::get(inputVectorType->getElementType()); -@@ -756,7 +756,7 @@ void ScalarizeFunction::scalarizeInstruction(ShuffleVectorInst* SI) - IGC_ASSERT(nullptr != sourceVector0Value); - Value* sourceVector1Value = SI->getOperand(1); - IGC_ASSERT(nullptr != sourceVector1Value); -- VectorType* const inputType = dyn_cast<VectorType>(sourceVector0Value->getType()); -+ IGCLLVM::FixedVectorType* const inputType = dyn_cast<IGCLLVM::FixedVectorType>(sourceVector0Value->getType()); - IGC_ASSERT_MESSAGE(nullptr != inputType, "vector input error"); - IGC_ASSERT_MESSAGE(inputType == sourceVector1Value->getType(), "vector input error"); - unsigned sourceVectorWidth = int_cast<unsigned>(inputType->getNumElements()); -@@ -778,7 +778,7 @@ void ScalarizeFunction::scalarizeInstruction(ShuffleVectorInst* SI) - - // Generate array for shuffled scalar values - SmallVector<Value*, MAX_INPUT_VECTOR_WIDTH>newVector; -- unsigned width = int_cast<unsigned>(SI->getType()->getNumElements()); -+ unsigned width = int_cast<unsigned>(dyn_cast<IGCLLVM::FixedVectorType>(SI->getType())->getNumElements()); - - // Generate undef value, which may be needed as some scalar elements - UndefValue* undef = UndefValue::get(inputType->getElementType()); -@@ -847,7 +847,7 @@ void ScalarizeFunction::scalarizeInstruction(GetElementPtrInst* GI) - - if (baseValue->getType()->isVectorTy()) - { -- width = int_cast<unsigned>(dyn_cast<VectorType>(baseValue->getType())->getNumElements()); -+ width = int_cast<unsigned>(dyn_cast<IGCLLVM::FixedVectorType>(baseValue->getType())->getNumElements()); - // Obtain the scalarized operands - obtainScalarizedValues(operand1, NULL, baseValue, GI); - ptrTy = dyn_cast<VectorType>(baseValue->getType())->getElementType(); -@@ -858,7 +858,7 @@ void ScalarizeFunction::scalarizeInstruction(GetElementPtrInst* GI) - } - if (indexValue->getType()->isVectorTy()) - { -- width = int_cast<unsigned>(dyn_cast<VectorType>(indexValue->getType())->getNumElements()); -+ width = int_cast<unsigned>(dyn_cast<IGCLLVM::FixedVectorType>(indexValue->getType())->getNumElements()); - // Obtain the scalarized operands - obtainScalarizedValues(operand2, NULL, indexValue, GI); - } -@@ -897,7 +897,7 @@ void ScalarizeFunction::scalarizeInstruction(LoadInst* LI) - V_PRINT(scalarizer, "\t\tLoad instruction\n"); - IGC_ASSERT_MESSAGE(LI, "instruction type dynamic cast failed"); - -- VectorType* dataType = dyn_cast<VectorType>(LI->getType()); -+ IGCLLVM::FixedVectorType* dataType = dyn_cast<IGCLLVM::FixedVectorType>(LI->getType()); - if (isScalarizableLoadStoreType(dataType) && m_pDL) - { - // Prepare empty SCM entry for the instruction -@@ -969,7 +969,7 @@ void ScalarizeFunction::scalarizeInstruction(StoreInst* SI) - - int indexPtr = SI->getPointerOperandIndex(); - int indexData = 1 - indexPtr; -- VectorType* dataType = dyn_cast<VectorType>(SI->getOperand(indexData)->getType()); -+ IGCLLVM::FixedVectorType* dataType = dyn_cast<IGCLLVM::FixedVectorType>(SI->getOperand(indexData)->getType()); - if (isScalarizableLoadStoreType(dataType) && m_pDL) - { - // Get additional info from instruction -@@ -1040,7 +1040,7 @@ void ScalarizeFunction::obtainScalarizedValues(SmallVectorImpl<Value*>& retValue - { - V_PRINT(scalarizer, "\t\t\tObtaining scalar value... " << *origValue << "\n"); - -- VectorType* origType = dyn_cast<VectorType>(origValue->getType()); -+ IGCLLVM::FixedVectorType* origType = dyn_cast<IGCLLVM::FixedVectorType>(origValue->getType()); - IGC_ASSERT_MESSAGE(origType, "Value must have a vector type!"); - unsigned width = int_cast<unsigned>(origType->getNumElements()); - -@@ -1202,7 +1202,7 @@ void ScalarizeFunction::obtainVectorValueWhichMightBeScalarizedImpl(Value* vecto - } - - Value* assembledVector = UndefValue::get(vectorVal->getType()); -- unsigned width = int_cast<unsigned>(dyn_cast<VectorType>(vectorVal->getType())->getNumElements()); -+ unsigned width = int_cast<unsigned>(dyn_cast<IGCLLVM::FixedVectorType>(vectorVal->getType())->getNumElements()); - for (unsigned i = 0; i < width; i++) - { - IGC_ASSERT_MESSAGE(NULL != valueEntry->scalarValues[i], "SCM entry has NULL value"); -@@ -1262,7 +1262,7 @@ void ScalarizeFunction::updateSCMEntryWithValues(ScalarizeFunction::SCMEntry* en - bool matchDbgLoc) - { - IGC_ASSERT_MESSAGE((origValue->getType()->isArrayTy() || origValue->getType()->isVectorTy()), "only Vector values are supported"); -- unsigned width = int_cast<unsigned>(dyn_cast<VectorType>(origValue->getType())->getNumElements()); -+ unsigned width = int_cast<unsigned>(dyn_cast<IGCLLVM::FixedVectorType>(origValue->getType())->getNumElements()); - - entry->isOriginalVectorRemoved = isOrigValueRemoved; - -@@ -1326,7 +1326,7 @@ void ScalarizeFunction::resolveDeferredInstructions() - Instruction* vectorInst = dyn_cast<Instruction>(current.unresolvedInst); - IGC_ASSERT_MESSAGE(vectorInst, "DRL only handles unresolved instructions"); - -- VectorType* currType = dyn_cast<VectorType>(vectorInst->getType()); -+ IGCLLVM::FixedVectorType* currType = dyn_cast<IGCLLVM::FixedVectorType>(vectorInst->getType()); - IGC_ASSERT_MESSAGE(currType, "Cannot have DRL of non-vector value"); - unsigned width = int_cast<unsigned>(currType->getNumElements()); - -diff --git a/IGC/Compiler/Optimizer/ValueTracker.cpp b/IGC/Compiler/Optimizer/ValueTracker.cpp -index c06f7312..d8f0d385 100644 ---- a/IGC/Compiler/Optimizer/ValueTracker.cpp -+++ b/IGC/Compiler/Optimizer/ValueTracker.cpp -@@ -178,7 +178,7 @@ Value* ValueTracker::handleExtractElement(ExtractElementInst* E) - { - auto mask = I->getShuffleMask(); - uint shuffleidx = int_cast<uint>(mask[(uint)idx]); -- auto vType = dyn_cast<VectorType>(I->getOperand(0)->getType()); -+ auto vType = dyn_cast<IGCLLVM::FixedVectorType>(I->getOperand(0)->getType()); - baseValue = (shuffleidx < vType->getNumElements()) ? - I->getOperand(0) : I->getOperand(1); - } -diff --git a/IGC/Compiler/VerificationPass.cpp b/IGC/Compiler/VerificationPass.cpp -index 98437d79..40d50f5a 100644 ---- a/IGC/Compiler/VerificationPass.cpp -+++ b/IGC/Compiler/VerificationPass.cpp -@@ -302,7 +302,7 @@ bool VerificationPass::verifyType(Type* type, Value* val) - - case IGCLLVM::VectorTyID: - { -- auto VType = cast<VectorType>(type); -+ auto VType = cast<IGCLLVM::FixedVectorType>(type); - unsigned typeSize = (unsigned)VType->getNumElements(); - if (!m_IGC_IR_spec.vectorTypeSizes.count(typeSize)) - { -diff --git a/IGC/DebugInfo/DebugInfoUtils.hpp b/IGC/DebugInfo/DebugInfoUtils.hpp -index 7f2e288a..b77a550d 100644 ---- a/IGC/DebugInfo/DebugInfoUtils.hpp -+++ b/IGC/DebugInfo/DebugInfoUtils.hpp -@@ -108,7 +108,11 @@ namespace IGC - IGCLLVM::DIBuilder Builder(M); - llvm::DIGlobalVariable* GV = GVs[j]->getVariable(); - llvm::DIScope* scopeToUse = GV->getScope(); -+#if LLVM_VERSION_MAJOR >= 12 -+ llvm::DILocation* locToUse = llvm::DILocation::get(scopeToUse->getContext(), GV->getLine(), 0, scopeToUse, loc); -+#else - llvm::DILocation* locToUse = llvm::DebugLoc::get(GV->getLine(), 0, scopeToUse, loc); -+#endif - if (llvm::isa<llvm::DICompileUnit>(GV->getScope())) - { - // Function has no DebugLoc so it is either internal -diff --git a/IGC/DebugInfo/DwarfDebug.cpp b/IGC/DebugInfo/DwarfDebug.cpp -index 683b83f7..bd9f17b7 100644 ---- a/IGC/DebugInfo/DwarfDebug.cpp -+++ b/IGC/DebugInfo/DwarfDebug.cpp -@@ -2102,9 +2102,17 @@ static DebugLoc getFnDebugLoc(DebugLoc DL, const LLVMContext& Ctx) - // Check for number of operands since the compatibility is cheap here. - if (SP->getNumOperands() > 19) - { -+#if LLVM_VERSION_MAJOR >= 12 -+ return DILocation::get(SP->getContext(), SP->getScopeLine(), 0, SP); -+#else - return DebugLoc::get(SP->getScopeLine(), 0, SP); -+#endif - } -+#if LLVM_VERSION_MAJOR >= 12 -+ return DILocation::get(SP->getContext(), SP->getLine(), 0, SP); -+#else - return DebugLoc::get(SP->getLine(), 0, SP); -+#endif - } - - return DebugLoc(); -diff --git a/IGC/GenISAIntrinsics/GenIntrinsics.cpp b/IGC/GenISAIntrinsics/GenIntrinsics.cpp -index 708cfdaa..772155b2 100644 ---- a/IGC/GenISAIntrinsics/GenIntrinsics.cpp -+++ b/IGC/GenISAIntrinsics/GenIntrinsics.cpp -@@ -204,7 +204,7 @@ static Type *DecodeFixedType(ArrayRef<GenISAIntrinsic::IITDescriptor> &Infos, - case IITDescriptor::SameVecWidthArgument: { - Type *EltTy = DecodeFixedType(Infos, Tys, Context); - Type *Ty = Tys[D.getArgumentNumber()]; -- if (VectorType *VTy = dyn_cast<VectorType>(Ty)) { -+ if (IGCLLVM::FixedVectorType *VTy = dyn_cast<IGCLLVM::FixedVectorType>(Ty)) { - return IGCLLVM::FixedVectorType::get(EltTy, int_cast<unsigned int>(VTy->getNumElements())); - } - IGC_ASSERT_EXIT_MESSAGE(0, "unhandled"); -@@ -215,7 +215,7 @@ static Type *DecodeFixedType(ArrayRef<GenISAIntrinsic::IITDescriptor> &Infos, - } - case IITDescriptor::VecOfPtrsToElt: { - Type *Ty = Tys[D.getArgumentNumber()]; -- VectorType *VTy = dyn_cast<VectorType>(Ty); -+ IGCLLVM::FixedVectorType *VTy = dyn_cast<IGCLLVM::FixedVectorType>(Ty); - if (!VTy) - IGC_ASSERT_EXIT_MESSAGE(0, "Expected an argument of Vector Type"); - Type *EltTy = cast<VectorType>(VTy)->getElementType(); -@@ -434,7 +434,7 @@ static std::string getMangledTypeStr(Type* Ty) { - Result += "f"; - } - else if (isa<VectorType>(Ty)) -- Result += "v" + utostr(cast<VectorType>(Ty)->getNumElements()) + -+ Result += "v" + utostr(cast<IGCLLVM::FixedVectorType>(Ty)->getNumElements()) + - getMangledTypeStr(cast<VectorType>(Ty)->getElementType()); - else if (Ty) - Result += EVT::getEVT(Ty).getEVTString(); -diff --git a/IGC/LLVM3DBuilder/BuiltinsFrontendDefinitions.hpp b/IGC/LLVM3DBuilder/BuiltinsFrontendDefinitions.hpp -index a89d63b9..a1bc5d2c 100644 ---- a/IGC/LLVM3DBuilder/BuiltinsFrontendDefinitions.hpp -+++ b/IGC/LLVM3DBuilder/BuiltinsFrontendDefinitions.hpp -@@ -4929,7 +4929,7 @@ void LLVM3DBuilder<preserveNames, T, Inserter>::VectorToScalars( - IGC_ASSERT(nullptr != vector->getType()); - IGC_ASSERT(vector->getType()->isVectorTy()); - -- const unsigned count = (unsigned)llvm::cast<llvm::VectorType>(vector->getType())->getNumElements(); -+ const unsigned count = (unsigned)llvm::cast<IGCLLVM::FixedVectorType>(vector->getType())->getNumElements(); - IGC_ASSERT(1 < count); - IGC_ASSERT(count <= 4); - IGC_ASSERT(count <= maxSize); -diff --git a/IGC/WrapperLLVM/include/llvmWrapper/Support/TypeSize.h b/IGC/WrapperLLVM/include/llvmWrapper/Support/TypeSize.h -index 08e5d6cf..30e29720 100644 ---- a/IGC/WrapperLLVM/include/llvmWrapper/Support/TypeSize.h -+++ b/IGC/WrapperLLVM/include/llvmWrapper/Support/TypeSize.h -@@ -35,10 +35,16 @@ using namespace llvm; - namespace IGCLLVM { - #if LLVM_VERSION_MAJOR < 11 - inline unsigned getElementCount(unsigned EC) { return EC; } --#else -+#elif LLVM_VERSION_MAJOR == 11 - inline ElementCount getElementCount(unsigned EC) { - return ElementCount(EC, false); - } -+#elif LLVM_VERSION_MAJOR == 12 -+inline ElementCount getElementCount(unsigned EC) { -+ return ElementCount::get(EC, false); -+} -+#else -+#error "unsupported LLVM version" - #endif - } // namespace IGCLLVM - -diff --git a/IGC/WrapperLLVM/include/llvmWrapper/Transforms/Scalar.h b/IGC/WrapperLLVM/include/llvmWrapper/Transforms/Scalar.h -index f6e54fb1..5266ee87 100644 ---- a/IGC/WrapperLLVM/include/llvmWrapper/Transforms/Scalar.h -+++ b/IGC/WrapperLLVM/include/llvmWrapper/Transforms/Scalar.h -@@ -44,7 +44,7 @@ namespace IGCLLVM - { - return llvm::createLoopUnrollPass(OptLevel, false, Threshold, Count, AllowPartial, Runtime, UpperBound, AllowPeeling); - } --#elif LLVM_VERSION_MAJOR == 9 || LLVM_VERSION_MAJOR == 10 || LLVM_VERSION_MAJOR == 11 -+#elif LLVM_VERSION_MAJOR >= 9 && LLVM_VERSION_MAJOR <= 12 - inline static llvm::Pass * createLoopUnrollPass( - int OptLevel = 2, int Threshold = -1, int Count = -1, - int AllowPartial = -1, int Runtime = -1, -@@ -52,7 +52,7 @@ namespace IGCLLVM - { - return llvm::createLoopUnrollPass(OptLevel, false, false, Threshold, Count, AllowPartial, Runtime, UpperBound, AllowPeeling); - } --#elif LLVM_VERSION_MAJOR >= 12 -+#else - //DO NOT assume same function signature for all incoming llvm versions! Double check to upgrade! - #error Not supported llvm version. - #endif -diff --git a/IGC/common/igc_resourceDimTypes.h b/IGC/common/igc_resourceDimTypes.h -index 6831b06c..d790330f 100644 ---- a/IGC/common/igc_resourceDimTypes.h -+++ b/IGC/common/igc_resourceDimTypes.h -@@ -66,6 +66,11 @@ namespace IGC - resourceDimTypeId == DIM_2D_TYPE || resourceDimTypeId == DIM_2D_ARRAY_TYPE || - resourceDimTypeId == DIM_3D_TYPE || resourceDimTypeId == DIM_CUBE_TYPE || resourceDimTypeId == DIM_CUBE_ARRAY_TYPE)); - -+#if LLVM_VERSION_MAJOR >= 12 -+ llvm::LLVMContext& llvmCtx = module.getContext(); -+ return llvm::StructType::getTypeByName(llvmCtx, ResourceDimensionTypeName[resourceDimTypeId]); -+#else - return module.getTypeByName(ResourceDimensionTypeName[resourceDimTypeId]); -+#endif - } - } -\ No newline at end of file --- -2.17.1 - diff --git a/dynamic-layers/clang-layer/recipes-opencl/igc/files/0001-external-SPIRV-Tools-change-path-to-tools-and-header.patch b/dynamic-layers/clang-layer/recipes-opencl/igc/files/0001-external-SPIRV-Tools-change-path-to-tools-and-header.patch new file mode 100644 index 00000000..dca75e22 --- /dev/null +++ b/dynamic-layers/clang-layer/recipes-opencl/igc/files/0001-external-SPIRV-Tools-change-path-to-tools-and-header.patch @@ -0,0 +1,31 @@ +From e69a3181482e5f442756a61c7b683914072884f1 Mon Sep 17 00:00:00 2001 +From: Anuj Mittal <anuj.mittal@intel.com> +Date: Mon, 9 Jan 2023 11:43:05 +0800 +Subject: [PATCH] external/SPIRV-Tools: change path to tools and headers + +We clone the SPIRV headers and tools in a different directory to ensure +file path substitutions take place. + +Upstream-Status: Inappropriate + +Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> + +--- + external/SPIRV-Tools/CMakeLists.txt | 4 ++-- + 1 file changed, 2 insertions(+), 2 deletions(-) + +diff --git a/external/SPIRV-Tools/CMakeLists.txt b/external/SPIRV-Tools/CMakeLists.txt +index 9afa5746c..7ca24d5dc 100644 +--- a/external/SPIRV-Tools/CMakeLists.txt ++++ b/external/SPIRV-Tools/CMakeLists.txt +@@ -43,8 +43,8 @@ else() #By default use build from sources + message(STATUS "[SPIRV-Tools] : Building from source") + message(STATUS "[SPIRV-Tools] : Current source dir: ${CMAKE_CURRENT_SOURCE_DIR}") + +- set(SPIRV-Headers_SOURCE_DIR "${CMAKE_CURRENT_SOURCE_DIR}/../../SPIRV-Headers") # used in subdirectory +- set(SPIRV-Tools_SOURCE_DIR "${CMAKE_CURRENT_SOURCE_DIR}/../../SPIRV-Tools") ++ set(SPIRV-Headers_SOURCE_DIR "${CMAKE_CURRENT_SOURCE_DIR}/../SPIRV-Headers") # used in subdirectory ++ set(SPIRV-Tools_SOURCE_DIR "${CMAKE_CURRENT_SOURCE_DIR}/../SPIRV-Tools") + + set(SPIRV-Tools_OUTPUT_DIR "${IGC_OPTION__OUTPUT_DIR}/external/SPIRV-Tools/build") + set(IGC_BUILD__SPIRV-Headers_DIR "${SPIRV-Headers_SOURCE_DIR}") diff --git a/dynamic-layers/clang-layer/recipes-opencl/igc/files/0001-fix-tblgen.patch b/dynamic-layers/clang-layer/recipes-opencl/igc/files/0001-fix-tblgen.patch new file mode 100644 index 00000000..39443931 --- /dev/null +++ b/dynamic-layers/clang-layer/recipes-opencl/igc/files/0001-fix-tblgen.patch @@ -0,0 +1,24 @@ +From 5648568e597acd0fed82aac3e6aef0f95a1b78d1 Mon Sep 17 00:00:00 2001 +From: Anuj Mittal <anuj.mittal@intel.com> +Date: Thu, 19 May 2022 22:50:09 +0800 +Subject: [PATCH] fix tblgen + +Upstream-Status: Inappropriate [OE specific] + +--- + IGC/cmake/igc_llvm.cmake | 2 +- + 1 file changed, 1 insertion(+), 1 deletion(-) + +diff --git a/IGC/cmake/igc_llvm.cmake b/IGC/cmake/igc_llvm.cmake +index 10322533c..9020cb3c8 100644 +--- a/IGC/cmake/igc_llvm.cmake ++++ b/IGC/cmake/igc_llvm.cmake +@@ -52,7 +52,7 @@ else() + set(LLVM_OPT_EXE "opt" CACHE STRING "") + + set(LLVM_TABLEGEN_EXE "llvm-tblgen") +- if(CMAKE_CROSSCOMPILING) ++ if(TRUE) + if(DEFINED LLVM_TABLEGEN) + set(LLVM_TABLEGEN_EXE ${LLVM_TABLEGEN}) + else() diff --git a/dynamic-layers/clang-layer/recipes-opencl/igc/files/0001-skip-execution-of-ElfPackager.patch b/dynamic-layers/clang-layer/recipes-opencl/igc/files/0001-skip-execution-of-ElfPackager.patch deleted file mode 100644 index 65342d9c..00000000 --- a/dynamic-layers/clang-layer/recipes-opencl/igc/files/0001-skip-execution-of-ElfPackager.patch +++ /dev/null @@ -1,43 +0,0 @@ -From 5501fab763aa2b11437fbdb19e07a11157a7d3e6 Mon Sep 17 00:00:00 2001 -From: Naveen Saini <naveen.kumar.saini@intel.com> -Date: Thu, 6 Feb 2020 14:56:56 +0800 -Subject: [PATCH] skip execution of ElfPackager - -ElfPackager adds the ability to convert llvm bitcode into elf files for -easier partitioning. Skip for now until we start building a native -version for this. - -Upstream-Status: Inappropriate [configuration specific] - -Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> -Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> ---- - IGC/ElfPackager/CMakeLists.txt | 12 ++++++------ - 1 file changed, 6 insertions(+), 6 deletions(-) - -diff --git a/IGC/ElfPackager/CMakeLists.txt b/IGC/ElfPackager/CMakeLists.txt -index 563ac9d6..c6c74ce9 100644 ---- a/IGC/ElfPackager/CMakeLists.txt -+++ b/IGC/ElfPackager/CMakeLists.txt -@@ -71,12 +71,12 @@ include_directories ( - ${IGC_SOURCE_DIR}/AdaptorOCL/CLElfLib/ - ) - --if(NOT ANDROID) -- add_custom_command(TARGET ${IGC_BUILD__PROJ__ElfPackager} -- POST_BUILD -- COMMAND $<TARGET_FILE:${IGC_BUILD__PROJ__ElfPackager}> -includeSizet -funcList ${CMAKE_CURRENT_SOURCE_DIR}/function_bin.txt ${IGC_BUILD__BIF_DIR}/OCLBiFImpl.bc ${IGC_BUILD__BIF_DIR}/igdclbif.bin -- ) --endif() -+#if(NOT ANDROID) -+# add_custom_command(TARGET ${IGC_BUILD__PROJ__ElfPackager} -+# POST_BUILD -+# COMMAND $<TARGET_FILE:${IGC_BUILD__PROJ__ElfPackager}> -includeSizet -funcList ${CMAKE_CURRENT_SOURCE_DIR}/function_bin.txt ${IGC_BUILD__BIF_DIR}/OCLBiFImpl.bc ${IGC_BUILD__BIF_DIR}/igdclbif.bin -+# ) -+#endif() - - - add_dependencies("${IGC_BUILD__PROJ__ElfPackager}" "${IGC_BUILD__PROJ__BiFModule_OCL}") --- -2.17.1 - diff --git a/dynamic-layers/clang-layer/recipes-opencl/igc/files/0002-Review-fixes-for-LLVM-12-phase-1.patch b/dynamic-layers/clang-layer/recipes-opencl/igc/files/0002-Review-fixes-for-LLVM-12-phase-1.patch deleted file mode 100644 index 6580df75..00000000 --- a/dynamic-layers/clang-layer/recipes-opencl/igc/files/0002-Review-fixes-for-LLVM-12-phase-1.patch +++ /dev/null @@ -1,318 +0,0 @@ -From 60136b453bb3a109bfc88c4040b871af9d522ed5 Mon Sep 17 00:00:00 2001 -From: =?UTF-8?q?Zolt=C3=A1n=20B=C3=B6sz=C3=B6rm=C3=A9nyi?= <zboszor@pr.hu> -Date: Thu, 25 Feb 2021 19:40:21 +0100 -Subject: [PATCH 2/3] Review fixes for LLVM 12 phase 1 -MIME-Version: 1.0 -Content-Type: text/plain; charset=UTF-8 -Content-Transfer-Encoding: 8bit - -Upstream-Status: Pending - -Signed-off-by: Zoltán Böszörményi <zboszor@gmail.com> -Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> ---- - IGC/AdaptorOCL/SPIRV/SPIRVReader.cpp | 6 +----- - IGC/AdaptorOCL/SPIRV/SPIRVUtil.cpp | 6 +----- - IGC/Compiler/CISACodeGen/AdvMemOpt.cpp | 7 ++----- - IGC/Compiler/CISACodeGen/Simd32Profitability.cpp | 7 ++----- - IGC/Compiler/ConvertMSAAPayloadTo16Bit.cpp | 11 +---------- - IGC/Compiler/GenTTI.cpp | 9 ++------- - IGC/Compiler/Legalizer/InstPromoter.cpp | 8 +------- - .../DeviceEnqueueFuncs/TransformBlocks.cpp | 6 +----- - .../OpenCLPasses/WIFuncs/WIFuncResolution.cpp | 4 ---- - IGC/Compiler/Optimizer/Scalarizer.cpp | 2 +- - IGC/WrapperLLVM/include/llvmWrapper/IR/DerivedTypes.h | 10 ++++++++++ - .../include/llvmWrapper/Support/TypeSize.h | 4 +--- - .../include/llvmWrapper/Transforms/Utils/LoopUtils.h | 8 ++++++++ - IGC/common/igc_resourceDimTypes.h | 5 ++--- - 14 files changed, 33 insertions(+), 60 deletions(-) - -diff --git a/IGC/AdaptorOCL/SPIRV/SPIRVReader.cpp b/IGC/AdaptorOCL/SPIRV/SPIRVReader.cpp -index 725a1512..12f42be8 100644 ---- a/IGC/AdaptorOCL/SPIRV/SPIRVReader.cpp -+++ b/IGC/AdaptorOCL/SPIRV/SPIRVReader.cpp -@@ -1929,11 +1929,7 @@ SPIRVToLLVM::transType(SPIRVType *T) { - auto name = isSubgroupAvcINTELTypeOpCode(OC) ? - OCLSubgroupINTELTypeOpCodeMap::rmap(OC) : - BuiltinOpaqueGenericTypeOpCodeMap::rmap(OC); --#if LLVM_VERSION_MAJOR >= 12 -- auto *pST = llvm::StructType::getTypeByName(M->getContext(), name); --#else -- auto *pST = M->getTypeByName(name); --#endif -+ auto *pST = IGCLLVM::getTypeByName(M, name); - pST = pST ? pST : StructType::create(*Context, name); - - return mapType(T, PointerType::get(pST, getOCLOpaqueTypeAddrSpace(OC))); -diff --git a/IGC/AdaptorOCL/SPIRV/SPIRVUtil.cpp b/IGC/AdaptorOCL/SPIRV/SPIRVUtil.cpp -index 91b4623c..f4f0dee5 100644 ---- a/IGC/AdaptorOCL/SPIRV/SPIRVUtil.cpp -+++ b/IGC/AdaptorOCL/SPIRV/SPIRVUtil.cpp -@@ -93,11 +93,7 @@ saveLLVMModule(Module *M, const std::string &OutputFile) { - PointerType* - getOrCreateOpaquePtrType(Module *M, const std::string &Name, - unsigned AddrSpace) { --#if LLVM_VERSION_MAJOR >= 12 -- auto OpaqueType = llvm::StructType::getTypeByName(M->getContext(), Name); --#else -- auto OpaqueType = M->getTypeByName(Name); --#endif -+ auto OpaqueType = IGCLLVM::getTypeByName(M, Name); - if (!OpaqueType) - OpaqueType = StructType::create(M->getContext(), Name); - return PointerType::get(OpaqueType, AddrSpace); -diff --git a/IGC/Compiler/CISACodeGen/AdvMemOpt.cpp b/IGC/Compiler/CISACodeGen/AdvMemOpt.cpp -index fc45a510..a612a473 100644 ---- a/IGC/Compiler/CISACodeGen/AdvMemOpt.cpp -+++ b/IGC/Compiler/CISACodeGen/AdvMemOpt.cpp -@@ -33,6 +33,7 @@ IN THE SOFTWARE. - #include <llvm/Support/Debug.h> - #include <llvm/Support/raw_ostream.h> - #include <llvm/Transforms/Utils/Local.h> -+#include "llvmWrapper/Transforms/Utils/LoopUtils.h" - #include "common/LLVMWarningsPop.hpp" - #include "GenISAIntrinsics/GenIntrinsics.h" - #include "Compiler/CISACodeGen/ShaderCodeGen.hpp" -@@ -134,11 +135,7 @@ bool AdvMemOpt::runOnFunction(Function& F) { - for (auto I = LI->begin(), E = LI->end(); I != E; ++I) - for (auto DFI = df_begin(*I), DFE = df_end(*I); DFI != DFE; ++DFI) { - Loop* L = *DFI; --#if LLVM_VERSION_MAJOR >= 12 -- if (L->isInnermost()) --#else -- if (L->empty()) --#endif -+ if (IGCLLVM::isInnermost(L)) - InnermostLoops.push_back(L); - } - -diff --git a/IGC/Compiler/CISACodeGen/Simd32Profitability.cpp b/IGC/Compiler/CISACodeGen/Simd32Profitability.cpp -index c1f4a419..5393d4e8 100644 ---- a/IGC/Compiler/CISACodeGen/Simd32Profitability.cpp -+++ b/IGC/Compiler/CISACodeGen/Simd32Profitability.cpp -@@ -28,6 +28,7 @@ IN THE SOFTWARE. - #include "Compiler/CISACodeGen/Platform.hpp" - #include "common/LLVMWarningsPush.hpp" - #include <llvmWrapper/IR/DerivedTypes.h> -+#include <llvmWrapper/Transforms/Utils/LoopUtils.h> - #include <llvm/IR/InstIterator.h> - #include <llvm/IR/Operator.h> - #include <llvmWrapper/IR/DerivedTypes.h> -@@ -995,11 +996,7 @@ static bool hasLongStridedLdStInLoop(Function* F, LoopInfo* LI, WIAnalysis* WI) - // Collect innermost simple loop. - for (auto I = LI->begin(), E = LI->end(); I != E; ++I) { - auto L = *I; --#if LLVM_VERSION_MAJOR >= 12 -- if (!L->isInnermost()) --#else -- if (!L->empty()) --#endif -+ if (!IGCLLVM::isInnermost(L)) - continue; - if (L->getNumBlocks() != 2) - continue; -diff --git a/IGC/Compiler/ConvertMSAAPayloadTo16Bit.cpp b/IGC/Compiler/ConvertMSAAPayloadTo16Bit.cpp -index adf992cb..33473c23 100644 ---- a/IGC/Compiler/ConvertMSAAPayloadTo16Bit.cpp -+++ b/IGC/Compiler/ConvertMSAAPayloadTo16Bit.cpp -@@ -153,21 +153,12 @@ void ConvertMSAAPayloadTo16Bit::visitCallInst(CallInst& I) - // In OGL there are uses of ldmcs other then ldms, using vec4float type. - // Fix them to use newly created 16bit ldmcs. - if (ldmcs->getType()->isVectorTy() && --#if LLVM_VERSION_MAJOR >= 12 -- ldmcs->getType()->getScalarType()->isFloatTy()) --#else -- ldmcs->getType()->getVectorElementType()->isFloatTy()) --#endif -+ cast<IGCLLVM::FixedVectorType>(ldmcs->getType())->getElementType()->isFloatTy()) - { - m_builder->SetInsertPoint(ldmcs); - --#if LLVM_VERSION_MAJOR >= 12 - uint ldmcsNumOfElements = cast<IGCLLVM::FixedVectorType>(ldmcs->getType())->getNumElements(); - uint new_mcs_callNumOfElements = cast<IGCLLVM::FixedVectorType>(new_mcs_call->getType())->getNumElements(); --#else -- uint ldmcsNumOfElements = ldmcs->getType()->getVectorNumElements(); -- uint new_mcs_callNumOfElements = new_mcs_call->getType()->getVectorNumElements(); --#endif - - // vec of 16bit ints to vec of 32bit ints - Type* new_mcs_callVecType = IGCLLVM::FixedVectorType::get(m_builder->getInt32Ty(), new_mcs_callNumOfElements); -diff --git a/IGC/Compiler/GenTTI.cpp b/IGC/Compiler/GenTTI.cpp -index 9e4d2f26..53e3ec9e 100644 ---- a/IGC/Compiler/GenTTI.cpp -+++ b/IGC/Compiler/GenTTI.cpp -@@ -37,6 +37,7 @@ IN THE SOFTWARE. - #include "llvm/Analysis/LoopInfo.h" - #include "llvm/Analysis/ScalarEvolution.h" - #include "llvm/Support/raw_ostream.h" -+#include "llvmWrapper/Transforms/Utils/LoopUtils.h" - #include "common/LLVMWarningsPop.hpp" - - using namespace llvm; -@@ -216,13 +217,7 @@ namespace llvm { - - // Skip non-simple loop. - if (L->getNumBlocks() != 1) { -- if (IGC_IS_FLAG_ENABLED(EnableAdvRuntimeUnroll) && --#if LLVM_VERSION_MAJOR >= 12 -- L->isInnermost() --#else -- L->empty() --#endif -- ) { -+ if (IGC_IS_FLAG_ENABLED(EnableAdvRuntimeUnroll) && IGCLLVM::isInnermost(L)) { - auto countNonPHI = [](BasicBlock* BB) { - unsigned Total = BB->size(); - unsigned PHIs = 0; -diff --git a/IGC/Compiler/Legalizer/InstPromoter.cpp b/IGC/Compiler/Legalizer/InstPromoter.cpp -index 8fadf89f..63cbccb5 100644 ---- a/IGC/Compiler/Legalizer/InstPromoter.cpp -+++ b/IGC/Compiler/Legalizer/InstPromoter.cpp -@@ -398,13 +398,7 @@ bool InstPromoter::visitBitCastInst(BitCastInst& I) { - IRB->CreateBitCast(Val, IGCLLVM::FixedVectorType::get(DestTy->getScalarType(), N)); - - std::vector<Constant*> Vals; -- for (unsigned i = 0; --#if LLVM_VERSION_MAJOR >= 12 -- i < cast<IGCLLVM::FixedVectorType>(DestTy)->getNumElements(); --#else -- i < DestTy->getVectorNumElements(); --#endif -- i++) -+ for (unsigned i = 0; i < cast<IGCLLVM::FixedVectorType>(DestTy)->getNumElements(); i++) - Vals.push_back(IRB->getInt32(i)); - - Value* Mask = ConstantVector::get(Vals); -diff --git a/IGC/Compiler/Optimizer/OpenCLPasses/DeviceEnqueueFuncs/TransformBlocks.cpp b/IGC/Compiler/Optimizer/OpenCLPasses/DeviceEnqueueFuncs/TransformBlocks.cpp -index 119520ed..a3681b79 100644 ---- a/IGC/Compiler/Optimizer/OpenCLPasses/DeviceEnqueueFuncs/TransformBlocks.cpp -+++ b/IGC/Compiler/Optimizer/OpenCLPasses/DeviceEnqueueFuncs/TransformBlocks.cpp -@@ -952,11 +952,7 @@ namespace //Anonymous - { - auto ndrangeStructName = "struct.ndrange_t"; - auto module = _deviceExecCall->getModule(); --#if LLVM_VERSION_MAJOR >= 12 -- auto ndrangeTy = llvm::StructType::getTypeByName(module->getContext(), ndrangeStructName); --#else -- auto ndrangeTy = module->getTypeByName(ndrangeStructName); --#endif -+ auto ndrangeTy = IGCLLVM::getTypeByName(module, ndrangeStructName); - if (ndrangeTy == nullptr) - { - //create struct type -diff --git a/IGC/Compiler/Optimizer/OpenCLPasses/WIFuncs/WIFuncResolution.cpp b/IGC/Compiler/Optimizer/OpenCLPasses/WIFuncs/WIFuncResolution.cpp -index 535d6268..c23c661d 100644 ---- a/IGC/Compiler/Optimizer/OpenCLPasses/WIFuncs/WIFuncResolution.cpp -+++ b/IGC/Compiler/Optimizer/OpenCLPasses/WIFuncs/WIFuncResolution.cpp -@@ -303,11 +303,7 @@ static Value* BuildLoadInst(CallInst& CI, unsigned int Offset, Type* DataType) - auto Size = ElemByteSize; - if (DataType->isVectorTy()) - { --#if LLVM_VERSION_MAJOR >= 12 - Size *= cast<IGCLLVM::FixedVectorType>(DataType)->getNumElements(); --#else -- Size *= DataType->getVectorNumElements(); --#endif - } - unsigned int AlignedOffset = (Offset / ElemByteSize) * ElemByteSize; - unsigned int LoadByteSize = (Offset == AlignedOffset) ? Size : Size * 2; -diff --git a/IGC/Compiler/Optimizer/Scalarizer.cpp b/IGC/Compiler/Optimizer/Scalarizer.cpp -index a4e73a6d..38627553 100644 ---- a/IGC/Compiler/Optimizer/Scalarizer.cpp -+++ b/IGC/Compiler/Optimizer/Scalarizer.cpp -@@ -778,7 +778,7 @@ void ScalarizeFunction::scalarizeInstruction(ShuffleVectorInst* SI) - - // Generate array for shuffled scalar values - SmallVector<Value*, MAX_INPUT_VECTOR_WIDTH>newVector; -- unsigned width = int_cast<unsigned>(dyn_cast<IGCLLVM::FixedVectorType>(SI->getType())->getNumElements()); -+ unsigned width = int_cast<unsigned>(cast<IGCLLVM::FixedVectorType>(SI->getType())->getNumElements()); - - // Generate undef value, which may be needed as some scalar elements - UndefValue* undef = UndefValue::get(inputType->getElementType()); -diff --git a/IGC/WrapperLLVM/include/llvmWrapper/IR/DerivedTypes.h b/IGC/WrapperLLVM/include/llvmWrapper/IR/DerivedTypes.h -index a3f5a0b8..6a5407bb 100644 ---- a/IGC/WrapperLLVM/include/llvmWrapper/IR/DerivedTypes.h -+++ b/IGC/WrapperLLVM/include/llvmWrapper/IR/DerivedTypes.h -@@ -29,6 +29,7 @@ SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. - - #include "llvm/Config/llvm-config.h" - #include "llvm/IR/DerivedTypes.h" -+#include "llvm/IR/Module.h" - - namespace IGCLLVM - { -@@ -62,6 +63,15 @@ namespace IGCLLVM - return false; - #endif - } -+ -+ inline llvm::StructType *getTypeByName(llvm::Module *M, llvm::StringRef Name) { -+#if LLVM_VERSION_MAJOR >= 12 -+ return llvm::StructType::getTypeByName(M->getContext(), Name); -+#else -+ return M->getTypeByName(Name); -+#endif -+ } -+ - } - - #endif -diff --git a/IGC/WrapperLLVM/include/llvmWrapper/Support/TypeSize.h b/IGC/WrapperLLVM/include/llvmWrapper/Support/TypeSize.h -index 30e29720..7021820c 100644 ---- a/IGC/WrapperLLVM/include/llvmWrapper/Support/TypeSize.h -+++ b/IGC/WrapperLLVM/include/llvmWrapper/Support/TypeSize.h -@@ -39,12 +39,10 @@ inline unsigned getElementCount(unsigned EC) { return EC; } - inline ElementCount getElementCount(unsigned EC) { - return ElementCount(EC, false); - } --#elif LLVM_VERSION_MAJOR == 12 -+#else - inline ElementCount getElementCount(unsigned EC) { - return ElementCount::get(EC, false); - } --#else --#error "unsupported LLVM version" - #endif - } // namespace IGCLLVM - -diff --git a/IGC/WrapperLLVM/include/llvmWrapper/Transforms/Utils/LoopUtils.h b/IGC/WrapperLLVM/include/llvmWrapper/Transforms/Utils/LoopUtils.h -index db47b00b..bce9cfc1 100644 ---- a/IGC/WrapperLLVM/include/llvmWrapper/Transforms/Utils/LoopUtils.h -+++ b/IGC/WrapperLLVM/include/llvmWrapper/Transforms/Utils/LoopUtils.h -@@ -41,6 +41,14 @@ namespace IGCLLVM - return llvm::InsertPreheaderForLoop(L, DT, LI, nullptr, PreserveLCSSA); - } - #endif -+ -+ inline bool isInnermost(llvm::Loop *L) { -+#if LLVM_VERSION_MAJOR >= 12 -+ return L->isInnermost(); -+#else -+ return L->empty(); -+#endif -+ } - } - - #endif -diff --git a/IGC/common/igc_resourceDimTypes.h b/IGC/common/igc_resourceDimTypes.h -index d790330f..2d675969 100644 ---- a/IGC/common/igc_resourceDimTypes.h -+++ b/IGC/common/igc_resourceDimTypes.h -@@ -67,10 +67,9 @@ namespace IGC - resourceDimTypeId == DIM_3D_TYPE || resourceDimTypeId == DIM_CUBE_TYPE || resourceDimTypeId == DIM_CUBE_ARRAY_TYPE)); - - #if LLVM_VERSION_MAJOR >= 12 -- llvm::LLVMContext& llvmCtx = module.getContext(); -- return llvm::StructType::getTypeByName(llvmCtx, ResourceDimensionTypeName[resourceDimTypeId]); -+ return llvm::StructType::getTypeByName(module.getContext(), ResourceDimensionTypeName[resourceDimTypeId]); - #else - return module.getTypeByName(ResourceDimensionTypeName[resourceDimTypeId]); - #endif - } --} -\ No newline at end of file -+} --- -2.17.1 - diff --git a/dynamic-layers/clang-layer/recipes-opencl/igc/files/improve_src_package_reproducibility.patch b/dynamic-layers/clang-layer/recipes-opencl/igc/files/0003-Improve-Reproducibility-for-src-package.patch index c694836f..650130a8 100644 --- a/dynamic-layers/clang-layer/recipes-opencl/igc/files/improve_src_package_reproducibility.patch +++ b/dynamic-layers/clang-layer/recipes-opencl/igc/files/0003-Improve-Reproducibility-for-src-package.patch @@ -1,4 +1,4 @@ -From 2fa3f2da1179bd5b2eeac82839d386f0111e611c Mon Sep 17 00:00:00 2001 +From 0559332abd04b6c8bc70171d201f43d2e4735336 Mon Sep 17 00:00:00 2001 From: Lee Chee Yang <chee.yang.lee@intel.com> Date: Wed, 2 Sep 2020 08:28:35 +0800 Subject: [PATCH] Improve Reproducibility for src package @@ -9,15 +9,16 @@ this only works on bison 3.7 onward, hence check for bison version before adding the flags. Upstream-Status: Inappropriate [applying --file-prefix-map in such way does not work for upstream] Signed-off-by: Lee Chee Yang <chee.yang.lee@intel.com> + --- visa/CMakeLists.txt | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/visa/CMakeLists.txt b/visa/CMakeLists.txt -index 981c35d2..d5944370 100644 +index 6be467587..930c386a6 100644 --- a/visa/CMakeLists.txt +++ b/visa/CMakeLists.txt -@@ -109,8 +109,11 @@ endif() +@@ -123,8 +123,11 @@ endif() set(bison_output_file ${CMAKE_CURRENT_BINARY_DIR}/CISA.tab.cpp) set(flex_output_file ${CMAKE_CURRENT_BINARY_DIR}/lex.CISA.cpp) @@ -31,6 +32,3 @@ index 981c35d2..d5944370 100644 ADD_FLEX_BISON_DEPENDENCY(CISAScanner CISAParser) set(CISAScanner_dependencies) --- -2.28.0 - diff --git a/dynamic-layers/clang-layer/recipes-opencl/igc/files/0003-Review-fixes-for-LLVM-12-phase-2.patch b/dynamic-layers/clang-layer/recipes-opencl/igc/files/0003-Review-fixes-for-LLVM-12-phase-2.patch deleted file mode 100644 index 189c2b3f..00000000 --- a/dynamic-layers/clang-layer/recipes-opencl/igc/files/0003-Review-fixes-for-LLVM-12-phase-2.patch +++ /dev/null @@ -1,123 +0,0 @@ -From c6d333637537263930acb1b6c5dadb0467d745f6 Mon Sep 17 00:00:00 2001 -From: =?UTF-8?q?Zolt=C3=A1n=20B=C3=B6sz=C3=B6rm=C3=A9nyi?= <zboszor@pr.hu> -Date: Fri, 26 Feb 2021 06:39:35 +0100 -Subject: [PATCH 3/3] Review fixes for LLVM 12 phase 2 -MIME-Version: 1.0 -Content-Type: text/plain; charset=UTF-8 -Content-Transfer-Encoding: 8bit - -Upstream-Status: Pending - -Signed-off-by: Zoltán Böszörményi <zboszor@gmail.com> -Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> ---- - IGC/AdaptorOCL/SPIRV/SPIRVReader.cpp | 4 ---- - .../AddressSpaceAliasAnalysis.cpp | 10 +++++----- - .../PrivateMemory/PrivateMemoryResolution.cpp | 4 ---- - IGC/DebugInfo/DebugInfoUtils.hpp | 4 ---- - IGC/DebugInfo/DwarfDebug.cpp | 8 -------- - 5 files changed, 5 insertions(+), 25 deletions(-) - -diff --git a/IGC/AdaptorOCL/SPIRV/SPIRVReader.cpp b/IGC/AdaptorOCL/SPIRV/SPIRVReader.cpp -index 12f42be8..c4f9d1ea 100644 ---- a/IGC/AdaptorOCL/SPIRV/SPIRVReader.cpp -+++ b/IGC/AdaptorOCL/SPIRV/SPIRVReader.cpp -@@ -1576,11 +1576,7 @@ void SPIRVToLLVMDbgTran::transDbgInfo(SPIRVValue *SV, Value *V) { - Line->getColumn(), scope, iat); - - if(scope && !isa<DIFile>(scope)) --#if LLVM_VERSION_MAJOR >= 12 - I->setDebugLoc(DILocation::get(scope->getContext(), Line->getLine(), Line->getColumn(), --#else -- I->setDebugLoc(DebugLoc::get(Line->getLine(), Line->getColumn(), --#endif - scope, iat)); - } - } -diff --git a/IGC/Compiler/Optimizer/OpenCLPasses/AddressSpaceAliasAnalysis/AddressSpaceAliasAnalysis.cpp b/IGC/Compiler/Optimizer/OpenCLPasses/AddressSpaceAliasAnalysis/AddressSpaceAliasAnalysis.cpp -index e9c07b34..b6b779da 100644 ---- a/IGC/Compiler/Optimizer/OpenCLPasses/AddressSpaceAliasAnalysis/AddressSpaceAliasAnalysis.cpp -+++ b/IGC/Compiler/Optimizer/OpenCLPasses/AddressSpaceAliasAnalysis/AddressSpaceAliasAnalysis.cpp -@@ -23,8 +23,7 @@ IN THE SOFTWARE. - ============================= end_copyright_notice ===========================*/ - - #include "llvm/Config/llvm-config.h" --#include "llvmWrapper/IR/DerivedTypes.h" --#include "llvmWrapper/Analysis/TargetLibraryInfo.h" -+#include <llvm/Analysis/TargetLibraryInfo.h> - #include "Compiler/Optimizer/OpenCLPasses/AddressSpaceAliasAnalysis/AddressSpaceAliasAnalysis.h" - #include "Compiler/CodeGenPublic.h" - #include "Compiler/IGCPassSupport.h" -@@ -180,11 +179,12 @@ namespace { - bool doInitialization(Module& M) override { - if(M.size() > 0) - { -+ Result.reset(new AddressSpaceAAResult( -+ getAnalysis<TargetLibraryInfoWrapperPass>().getTLI( - #if LLVM_VERSION_MAJOR >= 10 -- Function &F = *M.begin(); -+ *M.begin() - #endif -- Result.reset(new AddressSpaceAAResult( -- getAnalysis<TargetLibraryInfoWrapperPass>().getTLI(), -+ ), - *getAnalysis<CodeGenContextWrapper>().getCodeGenContext())); - } - return false; -diff --git a/IGC/Compiler/Optimizer/OpenCLPasses/PrivateMemory/PrivateMemoryResolution.cpp b/IGC/Compiler/Optimizer/OpenCLPasses/PrivateMemory/PrivateMemoryResolution.cpp -index 07f85f4c..98ea616f 100644 ---- a/IGC/Compiler/Optimizer/OpenCLPasses/PrivateMemory/PrivateMemoryResolution.cpp -+++ b/IGC/Compiler/Optimizer/OpenCLPasses/PrivateMemory/PrivateMemoryResolution.cpp -@@ -816,11 +816,7 @@ bool PrivateMemoryResolution::resolveAllocaInstructions(bool privateOnStack) - // Construct an empty DebugLoc. - IF_DEBUG_INFO(DebugLoc entryDebugLoc); - // Assign with the function location if available. --#if LLVM_VERSION_MAJOR >= 12 - IF_DEBUG_INFO_IF(DISubprogram *subprogram = m_currFunction->getSubprogram(), entryDebugLoc = DILocation::get(subprogram->getContext(), subprogram->getLine(), 0, subprogram);); --#else -- IF_DEBUG_INFO_IF(DISubprogram *subprogram = m_currFunction->getSubprogram(), entryDebugLoc = DebugLoc::get(subprogram->getLine(), 0, subprogram);); --#endif - IF_DEBUG_INFO(entryBuilder.SetCurrentDebugLocation(entryDebugLoc)); - - if (privateOnStack) -diff --git a/IGC/DebugInfo/DebugInfoUtils.hpp b/IGC/DebugInfo/DebugInfoUtils.hpp -index b77a550d..88b30a75 100644 ---- a/IGC/DebugInfo/DebugInfoUtils.hpp -+++ b/IGC/DebugInfo/DebugInfoUtils.hpp -@@ -108,11 +108,7 @@ namespace IGC - IGCLLVM::DIBuilder Builder(M); - llvm::DIGlobalVariable* GV = GVs[j]->getVariable(); - llvm::DIScope* scopeToUse = GV->getScope(); --#if LLVM_VERSION_MAJOR >= 12 - llvm::DILocation* locToUse = llvm::DILocation::get(scopeToUse->getContext(), GV->getLine(), 0, scopeToUse, loc); --#else -- llvm::DILocation* locToUse = llvm::DebugLoc::get(GV->getLine(), 0, scopeToUse, loc); --#endif - if (llvm::isa<llvm::DICompileUnit>(GV->getScope())) - { - // Function has no DebugLoc so it is either internal -diff --git a/IGC/DebugInfo/DwarfDebug.cpp b/IGC/DebugInfo/DwarfDebug.cpp -index bd9f17b7..3d9f0835 100644 ---- a/IGC/DebugInfo/DwarfDebug.cpp -+++ b/IGC/DebugInfo/DwarfDebug.cpp -@@ -2102,17 +2102,9 @@ static DebugLoc getFnDebugLoc(DebugLoc DL, const LLVMContext& Ctx) - // Check for number of operands since the compatibility is cheap here. - if (SP->getNumOperands() > 19) - { --#if LLVM_VERSION_MAJOR >= 12 - return DILocation::get(SP->getContext(), SP->getScopeLine(), 0, SP); --#else -- return DebugLoc::get(SP->getScopeLine(), 0, SP); --#endif - } --#if LLVM_VERSION_MAJOR >= 12 - return DILocation::get(SP->getContext(), SP->getLine(), 0, SP); --#else -- return DebugLoc::get(SP->getLine(), 0, SP); --#endif - } - - return DebugLoc(); --- -2.17.1 - diff --git a/dynamic-layers/clang-layer/recipes-opencl/igc/files/link-to-LLVMGenXIntrinsics.patch b/dynamic-layers/clang-layer/recipes-opencl/igc/files/link-to-LLVMGenXIntrinsics.patch deleted file mode 100644 index aead2d1c..00000000 --- a/dynamic-layers/clang-layer/recipes-opencl/igc/files/link-to-LLVMGenXIntrinsics.patch +++ /dev/null @@ -1,71 +0,0 @@ -From 11b923c99cbe3580885ce40c322277fc823107a0 Mon Sep 17 00:00:00 2001 -From: Naveen Saini <naveen.kumar.saini@intel.com> -Date: Tue, 2 Feb 2021 13:39:53 +0800 -Subject: [PATCH] IGC/VectorCompiler/CMakeLists.txt: link to external - LLVMGenXIntrinsics - -By default LLVMGenXIntrinsics is to be build In-tree, but we want to -link externally. - -Fix llvm-tblgen path. - -Upstream-Status: Inappropriate [configuration specific] - -Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> ---- - IGC/VectorCompiler/CMakeLists.txt | 32 ++----------------------------- - 1 file changed, 2 insertions(+), 30 deletions(-) - -diff --git a/IGC/VectorCompiler/CMakeLists.txt b/IGC/VectorCompiler/CMakeLists.txt -index 86f343ee..b4268160 100644 ---- a/IGC/VectorCompiler/CMakeLists.txt -+++ b/IGC/VectorCompiler/CMakeLists.txt -@@ -53,10 +53,7 @@ if(IGC_BUILD__USING_SYSTEM_LLVM OR (WIN32 AND LLVM_USE_PREBUILT)) - message(STATUS "[VC] Using system llvm") - - # Need to search for llvm-tblgen -- find_program(LLVM_TABLEGEN_EXE "llvm-tblgen" -- ${LLVM_TOOLS_BINARY_DIR} -- NO_DEFAULT_PATH -- ) -+ find_program(LLVM_TABLEGEN_EXE "llvm-tblgen") - if(LLVM_TABLEGEN_EXE-NOTFOUND) - message(FATAL_ERROR "[VC] llvm-tblgen is not found") - endif() -@@ -140,32 +137,7 @@ if(LLVM_ON_WIN32) - add_compile_options(/wd4141) - endif() - --if(DEFINED VC_INTRINSICS_SRC) -- set(INTRSRC "${VC_INTRINSICS_SRC}/GenXIntrinsics") --endif() -- --if(NOT DEFINED INTRSRC) -- set(INTRSRC "${CMAKE_CURRENT_SOURCE_DIR}/../../../vc-intrinsics/GenXIntrinsics") --endif() -- --message(STATUS "[VC] Using vc-intrinsics source from: ${INTRSRC}") -- --# We are using prebuilt SPIRV and building intrinsics. --set(INTRBUILD "${CMAKE_CURRENT_BINARY_DIR}/intrbuild") --add_subdirectory(${INTRSRC} ${INTRBUILD}) --# Make separate target for intrinstics headers. --# Ideally, this should be inside intrinsics, but --# this breaks in-tree build of intrinsics with LLVM. --add_library(VCIntrinsicsHeaders INTERFACE) --target_include_directories(VCIntrinsicsHeaders -- INTERFACE -- ${INTRSRC}/include -- ${INTRBUILD}/include -- ) --add_dependencies(VCIntrinsicsHeaders -- GenXIntrinsics -- ) -- -+find_package(LLVMGenXIntrinsics REQUIRED) - include(cmake/spirv.cmake) - include(${IGC_SOURCE_DIR}/cmake/utils.cmake) - --- -2.17.1 - diff --git a/dynamic-layers/clang-layer/recipes-opencl/igc/intel-graphics-compiler_1.0.15985.7.bb b/dynamic-layers/clang-layer/recipes-opencl/igc/intel-graphics-compiler_1.0.15985.7.bb new file mode 100644 index 00000000..96d828b6 --- /dev/null +++ b/dynamic-layers/clang-layer/recipes-opencl/igc/intel-graphics-compiler_1.0.15985.7.bb @@ -0,0 +1,75 @@ +SUMMARY = "The Intel(R) Graphics Compiler for OpenCL(TM)" +DESCRIPTION = "The Intel(R) Graphics Compiler for OpenCL(TM) is an \ +llvm based compiler for OpenCL(TM) targeting Intel Gen graphics \ +hardware architecture." + +LICENSE = "MIT & Apache-2.0" +LIC_FILES_CHKSUM = "file://IGC/BiFModule/Implementation/ExternalLibraries/libclc/LICENSE.TXT;md5=311cfc1a5b54bab8ed34a0b5fba4373e \ + file://LICENSE.md;md5=488d74376edf2765f6e78d271543dde3 \ + file://NOTICES.txt;md5=7f4fbc3eb2c34807465e63b1ec3c9d1a" + +SRC_URI = "git://github.com/intel/intel-graphics-compiler.git;protocol=https;name=igc;branch=releases/igc-1.0.15985 \ + git://github.com/intel/vc-intrinsics.git;protocol=https;destsuffix=git/vc-intrinsics;name=vc;nobranch=1 \ + git://github.com/KhronosGroup/SPIRV-Tools.git;protocol=https;destsuffix=git/SPIRV-Tools;name=spirv-tools;branch=main \ + git://github.com/KhronosGroup/SPIRV-Headers.git;protocol=https;destsuffix=git/SPIRV-Headers;name=spirv-headers;branch=main \ + file://0003-Improve-Reproducibility-for-src-package.patch \ + file://0001-BiF-CMakeLists.txt-remove-opt-from-DEPENDS.patch \ + file://0001-external-SPIRV-Tools-change-path-to-tools-and-header.patch \ + " + +SRC_URI:append:class-native = " file://0001-fix-tblgen.patch" + +SRCREV_igc = "6cc111d262e1c3abcf4bc6b6d8a589ebf821a5c0" +SRCREV_vc = "da892e1982b6c25b9a133f85b4ac97142d8a3def" +SRCREV_spirv-tools = "f0cc85efdbbe3a46eae90e0f915dc1509836d0fc" +SRCREV_spirv-headers = "cca08c63cefa129d082abca0302adcb81610b465" + +SRCREV_FORMAT = "igc_vc_spirv-tools_spirv-headers" + +# Used to replace with relative path in reproducibility patch +export B + +S = "${WORKDIR}/git" + +inherit cmake pkgconfig qemu python3native + +CXXFLAGS:append = " -Wno-error=nonnull" + +COMPATIBLE_HOST = '(x86_64).*-linux' +COMPATIBLE_HOST:libc-musl = "null" + +DEPENDS += " flex-native bison-native clang clang-cross-x86_64 opencl-clang qemu-native python3-mako-native" + +RDEPENDS:${PN} += "opencl-clang" + +PACKAGECONFIG ??= "vc" +PACKAGECONFIG[vc] = "-DIGC_BUILD__VC_ENABLED=ON -DIGC_OPTION__LINK_KHRONOS_SPIRV_TRANSLATOR=ON -DIGC_OPTION__SPIRV_TRANSLATOR_MODE=Prebuilds,-DIGC_BUILD__VC_ENABLED=OFF," + +EXTRA_OECMAKE = " \ + -DIGC_OPTION__LLVM_PREFERRED_VERSION=${LLVMVERSION} \ + -DVC_INTRINSICS_SRC="${S}/vc-intrinsics" \ + -DIGC_OPTION__LLVM_MODE=Prebuilds \ + -DLLVM_TABLEGEN=${STAGING_BINDIR_NATIVE}/llvm-tblgen \ + -DLLVM_LINK_EXE=${STAGING_BINDIR_NATIVE}/llvm-link \ + -DCLANG_EXE=${STAGING_BINDIR_NATIVE}/clang \ + -DCMAKE_CROSSCOMPILING_EMULATOR=${WORKDIR}/qemuwrapper \ + " + +do_configure:prepend:class-target () { + # Write out a qemu wrapper that will be used by cmake. + qemu_binary="${@qemu_wrapper_cmdline(d, d.getVar('STAGING_DIR_HOST'), [d.expand('${STAGING_DIR_HOST}${libdir}'),d.expand('${STAGING_DIR_HOST}${base_libdir}')])}" + cat > ${WORKDIR}/qemuwrapper << EOF +#!/bin/sh +$qemu_binary "\$@" +EOF + chmod +x ${WORKDIR}/qemuwrapper +} + +UPSTREAM_CHECK_GITTAGREGEX = "^igc-(?P<pver>(?!19\..*)\d+(\.\d+)+)$" + +FILES:${PN} += " \ + ${libdir}/igc/NOTICES.txt \ + " + +# libigc.so contains buildpaths +INSANE_SKIP:${PN} += "buildpaths" diff --git a/dynamic-layers/clang-layer/recipes-opencl/igc/intel-graphics-compiler_1.0.6646.bb b/dynamic-layers/clang-layer/recipes-opencl/igc/intel-graphics-compiler_1.0.6646.bb deleted file mode 100644 index 582db553..00000000 --- a/dynamic-layers/clang-layer/recipes-opencl/igc/intel-graphics-compiler_1.0.6646.bb +++ /dev/null @@ -1,45 +0,0 @@ -SUMMARY = "The Intel(R) Graphics Compiler for OpenCL(TM)" -DESCRIPTION = "The Intel(R) Graphics Compiler for OpenCL(TM) is an \ -llvm based compiler for OpenCL(TM) targeting Intel Gen graphics \ -hardware architecture." - -LICENSE = "MIT & BSD-3-Clause" -LIC_FILES_CHKSUM = "file://IGC/BiFModule/Implementation/ExternalLibraries/libclc/LICENSE.TXT;md5=311cfc1a5b54bab8ed34a0b5fba4373e \ - file://IGC/Compiler/LegalizationPass.cpp;beginline=1;endline=23;md5=8b19c5999abc484f18232b0905367f9f \ - file://NOTICES.txt;md5=b12e73994de4fbe0f688cf0bc91512a0" - -SRC_URI = "git://github.com/intel/intel-graphics-compiler.git;protocol=https; \ - file://0001-skip-execution-of-ElfPackager.patch \ - file://link-to-LLVMGenXIntrinsics.patch \ - file://improve_src_package_reproducibility.patch \ - file://0001-Fix-build-with-LLVM-12.patch \ - file://0002-Review-fixes-for-LLVM-12-phase-1.patch \ - file://0003-Review-fixes-for-LLVM-12-phase-2.patch \ - " - -SRCREV = "535aaaef03ce338e05e6162118082e6e007e8c10" - -# Used to replace with relative path in reproducibility patch -export B - -S = "${WORKDIR}/git" - -inherit cmake - -COMPATIBLE_HOST = '(x86_64).*-linux' -COMPATIBLE_HOST_libc-musl = "null" - -DEPENDS += " flex-native bison-native clang opencl-clang vc-intrinsics" -DEPENDS_append_class-target = " clang-cross-x86_64" - -RDEPENDS_${PN} += "opencl-clang" - -EXTRA_OECMAKE = "-DIGC_OPTION__LLVM_PREFERRED_VERSION=${LLVMVERSION} -DPYTHON_EXECUTABLE=${HOSTTOOLS_DIR}/python3 -DIGC_BUILD__VC_ENABLED=OFF -DIGC_BUILD__USE_KHRONOS_SPIRV_TRANSLATOR=ON" - -BBCLASSEXTEND = "native nativesdk" - -UPSTREAM_CHECK_GITTAGREGEX = "^igc-(?P<pver>(?!19\..*)\d+(\.\d+)+)$" - -FILES_${PN} += " \ - ${libdir}/igc/NOTICES.txt \ - " diff --git a/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/files/0001-Building-in-tree-with-LLVM-10.0-with-the-LLVM_LINK_L.patch b/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/files/0001-Building-in-tree-with-LLVM-10.0-with-the-LLVM_LINK_L.patch deleted file mode 100644 index 8ffa853b..00000000 --- a/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/files/0001-Building-in-tree-with-LLVM-10.0-with-the-LLVM_LINK_L.patch +++ /dev/null @@ -1,35 +0,0 @@ -From 7fc05c52dd91902fa324a7aac9b90715cfca4717 Mon Sep 17 00:00:00 2001 -From: Naveen Saini <naveen.kumar.saini@intel.com> -Date: Wed, 15 Apr 2020 17:55:32 +0800 -Subject: [PATCH] Building in-tree with LLVM 10.0 with the LLVM_LINK_LLVM_DYLIB - -Failed to link with the LLVMSPIRVLib library. - -Add an explicit dependency to force the correct build order and linking. - -Reference: -https://github.com/KhronosGroup/SPIRV-LLVM-Translator/commit/a6d4ccf082858e63e139ca06c02a071c343d2657 - -Upstream-Status: Submitted [https://github.com/intel/opencl-clang/pull/118] - -Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> ---- - CMakeLists.txt | 2 +- - 1 file changed, 1 insertion(+), 1 deletion(-) - -diff --git a/CMakeLists.txt b/CMakeLists.txt -index 51c140d..b8b514e 100644 ---- a/CMakeLists.txt -+++ b/CMakeLists.txt -@@ -208,7 +208,7 @@ link_directories( - - set(OPENCL_CLANG_LINK_LIBS ${CMAKE_DL_LIBS}) - --if(NOT LLVMSPIRVLib IN_LIST LLVM_AVAILABLE_LIBS) -+if(NOT LLVMSPIRVLib IN_LIST LLVM_AVAILABLE_LIBS OR (USE_PREBUILT_LLVM AND LLVM_LINK_LLVM_DYLIB)) - # SPIRV-LLVM-Translator is not included into LLVM as a component. - # So, we need to list it here explicitly as an external library - list(APPEND OPENCL_CLANG_LINK_LIBS LLVMSPIRVLib) --- -2.17.1 - diff --git a/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/files/0001-cl_headers-CMakeLists.txt-use-clang-from-native-sysr.patch b/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/files/0001-cl_headers-CMakeLists.txt-use-clang-from-native-sysr.patch new file mode 100644 index 00000000..031a77c7 --- /dev/null +++ b/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/files/0001-cl_headers-CMakeLists.txt-use-clang-from-native-sysr.patch @@ -0,0 +1,49 @@ +From 5aea653e611b59c70e529a1bd71885a509831557 Mon Sep 17 00:00:00 2001 +From: Anuj Mittal <anuj.mittal@intel.com> +Date: Tue, 1 Aug 2023 11:15:31 +0800 +Subject: [PATCH] cl_headers/CMakeLists.txt: use clang from native sysroot + +Allow clang to be found in target sysroot for target builds and dont try +to compile cross binaries, we do that ourselves. + +Upstream-Status: Inappropriate [oe-specific] +Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> +--- + CMakeLists.txt | 8 ++++---- + cl_headers/CMakeLists.txt | 2 +- + 2 files changed, 5 insertions(+), 5 deletions(-) + +diff --git a/CMakeLists.txt b/CMakeLists.txt +index 5864009..60ba39e 100644 +--- a/CMakeLists.txt ++++ b/CMakeLists.txt +@@ -35,10 +35,10 @@ set(CMAKE_MODULE_PATH + + include(CMakeFunctions) + +-if(CMAKE_CROSSCOMPILING AND OPENCL_CLANG_BUILD_EXTERNAL) +- include(CrossCompile) +- llvm_create_cross_target(${PROJECT_NAME} NATIVE "" Release) +-endif() ++#if(CMAKE_CROSSCOMPILING AND OPENCL_CLANG_BUILD_EXTERNAL) ++# include(CrossCompile) ++# llvm_create_cross_target(${PROJECT_NAME} NATIVE "" Release) ++#endif() + + if(CMAKE_SOURCE_DIR STREQUAL CMAKE_CURRENT_SOURCE_DIR) + set(USE_PREBUILT_LLVM ON) +diff --git a/cl_headers/CMakeLists.txt b/cl_headers/CMakeLists.txt +index 16cabb7..4423536 100644 +--- a/cl_headers/CMakeLists.txt ++++ b/cl_headers/CMakeLists.txt +@@ -1,6 +1,6 @@ + set(CL_HEADERS_LIB cl_headers) + if(USE_PREBUILT_LLVM) +- find_program(CLANG_COMMAND clang PATHS ${LLVM_TOOLS_BINARY_DIR} NO_DEFAULT_PATH) ++ find_program(CLANG_COMMAND clang PATHS ${LLVM_TOOLS_BINARY_DIR}) + else() + set(CLANG_COMMAND $<TARGET_FILE:clang>) + endif() +-- +2.37.3 + diff --git a/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/files/0001-don-t-redefine-LLVM_TABLEGEN_EXE.patch b/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/files/0001-don-t-redefine-LLVM_TABLEGEN_EXE.patch deleted file mode 100644 index 8df7e3ab..00000000 --- a/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/files/0001-don-t-redefine-LLVM_TABLEGEN_EXE.patch +++ /dev/null @@ -1,32 +0,0 @@ -From f3ef79a6301bab0b3a447f07ceb94c39a95009df Mon Sep 17 00:00:00 2001 -From: Anuj Mittal <anuj.mittal@intel.com> -Date: Thu, 2 Apr 2020 08:59:20 +0800 -Subject: [PATCH] don't redefine LLVM_TABLEGEN_EXE - -Use the value that has been passed by the user. - -Upstream-Status: Submitted -Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> ---- - CMakeLists.txt | 5 ++++- - 1 file changed, 4 insertions(+), 1 deletion(-) - -diff --git a/CMakeLists.txt b/CMakeLists.txt -index 6893e97..941b0ae 100644 ---- a/CMakeLists.txt -+++ b/CMakeLists.txt -@@ -137,7 +137,10 @@ endif(NOT USE_PREBUILT_LLVM) - set (COMPILE_OPTIONS_TD opencl_clang_options.td) - set (COMPILE_OPTIONS_INC opencl_clang_options.inc) - --set(LLVM_TABLEGEN_EXE "llvm-tblgen") -+if(NOT DEFINED LLVM_TABLEGEN_EXE) -+ set(LLVM_TABLEGEN_EXE "llvm-tblgen") -+endif() -+ - set(LLVM_TARGET_DEFINITIONS ${COMPILE_OPTIONS_TD}) - if(USE_PREBUILT_LLVM) - set(TABLEGEN_ADDITIONAL -I ${LLVM_INCLUDE_DIRS}) --- -2.25.1 - diff --git a/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/files/0002-Request-native-clang-only-when-cross-compiling-464.patch b/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/files/0002-Request-native-clang-only-when-cross-compiling-464.patch new file mode 100644 index 00000000..2f1814f8 --- /dev/null +++ b/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/files/0002-Request-native-clang-only-when-cross-compiling-464.patch @@ -0,0 +1,60 @@ +From 43c806ef321b1f677a49d28c89fb7ffecf539c2d Mon Sep 17 00:00:00 2001 +From: Tim Creech <timothy.m.creech@intel.com> +Date: Wed, 28 Jun 2023 03:45:51 -0400 +Subject: [PATCH 2/2] Request native clang only when cross-compiling (#464) + +* Request native clang only when cross-compiling + +LLVM_USE_HOST_TOOLS may be set if LLVM is configured with +LLVM_OPTIMIZED_TABLEGEN, which does not necessarily indicate +cross-compilation or that clang will only execute on the target. + +By checking that CMAKE_CROSSCOMPILING is set, we ensure that we only +build/use clang again if necessary for host execution. + +* fixup: CMAKE_CROSSCOMPILING implies LLVM_USE_HOST_TOOLS + +Co-authored-by: Wenju He <wenju.he@intel.com> + +* fixup: also use CMAKE_CROSSCOMPILING in top-level CMakeLists.txt + +--------- + +Co-authored-by: Wenju He <wenju.he@intel.com> + +Upstream-Status: Backport [https://github.com/intel/opencl-clang/commit/53843eee13cfb2357919ee02714a43bef1af0f86] +Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> +--- + CMakeLists.txt | 2 +- + cl_headers/CMakeLists.txt | 2 +- + 2 files changed, 2 insertions(+), 2 deletions(-) + +diff --git a/CMakeLists.txt b/CMakeLists.txt +index e772de9..5864009 100644 +--- a/CMakeLists.txt ++++ b/CMakeLists.txt +@@ -35,7 +35,7 @@ set(CMAKE_MODULE_PATH + + include(CMakeFunctions) + +-if(LLVM_USE_HOST_TOOLS AND OPENCL_CLANG_BUILD_EXTERNAL) ++if(CMAKE_CROSSCOMPILING AND OPENCL_CLANG_BUILD_EXTERNAL) + include(CrossCompile) + llvm_create_cross_target(${PROJECT_NAME} NATIVE "" Release) + endif() +diff --git a/cl_headers/CMakeLists.txt b/cl_headers/CMakeLists.txt +index 18296c2..16cabb7 100644 +--- a/cl_headers/CMakeLists.txt ++++ b/cl_headers/CMakeLists.txt +@@ -4,7 +4,7 @@ if(USE_PREBUILT_LLVM) + else() + set(CLANG_COMMAND $<TARGET_FILE:clang>) + endif() +-if(LLVM_USE_HOST_TOOLS AND NOT OPENCL_CLANG_BUILD_EXTERNAL) ++if(CMAKE_CROSSCOMPILING AND NOT OPENCL_CLANG_BUILD_EXTERNAL) + build_native_tool(clang CLANG_COMMAND) + endif() + +-- +2.37.3 + diff --git a/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/files/0002-make-sure-only-static-libraries-linked-for-native-bu.patch b/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/files/0002-make-sure-only-static-libraries-linked-for-native-bu.patch deleted file mode 100644 index 473f4d24..00000000 --- a/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/files/0002-make-sure-only-static-libraries-linked-for-native-bu.patch +++ /dev/null @@ -1,42 +0,0 @@ -From b29e00e6fe428a031cf577dfb703cf13eff837f6 Mon Sep 17 00:00:00 2001 -From: Naveen Saini <naveen.kumar.saini@intel.com> -Date: Wed, 15 Apr 2020 18:05:14 +0800 -Subject: [PATCH 2/2] make sure only static libraries linked for native build - -LINK_COMPONENTS=all isn't working for static libs for out of tree builds. Use -LLVM_AVAILABLE_LIBS instead. Reported: - -https://github.com/intel/opencl-clang/issues/114 - -Upstream-Status: Pending - -Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> -Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> ---- - CMakeLists.txt | 3 ++- - 1 file changed, 2 insertions(+), 1 deletion(-) - -diff --git a/CMakeLists.txt b/CMakeLists.txt -index 8707487..ad2dbda 100644 ---- a/CMakeLists.txt -+++ b/CMakeLists.txt -@@ -218,7 +218,7 @@ add_subdirectory(cl_headers) - - set(LLVM_REQUIRES_EH ON) - --if(USE_PREBUILT_LLVM OR CLANG_LINK_CLANG_DYLIB) -+if(false) - list(APPEND OPENCL_CLANG_LINK_LIBS clang-cpp) - else() - list(APPEND OPENCL_CLANG_LINK_LIBS -@@ -266,6 +266,7 @@ add_llvm_library(${TARGET_NAME} SHARED - all - LINK_LIBS - ${OPENCL_CLANG_LINK_LIBS} -+ ${LLVM_AVAILABLE_LIBS} - ) - - # Configure resource file on Windows --- -2.17.1 - diff --git a/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/opencl-clang_11.0.0.bb b/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/opencl-clang.inc index 0c5d0508..31a3fb21 100644 --- a/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/opencl-clang_11.0.0.bb +++ b/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/opencl-clang.inc @@ -5,31 +5,30 @@ DESCRIPTION = "Common clang has OpenCL-oriented API and is capable \ LICENSE = "NCSA" LIC_FILES_CHKSUM = "file://LICENSE;md5=e8a15bf1416762a09ece07e44c79118c" -SRC_URI = "git://github.com/intel/opencl-clang.git;branch=ocl-open-110;protocol=https \ - file://0001-don-t-redefine-LLVM_TABLEGEN_EXE.patch \ +SRC_URI = "git://github.com/intel/opencl-clang.git;branch=${BRANCH};protocol=https \ + file://0002-Request-native-clang-only-when-cross-compiling-464.patch \ + file://0001-cl_headers-CMakeLists.txt-use-clang-from-native-sysr.patch \ " -SRC_URI_append_class-native = " file://0002-make-sure-only-static-libraries-linked-for-native-bu.patch" - -SRCREV = "c67648d41df00ea8ee9d701d17299b86f86f0321" - S = "${WORKDIR}/git" inherit cmake DEPENDS += "clang" -DEPENDS_append_class-target = " opencl-clang-native" +DEPENDS:append:class-target = " opencl-clang-native" COMPATIBLE_HOST = '(x86_64).*-linux' -COMPATIBLE_HOST_libc-musl = "null" +COMPATIBLE_HOST:libc-musl = "null" + +DEPENDS += " spirv-llvm-translator" EXTRA_OECMAKE += "\ -DLLVM_TABLEGEN_EXE=${STAGING_BINDIR_NATIVE}/llvm-tblgen \ -DCMAKE_SKIP_RPATH=TRUE \ - -DPREFERRED_LLVM_VERSION="11.1.0" \ + -DPREFERRED_LLVM_VERSION=${LLVMVERSION} \ " -do_install_append_class-native() { +do_install:append:class-native() { install -d ${D}${bindir} - install -m 0755 ${B}/linux_linker/linux_resource_linker ${D}${bindir}/ + install -m 0755 ${B}/bin/linux_resource_linker ${D}${bindir}/ } BBCLASSEXTEND = "native nativesdk" diff --git a/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/opencl-clang_10.0.0.bb b/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/opencl-clang_10.0.0.bb deleted file mode 100644 index c5ba96c1..00000000 --- a/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/opencl-clang_10.0.0.bb +++ /dev/null @@ -1,35 +0,0 @@ -SUMMARY = "Common clang is a thin wrapper library around clang" -DESCRIPTION = "Common clang has OpenCL-oriented API and is capable \ - to compile OpenCL C kernels to SPIR-V modules." - -LICENSE = "NCSA" -LIC_FILES_CHKSUM = "file://LICENSE;md5=e8a15bf1416762a09ece07e44c79118c" - -SRC_URI = "git://github.com/intel/opencl-clang.git;branch=ocl-open-100;protocol=https \ - file://0001-don-t-redefine-LLVM_TABLEGEN_EXE.patch \ - file://0001-Building-in-tree-with-LLVM-10.0-with-the-LLVM_LINK_L.patch \ - " -SRC_URI_append_class-native = " file://0002-make-sure-only-static-libraries-linked-for-native-bu.patch" - -SRCREV = "92f3f7f1a06f25fb13708f87c26b0fbf50924c96" - -S = "${WORKDIR}/git" - -inherit cmake -DEPENDS += "clang" -DEPENDS_append_class-target = " opencl-clang-native" - -COMPATIBLE_HOST = '(x86_64).*-linux' -COMPATIBLE_HOST_libc-musl = "null" - -EXTRA_OECMAKE += "\ - -DLLVM_TABLEGEN_EXE=${STAGING_BINDIR_NATIVE}/llvm-tblgen \ - -DCMAKE_SKIP_RPATH=TRUE \ - " - -do_install_append_class-native() { - install -d ${D}${bindir} - install -m 0755 ${B}/linux_linker/linux_resource_linker ${D}${bindir}/ -} - -BBCLASSEXTEND = "native nativesdk" diff --git a/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/opencl-clang_12.0.0.bb b/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/opencl-clang_12.0.0.bb deleted file mode 100644 index 6df10912..00000000 --- a/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/opencl-clang_12.0.0.bb +++ /dev/null @@ -1,34 +0,0 @@ -SUMMARY = "Common clang is a thin wrapper library around clang" -DESCRIPTION = "Common clang has OpenCL-oriented API and is capable \ - to compile OpenCL C kernels to SPIR-V modules." - -LICENSE = "NCSA" -LIC_FILES_CHKSUM = "file://LICENSE;md5=e8a15bf1416762a09ece07e44c79118c" - -SRC_URI = "git://github.com/intel/opencl-clang.git;branch=ocl-open-120;protocol=https \ - file://0001-don-t-redefine-LLVM_TABLEGEN_EXE.patch \ - " - -SRCREV = "cea544b04460a47848e585bfe75cd774c0324b45" - -S = "${WORKDIR}/git" - -inherit cmake -DEPENDS += " clang spirv-llvm-translator" -DEPENDS_append_class-target = " opencl-clang-native" - -COMPATIBLE_HOST = '(x86_64).*-linux' -COMPATIBLE_HOST_libc-musl = "null" - -EXTRA_OECMAKE += "\ - -DLLVM_TABLEGEN_EXE=${STAGING_BINDIR_NATIVE}/llvm-tblgen \ - -DCMAKE_SKIP_RPATH=TRUE \ - -DPREFERRED_LLVM_VERSION="12.0.0" \ - " - -do_install_append_class-native() { - install -d ${D}${bindir} - install -m 0755 ${B}/linux_linker/linux_resource_linker ${D}${bindir}/ -} - -BBCLASSEXTEND = "native nativesdk" diff --git a/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/opencl-clang_14.0.0.bb b/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/opencl-clang_14.0.0.bb new file mode 100644 index 00000000..7e56af75 --- /dev/null +++ b/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/opencl-clang_14.0.0.bb @@ -0,0 +1,5 @@ +require opencl-clang.inc + +SRCREV = "980f1691c5babcf824ee10375a04a0d0c5d7d44a" + +BRANCH = "ocl-open-140" diff --git a/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/opencl-clang_15.0.0.bb b/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/opencl-clang_15.0.0.bb new file mode 100644 index 00000000..e946c31c --- /dev/null +++ b/dynamic-layers/clang-layer/recipes-opencl/opencl-clang/opencl-clang_15.0.0.bb @@ -0,0 +1,5 @@ +require opencl-clang.inc + +SRCREV = "60fd799cc58755c16d951f9ebfde6d0f9b8554dd" + +BRANCH = "ocl-open-150" diff --git a/dynamic-layers/clang-layer/recipes-opencl/vc-intrinsics/vc-intrinsics_1.0.bb b/dynamic-layers/clang-layer/recipes-opencl/vc-intrinsics/vc-intrinsics_1.0.bb deleted file mode 100644 index eca1cb4d..00000000 --- a/dynamic-layers/clang-layer/recipes-opencl/vc-intrinsics/vc-intrinsics_1.0.bb +++ /dev/null @@ -1,24 +0,0 @@ -SUMMARY = "VC Intrinsics" -DESCRIPTION = "VC Intrinsics project contains a set of new intrinsics on \ -top of core LLVM IR instructions that represent SIMD semantics of a program \ -targeting GPU" - -LICENSE = "MIT" -LIC_FILES_CHKSUM = "file://License.md;md5=c18ea6bb4786a26bf4eee88a7424a408" - -SRC_URI = "git://github.com/intel/vc-intrinsics.git;protocol=https; \ - " -SRCREV = "77974f5a32e19475125531483bede9ee23935257" - -S = "${WORKDIR}/git" - -inherit cmake - -COMPATIBLE_HOST = '(x86_64).*-linux' -COMPATIBLE_HOST_libc-musl = "null" - -DEPENDS += " clang" - -EXTRA_OECMAKE = "-DLLVM_DIR=${STAGING_LIBDIR}" - -BBCLASSEXTEND = "native nativesdk" diff --git a/dynamic-layers/meta-python/recipes-opencv/dldt/openvino-model-optimizer_2021.2.bb b/dynamic-layers/meta-python/recipes-opencv/dldt/openvino-model-optimizer_2024.0.0.bb index 9d8c1e18..70177c28 100644 --- a/dynamic-layers/meta-python/recipes-opencv/dldt/openvino-model-optimizer_2021.2.bb +++ b/dynamic-layers/meta-python/recipes-opencv/dldt/openvino-model-optimizer_2024.0.0.bb @@ -5,29 +5,29 @@ environment, performs static model analysis, and adjusts deep \ learning models for optimal execution on end-point target devices." HOMEPAGE = "https://01.org/openvinotoolkit" -SRC_URI = "git://github.com/openvinotoolkit/openvino.git;protocol=https;branch=releases/2021/2;lfs=0 \ +SRC_URI = "git://github.com/openvinotoolkit/openvino.git;protocol=https;branch=releases/2024/0;lfs=0 \ " -SRCREV = "4795391b73381660b69b4cd3986c7a0bf902e868" +SRCREV = "34caeefd07800b59065345d651949efbe8ab6649" LICENSE = "Apache-2.0" LIC_FILES_CHKSUM = "file://LICENSE;md5=86d3f3a95c324c9479bd8986968f4327" CVE_PRODUCT = "intel:openvino" - S = "${WORKDIR}/git" -do_install() { - mkdir -p ${D}${datadir}/openvino/model-optimizer - cp -r model-optimizer ${D}${datadir}/openvino/ -} +inherit setuptools3 -RDEPENDS_${PN} += " \ - python3-numpy \ - python3-protobuf \ +SETUPTOOLS_SETUP_PATH = "${WORKDIR}/git/tools/mo" + +RDEPENDS:${PN} += " \ python3-defusedxml \ + python3-fastjsonschema \ python3-networkx \ - python3-test-generator \ + python3-numpy \ + python3-protobuf \ + python3-requests \ + python3-urllib3 \ bash \ " -FILES_${PN} += "${datadir}/openvino" +UPSTREAM_CHECK_GITTAGREGEX = "(?P<pver>(\d+\.\d+\.\d+))$" diff --git a/dynamic-layers/openembedded-layer/recipes-bsp/amt/lms_2102.1.0.0.bb b/dynamic-layers/openembedded-layer/recipes-bsp/amt/lms_2406.0.0.0.bb index c4cf8052..bdf32576 100644 --- a/dynamic-layers/openembedded-layer/recipes-bsp/amt/lms_2102.1.0.0.bb +++ b/dynamic-layers/openembedded-layer/recipes-bsp/amt/lms_2406.0.0.0.bb @@ -8,13 +8,11 @@ LIC_FILES_CHKSUM = "file://COPYING;md5=2ee41112a44fe7014dce33e26468ba93" COMPATIBLE_HOST = '(i.86|x86_64).*-linux' -COMPATIBLE_HOST_libc-musl = "null" +COMPATIBLE_HOST:libc-musl = "null" -inherit cmake systemd features_check +inherit cmake systemd features_check python3native -DEPENDS = "metee ace xerces-c libnl libxml2 glib-2.0 glib-2.0-native" - -EXTRA_OECMAKE += "-DPYTHON_EXECUTABLE=${HOSTTOOLS_DIR}/python3" +DEPENDS = "metee ace xerces-c libnl libxml2 glib-2.0 glib-2.0-native pkgconfig-native python3-packaging-native" # Enable either connman or networkmanager or none but not both. PACKAGECONFIG ??= "connman" @@ -23,19 +21,24 @@ PACKAGECONFIG[networkmanager] = "-DNETWORK_NM=ON, -DNETWORK_NM=OFF, networkmanag REQUIRED_DISTRO_FEATURES= "systemd" -FILES_${PN} += "${datadir}/dbus-1/system-services/*.service" +FILES:${PN} += "${datadir}/dbus-1/system-services/*.service" S = "${WORKDIR}/git" -SYSTEMD_SERVICE_${PN} = "lms.service" +SYSTEMD_SERVICE:${PN} = "lms.service" -SRC_URI = "git://github.com/intel/lms.git \ +SRC_URI = "git://github.com/intel/lms.git;branch=master;protocol=https \ " -SRCREV = "2d400ec587470fc960dceecc84c123a38415d137" +SRCREV = "388f115b2aeb3ea11499971c65f828daefd32c47" -do_install_append() { +do_install:append() { + install -d ${D}${sysconfdir}/lms install -d ${D}${systemd_system_unitdir} install -m 0644 ${B}/UNS/lms.service ${D}${systemd_system_unitdir} + install -d ${D}${sysconfdir}/udev/rules.d + install -m 0644 ${S}/UNS/linux_scripts/70-mei-wdt.rules ${D}${sysconfdir}/udev/rules.d/70-mei-wdt.rules } -RDEPENDS_${PN} += "ace" +RDEPENDS:${PN} += "ace" + +CVE_STATUS[CVE-2018-1000535] = "cpe-incorrect: This CVE is for a different LMS - Lan Management System." diff --git a/dynamic-layers/openembedded-layer/recipes-bsp/thermald/files/0001-Makefile-Fix-build-Issue.patch b/dynamic-layers/openembedded-layer/recipes-bsp/thermald/files/0001-Makefile-Fix-build-Issue.patch new file mode 100644 index 00000000..270cc84d --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-bsp/thermald/files/0001-Makefile-Fix-build-Issue.patch @@ -0,0 +1,39 @@ +From c4c049481d48f33319b1d04cad56d622c4eed65f Mon Sep 17 00:00:00 2001 +From: Yogesh Tyagi <yogesh.tyagi@intel.com> +Date: Mon, 11 Mar 2024 13:13:32 +0800 +Subject: [PATCH] Makefile: Fix build Issue +MIME-Version: 1.0 +Content-Type: text/plain; charset=UTF-8 +Content-Transfer-Encoding: 8bit + +In case build directory is different from source, make sure make is able to find the correct input files. + +Fixes: +| dbus-binding-tool --prefix=thd_dbus_interface --mode=glib-server --output=thd_dbus_interface.h ../git/src/thd_dbus_interface.xml +| glib-compile-resources --generate-source thermald-resource.gresource.xml +| Failed to open file “thermald-resource.gresource.xml”: No such file or directory + + +Upstream-Status: Submitted +https://github.com/intel/thermal_daemon/pull/436 + +Signed-off-by: Yogesh Tyagi <yogesh.tyagi@intel.com> +--- + Makefile.am | 2 +- + 1 file changed, 1 insertion(+), 1 deletion(-) + +diff --git a/Makefile.am b/Makefile.am +index a0c051a..772e280 100644 +--- a/Makefile.am ++++ b/Makefile.am +@@ -95,6 +95,6 @@ thd_dbus_interface.h: $(top_srcdir)/src/thd_dbus_interface.xml + $(AM_V_GEN) dbus-binding-tool --prefix=thd_dbus_interface --mode=glib-server --output=$@ $< + + thermald-resource.c: $(top_srcdir)/thermald-resource.gresource.xml +- $(AM_V_GEN) glib-compile-resources --generate-source thermald-resource.gresource.xml ++ $(AM_V_GEN) glib-compile-resources --generate-source --sourcedir=${top_srcdir} $< + + CLEANFILES = $(BUILT_SOURCES) +-- +2.34.1 + diff --git a/dynamic-layers/openembedded-layer/recipes-bsp/thermald/thermald_2.4.3.bb b/dynamic-layers/openembedded-layer/recipes-bsp/thermald/thermald_2.5.6.bb index add00e3b..df242cdf 100644 --- a/dynamic-layers/openembedded-layer/recipes-bsp/thermald/thermald_2.4.3.bb +++ b/dynamic-layers/openembedded-layer/recipes-bsp/thermald/thermald_2.5.6.bb @@ -7,20 +7,23 @@ compensation using available cooling methods." HOMEPAGE = "https://github.com/01org/thermal_daemon" DEPENDS = "dbus dbus-glib dbus-glib-native libxml2 glib-2.0 glib-2.0-native upower libevdev" -LICENSE = "GPLv2" +DEPENDS += "autoconf-archive-native" + +LICENSE = "GPL-2.0-only" LIC_FILES_CHKSUM = "file://COPYING;md5=ea8831610e926e2e469075b52bf08848" -SRC_URI = "git://github.com/intel/thermal_daemon/ \ +SRC_URI = "git://github.com/intel/thermal_daemon/;branch=master;protocol=https \ + file://0001-Makefile-Fix-build-Issue.patch \ " -SRCREV = "9e731de1bd7bccd2c7c2e7923f01fbc4287ed80a" +SRCREV = "3990ce2391e5c07293758c056e09e4f2453691b4" S = "${WORKDIR}/git" inherit pkgconfig autotools systemd gtk-doc # gtkdocsize runs before autotools do_configure and it copies gtk-doc.m4 and fails # to copy becuase there is no m4 dir yet. -do_configure_prepend () { +do_configure:prepend () { mkdir -p ${S}/m4 } @@ -28,13 +31,13 @@ EXTRA_OECONF = " \ --with-systemdsystemunitdir=${systemd_system_unitdir} \ " -FILES_${PN} += "${datadir}/dbus-1/system-services/*.service" +FILES:${PN} += "${datadir}/dbus-1" -SYSTEMD_SERVICE_${PN} = "thermald.service" +SYSTEMD_SERVICE:${PN} = "thermald.service" COMPATIBLE_HOST = '(i.86|x86_64).*-linux' -CONFFILES_${PN} = " \ +CONFFILES:${PN} = " \ ${sysconfdir}/thermald/thermal-conf.xml \ ${sysconfdir}/thermald/thermal-cpu-cdev-order.xml \ " diff --git a/dynamic-layers/openembedded-layer/recipes-core/ispc/ispc/0001-Add-print-function-to-print-test-run-status-in-ptest.patch b/dynamic-layers/openembedded-layer/recipes-core/ispc/ispc/0001-Add-print-function-to-print-test-run-status-in-ptest.patch new file mode 100644 index 00000000..4d583657 --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-core/ispc/ispc/0001-Add-print-function-to-print-test-run-status-in-ptest.patch @@ -0,0 +1,53 @@ +From deccc0c69c2c8759c52885be8bdda54d3cee481c Mon Sep 17 00:00:00 2001 +From: Yogesh Tyagi <yogesh.tyagi@intel.com> +Date: Sun, 11 Dec 2022 22:34:15 +0800 +Subject: [PATCH] Add print function to print test run status in ptest format + +Upstream-Status: Inappropriate [OE ptest specific] + +Signed-off-by: Yogesh Tyagi <yogesh.tyagi@intel.com> +--- + run_tests.py | 16 ++++++++++++++++ + 1 file changed, 16 insertions(+) + +diff --git a/run_tests.py b/run_tests.py +index 1cd796dd..e3ffd1ab 100755 +--- a/run_tests.py ++++ b/run_tests.py +@@ -327,6 +327,9 @@ def run_test(testname, host, target): + else: + ispc_exe_rel = add_prefix(host.ispc_cmd, host, target) + ++ # to reslove the error '.rodata' can not be used when making a PIE object ++ ispc_exe_rel = ispc_exe_rel + " --pic" ++ + # is this a test to make sure an error is issued? + want_error = (filename.find("tests_errors") != -1) + if want_error == True: +@@ -795,6 +798,17 @@ def check_compiler_exists(compiler_exe): + return + error("missing the required compiler: %s \n" % compiler_exe, 1) + ++def print_test_run_status(results): ++ for fstatus in results: ++ if (fstatus[1] == Status.Success): ++ print( "%s: %s" % ("PASS", fstatus[0])) ++ elif (fstatus[1] == Status.Compfail): ++ print( "%s: %s" % ("FAIL", fstatus[0])) ++ elif (fstatus[1] == Status.Runfail): ++ print( "%s: %s" % ("FAIL", fstatus[0])) ++ elif (fstatus[1] == Status.Skip): ++ print( "%s: %s" % ("SKIP", fstatus[0])) ++ + def print_result(status, results, s, run_tests_log, csv): + title = StatusStr[status] + file_list = [fname for fname, fstatus in results if status == fstatus] +@@ -938,6 +952,8 @@ def run_tests(options1, args, print_version): + pass_rate = -1 + print_debug("PASSRATE (%d/%d) = %d%% \n\n" % (len(run_succeed_files), total_tests_executed, pass_rate), s, run_tests_log) + ++ print_test_run_status(results) ++ + for status in Status: + print_result(status, results, s, run_tests_log, options.csv) + fails = [status != Status.Compfail and status != Status.Runfail for _, status in results] diff --git a/dynamic-layers/openembedded-layer/recipes-core/ispc/ispc/0001-Fix-QA-Issues.patch b/dynamic-layers/openembedded-layer/recipes-core/ispc/ispc/0001-Fix-QA-Issues.patch new file mode 100644 index 00000000..b0a76ff9 --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-core/ispc/ispc/0001-Fix-QA-Issues.patch @@ -0,0 +1,36 @@ +From 7beff95c11071170eb27b6fa7d0cc77588caee8e Mon Sep 17 00:00:00 2001 +From: Yogesh Tyagi <yogesh.tyagi@intel.com> +Date: Tue, 26 Jul 2022 15:25:10 +0800 +Subject: [PATCH] Fix QA Issues + +Stop ispc from inserting host file path in generated headers which leads to reproducibility problems. + +Upstream-Status: Inappropriate [OE build specific] + +Signed-off-by: Yogesh Tyagi <yogesh.tyagi@intel.com> +--- + src/module.cpp | 4 ++-- + 1 file changed, 2 insertions(+), 2 deletions(-) + +diff --git a/src/module.cpp b/src/module.cpp +index e2084d2e..e2626865 100644 +--- a/src/module.cpp ++++ b/src/module.cpp +@@ -2555,7 +2555,7 @@ bool Module::writeHeader(const char *fn) { + perror("fopen"); + return false; + } +- fprintf(f, "//\n// %s\n// (Header automatically generated by the ispc compiler.)\n", fn); ++ fprintf(f, "//\n// \n// (Header automatically generated by the ispc compiler.)\n"); + fprintf(f, "// DO NOT EDIT THIS FILE.\n//\n\n"); + + // Create a nice guard string from the filename, turning any +@@ -2677,7 +2677,7 @@ bool Module::writeDispatchHeader(DispatchHeaderInfo *DHI) { + FILE *f = DHI->file; + + if (DHI->EmitFrontMatter) { +- fprintf(f, "//\n// %s\n// (Header automatically generated by the ispc compiler.)\n", DHI->fn); ++ fprintf(f, "//\n// \n// (Header automatically generated by the ispc compiler.)\n"); + fprintf(f, "// DO NOT EDIT THIS FILE.\n//\n\n"); + } + // Create a nice guard string from the filename, turning any diff --git a/dynamic-layers/openembedded-layer/recipes-core/ispc/ispc/0002-cmake-don-t-build-for-32-bit-targets.patch b/dynamic-layers/openembedded-layer/recipes-core/ispc/ispc/0002-cmake-don-t-build-for-32-bit-targets.patch new file mode 100644 index 00000000..f452dc50 --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-core/ispc/ispc/0002-cmake-don-t-build-for-32-bit-targets.patch @@ -0,0 +1,52 @@ +From 16a2c22339287122d2c25d8bb33a5a51b4e6ee51 Mon Sep 17 00:00:00 2001 +From: Naveen Saini <naveen.kumar.saini@intel.com> +Date: Thu, 24 Feb 2022 20:01:11 +0530 +Subject: [PATCH] cmake: don't build for 32-bit targets + +Error log: +| tmp/work/corei7-64-poky-linux/ispc/1.16.0-r0/recipe-sysroot/usr/include/bits/long-double.h:23:10: fatal error: 'bits/long-double-32.h' file not found +| #include <bits/long-double-32.h> +| ^~~~~~~~~~~~~~~~~~~~~~~ +| 1 error generated. + +Remove SYSTEM include search path and set -isysroot dir path +for root dir for cross compilation. + +Upstream-Status: Inappropriate + +Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> +--- + cmake/GenerateBuiltins.cmake | 6 +++--- + 1 file changed, 3 insertions(+), 3 deletions(-) + +diff --git a/cmake/GenerateBuiltins.cmake b/cmake/GenerateBuiltins.cmake +index f84494ed..d90cb1ec 100644 +--- a/cmake/GenerateBuiltins.cmake ++++ b/cmake/GenerateBuiltins.cmake +@@ -253,7 +253,7 @@ function(builtin_to_cpp bit os_name arch supported_archs supported_oses resultFi + # In this case headers will be installed in /usr/arm-linux-gnueabihf/include and will not be picked up + # by clang by default. So the following line adds such path explicitly. If this path doesn't exist and + # the headers can be found in other locations, this should not be a problem. +- set(includePath -isystem/usr/${debian_triple}/include) ++ set(includePath -isysroot${SYSROOT_DIR}) + endif() + endif() + +@@ -339,7 +339,7 @@ function (generate_target_builtins resultList) + set(regular_targets ${ARGN}) + list(FILTER regular_targets EXCLUDE REGEX wasm) + foreach (ispc_target ${regular_targets}) +- foreach (bit 32 64) ++ foreach (bit 64) + foreach (os_name ${TARGET_OS_LIST_FOR_LL}) + target_ll_to_cpp(target-${ispc_target} ${bit} ${os_name} output${os_name}${bit}) + list(APPEND tmpList ${output${os_name}${bit}}) +@@ -405,7 +405,7 @@ function (generate_common_builtins resultList) + endif() + + message (STATUS "ISPC will be built with support of ${supported_oses} for ${supported_archs}") +- foreach (bit 32 64) ++ foreach (bit 64) + foreach (os_name "windows" "linux" "freebsd" "macos" "android" "ios" "ps4" "web") + foreach (arch "x86" "arm" "wasm") + builtin_to_cpp(${bit} ${os_name} ${arch} "${supported_archs}" "${supported_oses}" res${bit}${os_name}${arch}) diff --git a/dynamic-layers/openembedded-layer/recipes-core/ispc/ispc/run-ptest b/dynamic-layers/openembedded-layer/recipes-core/ispc/ispc/run-ptest new file mode 100644 index 00000000..77d13bb4 --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-core/ispc/ispc/run-ptest @@ -0,0 +1,2 @@ +#!/bin/sh +python3 run_tests.py diff --git a/dynamic-layers/openembedded-layer/recipes-core/ispc/ispc_1.23.0.bb b/dynamic-layers/openembedded-layer/recipes-core/ispc/ispc_1.23.0.bb new file mode 100644 index 00000000..ed8df859 --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-core/ispc/ispc_1.23.0.bb @@ -0,0 +1,64 @@ +SUMMARY = "Intel(R) Implicit SPMD Program Compiler" +DESCRIPTION = "ispc is a compiler for a variant of the C programming language, \ +with extensions for single program, multiple data programming." +HOMEPAGE = "https://github.com/ispc/ispc" + +LICENSE = "BSD-3-Clause & Apache-2.0-with-LLVM-exception" +LIC_FILES_CHKSUM = "file://LICENSE.txt;md5=da5ecffdd210b3cf776b32b41c182e87 \ + file://third-party-programs.txt;md5=2061218c7be521556719c8b504bf9ddd" + +inherit cmake python3native ptest + +S = "${WORKDIR}/git" + +SRC_URI = "git://github.com/ispc/ispc.git;protocol=https;branch=main \ + file://0002-cmake-don-t-build-for-32-bit-targets.patch \ + file://0001-Fix-QA-Issues.patch \ + file://0001-Add-print-function-to-print-test-run-status-in-ptest.patch \ + file://run-ptest \ + " + +SRCREV = "bcb2cf896c00f9a802a11cbf291ef6e44b205416" + +COMPATIBLE_HOST = '(x86_64).*-linux' + +DEPENDS += " clang-native bison-native flex-native" +DEPENDS:append:class-target = " clang" +RDEPENDS:${PN}-ptest += " python3-multiprocessing" + +PACKAGECONFIG ??= "tbb" +PACKAGECONFIG[tbb] = "-DISPCRT_BUILD_TASK_MODEL=TBB, -DISPCRT_BUILD_TASK_MODEL=OpenMP, tbb" + +YFLAGS='-d -t -v -y --file-prefix-map=${WORKDIR}=/usr/src/debug/${PN}/${EXTENDPE}${PV}-${PR}' + +do_configure:prepend() { + sed -i -e 's#\${BISON_EXECUTABLE}.*#\${BISON_EXECUTABLE} ${YFLAGS} #g' ${S}/CMakeLists.txt + sed -i -e 's#\${FLEX_EXECUTABLE}.*#\${FLEX_EXECUTABLE} \-L #g' ${S}/CMakeLists.txt +} + +do_install_ptest() { + cp -rf ${S}/run_tests.py ${D}${PTEST_PATH} + cp -rf ${S}/common.py ${D}${PTEST_PATH} + cp -rf ${S}/tests ${D}${PTEST_PATH} + cp -rf ${S}/test_static.isph ${D}${PTEST_PATH} + cp -rf ${S}/fail_db.txt ${D}${PTEST_PATH} + cp -rf ${S}/test_static.cpp ${D}${PTEST_PATH} +} + +EXTRA_OECMAKE += " \ + -DISPC_INCLUDE_TESTS=OFF \ + -DISPC_INCLUDE_EXAMPLES=OFF \ + -DARM_ENABLED=OFF \ + -DISPC_CROSS=ON \ + -DISPC_ANDROID_TARGET=OFF \ + -DISPC_FREEBSD_TARGET=OFF \ + -DISPC_WINDOWS_TARGET=OFF \ + -DISPC_IOS_TARGET=OFF \ + -DISPC_PS_TARGET=OFF \ + -DSYSROOT_DIR=${STAGING_DIR} \ + -DCLANG_EXECUTABLE=${STAGING_BINDIR_NATIVE}/clang \ + -DCLANGPP_EXECUTABLE=${STAGING_BINDIR_NATIVE}/clang++ \ + -DLLVM_AS_EXECUTABLE=${STAGING_BINDIR_NATIVE}/llvm-as \ + " + +BBCLASSEXTEND = "native nativesdk" diff --git a/dynamic-layers/openembedded-layer/recipes-core/levelzero/level-zero_1.2.3.bb b/dynamic-layers/openembedded-layer/recipes-core/levelzero/level-zero_1.15.8.bb index 9ab0804f..8bade9f4 100644 --- a/dynamic-layers/openembedded-layer/recipes-core/levelzero/level-zero_1.2.3.bb +++ b/dynamic-layers/openembedded-layer/recipes-core/levelzero/level-zero_1.15.8.bb @@ -1,10 +1,10 @@ SUMMARY = "oneAPI Level Zero Specification Headers and Loader" HOMEPAGE = "https://github.com/oneapi-src/level-zero" LICENSE = "MIT" -LIC_FILES_CHKSUM = "file://LICENSE;md5=8956dfdba7f8169c4005d1e9753ffddc" +LIC_FILES_CHKSUM = "file://LICENSE;md5=97957beb2f7808ffa247e5d93e6442cc" -SRC_URI = "git://github.com/oneapi-src/level-zero.git;protocol=https" -SRCREV = "0d30b1fa712253c68bfdfa3863d380df4301b8a4" +SRC_URI = "git://github.com/oneapi-src/level-zero.git;protocol=https;branch=master" +SRCREV = "1685d01497428ca4d8b99200972b64685424d5c9" S = "${WORKDIR}/git" inherit cmake @@ -14,7 +14,7 @@ UPSTREAM_CHECK_GITTAGREGEX = "^v(?P<pver>(\d+(\.\d+)+))$" PACKAGES =+ "${PN}-headers ${PN}-samples ${PN}-loader" -do_install_append () { +do_install:append () { install -d ${D}${bindir} ${D}${libdir} install -m 755 ${B}/bin/zello* ${D}${bindir} @@ -22,10 +22,12 @@ do_install_append () { } -FILES_${PN}-headers = "${includedir}" -FILES_${PN}-samples = "${bindir} ${libdir}/libze_null* ${libdir}/libze_validation*" -FILES_${PN}-loader = "${libdir}" +FILES:${PN}-headers = "${includedir}" +FILES:${PN}-samples = "${bindir} ${libdir}/libze_null* ${libdir}/libze_validation*" +FILES:${PN}-loader = "${libdir}" # PN-loader (non -dev/-dbg/nativesdk- package) contains symlink .so -INSANE_SKIP_${PN}-loader = "dev-so" -INSANE_SKIP_${PN}-samples = "dev-so" +INSANE_SKIP:${PN}-loader = "dev-so" +INSANE_SKIP:${PN}-samples = "dev-so" +ALLOW_EMPTY:${PN} = "1" +BBCLASSEXTEND = "native nativesdk" diff --git a/dynamic-layers/openembedded-layer/recipes-oneapi/dpcpp-compiler/intel-oneapi-dpcpp-cpp-runtime_2024.0.0-49819.bb b/dynamic-layers/openembedded-layer/recipes-oneapi/dpcpp-compiler/intel-oneapi-dpcpp-cpp-runtime_2024.0.0-49819.bb new file mode 100644 index 00000000..cedbae1e --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-oneapi/dpcpp-compiler/intel-oneapi-dpcpp-cpp-runtime_2024.0.0-49819.bb @@ -0,0 +1,54 @@ +SUMMARY = "Intel® oneAPI DPC++/C++ Compiler runtime files" +DESCRIPTION = "The Intel® oneAPI DPC++/C++ Compiler provides optimizations \ +that help your applications run faster on Intel® 64 architectures with support \ +for the latest C, C++, and SYCL language standards. This compiler produces \ +optimized code that can run significantly faster by taking advantage of the \ +ever-increasing core count and vector register width in Intel® Xeon® processors \ +and compatible processors." + +HOMEPAGE = "https://www.intel.com/content/www/us/en/developer/tools/oneapi/dpc-compiler.html" + +LICENSE="EULA" + +COMPILERMAINVER = "2024.0" + +LIC_FILES_CHKSUM = " \ + file://opt/intel/oneapi/licensing/${COMPILERMAINVER}/licensing/${COMPILERMAINVER}/license.htm;md5=5ff64c6ff3ef98089ed69360e7a84c39 \ + " +COMPILERDOTVER = "2024.0.0-49406" +DEVUTILITVERSION = "2024.0-2024.0.0-49320" +SRC_URI = " \ + https://apt.repos.intel.com/oneapi/pool/main/intel-oneapi-compiler-dpcpp-cpp-runtime-${COMPILERMAINVER}-${PV}_amd64.deb;subdir=${BPN};name=dpcpp-runtime \ + https://apt.repos.intel.com/oneapi/pool/main/intel-oneapi-compiler-dpcpp-cpp-common-${COMPILERMAINVER}-${PV}_all.deb;subdir=${BPN};name=dpcpp-common \ + https://apt.repos.intel.com/oneapi/pool/main/intel-oneapi-compiler-shared-runtime-${COMPILERMAINVER}-${PV}_amd64.deb;subdir=${BPN};name=compiler-shared-runtime \ + https://apt.repos.intel.com/oneapi/pool/main/intel-oneapi-common-vars-${COMPILERDOTVER}_all.deb;subdir=${BPN};name=common-vars \ + https://apt.repos.intel.com/oneapi/pool/main/intel-oneapi-openmp-${COMPILERMAINVER}-${PV}_amd64.deb;subdir=${BPN};name=openmp \ + https://apt.repos.intel.com/oneapi/pool/main/intel-oneapi-openmp-common-${COMPILERMAINVER}-${PV}_all.deb;subdir=${BPN};name=openmp-common \ + https://apt.repos.intel.com/oneapi/pool/main/intel-oneapi-common-licensing-${COMPILERMAINVER}-${COMPILERDOTVER}_all.deb;subdir=${BPN};name=license \ + https://apt.repos.intel.com/oneapi/pool/main/intel-oneapi-dev-utilities-${DEVUTILITVERSION}_amd64.deb;subdir=${BPN};name=dev-utils \ + " + +SRC_URI[dpcpp-runtime.sha256sum] = "e24f0ba69daf3f66ceaf23d5e632f183cdb90bac69f65407fdb4407fc9034f33" +SRC_URI[dpcpp-common.sha256sum] = "f5a3db6a725598224edf1099260955aee3e36beadcaed2af5b8b453e873a82fa" +SRC_URI[compiler-shared-runtime.sha256sum] = "bce010cbe076259ddd3feb8e69792869e22fccd5b4e2c9af9e815826f2c1a394" +SRC_URI[common-vars.sha256sum] = "368553c99db1b52060b8225355336778be0b00e5991d0f769c42f891c6328750" +SRC_URI[openmp.sha256sum] = "154ff1e81adfdc872ba1d47bd860de70d62188417c7128422435dfd0ceca62fe" +SRC_URI[openmp-common.sha256sum] = "8217001d78311cbef97dd139e684c6767932b532309c3843ba57d7894d15c07d" +SRC_URI[license.sha256sum] = "9f9c8a12fc0bc82ab5b71e118e66745eff23f42224eba304068225b366cd74b6" +SRC_URI[dev-utils.sha256sum] = "c675d960a5abca361cead9217d6e74adee499ee0a095c4e44092bd710b304d50" + +S = "${WORKDIR}/${BPN}" + +inherit bin_package + +RDEPENDS:${PN} += "virtual-opencl-icd zlib tbb level-zero-loader bash tcsh" +SKIP_FILEDEPS:${PN} = '1' + +INHIBIT_PACKAGE_STRIP = "1" +INHIBIT_PACKAGE_DEBUG_SPLIT = "1" + +# doesn't have GNU_HASH (didn't pass LDFLAGS?) +INSANE_SKIP:${PN} += "textrel dev-so dev-elf ldflags already-stripped staticdev rpaths arch useless-rpaths file-rdeps" + +FILES_SOLIBSDEV = "" +BBCLASSEXTEND = "native nativesdk" diff --git a/dynamic-layers/openembedded-layer/recipes-oneapi/dpcpp-compiler/intel-oneapi-dpcpp-cpp_2024.0.0-49819.bb b/dynamic-layers/openembedded-layer/recipes-oneapi/dpcpp-compiler/intel-oneapi-dpcpp-cpp_2024.0.0-49819.bb new file mode 100644 index 00000000..90ada087 --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-oneapi/dpcpp-compiler/intel-oneapi-dpcpp-cpp_2024.0.0-49819.bb @@ -0,0 +1,45 @@ +SUMMARY = "Intel® oneAPI DPC++/C++ Compiler" +DESCRIPTION = "The Intel® oneAPI DPC++/C++ Compiler provides optimizations \ +that help your applications run faster on Intel® 64 architectures with support \ +for the latest C, C++, and SYCL language standards. This compiler produces \ +optimized code that can run significantly faster by taking advantage of the \ +ever-increasing core count and vector register width in Intel® Xeon® processors \ +and compatible processors." + +HOMEPAGE = "https://www.intel.com/content/www/us/en/developer/tools/oneapi/dpc-compiler.html" + +LICENSE="EULA" + +COMPILERMAINVER = "2024.0" + +LIC_FILES_CHKSUM = " \ + file://opt/intel/oneapi/compiler/${COMPILERMAINVER}/share/doc/compiler/credist.txt;md5=b41f55af9f479b9570fc35b955d5ba1a \ + " + +SRC_URI = " \ + https://apt.repos.intel.com/oneapi/pool/main/intel-oneapi-dpcpp-cpp-${COMPILERMAINVER}-${PV}_amd64.deb;subdir=${BPN};name=icx-compiler \ + https://apt.repos.intel.com/oneapi/pool/main/intel-oneapi-compiler-shared-${COMPILERMAINVER}-${PV}_amd64.deb;subdir=${BPN};name=compiler-linker \ + https://apt.repos.intel.com/oneapi/pool/main/intel-oneapi-compiler-shared-common-${COMPILERMAINVER}-${PV}_all.deb;subdir=${BPN};name=shared-common \ + " + +SRC_URI[icx-compiler.sha256sum] = "0dcbac766d5a1519d4cf393f5a85e71d19024fef65f77638f3f849796b62cd82" +SRC_URI[compiler-linker.sha256sum] = "e00faea6d797934d62143e4aa70b727ce30f80fdf30769d22122b3051140c236" +SRC_URI[shared-common.sha256sum] = "cf490a4a790f349da79e618359598d3b32312ca3b2639e5d4c84e1cfa2745558" + +S = "${WORKDIR}/${BPN}" + +inherit bin_package + +RDEPENDS:${PN} += "intel-oneapi-dpcpp-cpp-runtime" +SKIP_FILEDEPS:${PN} = '1' + +INHIBIT_PACKAGE_STRIP = "1" +INHIBIT_PACKAGE_DEBUG_SPLIT = "1" + +# doesn't have GNU_HASH (didn't pass LDFLAGS?) +INSANE_SKIP:${PN} += "textrel dev-so dev-elf ldflags already-stripped file-rdeps staticdev rpaths arch useless-rpaths" + +FILES_SOLIBSDEV = "" + +EXCLUDE_FROM_SHLIBS = "1" +BBCLASSEXTEND = "native nativesdk" diff --git a/dynamic-layers/openembedded-layer/recipes-oneapi/embree/embree_4.3.1.bb b/dynamic-layers/openembedded-layer/recipes-oneapi/embree/embree_4.3.1.bb new file mode 100644 index 00000000..fb341b01 --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-oneapi/embree/embree_4.3.1.bb @@ -0,0 +1,30 @@ +SUMMARY = "Collection of high-performance ray tracing kernels" +DESCRIPTION = "A collection of high-performance ray tracing kernels \ +intended to graphics application engineers that want to improve the \ +performance of their application." +HOMEPAGE = "https://github.com/embree/embree" + +LICENSE = "Apache-2.0 & syrah" +LIC_FILES_CHKSUM = "file://LICENSE.txt;md5=3b83ef96387f14655fc854ddc3c6bd57 \ + file://third-party-programs.txt;md5=f989f5b74cfff0d45d3ccf0e1366cbdc \ + file://common/math/transcendental.h;beginline=6;endline=8;md5=73380bb2ab6613b30b8464f114bd0ca8" + +inherit pkgconfig cmake + +S = "${WORKDIR}/git" + +SRC_URI = "git://github.com/embree/embree.git;protocol=https;branch=master" +SRCREV = "daa8de0e714e18ad5e5c9841b67c1950d9c91c51" + +COMPATIBLE_HOST = '(x86_64).*-linux' +COMPATIBLE_HOST:libc-musl = "null" + +DEPENDS = "tbb jpeg libpng glfw ispc-native" + +EXTRA_OECMAKE += " \ + -DEMBREE_IGNORE_CMAKE_CXX_FLAGS=OFF \ + -DEMBREE_MAX_ISA=DEFAULT \ + -DEMBREE_TUTORIALS=OFF \ + -DEMBREE_ISPC_SUPPORT=ON \ + -DEMBREE_ZIP_MODE=OFF \ + " diff --git a/dynamic-layers/openembedded-layer/recipes-oneapi/ipp/intel-oneapi-ipp_2021.10.0-653.bb b/dynamic-layers/openembedded-layer/recipes-oneapi/ipp/intel-oneapi-ipp_2021.10.0-653.bb new file mode 100644 index 00000000..1841253e --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-oneapi/ipp/intel-oneapi-ipp_2021.10.0-653.bb @@ -0,0 +1,35 @@ +DESCRIPTION = "Intel® Integrated Performance Primitives are production-ready \ + building blocks for cross-platform performance. Develop high-performance vision, \ + signal, security, and storage applications with this multithreaded software library." +HOMEPAGE = "https://software.intel.com/content/www/us/en/develop/tools/oneapi/components/ipp.html" + +LICENSE = "ISSL" + +MAXVER = "2021.10" + +LIC_FILES_CHKSUM = " \ + file://opt/intel/oneapi/ipp/${MAXVER}/share/doc/ipp/licensing/license.txt;md5=d7cdc92ed6c4de1263da879599ddc3e2 \ + file://opt/intel/oneapi/ipp/${MAXVER}/share/doc/ipp/licensing/third-party-programs.txt;md5=22bd13987911dcf790907041b43081f3 \ + " +SRC_URI = " \ + https://apt.repos.intel.com/oneapi/pool/main/intel-oneapi-ipp-${MAXVER}-${PV}_amd64.deb;subdir=${BPN};name=ipp \ + https://apt.repos.intel.com/oneapi/pool/main/intel-oneapi-ipp-common-devel-${MAXVER}-${PV}_all.deb;subdir=${BPN};name=headers \ + https://apt.repos.intel.com/oneapi/pool/main/intel-oneapi-ipp-common-${MAXVER}-${PV}_all.deb;subdir=${BPN};name=env \ + " + +SRC_URI[ipp.sha256sum] = "b51e45c6e691aa46c7136b0ab61f5abe346388433e017a30cf53fd23e92bea07" +SRC_URI[headers.sha256sum] = "342f37ab2f82bc9f4498435f848ee660591c2488b44d988bf6ee96b2a71fd005" +SRC_URI[env.sha256sum] = "731e8c28a3b8b757730cd874d0941de2eb744856128f24ade59d36c12b415bf6" + +S = "${WORKDIR}/${BPN}" + +COMPATIBLE_HOST:libc-musl = "null" + +inherit bin_package + +INHIBIT_PACKAGE_STRIP = "1" +INHIBIT_PACKAGE_DEBUG_SPLIT = "1" +INHIBIT_DEFAULT_DEPS = "" + +RDEPENDS:${PN} += "tbb setup-intel-oneapi-env" +INSANE_SKIP:${PN} += "ldflags dev-so" diff --git a/dynamic-layers/openembedded-layer/recipes-oneapi/mkl/intel-oneapi-mkl_2024.0.0-49656.bb b/dynamic-layers/openembedded-layer/recipes-oneapi/mkl/intel-oneapi-mkl_2024.0.0-49656.bb new file mode 100644 index 00000000..e3e93028 --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-oneapi/mkl/intel-oneapi-mkl_2024.0.0-49656.bb @@ -0,0 +1,74 @@ +SUMMARY = "Intel® oneAPI Math Kernel Library (oneMKL)" +DESCRIPTION = "The Intel® oneAPI Math Kernel Library (oneMKL) is a computing \ + math library of highly optimized and extensively parallelized routines \ + for applications that require maximum performance. oneMKL contains \ + the high-performance optimizations from the full Intel® Math Kernel Library \ + for CPU architectures (with C/Fortran programming language interfaces)\ + and adds to them a set of DPC++ programming language interfaces for \ + achieving performance on various CPU architectures \ + and Intel Graphics Technology for certain key functionalities." +HOMEPAGE = "https://software.intel.com/content/www/us/en/develop/tools/oneapi/components/onemkl.html" + +LICENSE = "ISSL" + +MKLMAINVER = "2024.0" + +LIC_FILES_CHKSUM = " \ + file://opt/intel/oneapi/mkl/${MKLMAINVER}/share/doc/mkl/licensing/license.txt;md5=8510d21bf355a76e378c3216c3929ccd \ + file://opt/intel/oneapi/mkl/${MKLMAINVER}/share/doc/mkl/licensing/third-party-programs-benchmarks.txt;md5=cb98e1a1f14c05ea85a979ea8982e7a4 \ + file://opt/intel/oneapi/mkl/${MKLMAINVER}/share/doc/mkl/licensing/third-party-programs-ipp.txt;md5=a4b2bf15e38f5c1267cdafed18bc0b09 \ + file://opt/intel/oneapi/mkl/${MKLMAINVER}/share/doc/mkl/licensing/third-party-programs-openmp.txt;md5=6b3c1aa2a11393060074c0346ce21e49 \ + file://opt/intel/oneapi/mkl/${MKLMAINVER}/share/doc/mkl/licensing/third-party-programs-safestring.txt;md5=c3aeee91c6d35a0f0753aed6c2633b82 \ + file://opt/intel/oneapi/mkl/${MKLMAINVER}/share/doc/mkl/licensing/third-party-programs.txt;md5=27de873e4084d62530fe828406b33ca9 \ + " + +SRC_URI = " \ + https://apt.repos.intel.com/oneapi/pool/main/intel-oneapi-mkl-${MKLMAINVER}-${PV}_amd64.deb;subdir=${BPN};name=runtime \ + https://apt.repos.intel.com/oneapi/pool/main/intel-oneapi-mkl-common-devel-${MKLMAINVER}-${PV}_all.deb;subdir=${BPN};name=common-devel \ + https://apt.repos.intel.com/oneapi/pool/main/intel-oneapi-mkl-devel-${MKLMAINVER}-${PV}_amd64.deb;subdir=${BPN};name=devel \ + https://apt.repos.intel.com/oneapi/pool/main/intel-oneapi-mkl-common-${MKLMAINVER}-${PV}_all.deb;subdir=${BPN};name=common-vars \ + " + +SRC_URI[runtime.sha256sum] = "10a86e24051d6ef4a80fd839c570e629190638a3c0ac9bcca99ab855f534b959" +SRC_URI[common-devel.sha256sum] = "adbf0ea946f63946d29b7f9c750c38a42ea7a65d8c81655b268aa2c7bb908192" +SRC_URI[devel.sha256sum] = "fab2a6f15e18bfd9b4d425f2703e4e98928c57f52c4feebc9ed886f097062e84" +SRC_URI[common-vars.sha256sum] = "ec2b67813739fa4a2895f63479a41acba2174afe2d0cb8a0c1c9119d1317d8ef" + +S = "${WORKDIR}/${BPN}" + +inherit bin_package + +do_install:append () { + install -d ${D}${bindir} + (cd ${D}${bindir} ; ln -s ../../opt/intel/oneapi/mkl/${MKLMAINVER}/bin/* .) + install -d ${D}${libdir} + (cd ${D}${libdir} ; ln -s ../../opt/intel/oneapi/mkl/${MKLMAINVER}/lib/intel64/*.so* .) + (cd ${D}${libdir} ; ln -s ../../opt/intel/oneapi/mkl/${MKLMAINVER}/lib/intel64/*.a* .) + install -d ${D}${libdir}/pkgconfig + (cd ${D}${libdir}/pkgconfig ; ln -s ../../../opt/intel/oneapi/mkl/${MKLMAINVER}/lib/pkgconfig/* .) + install -d ${D}${libdir}/cmake + (cd ${D}${libdir}/cmake ; ln -s ../../../opt/intel/oneapi/mkl/${MKLMAINVER}/lib/cmake/* .) + + install -d ${D}${includedir} + find ${D}/opt/intel/oneapi/mkl/${MKLMAINVER}/include/ -mindepth 1 -maxdepth 1 -type d -printf '%f\n' | while read srcdir; do + install -d ${D}${includedir}/$srcdir + (cd ${D}${includedir} ; ln -s ../../opt/intel/oneapi/mkl/${MKLMAINVER}/include/$srcdir/* ./$srcdir/) + done + + find ${D}/opt/intel/oneapi/mkl/${MKLMAINVER}/include/ -mindepth 1 -maxdepth 1 -type f -printf '%f\n' | while read srcfile; do + (cd ${D}${includedir} ; ln -s ../../opt/intel/oneapi/mkl/${MKLMAINVER}/include/$srcfile .) + done +} + +AUTO_LIBNAME_PKGS = "" +INHIBIT_PACKAGE_STRIP = "1" +INHIBIT_PACKAGE_DEBUG_SPLIT = "1" + +RDEPENDS:${PN} += "bash tbb intel-oneapi-dpcpp-cpp-runtime setup-intel-oneapi-env virtual-opencl-icd" +INSANE_SKIP:${PN} = "ldflags textrel dev-so staticdev arch already-stripped" + +FILES:${PN}-staticdev += "/opt/intel/oneapi/mkl/${MKLMAINVER}/lib/*.a*" + +SKIP_FILEDEPS:${PN} = '1' + +SYSROOT_DIRS += "/opt" diff --git a/dynamic-layers/openembedded-layer/recipes-oneapi/oidn/oidn_2.1.0.bb b/dynamic-layers/openembedded-layer/recipes-oneapi/oidn/oidn_2.1.0.bb new file mode 100644 index 00000000..4ea6ffbe --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-oneapi/oidn/oidn_2.1.0.bb @@ -0,0 +1,23 @@ +SUMMARY = "Intel® Open Image Denoise" +DESCRIPTION = "Intel Open Image Denoise is an open source library of \ +high-performance, high-quality denoising filters for images \ +rendered with ray tracing. Intel Open Image Denoise is part \ +of the Intel® oneAPI Rendering Toolkit" +HOMEPAGE = "https://www.openimagedenoise.org/" + +LICENSE = "Apache-2.0 & BSD-3-Clause" +LIC_FILES_CHKSUM = "file://LICENSE.txt;md5=3b83ef96387f14655fc854ddc3c6bd57 \ + file://external/mkl-dnn/LICENSE;md5=b48e3de3bfd47c27882a0d85b20823f5 \ + file://external/mkl-dnn/src/cpu/x64/xbyak/COPYRIGHT;md5=3b9bf048d063d54cdb28964db558bcc7 \ + file://external/mkl-dnn/src/common/ittnotify/LICENSE.BSD;md5=e671ff178b24a95a382ba670503c66fb \ + file://weights/LICENSE.txt;md5=3b83ef96387f14655fc854ddc3c6bd57" + +SRC_URI = "https://github.com/OpenImageDenoise/${BPN}/releases/download/v${PV}/${BP}.src.tar.gz\ + " +SRC_URI[sha256sum] = "ce144ba582ff36563d9442ee07fa2a4d249bc85aa93e5b25fc527ff4ee755ed6" + +inherit cmake + +DEPENDS += "tbb ispc-native" + +UPSTREAM_CHECK_URI = "https://github.com/OpenImageDenoise/oidn/releases" diff --git a/dynamic-layers/openembedded-layer/recipes-oneapi/openvkl/openvkl_2.0.1.bb b/dynamic-layers/openembedded-layer/recipes-oneapi/openvkl/openvkl_2.0.1.bb new file mode 100644 index 00000000..a2d709a1 --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-oneapi/openvkl/openvkl_2.0.1.bb @@ -0,0 +1,42 @@ +SUMMARY = "Intel(R) Open Volume Kernel Library" +DESCRIPTION = "Intel(R) Open Volume Kernel Library (Intel(R) Open VKL) is a \ +collection of high-performance volume computation kernels. The target users \ +of Open VKL are graphics application engineers who want to improve the \ +performance of their volume rendering applications by leveraging Open VKL’s \ +performance-optimized kernels, which include volume traversal and sampling \ +functionality for a variety of volumetric data formats. The kernels are optimized \ +for the latest Intel(R) processors with support for SSE, AVX, AVX2, and AVX-512 \ +instructions." +HOMEPAGE = "https://www.openvkl.org/" + +LICENSE = "Apache-2.0 & BSD-3-Clause & MIT & Zlib" +LIC_FILES_CHKSUM = "file://LICENSE.txt;md5=3b83ef96387f14655fc854ddc3c6bd57 \ + file://third-party-programs.txt;md5=69ec7caf49616c471161b921f53d5ec0 \ + file://testing/external/half.hpp;beginline=1;endline=17;md5=4b60058493630c3bd0ef145470f04a7b" + +inherit pkgconfig cmake + +S = "${WORKDIR}/git" + +SRC_URI = "git://github.com/openvkl/openvkl.git;protocol=https;branch=master \ + " +SRCREV = "8c6ba526813b871a624cb9d73d4cbb689ac7f4ce" + +COMPATIBLE_HOST = '(x86_64).*-linux' +COMPATIBLE_HOST:libc-musl = "null" + +DEPENDS = "ispc ispc-native rkcommon embree" + +EXTRA_OECMAKE += " \ + -DISPC_EXECUTABLE=${STAGING_BINDIR_NATIVE}/ispc \ + " +EXTRA_OECMAKE:intel-corei7-64 += " \ + -DOPENVKL_ISA_AVX=OFF \ + -DOPENVKL_ISA_AVX2=OFF \ + -DOPENVKL_ISA_AVX512SKX=OFF \ + " + +PACKAGES =+ "${PN}-examples" +FILES:${PN}-examples = "\ + ${bindir} \ + " diff --git a/dynamic-layers/openembedded-layer/recipes-oneapi/ospray/ospray/0001-Fix-GCC11-Compile-Error-in-benchmark_register.h.patch b/dynamic-layers/openembedded-layer/recipes-oneapi/ospray/ospray/0001-Fix-GCC11-Compile-Error-in-benchmark_register.h.patch new file mode 100644 index 00000000..c2136321 --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-oneapi/ospray/ospray/0001-Fix-GCC11-Compile-Error-in-benchmark_register.h.patch @@ -0,0 +1,28 @@ +From 09e6b6615d9a16e9555eff9d569356e32bd26aa4 Mon Sep 17 00:00:00 2001 +From: benradel <86351445+benradel@users.noreply.github.com> +Date: Wed, 23 Jun 2021 11:56:03 +0200 +Subject: [PATCH] Fix GCC11 Compile Error in benchmark_register.h + +Fix GCC11 compilation error due to missing header "limits" in benchmark_register.h + +Upstream-Status: Submitted [https://github.com/ospray/ospray/pull/484/files] +Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> +--- + apps/common/external/benchmark/src/benchmark_register.h | 1 + + 1 file changed, 1 insertion(+) + +diff --git a/apps/common/external/benchmark/src/benchmark_register.h b/apps/common/external/benchmark/src/benchmark_register.h +index 61377d742..8f1bb7795 100644 +--- a/apps/common/external/benchmark/src/benchmark_register.h ++++ b/apps/common/external/benchmark/src/benchmark_register.h +@@ -2,6 +2,7 @@ + #define BENCHMARK_REGISTER_H + + #include <vector> ++#include <limits> + + #include "check.h" + +-- +2.32.0 + diff --git a/dynamic-layers/openembedded-layer/recipes-oneapi/ospray/ospray_3.1.0.bb b/dynamic-layers/openembedded-layer/recipes-oneapi/ospray/ospray_3.1.0.bb new file mode 100644 index 00000000..3e03dcc2 --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-oneapi/ospray/ospray_3.1.0.bb @@ -0,0 +1,37 @@ +SUMMARY = "Intel OSPray, Ray Tracing based Rendering Engine" +DESCRIPTION = "Intel OSPRay is an open source, scalable, and portable ray \ +tracing engine for high-performance, high-fidelity visualization on \ +Intel Architecture CPUs." +HOMEPAGE = "https://www.ospray.org/" + +LICENSE = "Apache-2.0 & BSD-3-Clause & MIT & Zlib" +LIC_FILES_CHKSUM = "file://LICENSE.txt;md5=1dece7821bf3fd70fe1309eaa37d52a2 \ + file://third-party-programs.txt;md5=e37b77e3bd997abccc359c710fb1f1db \ + " + +inherit pkgconfig cmake + +S = "${WORKDIR}/git" + +SRC_URI = "git://github.com/ospray/ospray.git;protocol=https;branch=master \ + " +SRCREV = "f2a61c2eb58ccd666e34abfdb0fd7049ea073194" + +COMPATIBLE_HOST = '(x86_64).*-linux' +COMPATIBLE_HOST:libc-musl = "null" + +DEPENDS = "rkcommon ispc ispc-native openvkl embree" + +EXTRA_OECMAKE += " \ + -DISPC_EXECUTABLE=${STAGING_BINDIR_NATIVE}/ispc \ + -DOSPRAY_ENABLE_APPS_BENCHMARK=OFF \ + -DOSPRAY_ENABLE_APPS_TESTING=OFF \ + -DOSPRAY_ENABLE_APPS_EXAMPLES=OFF \ + " + +PACKAGES =+ "${PN}-apps" +FILES:${PN}-apps = "\ + ${bindir} \ + " + +UPSTREAM_CHECK_GITTAGREGEX = "^v(?P<pver>(\d+(\.\d+)+))$" diff --git a/dynamic-layers/openembedded-layer/recipes-oneapi/rkcommon/rkcommon_1.13.0.bb b/dynamic-layers/openembedded-layer/recipes-oneapi/rkcommon/rkcommon_1.13.0.bb new file mode 100644 index 00000000..fe6b23ea --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-oneapi/rkcommon/rkcommon_1.13.0.bb @@ -0,0 +1,19 @@ +SUMMARY = "rkcommon - C++/CMake infrastructure" +DESCRIPTION = "A common set of C++ infrastructure and CMake utilities \ +used by various components of Intel® oneAPI Rendering Toolkit." +HOMEPAGE = "https://github.com/ospray/rkcommon" + +LICENSE = "Apache-2.0" +LIC_FILES_CHKSUM = "file://LICENSE.txt;md5=3b83ef96387f14655fc854ddc3c6bd57" + +inherit pkgconfig cmake + +S = "${WORKDIR}/git" + +SRC_URI = "git://github.com/ospray/rkcommon.git;protocol=https;branch=master \ + " +SRCREV = "7ebfa0765ea590767202b328e7da38102c2f5a15" + +DEPENDS = "tbb" + +EXTRA_OECMAKE += " -DBUILD_TESTING=OFF" diff --git a/dynamic-layers/openembedded-layer/recipes-support/ipmctl/ipmctl/0001-CMakeLists-disable-Werror.patch b/dynamic-layers/openembedded-layer/recipes-support/ipmctl/ipmctl/0001-CMakeLists-disable-Werror.patch new file mode 100644 index 00000000..4474aa3c --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-support/ipmctl/ipmctl/0001-CMakeLists-disable-Werror.patch @@ -0,0 +1,38 @@ +From ef56be8e6bf2ea273cbeb960f3131164e7b517b6 Mon Sep 17 00:00:00 2001 +From: Anuj Mittal <anuj.mittal@intel.com> +Date: Wed, 18 May 2022 12:24:03 +0800 +Subject: [PATCH] CMakeLists: disable Werror + +gcc12 highlights minor warnings that cause the build to fail. Ignore +those until they aren't fixed upstream. + +Upstream-Status: Inappropriate + +Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> +--- + CMakeLists.txt | 4 ++-- + 1 file changed, 2 insertions(+), 2 deletions(-) + +diff --git a/CMakeLists.txt b/CMakeLists.txt +index 418483f7..c019fea0 100644 +--- a/CMakeLists.txt ++++ b/CMakeLists.txt +@@ -146,7 +146,7 @@ else() + set(CMAKE_CXX_FLAGS_RELEASE "${CMAKE_CXX_FLAGS_RELEASE} -O2 -fno-strict-aliasing -D_FORTIFY_SOURCE=2") + if(LNX_BUILD) + #A few warnings yet to resolve under esx +- set(CMAKE_C_FLAGS_RELEASE "${CMAKE_C_FLAGS_RELEASE} -Werror") ++ #set(CMAKE_C_FLAGS_RELEASE "${CMAKE_C_FLAGS_RELEASE} -Werror") + set(CMAKE_CXX_FLAGS_RELEASE "${CMAKE_CXX_FLAGS_RELEASE} -Werror") + endif() + set(CMAKE_C_FLAGS_DEBUG "${CMAKE_C_FLAGS_DEBUG} -O0 -ggdb") +@@ -791,4 +791,4 @@ if(ESX_BUILD) + include(CMake/esx.cmake) + endif() + +-add_subdirectory(src/os/nvm_api_sample) +\ No newline at end of file ++add_subdirectory(src/os/nvm_api_sample) +-- +2.35.3 + diff --git a/dynamic-layers/openembedded-layer/recipes-support/ipmctl/ipmctl/0001-Ignore-STATIC_ASSERT-and-NULL-definition-so-we-can-c.patch b/dynamic-layers/openembedded-layer/recipes-support/ipmctl/ipmctl/0001-Ignore-STATIC_ASSERTs-and-NULL-define-for-os-and-ut-builds.patch index 4a3b26c7..11305e83 100644 --- a/dynamic-layers/openembedded-layer/recipes-support/ipmctl/ipmctl/0001-Ignore-STATIC_ASSERT-and-NULL-definition-so-we-can-c.patch +++ b/dynamic-layers/openembedded-layer/recipes-support/ipmctl/ipmctl/0001-Ignore-STATIC_ASSERTs-and-NULL-define-for-os-and-ut-builds.patch @@ -1,56 +1,50 @@ -From cc0e412ad05574d22938586172c56863666eb2a5 Mon Sep 17 00:00:00 2001 -From: Nolan Hergert <nolan.hergert@intel.com> -Date: Mon, 5 Apr 2021 11:15:09 +0800 -Subject: [PATCH] Ignore STATIC_ASSERT and NULL definition so we can compile in - OS - -Upstream-Status: Backport [from ipmctl patches; https://github.com/intel/ipmctl/blob/development/src/os/patches/0001-Ignore-STATIC_ASSERTs-and-NULL-define-for-os-and-ut-builds.patch] - -Signed-off-by: Nolan Hergert <nolan.hergert@intel.com> -Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> +Subject: [PATCH] [PATCH]: ignore static asserts and null define for os and ut + builds +Upstream-Status: Backport +Signed-off-by: Teoh Suh Haw <suh.haw.teoh@intel.com> --- - MdePkg/Include/Base.h | 13 ++++++++++++- - 1 file changed, 12 insertions(+), 1 deletion(-) + MdePkg/Include/Base.h | 12 ++++++++++++ + 1 file changed, 12 insertions(+) diff --git a/MdePkg/Include/Base.h b/MdePkg/Include/Base.h -index 8e4271f6ea..c39511a064 100644 +index d209e6de28..6e61b8f3d3 100644 --- a/MdePkg/Include/Base.h +++ b/MdePkg/Include/Base.h -@@ -318,7 +318,11 @@ struct _LIST_ENTRY { - /// - /// NULL pointer (VOID *) - /// +@@ -316,8 +316,12 @@ struct _LIST_ENTRY { + #define NULL __null + #endif + #else +#ifndef OS_BUILD +#ifndef UNIT_TEST_UEFI_BUILD #define NULL ((VOID *) 0) + #endif +#endif +#endif - + // // Null character -@@ -809,7 +813,8 @@ typedef UINTN *BASE_LIST; - // Verify that ProcessorBind.h produced UEFI Data Types that are compliant with +@@ -779,6 +783,8 @@ typedef UINTN *BASE_LIST; // Section 2.3.1 of the UEFI 2.3 Specification. // -- + +#ifndef OS_BUILD +#ifndef UNIT_TEST_UEFI_BUILD STATIC_ASSERT (sizeof (BOOLEAN) == 1, "sizeof (BOOLEAN) does not meet UEFI Specification Data Type requirements"); STATIC_ASSERT (sizeof (INT8) == 1, "sizeof (INT8) does not meet UEFI Specification Data Type requirements"); STATIC_ASSERT (sizeof (UINT8) == 1, "sizeof (UINT8) does not meet UEFI Specification Data Type requirements"); -@@ -821,6 +826,8 @@ STATIC_ASSERT (sizeof (INT64) == 8, "sizeof (INT64) does not meet UEFI Specifi - STATIC_ASSERT (sizeof (UINT64) == 8, "sizeof (UINT64) does not meet UEFI Specification Data Type requirements"); - STATIC_ASSERT (sizeof (CHAR8) == 1, "sizeof (CHAR8) does not meet UEFI Specification Data Type requirements"); +@@ -792,6 +798,8 @@ STATIC_ASSERT (sizeof (CHAR8) == 1, "sizeof (CHAR8) does not meet UEFI Specifi STATIC_ASSERT (sizeof (CHAR16) == 2, "sizeof (CHAR16) does not meet UEFI Specification Data Type requirements"); + STATIC_ASSERT (sizeof (L'A') == 2, "sizeof (L'A') does not meet UEFI Specification Data Type requirements"); + STATIC_ASSERT (sizeof (L"A") == 4, "sizeof (L\"A\") does not meet UEFI Specification Data Type requirements"); +#endif +#endif - + // // The following three enum types are used to verify that the compiler -@@ -841,9 +848,13 @@ typedef enum { +@@ -812,9 +820,13 @@ typedef enum { __VerifyUint32EnumValue = 0xffffffff } __VERIFY_UINT32_ENUM_SIZE; - + +#ifndef OS_BUILD +#ifndef UNIT_TEST_UEFI_BUILD STATIC_ASSERT (sizeof (__VERIFY_UINT8_ENUM_SIZE) == 4, "Size of enum does not meet UEFI Specification Data Type requirements"); @@ -58,9 +52,8 @@ index 8e4271f6ea..c39511a064 100644 STATIC_ASSERT (sizeof (__VERIFY_UINT32_ENUM_SIZE) == 4, "Size of enum does not meet UEFI Specification Data Type requirements"); +#endif +#endif - + /** Macro that returns a pointer to the data structure that contains a specified field of --- -2.17.1 - +-- +2.37.3 diff --git a/dynamic-layers/openembedded-layer/recipes-support/ipmctl/ipmctl_03.00.00.0302.bb b/dynamic-layers/openembedded-layer/recipes-support/ipmctl/ipmctl_03.00.00.0485.bb index 688ad108..c4743d1c 100644 --- a/dynamic-layers/openembedded-layer/recipes-support/ipmctl/ipmctl_03.00.00.0302.bb +++ b/dynamic-layers/openembedded-layer/recipes-support/ipmctl/ipmctl_03.00.00.0485.bb @@ -13,27 +13,35 @@ BUGTRACKER = "https://github.com/intel/ipmctl/issues" LICENSE = "BSD-3-Clause | BSD-2-Clause" LIC_FILES_CHKSUM = "file://LICENSE;md5=72b9da60da6219d612ce30b746a0fe71 \ - file://../edk2/License.txt;md5=2b415520383f7964e96700ae12b4570a" + file://edk2/License.txt;md5=6123e5bf044a66db96c4ce88a36b2d08" -SRC_URI = "git://github.com/intel/ipmctl.git;protocol=https;branch=development;name=ipmctl; \ - git://github.com/tianocore/edk2.git;protocol=https;name=edk2;destsuffix=edk2; \ - file://0001-Ignore-STATIC_ASSERT-and-NULL-definition-so-we-can-c.patch;patchdir=../edk2 \ +SRC_URI = "git://github.com/intel/ipmctl.git;protocol=https;branch=master;name=ipmctl; \ + git://github.com/tianocore/edk2.git;protocol=https;name=edk2;destsuffix=git/edk2;branch=master \ + file://0001-Ignore-STATIC_ASSERTs-and-NULL-define-for-os-and-ut-builds.patch;patchdir=edk2 \ + file://0001-CMakeLists-disable-Werror.patch \ " -SRCREV_ipmctl = "4579c7ec127d6c678ce25136f928b66c7ba51097" -#tag: edk2-stable202102 -SRCREV_edk2 = "ef91b07388e1c0a50c604e5350eeda98428ccea6" +SRCREV_ipmctl = "c75bd840ea7820c8f93a5488fcff75d08beedd51" +#tag edk2-stable202302 +SRCREV_edk2 = "f80f052277c88a67c55e107b550f504eeea947d3" +SRCREV_FORMAT = "ipmctl_edk2" S = "${WORKDIR}/git" -inherit cmake +inherit cmake dos2unix -DEPENDS = "ndctl" +DEPENDS = "ndctl pkgconfig-native" EXTRA_OECMAKE = "-DRELEASE=ON" -do_configure_prepend() { +do_configure:prepend() { for dir in BaseTools MdeModulePkg MdePkg ShellPkg ; do - ln -sf ../edk2/${dir} ${S} + ln -sf edk2/${dir} ${S} done } + +do_install:append() { + # Remove /var/log/ipmctl as anything created in /var/log will not be + # available when tmpfs is mounted at /var/volatile/log. + rm -rf ${D}${localstatedir}/log +} diff --git a/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0001-CPU-Solving-the-build-failure-caused-by-setting-the-.patch b/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0001-CPU-Solving-the-build-failure-caused-by-setting-the-.patch new file mode 100644 index 00000000..d1851406 --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0001-CPU-Solving-the-build-failure-caused-by-setting-the-.patch @@ -0,0 +1,43 @@ +From b8c3bae71e6d3417ade5cb537cb1785fd75a75c8 Mon Sep 17 00:00:00 2001 +From: Razvan Apetroaie <117895604+razvanapetroaie@users.noreply.github.com> +Date: Tue, 20 Feb 2024 02:28:14 +0200 +Subject: [PATCH] [CPU] Solving the build failure caused by setting the + "ENABLE_OV_ONNX_FRONTEND" option to "OFF" (#22934) + +### Details: +See the [ticket +description](https://jira.devtools.intel.com/browse/CVS-132119). The +solution was checked only on a local setup on Ubuntu, if there's a way +to check that using the CI please let me know (or if you could run the +job and paste the link in the comments I would be grateful). + +Disclaimer: I'm not a CPU plugin developer, so I can't tell for sure if +this is the ideal fix and no side effects are introduced. Please take +that into account when reviewing/merging. + +### Tickets: + - [CVS-132119](https://jira.devtools.intel.com/browse/CVS-132119) + +Upstream-Status: Backport [https://github.com/openvinotoolkit/openvino/commit/b8c3bae71e6d3417ade5cb537cb1785fd75a75c8] + +Signed-off-by: Yogesh Tyagi <yogesh.tyagi@intel.com> +--- + src/plugins/intel_cpu/tests/functional/CMakeLists.txt | 2 +- + 1 file changed, 1 insertion(+), 1 deletion(-) + +diff --git a/src/plugins/intel_cpu/tests/functional/CMakeLists.txt b/src/plugins/intel_cpu/tests/functional/CMakeLists.txt +index db5ae8d01c..6941cb528d 100644 +--- a/src/plugins/intel_cpu/tests/functional/CMakeLists.txt ++++ b/src/plugins/intel_cpu/tests/functional/CMakeLists.txt +@@ -16,7 +16,7 @@ set(LINK_LIBRARIES funcSharedTests cpuSpecificRtInfo openvino::snippets ov_snipp + if(ENABLE_OV_ONNX_FRONTEND) + list(APPEND DEFINES TEST_MODELS="${TEST_MODEL_ZOO}") + else() +- set(EXCLUDED_SOURCE_PATHS ${CMAKE_CURRENT_SOURCE_DIR}/extension ${CMAKE_CURRENT_SOURCE_DIR}/shared_tests_instances/onnx) ++ set(EXCLUDED_SOURCE_PATHS ${CMAKE_CURRENT_SOURCE_DIR}/custom/extension ${CMAKE_CURRENT_SOURCE_DIR}/shared_tests_instances/onnx) + endif() + + if(NOT (ARM OR AARCH64)) +-- +2.34.1 + diff --git a/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0001-cmake-fix-build-when-using-sysroot.patch b/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0001-cmake-fix-build-when-using-sysroot.patch new file mode 100644 index 00000000..d251f21b --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0001-cmake-fix-build-when-using-sysroot.patch @@ -0,0 +1,32 @@ +From b5dfcf8bc1245e804c847745c237068eb6f19931 Mon Sep 17 00:00:00 2001 +From: Yogesh Tyagi <yogesh.tyagi@intel.com> +Date: Fri, 15 Mar 2024 16:28:41 +0800 +Subject: [PATCH] cmake: fix build when using sysroot + +When cross-compiling against a sysroot, system headers will not be at a place +that starts with /usr. Update conditional check to exclude directories +which have "/usr/include" in them to not add <sysroot>/usr/include as well. + +Upstream-Status: Submitted [https://github.com/openvinotoolkit/openvino/pull/23486] + +Signed-off-by: Yogesh Tyagi <yogesh.tyagi@intel.com> +--- + src/cmake/ov_parallel.cmake | 2 +- + 1 file changed, 1 insertion(+), 1 deletion(-) + +diff --git a/src/cmake/ov_parallel.cmake b/src/cmake/ov_parallel.cmake +index 3793db98e30..819d3410531 100644 +--- a/src/cmake/ov_parallel.cmake ++++ b/src/cmake/ov_parallel.cmake +@@ -296,7 +296,7 @@ function(ov_set_threading_interface_for TARGET_NAME) + if(include_directories) + foreach(include_directory IN LISTS include_directories) + # cannot include /usr/include headers as SYSTEM +- if(NOT "${include_directory}" MATCHES "^/usr.*$") ++ if(NOT "${include_directory}" MATCHES ".*/usr/include.*$") + target_include_directories(${TARGET_NAME} SYSTEM + ${LINK_TYPE} $<BUILD_INTERFACE:${include_directory}>) + else() +-- +2.34.1 + diff --git a/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0001-cmake-yocto-specific-tweaks-to-the-build-process.patch b/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0001-cmake-yocto-specific-tweaks-to-the-build-process.patch new file mode 100644 index 00000000..7ab31309 --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0001-cmake-yocto-specific-tweaks-to-the-build-process.patch @@ -0,0 +1,90 @@ +From bfcf5ae581ca4e7266cf7dc65b1c71754cd78cc0 Mon Sep 17 00:00:00 2001 +From: Anuj Mittal <anuj.mittal@intel.com> +Date: Wed, 29 Nov 2023 12:42:57 +0530 +Subject: [PATCH 1/4] cmake: yocto specific tweaks to the build process + +* Dont try to detect glibc version as that doesn't work when cross compiling. +* Install sample binaries as well. +* Dont strip binaries. +* Dont try to write triggers for CPack. We package ourselves. +* Fix the installation path for Python modules when baselib = lib64. + +Upstream-Status: Inappropriate + +Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> +--- + cmake/developer_package/packaging/rpm/rpm.cmake | 2 +- + cmake/developer_package/target_flags.cmake | 2 +- + samples/cpp/CMakeLists.txt | 6 +++--- + src/bindings/python/CMakeLists.txt | 2 +- + src/bindings/python/wheel/setup.py | 1 - + 5 files changed, 6 insertions(+), 7 deletions(-) + +diff --git a/cmake/developer_package/packaging/rpm/rpm.cmake b/cmake/developer_package/packaging/rpm/rpm.cmake +index a7c0ec2cf61..40448e8a962 100644 +--- a/cmake/developer_package/packaging/rpm/rpm.cmake ++++ b/cmake/developer_package/packaging/rpm/rpm.cmake +@@ -154,7 +154,7 @@ ov_rpm_specific_settings() + # needed to add triggers for packages with libraries + set(def_triggers "${OpenVINO_BINARY_DIR}/_CPack_Packages/triggers") + set(triggers_content "# /bin/sh -p\n/sbin/ldconfig\n") +-file(WRITE "${def_triggers}" "${triggers_content}") ++#file(WRITE "${def_triggers}" "${triggers_content}") + + # + # Functions helpful for packaging your modules with RPM cpack +diff --git a/cmake/developer_package/target_flags.cmake b/cmake/developer_package/target_flags.cmake +index 29f23e713e1..942dd445b19 100644 +--- a/cmake/developer_package/target_flags.cmake ++++ b/cmake/developer_package/target_flags.cmake +@@ -145,4 +145,4 @@ function(ov_glibc_version) + endif() + endfunction() + +-ov_glibc_version() ++#ov_glibc_version() +diff --git a/samples/cpp/CMakeLists.txt b/samples/cpp/CMakeLists.txt +index c814cc37e2c..431e7bd2ed3 100644 +--- a/samples/cpp/CMakeLists.txt ++++ b/samples/cpp/CMakeLists.txt +@@ -206,9 +206,9 @@ macro(ov_add_sample) + target_link_libraries(${SAMPLE_NAME} PRIVATE ${ov_link_libraries} Threads::Threads ${SAMPLE_DEPENDENCIES}) + + install(TARGETS ${SAMPLE_NAME} +- RUNTIME DESTINATION samples_bin/ +- COMPONENT samples_bin +- EXCLUDE_FROM_ALL) ++ DESTINATION ${CMAKE_INSTALL_BINDIR} ++ COMPONENT samples_bin) ++ + + # create global target with all samples / demo apps + if(NOT TARGET ov_samples) +diff --git a/src/bindings/python/CMakeLists.txt b/src/bindings/python/CMakeLists.txt +index 58ff9b74302..4763994ba56 100644 +--- a/src/bindings/python/CMakeLists.txt ++++ b/src/bindings/python/CMakeLists.txt +@@ -356,7 +356,7 @@ if(ENABLE_PYTHON_PACKAGING) + endif() + + set(python_package_prefix "${CMAKE_CURRENT_BINARY_DIR}/install_${pyversion}") +- set(install_lib "${python_package_prefix}/lib/${python_versioned_folder}/${ov_site_packages}") ++ set(install_lib "${python_package_prefix}/${CMAKE_INSTALL_LIBDIR}/${python_versioned_folder}/${ov_site_packages}") + set(meta_info_subdir "openvino-${OpenVINO_VERSION}-py${python_xy}.egg-info") + set(meta_info_file "${install_lib}/${meta_info_subdir}/PKG-INFO") + +diff --git a/src/bindings/python/wheel/setup.py b/src/bindings/python/wheel/setup.py +index 4b056912212..5f05d891310 100644 +--- a/src/bindings/python/wheel/setup.py ++++ b/src/bindings/python/wheel/setup.py +@@ -270,7 +270,6 @@ class CustomBuild(build): + self.spawn(["cmake", "--install", binary_dir, + "--prefix", prefix, + "--config", CONFIG, +- "--strip", + "--component", cpack_comp_name]) + + def run(self): +-- +2.34.1 + diff --git a/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0001-dont-install-licenses-and-version-file.patch b/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0001-dont-install-licenses-and-version-file.patch deleted file mode 100644 index 2a2c8b87..00000000 --- a/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0001-dont-install-licenses-and-version-file.patch +++ /dev/null @@ -1,54 +0,0 @@ -From a60710c910e6c995aef9dd36757bf040bf24acf5 Mon Sep 17 00:00:00 2001 -From: Anuj Mittal <anuj.mittal@intel.com> -Date: Tue, 9 Jun 2020 11:35:59 +0800 -Subject: [PATCH] don't install licenses and version - -Upstream-Status: Inappropriate - -Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> ---- - ngraph/CMakeLists.txt | 6 ++---- - ngraph/test/runtime/CMakeLists.txt | 2 ++ - 2 files changed, 4 insertions(+), 4 deletions(-) - -diff --git a/ngraph/CMakeLists.txt b/ngraph/CMakeLists.txt -index 99f32a6181..b90aa382b9 100644 ---- a/ngraph/CMakeLists.txt -+++ b/ngraph/CMakeLists.txt -@@ -490,7 +490,7 @@ if (NGRAPH_EXPORT_TARGETS_ENABLE) - install(EXPORT ngraphTargets - FILE ngraphTargets.cmake - NAMESPACE ngraph:: -- DESTINATION ${NGRAPH_COMPONENT_PREFIX}cmake -+ DESTINATION ${CMAKE_INSTALL_LIBDIR}/cmake/ngraph - COMPONENT ngraph) - - configure_package_config_file(${CMAKE_CURRENT_SOURCE_DIR}/cmake/share/ngraphConfig.cmake.in -@@ -503,7 +503,7 @@ if (NGRAPH_EXPORT_TARGETS_ENABLE) - - install(FILES ${CMAKE_CURRENT_BINARY_DIR}/ngraphConfig.cmake - ${CMAKE_CURRENT_BINARY_DIR}/ngraphConfigVersion.cmake -- DESTINATION ${NGRAPH_COMPONENT_PREFIX}cmake -+ DESTINATION ${CMAKE_INSTALL_LIBDIR}/cmake/ngraph - COMPONENT ngraph) - endif() - -@@ -518,5 +518,3 @@ add_subdirectory(test) - if (NGRAPH_PYTHON_BUILD_ENABLE) - add_subdirectory(python) - endif() -- --install(FILES ${CMAKE_CURRENT_BINARY_DIR}/VERSION DESTINATION "${NGRAPH_COMPONENT_PREFIX}." COMPONENT ngraph) -diff --git a/ngraph/test/runtime/CMakeLists.txt b/ngraph/test/runtime/CMakeLists.txt -index d27ea8090c..17a3843f7d 100644 ---- a/ngraph/test/runtime/CMakeLists.txt -+++ b/ngraph/test/runtime/CMakeLists.txt -@@ -80,3 +80,5 @@ target_include_directories(ngraph_backend PUBLIC ${CMAKE_CURRENT_SOURCE_DIR}) - - add_subdirectory(interpreter) - add_subdirectory(ie) -+ -+install(TARGETS ngraph_backend DESTINATION ${NGRAPH_INSTALL_LIB}) --- -2.29.2 - diff --git a/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0001-inference-engine-use-system-installed-packages.patch b/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0001-inference-engine-use-system-installed-packages.patch deleted file mode 100644 index 1f64ce21..00000000 --- a/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0001-inference-engine-use-system-installed-packages.patch +++ /dev/null @@ -1,172 +0,0 @@ -From a8b6a2e251c42a393218174c6737208760d2e1c1 Mon Sep 17 00:00:00 2001 -From: Anuj Mittal <anuj.mittal@intel.com> -Date: Wed, 10 Jun 2020 09:43:55 +0800 -Subject: [PATCH] inference-engine: use system installed packages - -Use installed versions of ade, protobuf and tbb. - -Upstream-Status: Pending - -Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> ---- - cmake/dependencies.cmake | 18 ------------------ - inference-engine/CMakeLists.txt | 2 +- - inference-engine/cmake/ie_parallel.cmake | 7 +------ - .../src/inference_engine/CMakeLists.txt | 19 ------------------- - .../common_test_utils/CMakeLists.txt | 4 +--- - .../unit_test_utils/CMakeLists.txt | 3 +-- - inference-engine/thirdparty/CMakeLists.txt | 5 ++--- - 7 files changed, 6 insertions(+), 52 deletions(-) - -diff --git a/cmake/dependencies.cmake b/cmake/dependencies.cmake -index 6a44c8769e..a128eb63a1 100644 ---- a/cmake/dependencies.cmake -+++ b/cmake/dependencies.cmake -@@ -13,25 +13,7 @@ if(CMAKE_CROSSCOMPILING AND NGRAPH_ONNX_IMPORT_ENABLE) - - set(protoc_version "3.7.1") - if(CMAKE_HOST_SYSTEM_NAME MATCHES Linux) -- RESOLVE_DEPENDENCY(SYSTEM_PROTOC_ROOT -- ARCHIVE_LIN "protoc-${protoc_version}-linux-x86_64.tar.gz" -- TARGET_PATH "${TEMP}/protoc-${protoc_version}-linux-x86_64") -- debug_message(STATUS "host protoc-${protoc_version} root path = " ${SYSTEM_PROTOC_ROOT}) - else() - message(FATAL_ERROR "Unsupported host system (${CMAKE_HOST_SYSTEM_NAME}) and arch (${CMAKE_HOST_SYSTEM_PROCESSOR}) for cross-compilation") - endif() -- -- reset_deps_cache(SYSTEM_PROTOC) -- -- message("${SYSTEM_PROTOC_ROOT}/bin") -- find_program( -- SYSTEM_PROTOC -- NAMES protoc -- PATHS "${SYSTEM_PROTOC_ROOT}/bin" -- NO_DEFAULT_PATH) -- if(NOT SYSTEM_PROTOC) -- message(FATAL_ERROR "[ONNX IMPORTER] Missing host protoc binary") -- endif() -- -- update_deps_cache(SYSTEM_PROTOC "${SYSTEM_PROTOC}" "Path to host protoc for ONNX Importer") - endif() -diff --git a/inference-engine/CMakeLists.txt b/inference-engine/CMakeLists.txt -index a3e4f74c40..76b09683dd 100644 ---- a/inference-engine/CMakeLists.txt -+++ b/inference-engine/CMakeLists.txt -@@ -56,7 +56,7 @@ function(ie_developer_export) - APPEND FILE "${CMAKE_BINARY_DIR}/targets_developer.cmake") - - # Custom target to build only Inference Engine Developer Package targets -- add_custom_target(ie_dev_targets ALL DEPENDS ${OpenVINODeveloperPackageTargets} ${IEDeveloperPackageTargets} gflags -+ add_custom_target(ie_dev_targets ALL DEPENDS ${OpenVINODeveloperPackageTargets} ${IEDeveloperPackageTargets} - inference_engine_ir_reader inference_engine_ir_v7_reader) - endfunction() - -diff --git a/inference-engine/cmake/ie_parallel.cmake b/inference-engine/cmake/ie_parallel.cmake -index f8988aa105..28e79417f4 100644 ---- a/inference-engine/cmake/ie_parallel.cmake -+++ b/inference-engine/cmake/ie_parallel.cmake -@@ -67,13 +67,8 @@ function(set_ie_threading_interface_for TARGET_NAME) - set(IE_THREAD_DEFINE "IE_THREAD_SEQ") - - if (THREADING STREQUAL "TBB" OR THREADING STREQUAL "TBB_AUTO") -- if (TBB_FOUND) - set(IE_THREAD_DEFINE "IE_THREAD_TBB") -- ie_target_link_libraries(${TARGET_NAME} ${LINK_TYPE} ${TBB_IMPORTED_TARGETS}) -- else () -- ext_message(WARNING "TBB was not found by the configured TBB_DIR path.\ -- SEQ method will be used for ${TARGET_NAME}") -- endif () -+ target_link_libraries(${TARGET_NAME} INTERFACE tbb tbbmalloc) - elseif (THREADING STREQUAL "OMP") - if (WIN32) - set(omp_lib_name libiomp5md) -diff --git a/inference-engine/src/inference_engine/CMakeLists.txt b/inference-engine/src/inference_engine/CMakeLists.txt -index f012a038a2..3bba22215e 100644 ---- a/inference-engine/src/inference_engine/CMakeLists.txt -+++ b/inference-engine/src/inference_engine/CMakeLists.txt -@@ -216,25 +216,6 @@ list(APPEND core_components ngraph) - if((THREADING STREQUAL "TBB" OR THREADING STREQUAL "TBB_AUTO") AND TBBROOT MATCHES ${TEMP}) - ie_cpack_add_component(tbb REQUIRED) - list(APPEND core_components tbb) -- -- install(DIRECTORY "${TBB}/include" -- DESTINATION ${IE_CPACK_IE_DIR}/external/tbb -- COMPONENT tbb) -- install(DIRECTORY "${TBB}/lib" -- DESTINATION ${IE_CPACK_IE_DIR}/external/tbb -- COMPONENT tbb) -- if(EXISTS "${TBB}/bin") -- install(DIRECTORY "${TBB}/bin" -- DESTINATION ${IE_CPACK_IE_DIR}/external/tbb -- COMPONENT tbb) -- endif() -- install(FILES "${TBB}/LICENSE" -- DESTINATION ${IE_CPACK_IE_DIR}/external/tbb -- COMPONENT tbb) -- install(FILES "${TBB}/cmake/TBBConfig.cmake" -- "${TBB}/cmake/TBBConfigVersion.cmake" -- DESTINATION ${IE_CPACK_IE_DIR}/external/tbb/cmake -- COMPONENT tbb) - endif() - - ie_cpack_add_component(core REQUIRED DEPENDS ${core_components}) -diff --git a/inference-engine/tests/ie_test_utils/common_test_utils/CMakeLists.txt b/inference-engine/tests/ie_test_utils/common_test_utils/CMakeLists.txt -index ca58264bd5..11df150845 100644 ---- a/inference-engine/tests/ie_test_utils/common_test_utils/CMakeLists.txt -+++ b/inference-engine/tests/ie_test_utils/common_test_utils/CMakeLists.txt -@@ -26,7 +26,7 @@ function(add_gtest_libraries) - PROPERTIES FOLDER thirdparty) - endfunction() - --add_gtest_libraries() -+#add_gtest_libraries() - - if (MSVC) - set(PUGI pugixml_mt) -@@ -37,8 +37,6 @@ endif () - list(APPEND EXPORT_DEPENDENCIES - ${PUGI} - ${NGRAPH_LIBRARIES} -- gtest -- gtest_main - ) - - set(TARGET_NAME commonTestUtils) -diff --git a/inference-engine/tests/ie_test_utils/unit_test_utils/CMakeLists.txt b/inference-engine/tests/ie_test_utils/unit_test_utils/CMakeLists.txt -index 5413d13895..3f2803750f 100644 ---- a/inference-engine/tests/ie_test_utils/unit_test_utils/CMakeLists.txt -+++ b/inference-engine/tests/ie_test_utils/unit_test_utils/CMakeLists.txt -@@ -8,8 +8,7 @@ add_subdirectory(mocks/mock_engine) - - list(APPEND EXPORT_DEPENDENCIES - commonTestUtils_s -- inference_engine_s -- gmock) -+ inference_engine_s) - - addIeTarget( - NAME ${TARGET_NAME} -diff --git a/inference-engine/thirdparty/CMakeLists.txt b/inference-engine/thirdparty/CMakeLists.txt -index fa2a4d02d8..67d3b7ad05 100644 ---- a/inference-engine/thirdparty/CMakeLists.txt -+++ b/inference-engine/thirdparty/CMakeLists.txt -@@ -62,14 +62,13 @@ else() - endif() - - add_subdirectory(stb_lib) --add_subdirectory(ade) - add_subdirectory(fluid/modules/gapi) - --set_target_properties(ade fluid stb_image PROPERTIES FOLDER thirdparty) -+set_target_properties(fluid stb_image PROPERTIES FOLDER thirdparty) - - # developer package - --ie_developer_export_targets(ade fluid) -+ie_developer_export_targets(fluid) - - if (NOT USE_SYSTEM_PUGIXML) - set_target_properties(pugixml PROPERTIES FOLDER thirdparty) --- -2.29.2 - diff --git a/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0002-cldNN-disable-Werror.patch b/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0002-cldNN-disable-Werror.patch deleted file mode 100644 index a4e58b8e..00000000 --- a/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0002-cldNN-disable-Werror.patch +++ /dev/null @@ -1,30 +0,0 @@ -From 65014a2703a6a3892fdebc86fe1c5a3a589dbf56 Mon Sep 17 00:00:00 2001 -From: Anuj Mittal <anuj.mittal@intel.com> -Date: Wed, 10 Jun 2020 09:45:48 +0800 -Subject: [PATCH 2/5] cldNN: disable Werror - -Don't treat warnings as errors. This just leads to failures every time -we upgrade the compiler. - -Upstream-Status: Inappropriate - -Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> ---- - inference-engine/thirdparty/clDNN/CMakeLists.txt | 1 - - 1 file changed, 1 deletion(-) - -diff --git a/inference-engine/thirdparty/clDNN/CMakeLists.txt b/inference-engine/thirdparty/clDNN/CMakeLists.txt -index 4b444eca..8610b9b8 100644 ---- a/inference-engine/thirdparty/clDNN/CMakeLists.txt -+++ b/inference-engine/thirdparty/clDNN/CMakeLists.txt -@@ -770,7 +770,6 @@ foreach(__CLDNN_CompilerFlagName IN ITEMS "CMAKE_CXX_FLAGS" "CMAKE_C_FLAGS") - MultiProcessorCompilation - DeadCodeEliminate - ExtensionsEnabled -- TreatWarnAsErrorEnabled - WarnLevel4 - NoFastMath - StackProtector --- -2.25.4 - diff --git a/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0003-cmake-Fix-overloaded-virtual-error.patch b/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0003-cmake-Fix-overloaded-virtual-error.patch new file mode 100644 index 00000000..e0967d55 --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0003-cmake-Fix-overloaded-virtual-error.patch @@ -0,0 +1,33 @@ +From 900eeeb2953095e651270c0f42ccd8b26fd7885c Mon Sep 17 00:00:00 2001 +From: Anuj Mittal <anuj.mittal@intel.com> +Date: Wed, 29 Nov 2023 12:49:35 +0530 +Subject: [PATCH 3/4] cmake: Fix overloaded-virtual error + +* Remove -Werror for: +|git/src/plugins/intel_gpu/src/kernel_selector/jitter.h:129:28: error: 'virtual kernel_selector::JitDefinitions kernel_selector::JitConstant::GetDefinitions() const' was hidden [-Werror=overloaded-virtual=] +| 129 | virtual JitDefinitions GetDefinitions() const = 0; +| | + +Upstream-Status: Pending + +Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> +--- + src/plugins/intel_gpu/CMakeLists.txt | 2 +- + 1 file changed, 1 insertion(+), 1 deletion(-) + +diff --git a/src/plugins/intel_gpu/CMakeLists.txt b/src/plugins/intel_gpu/CMakeLists.txt +index b0c66a435d6..a3037147cc2 100644 +--- a/src/plugins/intel_gpu/CMakeLists.txt ++++ b/src/plugins/intel_gpu/CMakeLists.txt +@@ -38,7 +38,7 @@ add_subdirectory(thirdparty) + include(thirdparty/cmake/rapidjson.cmake) + + if(CMAKE_COMPILER_IS_GNUCXX) +- ov_add_compiler_flags(-Werror) ++ #ov_add_compiler_flags(-Werror) + endif() + + add_subdirectory(src/runtime) +-- +2.34.1 + diff --git a/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0003-inference-engine-installation-fixes.patch b/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0003-inference-engine-installation-fixes.patch deleted file mode 100644 index 52049eb0..00000000 --- a/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0003-inference-engine-installation-fixes.patch +++ /dev/null @@ -1,223 +0,0 @@ -From cb9a755218ff033c8c42e9dc46a845310a3331fa Mon Sep 17 00:00:00 2001 -From: Anuj Mittal <anuj.mittal@intel.com> -Date: Thu, 11 Jun 2020 14:24:04 +0800 -Subject: [PATCH] cmake installation fixes - -Make sure that libraries/samples/binaries are installed correctly. - -Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> ---- - CMakeLists.txt | 14 -------------- - cmake/developer_package.cmake | 10 ++++++---- - .../samples/common/opencv_c_wraper/CMakeLists.txt | 2 ++ - inference-engine/samples/CMakeLists.txt | 9 +++------ - .../samples/common/format_reader/CMakeLists.txt | 2 ++ - .../src/inference_engine/CMakeLists.txt | 4 ++-- - .../src/vpu/myriad_plugin/CMakeLists.txt | 2 +- - inference-engine/tests/unit/cpu/CMakeLists.txt | 2 ++ - .../tests/unit/inference_engine/CMakeLists.txt | 2 ++ - inference-engine/tests/unit/vpu/CMakeLists.txt | 2 ++ - inference-engine/tools/compile_tool/CMakeLists.txt | 2 +- - 11 files changed, 23 insertions(+), 28 deletions(-) - -diff --git a/CMakeLists.txt b/CMakeLists.txt -index 5c3585a3e9..2b8c99f862 100644 ---- a/CMakeLists.txt -+++ b/CMakeLists.txt -@@ -176,12 +176,8 @@ ie_shellcheck_process(DIRECTORY "${OpenVINO_MAIN_SOURCE_DIR}" - - # install setupvars - --ie_cpack_add_component(setupvars REQUIRED) - - if(UNIX) -- install(PROGRAMS scripts/setupvars/setupvars.sh -- DESTINATION bin -- COMPONENT setupvars) - elseif(WIN32) - install(PROGRAMS scripts/setupvars/setupvars.bat - DESTINATION bin -@@ -191,22 +187,12 @@ endif() - # install install_dependencies - - if(UNIX) -- ie_cpack_add_component(install_dependencies REQUIRED) -- install(DIRECTORY scripts/install_dependencies/ -- DESTINATION install_dependencies -- COMPONENT install_dependencies) - endif() - - # install files for demo - --ie_cpack_add_component(demo_scripts REQUIRED DEPENDS core) - - if(UNIX) -- install(DIRECTORY scripts/demo/ -- DESTINATION deployment_tools/demo -- COMPONENT demo_scripts -- USE_SOURCE_PERMISSIONS -- PATTERN *.bat EXCLUDE) - elseif(WIN32) - install(DIRECTORY scripts/demo/ - DESTINATION deployment_tools/demo -diff --git a/cmake/developer_package.cmake b/cmake/developer_package.cmake -index cda7afd294..ec51636e96 100644 ---- a/cmake/developer_package.cmake -+++ b/cmake/developer_package.cmake -@@ -29,7 +29,9 @@ list(APPEND CMAKE_MODULE_PATH - include(CPackComponent) - unset(IE_CPACK_COMPONENTS_ALL CACHE) - --set(IE_CPACK_IE_DIR deployment_tools/inference_engine) -+if (NOT DEFINED IE_CPACK_IE_DIR) -+ set(IE_CPACK_IE_DIR deployment_tools/inference_engine) -+endif() - - # Search packages for the host system instead of packages for the target system - # in case of cross compilation these macros should be defined by the toolchain file -@@ -55,8 +57,8 @@ function(ie_cpack_set_library_dir) - set(IE_CPACK_RUNTIME_PATH ${IE_CPACK_IE_DIR}/bin/${ARCH_FOLDER}/${CMAKE_BUILD_TYPE} PARENT_SCOPE) - set(IE_CPACK_ARCHIVE_PATH ${IE_CPACK_IE_DIR}/lib/${ARCH_FOLDER}/${CMAKE_BUILD_TYPE} PARENT_SCOPE) - else() -- set(IE_CPACK_LIBRARY_PATH ${IE_CPACK_IE_DIR}/lib/${ARCH_FOLDER} PARENT_SCOPE) -- set(IE_CPACK_RUNTIME_PATH ${IE_CPACK_IE_DIR}/lib/${ARCH_FOLDER} PARENT_SCOPE) -+ set(IE_CPACK_LIBRARY_PATH ${CMAKE_INSTALL_LIBDIR} PARENT_SCOPE) -+ set(IE_CPACK_RUNTIME_PATH ${CMAKE_INSTALL_BINDIR} PARENT_SCOPE) - set(IE_CPACK_ARCHIVE_PATH ${IE_CPACK_IE_DIR}/lib/${ARCH_FOLDER} PARENT_SCOPE) - endif() - endfunction() -@@ -151,7 +153,7 @@ endif() - - # allow to override default OUTPUT_ROOT root - if(NOT DEFINED OUTPUT_ROOT) -- set(OUTPUT_ROOT ${OpenVINO_MAIN_SOURCE_DIR}) -+ set(OUTPUT_ROOT ${CMAKE_CURRENT_BINARY_DIR}) - endif() - - # Enable postfixes for Debug/Release builds -diff --git a/inference-engine/ie_bridges/c/samples/common/opencv_c_wraper/CMakeLists.txt b/inference-engine/ie_bridges/c/samples/common/opencv_c_wraper/CMakeLists.txt -index b8d5ddf622..d086478f6d 100644 ---- a/inference-engine/ie_bridges/c/samples/common/opencv_c_wraper/CMakeLists.txt -+++ b/inference-engine/ie_bridges/c/samples/common/opencv_c_wraper/CMakeLists.txt -@@ -29,3 +29,5 @@ set_target_properties(${TARGET_NAME} PROPERTIES FOLDER c_samples) - if(COMMAND add_cpplint_target) - add_cpplint_target(${TARGET_NAME}_cpplint FOR_TARGETS ${TARGET_NAME}) - endif() -+ -+install(TARGETS ${TARGET_NAME} DESTINATION ${CMAKE_INSTALL_LIBDIR}) -diff --git a/inference-engine/samples/CMakeLists.txt b/inference-engine/samples/CMakeLists.txt -index 236c17e6af..22f410d1fd 100644 ---- a/inference-engine/samples/CMakeLists.txt -+++ b/inference-engine/samples/CMakeLists.txt -@@ -34,7 +34,7 @@ endif() - - if(IE_MAIN_SOURCE_DIR) - # in case if samples are built from IE repo -- set(IE_MAIN_SAMPLES_DIR ${OpenVINO_MAIN_SOURCE_DIR}) -+ set(IE_MAIN_SAMPLES_DIR ${CMAKE_BINARY_DIR}) - # hint for find_package(InferenceEngine in the samples folder) - set(InferenceEngine_DIR "${CMAKE_BINARY_DIR}") - # hint for find_package(ngraph in the samples folder) -@@ -118,11 +118,6 @@ set (HAVE_INTTYPES_H 1) - set (INTTYPES_FORMAT C99) - set (BUILD_TESTING OFF) - --if(EXISTS "${CMAKE_CURRENT_SOURCE_DIR}/thirdparty/gflags") -- add_subdirectory(thirdparty/gflags) -- set_target_properties(gflags_nothreads_static PROPERTIES FOLDER thirdparty) --endif() -- - if (CMAKE_CXX_COMPILER_ID STREQUAL "GNU") - set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wall") - endif() -@@ -245,6 +240,8 @@ macro(ie_add_sample) - add_cpplint_target(${IE_SAMPLE_NAME}_cpplint FOR_TARGETS ${IE_SAMPLE_NAME} - CUSTOM_FILTERS ${custom_filters}) - endif() -+ -+ install(TARGETS ${IE_SAMPLE_NAME} DESTINATION bin) - endmacro() - - # collect all samples subdirectories -diff --git a/inference-engine/samples/common/format_reader/CMakeLists.txt b/inference-engine/samples/common/format_reader/CMakeLists.txt -index 48dbed9f2b..76532fd047 100644 ---- a/inference-engine/samples/common/format_reader/CMakeLists.txt -+++ b/inference-engine/samples/common/format_reader/CMakeLists.txt -@@ -41,3 +41,5 @@ target_include_directories(${TARGET_NAME} PUBLIC "${CMAKE_CURRENT_SOURCE_DIR}" - - set_target_properties(${TARGET_NAME} PROPERTIES COMPILE_PDB_NAME ${TARGET_NAME} - FOLDER cpp_samples) -+ -+install(TARGETS ${TARGET_NAME} DESTINATION ${CMAKE_INSTALL_LIBDIR}) -diff --git a/inference-engine/src/inference_engine/CMakeLists.txt b/inference-engine/src/inference_engine/CMakeLists.txt -index 3bba22215e..b6f6b7b203 100644 ---- a/inference-engine/src/inference_engine/CMakeLists.txt -+++ b/inference-engine/src/inference_engine/CMakeLists.txt -@@ -229,8 +229,8 @@ install(TARGETS ${TARGET_NAME} - install(FILES "${OpenVINO_BINARY_DIR}/share/ie_parallel.cmake" - "${OpenVINO_BINARY_DIR}/share/InferenceEngineConfig.cmake" - "${OpenVINO_BINARY_DIR}/share/InferenceEngineConfig-version.cmake" -- DESTINATION ${IE_CPACK_IE_DIR}/share -+ DESTINATION ${CMAKE_INSTALL_LIBDIR}/cmake/InferenceEngine - COMPONENT core) - install(FILES $<TARGET_FILE_DIR:${TARGET_NAME}>/plugins.xml -- DESTINATION ${IE_CPACK_RUNTIME_PATH} -+ DESTINATION ${CMAKE_INSTALL_LIBDIR} - COMPONENT core) -diff --git a/inference-engine/src/vpu/myriad_plugin/CMakeLists.txt b/inference-engine/src/vpu/myriad_plugin/CMakeLists.txt -index 889f2e6e66..67de9c3a32 100644 ---- a/inference-engine/src/vpu/myriad_plugin/CMakeLists.txt -+++ b/inference-engine/src/vpu/myriad_plugin/CMakeLists.txt -@@ -50,5 +50,5 @@ set_target_properties(${TARGET_NAME} PROPERTIES INTERPROCEDURAL_OPTIMIZATION_REL - # install - - install(FILES ${IE_MAIN_SOURCE_DIR}/thirdparty/movidius/mvnc/src/97-myriad-usbboot.rules -- DESTINATION ${IE_CPACK_IE_DIR}/external -+ DESTINATION ${CMAKE_INSTALL_SYSCONFDIR}/udev/rules.d - COMPONENT myriad) -diff --git a/inference-engine/tests/unit/cpu/CMakeLists.txt b/inference-engine/tests/unit/cpu/CMakeLists.txt -index c9a92b4857..2b4fcc4d01 100644 ---- a/inference-engine/tests/unit/cpu/CMakeLists.txt -+++ b/inference-engine/tests/unit/cpu/CMakeLists.txt -@@ -25,3 +25,5 @@ addIeTargetTest( - ie_faster_build(${TARGET_NAME} - UNITY - ) -+ -+install(TARGETS ${TARGET_NAME} DESTINATION ${CMAKE_INSTALL_BINDIR}) -diff --git a/inference-engine/tests/unit/inference_engine/CMakeLists.txt b/inference-engine/tests/unit/inference_engine/CMakeLists.txt -index 76cc67cca1..04ff11cc05 100644 ---- a/inference-engine/tests/unit/inference_engine/CMakeLists.txt -+++ b/inference-engine/tests/unit/inference_engine/CMakeLists.txt -@@ -26,3 +26,5 @@ addIeTargetTest( - LABELS - IE - ) -+ -+install(TARGETS ${TARGET_NAME} DESTINATION ${CMAKE_INSTALL_BINDIR}) -diff --git a/inference-engine/tests/unit/vpu/CMakeLists.txt b/inference-engine/tests/unit/vpu/CMakeLists.txt -index 215562f2af..215ec9f4fb 100644 ---- a/inference-engine/tests/unit/vpu/CMakeLists.txt -+++ b/inference-engine/tests/unit/vpu/CMakeLists.txt -@@ -35,3 +35,5 @@ addIeTargetTest( - VPU - MYRIAD - ) -+ -+install(TARGETS ${TARGET_NAME} DESTINATION ${CMAKE_INSTALL_BINDIR}) -diff --git a/inference-engine/tools/compile_tool/CMakeLists.txt b/inference-engine/tools/compile_tool/CMakeLists.txt -index 58614d3ced..9dfa068e4d 100644 ---- a/inference-engine/tools/compile_tool/CMakeLists.txt -+++ b/inference-engine/tools/compile_tool/CMakeLists.txt -@@ -49,7 +49,7 @@ add_cpplint_target(${TARGET_NAME}_cpplint FOR_TARGETS ${TARGET_NAME}) - # install - - install(TARGETS compile_tool -- RUNTIME DESTINATION deployment_tools/tools/compile_tool -+ RUNTIME DESTINATION ${CMAKE_INSTALL_BINDIR} - COMPONENT core) - - install(FILES README.md --- -2.29.2 - diff --git a/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0004-protobuf-allow-target-protoc-to-be-built.patch b/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0004-protobuf-allow-target-protoc-to-be-built.patch new file mode 100644 index 00000000..59095133 --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0004-protobuf-allow-target-protoc-to-be-built.patch @@ -0,0 +1,45 @@ +From 3e288ed876c6bcb6aa3174e52446b479255ddf22 Mon Sep 17 00:00:00 2001 +From: Anuj Mittal <anuj.mittal@intel.com> +Date: Wed, 29 Nov 2023 12:55:19 +0530 +Subject: [PATCH 4/4] protobuf: allow target protoc to be built + +We can run target binaries using a qemu wrapper so allow these to be +built and run. + +Upstream-Status: Inappropriate + +Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> +--- + cmake/developer_package/frontends/frontends.cmake | 2 +- + thirdparty/protobuf/CMakeLists.txt | 2 +- + 2 files changed, 2 insertions(+), 2 deletions(-) + +diff --git a/cmake/developer_package/frontends/frontends.cmake b/cmake/developer_package/frontends/frontends.cmake +index 49c5b881030..2a1ea8562bc 100644 +--- a/cmake/developer_package/frontends/frontends.cmake ++++ b/cmake/developer_package/frontends/frontends.cmake +@@ -143,7 +143,7 @@ macro(ov_add_frontend) + set(OUTPUT_PB_HEADER ${CMAKE_CURRENT_BINARY_DIR}/${relative_path}/${FILE_WE}.pb.h) + add_custom_command( + OUTPUT "${OUTPUT_PB_SRC}" "${OUTPUT_PB_HEADER}" +- COMMAND ${PROTOC_EXECUTABLE} ARGS --cpp_out ${CMAKE_CURRENT_BINARY_DIR} -I ${protofiles_root_dir} ${proto_file} ++ COMMAND protoc ARGS --cpp_out ${CMAKE_CURRENT_BINARY_DIR} -I ${protofiles_root_dir} ${proto_file} + DEPENDS ${PROTOC_DEPENDENCY} ${proto_file} + COMMENT "Running C++ protocol buffer compiler (${PROTOC_EXECUTABLE}) on ${proto_file_relative}" + VERBATIM +diff --git a/thirdparty/protobuf/CMakeLists.txt b/thirdparty/protobuf/CMakeLists.txt +index 4b6d6da87f3..409e492a5b3 100644 +--- a/thirdparty/protobuf/CMakeLists.txt ++++ b/thirdparty/protobuf/CMakeLists.txt +@@ -28,7 +28,7 @@ set(ABSL_PROPAGATE_CXX_STD ON CACHE BOOL "Abseil protogate CXX standard to depen + if(CMAKE_CROSSCOMPILING OR + (APPLE AND (HOST_X86_64 AND AARCH64)) OR + (MSVC AND (HOST_X86_64 AND (AARCH64 OR ARM)))) +- set(protobuf_BUILD_PROTOC_BINARIES OFF CACHE BOOL "Build protoc binaries" FORCE) ++ set(protobuf_BUILD_PROTOC_BINARIES ON CACHE BOOL "Build protoc binaries" FORCE) + else() + set(protobuf_BUILD_PROTOC_BINARIES ON CACHE BOOL "Build protoc binaries" FORCE) + endif() +-- +2.34.1 + diff --git a/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0005-cldnn-fix-inclusion-of-headers.patch b/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0005-cldnn-fix-inclusion-of-headers.patch deleted file mode 100644 index d4819077..00000000 --- a/dynamic-layers/openembedded-layer/recipes-support/opencv/files/0005-cldnn-fix-inclusion-of-headers.patch +++ /dev/null @@ -1,53 +0,0 @@ -From aa515686548d74a68456b9ba7a295b6ae4be343a Mon Sep 17 00:00:00 2001 -From: Anuj Mittal <anuj.mittal@intel.com> -Date: Wed, 10 Jun 2020 09:50:19 +0800 -Subject: [PATCH] cldnn: fix inclusion of headers - -Upstream-Status: Inappropriate - -Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> - ---- - inference-engine/include/gpu/gpu_ocl_wrapper.hpp | 2 +- - inference-engine/src/cldnn_engine/CMakeLists.txt | 1 + - .../thirdparty/clDNN/common/khronos_ocl_clhpp/cl2_ext.hpp | 2 +- - 3 files changed, 3 insertions(+), 2 deletions(-) - -diff --git a/inference-engine/include/gpu/gpu_ocl_wrapper.hpp b/inference-engine/include/gpu/gpu_ocl_wrapper.hpp -index 68ccf3217..c5b0ce6be 100644 ---- a/inference-engine/include/gpu/gpu_ocl_wrapper.hpp -+++ b/inference-engine/include/gpu/gpu_ocl_wrapper.hpp -@@ -39,7 +39,7 @@ - # pragma GCC system_header - #endif - --#include <CL/cl2.hpp> -+#include <cl2.hpp> - - #ifdef __GNUC__ - # pragma GCC diagnostic pop -diff --git a/inference-engine/src/cldnn_engine/CMakeLists.txt b/inference-engine/src/cldnn_engine/CMakeLists.txt -index 7e15abbed..46d43e0e5 100644 ---- a/inference-engine/src/cldnn_engine/CMakeLists.txt -+++ b/inference-engine/src/cldnn_engine/CMakeLists.txt -@@ -31,6 +31,7 @@ target_include_directories(${TARGET_NAME} PRIVATE - ${CMAKE_CURRENT_SOURCE_DIR} - $<TARGET_PROPERTY:inference_engine_transformations,INTERFACE_INCLUDE_DIRECTORIES> - ${CLDNN__IOCL_ICD_INCDIRS} -+ ${CLDNN_TOP_FOLDER}/common/khronos_ocl_clhpp - ${CLDNN_TOP_FOLDER}) - - set_target_properties(${TARGET_NAME} PROPERTIES INTERPROCEDURAL_OPTIMIZATION_RELEASE ${ENABLE_LTO}) -diff --git a/inference-engine/thirdparty/clDNN/common/khronos_ocl_clhpp/cl2_ext.hpp b/inference-engine/thirdparty/clDNN/common/khronos_ocl_clhpp/cl2_ext.hpp -index 97608c30a..3a04096c0 100644 ---- a/inference-engine/thirdparty/clDNN/common/khronos_ocl_clhpp/cl2_ext.hpp -+++ b/inference-engine/thirdparty/clDNN/common/khronos_ocl_clhpp/cl2_ext.hpp -@@ -29,7 +29,7 @@ typedef cl_d3d11_device_set_khr cl_device_set_intel; - typedef cl_va_api_device_source_intel cl_device_source_intel; - typedef cl_va_api_device_set_intel cl_device_set_intel; - #endif --#include <CL/cl_intel_planar_yuv.h> -+#include <CL/cl_ext_intel.h> - #include "cl_intel_usm_defs.h" - - #include <memory> diff --git a/dynamic-layers/openembedded-layer/recipes-support/opencv/open-model-zoo/0001-use-oe-gflags.patch b/dynamic-layers/openembedded-layer/recipes-support/opencv/open-model-zoo/0001-use-oe-gflags.patch index 864c735e..816a98a3 100644 --- a/dynamic-layers/openembedded-layer/recipes-support/opencv/open-model-zoo/0001-use-oe-gflags.patch +++ b/dynamic-layers/openembedded-layer/recipes-support/opencv/open-model-zoo/0001-use-oe-gflags.patch @@ -1,33 +1,27 @@ -From 56078933a8397b7fae146156743408319eac1ae0 Mon Sep 17 00:00:00 2001 -From: Anuj Mittal <anuj.mittal@intel.com> -Date: Mon, 19 Oct 2020 23:29:23 +0800 +From 804b08023b3f8e72b8e3eb09e464d6775c11d966 Mon Sep 17 00:00:00 2001 +From: Naveen Saini <naveen.kumar.saini@intel.com> +Date: Fri, 21 Oct 2022 11:38:23 +0800 Subject: [PATCH] demos: use gflags from meta-oe Upstream-Status: Inappropriate Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> +Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> + --- - demos/CMakeLists.txt | 4 ++-- - 1 file changed, 2 insertions(+), 2 deletions(-) + demos/CMakeLists.txt | 2 +- + 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/demos/CMakeLists.txt b/demos/CMakeLists.txt -index e5f17231..08d4e4e6 100644 +index 51767051c..fb7e3d22f 100644 --- a/demos/CMakeLists.txt +++ b/demos/CMakeLists.txt -@@ -90,11 +90,11 @@ if (${CMAKE_CXX_COMPILER_ID} STREQUAL GNU) - endif() - #################################### - --set (GFLAGS_IS_SUBPROJECT TRUE) -+set (GFLAGS_IS_SUBPROJECT FALSE) - set (HAVE_SYS_STAT_H 1) - set (HAVE_INTTYPES_H 1) +@@ -141,7 +141,7 @@ endmacro() + find_package(OpenCV REQUIRED COMPONENTS core highgui videoio imgproc imgcodecs) + find_package(OpenVINO REQUIRED COMPONENTS Runtime) -add_subdirectory(thirdparty/gflags) +#add_subdirectory(thirdparty/gflags) + add_subdirectory(common/cpp) - if (${CMAKE_CXX_COMPILER_ID} STREQUAL GNU) - set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wall") --- -2.26.2 - + find_package(OpenCV QUIET COMPONENTS gapi) diff --git a/dynamic-layers/openembedded-layer/recipes-support/opencv/open-model-zoo_2021.2.bb b/dynamic-layers/openembedded-layer/recipes-support/opencv/open-model-zoo_2021.2.bb deleted file mode 100644 index 6784d0d8..00000000 --- a/dynamic-layers/openembedded-layer/recipes-support/opencv/open-model-zoo_2021.2.bb +++ /dev/null @@ -1,63 +0,0 @@ -SUMMARY = "OpenVINO(TM) Toolkit - Open Model Zoo repository" -HOMEPAGE = "https://github.com/opencv/open_model_zoo" -DESCRIPTION = "This repository includes optimized deep learning \ -models and a set of demos to expedite development of high-performance \ -deep learning inference applications." - -SRC_URI = "git://github.com/opencv/open_model_zoo.git;protocol=git;branch=master \ - file://0001-use-oe-gflags.patch;striplevel=2 \ - " - -SRCREV = "338630987b403a6981d03ab6d04c2d5ad367793a" - -LICENSE = "Apache-2.0" -LIC_FILES_CHKSUM = "file://../LICENSE;md5=86d3f3a95c324c9479bd8986968f4327 \ -" - -inherit cmake - -S = "${WORKDIR}/git/demos" - -DEPENDS += "openvino-inference-engine opencv gflags" - -RDEPENDS_${PN} += " \ - python3-decorator \ - python3-defusedxml \ - python3-networkx \ - python3-protobuf \ - python3-test-generator \ - python3-requests \ - python3-pyyaml \ -" - -COMPATIBLE_HOST = '(x86_64).*-linux' - -EXTRA_OECMAKE += " \ - -DIE_MAIN_SOURCE_DIR=${B} \ - -DENABLE_SAMPLES=ON \ - -DIE_INCLUDE_DIR=${STAGING_EXECPREFIXDIR} \ - -DIE_RELEASE_LIBRARY=${STAGING_LIBDIR}/libinference_engine.so \ - -DIE_C_API_RELEASE_LIBRARY=${STAGING_LIBDIR}/libinference_engine_c_api.so \ - -DIE_LEGACY_RELEASE_LIBRARY=${STAGING_LIBDIR}/libinference_engine_legacy.so \ - -DIE_NN_BUILDER_RELEASE_LIBRARY=${STAGING_LIBDIR}/libinference_engine_nn_builder.so \ - -DIE_ROOT_DIR=${WORKDIR}/InferenceEngine \ -" - -do_configure_prepend(){ - mkdir -p ${WORKDIR}/InferenceEngine/share - cp ${STAGING_LIBDIR}/cmake/InferenceEngine/* ${WORKDIR}/InferenceEngine/share/ -} - -do_install(){ - install -d ${D}${libdir} - install -d ${D}${bindir} - install -d ${D}${datadir}/openvino/open-model-zoo/tools - install -d ${D}${datadir}/openvino/open-model-zoo/demos/python_demos - cp -rf ${WORKDIR}/build/intel64/Release/lib/*.a ${D}${libdir} - cp -rf ${WORKDIR}/build/intel64/Release/*_demo* ${D}${bindir} - cp -rf ${WORKDIR}/git/models ${D}${datadir}/openvino/open-model-zoo - cp -rf ${WORKDIR}/git/tools/downloader ${D}${datadir}/openvino/open-model-zoo/tools - cp -rf ${WORKDIR}/git/demos/python_demos ${D}${datadir}/openvino/open-model-zoo/demos -} - -FILES_${PN} += "${datadir}/openvino" diff --git a/dynamic-layers/openembedded-layer/recipes-support/opencv/open-model-zoo_2024.0.0.bb b/dynamic-layers/openembedded-layer/recipes-support/opencv/open-model-zoo_2024.0.0.bb new file mode 100644 index 00000000..495a4786 --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-support/opencv/open-model-zoo_2024.0.0.bb @@ -0,0 +1,54 @@ +SUMMARY = "OpenVINO(TM) Toolkit - Open Model Zoo repository" +HOMEPAGE = "https://github.com/opencv/open_model_zoo" +DESCRIPTION = "This repository includes optimized deep learning \ +models and a set of demos to expedite development of high-performance \ +deep learning inference applications." + +SRC_URI = "git://github.com/opencv/open_model_zoo.git;protocol=https;branch=master \ + file://0001-use-oe-gflags.patch \ + " + +SRCREV = "37f60eb7fe1dcdedc552b2fb184d646723ed5e80" + +LICENSE = "Apache-2.0" +LIC_FILES_CHKSUM = "file://LICENSE;md5=86d3f3a95c324c9479bd8986968f4327 \ +" + +inherit cmake + +S = "${WORKDIR}/git" +OECMAKE_SOURCEPATH = "${S}/demos" + +DEPENDS += "openvino-inference-engine opencv gflags" + +RDEPENDS:${PN} += " \ + python3-decorator \ + python3-defusedxml \ + python3-networkx \ + python3-protobuf \ + python3-requests \ + python3-pyyaml \ + python3-numpy \ + bash \ +" + +COMPATIBLE_HOST = '(x86_64).*-linux' +COMPATIBLE_HOST:libc-musl = "null" + +EXTRA_OECMAKE += " \ + -DENABLE_SAMPLES=ON \ + " + +do_install(){ + install -d ${D}${libdir} + install -d ${D}${bindir} + install -d ${D}${datadir}/openvino/open-model-zoo/tools + install -d ${D}${datadir}/openvino/open-model-zoo/demos + cp -rf ${WORKDIR}/build/intel64/Release/*.a ${D}${libdir} + cp -rf ${WORKDIR}/build/intel64/Release/*_demo* ${D}${bindir} + cp -rf ${WORKDIR}/git/models ${D}${datadir}/openvino/open-model-zoo + cp -rf ${WORKDIR}/git/demos ${D}${datadir}/openvino/open-model-zoo + cp -rf ${WORKDIR}/git/tools/model_tools ${D}${datadir}/openvino/open-model-zoo/tools +} + +FILES:${PN} += "${datadir}/openvino" diff --git a/dynamic-layers/openembedded-layer/recipes-support/opencv/openvino-inference-engine_2021.2.bb b/dynamic-layers/openembedded-layer/recipes-support/opencv/openvino-inference-engine_2021.2.bb deleted file mode 100644 index 22194467..00000000 --- a/dynamic-layers/openembedded-layer/recipes-support/opencv/openvino-inference-engine_2021.2.bb +++ /dev/null @@ -1,123 +0,0 @@ -SUMMARY = "OpenVINO(TM) Toolkit - Deep Learning Deployment Toolkit" -HOMEPAGE = "https://github.com/opencv/dldt" -DESCRIPTION = "This toolkit allows developers to deploy pre-trained \ -deep learning models through a high-level C++ Inference Engine API \ -integrated with application logic." - -SRC_URI = "git://github.com/openvinotoolkit/openvino.git;protocol=git;branch=releases/2021/2;lfs=0 \ - https://download.01.org/opencv/master/openvinotoolkit/thirdparty/unified/VPU/usb-ma2x8x/firmware_usb-ma2x8x_1522.zip;name=usb_ma2x8x \ - https://download.01.org/opencv/master/openvinotoolkit/thirdparty/unified/VPU/pcie-ma2x8x/firmware_pcie-ma2x8x_1522.zip;name=pcie_ma2x8x \ - git://github.com/openvinotoolkit/oneDNN.git;protocol=https;destsuffix=git/inference-engine/thirdparty/mkl-dnn;name=mkl;nobranch=1 \ - file://0001-inference-engine-use-system-installed-packages.patch \ - file://0002-cldNN-disable-Werror.patch \ - file://0003-inference-engine-installation-fixes.patch \ - file://0005-cldnn-fix-inclusion-of-headers.patch \ - file://0001-dont-install-licenses-and-version-file.patch \ - " - -SRCREV = "4795391b73381660b69b4cd3986c7a0bf902e868" -SRCREV_mkl = "5ef085d5af65e8966e03cdfcbaa65761d61a5c9a" - -SRC_URI[usb_ma2x8x.sha256sum] = "95a93144f0bbfe6e35d3830e93e6b63e1e109f849a6a7c307cae9030e3a662aa" -SRC_URI[pcie_ma2x8x.sha256sum] = "6d061d21d90f1919ef375138066ba7a20ceb663901d2729d9cb1b639169df5da" - -LICENSE = "Apache-2.0 & ISSL & MIT" -LIC_FILES_CHKSUM = "file://LICENSE;md5=86d3f3a95c324c9479bd8986968f4327 \ - file://inference-engine/thirdparty/mkl-dnn/LICENSE;md5=afa44a3d001cc203032135324f9636b7 \ - file://inference-engine/thirdparty/mkl-dnn/src/cpu/xbyak/COPYRIGHT;md5=03532861dad9003cc2c17f14fc7a4efa \ - file://inference-engine/thirdparty/clDNN/common/khronos_ocl_clhpp/LICENSE.txt;md5=88b295a48d2b3244ba65d3c055472c8a \ -" -LICENSE_${PN}-vpu-firmware = "ISSL" - -inherit cmake python3native - -S = "${WORKDIR}/git" - -EXTRA_OECMAKE += " \ - -DENABLE_OPENCV=0 \ - -DENABLE_PLUGIN_RPATH=0 \ - -DENABLE_GNA=0 \ - -DPYTHON_EXECUTABLE=${PYTHON} \ - -DCMAKE_BUILD_TYPE=RelWithDebInfo \ - -DTHREADING=TBB -DTBB_DIR=${STAGING_LIBDIR} \ - -DENABLE_SAMPLES=1 \ - -DIE_CPACK_IE_DIR=${prefix} \ - -DNGRAPH_UNIT_TEST_ENABLE=FALSE \ - -DNGRAPH_TEST_UTIL_ENABLE=FALSE \ - -DNGRAPH_ONNX_IMPORT_ENABLE=OFF \ - -DNGRAPH_JSON_ENABLE=FALSE \ - -DTREAT_WARNING_AS_ERROR=FALSE \ - -DENABLE_SPEECH_DEMO=FALSE \ - -DENABLE_DATA=FALSE \ - -DUSE_SYSTEM_PUGIXML=TRUE \ - " - -DEPENDS += "libusb1 \ - ade \ - opencv \ - pugixml \ - protobuf-native \ - tbb \ - onednn \ - " - -COMPATIBLE_HOST = '(x86_64).*-linux' -COMPATIBLE_HOST_libc-musl = "null" - -PACKAGECONFIG ?= "vpu opencl" -PACKAGECONFIG[opencl] = "-DENABLE_CLDNN=1 -DCLDNN__IOCL_ICD_INCDIRS=${STAGING_INCDIR} -DCLDNN__IOCL_ICD_STLDIRS=${STAGING_LIBDIR} -DCLDNN__IOCL_ICD_SHLDIRS=${STAGING_LIBDIR}, -DENABLE_CLDNN=0, ocl-icd opencl-headers libva, intel-compute-runtime" -PACKAGECONFIG[python3] = "-DENABLE_PYTHON=ON -DPYTHON_LIBRARY=${PYTHON_LIBRARY} -DPYTHON_INCLUDE_DIR=${PYTHON_INCLUDE_DIR}, -DENABLE_PYTHON=OFF, python3-cython-native, python3 python3-numpy python3-opencv python3-progress python3-cython" -PACKAGECONFIG[vpu] = "-DENABLE_VPU=ON -DVPU_FIRMWARE_USB-MA2X8X_FILE=../mvnc/usb-ma2x8x.mvcmd -DVPU_FIRMWARE_PCIE-MA2X8X_FILE=../mvnc/pcie-ma2x8x.mvcmd,-DENABLE_VPU=OFF,,${PN}-vpu-firmware" -PACKAGECONFIG[verbose] = "-DVERBOSE_BUILD=1,-DVERBOSE_BUILD=0" - -do_install_append() { - if ${@bb.utils.contains('PACKAGECONFIG', 'vpu', 'true', 'false', d)}; then - install -m0644 ${WORKDIR}/mvnc/usb-ma2x8x.mvcmd ${D}${libdir}/ - install -m0644 ${WORKDIR}/mvnc/pcie-ma2x8x.mvcmd ${D}${libdir}/ - fi - - if ${@bb.utils.contains('PACKAGECONFIG', 'python3', 'true', 'false', d)}; then - install -d ${D}${datadir}/inference_engine - mv ${D}/usr/samples/python ${D}${datadir}/inference_engine/ - - install -d ${D}${PYTHON_SITEPACKAGES_DIR} - mv ${D}${prefix}/python/${PYTHON_DIR}/openvino ${D}${PYTHON_SITEPACKAGES_DIR}/ - mv ${D}${prefix}/deployment_tools/tools/benchmark_tool ${D}${PYTHON_SITEPACKAGES_DIR}/openvino/ - mv ${D}${prefix}/deployment_tools/tools/cross_check_tool ${D}${PYTHON_SITEPACKAGES_DIR}/openvino/ - - rm -rf ${D}${prefix}/python - fi - - rm -rf ${D}${prefix}/deployment_tools - - # Remove the samples source directory. We install the built samples. - rm -rf ${D}/usr/samples -} - -# Otherwise e.g. ros-openvino-toolkit-dynamic-vino-sample when using dldt-inference-engine uses dldt-inference-engine WORKDIR -# instead of RSS -SSTATE_SCAN_FILES_append = " *.cmake" - -FILES_${PN}-dev = "${includedir} \ - ${libdir}/cmake \ - " - -FILES_${PN} += "${libdir}/lib*${SOLIBSDEV} \ - ${datadir}/openvino \ - ${libdir}/custom_kernels \ - ${libdir}/plugins.xml \ - ${libdir}/cache.json \ - " - -# Move inference engine samples into a separate package -PACKAGES =+ "${PN}-samples ${PN}-vpu-firmware" - -FILES_${PN}-samples = "${datadir}/inference_engine \ - ${bindir} \ - " -FILES_${PN}-vpu-firmware += "${libdir}/*.mvcmd" - -# Package for inference engine python API -PACKAGES =+ "${PN}-${PYTHON_PN}" - -FILES_${PN}-${PYTHON_PN} = "${PYTHON_SITEPACKAGES_DIR}/openvino" diff --git a/dynamic-layers/openembedded-layer/recipes-support/opencv/openvino-inference-engine_2024.0.0.bb b/dynamic-layers/openembedded-layer/recipes-support/opencv/openvino-inference-engine_2024.0.0.bb new file mode 100644 index 00000000..94edd1b8 --- /dev/null +++ b/dynamic-layers/openembedded-layer/recipes-support/opencv/openvino-inference-engine_2024.0.0.bb @@ -0,0 +1,143 @@ +SUMMARY = "OpenVINO(TM) Toolkit - Deep Learning Deployment Toolkit" +HOMEPAGE = "https://github.com/opencv/dldt" +DESCRIPTION = "This toolkit allows developers to deploy pre-trained \ +deep learning models through a high-level C++ Inference Engine API \ +integrated with application logic." + +SRC_URI = "git://github.com/openvinotoolkit/openvino.git;protocol=https;name=openvino;branch=releases/2024/0;lfs=0 \ + git://github.com/openvinotoolkit/oneDNN.git;protocol=https;destsuffix=git/src/plugins/intel_cpu/thirdparty/onednn;name=mkl;nobranch=1 \ + git://github.com/oneapi-src/oneDNN.git;protocol=https;destsuffix=git/src/plugins/intel_gpu/thirdparty/onednn_gpu;name=onednn;nobranch=1 \ + git://github.com/herumi/xbyak.git;protocol=https;destsuffix=git/thirdparty/xbyak;name=xbyak;branch=master \ + git://github.com/nlohmann/json.git;protocol=https;destsuffix=git/thirdparty/json/nlohmann_json;name=json;branch=develop \ + git://github.com/opencv/ade.git;protocol=https;destsuffix=git/thirdparty/ade;name=ade;nobranch=1 \ + git://github.com/protocolbuffers/protobuf.git;protocol=https;destsuffix=git/thirdparty/protobuf/protobuf;name=protobuf;branch=3.20.x \ + git://github.com/gflags/gflags.git;protocol=https;destsuffix=git/thirdparty/gflags/gflags;name=gflags;nobranch=1 \ + git://github.com/madler/zlib.git;protocol=https;destsuffix=git/thirdparty/zlib/zlib;name=zlib;nobranch=1 \ + git://github.com/openvinotoolkit/mlas.git;protocol=https;destsuffix=git/src/plugins/intel_cpu/thirdparty/mlas;name=mlas;nobranch=1 \ + git://github.com/nodejs/node-api-headers.git;protocol=https;destsuffix=git/node-api-headers-src;name=node-api-headers;nobranch=1 \ + git://github.com/nodejs/node-addon-api.git;protocol=https;destsuffix=git/node-addon-api-src;name=node-addon-api;nobranch=1 \ + file://0001-cmake-yocto-specific-tweaks-to-the-build-process.patch \ + file://0003-cmake-Fix-overloaded-virtual-error.patch \ + file://0004-protobuf-allow-target-protoc-to-be-built.patch \ + file://0001-cmake-fix-build-when-using-sysroot.patch \ + file://0001-CPU-Solving-the-build-failure-caused-by-setting-the-.patch \ + " + +SRCREV_openvino = "34caeefd07800b59065345d651949efbe8ab6649" +SRCREV_mkl = "f82148befdbdc9576ec721c9d500155ee4de8060" +SRCREV_onednn = "494af5f9921bdae98f1a0e2955fa7d76ff386c4f" +SRCREV_xbyak = "740dff2e866f3ae1a70dd42d6e8836847ed95cc2" +SRCREV_json = "9cca280a4d0ccf0c08f47a99aa71d1b0e52f8d03" +SRCREV_ade = "0e8a2ccdd34f29dba55894f5f3c5179809888b9e" +SRCREV_protobuf = "fe271ab76f2ad2b2b28c10443865d2af21e27e0e" +SRCREV_gflags = "e171aa2d15ed9eb17054558e0b3a6a413bb01067" +SRCREV_zlib = "09155eaa2f9270dc4ed1fa13e2b4b2613e6e4851" +SRCREV_mlas = "d1bc25ec4660cddd87804fcf03b2411b5dfb2e94" +SRCREV_node-api-headers = "186e04b5e40e54d7fd1655bc67081cc483f12488" +SRCREV_node-addon-api = "39a25bf27788ff7a7ea5c64978c4dcd1e7b9d80d" +SRCREV_FORMAT = "openvino_mkl_onednn_xbyak_json_ade_protobuf_gflags_zlib_node-api-headers_node-addon-api_mlas" + +LICENSE = "Apache-2.0 & MIT & BSD-3-Clause & Zlib" +LIC_FILES_CHKSUM = "file://LICENSE;md5=86d3f3a95c324c9479bd8986968f4327 \ + file://thirdparty/xbyak/COPYRIGHT;md5=3c98edfaa50a86eeaef4c6109e803f16 \ + file://thirdparty/cnpy/LICENSE;md5=689f10b06d1ca2d4b1057e67b16cd580 \ + file://thirdparty/json/nlohmann_json/LICENSE.MIT;md5=f969127d7b7ed0a8a63c2bbeae002588 \ + file://thirdparty/ade/LICENSE;md5=3b83ef96387f14655fc854ddc3c6bd57 \ + file://thirdparty/gflags/gflags/COPYING.txt;md5=c80d1a3b623f72bb85a4c75b556551df \ + file://thirdparty/zlib/zlib/LICENSE;md5=b51a40671bc46e961c0498897742c0b8 \ + file://src/plugins/intel_cpu/thirdparty/mlas/LICENSE;md5=86d3f3a95c324c9479bd8986968f4327 \ + file://src/plugins/intel_cpu/thirdparty/onednn/LICENSE;md5=3b64000f6e7d52516017622a37a94ce9 \ + file://src/plugins/intel_gpu/thirdparty/onednn_gpu/LICENSE;md5=3b64000f6e7d52516017622a37a94ce9 \ + file://node-api-headers-src/LICENSE;md5=6adb2909701d4605b4b2ae1a9b25d8bd \ + file://node-addon-api-src/LICENSE.md;md5=0492ef29a9d558a3e9660e7accc9ca6a \ +" + +inherit cmake python3native pkgconfig qemu + +S = "${WORKDIR}/git" +EXTRA_OECMAKE += " \ + -DCMAKE_CROSSCOMPILING_EMULATOR=${WORKDIR}/qemuwrapper \ + -DENABLE_OPENCV=OFF \ + -DENABLE_INTEL_GNA=OFF \ + -DENABLE_SYSTEM_TBB=ON \ + -DPYTHON_EXECUTABLE=${PYTHON} \ + -DCMAKE_BUILD_TYPE=RelWithDebInfo \ + -DTHREADING=TBB -DTBB_DIR="${STAGING_LIBDIR}/cmake/TBB" \ + -DTREAT_WARNING_AS_ERROR=FALSE \ + -DENABLE_DATA=FALSE \ + -DENABLE_SYSTEM_PUGIXML=TRUE \ + -DENABLE_OV_ONNX_FRONTEND=FALSE \ + -DUSE_BUILD_TYPE_SUBFOLDER=OFF \ + -DENABLE_FUZZING=OFF \ + -DENABLE_TBBBIND_2_5=OFF \ + -DCPACK_GENERATOR=RPM \ + -DENABLE_SYSTEM_FLATBUFFERS=ON \ + -DENABLE_SYSTEM_SNAPPY=ON \ + -DFETCHCONTENT_BASE_DIR="${S}" \ + " + +DEPENDS += "\ + flatbuffers-native \ + pugixml \ + python3-pybind11 \ + python3-pybind11-native \ + qemu-native \ + snappy \ + tbb \ + " + +COMPATIBLE_HOST = '(x86_64).*-linux' +COMPATIBLE_HOST:libc-musl = "null" + +PACKAGECONFIG ?= "opencl samples" +PACKAGECONFIG[opencl] = "-DENABLE_INTEL_GPU=TRUE, -DENABLE_INTEL_GPU=FALSE, virtual/opencl-icd opencl-headers opencl-clhpp," +PACKAGECONFIG[python3] = "-DENABLE_PYTHON=ON -DPYTHON_LIBRARY=${PYTHON_LIBRARY} -DPYTHON_INCLUDE_DIR=${PYTHON_INCLUDE_DIR} -DENABLE_PYTHON_PACKAGING=ON, -DENABLE_PYTHON=OFF, patchelf-native, python3 python3-numpy python3-progress" +PACKAGECONFIG[samples] = "-DENABLE_SAMPLES=ON -DENABLE_COMPILE_TOOL=ON, -DENABLE_SAMPLES=OFF -DENABLE_COMPILE_TOOL=OFF, opencv" +PACKAGECONFIG[verbose] = "-DVERBOSE_BUILD=1,-DVERBOSE_BUILD=0" + +do_configure:prepend() { + # Dont set PROJECT_ROOT_DIR + sed -i -e 's:\${OpenVINO_SOURCE_DIR}::;' ${S}/src/CMakeLists.txt + + # qemu wrapper that can be used by cmake to run target binaries. + qemu_binary="${@qemu_wrapper_cmdline(d, d.getVar('STAGING_DIR_HOST'), [d.expand('${STAGING_DIR_HOST}${libdir}'),d.expand('${STAGING_DIR_HOST}${base_libdir}')])}" + cat > ${WORKDIR}/qemuwrapper << EOF +#!/bin/sh +$qemu_binary "\$@" +EOF + chmod +x ${WORKDIR}/qemuwrapper +} + +do_install:append() { + rm -rf ${D}${prefix}/install_dependencies + rm -rf ${D}${prefix}/setupvars.sh + + find ${B}/src/plugins/intel_cpu/cross-compiled/ -type f -name *_disp.cpp -exec sed -i -e 's%'"${S}"'%'"${TARGET_DBGSRC_DIR}"'%g' {} + +} + +# Otherwise e.g. ros-openvino-toolkit-dynamic-vino-sample when using dldt-inference-engine uses dldt-inference-engine WORKDIR +# instead of RSS +SSTATE_SCAN_FILES:append = " *.cmake" + +FILES:${PN} += "\ + ${libdir}/openvino-${PV}/lib*${SOLIBSDEV} \ + ${libdir}/openvino-${PV}/plugins.xml \ + ${libdir}/openvino-${PV}/cache.json \ + " + +# Move inference engine samples into a separate package +PACKAGES =+ "${PN}-samples" + +FILES:${PN}-samples = "${datadir}/openvino \ + ${bindir} \ + ${libdir}/libformat_reader.a \ + ${libdir}/libopencv_c_wrapper.a \ + " +RDEPENDS:${PN}-samples += "python3-core" + +# Package for inference engine python API +PACKAGES =+ "${PN}-python3" + +FILES:${PN}-python3 = "${PYTHON_SITEPACKAGES_DIR}" + +UPSTREAM_CHECK_GITTAGREGEX = "(?P<pver>(\d+\.\d+\.\d+))$" diff --git a/lib/oeqa/runtime/cases/jhi.py b/lib/oeqa/runtime/cases/jhi.py deleted file mode 100644 index 7a24b6da..00000000 --- a/lib/oeqa/runtime/cases/jhi.py +++ /dev/null @@ -1,41 +0,0 @@ -import os -from oeqa.runtime.decorator.package import OEHasPackage -from oeqa.runtime.case import OERuntimeTestCase -from oeqa.core.decorator.depends import OETestDepends - -class JhiTest(OERuntimeTestCase): - - @classmethod - def tearDownClass(cls): - _, output = cls.tc.target.run('pidof jhid') - cls.tc.target.run('kill %s' % output) - - @OEHasPackage(['openssh-sshd']) - @OEHasPackage(['jhi']) - def test_jhi_mei_driver(self): - command = 'ls /dev/mei*' - (status, output) = self.target.run(command) - self.assertEqual(status, 0, msg="Error messages: %s" % output) - - @OETestDepends(['jhi.JhiTest.test_jhi_mei_driver']) - def test_jhi_daemon_version(self): - command = 'jhid -v' - (status, output) = self.target.run(command) - self.assertEqual(status, 0, msg="Error messages: %s" % output) - - @OETestDepends(['jhi.JhiTest.test_jhi_mei_driver']) - def test_jhi_daemon_can_initialized(self): - command = 'jhid -d' - (status, output) = self.target.run(command) - self.assertEqual(status, 0, msg="Error messages: %s" % output) - - @OEHasPackage(['jhi-test']) - @OETestDepends(['jhi.JhiTest.test_jhi_daemon_can_initialized']) - def test_jhi_bist(self): - (status, output) = self.target.run('uname -m') - self.assertEqual(status, 0, msg='status and output: %s and %s' % (status, output)) - if 'x86_64' not in output: - self.skipTest("Skipped jhi bist test if not x86_64 machine (current machine: %s)." % output) - command = 'bist' - (status, output) = self.target.run(command) - self.assertEqual(status, 0, msg="Error messages: %s" % output) diff --git a/lib/oeqa/runtime/cases/parselogs-ignores-intel-core2-32.txt b/lib/oeqa/runtime/cases/parselogs-ignores-intel-core2-32.txt new file mode 100644 index 00000000..84ce8168 --- /dev/null +++ b/lib/oeqa/runtime/cases/parselogs-ignores-intel-core2-32.txt @@ -0,0 +1,9 @@ +# These should be reviewed to see if they are still needed +ACPI: No _BQC method, cannot determine initial brightness +[Firmware Bug]: ACPI: No _BQC method, cannot determine initial brightness +(EE) Failed to load module "psb" +(EE) Failed to load module "psbdrv" +(EE) open /dev/fb0: No such file or directory +(EE) AIGLX: reverting to software rendering +dmi: Firmware registration failed. +ioremap error for 0x78 diff --git a/lib/oeqa/runtime/cases/parselogs-ignores-intel-corei7-64.txt b/lib/oeqa/runtime/cases/parselogs-ignores-intel-corei7-64.txt new file mode 100644 index 00000000..5c9b4bc7 --- /dev/null +++ b/lib/oeqa/runtime/cases/parselogs-ignores-intel-corei7-64.txt @@ -0,0 +1,14 @@ +# These should be reviewed to see if they are still needed +can't set Max Payload Size to 256 +intel_punit_ipc: can't request region for resource +[drm] parse error at position 4 in video mode 'efifb' +ACPI Error: Could not enable RealTimeClock event +ACPI Warning: Could not enable fixed event - RealTimeClock +hci_intel INT33E1:00: Unable to retrieve gpio +hci_intel: probe of INT33E1:00 failed +can't derive routing for PCI INT A +failed to read out thermal zone +Bluetooth: hci0: Setting Intel event mask failed +ttyS2 - failed to request DMA +Bluetooth: hci0: Failed to send firmware data (-38) +atkbd serio0: Failed to enable keyboard on isa0060/serio0 diff --git a/recipes-bsp/formfactor/formfactor_0.0.bbappend b/recipes-bsp/formfactor/formfactor_0.0.bbappend index 6dd422ae..d21c3bee 100644 --- a/recipes-bsp/formfactor/formfactor_0.0.bbappend +++ b/recipes-bsp/formfactor/formfactor_0.0.bbappend @@ -1 +1 @@ -FILESEXTRAPATHS_prepend_intel-x86-common := "${THISDIR}/${PN}:" +FILESEXTRAPATHS:prepend:intel-x86-common := "${THISDIR}/${PN}:" diff --git a/recipes-bsp/intel-cmt-cat/intel-cmt-cat_23.11.1.bb b/recipes-bsp/intel-cmt-cat/intel-cmt-cat_23.11.1.bb new file mode 100644 index 00000000..60d0dfd2 --- /dev/null +++ b/recipes-bsp/intel-cmt-cat/intel-cmt-cat_23.11.1.bb @@ -0,0 +1,29 @@ +SUMMARY = "intel-cmt-cat" +DESCRIPTION = "Software package which provides basic support for Intel(R) \ +Resource Director Technology (Intel(R) RDT)" +HOMEPAGE = "https://github.com/intel/intel-cmt-cat" + +LICENSE = "BSD-3-Clause" +LIC_FILES_CHKSUM = "file://LICENSE;md5=4b63c65942e1c16fd897f8cd20abebf8" + +SRC_URI = "git://github.com/intel/intel-cmt-cat;protocol=https;branch=master" +SRCREV = "b26b31b0ae6980c5939a421cefe0316cae884626" + +S = "${WORKDIR}/git" + +COMPATIBLE_HOST = '(x86_64).*-linux' +COMPATIBLE_HOST:libc-musl = "null" + +do_install() { + oe_runmake install PREFIX=${D}${prefix} NOLDCONFIG=y +} + +FILES:${PN} += "${nonarch_libdir}/libpqos*" +FILES:${PN}-doc = "/usr/man*" + +INSANE_SKIP:${PN} += "ldflags" +INSANE_SKIP:${PN} += "dev-so" +INSANE_SKIP:${PN} += "libdir" +INSANE_SKIP:${PN} += "already-stripped" + +INHIBIT_PACKAGE_DEBUG_SPLIT = "1" diff --git a/recipes-bsp/metee/metee_3.0.0.bb b/recipes-bsp/metee/metee_3.2.4.bb index 0f970a87..da8220b9 100644 --- a/recipes-bsp/metee/metee_3.0.0.bb +++ b/recipes-bsp/metee/metee_3.2.4.bb @@ -10,8 +10,9 @@ COMPATIBLE_HOST = '(i.86|x86_64).*-linux' inherit cmake -SRC_URI = "git://github.com/intel/metee.git" -SRCREV = "92e1597181d7faec3196847a4bf35ea1f881383f" +SRC_URI = "git://github.com/intel/metee.git;branch=master;protocol=https \ +" +SRCREV = "db45e37e146fd9c06907a15ade55eba06ad1f951" S = "${WORKDIR}/git" diff --git a/recipes-core/images/core-image-minimal-initramfs.bbappend b/recipes-core/images/core-image-minimal-initramfs.bbappend index 132f15b5..90ac28fd 100644 --- a/recipes-core/images/core-image-minimal-initramfs.bbappend +++ b/recipes-core/images/core-image-minimal-initramfs.bbappend @@ -1,2 +1,2 @@ # Add i915 graphics firmware -PACKAGE_INSTALL_append_intel-x86-common = " linux-firmware-i915" +PACKAGE_INSTALL:append:intel-x86-common = " linux-firmware-i915" diff --git a/recipes-core/images/core-image-tiny.bb b/recipes-core/images/core-image-tiny.bb index 6b09b0c2..f521f668 100644 --- a/recipes-core/images/core-image-tiny.bb +++ b/recipes-core/images/core-image-tiny.bb @@ -32,5 +32,5 @@ python() { d.appendVarFlag('do_image', 'depends', ' %s:do_image_complete' % initrd_i) } -WKS_FILE_intel-corei7-64 = "core-image-tiny.wks.in" -WKS_FILE_intel-core2-32 = "core-image-tiny.wks.in" +WKS_FILE:intel-corei7-64 = "core-image-tiny.wks.in" +WKS_FILE:intel-core2-32 = "core-image-tiny.wks.in" diff --git a/recipes-core/libxcam/libxcam/0001-fake_v4l2_device.h-fix-narrowing-warning.patch b/recipes-core/libxcam/libxcam/0001-fake_v4l2_device.h-fix-narrowing-warning.patch new file mode 100644 index 00000000..f15d5cc7 --- /dev/null +++ b/recipes-core/libxcam/libxcam/0001-fake_v4l2_device.h-fix-narrowing-warning.patch @@ -0,0 +1,94 @@ +From d4c97b50b577ea16b9ff6d9a352ab474a119310e Mon Sep 17 00:00:00 2001 +From: Naveen Saini <naveen.kumar.saini@intel.com> +Date: Wed, 18 May 2022 15:41:24 +0800 +Subject: [PATCH] fake_v4l2_device.h: fix narrowing warning + +Use uint32_t instead of int for IOCTLs commands. + +Warning log: +| ../../../git/xcore/fake_v4l2_device.h: In member function 'virtual int XCam::FakeV4l2Device::io_control(int, void*)': +| ../../../git/xcore/fake_v4l2_device.h:42:14: error: narrowing conversion of '3225441794' from 'long unsigned int' to 'int' [-Wnarrowing] +| 42 | case VIDIOC_ENUM_FMT: +| | ^~~~~~~~~~~~~~~ +| make[4]: *** [Makefile:685: libgstxcamsrc_la-gstxcamsrc.lo] Error 1 + +Issue: https://github.com/intel/libxcam/issues/801 +Upstream-Status: Submitted [https://github.com/intel/libxcam/pull/802] + +Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> +--- + xcore/base/xcam_common.h | 2 +- + xcore/fake_v4l2_device.h | 2 +- + xcore/v4l2_device.cpp | 2 +- + xcore/v4l2_device.h | 2 +- + xcore/xcam_common.cpp | 2 +- + 5 files changed, 5 insertions(+), 5 deletions(-) + +diff --git a/xcore/base/xcam_common.h b/xcore/base/xcam_common.h +index 1f16e1e..4aa6cb9 100644 +--- a/xcore/base/xcam_common.h ++++ b/xcore/base/xcam_common.h +@@ -75,7 +75,7 @@ void xcam_free (void *ptr); + * return, 0 successfully + * else, check errno + */ +-int xcam_device_ioctl (int fd, int cmd, void *arg); ++int xcam_device_ioctl (int fd, uint32_t cmd, void *arg); + const char *xcam_fourcc_to_string (uint32_t fourcc); + + void xcam_set_log (const char* file_name); +diff --git a/xcore/fake_v4l2_device.h b/xcore/fake_v4l2_device.h +index f679c19..e29787d 100644 +--- a/xcore/fake_v4l2_device.h ++++ b/xcore/fake_v4l2_device.h +@@ -33,7 +33,7 @@ public: + : V4l2Device ("/dev/null") + {} + +- int io_control (int cmd, void *arg) ++ int io_control (uint32_t cmd, void *arg) + { + XCAM_UNUSED (arg); + +diff --git a/xcore/v4l2_device.cpp b/xcore/v4l2_device.cpp +index 395461e..66a8ac6 100644 +--- a/xcore/v4l2_device.cpp ++++ b/xcore/v4l2_device.cpp +@@ -185,7 +185,7 @@ V4l2Device::close () + } + + int +-V4l2Device::io_control (int cmd, void *arg) ++V4l2Device::io_control (uint32_t cmd, void *arg) + + { + if (_fd <= 0) +diff --git a/xcore/v4l2_device.h b/xcore/v4l2_device.h +index b4ad7ad..2551a92 100644 +--- a/xcore/v4l2_device.h ++++ b/xcore/v4l2_device.h +@@ -104,7 +104,7 @@ public: + XCamReturn queue_buffer (SmartPtr<V4l2Buffer> &buf); + + // use as less as possible +- virtual int io_control (int cmd, void *arg); ++ virtual int io_control (uint32_t cmd, void *arg); + + protected: + +diff --git a/xcore/xcam_common.cpp b/xcore/xcam_common.cpp +index 848884d..d4d5093 100644 +--- a/xcore/xcam_common.cpp ++++ b/xcore/xcam_common.cpp +@@ -53,7 +53,7 @@ void xcam_free(void *ptr) + free (ptr); + } + +-int xcam_device_ioctl (int fd, int cmd, void *arg) ++int xcam_device_ioctl (int fd, uint32_t cmd, void *arg) + { + int ret = 0; + int tried_time = 0; +-- +2.25.1 + diff --git a/recipes-core/libxcam/libxcam_1.5.0.bb b/recipes-core/libxcam/libxcam_1.5.0.bb index fcd14bd4..31e9344b 100644 --- a/recipes-core/libxcam/libxcam_1.5.0.bb +++ b/recipes-core/libxcam/libxcam_1.5.0.bb @@ -8,10 +8,11 @@ inherit autotools pkgconfig S = "${WORKDIR}/git" SRCREV = "231a1d5243cd45c7a6b511b667f1ec52178fdda8" -SRC_URI = "git://github.com/intel/libxcam.git;branch=1.5.0; \ +SRC_URI = "git://github.com/intel/libxcam.git;branch=1.5.0;protocol=https \ + file://0001-fake_v4l2_device.h-fix-narrowing-warning.patch \ " -COMPATIBLE_HOST_libc-musl = "null" +COMPATIBLE_HOST:libc-musl = "null" PACKAGECONFIG ??= " gst \ ${@bb.utils.contains("DISTRO_FEATURES", "opengl", "gles", "", d)} \ @@ -27,12 +28,12 @@ PACKAGECONFIG[gles] = "--enable-gles, --disable-gles, virtual/mesa" PACKAGECONFIG[vulkan] = "--enable-vulkan, --disable-vulkan, vulkan-loader virtual/mesa" PACKAGECONFIG[dnn] = "--enable-dnn, --disable-dnn," -do_install_append () { +do_install:append () { install -d ${D}${bindir}/libxcam cp -r ${WORKDIR}/build/tests/.libs/* ${D}${bindir}/libxcam/ } -FILES_${PN} += "${libdir}/gstreamer-*/*.so" -FILES_${PN}-test = "${bindir}/libxcam/*" +FILES:${PN} += "${libdir}/gstreamer-*/*.so" +FILES:${PN}-test = "${bindir}/libxcam/*" PACKAGES =+ "${PN}-test" -RDEPENDS_${PN}-test =+ "bash" +RDEPENDS:${PN}-test =+ "bash" diff --git a/recipes-core/meta/icx-environment.inc b/recipes-core/meta/icx-environment.inc new file mode 100644 index 00000000..65322a91 --- /dev/null +++ b/recipes-core/meta/icx-environment.inc @@ -0,0 +1,15 @@ +export ICX_LDFLAGS_OPTION = " -Wl,-dynamic-linker,/lib/ld-linux-x86-64.so.2" +export ICXSDK_PREFIX_OPTION = "-B ${TARGET_PREFIX}" +export ICXQSDK_PREFIX_OPTION = "-qgnu-prefix=${TARGET_PREFIX}" + +create_sdk_files:append() { + script=${SDK_OUTPUT}/${SDKPATH}/environment-setup-${REAL_MULTIMACH_TARGET_SYS} + if ${@bb.utils.contains('ICXSDK', '1', 'true', 'false', d)}; then + echo 'export ICX="icx ${ICXSDK_PREFIX_OPTION} --target=${TARGET_SYS} ${TARGET_CC_ARCH} --sysroot=$SDKTARGETSYSROOT ${ICX_LDFLAGS_OPTION}"' >> $script + echo 'export ICXCXX="icpx ${ICXSDK_PREFIX_OPTION} --target=${TARGET_SYS} ${TARGET_CC_ARCH} --sysroot=$SDKTARGETSYSROOT ${ICX_LDFLAGS_OPTION}"' >> $script + echo 'export ICXCPP="icx ${ICXSDK_PREFIX_OPTION} -E --target=${TARGET_SYS} ${TARGET_CC_ARCH} --sysroot=$SDKTARGETSYSROOT ${ICX_LDFLAGS_OPTION}"' >> $script + echo 'export ICXLD="xild --sysroot=$SDKTARGETSYSROOT "' >> $script + echo 'export ICXCCLD="icx ${ICXSDK_PREFIX_OPTION} --sysroot=$SDKTARGETSYSROOT"' >> $script + echo 'export ICXAR="xiar ${ICXQSDK_PREFIX_OPTION}"' >> $script + fi +} diff --git a/recipes-core/meta/meta-environment-extsdk.bbappend b/recipes-core/meta/meta-environment-extsdk.bbappend new file mode 100644 index 00000000..2cff6fc4 --- /dev/null +++ b/recipes-core/meta/meta-environment-extsdk.bbappend @@ -0,0 +1,3 @@ +FILESEXTRAPATHS:prepend := "${THISDIR}/${PN}:" + +require ${@bb.utils.contains('ICXSDK', '1', 'icx-environment.inc', '', d)} diff --git a/recipes-core/meta/meta-environment.bbappend b/recipes-core/meta/meta-environment.bbappend new file mode 100644 index 00000000..2cff6fc4 --- /dev/null +++ b/recipes-core/meta/meta-environment.bbappend @@ -0,0 +1,3 @@ +FILESEXTRAPATHS:prepend := "${THISDIR}/${PN}:" + +require ${@bb.utils.contains('ICXSDK', '1', 'icx-environment.inc', '', d)} diff --git a/recipes-core/microcode/intel-microcode_20210216.bb b/recipes-core/microcode/intel-microcode_20240312.bb index 69c8f08c..00b18231 100644 --- a/recipes-core/microcode/intel-microcode_20210216.bb +++ b/recipes-core/microcode/intel-microcode_20240312.bb @@ -16,7 +16,7 @@ LIC_FILES_CHKSUM = "file://license;md5=d8405101ec6e90c1d84b082b0c40c721" SRC_URI = "git://github.com/intel/Intel-Linux-Processor-Microcode-Data-Files.git;protocol=https;branch=main \ " -SRCREV = "4c5f3fdcc7f5d2b6c1bcc11001ed58923db481ab" +SRCREV = "41af34500598418150aa298bb04e7edacc547897" DEPENDS = "iucode-tool-native" S = "${WORKDIR}/git" @@ -34,12 +34,6 @@ do_compile() { ${STAGING_DIR_NATIVE}${sbindir_native}/iucode_tool \ ${UCODE_FILTER_PARAMETERS} \ --overwrite \ - --write-to=${WORKDIR}/microcode_${PV}.bin \ - ${S}/intel-ucode/* ${S}/intel-ucode-with-caveats/* - - ${STAGING_DIR_NATIVE}${sbindir_native}/iucode_tool \ - ${UCODE_FILTER_PARAMETERS} \ - --overwrite \ --write-earlyfw=${WORKDIR}/microcode_${PV}.cpio \ ${S}/intel-ucode/* ${S}/intel-ucode-with-caveats/* } @@ -47,6 +41,7 @@ do_compile() { do_install() { install -d ${D}${nonarch_base_libdir}/firmware/intel-ucode/ ${STAGING_DIR_NATIVE}${sbindir_native}/iucode_tool \ + ${UCODE_FILTER_PARAMETERS} \ --write-firmware=${D}${nonarch_base_libdir}/firmware/intel-ucode \ ${S}/intel-ucode/* ${S}/intel-ucode-with-caveats/* } @@ -63,6 +58,6 @@ addtask deploy before do_build after do_compile PACKAGES = "${PN}" -FILES_${PN} = "${nonarch_base_libdir}" +FILES:${PN} = "${nonarch_base_libdir}" UPSTREAM_CHECK_GITTAGREGEX = "^microcode-(?P<pver>(\d+)[a-z]*)$" diff --git a/recipes-core/microcode/iucode-tool_2.3.1.bb b/recipes-core/microcode/iucode-tool_2.3.1.bb index df74a8e9..19417b7c 100644 --- a/recipes-core/microcode/iucode-tool_2.3.1.bb +++ b/recipes-core/microcode/iucode-tool_2.3.1.bb @@ -12,14 +12,14 @@ DESCRIPTION = "iucode_tool is a program to manipulate Intel i686 and X86-64\ HOMEPAGE = "https://gitlab.com/iucode-tool/" BUGTRACKER = "https://bugs.debian.org/cgi-bin/pkgreport.cgi?ordering=normal;archive=0;src=iucode-tool;repeatmerged=0" -LICENSE = "GPLv2+" +LICENSE = "GPL-2.0-or-later" LIC_FILES_CHKSUM = "file://COPYING;md5=751419260aa954499f7abaabaa882bbe \ file://iucode_tool.c;beginline=1;endline=15;md5=71eeab3190360ff0267101b570874756" -DEPENDS_append_libc-musl = " argp-standalone" +DEPENDS:append:libc-musl = " argp-standalone" SRC_URI = "https://gitlab.com/iucode-tool/releases/raw/master/iucode-tool_${PV}.tar.xz" -SRC_URI_append_libc-musl = " file://0001-Makefile.am-Add-arg-parse-library-for-MUSL-support.patch" +SRC_URI:append:libc-musl = " file://0001-Makefile.am-Add-arg-parse-library-for-MUSL-support.patch" SRC_URI[md5sum] = "63b33cc0ea1f8c73b443412abbf39d6f" SRC_URI[sha256sum] = "12b88efa4d0d95af08db05a50b3dcb217c0eb2bfc67b483779e33d498ddb2f95" diff --git a/recipes-core/ovmf/files/0001-ovmf-RefkitTestCA-TEST-UEFI-SecureBoot.patch b/recipes-core/ovmf/files/0001-ovmf-RefkitTestCA-TEST-UEFI-SecureBoot.patch index c13363d2..7eb3bc69 100644 --- a/recipes-core/ovmf/files/0001-ovmf-RefkitTestCA-TEST-UEFI-SecureBoot.patch +++ b/recipes-core/ovmf/files/0001-ovmf-RefkitTestCA-TEST-UEFI-SecureBoot.patch @@ -1,6 +1,6 @@ -From 149b34de80e7f9f5331c285b862776cdd4b02ee3 Mon Sep 17 00:00:00 2001 +From b2099e7184d48a6d05c8713b6fd5dac0e2e70963 Mon Sep 17 00:00:00 2001 From: Mikko Ylinen <mikko.ylinen@linux.intel.com> -Date: Fri, 7 Apr 2017 12:06:14 +0300 +Date: Wed, 2 Mar 2022 10:55:35 +0800 Subject: [PATCH] ovmf: RefkitTestCA: TEST UEFI SecureBoot This patch adds refkit-db.cer (via xxd -i) in OVMF's db @@ -9,25 +9,23 @@ application. It's used for testing purposes only. Images signed with refkit-db keys are allowed to boot. -Signed-off-by: Mikko Ylinen <mikko.ylinen@linux.intel.com> - -[ Change location of key to AuthData.c ] -Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> +Upstream-Status: Inappropriate -%% original patch: 0001-ovmf-RefkitTestCA-TEST-UEFI-SecureBoot.patch +Signed-off-by: Mikko Ylinen <mikko.ylinen@linux.intel.com> +Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> --- - OvmfPkg/EnrollDefaultKeys/AuthData.c | 68 +++++++++++++++++++ - OvmfPkg/EnrollDefaultKeys/EnrollDefaultKeys.c | 1 + + OvmfPkg/EnrollDefaultKeys/AuthData.c | 69 +++++++++++++++++++ + OvmfPkg/EnrollDefaultKeys/EnrollDefaultKeys.c | 3 + OvmfPkg/EnrollDefaultKeys/EnrollDefaultKeys.h | 2 + - 3 files changed, 71 insertions(+) + 3 files changed, 74 insertions(+) diff --git a/OvmfPkg/EnrollDefaultKeys/AuthData.c b/OvmfPkg/EnrollDefaultKeys/AuthData.c -index 3b4856a01f..7109766296 100644 +index 53ee7f7003..127131cd05 100644 --- a/OvmfPkg/EnrollDefaultKeys/AuthData.c +++ b/OvmfPkg/EnrollDefaultKeys/AuthData.c -@@ -398,6 +398,74 @@ CONST UINT8 mMicrosoftUefiCa[] = { +@@ -395,6 +395,75 @@ CONST UINT8 mMicrosoftUefiCa[] = { - CONST UINTN mSizeOfMicrosoftUefiCa = sizeof mMicrosoftUefiCa;
+ CONST UINTN mSizeOfMicrosoftUefiCa = sizeof mMicrosoftUefiCa;
+CONST UINT8 mRefkitTestCA[] = {
+ 0x30, 0x82, 0x02, 0xfb, 0x30, 0x82, 0x01, 0xe3, 0xa0, 0x03, 0x02, 0x01,
@@ -97,32 +95,35 @@ index 3b4856a01f..7109766296 100644 +};
+
+CONST UINTN mSizeOfRefkitTestCA = sizeof mRefkitTestCA;
-
++
//
// The Microsoft.UefiSecureBootLogo.Tests.OutOfBoxConfirmDBXisPresent test case
+ // of the Secure Boot Logo Test in the Microsoft Hardware Certification Kit
diff --git a/OvmfPkg/EnrollDefaultKeys/EnrollDefaultKeys.c b/OvmfPkg/EnrollDefaultKeys/EnrollDefaultKeys.c -index f45cb799f7..fb1252e768 100644 +index 094e4c821b..0a7eef54dc 100644 --- a/OvmfPkg/EnrollDefaultKeys/EnrollDefaultKeys.c +++ b/OvmfPkg/EnrollDefaultKeys/EnrollDefaultKeys.c -@@ -615,6 +615,7 @@ ShellAppMain ( - &gEfiCertX509Guid,
- mMicrosoftPca, mSizeOfMicrosoftPca, &gMicrosoftVendorGuid,
- mMicrosoftUefiCa, mSizeOfMicrosoftUefiCa, &gMicrosoftVendorGuid,
-+ mRefkitTestCA, mSizeOfRefkitTestCA, &gEfiCallerIdGuid,
- NULL);
+@@ -702,6 +702,9 @@ ShellAppMain ( + mMicrosoftUefiCa,
+ mSizeOfMicrosoftUefiCa,
+ &gMicrosoftVendorGuid,
++ mRefkitTestCA,
++ mSizeOfRefkitTestCA,
++ &gEfiCallerIdGuid,
+ NULL
+ );
}
- if (EFI_ERROR (Status)) {
diff --git a/OvmfPkg/EnrollDefaultKeys/EnrollDefaultKeys.h b/OvmfPkg/EnrollDefaultKeys/EnrollDefaultKeys.h -index 8e61f0a77b..e3fa046f95 100644 +index 56da9c71d6..8de1dfe4e0 100644 --- a/OvmfPkg/EnrollDefaultKeys/EnrollDefaultKeys.h +++ b/OvmfPkg/EnrollDefaultKeys/EnrollDefaultKeys.h -@@ -135,4 +135,6 @@ extern CONST UINTN mSizeOfMicrosoftUefiCa; - extern CONST UINT8 mSha256OfDevNull[];
- extern CONST UINTN mSizeOfSha256OfDevNull;
+@@ -133,4 +133,6 @@ extern CONST UINTN mSizeOfMicrosoftUefiCa; + extern CONST UINT8 mSha256OfDevNull[];
+ extern CONST UINTN mSizeOfSha256OfDevNull;
+extern CONST UINT8 mRefkitTestCA[];
+extern CONST UINTN mSizeOfRefkitTestCA;
#endif /* ENROLL_DEFAULT_KEYS_H_ */
-- -2.21.0 +2.17.1 diff --git a/recipes-core/ovmf/ovmf-shell-image-enrollkeys.bb b/recipes-core/ovmf/ovmf-shell-image-enrollkeys.bb index b20f6e58..ca3cfc15 100644 --- a/recipes-core/ovmf/ovmf-shell-image-enrollkeys.bb +++ b/recipes-core/ovmf/ovmf-shell-image-enrollkeys.bb @@ -1,10 +1,10 @@ require recipes-core/ovmf/ovmf-shell-image.bb -WKS_SEARCH_PATH_append = ":${COREBASE}/meta/recipes-core/ovmf" +WKS_SEARCH_PATH:append = ":${COREBASE}/meta/recipes-core/ovmf" QB_DRIVE_TYPE = "/dev/vd" -do_image_append() { +do_image:append() { cat > ${IMAGE_ROOTFS}/startup.nsh << EOF EnrollDefaultKeys reset diff --git a/recipes-core/ovmf/ovmf_%.bbappend b/recipes-core/ovmf/ovmf_%.bbappend index bbf5fa32..34a9dd1b 100644 --- a/recipes-core/ovmf/ovmf_%.bbappend +++ b/recipes-core/ovmf/ovmf_%.bbappend @@ -1,6 +1,6 @@ -FILESEXTRAPATHS_prepend_intel-x86-common := "${THISDIR}/files:" +FILESEXTRAPATHS:prepend:intel-x86-common := "${THISDIR}/files:" -SRC_URI_append_intel-x86-common = " \ +SRC_URI:append:intel-x86-common = " \ file://0001-ovmf-RefkitTestCA-TEST-UEFI-SecureBoot.patch \ " -PACKAGECONFIG_append_intel-x86-common = " secureboot" +PACKAGECONFIG:append:intel-x86-common = " secureboot" diff --git a/recipes-core/packagegroups/nativesdk-packagegroup-sdk-host.bbappend b/recipes-core/packagegroups/nativesdk-packagegroup-sdk-host.bbappend new file mode 100644 index 00000000..b5a5fdd4 --- /dev/null +++ b/recipes-core/packagegroups/nativesdk-packagegroup-sdk-host.bbappend @@ -0,0 +1 @@ +RDEPENDS:${PN} += "${@bb.utils.contains('ICXSDK', '1', ' intel-oneapi-dpcpp-cpp intel-oneapi-dpcpp-cpp-dev intel-oneapi-dpcpp-cpp-runtime intel-oneapi-dpcpp-cpp-runtime-dev ', '', d)}" diff --git a/recipes-core/packagegroups/packagegroup-core-standalone-sdk-target.bbappend b/recipes-core/packagegroups/packagegroup-core-standalone-sdk-target.bbappend new file mode 100644 index 00000000..608ea3d3 --- /dev/null +++ b/recipes-core/packagegroups/packagegroup-core-standalone-sdk-target.bbappend @@ -0,0 +1 @@ +RRECOMMENDS:${PN}:append:intel-x86-common = "${@bb.utils.contains('ICXSDK', '1', ' intel-oneapi-dpcpp-cpp intel-oneapi-dpcpp-cpp-dev intel-oneapi-dpcpp-cpp-runtime intel-oneapi-dpcpp-cpp-runtime-dev ', '', d)}" diff --git a/recipes-core/zlib/zlib-intel_1.2.11.1.jtkv6.3.bb b/recipes-core/zlib/zlib-intel_1.2.11.1.jtkv6.3.bb deleted file mode 100644 index 65303bff..00000000 --- a/recipes-core/zlib/zlib-intel_1.2.11.1.jtkv6.3.bb +++ /dev/null @@ -1,18 +0,0 @@ -require recipes-core/zlib/zlib_1.2.11.bb - -FILESEXTRAPATHS_append = ":${COREBASE}/meta/recipes-core/zlib/zlib" - -SUMMARY = "Zlib Compression Library (Intel-tuned)" - -PROVIDES = "zlib" - -SRC_URI_remove = "${SOURCEFORGE_MIRROR}/libpng/${BPN}/${PV}/${BPN}-${PV}.tar.xz" -SRC_URI_prepend = "git://github.com/jtkukunas/zlib.git;protocol=git " - -SRCREV = "a43a247bfa16ec5368747b5b64f11ea5ca033010" -S = "${WORKDIR}/git" - -RPROVIDES_${PN} += "zlib" -RPROVIDES_${PN}-ptest += "zlib-ptest" - -UPSTREAM_CHECK_GITTAGREGEX = "^v(?P<pver>(\d+(\.\d+)+)_jtkv(\d+(\.\d+)+))$" diff --git a/recipes-devtools/slimboot/slimboot-tools_git.bb b/recipes-devtools/slimboot/slimboot-tools_git.bb index 7d8b66be..b65b182d 100644 --- a/recipes-devtools/slimboot/slimboot-tools_git.bb +++ b/recipes-devtools/slimboot/slimboot-tools_git.bb @@ -4,11 +4,11 @@ binary file format understood by Slim Bootloader to load and initialize\ Operating Systems or Hypervisors." HOMEPAGE = "https://slimbootloader.github.io/tools/index.html" -SRC_URI = "git://github.com/slimbootloader/slimbootloader;protocol=https" -SRCREV = "9f146afd47e0ca204521826a583d55388850b216" +SRC_URI = "git://github.com/slimbootloader/slimbootloader;protocol=https;branch=master" +SRCREV = "df5bd0bc2a522afcb8945a6797592b04838db753" PV = "0.0.0+git${SRCPV}" LICENSE = "BSD-2-Clause-Patent" -LIC_FILES_CHKSUM = "file://LICENSE;md5=d1ed89007e7aa232a4dc1c59b6c9efc4" +LIC_FILES_CHKSUM = "file://LICENSE;md5=ef7fba7be2819ac13aaf5d0f842ce5d9" S = "${WORKDIR}/git" inherit python3native diff --git a/recipes-graphics/gmmlib/files/0001-Add-new-DG2-device-IDs-194.patch b/recipes-graphics/gmmlib/files/0001-Add-new-DG2-device-IDs-194.patch new file mode 100644 index 00000000..21cffe24 --- /dev/null +++ b/recipes-graphics/gmmlib/files/0001-Add-new-DG2-device-IDs-194.patch @@ -0,0 +1,38 @@ +From 8729d66800995a9b7e693bd78249932bd12383e1 Mon Sep 17 00:00:00 2001 +From: John Machado <john.machado@intel.com> +Date: Mon, 25 Mar 2024 11:13:27 +0530 +Subject: [PATCH] Add new DG2 device IDs (#194) + +Upstream-status: Backport +https://github.com/intel/gmmlib/commit/dcc4b85ea3728c848b62fbb164c00f43b519b6fb + +Signed-off-by: Hoe, Sheng Yang <sheng.yang.hoe@intel.com> +--- + Source/inc/common/igfxfmid.h | 4 ++++ + 1 file changed, 4 insertions(+) + +diff --git a/Source/inc/common/igfxfmid.h b/Source/inc/common/igfxfmid.h +index 7969779..8e2c955 100644 +--- a/Source/inc/common/igfxfmid.h ++++ b/Source/inc/common/igfxfmid.h +@@ -1932,6 +1932,8 @@ typedef enum __NATIVEGTTYPE + #define DEV_ID_56BB 0x56BB + #define DEV_ID_56BC 0x56BC + #define DEV_ID_56BD 0x56BD ++#define DEV_ID_56BE 0x56BE ++#define DEV_ID_56BF 0x56BF + #define DEV_ID_56C0 0x56C0 + #define DEV_ID_56C1 0x56C1 + +@@ -1992,6 +1994,8 @@ typedef enum __NATIVEGTTYPE + ( d == DEV_ID_5690 ) || \ + ( d == DEV_ID_5691 ) || \ + ( d == DEV_ID_5692 ) || \ ++ ( d == DEV_ID_56BE ) || \ ++ ( d == DEV_ID_56BF ) || \ + ( d == DEV_ID_56C0 ) || \ + ( d == DEV_ID_4F80 ) || \ + ( d == DEV_ID_4F81 ) || \ +-- +2.34.1 + diff --git a/recipes-graphics/gmmlib/files/0001-Introduce-ARL-H-support-172.patch b/recipes-graphics/gmmlib/files/0001-Introduce-ARL-H-support-172.patch new file mode 100644 index 00000000..ab4cb178 --- /dev/null +++ b/recipes-graphics/gmmlib/files/0001-Introduce-ARL-H-support-172.patch @@ -0,0 +1,42 @@ +From 2b267b72d6aaa922861b120da9c1543161797046 Mon Sep 17 00:00:00 2001 +From: John Machado <john.machado@intel.com> +Date: Fri, 9 Feb 2024 13:10:18 +0530 +Subject: [PATCH] Introduce ARL-H support (#172) + +Upstream-Status: Backport [https://github.com/intel/gmmlib/commit/544a8be6f66631e085c86c453aafd78b0e7f5179] +Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com> +--- + Source/inc/common/igfxfmid.h | 10 ++++++++++ + 1 file changed, 10 insertions(+) + +diff --git a/Source/inc/common/igfxfmid.h b/Source/inc/common/igfxfmid.h +index 99fc624..19aa505 100644 +--- a/Source/inc/common/igfxfmid.h ++++ b/Source/inc/common/igfxfmid.h +@@ -1966,6 +1966,11 @@ typedef enum __NATIVEGTTYPE + // ARL-S + #define DEV_ID_7D67 0x7D67 + ++// ARL-H ++#define DEV_ID_7D41 0x7D41 ++#define DEV_ID_7D51 0x7D51 ++#define DEV_ID_7DD1 0x7DD1 ++ + #define MGM_HAS 0 + + //#define SDG_HAS 1 //Reserve place for Springdale-G HAS +@@ -2012,6 +2017,11 @@ typedef enum __NATIVEGTTYPE + // Macro to identify ARL-S Device ID + #define GFX_IS_ARL_S(d) ( ( d == DEV_ID_7D67 ) ) + ++// Macro to identify ARL-H Device ID ++#define GFX_IS_ARL_H(d) ( ( d == DEV_ID_7D41 ) || \ ++ ( d == DEV_ID_7D51 ) || \ ++ ( d == DEV_ID_7DD1 )) ++ + //we define the highest cap and lower cap of stepping IDs + #define SI_REV_ID(lo,hi) (lo | hi<<16) + +-- +2.40.1 + diff --git a/recipes-graphics/gmmlib/gmmlib_21.1.1.bb b/recipes-graphics/gmmlib/gmmlib_22.3.15.bb index f5de0ef5..875fbaf5 100644 --- a/recipes-graphics/gmmlib/gmmlib_21.1.1.bb +++ b/recipes-graphics/gmmlib/gmmlib_22.3.15.bb @@ -4,17 +4,19 @@ device specific and buffer management for the Intel(R) Graphics \ Compute Runtime for OpenCL(TM) and the Intel(R) Media Driver for VAAPI." LICENSE = "MIT" -LIC_FILES_CHKSUM = "file://LICENSE.md;md5=d9a6e772cd4c362ee4c8ef87c5aad843" +LIC_FILES_CHKSUM = "file://LICENSE.md;md5=465fe90caea3edd6a2cecb3f0c28a654" SRC_URI = " \ - git://github.com/intel/gmmlib.git;protocol=https \ + git://github.com/intel/gmmlib.git;protocol=https;branch=master \ + file://0001-Introduce-ARL-H-support-172.patch \ + file://0001-Add-new-DG2-device-IDs-194.patch \ " -SRCREV = "60a77187c43ff23a0888b2f297ae5c93cc50a666" +SRCREV = "5fb4180e22695ad4e3c155cf680119913c9f44bc" S = "${WORKDIR}/git" -COMPATIBLE_HOST_x86-x32 = "null" +COMPATIBLE_HOST:x86-x32 = "null" UPSTREAM_CHECK_GITTAGREGEX = "^intel-gmmlib-(?P<pver>(\d+(\.\d+)+))$" diff --git a/recipes-graphics/hdcp/hdcp/0001-main-fix-hdcpd-service-failure.patch b/recipes-graphics/hdcp/hdcp/0001-main-fix-hdcpd-service-failure.patch index f07d466f..6d32da43 100644 --- a/recipes-graphics/hdcp/hdcp/0001-main-fix-hdcpd-service-failure.patch +++ b/recipes-graphics/hdcp/hdcp/0001-main-fix-hdcpd-service-failure.patch @@ -9,7 +9,7 @@ exits, which causing hdcpd service failure. Added wait function to block parent process until child process exits. -Upstream-Status: Submitted[https://github.com/intel/hdcp/pull/17] +Upstream-Status: Submitted [https://github.com/intel/hdcp/pull/17] Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> --- diff --git a/recipes-graphics/hdcp/hdcp_20.4.0.bb b/recipes-graphics/hdcp/hdcp_21.2.0.bb index 016e8e0a..8529b5ab 100644 --- a/recipes-graphics/hdcp/hdcp_20.4.0.bb +++ b/recipes-graphics/hdcp/hdcp_21.2.0.bb @@ -9,7 +9,7 @@ SECTION = "lib" inherit pkgconfig cmake useradd systemd features_check -SRC_URI = "git://github.com/intel/hdcp.git \ +SRC_URI = "git://github.com/intel/hdcp.git;branch=master;protocol=https \ file://0001-main-fix-hdcpd-service-failure.patch \ " SRCREV = "cacc352159c5bdaf1c21846cecdd1eb7b031e1d3" @@ -20,7 +20,7 @@ REQUIRED_DISTRO_FEATURES = "systemd" DEPENDS = "libdrm virtual/mesa systemd" USERADD_PACKAGES = "${PN}" -USERADD_PARAM_${PN} = "--system --no-create-home --shell /bin/false --gid nogroup media" +USERADD_PARAM:${PN} = "--system --no-create-home --shell /bin/false --gid nogroup media" -SYSTEMD_SERVICE_${PN} = "hdcpd.service" +SYSTEMD_SERVICE:${PN} = "hdcpd.service" SYSTEMD_AUTO_ENABLE = "disable" diff --git a/recipes-graphics/libva/libva-intel-utils_2.11.1.bb b/recipes-graphics/libva/libva-intel-utils_2.20.1.bb index f9bea364..9c26f009 100644 --- a/recipes-graphics/libva/libva-intel-utils_2.11.1.bb +++ b/recipes-graphics/libva/libva-intel-utils_2.20.1.bb @@ -14,8 +14,8 @@ SECTION = "x11" LICENSE = "MIT" LIC_FILES_CHKSUM = "file://COPYING;md5=b148fc8adf19dc9aec17cf9cd29a9a5e" -SRC_URI = "git://github.com/intel/libva-utils.git;branch=v2.11-branch" -SRCREV = "959b2f20cf99d4611eab66164af3d7e4a84f674b" +SRC_URI = "git://github.com/intel/libva-utils.git;branch=v2.20-branch;protocol=https" +SRCREV = "2ad888bb463dc9bfb3deb512ec9faf78f1d3bfa8" S = "${WORKDIR}/git" UPSTREAM_CHECK_GITTAGREGEX = "(?P<pver>(\d+(\.\d+)+))" @@ -32,4 +32,4 @@ PACKAGECONFIG[x11] = "-Dx11=true, -Dx11=false,virtual/libx11 libxext libxfixes" PACKAGECONFIG[wayland] = "-Dwayland=true, -Dwayland=false,wayland-native wayland" PROVIDES = "libva-utils" -RPROVIDES_${PN} += "libva-utils" +RPROVIDES:${PN} += "libva-utils" diff --git a/recipes-graphics/libva/libva-intel_2.11.0.bb b/recipes-graphics/libva/libva-intel_2.20.0.bb index 406f6f68..c81ed65c 100644 --- a/recipes-graphics/libva/libva-intel_2.11.0.bb +++ b/recipes-graphics/libva/libva-intel_2.20.0.bb @@ -17,8 +17,9 @@ SECTION = "x11" LICENSE = "MIT" LIC_FILES_CHKSUM = "file://COPYING;md5=2e48940f94acb0af582e5ef03537800f" -SRC_URI = "https://github.com/intel/libva/releases/download/${PV}/libva-${PV}.tar.bz2" -SRC_URI[sha256sum] = "6e361117038b571ad4741d38c9280db8c140b17e76e8c01fc7a4d608d3ed7d5d" +SRC_URI = "https://github.com/intel/libva/releases/download/${PV}/libva-${PV}.tar.bz2 \ + " +SRC_URI[sha256sum] = "f72bdb4f48dfe71ad01f1cbefe069672a2c949a6abd51cf3c4d4784210badc49" S = "${WORKDIR}/libva-${PV}" @@ -40,15 +41,15 @@ PACKAGECONFIG[wayland] = "-Dwith_wayland=yes,-Dwith_wayland=no,wayland-native wa PACKAGES =+ "${PN}-x11 ${PN}-glx ${PN}-wayland" -RDEPENDS_${PN}-x11 =+ "${PN}" -RDEPENDS_${PN}-glx =+ "${PN}-x11" +RDEPENDS:${PN}-x11 =+ "${PN}" +RDEPENDS:${PN}-glx =+ "${PN}-x11" -FILES_${PN}-x11 =+ "${libdir}/libva-x11*${SOLIBS}" -FILES_${PN}-glx =+ "${libdir}/libva-glx*${SOLIBS}" -FILES_${PN}-wayland =+ "${libdir}/libva-wayland*${SOLIBS}" +FILES:${PN}-x11 =+ "${libdir}/libva-x11*${SOLIBS}" +FILES:${PN}-glx =+ "${libdir}/libva-glx*${SOLIBS}" +FILES:${PN}-wayland =+ "${libdir}/libva-wayland*${SOLIBS}" PROVIDES += "libva" -RPROVIDES_${PN} += "libva" -RPROVIDES_${PN}-x11 += "libva-x11" -RPROVIDES_${PN}-glx += "libva-glx" -RPROVIDES_${PN}-wayland += "libva-wayland" +RPROVIDES:${PN} += "libva" +RPROVIDES:${PN}-x11 += "libva-x11" +RPROVIDES:${PN}-glx += "libva-glx" +RPROVIDES:${PN}-wayland += "libva-wayland" diff --git a/recipes-graphics/metrics-discovery/metrics-discovery/0001-md_internal.h-Replace-string.h-with-string-C-header-.patch b/recipes-graphics/metrics-discovery/metrics-discovery/0001-md_internal.h-Replace-string.h-with-string-C-header-.patch deleted file mode 100644 index d97c6543..00000000 --- a/recipes-graphics/metrics-discovery/metrics-discovery/0001-md_internal.h-Replace-string.h-with-string-C-header-.patch +++ /dev/null @@ -1,34 +0,0 @@ -From 8c11096a0e5cf980309692ae3e272108da5f2b4e Mon Sep 17 00:00:00 2001 -From: Khem Raj <raj.khem@gmail.com> -Date: Wed, 9 Oct 2019 11:36:56 -0700 -Subject: [PATCH] md_internal.h: Replace string.h with <string> C++ header API - -This helps in compiling with libc++/Clang -Fixes -usr/include/c++/v1/type_traits:1561:38: error: implicit instantiation of undefined template 'std::__1::hash<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > >' -: public integral_constant<bool, __is_empty(_Tp)> {}; -^ - -Upstream-Status: Submitted [https://github.com/intel/metrics-discovery/pull/14] -Signed-off-by: Khem Raj <raj.khem@gmail.com> ---- - instrumentation/metrics_discovery/common/md_internal.cpp | 2 +- - 1 file changed, 1 insertion(+), 1 deletion(-) - -diff --git a/instrumentation/metrics_discovery/common/md_internal.cpp b/instrumentation/metrics_discovery/common/md_internal.cpp -index f9571ac..0767e81 100644 ---- a/instrumentation/metrics_discovery/common/md_internal.cpp -+++ b/instrumentation/metrics_discovery/common/md_internal.cpp -@@ -27,8 +27,8 @@ - \*****************************************************************************/ - #include "md_internal.h" - --#include <string.h> - #include <stdlib.h> -+#include <string> - #include <new> - #include <unordered_map> - --- -2.23.0 - diff --git a/recipes-graphics/metrics-discovery/metrics-discovery_1.5.108.bb b/recipes-graphics/metrics-discovery/metrics-discovery_1.12.170.bb index da7cb652..3ed1c9bd 100644 --- a/recipes-graphics/metrics-discovery/metrics-discovery_1.5.108.bb +++ b/recipes-graphics/metrics-discovery/metrics-discovery_1.12.170.bb @@ -8,9 +8,8 @@ SECTION = "lib" inherit pkgconfig cmake S = "${WORKDIR}/git" -SRCREV = "a495db4682b7318bc82b1cccfb17fafdf2e3a2ff" -SRC_URI = "git://github.com/intel/metrics-discovery.git \ - file://0001-md_internal.h-Replace-string.h-with-string-C-header-.patch \ +SRCREV = "75b292eb1309e4ea10dce403065ee8ca5bf29ecc" +SRC_URI = "git://github.com/intel/metrics-discovery.git;branch=master;protocol=https \ " EXTRA_OECMAKE += "-DMD_PLATFORM=linux" diff --git a/recipes-graphics/xorg-driver/xf86-video-ast_1.1.5.bb b/recipes-graphics/xorg-driver/xf86-video-ast_1.1.6.bb index 090efd62..d7e4eab4 100644 --- a/recipes-graphics/xorg-driver/xf86-video-ast_1.1.5.bb +++ b/recipes-graphics/xorg-driver/xf86-video-ast_1.1.6.bb @@ -7,6 +7,6 @@ DESCRIPTION = "ast is an Xorg driver for ASpeed Technologies video cards" LIC_FILES_CHKSUM = "file://COPYING;md5=0b8c242f0218eea5caa949b7910a774b" DEPENDS += "libpciaccess" +XORG_DRIVER_COMPRESSOR = ".tar.gz" -SRC_URI[md5sum] = "4f85febe48d51e53624550a96fc9e9ee" -SRC_URI[sha256sum] = "1edbbc55d47d3fd71dec99b15c2483e22738c642623a0fb86ef4a81a9067a2de" +SRC_URI[sha256sum] = "eeff52178ce2916a0e07b531bb23c8b105826b6e6e6c5f7c90e33f757827a7c7" diff --git a/recipes-kernel/intel-ethernet/ixgbe_5.11.3.bb b/recipes-kernel/intel-ethernet/ixgbe_5.19.6.bb index 73b13c9b..7ef38650 100644 --- a/recipes-kernel/intel-ethernet/ixgbe_5.11.3.bb +++ b/recipes-kernel/intel-ethernet/ixgbe_5.19.6.bb @@ -4,13 +4,13 @@ PCI Express* 10 Gigabit Network Connections." HOMEPAGE = "https://sourceforge.net/projects/e1000/" SECTION = "kernel/network" -LICENSE = "GPLv2" +LICENSE = "GPL-2.0-only" LIC_FILES_CHKSUM = "file://${WORKDIR}/${BP}/COPYING;md5=a216b4192dc6b777b6f0db560e9a8417" SRC_URI = "https://sourceforge.net/projects/e1000/files/ixgbe%20stable/${PV}/${BP}.tar.gz \ " -SRC_URI[sha256sum] = "b7e5de734c406d28ea146366a5cc3b58bb3b6be039977863e4e1b217a8185ae2" +SRC_URI[sha256sum] = "a844f1fea8064e30b276792455c3b286c1d7af26731e8f865d4a4e9ed1dcf4ab" UPSTREAM_CHECK_URI = "https://sourceforge.net/projects/e1000/files/ixgbe%20stable/" UPSTREAM_CHECK_REGEX = "ixgbe%20stable/(?P<pver>\d+(\.\d+)+)/" @@ -19,14 +19,14 @@ CVE_PRODUCT = "linux:linux_kernel_ixgbe" S = "${WORKDIR}/${BP}/src" -EXTRA_OEMAKE='KSRC="${STAGING_KERNEL_BUILDDIR}" KVER="${KERNEL_VERSION}" INSTALL_MOD_PATH="${D}"' +EXTRA_OEMAKE=' KSRC="${STAGING_KERNEL_DIR}" KOBJ="${STAGING_KERNEL_BUILDDIR}" KVER="${KERNEL_VERSION}" INSTALL_MOD_PATH="${D}"' -KERNEL_MODULE_AUTOLOAD_append_intel-core2-32 = " ixgbe" -KERNEL_MODULE_AUTOLOAD_append_intel-corei7-64 = " ixgbe" +KERNEL_MODULE_AUTOLOAD:append:intel-core2-32 = " ixgbe" +KERNEL_MODULE_AUTOLOAD:append:intel-corei7-64 = " ixgbe" inherit module -do_install_append () { +do_install:append () { # Install scripts/set_irq_affinity install -d ${D}${sysconfdir}/network install -m 0755 ${S}/../scripts/set_irq_affinity ${D}${sysconfdir}/network @@ -36,6 +36,8 @@ do_install_append () { PACKAGES += "${PN}-script" -FILES_${PN}-script += "${sysconfdir}/network/set_irq_affinity" +FILES:${PN}-script += "${sysconfdir}/network/set_irq_affinity" EXCLUDE_FROM_WORLD = "1" + +CVE_STATUS[CVE-2015-1142857] = "fixed-version: Fixed from version 4.4-rc1" diff --git a/recipes-kernel/intel-ethernet/ixgbevf_4.11.1.bb b/recipes-kernel/intel-ethernet/ixgbevf_4.18.7.bb index d1c5f299..b21796c6 100644 --- a/recipes-kernel/intel-ethernet/ixgbevf_4.11.1.bb +++ b/recipes-kernel/intel-ethernet/ixgbevf_4.18.7.bb @@ -7,13 +7,13 @@ The guest OS loading this driver must support MSI-X interrupts." HOMEPAGE = "https://sourceforge.net/projects/e1000/" SECTION = "kernel/network" -LICENSE = "GPLv2" +LICENSE = "GPL-2.0-only" LIC_FILES_CHKSUM = "file://${WORKDIR}/${BP}/COPYING;md5=a216b4192dc6b777b6f0db560e9a8417" SRC_URI = "https://sourceforge.net/projects/e1000/files/ixgbevf%20stable/${PV}/${BP}.tar.gz \ " -SRC_URI[sha256sum] = "0031d09a54af9ecf1216c185e4641e38527643afa09e78d1bd95752e49fe8985" +SRC_URI[sha256sum] = "90f6cd614008839b6fc748ae0f4ad3503435f8b788318d4f40cfc83c7029025e" UPSTREAM_CHECK_URI = "https://sourceforge.net/projects/e1000/files/ixgbevf%20stable/" UPSTREAM_CHECK_REGEX = "ixgbevf%20stable/(?P<pver>\d+(\.\d+)+)/" @@ -22,14 +22,14 @@ CVE_PRODUCT = "linux:linux_kernel_ixgbe" S = "${WORKDIR}/${BP}/src" -EXTRA_OEMAKE='KSRC="${STAGING_KERNEL_BUILDDIR}" KVER="${KERNEL_VERSION}" INSTALL_MOD_PATH="${D}"' +EXTRA_OEMAKE='KSRC="${STAGING_KERNEL_DIR}" KOBJ="${STAGING_KERNEL_BUILDDIR}" KVER="${KERNEL_VERSION}" INSTALL_MOD_PATH="${D}"' -KERNEL_MODULE_AUTOLOAD_append_intel-core2-32 = " ixgbevf" -KERNEL_MODULE_AUTOLOAD_append_intel-corei7-64 = " ixgbevf" +KERNEL_MODULE_AUTOLOAD:append:intel-core2-32 = " ixgbevf" +KERNEL_MODULE_AUTOLOAD:append:intel-corei7-64 = " ixgbevf" inherit module -do_install_append () { +do_install:append () { # Install scripts/set_irq_affinity install -d ${D}${sysconfdir}/network install -m 0755 ${S}/../scripts/set_irq_affinity ${D}${sysconfdir}/network @@ -39,4 +39,8 @@ do_install_append () { PACKAGES += "${PN}-script" -FILES_${PN}-script += "${sysconfdir}/network/set_irq_affinity" +FILES:${PN}-script += "${sysconfdir}/network/set_irq_affinity" + +CVE_STATUS[CVE-2015-1142857] = "fixed-version: Fixed from version 4.4-rc1" + +EXCLUDE_FROM_WORLD = "1" diff --git a/recipes-kernel/iwlwifi/backport-iwlwifi_git.bb b/recipes-kernel/iwlwifi/backport-iwlwifi_git.bb index 9eb0a45d..ea36cfc3 100644 --- a/recipes-kernel/iwlwifi/backport-iwlwifi_git.bb +++ b/recipes-kernel/iwlwifi/backport-iwlwifi_git.bb @@ -1,13 +1,13 @@ SUMMARY = "Intel Wireless LinuxCore kernel driver" DESCRIPTION = "Intel Wireless LinuxCore kernel driver" SECTION = "kernel" -LICENSE = "GPL-2.0" +LICENSE = "GPL-2.0-only" REQUIRED_DISTRO_FEATURES = "wifi" -LIC_FILES_CHKSUM = "file://${S}/COPYING;md5=bbea815ee2795b2f4230826c0c6b8814" +LIC_FILES_CHKSUM = "file://${S}/COPYING;md5=6bc538ed5bd9a7fc9398086aedcd7e46" -inherit module +inherit module features_check # For some iwfwifi LinuxCore supported wireless chips, the best/latest # firmware blobs are found in the iwlwifi's linux-firmware.git fork. @@ -17,8 +17,8 @@ inherit module # When updating this recipe, ensure that the proper firmware is included from # either the linux-firmware or iwlwifi-firmware repos. -PV = "45" -SRCREV = "a75c1de6b3fa87885556c67619429cfa87cc048f" +PV = "79" +SRCREV = "574631d89d736fd2c76b0e2ea489270c50903e52" SRC_URI = " \ git://git.kernel.org/pub/scm/linux/kernel/git/iwlwifi/backport-iwlwifi;branch=release/core${PV} \ @@ -36,16 +36,18 @@ do_configure() { MODULES_INSTALL_TARGET="install" -do_install_append() { +do_install:append() { ## install configs and service scripts install -d ${D}${sysconfdir}/modprobe.d install -m 0644 ${WORKDIR}/iwlwifi.conf ${D}${sysconfdir}/modprobe.d } -RDEPENDS_${PN} = "linux-firmware-iwlwifi" +RDEPENDS:${PN} = "linux-firmware-iwlwifi" -KERNEL_MODULE_AUTOLOAD_append_core2-32-intel-common = " iwlwifi" -KERNEL_MODULE_AUTOLOAD_append_corei7-64-intel-common = " iwlwifi" +FILES:${PN} += "${sysconfdir}/modprobe.d/iwlwifi.conf" + +KERNEL_MODULE_AUTOLOAD:append:core2-32-intel-common = " iwlwifi" +KERNEL_MODULE_AUTOLOAD:append:corei7-64-intel-common = " iwlwifi" KERNEL_MODULE_PACKAGE_PREFIX = "backport-iwlwifi" diff --git a/recipes-kernel/linux/linux-intel-dev.bb b/recipes-kernel/linux/linux-intel-dev.bb deleted file mode 100644 index f99b1f19..00000000 --- a/recipes-kernel/linux/linux-intel-dev.bb +++ /dev/null @@ -1,34 +0,0 @@ -require recipes-kernel/linux/linux-yocto.inc -require recipes-kernel/linux/meta-intel-compat-kernel.inc - -FILESEXTRAPATHS_prepend := "${THISDIR}/linux-intel:" - -SRC_URI = " \ - git://github.com/intel/mainline-tracking.git;protocol=https;name=machine;nobranch=1; \ - git://git.yoctoproject.org/yocto-kernel-cache;type=kmeta;name=meta;branch=${KMETA_BRANCH};destsuffix=${KMETA} \ - file://0001-menuconfig-mconf-cfg-Allow-specification-of-ncurses-.patch \ - " -SRC_URI_append_core2-32-intel-common = " file://disable_skylake_sound.cfg" - -KMETA = "kernel-meta" -KCONF_BSP_AUDIT_LEVEL = "2" - -KMETA_BRANCH = "master" - -LIC_FILES_CHKSUM = "file://COPYING;md5=6bc538ed5bd9a7fc9398086aedcd7e46" - -DEPENDS += "elfutils-native openssl-native util-linux-native" - -LINUX_VERSION ?= "5.11-rc6" -SRCREV_machine ?= "1048ba83fb1c00cd24172e23e8263972f6b5d9ac" -SRCREV_meta ?= "3356fb62fb5eadec1f5c3c4248f1f8dbab968aa1" - -LINUX_VERSION_EXTENSION ?= "-mainline-tracking-${LINUX_KERNEL_TYPE}" -PV = "${LINUX_VERSION}+git${SRCPV}" - -COMPATIBLE_MACHINE ?= "(intel-corei7-64|intel-core2-32)" - -# Functionality flags -KERNEL_FEATURES_append = " ${KERNEL_EXTRA_FEATURES}" -KERNEL_FEATURES_append = " ${@bb.utils.contains("TUNE_FEATURES", "mx32", " cfg/x32.scc", "" ,d)}" -KERNEL_EXTRA_FEATURES ?= "features/netfilter/netfilter.scc features/security/security.scc" diff --git a/recipes-kernel/linux/linux-intel-rt_5.4.bb b/recipes-kernel/linux/linux-intel-rt_5.4.bb deleted file mode 100644 index 6b61e168..00000000 --- a/recipes-kernel/linux/linux-intel-rt_5.4.bb +++ /dev/null @@ -1,30 +0,0 @@ -require linux-intel.inc - -# Skip processing of this recipe if it is not explicitly specified as the -# PREFERRED_PROVIDER for virtual/kernel. This avoids errors when trying -# to build multiple virtual/kernel providers, e.g. as dependency of -# core-image-rt-sdk, core-image-rt. -python () { - if d.getVar("KERNEL_PACKAGE_NAME", True) == "kernel" and d.getVar("PREFERRED_PROVIDER_virtual/kernel") != "linux-intel-rt": - raise bb.parse.SkipPackage("Set PREFERRED_PROVIDER_virtual/kernel to linux-intel-rt to enable it") -} - -LIC_FILES_CHKSUM = "file://COPYING;md5=bbea815ee2795b2f4230826c0c6b8814" -SRC_URI_append = " file://0001-menuconfig-mconf-cfg-Allow-specification-of-ncurses-.patch \ -" - -KBRANCH = "5.4/preempt-rt" -KMETA_BRANCH = "yocto-5.4" - -DEPENDS += "elfutils-native openssl-native util-linux-native" - -LINUX_VERSION ?= "5.4.102" -SRCREV_machine ?= "7bbaa89ee7f0cb8e912e52b17247d0c0a7b2db15" -SRCREV_meta ?= "19738ca97b999a3b150e2d34232bb44b6537348f" - -LINUX_KERNEL_TYPE = "preempt-rt" - -# Kernel config 'CONFIG_GPIO_LYNXPOINT' goes by a different name 'CONFIG_PINCTRL_LYNXPOINT' in -# linux-intel 5.4 specifically. This cause annoying warning during kernel config audit. So suppress the -# harmless warning for now. -KCONF_BSP_AUDIT_LEVEL = "0" diff --git a/recipes-kernel/linux/linux-intel-rt_5.10.bb b/recipes-kernel/linux/linux-intel-rt_6.6.bb index 7d67fb0a..23e79a43 100644 --- a/recipes-kernel/linux/linux-intel-rt_5.10.bb +++ b/recipes-kernel/linux/linux-intel-rt_6.6.bb @@ -1,5 +1,8 @@ require linux-intel.inc +SRC_URI:prepend = "git://github.com/intel/linux-intel-lts.git;protocol=https;name=machine;branch=${KBRANCH}; \ + " + # Skip processing of this recipe if it is not explicitly specified as the # PREFERRED_PROVIDER for virtual/kernel. This avoids errors when trying # to build multiple virtual/kernel providers, e.g. as dependency of @@ -9,17 +12,22 @@ python () { raise bb.parse.SkipPackage("Set PREFERRED_PROVIDER_virtual/kernel to linux-intel-rt to enable it") } -LIC_FILES_CHKSUM = "file://COPYING;md5=6bc538ed5bd9a7fc9398086aedcd7e46" -SRC_URI_append = " file://0001-menuconfig-mconf-cfg-Allow-specification-of-ncurses-.patch \ - " +KBRANCH = "6.6/preempt-rt" +KMETA_BRANCH = "yocto-6.6" -KBRANCH = "5.10/preempt-rt" -KMETA_BRANCH = "yocto-5.10" +LIC_FILES_CHKSUM = "file://COPYING;md5=6bc538ed5bd9a7fc9398086aedcd7e46" DEPENDS += "elfutils-native openssl-native util-linux-native" -LINUX_VERSION ?= "5.10.21" -SRCREV_machine ?= "774cd893ee3f4044305e0a7bf5090e32b72578b3" -SRCREV_meta ?= "eb09284047fec2f09d62068c338ae320c6681bd1" +LINUX_VERSION_EXTENSION ??= "-intel-pk-${LINUX_KERNEL_TYPE}" + +LINUX_VERSION ?= "6.6.23" +SRCREV_machine ?= "73113409624a0a68494db7b1c1d535d1c8b96689" +SRCREV_meta ?= "eb283ea577df80542d48f0c498365960b4c4ecd9" LINUX_KERNEL_TYPE = "preempt-rt" + +# Functionality flags +KERNEL_EXTRA_FEATURES ?= "features/netfilter/netfilter.scc features/security/security.scc" + +UPSTREAM_CHECK_GITTAGREGEX = "^lts-(?P<pver>v6.6.(\d+)-rt(\d)-preempt-rt-(\d+)T(\d+)Z)$" diff --git a/recipes-kernel/linux/linux-intel.inc b/recipes-kernel/linux/linux-intel.inc index c6163973..f78f5205 100644 --- a/recipes-kernel/linux/linux-intel.inc +++ b/recipes-kernel/linux/linux-intel.inc @@ -1,26 +1,24 @@ require recipes-kernel/linux/linux-yocto.inc require recipes-kernel/linux/meta-intel-compat-kernel.inc -FILESEXTRAPATHS_prepend := "${THISDIR}/linux-intel:" +FILESEXTRAPATHS:prepend := "${THISDIR}/linux-intel:" KERNEL_CONFIG_URI ?= "git://git.yoctoproject.org/yocto-kernel-cache;type=kmeta;name=meta;branch=${KMETA_BRANCH};destsuffix=${KMETA}" SRC_URI = " \ - git://github.com/intel/linux-intel-lts.git;protocol=https;name=machine;branch=${KBRANCH}; \ ${KERNEL_CONFIG_URI} \ + file://0001-vt-conmakehash-improve-reproducibility.patch \ + file://0001-lib-build_OID_registry-fix-reproducibility-issues.patch \ + file://fix-perf-reproducibility.patch \ + file://0001-menuconfig-mconf-cfg-Allow-specification-of-ncurses-.patch \ + file://0002-mconf-fix-output-of-cflags-and-libraries.patch \ " -SRC_URI_append = " file://enable_lynxpoint_gpio.cfg" -SRC_URI_append_core2-32-intel-common = " file://disable_skylake_sound.cfg" - -LINUX_VERSION_EXTENSION ??= "-intel-pk-${LINUX_KERNEL_TYPE}" - PV = "${LINUX_VERSION}+git${SRCPV}" KMETA = "kernel-meta" -KCONF_BSP_AUDIT_LEVEL = "2" +KCONF_BSP_AUDIT_LEVEL = "0" -COMPATIBLE_MACHINE ?= "(intel-corei7-64|intel-core2-32)" +COMPATIBLE_MACHINE ?= "(intel-corei7-64)" -KERNEL_FEATURES_append = " ${KERNEL_EXTRA_FEATURES}" -KERNEL_FEATURES_append = " ${@bb.utils.contains("TUNE_FEATURES", "mx32", " cfg/x32.scc", "" ,d)}" +KERNEL_FEATURES:append = " ${KERNEL_EXTRA_FEATURES}" diff --git a/recipes-kernel/linux/linux-intel/0001-lib-build_OID_registry-fix-reproducibility-issues.patch b/recipes-kernel/linux/linux-intel/0001-lib-build_OID_registry-fix-reproducibility-issues.patch new file mode 100644 index 00000000..d41c3f0b --- /dev/null +++ b/recipes-kernel/linux/linux-intel/0001-lib-build_OID_registry-fix-reproducibility-issues.patch @@ -0,0 +1,48 @@ +From 2fca0fd719812ea2ff67630b01355aa80481623e Mon Sep 17 00:00:00 2001 +From: Bruce Ashfield <bruce.ashfield@gmail.com> +Date: Sun, 10 Jul 2022 22:56:53 -0400 +Subject: [PATCH] lib/build_OID_registry: fix reproducibility issues + +The script build_OID_registry captures the full path of itself +in the generated data. This causes reproduciblity issues as the +path is captured and packaged. + +We use the basename of the script instead, and that allows us +to be reprodicible, with slightly less information captured in +the output data (but the generating script can still easily +be found). + +Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com> +Upstream-Status: Inappropriate + +Taken from linux-yocto, v5.15/standard/base. + +Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> +--- + lib/build_OID_registry | 3 ++- + 1 file changed, 2 insertions(+), 1 deletion(-) + +diff --git a/lib/build_OID_registry b/lib/build_OID_registry +index d7fc32ea8ac2..f6de0a7f7457 100755 +--- a/lib/build_OID_registry ++++ b/lib/build_OID_registry +@@ -8,6 +8,7 @@ + # + + use strict; ++use File::Basename; + + my @names = (); + my @oids = (); +@@ -35,7 +36,7 @@ close IN_FILE || die; + # + open C_FILE, ">$ARGV[1]" or die; + print C_FILE "/*\n"; +-print C_FILE " * Automatically generated by ", $0, ". Do not edit\n"; ++print C_FILE " * Automatically generated by ", basename $0, ". Do not edit\n"; + print C_FILE " */\n"; + + # +-- +2.36.1 + diff --git a/recipes-kernel/linux/linux-intel/0001-menuconfig-mconf-cfg-Allow-specification-of-ncurses-.patch b/recipes-kernel/linux/linux-intel/0001-menuconfig-mconf-cfg-Allow-specification-of-ncurses-.patch index 3dc71ff2..a601e783 100644 --- a/recipes-kernel/linux/linux-intel/0001-menuconfig-mconf-cfg-Allow-specification-of-ncurses-.patch +++ b/recipes-kernel/linux/linux-intel/0001-menuconfig-mconf-cfg-Allow-specification-of-ncurses-.patch @@ -1,7 +1,8 @@ -From bebd63730a433ba62549a80114a9851328aa8897 Mon Sep 17 00:00:00 2001 +From 1811da09f42ca5e82282970e8ad014707bc94c82 Mon Sep 17 00:00:00 2001 From: Bruce Ashfield <bruce.ashfield@windriver.com> Date: Mon, 2 Jul 2018 23:10:28 -0400 -Subject: [PATCH] menuconfig,mconf-cfg: Allow specification of ncurses location +Subject: [PATCH 1/2] menuconfig,mconf-cfg: Allow specification of ncurses + location In some cross build environments such as the Yocto Project build environment it provides an ncurses library that is compiled @@ -16,33 +17,36 @@ check-lxdialog.sh for environments such as the Yocto Project. Adding a CROSS_CURSES_LIB and CROSS_CURSES_INC solves the issue and allowing compiling and linking against the right headers and libraries. +Upstream-Status: Submitted + Signed-off-by: Jason Wessel <jason.wessel@windriver.com> cc: Michal Marek <mmarek@suse.cz> cc: linux-kbuild@vger.kernel.org Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com> +Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> --- scripts/kconfig/mconf-cfg.sh | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/scripts/kconfig/mconf-cfg.sh b/scripts/kconfig/mconf-cfg.sh -index c812872d7f9d..42d20819025c 100755 +index 1e61f50a5905..38cf8304bb31 100755 --- a/scripts/kconfig/mconf-cfg.sh +++ b/scripts/kconfig/mconf-cfg.sh -@@ -4,6 +4,14 @@ +@@ -7,6 +7,14 @@ libs=$2 PKG="ncursesw" PKG2="ncurses" +if [ "$CROSS_CURSES_LIB" != "" ]; then -+ echo libs=\'$CROSS_CURSES_LIB\' -+ if [ x"$CROSS_CURSES_INC" != x ]; then -+ echo cflags=\'$CROSS_CURSES_INC\' -+ fi -+ exit 0 ++ echo libs=\'$CROSS_CURSES_LIB\' ++ if [ x"$CROSS_CURSES_INC" != x ]; then ++ echo cflags=\'$CROSS_CURSES_INC\' ++ fi ++ exit 0 +fi + - if [ -n "$(command -v pkg-config)" ]; then - if pkg-config --exists $PKG; then - echo cflags=\"$(pkg-config --cflags $PKG)\" + if [ -n "$(command -v ${HOSTPKG_CONFIG})" ]; then + if ${HOSTPKG_CONFIG} --exists $PKG; then + ${HOSTPKG_CONFIG} --cflags ${PKG} > ${cflags} -- -2.17.1 +2.34.1 diff --git a/recipes-kernel/linux/linux-intel/0001-vt-conmakehash-improve-reproducibility.patch b/recipes-kernel/linux/linux-intel/0001-vt-conmakehash-improve-reproducibility.patch new file mode 100644 index 00000000..33280063 --- /dev/null +++ b/recipes-kernel/linux/linux-intel/0001-vt-conmakehash-improve-reproducibility.patch @@ -0,0 +1,58 @@ +From 0f586f4ee8adacac79b64d1f3d47799a5eb7fbea Mon Sep 17 00:00:00 2001 +From: Bruce Ashfield <bruce.ashfield@gmail.com> +Date: Sun, 10 Jul 2022 21:37:07 -0400 +Subject: [PATCH] vt/conmakehash: improve reproducibility + +The file generated by conmakehash capture the application +path used to generate the file. While that can be informative, +it varies based on where the kernel was built, as the full +path is captured. + +We tweak the application to use a second input as the "capture +name", and then modify the Makefile to pass the basename of +the source, making it reproducible. + +This could be improved by using some sort of path mapping, +or the application manipualing argv[1] itself, but for now +this solves the reprodicibility issue. + +Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com> + +Upstream-Status: Inappropriate + +Taken from linux-yocto, v5.15/standard/base +Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> +--- + drivers/tty/vt/Makefile | 2 +- + drivers/tty/vt/conmakehash.c | 2 +- + 2 files changed, 2 insertions(+), 2 deletions(-) + +diff --git a/drivers/tty/vt/Makefile b/drivers/tty/vt/Makefile +index fe30ce512819..cb51c21b58f9 100644 +--- a/drivers/tty/vt/Makefile ++++ b/drivers/tty/vt/Makefile +@@ -15,7 +15,7 @@ clean-files := consolemap_deftbl.c defkeymap.c + hostprogs += conmakehash + + quiet_cmd_conmk = CONMK $@ +- cmd_conmk = $(obj)/conmakehash $< > $@ ++ cmd_conmk = $(obj)/conmakehash $< $(shell basename $<) > $@ + + $(obj)/consolemap_deftbl.c: $(src)/$(FONTMAPFILE) $(obj)/conmakehash + $(call cmd,conmk) +diff --git a/drivers/tty/vt/conmakehash.c b/drivers/tty/vt/conmakehash.c +index cddd789fe46e..d62510b280e9 100644 +--- a/drivers/tty/vt/conmakehash.c ++++ b/drivers/tty/vt/conmakehash.c +@@ -253,7 +253,7 @@ int main(int argc, char *argv[]) + #include <linux/types.h>\n\ + \n\ + u8 dfont_unicount[%d] = \n\ +-{\n\t", argv[1], fontlen); ++{\n\t", argv[2], fontlen); + + for ( i = 0 ; i < fontlen ; i++ ) + { +-- +2.36.1 + diff --git a/recipes-kernel/linux/linux-intel/0002-mconf-fix-output-of-cflags-and-libraries.patch b/recipes-kernel/linux/linux-intel/0002-mconf-fix-output-of-cflags-and-libraries.patch new file mode 100644 index 00000000..a96b68d9 --- /dev/null +++ b/recipes-kernel/linux/linux-intel/0002-mconf-fix-output-of-cflags-and-libraries.patch @@ -0,0 +1,40 @@ +From 1b53d82a8152843afcddd7f16b0c43b6b4f22895 Mon Sep 17 00:00:00 2001 +From: Bruce Ashfield <bruce.ashfield@gmail.com> +Date: Mon, 17 Jul 2023 17:17:55 -0400 +Subject: [PATCH 2/2] mconf: fix output of cflags and libraries + +commit 3122c84409d578a5df8bcb1 [kconfig: refactor Makefile to reduce +process forks] changes the way that flags are detected. They are +no longer just echo'd and captured, they are written to a file and +later read. + +We adjust our CROSS ncurses patch accordingly. + +We'll eventually be able to drop this patch, but not quite yet. + +Upstream-Status: Inappropriate [OE-Specific] +Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com> +Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> +--- + scripts/kconfig/mconf-cfg.sh | 4 ++-- + 1 file changed, 2 insertions(+), 2 deletions(-) + +diff --git a/scripts/kconfig/mconf-cfg.sh b/scripts/kconfig/mconf-cfg.sh +index 38cf8304bb31..a5ae56e08b07 100755 +--- a/scripts/kconfig/mconf-cfg.sh ++++ b/scripts/kconfig/mconf-cfg.sh +@@ -8,9 +8,9 @@ PKG="ncursesw" + PKG2="ncurses" + + if [ "$CROSS_CURSES_LIB" != "" ]; then +- echo libs=\'$CROSS_CURSES_LIB\' ++ echo $CROSS_CURSES_LIB > ${libs} + if [ x"$CROSS_CURSES_INC" != x ]; then +- echo cflags=\'$CROSS_CURSES_INC\' ++ echo $CROSS_CURSES_INC > ${cflags} + fi + exit 0 + fi +-- +2.34.1 + diff --git a/recipes-kernel/linux/linux-intel/disable_skylake_sound.cfg b/recipes-kernel/linux/linux-intel/disable_skylake_sound.cfg deleted file mode 100644 index 312bcb1a..00000000 --- a/recipes-kernel/linux/linux-intel/disable_skylake_sound.cfg +++ /dev/null @@ -1,7 +0,0 @@ -# CONFIG_SND_SOC_INTEL_SKYLAKE is not set -# CONFIG_SND_SOC_INTEL_BXT_DA7219_MAX98357A_MACH is not set -# CONFIG_SND_SOC_INTEL_BXT_RT298_MACH is not set -# CONFIG_SND_SOC_INTEL_KBL_RT5663_MAX98927_MACH is not set -# CONFIG_SND_SOC_INTEL_SKL_RT286_MACH is not set -# CONFIG_SND_SOC_INTEL_SKL_NAU88L25_SSM4567_MACH is not set -# CONFIG_SND_SOC_INTEL_SKL_NAU88L25_MAX98357A_MACH is not set diff --git a/recipes-kernel/linux/linux-intel/enable_lynxpoint_gpio.cfg b/recipes-kernel/linux/linux-intel/enable_lynxpoint_gpio.cfg deleted file mode 100644 index 49e70e67..00000000 --- a/recipes-kernel/linux/linux-intel/enable_lynxpoint_gpio.cfg +++ /dev/null @@ -1 +0,0 @@ -CONFIG_PINCTRL_LYNXPOINT=m diff --git a/recipes-kernel/linux/linux-intel/fix-perf-reproducibility.patch b/recipes-kernel/linux/linux-intel/fix-perf-reproducibility.patch new file mode 100644 index 00000000..1a98b1db --- /dev/null +++ b/recipes-kernel/linux/linux-intel/fix-perf-reproducibility.patch @@ -0,0 +1,39 @@ +From b8cd0e429bf75b673c438a8277d4bc74327df992 Mon Sep 17 00:00:00 2001 +From: Tom Zanussi <tom.zanussi@intel.com> +Date: Tue, 3 Jul 2012 13:07:23 -0500 +Subject: perf: change --root to --prefix for python install + +Otherwise we get the sysroot path appended to the build path, not what +we want. + +Signed-off-by: Tom Zanussi <tom.zanussi@intel.com> + +Upstream-Status: Inappropriate + +Taken from linux-yocto, v5.15/standard/base +https://git.yoctoproject.org/linux-yocto/commit?id=b8cd0e429bf75b673c438a8277d4bc74327df992&h=v5.15%2Fstandard%2Fbase + +Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> +--- + tools/perf/Makefile.perf | 2 +- + 1 file changed, 1 insertion(+), 1 deletion(-) + +(limited to 'tools/perf/Makefile.perf') + +diff --git a/tools/perf/Makefile.perf b/tools/perf/Makefile.perf +index 8f738e11356d..ee945d8e3996 100644 +--- a/tools/perf/Makefile.perf ++++ b/tools/perf/Makefile.perf +@@ -1022,7 +1022,7 @@ install-bin: install-tools install-tests install-traceevent-plugins + install: install-bin try-install-man + + install-python_ext: +- $(PYTHON_WORD) util/setup.py --quiet install --root='/$(DESTDIR_SQ)' ++ $(PYTHON_WORD) util/setup.py --quiet install --prefix='$(DESTDIR_SQ)/usr' + + # 'make install-doc' should call 'make -C Documentation install' + $(INSTALL_DOC_TARGETS): +-- +cgit + + diff --git a/recipes-kernel/linux/linux-intel_5.10.bb b/recipes-kernel/linux/linux-intel_5.10.bb deleted file mode 100644 index 399ab014..00000000 --- a/recipes-kernel/linux/linux-intel_5.10.bb +++ /dev/null @@ -1,22 +0,0 @@ -require linux-intel.inc - -KBRANCH = "5.10/yocto" -KMETA_BRANCH = "yocto-5.10" - -LIC_FILES_CHKSUM = "file://COPYING;md5=6bc538ed5bd9a7fc9398086aedcd7e46" - -SRC_URI_append = " file://0001-menuconfig-mconf-cfg-Allow-specification-of-ncurses-.patch \ - " - -DEPENDS += "elfutils-native openssl-native util-linux-native" - -LINUX_VERSION ?= "5.10.25" -SRCREV_machine ?= "7c8f517558810be0c9470f78a06f9c8b909157b4" -SRCREV_meta ?= "eb09284047fec2f09d62068c338ae320c6681bd1" - -# For Crystalforest and Romley -KERNEL_MODULE_AUTOLOAD_append_core2-32-intel-common = " uio" -KERNEL_MODULE_AUTOLOAD_append_corei7-64-intel-common = " uio" - -# Functionality flags -KERNEL_EXTRA_FEATURES ?= "features/netfilter/netfilter.scc features/security/security.scc" diff --git a/recipes-kernel/linux/linux-intel_5.4.bb b/recipes-kernel/linux/linux-intel_5.4.bb deleted file mode 100644 index 5b8a4457..00000000 --- a/recipes-kernel/linux/linux-intel_5.4.bb +++ /dev/null @@ -1,26 +0,0 @@ -require linux-intel.inc - -KBRANCH = "5.4/yocto" -KMETA_BRANCH = "yocto-5.4" - -LIC_FILES_CHKSUM = "file://COPYING;md5=bbea815ee2795b2f4230826c0c6b8814" - -SRC_URI_append = " file://0001-menuconfig-mconf-cfg-Allow-specification-of-ncurses-.patch" - -DEPENDS += "elfutils-native openssl-native util-linux-native" - -LINUX_VERSION ?= "5.4.106" -SRCREV_machine ?= "4dbe3697c9b89500917d5b6ff3a42b7af769de00" -SRCREV_meta ?= "19738ca97b999a3b150e2d34232bb44b6537348f" - -# For Crystalforest and Romley -KERNEL_MODULE_AUTOLOAD_append_core2-32-intel-common = " uio" -KERNEL_MODULE_AUTOLOAD_append_corei7-64-intel-common = " uio" - -# Functionality flags -KERNEL_EXTRA_FEATURES ?= "features/netfilter/netfilter.scc features/security/security.scc" - -# Kernel config 'CONFIG_GPIO_LYNXPOINT' goes by a different name 'CONFIG_PINCTRL_LYNXPOINT' in -# linux-intel 5.4 specifically. This cause warning during kernel config audit. So suppress the -# harmless warning for now. -KCONF_BSP_AUDIT_LEVEL = "0" diff --git a/recipes-kernel/linux/linux-intel_6.6.bb b/recipes-kernel/linux/linux-intel_6.6.bb new file mode 100644 index 00000000..0d914b73 --- /dev/null +++ b/recipes-kernel/linux/linux-intel_6.6.bb @@ -0,0 +1,21 @@ +require linux-intel.inc + +SRC_URI:prepend = "git://github.com/intel/linux-intel-lts.git;protocol=https;name=machine;branch=${KBRANCH}; \ + " +KBRANCH = "6.6/linux" +KMETA_BRANCH = "yocto-6.6" + +LIC_FILES_CHKSUM = "file://COPYING;md5=6bc538ed5bd9a7fc9398086aedcd7e46" + +DEPENDS += "elfutils-native openssl-native util-linux-native" + +LINUX_VERSION_EXTENSION ??= "-intel-pk-${LINUX_KERNEL_TYPE}" + +LINUX_VERSION ?= "6.6.23" +SRCREV_machine ?= "lts-v6.6.23-linux-240407T055600Z" +SRCREV_meta ?= "eb283ea577df80542d48f0c498365960b4c4ecd9" + +# Functionality flags +KERNEL_EXTRA_FEATURES ?= "features/netfilter/netfilter.scc features/security/security.scc" + +UPSTREAM_CHECK_GITTAGREGEX = "^lts-(?P<pver>v6.6.(\d+)-linux-(\d+)T(\d+)Z)$" diff --git a/recipes-kernel/linux/linux-intel_6.8.bb b/recipes-kernel/linux/linux-intel_6.8.bb new file mode 100644 index 00000000..036879db --- /dev/null +++ b/recipes-kernel/linux/linux-intel_6.8.bb @@ -0,0 +1,20 @@ +require linux-intel.inc + +SRC_URI:prepend = "git://github.com/intel/mainline-tracking.git;protocol=https;name=machine;nobranch=1; \ + " +KMETA_BRANCH = "master" + +LIC_FILES_CHKSUM = "file://COPYING;md5=6bc538ed5bd9a7fc9398086aedcd7e46" + +DEPENDS += "elfutils-native openssl-native util-linux-native" + +LINUX_VERSION_EXTENSION ??= "-mainline-tracking-${LINUX_KERNEL_TYPE}" + +LINUX_VERSION ?= "6.8" +SRCREV_machine ?= "efbae83db36abbbbdb946d4f7bbdfda174107cd2" +SRCREV_meta ?= "27907f391a4fc508da21358b13419c6e86926c34" + +# Functionality flags +KERNEL_EXTRA_FEATURES ?= "features/netfilter/netfilter.scc features/security/security.scc" + +UPSTREAM_CHECK_GITTAGREGEX = "^mainline-tracking-v6.7-rc3-linux-(?P<pver>(\d+)T(\d+)Z)$" diff --git a/recipes-kernel/linux/linux-yocto-dev.bbappend b/recipes-kernel/linux/linux-yocto-dev.bbappend index 6173148c..deb19940 100644 --- a/recipes-kernel/linux/linux-yocto-dev.bbappend +++ b/recipes-kernel/linux/linux-yocto-dev.bbappend @@ -1,21 +1,21 @@ -FILESEXTRAPATHS_prepend_intel-x86-common := "${THISDIR}/${PN}:" +FILESEXTRAPATHS:prepend:intel-x86-common := "${THISDIR}/${PN}:" -COMPATIBLE_MACHINE_core2-32-intel-common = "${MACHINE}" -KMACHINE_core2-32-intel-common = "intel-core2-32" -KERNEL_FEATURES_append_core2-32-intel-common = "${KERNEL_FEATURES_INTEL_COMMON}" +COMPATIBLE_MACHINE:core2-32-intel-common = "${MACHINE}" +KMACHINE:core2-32-intel-common = "intel-core2-32" +KERNEL_FEATURES:append:core2-32-intel-common = "${KERNEL_FEATURES_INTEL_COMMON}" -COMPATIBLE_MACHINE_corei7-64-intel-common = "${MACHINE}" -KMACHINE_corei7-64-intel-common = "intel-corei7-64" -KERNEL_FEATURES_append_corei7-64-intel-common = "${KERNEL_FEATURES_INTEL_COMMON}" +COMPATIBLE_MACHINE:corei7-64-intel-common = "${MACHINE}" +KMACHINE:corei7-64-intel-common = "intel-corei7-64" +KERNEL_FEATURES:append:corei7-64-intel-common = "${KERNEL_FEATURES_INTEL_COMMON}" # NOTE: We do not set SRCREVs here as -dev is intended to be built with AUTOREV # and setting them here breaks the default mechanism to use AUTOREV if the # default SRCREV is set and linux-yocto-dev is the preferred provider. # For Crystalforest and Romley -KERNEL_MODULE_AUTOLOAD_append_core2-32-intel-common = " uio" -KERNEL_MODULE_AUTOLOAD_append_corei7-64-intel-common = " uio" +KERNEL_MODULE_AUTOLOAD:append:core2-32-intel-common = " uio" +KERNEL_MODULE_AUTOLOAD:append:corei7-64-intel-common = " uio" # For FRI2, NUC -KERNEL_MODULE_AUTOLOAD_append_core2-32-intel-common = " iwlwifi" -KERNEL_MODULE_AUTOLOAD_append_corei7-64-intel-common = " iwlwifi" +KERNEL_MODULE_AUTOLOAD:append:core2-32-intel-common = " iwlwifi" +KERNEL_MODULE_AUTOLOAD:append:corei7-64-intel-common = " iwlwifi" diff --git a/recipes-kernel/linux/linux-yocto-rt_5.%.bbappend b/recipes-kernel/linux/linux-yocto-rt_%.bbappend index daa7dd6b..daa7dd6b 100644 --- a/recipes-kernel/linux/linux-yocto-rt_5.%.bbappend +++ b/recipes-kernel/linux/linux-yocto-rt_%.bbappend diff --git a/recipes-kernel/linux/linux-yocto_%.bbappend b/recipes-kernel/linux/linux-yocto_%.bbappend index 7c29be14..9f3d43e3 100644 --- a/recipes-kernel/linux/linux-yocto_%.bbappend +++ b/recipes-kernel/linux/linux-yocto_%.bbappend @@ -1,4 +1,6 @@ +require meta-intel-compat-kernel.inc + # The kernel build is 64-bit regardless, so include both common overrides. # Without this, the kernel will be missing vars that make it buildable for the # intel-corei7-64 machine. -MACHINEOVERRIDES_prepend_corei7-64-x32-intel-common = "corei7-64-intel-common:" +MACHINEOVERRIDES:prepend:corei7-64-x32-intel-common = "corei7-64-intel-common:" diff --git a/recipes-kernel/linux/linux-yocto_5.%.bbappend b/recipes-kernel/linux/linux-yocto_5.%.bbappend deleted file mode 100644 index daa7dd6b..00000000 --- a/recipes-kernel/linux/linux-yocto_5.%.bbappend +++ /dev/null @@ -1 +0,0 @@ -require meta-intel-compat-kernel.inc diff --git a/recipes-kernel/linux/meta-intel-compat-kernel.inc b/recipes-kernel/linux/meta-intel-compat-kernel.inc index a985925f..0c861e13 100644 --- a/recipes-kernel/linux/meta-intel-compat-kernel.inc +++ b/recipes-kernel/linux/meta-intel-compat-kernel.inc @@ -3,12 +3,12 @@ KERNEL_FEATURES_INTEL_COMMON ?= "" -COMPATIBLE_MACHINE_intel-x86-common = "${MACHINE}" -KERNEL_FEATURES_intel-x86-common = "${KERNEL_FEATURES_INTEL_COMMON}" +COMPATIBLE_MACHINE:intel-x86-common = "${MACHINE}" +KERNEL_FEATURES:intel-x86-common = "${KERNEL_FEATURES_INTEL_COMMON}" -KMACHINE_corei7-64-intel-common = "intel-corei7-64" -KMACHINE_core2-32-intel-common = "intel-core2-32" -KMACHINE_skylake-64-intel-common = "intel-corei7-64" +KMACHINE:corei7-64-intel-common = "intel-corei7-64" +KMACHINE:core2-32-intel-common = "intel-core2-32" +KMACHINE:x86-64-v3-intel-common = "intel-corei7-64" INTEL_COMMON_AUTOLOAD ?= " uio iwlwifi i915" -KERNEL_MODULE_AUTOLOAD_intel-x86-common = " ${INTEL_COMMON_AUTOLOAD}" +KERNEL_MODULE_AUTOLOAD:intel-x86-common = " ${INTEL_COMMON_AUTOLOAD}" diff --git a/recipes-multimedia/itt/itt_3.18.10.bb b/recipes-multimedia/itt/itt_3.24.6.bb index c410977b..1c4759f9 100644 --- a/recipes-multimedia/itt/itt_3.18.10.bb +++ b/recipes-multimedia/itt/itt_3.24.6.bb @@ -3,18 +3,20 @@ DESCRIPTION = "The Instrumentation and Tracing Technology (ITT) API enables \ application to generate and control the collection of trace data during its \ execution across different Intel tools." -LICENSE = "BSD-3-Clause & GPLv2" +LICENSE = "BSD-3-Clause & GPL-2.0-only" LIC_FILES_CHKSUM = "file://LICENSES/BSD-3-Clause.txt;md5=c551872bcf41ce707df54c722edeca7b \ file://LICENSES/GPL-2.0-only.txt;md5=e2d76e7801260c21b90eea3605508ad6 \ " -SRC_URI = "git://github.com/intel/ittapi.git;protocol=https" -SRCREV = "66ebaee7a44fcdfb6ae052beb9507318781f3cd0" +SRC_URI = "git://github.com/intel/ittapi.git;protocol=https;branch=master" +SRCREV = "ec077431fc725dba590b09b2cec5137878b14783" S = "${WORKDIR}/git" PE = "1" UPSTREAM_CHECK_GITTAGREGEX = "^v(?P<pver>(\d+(\.\d+)+))$" +COMPATIBLE_HOST = '(i.86|x86_64).*-linux' + inherit pkgconfig cmake EXTRA_OECMAKE += "-DCMAKE_BUILD_TYPE=Release" @@ -34,4 +36,4 @@ do_install() { } -RDEPENDS_${PN}-dev_remove = "${PN} (= ${EXTENDPKGV})" +RDEPENDS:${PN}-dev:remove = "${PN} (= ${EXTENDPKGV})" diff --git a/recipes-multimedia/libva/files/0001-Disable-vp9-padding-on-mtl.patch b/recipes-multimedia/libva/files/0001-Disable-vp9-padding-on-mtl.patch new file mode 100644 index 00000000..90dff8d1 --- /dev/null +++ b/recipes-multimedia/libva/files/0001-Disable-vp9-padding-on-mtl.patch @@ -0,0 +1,35 @@ +From 1b303f417113ad1aa6b63fc024fbe4aa0c943f57 Mon Sep 17 00:00:00 2001 +From: Lim Siew Hoon <siew.hoon.lim@intel.com> +Date: Tue, 5 Sep 2023 16:13:42 +0800 +Subject: [PATCH 01/12] Disable vp9 padding on mtl. + +Upstream-Status: Submitted [https://github.com/intel/media-driver/pull/1720] + +Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com> +--- + media_softlet/linux/Xe_M_plus/ddi/media_sku_wa_mtl.cpp | 4 ++-- + 1 file changed, 2 insertions(+), 2 deletions(-) + +diff --git a/media_softlet/linux/Xe_M_plus/ddi/media_sku_wa_mtl.cpp b/media_softlet/linux/Xe_M_plus/ddi/media_sku_wa_mtl.cpp +index 844545a87..72265289c 100644 +--- a/media_softlet/linux/Xe_M_plus/ddi/media_sku_wa_mtl.cpp ++++ b/media_softlet/linux/Xe_M_plus/ddi/media_sku_wa_mtl.cpp +@@ -284,7 +284,7 @@ static bool InitMtlMediaWaExt(struct GfxDeviceInfo *devInfo, + + MEDIA_WR_WA(waTable, WaDisableSetObjectCapture, 1); + +- MEDIA_WR_WA(waTable, Wa_Vp9UnalignedHeight, 1); ++ MEDIA_WR_WA(waTable, Wa_Vp9UnalignedHeight, 0); + + MEDIA_WR_WA(waTable, Wa_15013355402, 1); + +@@ -337,4 +337,4 @@ static struct LinuxDeviceInit arlDeviceInit = + }; + + static bool arlDeviceRegister = DeviceInfoFactory<LinuxDeviceInit>:: +- RegisterDevice((uint32_t)IGFX_ARROWLAKE, &arlDeviceInit); +\ No newline at end of file ++ RegisterDevice((uint32_t)IGFX_ARROWLAKE, &arlDeviceInit); +-- +2.40.1 + diff --git a/recipes-multimedia/libva/files/0002-Force-ARGB-surface-to-tile4-for-ACM.patch b/recipes-multimedia/libva/files/0002-Force-ARGB-surface-to-tile4-for-ACM.patch new file mode 100644 index 00000000..df3d9805 --- /dev/null +++ b/recipes-multimedia/libva/files/0002-Force-ARGB-surface-to-tile4-for-ACM.patch @@ -0,0 +1,44 @@ +From 1580f01ec5ad5afdad58c39dded999494275be10 Mon Sep 17 00:00:00 2001 +From: Lim Siew Hoon <siew.hoon.lim@intel.com> +Date: Wed, 11 Oct 2023 15:36:21 +0800 +Subject: [PATCH 02/12] Force ARGB surface to tile4 for ACM + +Upstream-Status: Submitted [https://github.com/intel/media-driver/pull/1728] + +Signed-off-by: Soon, Thean Siew <thean.siew.soon@intel.com> +Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com> +--- + media_driver/linux/common/ddi/media_libva_util.cpp | 10 ++++++++++ + 1 file changed, 10 insertions(+) + +diff --git a/media_driver/linux/common/ddi/media_libva_util.cpp b/media_driver/linux/common/ddi/media_libva_util.cpp +index 63c173419..73be76366 100755 +--- a/media_driver/linux/common/ddi/media_libva_util.cpp ++++ b/media_driver/linux/common/ddi/media_libva_util.cpp +@@ -504,6 +504,7 @@ VAStatus DdiMediaUtil_AllocateSurface( + gmmCustomParams.Flags.Gpu.UnifiedAuxSurface = 0; + } + } ++ + break; + case TILING_X: + gmmCustomParams.Flags.Info.TiledX = true; +@@ -685,6 +686,15 @@ VAStatus DdiMediaUtil_AllocateSurface( + } + } + } ++ // There's no VAAPI interface for modifier query yet. As a workaround, always allocate ++ // RGB surface as tile4. ++ if ((format == Media_Format_A8R8G8B8 || ++ format == Media_Format_B10G10R10A2 | ++ format == Media_Format_A8B8G8R8 || ++ format == Media_Format_X8R8G8B8) && !MEDIA_IS_SKU(&mediaDrvCtx->SkuTable, FtrTileY)) ++ { ++ gmmParams.Flags.Info.Tile4 = true; ++ } + break; + case TILING_X: + gmmParams.Flags.Info.TiledX = true; +-- +2.40.1 + diff --git a/recipes-multimedia/libva/files/0004-Add-device-ID-for-ARL.patch b/recipes-multimedia/libva/files/0004-Add-device-ID-for-ARL.patch new file mode 100644 index 00000000..59fff127 --- /dev/null +++ b/recipes-multimedia/libva/files/0004-Add-device-ID-for-ARL.patch @@ -0,0 +1,45 @@ +From bb860c0bfe53060925b1dce50da4e6e94cfc8cae Mon Sep 17 00:00:00 2001 +From: huangli2018 <li.h.huang@intel.com> +Date: Wed, 20 Dec 2023 15:47:43 +0800 +Subject: [PATCH 04/13] Add device ID for ARL + +Add device ID for ARL + +Upstream-Status: Backport [https://github.com/intel/media-driver/commit/95cbfce81cafa3c24d9d4601d41a855093bc51fb] +Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com> +--- + .../linux/Xe_M_plus/ddi/media_sysinfo_mtl.cpp | 13 +++++++++++-- + 1 file changed, 11 insertions(+), 2 deletions(-) + +diff --git a/media_softlet/linux/Xe_M_plus/ddi/media_sysinfo_mtl.cpp b/media_softlet/linux/Xe_M_plus/ddi/media_sysinfo_mtl.cpp +index 426bd795f..7b73dfaff 100644 +--- a/media_softlet/linux/Xe_M_plus/ddi/media_sysinfo_mtl.cpp ++++ b/media_softlet/linux/Xe_M_plus/ddi/media_sysinfo_mtl.cpp +@@ -1,6 +1,6 @@ + /*===================== begin_copyright_notice ================================== + +-Copyright (c) 2022, Intel Corporation ++Copyright (c) 2023, Intel Corporation + + Permission is hereby granted, free of charge, to any person obtaining a + copy of this software and associated documentation files (the "Software"), +@@ -218,4 +218,13 @@ static struct GfxDeviceInfo arlInfo = { + }; + + static bool arlsDevice7d67 = DeviceInfoFactory<GfxDeviceInfo>:: +- RegisterDevice(0x7D67, &arlInfo); +\ No newline at end of file ++ RegisterDevice(0x7D67, &arlInfo); ++ ++static bool arlhDevice7d51 = DeviceInfoFactory<GfxDeviceInfo>:: ++ RegisterDevice(0x7D51, &arlInfo); ++ ++static bool arlhDevice7dd1 = DeviceInfoFactory<GfxDeviceInfo>:: ++ RegisterDevice(0x7DD1, &arlInfo); ++ ++static bool arlhDevice7d41 = DeviceInfoFactory<GfxDeviceInfo>:: ++ RegisterDevice(0x7D41, &arlInfo); +\ No newline at end of file +-- +2.40.1 + diff --git a/recipes-multimedia/libva/files/0005-Add-XR24-support-to-DMABuf.patch b/recipes-multimedia/libva/files/0005-Add-XR24-support-to-DMABuf.patch new file mode 100644 index 00000000..a9cd2317 --- /dev/null +++ b/recipes-multimedia/libva/files/0005-Add-XR24-support-to-DMABuf.patch @@ -0,0 +1,34 @@ +From bbd3af94a6146d702909ff7e25b0c0547ecbb895 Mon Sep 17 00:00:00 2001 +From: jlui <jonathan.ming.jun.lui@intel.com> +Date: Wed, 15 Nov 2023 10:48:34 +0800 +Subject: [PATCH 05/13] Add XR24 support to DMABuf + +This is to enable XR24/BGRX support for DMABuf + +Signed-off-by: Lui, Jonathan Ming Jun <jonathan.ming.jun.lui@intel.com> + +Upstream-Status: Backport [https://github.com/intel/media-driver/commit/3910b7029f5416fd0b5972f940c65d0a82c48c7b] + +Signed-off-by: Lim, Siew Hoon <siew.hoon.lim@intel.com> +--- + .../vp/ddi/capstable_data_vp_xe_lpm_plus_r0_specific.h | 3 ++- + 1 file changed, 2 insertions(+), 1 deletion(-) + +diff --git a/media_softlet/linux/xe_lpm_plus_r0/vp/ddi/capstable_data_vp_xe_lpm_plus_r0_specific.h b/media_softlet/linux/xe_lpm_plus_r0/vp/ddi/capstable_data_vp_xe_lpm_plus_r0_specific.h +index fdcca8873..2dcf440a4 100644 +--- a/media_softlet/linux/xe_lpm_plus_r0/vp/ddi/capstable_data_vp_xe_lpm_plus_r0_specific.h ++++ b/media_softlet/linux/xe_lpm_plus_r0/vp/ddi/capstable_data_vp_xe_lpm_plus_r0_specific.h +@@ -55,8 +55,9 @@ static ProfileSurfaceAttribInfo surfaceAttribInfo_VAEntrypointVideoProc_VAProfil + {VASurfaceAttribPixelFormat, VA_SURFACE_ATTRIB_GETTABLE | VA_SURFACE_ATTRIB_SETTABLE, {VAGenericValueTypeInteger, {VA_FOURCC('4', '2', '2', 'V')}}}, + {VASurfaceAttribPixelFormat, VA_SURFACE_ATTRIB_GETTABLE | VA_SURFACE_ATTRIB_SETTABLE, {VAGenericValueTypeInteger, {VA_FOURCC('R', 'G', 'B', 'A')}}}, + {VASurfaceAttribPixelFormat, VA_SURFACE_ATTRIB_GETTABLE | VA_SURFACE_ATTRIB_SETTABLE, {VAGenericValueTypeInteger, {VA_FOURCC('B', 'G', 'R', 'A')}}}, +- {VASurfaceAttribPixelFormat, VA_SURFACE_ATTRIB_GETTABLE | VA_SURFACE_ATTRIB_SETTABLE, {VAGenericValueTypeInteger, { VA_FOURCC('R', 'G', 'B', 'P')}}}, ++ {VASurfaceAttribPixelFormat, VA_SURFACE_ATTRIB_GETTABLE | VA_SURFACE_ATTRIB_SETTABLE, {VAGenericValueTypeInteger, {VA_FOURCC('R', 'G', 'B', 'P')}}}, + {VASurfaceAttribPixelFormat, VA_SURFACE_ATTRIB_GETTABLE | VA_SURFACE_ATTRIB_SETTABLE, {VAGenericValueTypeInteger, {VA_FOURCC('R', 'G', 'B', 'X')}}}, ++ {VASurfaceAttribPixelFormat, VA_SURFACE_ATTRIB_GETTABLE | VA_SURFACE_ATTRIB_SETTABLE, {VAGenericValueTypeInteger, {VA_FOURCC('X', 'R', 'G', 'B')}}}, + {VASurfaceAttribPixelFormat, VA_SURFACE_ATTRIB_GETTABLE | VA_SURFACE_ATTRIB_SETTABLE, {VAGenericValueTypeInteger, {VA_FOURCC('P', '0', '1', '0')}}}, + {VASurfaceAttribPixelFormat, VA_SURFACE_ATTRIB_GETTABLE | VA_SURFACE_ATTRIB_SETTABLE, {VAGenericValueTypeInteger, {VA_FOURCC('R', 'G', '2', '4')}}}, + {VASurfaceAttribPixelFormat, VA_SURFACE_ATTRIB_GETTABLE | VA_SURFACE_ATTRIB_SETTABLE, {VAGenericValueTypeInteger, {VA_FOURCC_ARGB}}}, +-- +2.40.1 + diff --git a/recipes-multimedia/libva/files/0006-add-INTEL-MEDIA-ALLOC-refineE-to-specify-the-memory-.patch b/recipes-multimedia/libva/files/0006-add-INTEL-MEDIA-ALLOC-refineE-to-specify-the-memory-.patch new file mode 100644 index 00000000..5b04201a --- /dev/null +++ b/recipes-multimedia/libva/files/0006-add-INTEL-MEDIA-ALLOC-refineE-to-specify-the-memory-.patch @@ -0,0 +1,214 @@ +From 67fb6128adf8fc03d429393e98f0982b42a40e64 Mon Sep 17 00:00:00 2001 +From: Zhang_Xinfeng <carl.zhang@intel.com> +Date: Wed, 27 Dec 2023 09:59:29 +0800 +Subject: [PATCH 06/12] add INTEL MEDIA ALLOC refineE to specify the memory + alignment + +this key is used to allocate bigger pages +env variable INTEL_MEDIA_ALLOC_refineE also could be used +and was not enabled by default +mode 0 is default mode +mode 1 is < 64 align to 64 +mode 2 is > 1M && <= 3M align to 1M, >3M align to 2M +mode 3 is mode 1 & mode 2 + +Upstream-Status: Backport [https://github.com/intel/media-driver/commit/765dd939dcc5562d18cca18e5eda505bda952797] +Signed-off-by: Lim, Siew Hoon <siew.hoon.lim@intel.com> +--- + .../linux/common/os/i915/mos_bufmgr.c | 75 +++++++++++++++---- + .../common/os/i915_production/mos_bufmgr.c | 1 + + .../common/os/mos_context_specific_next.cpp | 21 +++++- + .../common/os/mos_user_setting_specific.cpp | 7 ++ + 4 files changed, 90 insertions(+), 14 deletions(-) + +diff --git a/media_softlet/linux/common/os/i915/mos_bufmgr.c b/media_softlet/linux/common/os/i915/mos_bufmgr.c +index c0b3ba027..b623d0129 100644 +--- a/media_softlet/linux/common/os/i915/mos_bufmgr.c ++++ b/media_softlet/linux/common/os/i915/mos_bufmgr.c +@@ -138,7 +138,7 @@ struct mos_bufmgr_gem { + int exec_count; + + /** Array of lists of cached gem objects of power-of-two sizes */ +- struct mos_gem_bo_bucket cache_bucket[14 * 4]; ++ struct mos_gem_bo_bucket cache_bucket[64]; + int num_buckets; + time_t time; + +@@ -3852,9 +3852,9 @@ add_bucket(struct mos_bufmgr_gem *bufmgr_gem, int size) + } + + static void +-init_cache_buckets(struct mos_bufmgr_gem *bufmgr_gem) ++init_cache_buckets(struct mos_bufmgr_gem *bufmgr_gem, uint8_t alloc_mode) + { +- unsigned long size, cache_max_size = 64 * 1024 * 1024; ++ unsigned long size, cache_max_size = 64 * 1024 * 1024, unit_size; + + /* OK, so power of two buckets was too wasteful of memory. + * Give 3 other sizes between each power of two, to hopefully +@@ -3864,17 +3864,63 @@ init_cache_buckets(struct mos_bufmgr_gem *bufmgr_gem) + * width/height alignment and rounding of sizes to pages will + * get us useful cache hit rates anyway) + */ +- add_bucket(bufmgr_gem, 4096); +- add_bucket(bufmgr_gem, 4096 * 2); +- add_bucket(bufmgr_gem, 4096 * 3); ++ /* alloc_mode 0 is default alloc_mode ++ * alloc_mode 1 rounding up to 64K for all < 1M ++ * alloc_mode 2 rounding up to 2M for size> 1M ++ * alloc_mode 3 rounding up to 2M for size > 1M and 64K for size <= 1M */ ++ if( alloc_mode > 3 ) ++ alloc_mode = 0; ++ ++ if ( 0 == alloc_mode || 2 == alloc_mode) ++ { ++ // < 1M normal alloc_mode ++ add_bucket(bufmgr_gem, 4096); ++ add_bucket(bufmgr_gem, 4096 * 2); ++ add_bucket(bufmgr_gem, 4096 * 3); ++ /* Initialize the linked lists for BO reuse cache. */ ++ for (size = 4 * 4096; size < 1024 * 1024; size *= 2) { ++ add_bucket(bufmgr_gem, size); ++ add_bucket(bufmgr_gem, size + size * 1 / 4); ++ add_bucket(bufmgr_gem, size + size * 2 / 4); ++ add_bucket(bufmgr_gem, size + size * 3 / 4); ++ } + +- /* Initialize the linked lists for BO reuse cache. */ +- for (size = 4 * 4096; size <= cache_max_size; size *= 2) { +- add_bucket(bufmgr_gem, size); ++ add_bucket(bufmgr_gem, 1024 * 1024); ++ } ++ if (1 == alloc_mode || 3 == alloc_mode) ++ { ++ // < 1M 64k alignment ++ unit_size = 64 * 1024; ++ for (size = unit_size; size <= 1024 * 1024; size += unit_size) ++ { ++ add_bucket(bufmgr_gem, size); ++ } ++ } ++ if( 0 == alloc_mode || 1 == alloc_mode) ++ { ++ //> 1M is normal alloc_mode ++ add_bucket(bufmgr_gem, 1280 * 1024); ++ add_bucket(bufmgr_gem, 1536 * 1024); ++ add_bucket(bufmgr_gem, 1792 * 1024); ++ ++ for (size = 2 * 1024 * 1024; size < cache_max_size; size *= 2) { ++ add_bucket(bufmgr_gem, size); ++ add_bucket(bufmgr_gem, size + size * 1 / 4); ++ add_bucket(bufmgr_gem, size + size * 2 / 4); ++ add_bucket(bufmgr_gem, size + size * 3 / 4); ++ } ++ } ++ if( 2 == alloc_mode || 3 == alloc_mode) ++ { ++ //> 1M rolling to 2M ++ unit_size = 2 * 1024 * 1024; ++ add_bucket(bufmgr_gem, unit_size); ++ add_bucket(bufmgr_gem, 3 * 1024 * 1024); + +- add_bucket(bufmgr_gem, size + size * 1 / 4); +- add_bucket(bufmgr_gem, size + size * 2 / 4); +- add_bucket(bufmgr_gem, size + size * 3 / 4); ++ for (size = 4 * 1024 * 1024; size <= cache_max_size; size += unit_size) ++ { ++ add_bucket(bufmgr_gem, size); ++ } + } + } + +@@ -5100,6 +5146,7 @@ mos_bufmgr_gem_init_i915(int fd, int batch_size) + struct drm_i915_gem_get_aperture aperture; + drm_i915_getparam_t gp; + int ret, tmp; ++ uint8_t alloc_mode; + bool exec2 = false; + + pthread_mutex_lock(&bufmgr_list_mutex); +@@ -5352,10 +5399,12 @@ mos_bufmgr_gem_init_i915(int fd, int batch_size) + * + * Every 4 was too few for the blender benchmark. + */ ++ alloc_mode = (uint8_t)(batch_size & 0xff); ++ batch_size &= 0xffffff00; + bufmgr_gem->max_relocs = batch_size / sizeof(uint32_t) / 2 - 2; + + DRMINITLISTHEAD(&bufmgr_gem->named); +- init_cache_buckets(bufmgr_gem); ++ init_cache_buckets(bufmgr_gem,alloc_mode); + + DRMLISTADD(&bufmgr_gem->managers, &bufmgr_list); + +diff --git a/media_softlet/linux/common/os/i915_production/mos_bufmgr.c b/media_softlet/linux/common/os/i915_production/mos_bufmgr.c +index 90b5685b1..b3574f7d3 100644 +--- a/media_softlet/linux/common/os/i915_production/mos_bufmgr.c ++++ b/media_softlet/linux/common/os/i915_production/mos_bufmgr.c +@@ -5403,6 +5403,7 @@ mos_bufmgr_gem_init_i915(int fd, int batch_size) + * + * Every 4 was too few for the blender benchmark. + */ ++ batch_size &= 0xffffff00; + bufmgr_gem->max_relocs = batch_size / sizeof(uint32_t) / 2 - 2; + + DRMINITLISTHEAD(&bufmgr_gem->named); +diff --git a/media_softlet/linux/common/os/mos_context_specific_next.cpp b/media_softlet/linux/common/os/mos_context_specific_next.cpp +index 9e9e3ff7e..543e262d1 100644 +--- a/media_softlet/linux/common/os/mos_context_specific_next.cpp ++++ b/media_softlet/linux/common/os/mos_context_specific_next.cpp +@@ -64,6 +64,7 @@ MOS_STATUS OsContextSpecificNext::Init(DDI_DEVICE_CONTEXT ddiDriverContext) + uint32_t iDeviceId = 0; + MOS_STATUS eStatus = MOS_STATUS_SUCCESS; + uint32_t value = 0; ++ uint32_t mode = 0; + MediaUserSettingSharedPtr userSettingPtr = nullptr; + + MOS_OS_FUNCTION_ENTER; +@@ -89,7 +90,25 @@ MOS_STATUS OsContextSpecificNext::Init(DDI_DEVICE_CONTEXT ddiDriverContext) + + userSettingPtr = MosInterface::MosGetUserSettingInstance(osDriverContext); + +- m_bufmgr = mos_bufmgr_gem_init(m_fd, BATCH_BUFFER_SIZE, &m_deviceType); ++ mode = BATCH_BUFFER_SIZE; ++ ReadUserSetting( ++ userSettingPtr, ++ value, ++ "INTEL MEDIA ALLOC MODE", ++ MediaUserSetting::Group::Device); ++ ++ if (value) ++ { ++ mode |= (value & 0x000000ff); ++ } ++ value = 0; ++ /* no need to set batch buffer size after switch to softpin ++ * keep it, just for test during relocation to softpin transition ++ * now , it could be a debug method , but is actually useless ++ * so it is safe to reuse the lowest 8bit to convey addtional information ++ * more suitable solution is deleting it , or add additional parameter*/ ++ ++ m_bufmgr = mos_bufmgr_gem_init(m_fd, (int)mode, &m_deviceType); + if (nullptr == m_bufmgr) + { + MOS_OS_ASSERTMESSAGE("Not able to allocate buffer manager, fd=0x%d", m_fd); +diff --git a/media_softlet/linux/common/os/mos_user_setting_specific.cpp b/media_softlet/linux/common/os/mos_user_setting_specific.cpp +index 6be8b4298..caed584f4 100644 +--- a/media_softlet/linux/common/os/mos_user_setting_specific.cpp ++++ b/media_softlet/linux/common/os/mos_user_setting_specific.cpp +@@ -52,5 +52,12 @@ MOS_STATUS MosUserSetting::InitMosUserSettingSpecific(MediaUserSettingSharedPtr + 0, + true); //"Enable VM Bind." + ++ DeclareUserSettingKey( ++ userSettingPtr, ++ "INTEL MEDIA ALLOC MODE", ++ MediaUserSetting::Group::Device, ++ 0, ++ false); // ++ + return MOS_STATUS_SUCCESS; + } +-- +2.40.1 + diff --git a/recipes-multimedia/libva/files/0007-Skip-report-keys.patch b/recipes-multimedia/libva/files/0007-Skip-report-keys.patch new file mode 100644 index 00000000..459ff490 --- /dev/null +++ b/recipes-multimedia/libva/files/0007-Skip-report-keys.patch @@ -0,0 +1,78 @@ +From b1aebef4d342c77fc2b9c5efbf20aeb2aa9d634e Mon Sep 17 00:00:00 2001 +From: Wang_Pingli <pingli.wang@intel.com> +Date: Thu, 28 Dec 2023 16:44:40 +0800 +Subject: [PATCH 07/12] Skip report keys + +Skip to report keys + +Upstream-Status: Backport [https://github.com/intel/media-driver/commit/68ce25b0a6fa90614eb4734c8680aa4e149e8323] +Signed-off-by: Lim, Siew Hoon <siew.hoon.lim@intel.com> +--- + .../os/osservice/mos_utilities_specific.cpp | 30 +++++++++++++++++-- + 1 file changed, 28 insertions(+), 2 deletions(-) + +diff --git a/media_softlet/linux/common/os/osservice/mos_utilities_specific.cpp b/media_softlet/linux/common/os/osservice/mos_utilities_specific.cpp +index b5df29f06..96785e21e 100644 +--- a/media_softlet/linux/common/os/osservice/mos_utilities_specific.cpp ++++ b/media_softlet/linux/common/os/osservice/mos_utilities_specific.cpp +@@ -67,6 +67,8 @@ int32_t g_mosMemoryFailSimulateAllocCounter = 0; + int32_t *MosUtilities::m_mosAllocMemoryFailSimulateAllocCounter = &g_mosMemoryFailSimulateAllocCounter; + #endif + ++static bool s_skipToReportReg = false; ++ + double MosUtilities::MosGetTime() + { + struct timespec ts = {}; +@@ -1551,6 +1553,10 @@ MOS_STATUS MosUtilities::MosInitializeReg(RegBufferMap ®BufferMap) + { + std::string id = ""; + ++ static const char *disableReportRegKeyList[] = { ++ "INTEL MEDIA ALLOC MODE" ++ }; ++ static const uint32_t disableReportRegKeyListCount = sizeof(disableReportRegKeyList) / sizeof(disableReportRegKeyList[0]); + while(!regStream.eof()) + { + std::string line = ""; +@@ -1585,8 +1591,22 @@ MOS_STATUS MosUtilities::MosInitializeReg(RegBufferMap ®BufferMap) + { + std::string name = line.substr(0,pos); + std::string value = line.substr(pos+1); +- auto &keys = regBufferMap[id]; +- keys[name] = value; ++ if (name.size() > 0 && value.size() > 0) ++ { ++ auto &keys = regBufferMap[id]; ++ keys[name] = value; ++ if (s_skipToReportReg == false && id == USER_SETTING_CONFIG_PATH) ++ { ++ for (uint32_t i = 0; i < disableReportRegKeyListCount; i++) ++ { ++ if (strcmp(name.c_str(), disableReportRegKeyList[i]) == 0) ++ { ++ s_skipToReportReg = true; ++ break; ++ } ++ } ++ } ++ } + } + } + } +@@ -1606,6 +1626,12 @@ MOS_STATUS MosUtilities::MosInitializeReg(RegBufferMap ®BufferMap) + MOS_STATUS MosUtilities::MosUninitializeReg(RegBufferMap ®BufferMap) + { + MOS_STATUS status = MOS_STATUS_SUCCESS; ++ ++ if (s_skipToReportReg) ++ { ++ return MOS_STATUS_SUCCESS; ++ } ++ + if (regBufferMap.size() == 0) + { + return MOS_STATUS_SUCCESS; +-- +2.40.1 + diff --git a/recipes-multimedia/libva/files/0008-Limit-INTEL-MEDIA-ALLOC-MODE-to-MTL-and-ARL-only.patch b/recipes-multimedia/libva/files/0008-Limit-INTEL-MEDIA-ALLOC-MODE-to-MTL-and-ARL-only.patch new file mode 100644 index 00000000..ece5e00b --- /dev/null +++ b/recipes-multimedia/libva/files/0008-Limit-INTEL-MEDIA-ALLOC-MODE-to-MTL-and-ARL-only.patch @@ -0,0 +1,255 @@ +From ac2ffac8d4aade216cc361c6c7120c13a8780719 Mon Sep 17 00:00:00 2001 +From: "Soon, Thean Siew" <thean.siew.soon@intel.com> +Date: Thu, 28 Dec 2023 18:42:59 +0000 +Subject: [PATCH 08/12] Limit INTEL MEDIA ALLOC MODE to MTL and ARL only + +Upstream-Status: Submitted [https://github.com/intel/media-driver/pull/1754] + +Signed-off-by: Soon, Thean Siew <thean.siew.soon@intel.com> +--- + .../common/os/i915/include/mos_bufmgr_api.h | 1 + + .../common/os/i915/include/mos_bufmgr_priv.h | 2 +- + .../linux/common/os/i915/mos_bufmgr.c | 69 +++++++++++++++---- + .../linux/common/os/i915/mos_bufmgr_api.c | 19 +++++ + .../os/i915_production/mos_bufmgr_priv.h | 1 + + .../common/os/mos_context_specific_next.cpp | 36 +++++----- + 6 files changed, 94 insertions(+), 34 deletions(-) + +diff --git a/media_softlet/linux/common/os/i915/include/mos_bufmgr_api.h b/media_softlet/linux/common/os/i915/include/mos_bufmgr_api.h +index 71cc01099..ab4f1ba89 100644 +--- a/media_softlet/linux/common/os/i915/include/mos_bufmgr_api.h ++++ b/media_softlet/linux/common/os/i915/include/mos_bufmgr_api.h +@@ -299,6 +299,7 @@ void mos_bufmgr_enable_vmbind(struct mos_bufmgr *bufmgr); + void mos_bufmgr_disable_object_capture(struct mos_bufmgr *bufmgr); + int mos_bufmgr_get_memory_info(struct mos_bufmgr *bufmgr, char *info, uint32_t length); + int mos_bufmgr_get_devid(struct mos_bufmgr *bufmgr); ++void mos_bufmgr_realloc_cache(struct mos_bufmgr *bufmgr, uint8_t alloc_mode); + + int mos_bo_map_unsynchronized(struct mos_linux_bo *bo); + int mos_bo_map_gtt(struct mos_linux_bo *bo); +diff --git a/media_softlet/linux/common/os/i915/include/mos_bufmgr_priv.h b/media_softlet/linux/common/os/i915/include/mos_bufmgr_priv.h +index 63f69f985..6fd58f827 100644 +--- a/media_softlet/linux/common/os/i915/include/mos_bufmgr_priv.h ++++ b/media_softlet/linux/common/os/i915/include/mos_bufmgr_priv.h +@@ -358,9 +358,9 @@ struct mos_bufmgr { + void (*disable_object_capture)(struct mos_bufmgr *bufmgr) = nullptr; + int (*get_memory_info)(struct mos_bufmgr *bufmgr, char *info, uint32_t length) = nullptr; + int (*get_devid)(struct mos_bufmgr *bufmgr) = nullptr; ++ void (*realloc_cache)(struct mos_bufmgr *bufmgr, uint8_t alloc_mode) = nullptr; + int (*query_engines_count)(struct mos_bufmgr *bufmgr, + unsigned int *nengine) = nullptr; +- + int (*query_engines)(struct mos_bufmgr *bufmgr, + __u16 engine_class, + __u64 caps, +diff --git a/media_softlet/linux/common/os/i915/mos_bufmgr.c b/media_softlet/linux/common/os/i915/mos_bufmgr.c +index b623d0129..2814ddbb7 100644 +--- a/media_softlet/linux/common/os/i915/mos_bufmgr.c ++++ b/media_softlet/linux/common/os/i915/mos_bufmgr.c +@@ -2426,19 +2426,9 @@ mos_gem_bo_start_gtt_access(struct mos_linux_bo *bo, int write_enable) + } + + static void +-mos_bufmgr_gem_destroy(struct mos_bufmgr *bufmgr) ++mos_bufmgr_cleanup_cache(struct mos_bufmgr_gem *bufmgr_gem) + { +- struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bufmgr; +- struct drm_gem_close close_bo; +- int i, ret; +- +- free(bufmgr_gem->exec2_objects); +- free(bufmgr_gem->exec_objects); +- free(bufmgr_gem->exec_bos); +- pthread_mutex_destroy(&bufmgr_gem->lock); +- +- /* Free any cached buffer objects we were going to reuse */ +- for (i = 0; i < bufmgr_gem->num_buckets; i++) { ++ for (int i = 0; i < bufmgr_gem->num_buckets; i++) { + struct mos_gem_bo_bucket *bucket = + &bufmgr_gem->cache_bucket[i]; + struct mos_bo_gem *bo_gem; +@@ -2450,7 +2440,25 @@ mos_bufmgr_gem_destroy(struct mos_bufmgr *bufmgr) + + mos_gem_bo_free(&bo_gem->bo); + } ++ bufmgr_gem->cache_bucket[i].size = 0; + } ++ bufmgr_gem->num_buckets = 0; ++} ++ ++static void ++mos_bufmgr_gem_destroy(struct mos_bufmgr *bufmgr) ++{ ++ struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *)bufmgr; ++ struct drm_gem_close close_bo; ++ int ret; ++ ++ free(bufmgr_gem->exec2_objects); ++ free(bufmgr_gem->exec_objects); ++ free(bufmgr_gem->exec_bos); ++ pthread_mutex_destroy(&bufmgr_gem->lock); ++ ++ /* Free any cached buffer objects we were going to reuse */ ++ mos_bufmgr_cleanup_cache(bufmgr_gem); + + /* Release userptr bo kept hanging around for optimisation. */ + if (bufmgr_gem->userptr_active.ptr) { +@@ -3852,9 +3860,41 @@ add_bucket(struct mos_bufmgr_gem *bufmgr_gem, int size) + } + + static void +-init_cache_buckets(struct mos_bufmgr_gem *bufmgr_gem, uint8_t alloc_mode) ++init_cache_buckets(struct mos_bufmgr_gem *bufmgr_gem) ++{ ++ unsigned long size, cache_max_size = 64 * 1024 * 1024; ++ ++ /* OK, so power of two buckets was too wasteful of memory. ++ * Give 3 other sizes between each power of two, to hopefully ++ * cover things accurately enough. (The alternative is ++ * probably to just go for exact matching of sizes, and assume ++ * that for things like composited window resize the tiled ++ * width/height alignment and rounding of sizes to pages will ++ * get us useful cache hit rates anyway) ++ */ ++ add_bucket(bufmgr_gem, 4096); ++ add_bucket(bufmgr_gem, 4096 * 2); ++ add_bucket(bufmgr_gem, 4096 * 3); ++ ++ /* Initialize the linked lists for BO reuse cache. */ ++ for (size = 4 * 4096; size <= cache_max_size; size *= 2) { ++ add_bucket(bufmgr_gem, size); ++ ++ add_bucket(bufmgr_gem, size + size * 1 / 4); ++ add_bucket(bufmgr_gem, size + size * 2 / 4); ++ add_bucket(bufmgr_gem, size + size * 3 / 4); ++ } ++} ++ ++static void ++mos_gem_realloc_cache(struct mos_bufmgr *bufmgr, uint8_t alloc_mode) + { + unsigned long size, cache_max_size = 64 * 1024 * 1024, unit_size; ++ struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *)bufmgr; ++ ++ // Clean up the pre-allocated cache before re-allocating according ++ // to alloc_mode ++ mos_bufmgr_cleanup_cache(bufmgr_gem); + + /* OK, so power of two buckets was too wasteful of memory. + * Give 3 other sizes between each power of two, to hopefully +@@ -5218,6 +5258,7 @@ mos_bufmgr_gem_init_i915(int fd, int batch_size) + bufmgr_gem->bufmgr.disable_object_capture = mos_gem_disable_object_capture; + bufmgr_gem->bufmgr.get_memory_info = mos_gem_get_memory_info; + bufmgr_gem->bufmgr.get_devid = mos_gem_get_devid; ++ bufmgr_gem->bufmgr.realloc_cache = mos_gem_realloc_cache; + bufmgr_gem->bufmgr.set_context_param = mos_gem_set_context_param; + bufmgr_gem->bufmgr.set_context_param_parallel = mos_gem_set_context_param_parallel; + bufmgr_gem->bufmgr.set_context_param_load_balance = mos_gem_set_context_param_load_balance; +@@ -5404,7 +5445,7 @@ mos_bufmgr_gem_init_i915(int fd, int batch_size) + bufmgr_gem->max_relocs = batch_size / sizeof(uint32_t) / 2 - 2; + + DRMINITLISTHEAD(&bufmgr_gem->named); +- init_cache_buckets(bufmgr_gem,alloc_mode); ++ init_cache_buckets(bufmgr_gem); + + DRMLISTADD(&bufmgr_gem->managers, &bufmgr_list); + +diff --git a/media_softlet/linux/common/os/i915/mos_bufmgr_api.c b/media_softlet/linux/common/os/i915/mos_bufmgr_api.c +index abe8ef96a..d0e02c267 100644 +--- a/media_softlet/linux/common/os/i915/mos_bufmgr_api.c ++++ b/media_softlet/linux/common/os/i915/mos_bufmgr_api.c +@@ -1204,6 +1204,25 @@ mos_bufmgr_get_devid(struct mos_bufmgr *bufmgr) + } + } + ++void ++mos_bufmgr_realloc_cache(struct mos_bufmgr *bufmgr, uint8_t alloc_mode) ++{ ++ if(!bufmgr) ++ { ++ MOS_OS_CRITICALMESSAGE("Input null ptr\n"); ++ return; ++ } ++ ++ if (bufmgr->realloc_cache) ++ { ++ return bufmgr->realloc_cache(bufmgr, alloc_mode); ++ } ++ else ++ { ++ MOS_OS_CRITICALMESSAGE("Unsupported\n"); ++ } ++} ++ + int + mos_query_engines_count(struct mos_bufmgr *bufmgr, + unsigned int *nengine) +diff --git a/media_softlet/linux/common/os/i915_production/mos_bufmgr_priv.h b/media_softlet/linux/common/os/i915_production/mos_bufmgr_priv.h +index c3b765a4c..1d7f7edb0 100644 +--- a/media_softlet/linux/common/os/i915_production/mos_bufmgr_priv.h ++++ b/media_softlet/linux/common/os/i915_production/mos_bufmgr_priv.h +@@ -358,6 +358,7 @@ struct mos_bufmgr { + void (*disable_object_capture)(struct mos_bufmgr *bufmgr) = nullptr; + int (*get_memory_info)(struct mos_bufmgr *bufmgr, char *info, uint32_t length) = nullptr; + int (*get_devid)(struct mos_bufmgr *bufmgr) = nullptr; ++ void (*realloc_cache)(struct mos_bufmgr *bufmgr, uint8_t alloc_mode) = nullptr; + int (*query_engines_count)(struct mos_bufmgr *bufmgr, + unsigned int *nengine) = nullptr; + +diff --git a/media_softlet/linux/common/os/mos_context_specific_next.cpp b/media_softlet/linux/common/os/mos_context_specific_next.cpp +index 543e262d1..87059f538 100644 +--- a/media_softlet/linux/common/os/mos_context_specific_next.cpp ++++ b/media_softlet/linux/common/os/mos_context_specific_next.cpp +@@ -90,25 +90,7 @@ MOS_STATUS OsContextSpecificNext::Init(DDI_DEVICE_CONTEXT ddiDriverContext) + + userSettingPtr = MosInterface::MosGetUserSettingInstance(osDriverContext); + +- mode = BATCH_BUFFER_SIZE; +- ReadUserSetting( +- userSettingPtr, +- value, +- "INTEL MEDIA ALLOC MODE", +- MediaUserSetting::Group::Device); +- +- if (value) +- { +- mode |= (value & 0x000000ff); +- } +- value = 0; +- /* no need to set batch buffer size after switch to softpin +- * keep it, just for test during relocation to softpin transition +- * now , it could be a debug method , but is actually useless +- * so it is safe to reuse the lowest 8bit to convey addtional information +- * more suitable solution is deleting it , or add additional parameter*/ +- +- m_bufmgr = mos_bufmgr_gem_init(m_fd, (int)mode, &m_deviceType); ++ m_bufmgr = mos_bufmgr_gem_init(m_fd, BATCH_BUFFER_SIZE, &m_deviceType); + if (nullptr == m_bufmgr) + { + MOS_OS_ASSERTMESSAGE("Not able to allocate buffer manager, fd=0x%d", m_fd); +@@ -151,6 +133,22 @@ MOS_STATUS OsContextSpecificNext::Init(DDI_DEVICE_CONTEXT ddiDriverContext) + return eStatus; + } + ++ if (m_platformInfo.eProductFamily == IGFX_METEORLAKE || ++ m_platformInfo.eProductFamily == IGFX_ARROWLAKE) ++ { ++ ReadUserSetting( ++ userSettingPtr, ++ value, ++ "INTEL MEDIA ALLOC MODE", ++ MediaUserSetting::Group::Device); ++ ++ if (value) ++ { ++ mode = (value & 0x000000ff); ++ } ++ mos_bufmgr_realloc_cache(m_bufmgr, mode); ++ } ++ + ReadUserSetting( + userSettingPtr, + value, +-- +2.40.1 + diff --git a/recipes-multimedia/libva/files/0009-Skip-cache-bucket-realloc-for-default-mode-0.patch b/recipes-multimedia/libva/files/0009-Skip-cache-bucket-realloc-for-default-mode-0.patch new file mode 100644 index 00000000..8ee7ef14 --- /dev/null +++ b/recipes-multimedia/libva/files/0009-Skip-cache-bucket-realloc-for-default-mode-0.patch @@ -0,0 +1,33 @@ +From 18c32247cbc0e9fabff7a847add099496af9e53f Mon Sep 17 00:00:00 2001 +From: "Soon, Thean Siew" <thean.siew.soon@intel.com> +Date: Mon, 8 Jan 2024 18:12:30 +0000 +Subject: [PATCH 09/12] Skip cache bucket realloc for default mode 0 + +Upstream-Status: Submitted [https://github.com/intel/media-driver/pull/1754] + +Signed-off-by: Soon, Thean Siew <thean.siew.soon@intel.com> +--- + .../linux/common/os/mos_context_specific_next.cpp | 7 ++++++- + 1 file changed, 6 insertions(+), 1 deletion(-) + +diff --git a/media_softlet/linux/common/os/mos_context_specific_next.cpp b/media_softlet/linux/common/os/mos_context_specific_next.cpp +index 87059f538..9dbd5da79 100644 +--- a/media_softlet/linux/common/os/mos_context_specific_next.cpp ++++ b/media_softlet/linux/common/os/mos_context_specific_next.cpp +@@ -146,7 +146,12 @@ MOS_STATUS OsContextSpecificNext::Init(DDI_DEVICE_CONTEXT ddiDriverContext) + { + mode = (value & 0x000000ff); + } +- mos_bufmgr_realloc_cache(m_bufmgr, mode); ++ ++ // Realloc cache only if it's not mode 0 ++ if (mode) ++ { ++ mos_bufmgr_realloc_cache(m_bufmgr, mode); ++ } + } + + ReadUserSetting( +-- +2.40.1 + diff --git a/recipes-multimedia/libva/files/0010-Fix-failed-4k-video-wall-test-case-and-color-corrupt.patch b/recipes-multimedia/libva/files/0010-Fix-failed-4k-video-wall-test-case-and-color-corrupt.patch new file mode 100644 index 00000000..0100df77 --- /dev/null +++ b/recipes-multimedia/libva/files/0010-Fix-failed-4k-video-wall-test-case-and-color-corrupt.patch @@ -0,0 +1,195 @@ +From 7291ab5206c5e172c6c94dcfbe4f1bc258f3b7a5 Mon Sep 17 00:00:00 2001 +From: WenshengZhang <wensheng.zhang@intel.com> +Date: Wed, 10 Jan 2024 14:08:37 +0800 +Subject: [PATCH 10/12] Fix failed 4k video wall test case and color corruption + of video composition + +Fix failed 4k video wall test case from 16CH video only show 1CH output and +corruption observed on certain number of video composition when doing +sample_multi_transcode. + +Upstream-Status: Backport [https://github.com/intel/media-driver/commit/38e14b1b97170b51586d00d1fa607e496b0c5825] +Signed-off-by: Lim, Siew Hoon <siew.hoon.lim@intel.com> +--- + .../common/vp/kdll/hal_kerneldll_next.h | 1 + + .../vp/hal/packet/vp_render_fc_kernel.cpp | 7 ++ + .../common/vp/kdll/hal_kerneldll_next.c | 70 +++++++++++-------- + .../linux/common/vp/ddi/ddi_vp_functions.cpp | 20 ++---- + 4 files changed, 55 insertions(+), 43 deletions(-) + +diff --git a/media_common/agnostic/common/vp/kdll/hal_kerneldll_next.h b/media_common/agnostic/common/vp/kdll/hal_kerneldll_next.h +index 6447999b7..8e2ab371e 100644 +--- a/media_common/agnostic/common/vp/kdll/hal_kerneldll_next.h ++++ b/media_common/agnostic/common/vp/kdll/hal_kerneldll_next.h +@@ -588,6 +588,7 @@ typedef struct tagKdll_FilterEntry + Kdll_Scalingratio ScalingRatio; + Kdll_RenderMethod RenderMethod; + Kdll_SetCSCCoeffMethod SetCSCCoeffMode; ++ bool forceToTargetColorSpace; + } Kdll_FilterEntry, *PKdll_FilterEntry; + + // Structure that defines a compositing filter +diff --git a/media_softlet/agnostic/common/vp/hal/packet/vp_render_fc_kernel.cpp b/media_softlet/agnostic/common/vp/hal/packet/vp_render_fc_kernel.cpp +index 8e982f67e..f8308fed1 100644 +--- a/media_softlet/agnostic/common/vp/hal/packet/vp_render_fc_kernel.cpp ++++ b/media_softlet/agnostic/common/vp/hal/packet/vp_render_fc_kernel.cpp +@@ -775,6 +775,13 @@ MOS_STATUS VpRenderFcKernel::BuildFilter( + + for (i = 0; (i < (int)compParams->sourceCount) && (iMaxFilterSize > 0); i++) + { ++ if (i > 0) ++ { ++ if (!RECT1_CONTAINS_RECT2(compParams->source[0].surf->rcDst, compParams->source[i].surf->rcDst)) ++ { ++ pFilter->forceToTargetColorSpace = true; ++ } ++ } + src = &compParams->source[i]; + + //-------------------------------- +diff --git a/media_softlet/agnostic/common/vp/kdll/hal_kerneldll_next.c b/media_softlet/agnostic/common/vp/kdll/hal_kerneldll_next.c +index d8f7d429a..151f3b039 100644 +--- a/media_softlet/agnostic/common/vp/kdll/hal_kerneldll_next.c ++++ b/media_softlet/agnostic/common/vp/kdll/hal_kerneldll_next.c +@@ -2378,6 +2378,7 @@ bool KernelDll_SetupCSC( + Kdll_CSC_Matrix curr_matrix; + Kdll_CSC_Matrix *matrix = pCSC->Matrix; // Color Space conversion matrix + uint8_t * matrixID = pCSC->MatrixID; // CSC coefficient allocation table ++ bool forceToTargetColorSpace = false; + + // Clear all CSC matrices + MOS_ZeroMemory(matrix, sizeof(pCSC->Matrix)); +@@ -2395,6 +2396,10 @@ bool KernelDll_SetupCSC( + //---------------------------------------------------------------// + for (i = iFilterSize, pFilter = pSearchState->Filter; i > 0; i--, pFilter++) + { ++ if (pFilter->forceToTargetColorSpace) ++ { ++ forceToTargetColorSpace = true; ++ } + // Disable Procamp for all layers except Main Video + // Disable Procamp if source is RGB + if (pFilter->layer != Layer_MainVideo || +@@ -2456,44 +2461,51 @@ bool KernelDll_SetupCSC( + //---------------------------------------------------------------// + if (sel_cspace == CSpace_Any) + { +- int cs; +- for (cs = (CSpace_Any + 1); cs < CSpace_Count; cs++) ++ if (forceToTargetColorSpace) + { +- // Skip color spaces not in use +- cspace = (VPHAL_CSPACE)cs; +- if (!cspace_in_use[cspace]) +- { +- continue; +- } +- +- // xvYCC and BT are treated as same for CSC considerations (BT.x to xvYCC.x matrix is I) +- cspace = KernelDll_TranslateCspace(cspace); +- +- // Count # of CS conversions and matrices +- csc_count = 0; +- for (i = iFilterSize, pFilter = pSearchState->Filter; i > 0; i--, pFilter++) ++ sel_cspace = out_cspace; ++ } ++ else ++ { ++ int cs; ++ for (cs = (CSpace_Any + 1); cs < CSpace_Count; cs++) + { +- // Ignore layers where the Color Space may be set in software (colorfill, palletized) +- if (pFilter->cspace == CSpace_Any) ++ // Skip color spaces not in use ++ cspace = (VPHAL_CSPACE)cs; ++ if (!cspace_in_use[cspace]) + { + continue; + } + +- // Check if CSC/PA is required +- if (KernelDll_TranslateCspace(pFilter->cspace) != cspace || +- pFilter->procamp != DL_PROCAMP_DISABLED) ++ // xvYCC and BT are treated as same for CSC considerations (BT.x to xvYCC.x matrix is I) ++ cspace = KernelDll_TranslateCspace(cspace); ++ ++ // Count # of CS conversions and matrices ++ csc_count = 0; ++ for (i = iFilterSize, pFilter = pSearchState->Filter; i > 0; i--, pFilter++) + { +- csc_count++; ++ // Ignore layers where the Color Space may be set in software (colorfill, palletized) ++ if (pFilter->cspace == CSpace_Any) ++ { ++ continue; ++ } ++ ++ // Check if CSC/PA is required ++ if (KernelDll_TranslateCspace(pFilter->cspace) != cspace || ++ pFilter->procamp != DL_PROCAMP_DISABLED) ++ { ++ csc_count++; ++ } + } +- } + +- // Save best choice as requiring minimum number of CSC operations +- if ((sel_csc_count < 0) || // Initial value +- (csc_count < sel_csc_count) || // Minimum number of CSC operations +- (csc_count == sel_csc_count && cs == main_cspace)) // Use main cspace as default if same CSC count +- { +- sel_cspace = cspace; +- sel_csc_count = csc_count; ++ // Save best choice as requiring minimum number of CSC operations ++ if ((sel_csc_count < 0) || // Initial value ++ (csc_count < sel_csc_count) || // Minimum number of CSC operations ++ (csc_count == sel_csc_count && cs == main_cspace)) // Use main cspace as default if same CSC count ++ { ++ sel_cspace = cspace; ++ sel_csc_count = csc_count; ++ } + } + } + } +diff --git a/media_softlet/linux/common/vp/ddi/ddi_vp_functions.cpp b/media_softlet/linux/common/vp/ddi/ddi_vp_functions.cpp +index 40d0b3fad..e60a6c69c 100644 +--- a/media_softlet/linux/common/vp/ddi/ddi_vp_functions.cpp ++++ b/media_softlet/linux/common/vp/ddi/ddi_vp_functions.cpp +@@ -2115,9 +2115,8 @@ VAStatus DdiVpFunctions::SetBackgroundColorfill( + { + DDI_VP_FUNC_ENTER; + DDI_VP_CHK_NULL(vpHalRenderParams, "nullptr vpHalRenderParams.", VA_STATUS_ERROR_INVALID_PARAMETER); +- DDI_VP_CHK_NULL(vpHalRenderParams->pTarget[0],"nullptr pTarget[0].", VA_STATUS_ERROR_INVALID_PARAMETER); + +- if ((outBackGroundcolor >> 24) != 0 || vpHalRenderParams->pTarget[0]->ColorSpace == CSpace_sRGB) ++ if ((outBackGroundcolor >> 24) != 0) + { + if (vpHalRenderParams->pColorFillParams == nullptr) + { +@@ -2126,18 +2125,11 @@ VAStatus DdiVpFunctions::SetBackgroundColorfill( + + DDI_VP_CHK_NULL(vpHalRenderParams->pColorFillParams, "nullptr pColorFillParams.", VA_STATUS_ERROR_UNKNOWN); + +- if (vpHalRenderParams->pTarget[0]->ColorSpace == CSpace_sRGB && (outBackGroundcolor >> 24) == 0) +- { +- // set color space for sRGB output +- vpHalRenderParams->pColorFillParams->CSpace = CSpace_sRGB; +- } +- else +- { +- // set background colorfill option +- vpHalRenderParams->pColorFillParams->Color = outBackGroundcolor; +- vpHalRenderParams->pColorFillParams->bYCbCr = false; +- vpHalRenderParams->pColorFillParams->CSpace = CSpace_sRGB; +- } ++ // set background colorfill option ++ vpHalRenderParams->pColorFillParams->Color = outBackGroundcolor; ++ vpHalRenderParams->pColorFillParams->bYCbCr = false; ++ vpHalRenderParams->pColorFillParams->CSpace = CSpace_sRGB; ++ + } + else + { +-- +2.40.1 + diff --git a/recipes-multimedia/libva/files/0011-Disable-422H-format-output.patch b/recipes-multimedia/libva/files/0011-Disable-422H-format-output.patch new file mode 100644 index 00000000..2f4aec1e --- /dev/null +++ b/recipes-multimedia/libva/files/0011-Disable-422H-format-output.patch @@ -0,0 +1,33 @@ +From f8b4d982232d4c1330e1e5f5dcd7cd8ca94decba Mon Sep 17 00:00:00 2001 +From: WenshengZhang <wensheng.zhang@intel.com> +Date: Thu, 25 Jan 2024 16:35:10 +0800 +Subject: [PATCH 11/12] Disable 422H format output + +Disable 422H format output. + +Upstream-Status: Backport [https://github.com/intel/media-driver/commit/eab411768e61a46e096793291f5f992eecf76cd8] +Signed-off-by: Lim, Siew Hoon <siew.hoon.lim@intel.com> +--- + .../agnostic/common/vp/hal/feature_manager/policy.cpp | 6 ++++++ + 1 file changed, 6 insertions(+) + +diff --git a/media_softlet/agnostic/common/vp/hal/feature_manager/policy.cpp b/media_softlet/agnostic/common/vp/hal/feature_manager/policy.cpp +index 317cec25a..12f1124fa 100644 +--- a/media_softlet/agnostic/common/vp/hal/feature_manager/policy.cpp ++++ b/media_softlet/agnostic/common/vp/hal/feature_manager/policy.cpp +@@ -1140,6 +1140,12 @@ MOS_STATUS Policy::GetScalingExecutionCaps(SwFilter *feature, bool isHdrEnabled) + bool isAlphaSettingSupportedByVebox = + IsAlphaSettingSupportedByVebox(scalingParams->formatInput, scalingParams->formatOutput, scalingParams->pCompAlpha); + ++ if (scalingParams->formatOutput == Format_422H) ++ { ++ VP_PUBLIC_ASSERTMESSAGE("Scaling not support 422H format output."); ++ return MOS_STATUS_UNIMPLEMENTED; ++ } ++ + // Clean usedForNextPass flag. + if (scalingEngine->usedForNextPass) + { +-- +2.40.1 + diff --git a/recipes-multimedia/libva/files/0012-Decode-Fix-AVC-decode-SFC-4K-hang-issue.patch b/recipes-multimedia/libva/files/0012-Decode-Fix-AVC-decode-SFC-4K-hang-issue.patch new file mode 100644 index 00000000..1be9139d --- /dev/null +++ b/recipes-multimedia/libva/files/0012-Decode-Fix-AVC-decode-SFC-4K-hang-issue.patch @@ -0,0 +1,32 @@ +From b93904533d3ee66c7919fa8e23d26b852a37a917 Mon Sep 17 00:00:00 2001 +From: Lim Siew Hoon <siew.hoon.lim@intel.com> +Date: Tue, 30 Jan 2024 13:27:52 +0800 +Subject: [PATCH 12/12] [Decode] Fix AVC decode SFC 4K hang issue + +Updated VDSFC input width and height for AVC decode + +Upstream-Status: Submitted [https://github.com/intel-innersource/drivers.gpu.unified/pull/151401] + +Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com> +--- + .../hal/dec/avc/packet/decode_avc_downsampling_packet.cpp | 4 ++++ + 1 file changed, 4 insertions(+) + +diff --git a/media_softlet/agnostic/common/codec/hal/dec/avc/packet/decode_avc_downsampling_packet.cpp b/media_softlet/agnostic/common/codec/hal/dec/avc/packet/decode_avc_downsampling_packet.cpp +index e75fd8137..9cc236bd7 100644 +--- a/media_softlet/agnostic/common/codec/hal/dec/avc/packet/decode_avc_downsampling_packet.cpp ++++ b/media_softlet/agnostic/common/codec/hal/dec/avc/packet/decode_avc_downsampling_packet.cpp +@@ -59,6 +59,10 @@ MOS_STATUS AvcDownSamplingPkt::InitSfcParams(VDBOX_SFC_PARAMS &sfcParams) + + AvcBasicFeature *avcBasicFeature = dynamic_cast<AvcBasicFeature*>(m_basicFeature); + DECODE_CHK_NULL(avcBasicFeature); ++ ++ sfcParams.input.width = avcBasicFeature->m_width; ++ sfcParams.input.height = avcBasicFeature->m_height; ++ + CODEC_PICTURE curPic = avcBasicFeature->m_avcPicParams->CurrPic; + + if (avcBasicFeature->m_avcPicParams->seq_fields.mb_adaptive_frame_field_flag == true) +-- +2.40.1 + diff --git a/recipes-multimedia/libva/intel-media-driver_21.1.3.bb b/recipes-multimedia/libva/intel-media-driver_23.4.3.bb index 637d8f0a..6513d574 100644 --- a/recipes-multimedia/libva/intel-media-driver_21.1.3.bb +++ b/recipes-multimedia/libva/intel-media-driver_23.4.3.bb @@ -19,19 +19,30 @@ REQUIRED_DISTRO_FEATURES = "opengl" DEPENDS += "libva gmmlib" SRC_URI = "git://github.com/intel/media-driver.git;protocol=https;nobranch=1 \ + file://0001-Disable-vp9-padding-on-mtl.patch \ + file://0002-Force-ARGB-surface-to-tile4-for-ACM.patch \ + file://0004-Add-device-ID-for-ARL.patch \ + file://0005-Add-XR24-support-to-DMABuf.patch \ + file://0006-add-INTEL-MEDIA-ALLOC-refineE-to-specify-the-memory-.patch \ + file://0007-Skip-report-keys.patch \ + file://0008-Limit-INTEL-MEDIA-ALLOC-MODE-to-MTL-and-ARL-only.patch \ + file://0009-Skip-cache-bucket-realloc-for-default-mode-0.patch \ + file://0010-Fix-failed-4k-video-wall-test-case-and-color-corrupt.patch \ + file://0011-Disable-422H-format-output.patch \ + file://0012-Decode-Fix-AVC-decode-SFC-4K-hang-issue.patch \ " -SRCREV = "bec8e1382637589bc3f302edf6c866bec3d754dd" +SRCREV = "a9f272496fc0f3e823cc4a814666ea27b443d902" S = "${WORKDIR}/git" -COMPATIBLE_HOST_x86-x32 = "null" +COMPATIBLE_HOST:x86-x32 = "null" UPSTREAM_CHECK_GITTAGREGEX = "^intel-media-(?P<pver>(?!600\..*)\d+(\.\d+)+)$" inherit cmake pkgconfig -MEDIA_DRIVER_ARCH_x86 = "32" -MEDIA_DRIVER_ARCH_x86-64 = "64" +MEDIA_DRIVER_ARCH:x86 = "32" +MEDIA_DRIVER_ARCH:x86-64 = "64" EXTRA_OECMAKE += " \ -DMEDIA_RUN_TEST_SUITE=OFF \ @@ -39,12 +50,12 @@ EXTRA_OECMAKE += " \ -DMEDIA_BUILD_FATAL_WARNINGS=OFF \ " -CXXFLAGS_append_x86 = " -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE" +CXXFLAGS:append:x86 = " -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE" -do_configure_prepend_toolchain-clang() { +do_configure:prepend:toolchain-clang() { sed -i -e '/-fno-tree-pre/d' ${S}/media_driver/cmake/linux/media_compile_flags_linux.cmake } -FILES_${PN} += " \ +FILES:${PN} += " \ ${libdir}/dri/ \ " diff --git a/recipes-multimedia/libva/intel-vaapi-driver_2.4.1.bb b/recipes-multimedia/libva/intel-vaapi-driver_2.4.1.bb index ccb9fec1..5038d8d7 100644 --- a/recipes-multimedia/libva/intel-vaapi-driver_2.4.1.bb +++ b/recipes-multimedia/libva/intel-vaapi-driver_2.4.1.bb @@ -29,6 +29,6 @@ PACKAGECONFIG ??= "${@bb.utils.contains("DISTRO_FEATURES", "x11", "x11", "", d)} PACKAGECONFIG[x11] = "-Dwith_x11=yes, -Dwith_x11=no" PACKAGECONFIG[wayland] = "-Dwith_wayland=yes, -Dwith_wayland=no, wayland wayland-native virtual/egl" -FILES_${PN} += "${libdir}/dri/*.so" -FILES_${PN}-dev += "${libdir}/dri/*.la" -FILES_${PN}-dbg += "${libdir}/dri/.debug" +FILES:${PN} += "${libdir}/dri/*.so" +FILES:${PN}-dev += "${libdir}/dri/*.la" +FILES:${PN}-dbg += "${libdir}/dri/.debug" diff --git a/recipes-multimedia/libyami/libyami-utils/0001-Fix-build-with-clang.patch b/recipes-multimedia/libyami/libyami-utils/0001-Fix-build-with-clang.patch deleted file mode 100644 index f7de9d16..00000000 --- a/recipes-multimedia/libyami/libyami-utils/0001-Fix-build-with-clang.patch +++ /dev/null @@ -1,96 +0,0 @@ -From f2e6d2ecfea635ab952649156e31ca893d4b1a47 Mon Sep 17 00:00:00 2001 -From: Khem Raj <raj.khem@gmail.com> -Date: Sat, 12 Aug 2017 08:49:20 -0700 -Subject: [PATCH] Fix build with clang - -Fix errors e.g. -error: comparison of constant -1 with expression of type 'char' is always true [-Werror,-Wtautological-constant-out-of-range-compare]error: comparison of constant -1 with expression of type 'char' is always true [-Werror,-Wtautological-constant-out-of-range-compare] - -and - -psnr.cpp:225:17: error: bool literal returned from 'main' [-Werror,-Wmain] - -Signed-off-by: Khem Raj <raj.khem@gmail.com> -Upstream-Status: Pending - ---- - tests/decodehelp.cpp | 2 +- - tests/encodehelp.h | 2 +- - tests/vpp.cpp | 2 +- - tests/yamitranscode.cpp | 2 +- - testscripts/psnr.cpp | 4 ++-- - 5 files changed, 6 insertions(+), 6 deletions(-) - -diff --git a/tests/decodehelp.cpp b/tests/decodehelp.cpp -index b27b977..62cca63 100644 ---- a/tests/decodehelp.cpp -+++ b/tests/decodehelp.cpp -@@ -81,7 +81,7 @@ bool processCmdLine(int argc, char** argv, DecodeParameter* parameters) - { NULL, no_argument, NULL, 0 } - }; - -- char opt; -+ int opt; - while ((opt = getopt_long_only(argc, argv, "h:m:n:i:f:o:w:?", long_opts,&option_index)) != -1){ - switch (opt) { - case 'h': -diff --git a/tests/encodehelp.h b/tests/encodehelp.h -index 63580a4..651ec08 100644 ---- a/tests/encodehelp.h -+++ b/tests/encodehelp.h -@@ -109,7 +109,7 @@ static VideoRateControl string_to_rc_mode(char *str) - - static bool process_cmdline(int argc, char *argv[]) - { -- char opt; -+ int opt; - const struct option long_opts[] = { - { "help", no_argument, NULL, 'h' }, - { "qp", required_argument, NULL, 0 }, -diff --git a/tests/vpp.cpp b/tests/vpp.cpp -index 5a60c0a..52da43b 100644 ---- a/tests/vpp.cpp -+++ b/tests/vpp.cpp -@@ -151,7 +151,7 @@ public: - private: - bool processCmdLine(int argc, char* argv[]) - { -- char opt; -+ int opt; - const struct option long_opts[] = { - { "help", no_argument, NULL, 'h' }, - { "sharpening", required_argument, NULL, 's' }, -diff --git a/tests/yamitranscode.cpp b/tests/yamitranscode.cpp -index 6207209..00164ce 100755 ---- a/tests/yamitranscode.cpp -+++ b/tests/yamitranscode.cpp -@@ -100,7 +100,7 @@ static VideoRateControl string_to_rc_mode(char *str) - - static bool processCmdLine(int argc, char *argv[], TranscodeParams& para) - { -- char opt; -+ int opt; - const struct option long_opts[] = { - { "help", no_argument, NULL, 'h' }, - { "qp", required_argument, NULL, 0 }, -diff --git a/testscripts/psnr.cpp b/testscripts/psnr.cpp -index 5cc24c9..68bd668 100644 ---- a/testscripts/psnr.cpp -+++ b/testscripts/psnr.cpp -@@ -215,14 +215,14 @@ int main(int argc, char *argv[]) - const char* psnrresult = "average_psnr.txt"; - int width=0,height=0; - int standardpsnr = NORMAL_PSNR; -- char opt; -+ int opt; - while ((opt = getopt(argc, argv, "h:W:H:i:o:s:?")) != -1) - { - switch (opt) { - case 'h': - case '?': - print_help(argv[0]); -- return false; -+ return -1; - case 'i': - filename1 = optarg; - break; diff --git a/recipes-multimedia/libyami/libyami-utils_1.3.2.bb b/recipes-multimedia/libyami/libyami-utils_1.3.2.bb deleted file mode 100644 index f9333f3d..00000000 --- a/recipes-multimedia/libyami/libyami-utils_1.3.2.bb +++ /dev/null @@ -1,26 +0,0 @@ -SUMMARY = "Applications and Scripts for libyami." -DESCRIPTION = "Applications and Scripts for libyami." - -HOMEPAGE = "https://github.com/intel/libyami-utils" -BUGTRACKER = "https://github.com/intel/libyami-utils/issues/new" - -LICENSE = "Apache-2.0" -LIC_FILES_CHKSUM = "file://LICENSE;md5=e3fc50a88d0a364313df4b21ef20c29e" - -SRC_URI = "git://github.com/intel/libyami-utils.git \ - file://0001-Fix-build-with-clang.patch \ - " -SRCREV = "c3d25b64b05aeb0c4eecc140aef617cfeced6b8e" -S = "${WORKDIR}/git" - -DEPENDS = "libva libyami" - -inherit autotools pkgconfig features_check - -REQUIRED_DISTRO_FEATURES = "opengl" - -PACKAGECONFIG = "${@bb.utils.filter('DISTRO_FEATURES', 'x11', d)}" - -# --enable-x11 needs libva-x11 -# gles-tests fail to build without x11: see https://github.com/intel/libyami-utils/issues/91 -PACKAGECONFIG[x11] = "--enable-x11 --enable-egl,--disable-x11 --disable-egl, virtual/libx11" diff --git a/recipes-multimedia/libyami/libyami_1.3.2.bb b/recipes-multimedia/libyami/libyami_1.3.2.bb deleted file mode 100644 index c8628141..00000000 --- a/recipes-multimedia/libyami/libyami_1.3.2.bb +++ /dev/null @@ -1,24 +0,0 @@ -SUMMARY = "Yami is media infrastructure base on libva" -DESCRIPTION = "Yet Another Media Infrastructure \ -light weight hardware codec library base on VA-API " - -HOMEPAGE = "https://github.com/intel/libyami" -BUGTRACKER = "https://github.com/intel/libyami/issues/new" - -LICENSE = "Apache-2.0" -LIC_FILES_CHKSUM = "file://LICENSE.md;md5=3b83ef96387f14655fc854ddc3c6bd57" - -SRC_URI = "git://github.com/intel/libyami.git;branch=apache \ -" -SRCREV = "08606d0a43e0ef15e5b61cc13563169370ce8715" -S = "${WORKDIR}/git" - -CXXFLAGS_append = " -Wno-error" - -PACKAGECONFIG ??= "${@bb.utils.contains("DISTRO_FEATURES", "x11", "x11", "", d)}" -PACKAGECONFIG[x11] = "--enable-x11,--disable-x11,virtual/libx11 libxrandr libxrender" - -DEPENDS = "libva" -inherit autotools pkgconfig features_check - -REQUIRED_DISTRO_FEATURES = "opengl" diff --git a/recipes-multimedia/mediasdk/files/0001-Don-t-try-to-autheticate-with-render-node.patch b/recipes-multimedia/mediasdk/files/0001-Don-t-try-to-autheticate-with-render-node.patch deleted file mode 100644 index d424b68d..00000000 --- a/recipes-multimedia/mediasdk/files/0001-Don-t-try-to-autheticate-with-render-node.patch +++ /dev/null @@ -1,44 +0,0 @@ -From 6945cd0ad84855ab4a2dc40016879fb5fc7f2b6a Mon Sep 17 00:00:00 2001 -From: Vincent Cheah Beng Keat <vincent.beng.keat.cheah@intel.com> -Date: Wed, 7 Apr 2021 05:30:15 +0000 -Subject: [PATCH] Don't try to autheticate with render node - -Replicate similar libVA wayland change. -https://github.com/intel/libva/commit/283f776a9649dcef58b47958c1269499adfa1cd4 - -Platform: ADL-S, EHL, TGL -OS: Yocto native wayland weston -Tested: - ./sample_decode h265 -i Puppies_3840x2160.265 -rwld -rgb4 - -Issue: #2592 - -Upstream-Status: Backport [https://github.com/Intel-Media-SDK/MediaSDK/commit/6945cd0ad84855ab4a2dc40016879fb5fc7f2b6a] - -Signed-off-by: Cheah, Vincent Beng Keat <vincent.beng.keat.cheah@intel.com> ---- - samples/sample_misc/wayland/src/class_wayland.cpp | 8 ++++++-- - 1 file changed, 6 insertions(+), 2 deletions(-) - -diff --git a/samples/sample_misc/wayland/src/class_wayland.cpp b/samples/sample_misc/wayland/src/class_wayland.cpp -index 62f326e3..4a0beed3 100644 ---- a/samples/sample_misc/wayland/src/class_wayland.cpp -+++ b/samples/sample_misc/wayland/src/class_wayland.cpp -@@ -433,8 +433,12 @@ void Wayland::DrmHandleDevice(const char *name) - m_device_name << "\n"; - return; - } -- drmGetMagic(m_fd, &magic); -- wl_drm_authenticate(m_drm, magic); -+ -+ int type = drmGetNodeTypeFromFd(m_fd); -+ if (type != DRM_NODE_RENDER) { -+ drmGetMagic(m_fd, &magic); -+ wl_drm_authenticate(m_drm, magic); -+ } - } - - void Wayland::DrmHandleAuthenticated() --- -2.24.3 - diff --git a/recipes-multimedia/mediasdk/files/fix-gcc13.patch b/recipes-multimedia/mediasdk/files/fix-gcc13.patch new file mode 100644 index 00000000..fb973870 --- /dev/null +++ b/recipes-multimedia/mediasdk/files/fix-gcc13.patch @@ -0,0 +1,15 @@ +Upstream-Status: Inactive-Upstream +Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> + +diff --git a/api/mfx_dispatch/linux/mfxparser.cpp b/api/mfx_dispatch/linux/mfxparser.cpp +index 9d3823ec3e..12e46d1881 100644 +--- a/api/mfx_dispatch/linux/mfxparser.cpp ++++ b/api/mfx_dispatch/linux/mfxparser.cpp +@@ -20,6 +20,7 @@ + + #include <ctype.h> + #include <stdio.h> ++#include <stdint.h> + #include <stdlib.h> + #include <string.h> + diff --git a/recipes-multimedia/mediasdk/intel-mediasdk_21.1.3.bb b/recipes-multimedia/mediasdk/intel-mediasdk_23.2.2.bb index b3a30bc4..f00a4404 100644 --- a/recipes-multimedia/mediasdk/intel-mediasdk_21.1.3.bb +++ b/recipes-multimedia/mediasdk/intel-mediasdk_23.2.2.bb @@ -12,12 +12,14 @@ CVE_DETAILS = "intel:media_sdk" # Only for 64 bit until media-driver issues aren't fixed COMPATIBLE_HOST = '(x86_64).*-linux' -COMPATIBLE_HOST_x86-x32 = "null" +COMPATIBLE_HOST:x86-x32 = "null" inherit features_check REQUIRED_DISTRO_FEATURES = "opengl" -DEPENDS += "libdrm libva intel-media-driver" +DEPENDS += "libva" + +RDEPENDS:${PN} += "intel-media-driver" PACKAGECONFIG ??= "${@bb.utils.contains("DISTRO_FEATURES", "x11", "dri3", "", d)} \ ${@bb.utils.contains("DISTRO_FEATURES", "wayland", "wayland", "", d)} \ @@ -27,16 +29,16 @@ PACKAGECONFIG ??= "${@bb.utils.contains("DISTRO_FEATURES", "x11", "dri3", "", d) PACKAGECONFIG[dri3] = "-DENABLE_X11_DRI3=ON, -DENABLE_X11_DRI3=OFF" PACKAGECONFIG[itt] = "-DENABLE_ITT=ON, -DENABLE_ITT=OFF, itt" -PACKAGECONFIG[opencl] = "-DENABLE_OPENCL=ON, -DENABLE_OPENCL=OFF, ocl-icd opencl-clhpp opencl-headers" +PACKAGECONFIG[opencl] = "-DENABLE_OPENCL=ON, -DENABLE_OPENCL=OFF, virtual/opencl-icd opencl-clhpp opencl-headers" PACKAGECONFIG[samples] = "-DBUILD_SAMPLES=ON, -DBUILD_SAMPLES=OFF" PACKAGECONFIG[wayland] = "-DENABLE_WAYLAND=ON, -DENABLE_WAYLAND=OFF, wayland wayland-native" SRC_URI = "git://github.com/Intel-Media-SDK/MediaSDK.git;protocol=https;nobranch=1;lfs=0 \ file://0001-FindITT.cmake-fix-detection-of-header-library.patch \ - file://0001-Don-t-try-to-autheticate-with-render-node.patch \ + file://fix-gcc13.patch \ " -SRCREV = "03164a0b491651c210abd4549ca5275ac974d105" +SRCREV = "869b60a6c3d7b5e9f7c3b3b914986322dca4bbae" S = "${WORKDIR}/git" UPSTREAM_CHECK_GITTAGREGEX = "^intel-mediasdk-(?P<pver>(\d+(\.\d+)+))$" @@ -45,17 +47,17 @@ inherit cmake pkgconfig EXTRA_OECMAKE += "-DMFX_INCLUDE=${S}/api/include" -do_install_append() { +do_install:append() { mv ${D}${datadir}/mfx/samples ${D}${libdir}/mfx/samples } PACKAGE_BEFORE_PN = " ${PN}-samples" -FILES_${PN} += " \ +FILES:${PN} += " \ ${libdir}/mfx \ ${datadir}/mfx/plugins.cfg \ " -FILES_${PN}-samples = "${libdir}/mfx/samples" +FILES:${PN}-samples = "${libdir}/mfx/samples" -INSANE_SKIP_${PN}-samples += "staticdev" +INSANE_SKIP:${PN}-samples += "staticdev" diff --git a/recipes-multimedia/onevpl/files/0001-ARLH-DID-open-source-6286.patch b/recipes-multimedia/onevpl/files/0001-ARLH-DID-open-source-6286.patch new file mode 100644 index 00000000..8d1f3380 --- /dev/null +++ b/recipes-multimedia/onevpl/files/0001-ARLH-DID-open-source-6286.patch @@ -0,0 +1,39 @@ +From a425927469b36e7bc597bff4fdc24d42a1168d1a Mon Sep 17 00:00:00 2001 +From: gfxVPLsdm <gfxvplsdm@intel.com> +Date: Mon, 18 Dec 2023 09:58:35 +0800 +Subject: [PATCH] ARLH DID open source (#6286) + +Co-authored-by: huangli <li.h.huang@intel.com> + +Upstream-Status: Backport [https://github.com/oneapi-src/oneVPL-intel-gpu/commit/66f1b4d60a367eb1b0774e0b4369f906e0e632ae] +Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com> + +--- + _studio/shared/include/mfxstructures-int.h | 7 ++++++- + 1 file changed, 6 insertions(+), 1 deletion(-) + +diff --git a/_studio/shared/include/mfxstructures-int.h b/_studio/shared/include/mfxstructures-int.h +index 363d9a63..8f16b0a8 100644 +--- a/_studio/shared/include/mfxstructures-int.h ++++ b/_studio/shared/include/mfxstructures-int.h +@@ -1,4 +1,4 @@ +-// Copyright (c) 2007-2021 Intel Corporation ++// Copyright (c) 2007-2023 Intel Corporation + // + // Permission is hereby granted, free of charge, to any person obtaining a copy + // of this software and associated documentation files (the "Software"), to deal +@@ -569,6 +569,11 @@ typedef struct { + /* ARL S */ + { 0x7D67, MFX_HW_ARL, MFX_GT2 }, + ++ /* ARL H*/ ++ { 0x7D51, MFX_HW_ARL, MFX_GT2 }, ++ { 0x7DD1, MFX_HW_ARL, MFX_GT2 }, ++ { 0x7D41, MFX_HW_ARL, MFX_GT2 }, ++ + }; + + /* +-- +2.40.1 + diff --git a/recipes-multimedia/onevpl/files/0001-Fix-SetBuffersYV12-V-U-plane-offsets-calculation-647.patch b/recipes-multimedia/onevpl/files/0001-Fix-SetBuffersYV12-V-U-plane-offsets-calculation-647.patch new file mode 100644 index 00000000..8e014c97 --- /dev/null +++ b/recipes-multimedia/onevpl/files/0001-Fix-SetBuffersYV12-V-U-plane-offsets-calculation-647.patch @@ -0,0 +1,32 @@ +From e118062518c95d8fb0080cc8276b123a90bd0eb9 Mon Sep 17 00:00:00 2001 +From: gfxVPLsdm <gfxvplsdm@intel.com> +Date: Wed, 21 Feb 2024 17:11:52 +0800 +Subject: [PATCH] Fix SetBuffersYV12() V & U plane offsets calculation (#6475) + +Signed-off-by: Hoe, Sheng Yang <sheng.yang.hoe@intel.com> +Co-authored-by: Hoe, Sheng Yang <sheng.yang.hoe@intel.com> + +Upstream-Status: Backport [https://github.com/oneapi-src/oneVPL-intel-gpu/commit/ce704ddfe11724767ba86c9425a3917dc3100d4a] +Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com> +--- + _studio/shared/src/libmfx_core_vaapi.cpp | 4 ++-- + 1 file changed, 2 insertions(+), 2 deletions(-) + +diff --git a/_studio/shared/src/libmfx_core_vaapi.cpp b/_studio/shared/src/libmfx_core_vaapi.cpp +index 2cad86f8..fd5fc269 100644 +--- a/_studio/shared/src/libmfx_core_vaapi.cpp ++++ b/_studio/shared/src/libmfx_core_vaapi.cpp +@@ -462,8 +462,8 @@ public: + } + else + { +- eb.offsets[1] = eb.offsets[0] + uint32_t(eb.height * eb.pitches[1] / 2); +- eb.offsets[2] = eb.offsets[1] + uint32_t(eb.height * eb.pitches[2] / 2); ++ eb.offsets[1] = eb.offsets[0] + uint32_t(eb.height * eb.pitches[0]); ++ eb.offsets[2] = eb.offsets[1] + uint32_t(eb.height * eb.pitches[1] / 2); + } + eb.num_planes = 3; + eb.data_size = eb.offsets[2] + (eb.height * eb.pitches[2] / 2); +-- +2.40.1 + diff --git a/recipes-multimedia/onevpl/files/0001-JPEGe-Enable-BGR4-JPEG-Enc-support-6470.patch b/recipes-multimedia/onevpl/files/0001-JPEGe-Enable-BGR4-JPEG-Enc-support-6470.patch new file mode 100644 index 00000000..e3854407 --- /dev/null +++ b/recipes-multimedia/onevpl/files/0001-JPEGe-Enable-BGR4-JPEG-Enc-support-6470.patch @@ -0,0 +1,96 @@ +From 40c0bc6d3dfe172dd25908df2c16de29c34fa4d4 Mon Sep 17 00:00:00 2001 +From: gfxVPLsdm <gfxvplsdm@intel.com> +Date: Mon, 5 Feb 2024 17:14:07 +0800 +Subject: [PATCH] [JPEGe] Enable BGR4 JPEG Enc support (#6470) + +Co-authored-by: vcheah <vincent.beng.keat.cheah@intel.com> + +Upstream-Status: Backport [https://github.com/oneapi-src/oneVPL-intel-gpu/commit/e73763d05f8fdc348c356b58dcb83075b5e868f8] +Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com> +--- + .../mfx_lib/encode_hw/mjpeg/src/mfx_mjpeg_encode_hw.cpp | 6 ++++-- + .../encode_hw/mjpeg/src/mfx_mjpeg_encode_hw_utils.cpp | 9 +++++---- + 2 files changed, 9 insertions(+), 6 deletions(-) + +diff --git a/_studio/mfx_lib/encode_hw/mjpeg/src/mfx_mjpeg_encode_hw.cpp b/_studio/mfx_lib/encode_hw/mjpeg/src/mfx_mjpeg_encode_hw.cpp +index 141564b1..332f6372 100644 +--- a/_studio/mfx_lib/encode_hw/mjpeg/src/mfx_mjpeg_encode_hw.cpp ++++ b/_studio/mfx_lib/encode_hw/mjpeg/src/mfx_mjpeg_encode_hw.cpp +@@ -212,7 +212,8 @@ mfxStatus MFXVideoENCODEMJPEG_HW::QueryImplsDescription( + ah.PushBack(memCaps.ColorFormats) = MFX_FOURCC_YV12; + ah.PushBack(memCaps.ColorFormats) = MFX_FOURCC_YUY2; + ah.PushBack(memCaps.ColorFormats) = MFX_FOURCC_RGB4; +- memCaps.NumColorFormats = 4; ++ ah.PushBack(memCaps.ColorFormats) = MFX_FOURCC_BGR4; ++ memCaps.NumColorFormats = 5; + + ah.PushBack(profileCaps.MemDesc); + profileCaps.MemDesc[1] = profileCaps.MemDesc[0]; +@@ -392,7 +393,7 @@ mfxStatus MFXVideoENCODEMJPEG_HW::Query(VideoCORE * core, mfxVideoParam *in, mfx + if ((fourCC == 0 && chromaFormat == 0) || + (fourCC == MFX_FOURCC_NV12 && (chromaFormat == MFX_CHROMAFORMAT_YUV420 || chromaFormat == MFX_CHROMAFORMAT_YUV400)) || + (fourCC == MFX_FOURCC_YUY2 && chromaFormat == MFX_CHROMAFORMAT_YUV422H) || +- (fourCC == MFX_FOURCC_RGB4 && chromaFormat == MFX_CHROMAFORMAT_YUV444)) ++ ((fourCC == MFX_FOURCC_RGB4 || fourCC == MFX_FOURCC_BGR4) && chromaFormat == MFX_CHROMAFORMAT_YUV444)) + { + out->mfx.FrameInfo.FourCC = in->mfx.FrameInfo.FourCC; + out->mfx.FrameInfo.ChromaFormat = in->mfx.FrameInfo.ChromaFormat; +@@ -698,6 +699,7 @@ mfxStatus MFXVideoENCODEMJPEG_HW::Init(mfxVideoParam *par) + doubleBytesPerPx = 4; + break; + case MFX_FOURCC_RGB4: ++ case MFX_FOURCC_BGR4: + default: + doubleBytesPerPx = 8; + break; +diff --git a/_studio/mfx_lib/encode_hw/mjpeg/src/mfx_mjpeg_encode_hw_utils.cpp b/_studio/mfx_lib/encode_hw/mjpeg/src/mfx_mjpeg_encode_hw_utils.cpp +index 3fc619d1..b1ab556e 100644 +--- a/_studio/mfx_lib/encode_hw/mjpeg/src/mfx_mjpeg_encode_hw_utils.cpp ++++ b/_studio/mfx_lib/encode_hw/mjpeg/src/mfx_mjpeg_encode_hw_utils.cpp +@@ -108,6 +108,7 @@ mfxStatus MfxHwMJpegEncode::CheckJpegParam(VideoCORE *core, mfxVideoParam & par, + BytesPerPx = 2; + break; + case MFX_FOURCC_RGB4: ++ case MFX_FOURCC_BGR4: + default: + BytesPerPx = 4; + } +@@ -157,7 +158,7 @@ mfxStatus ExecuteBuffers::Init(mfxVideoParam const *par, mfxEncodeCtrl const * c + + m_payload_base.length = 0; + m_payload_list.clear(); +- if (fourCC == MFX_FOURCC_RGB4 && chromaFormat == MFX_CHROMAFORMAT_YUV444) ++ if ((fourCC == MFX_FOURCC_RGB4 || fourCC == MFX_FOURCC_BGR4) && chromaFormat == MFX_CHROMAFORMAT_YUV444) + { + m_app14_data.header = 0xEEFF;//APP14 + m_app14_data.lenH = 0; +@@ -287,7 +288,7 @@ mfxStatus ExecuteBuffers::Init(mfxVideoParam const *par, mfxEncodeCtrl const * c + m_pps.num_components = 3; + else if (fourCC == MFX_FOURCC_NV12 && chromaFormat == MFX_CHROMAFORMAT_YUV400) + m_pps.num_components = 1; +- else if (fourCC == MFX_FOURCC_RGB4 && chromaFormat == MFX_CHROMAFORMAT_YUV444) ++ else if ((fourCC == MFX_FOURCC_RGB4 || fourCC == MFX_FOURCC_BGR4) && chromaFormat == MFX_CHROMAFORMAT_YUV444) + m_pps.num_components = 3; + else + MFX_RETURN(MFX_ERR_UNDEFINED_BEHAVIOR); +@@ -341,7 +342,7 @@ mfxStatus ExecuteBuffers::Init(mfxVideoParam const *par, mfxEncodeCtrl const * c + { + // No external tables - use Quality parameter + m_dqt_list.resize(0); +- if (fourCC == MFX_FOURCC_RGB4) ++ if (fourCC == MFX_FOURCC_RGB4 || fourCC == MFX_FOURCC_BGR4) + { + m_pps.quantiser_table_selector[0] = 0; + m_pps.quantiser_table_selector[1] = 0; +@@ -393,7 +394,7 @@ mfxStatus ExecuteBuffers::Init(mfxVideoParam const *par, mfxEncodeCtrl const * c + { + m_dht_list.resize(0); + } +- else if (hwCaps->MaxNumHuffTable == 1 || fourCC == MFX_FOURCC_RGB4) ++ else if (hwCaps->MaxNumHuffTable == 1 || (fourCC == MFX_FOURCC_RGB4 || fourCC == MFX_FOURCC_BGR4)) + { + m_dht_list.resize(1); + +-- +2.40.1 + diff --git a/recipes-multimedia/onevpl/files/0001-vpl.pc.in-dont-pass-pcfiledir-to-cflags.patch b/recipes-multimedia/onevpl/files/0001-vpl.pc.in-dont-pass-pcfiledir-to-cflags.patch new file mode 100644 index 00000000..762a4902 --- /dev/null +++ b/recipes-multimedia/onevpl/files/0001-vpl.pc.in-dont-pass-pcfiledir-to-cflags.patch @@ -0,0 +1,28 @@ +From 630c32194f53c70f4f654fb3d198790df253ba1f Mon Sep 17 00:00:00 2001 +From: Markus Volk <f_l_k@t-online.de> +Date: Thu, 15 Jun 2023 13:43:32 +0200 +Subject: [PATCH] vpl.pc.in: dont pass pcfiledir to cflags + +Signed-off-by: Markus Volk <f_l_k@t-online.de> + +Upstream-Status: Inappropriate [oe specific] +--- + libvpl/pkgconfig/vpl.pc.in | 4 ++-- + 1 file changed, 2 insertions(+), 2 deletions(-) + +diff --git a/libvpl/pkgconfig/vpl.pc.in b/libvpl/pkgconfig/vpl.pc.in +index ab468a9..05c020e 100644 +--- a/libvpl/pkgconfig/vpl.pc.in ++++ b/libvpl/pkgconfig/vpl.pc.in +@@ -8,6 +8,6 @@ Description: oneAPI Video Processing Library + Version: @API_VERSION_MAJOR@.@API_VERSION_MINOR@ + URL: https://software.intel.com/content/www/us/en/develop/tools/oneapi/components/onevpl.html + +-Libs: -L${libdir} -l@OUTPUT_NAME@ @VPL_PKGCONFIG_DEPENDENT_LIBS@ ++Libs: -L@CMAKE_INSTALL_PREFIX@/@CMAKE_INSTALL_LIBDIR@ -l@OUTPUT_NAME@ @VPL_PKGCONFIG_DEPENDENT_LIBS@ + Libs.private: @VPL_PKGCONFIG_PRIVATE_LIBS@ +-Cflags: -I${includedir} -I${includedir}/vpl ++Cflags: -I@CMAKE_INSTALL_PREFIX@/@CMAKE_INSTALL_INCLUDEDIR@ -I@CMAKE_INSTALL_PREFIX@/@CMAKE_INSTALL_INCLUDEDIR@/vpl +-- +2.40.1 + diff --git a/recipes-multimedia/onevpl/onevpl-intel-gpu_23.4.3.bb b/recipes-multimedia/onevpl/onevpl-intel-gpu_23.4.3.bb new file mode 100644 index 00000000..0b849bf1 --- /dev/null +++ b/recipes-multimedia/onevpl/onevpl-intel-gpu_23.4.3.bb @@ -0,0 +1,33 @@ +SUMMARY = "Intel(R) oneVPL runtime for Intel GPU accelerated media processing" +DESCRIPTION = "Intel(R) oneVPL runtime provides an runtime to access hardware-accelerated \ +video decode, encode and filtering on Intel® graphics." + +HOMEPAGE = "https://github.com/oneapi-src/oneVPL-intel-gpu" +BUGTRACKER = "https://github.com/oneapi-src/oneVPL-intel-gpu/issues" + +LICENSE = "MIT" +LIC_FILES_CHKSUM = "file://LICENSE;md5=eb8cb45b9b57dbaa9fcc9adc4230202b" + +PE = "1" + +# Only for 64 bit +COMPATIBLE_HOST = '(x86_64).*-linux' +COMPATIBLE_HOST:x86-x32 = "null" + +DEPENDS += "libdrm libva intel-media-driver onevpl pkgconfig-native" +RDEPENDS:${PN} += "intel-media-driver" + +SRC_URI = "git://github.com/oneapi-src/oneVPL-intel-gpu.git;protocol=https;nobranch=1;lfs=0 \ + file://0001-ARLH-DID-open-source-6286.patch \ + file://0001-JPEGe-Enable-BGR4-JPEG-Enc-support-6470.patch \ + file://0001-Fix-SetBuffersYV12-V-U-plane-offsets-calculation-647.patch \ + " + +SRCREV = "852fa9f705ef44c004d014548601f3804a6de705" +S = "${WORKDIR}/git" + +FILES:${PN} += " \ + ${libdir}/libmfx-gen/enctools.so \ + " + +inherit cmake diff --git a/recipes-multimedia/onevpl/onevpl_2023.4.0.bb b/recipes-multimedia/onevpl/onevpl_2023.4.0.bb new file mode 100644 index 00000000..513c7f2c --- /dev/null +++ b/recipes-multimedia/onevpl/onevpl_2023.4.0.bb @@ -0,0 +1,40 @@ +SUMMARY = "oneAPI Video Processing Library" +DESCRIPTION = "The oneAPI Video Processing Library (oneVPL) provides \ +a single video processing API for encode, decode, and video processing \ +that works across a wide range of accelerators." + +HOMEPAGE = "https://github.com/oneapi-src/oneVPL" +LICENSE = "MIT" +LIC_FILES_CHKSUM = "file://LICENSE;md5=c18ea6bb4786a26bf4eee88a7424a408 \ + file://third-party-programs.txt;md5=0e35a23482445dd089b4eabe19103a06" + +SRC_URI = "git://github.com/oneapi-src/oneVPL.git;protocol=https;branch=master \ + file://0001-vpl.pc.in-dont-pass-pcfiledir-to-cflags.patch \ + " +SRCREV = "2274efcd3672b43297ef774f332e1fed6781381c" +S = "${WORKDIR}/git" + +inherit cmake +DEPENDS += "libva pkgconfig-native" + +PACKAGECONFIG ??= "tools" +PACKAGECONFIG[tools] = "-DBUILD_TOOLS=ON, -DBUILD_TOOLS=OFF, wayland wayland-native wayland-protocols" + +do_install:append() { + mkdir -p ${D}${datadir}/oneVPL/samples + mv ${D}${bindir}/sample_* ${D}${datadir}/oneVPL/samples +} + +COMPATIBLE_HOST = '(x86_64).*-linux' + +PACKAGES =+ "${PN}-examples" + +FILES:${PN}-examples = "${datadir}/vpl \ + " + +FILES_SOLIBSDEV = "" +FILES:${PN}-dev += "${libdir}/libvpl.so" + +FILES:${PN} += " ${datadir}/oneVPL/samples \ + ${libdir}/vpl/libvpl_wayland.so \ + " diff --git a/recipes-oneapi/crypto/intel-crypto-mb/0001-CMakeLists.txt-exclude-host-system-headers.patch b/recipes-oneapi/crypto/intel-crypto-mb/0001-CMakeLists.txt-exclude-host-system-headers.patch new file mode 100644 index 00000000..58ed1c9a --- /dev/null +++ b/recipes-oneapi/crypto/intel-crypto-mb/0001-CMakeLists.txt-exclude-host-system-headers.patch @@ -0,0 +1,27 @@ +From efedbf9080c19241c2aa9ee7ba901245d38c8fa2 Mon Sep 17 00:00:00 2001 +From: Naveen Saini <naveen.kumar.saini@intel.com> +Date: Mon, 7 Mar 2022 16:44:24 +0800 +Subject: [PATCH 1/2] CMakeLists.txt: exclude host system headers + +Upstream-Status: Inappropriate + +Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> +--- + sources/ippcp/crypto_mb/CMakeLists.txt | 1 - + 1 file changed, 1 deletion(-) + +diff --git a/sources/ippcp/crypto_mb/CMakeLists.txt b/sources/ippcp/crypto_mb/CMakeLists.txt +index c4cc82a..4094f34 100644 +--- a/sources/ippcp/crypto_mb/CMakeLists.txt ++++ b/sources/ippcp/crypto_mb/CMakeLists.txt +@@ -82,7 +82,6 @@ include_directories( + ${CRYPTO_MB_INCLUDE_DIR} + ${OPENSSL_INCLUDE_DIR} + $<$<C_COMPILER_ID:Intel>:$ENV{ROOT}/compiler/include $ENV{ROOT}/compiler/include/icc> +- $<$<NOT:$<C_COMPILER_ID:Intel>>:${CMAKE_SYSTEM_INCLUDE_PATH}> + $<$<OR:$<C_COMPILER_ID:Intel>,$<BOOL:${MSVC_IDE}>>:$ENV{INCLUDE}> + ) + +-- +2.17.1 + diff --git a/recipes-oneapi/crypto/intel-crypto-mb/0001-crypto-mb-Make-sure-libs-are-installed-correctly.patch b/recipes-oneapi/crypto/intel-crypto-mb/0001-crypto-mb-Make-sure-libs-are-installed-correctly.patch new file mode 100644 index 00000000..c47c8de5 --- /dev/null +++ b/recipes-oneapi/crypto/intel-crypto-mb/0001-crypto-mb-Make-sure-libs-are-installed-correctly.patch @@ -0,0 +1,42 @@ +From 3b7b4eca54aa8d851e37f60c74e4a027fa7b21f2 Mon Sep 17 00:00:00 2001 +From: Anuj Mittal <anuj.mittal@intel.com> +Date: Wed, 14 Jun 2023 13:18:45 +0800 +Subject: [PATCH] crypto-mb: Make sure libs are installed correctly + +Dont assume that "lib" is always the correct destination. This fixes +multilib builds when libdir != /usr/lib. + +Upstream-Status: Pending + +Signed-off-by: Anuj Mittal <anuj.mittal@intel.com> +--- + sources/ippcp/crypto_mb/src/CMakeLists.txt | 6 +++--- + 1 file changed, 3 insertions(+), 3 deletions(-) + +diff --git a/sources/ippcp/crypto_mb/src/CMakeLists.txt b/sources/ippcp/crypto_mb/src/CMakeLists.txt +index c9cad06..31f52bd 100644 +--- a/sources/ippcp/crypto_mb/src/CMakeLists.txt ++++ b/sources/ippcp/crypto_mb/src/CMakeLists.txt +@@ -123,8 +123,8 @@ endif(DYNAMIC_LIB OR MB_STANDALONE) + # Installation of the shared library + if (MB_STANDALONE) # standalone crypto_mb's cmake run + install(TARGETS ${MB_DYN_LIB_TARGET} +- LIBRARY DESTINATION "lib" +- RUNTIME DESTINATION "lib" ++ LIBRARY DESTINATION ${CMAKE_INSTALL_LIBDIR} ++ RUNTIME DESTINATION ${CMAKE_INSTALL_LIBDIR} + PUBLIC_HEADER DESTINATION "include/crypto_mb") + elseif (DYNAMIC_LIB) # build from ippcp's cmake + install(TARGETS ${MB_DYN_LIB_TARGET} +@@ -154,7 +154,7 @@ endif() + # Static lib installation + if(MB_STANDALONE) + install(TARGETS ${MB_STATIC_LIB_TARGET} +- ARCHIVE DESTINATION "lib" ++ ARCHIVE DESTINATION ${CMAKE_INSTALL_LIBDIR} + PUBLIC_HEADER DESTINATION "include/crypto_mb") + else() + install(TARGETS ${MB_STATIC_LIB_TARGET} +-- +2.37.3 + diff --git a/recipes-oneapi/crypto/intel-crypto-mb/0002-cmake-exclude-Yocto-build-flags.patch b/recipes-oneapi/crypto/intel-crypto-mb/0002-cmake-exclude-Yocto-build-flags.patch new file mode 100644 index 00000000..62b56ea2 --- /dev/null +++ b/recipes-oneapi/crypto/intel-crypto-mb/0002-cmake-exclude-Yocto-build-flags.patch @@ -0,0 +1,45 @@ +From b4549bb765d279b5ba042c6340e1dd69d0890b64 Mon Sep 17 00:00:00 2001 +From: Naveen Saini <naveen.kumar.saini@intel.com> +Date: Thu, 10 Mar 2022 14:30:01 +0800 +Subject: [PATCH 2/2] cmake: exclude Yocto build flags + +Ipp-crypto has its own set of flags and -march values, +which causes conflict with default -march=nehalem in gcc + +Upstream-Status: Inappropriate + +Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com> +--- + sources/ippcp/crypto_mb/src/CMakeLists.txt | 2 +- + sources/ippcp/crypto_mb/src/cmake/linux/GNU.cmake | 2 +- + 2 files changed, 2 insertions(+), 2 deletions(-) + +diff --git a/sources/ippcp/crypto_mb/src/CMakeLists.txt b/sources/ippcp/crypto_mb/src/CMakeLists.txt +index 1ca1a8c..5284d62 100644 +--- a/sources/ippcp/crypto_mb/src/CMakeLists.txt ++++ b/sources/ippcp/crypto_mb/src/CMakeLists.txt +@@ -81,7 +81,7 @@ endif() + set(MB_LIB_TARGET ${MB_DYN_LIB_TARGET}) + + set_source_files_properties(${CRYPTO_MB_SOURCES} PROPERTIES COMPILE_DEFINITIONS "${AVX512_LIBRARY_DEFINES}" +- COMPILE_FLAGS "${AVX512_CFLAGS} ${CMAKE_ASM_FLAGS} ${CMAKE_C_FLAGS_SECURITY}") ++ COMPILE_FLAGS "${AVX512_CFLAGS} ${CMAKE_C_FLAGS_SECURITY}") + + # Don't specify architectural flags for the assembler for this sources, because of the bug in Intel® C Compiler under MacOS: error: invalid instruction mnemonic 'vkmovb' + # The bug has been fixed since version 2021.3. This is a workaround to support older versions of Intel® C Compiler. +diff --git a/sources/ippcp/crypto_mb/src/cmake/linux/GNU.cmake b/sources/ippcp/crypto_mb/src/cmake/linux/GNU.cmake +index a2abeeb..aadd6e2 100644 +--- a/sources/ippcp/crypto_mb/src/cmake/linux/GNU.cmake ++++ b/sources/ippcp/crypto_mb/src/cmake/linux/GNU.cmake +@@ -58,7 +58,7 @@ set(LINK_FLAGS_DYNAMIC "${LINK_FLAGS_DYNAMIC} ${CRYPTO_MB_SOURCES_DIR}/cmake/dll + # Compiler flags + + # Tells the compiler to align functions and loops +-set(CMAKE_C_FLAGS " -falign-functions=32 -falign-loops=32") ++set(CMAKE_C_FLAGS " -falign-functions=32 -falign-loops=32 ${TOOLCHAIN_OPTIONS}") + # Ensures that compilation takes place in a freestanding environment + set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -ffreestanding") + +-- +2.17.1 + diff --git a/recipes-oneapi/crypto/intel-crypto-mb_2021.11.1.bb b/recipes-oneapi/crypto/intel-crypto-mb_2021.11.1.bb new file mode 100644 index 00000000..be72fd85 --- /dev/null +++ b/recipes-oneapi/crypto/intel-crypto-mb_2021.11.1.bb @@ -0,0 +1,28 @@ +SUMMARY = "Crypto Multi-buffer Library" +DESCRIPTION = "Intel® Integrated Performance Primitives (Intel® IPP) Cryptography \ +is a secure, fast and lightweight library of building blocks for cryptography, \ +highly-optimized for various Intel® CPUs." +HOMEPAGE = "https://github.com/intel/ipp-crypto" + +LICENSE = "Apache-2.0" + +LIC_FILES_CHKSUM = "file://../../../LICENSE;md5=e787af283468feca985d6b865d27d95b" + +IPP_BRANCH = "ipp-crypto_${@'_'.join(d.getVar('PV').rsplit('.')[-3:])}" + +SRC_URI = "git://github.com/intel/ipp-crypto;protocol=https;branch=${IPP_BRANCH} \ + file://0001-CMakeLists.txt-exclude-host-system-headers.patch;striplevel=4 \ + file://0002-cmake-exclude-Yocto-build-flags.patch;striplevel=4 \ + file://0001-crypto-mb-Make-sure-libs-are-installed-correctly.patch;striplevel=4 \ + " +SRCREV = "d02611e34328898d16025467d3deeb7d62b6076b" + +S = "${WORKDIR}/git/sources/ippcp/crypto_mb" + +DEPENDS = "openssl" + +inherit cmake pkgconfig +COMPATIBLE_HOST = '(x86_64).*-linux' + +EXTRA_OECMAKE += " -DARCH=intel64" +EXTRA_OECMAKE += " -DTOOLCHAIN_OPTIONS='${TOOLCHAIN_OPTIONS}'" diff --git a/recipes-oneapi/onedpl/onedpl_2022.3.0.bb b/recipes-oneapi/onedpl/onedpl_2022.3.0.bb new file mode 100644 index 00000000..d80b64c9 --- /dev/null +++ b/recipes-oneapi/onedpl/onedpl_2022.3.0.bb @@ -0,0 +1,26 @@ +SUMMARY = "oneAPI DPC++ Library (oneDPL)" +DESCRIPTION = "The oneAPI DPC++ Library (oneDPL) aims to work with the \ +oneAPI DPC++ Compiler to provide high-productivity APIs to developers, \ +which can minimize DPC++ programming efforts across devices for high \ +performance parallel applications." +HOMEPAGE = "https://github.com/oneapi-src/oneDPL" + +LICENSE = "Apache-2.0-with-LLVM-exception" +LIC_FILES_CHKSUM = "file://licensing/LICENSE.txt;md5=2e982d844baa4df1c80de75470e0c5cb \ + file://licensing/third-party-programs.txt;md5=cfb8d6b1f04a8fcc7d0dddd817619634" + +S = "${WORKDIR}/git" + +SRC_URI = "git://github.com/oneapi-src/oneDPL.git;protocol=https;branch=release/2022.3 \ + " +SRCREV = "180f18ad25fbc39fa13bad43f1df7d54ee8f5609" + +do_compile[noexec] = "1" +do_configure[noexec] = "1" + +do_install() { + install -d -m 755 ${D}${includedir}/onedpl + cp -r ${S}/include/* ${D}${includedir}/onedpl +} + +UPSTREAM_CHECK_GITTAGREGEX = "^oneDPL-(?P<pver>(\d+(\.\d+)+))-release$" diff --git a/recipes-oneapi/setup-oneapi-env/files/intel-oneapi-runtime.conf b/recipes-oneapi/setup-oneapi-env/files/intel-oneapi-runtime.conf new file mode 100644 index 00000000..5faf9854 --- /dev/null +++ b/recipes-oneapi/setup-oneapi-env/files/intel-oneapi-runtime.conf @@ -0,0 +1,4 @@ +/opt/intel/oneapi/lib +/opt/intel/oneapi/lib/intel64 +/opt/intel/oneapi/lib/ia32 +/opt/intel/oneapi/lib/emu diff --git a/recipes-oneapi/setup-oneapi-env/setup-intel-oneapi-env_2023.0.0-25370.bb b/recipes-oneapi/setup-oneapi-env/setup-intel-oneapi-env_2023.0.0-25370.bb new file mode 100644 index 00000000..702f794e --- /dev/null +++ b/recipes-oneapi/setup-oneapi-env/setup-intel-oneapi-env_2023.0.0-25370.bb @@ -0,0 +1,20 @@ +SUMMARY = "Intel® oneAPI runtime libraries environment set up" +DESCRIPTION = "Recipe to set up the environment for Intel® oneAPI runtime libraries \ +via configuration file in ld.so.conf.d directory." + +LICENSE = "EULA" +LIC_FILES_CHKSUM = "file://${CUSTOM_LICENSES_PATH}/EULA;md5=7bfc91523de2e84e7131d0eacf2827d4" + +SRC_URI = "file://intel-oneapi-runtime.conf" + +do_install() { + mkdir -p ${D}${sysconfdir}/ld.so.conf.d/ + install -m 644 ${WORKDIR}/intel-oneapi-runtime.conf ${D}${sysconfdir}/ld.so.conf.d/ +} + +pkg_postinst_ontarget:${PN}() { +if [ x"$D" = "x" ]; then + if [ -x /sbin/ldconfig ]; then /sbin/ldconfig ; fi +fi +} +BBCLASSEXTEND = "native nativesdk" diff --git a/recipes-selftest/images/secureboot-selftest-image-unsigned.bb b/recipes-selftest/images/secureboot-selftest-image-unsigned.bb index e03e7b47..d0fa6405 100644 --- a/recipes-selftest/images/secureboot-selftest-image-unsigned.bb +++ b/recipes-selftest/images/secureboot-selftest-image-unsigned.bb @@ -1,12 +1,12 @@ require recipes-core/images/core-image-minimal.bb -DEPENDS_remove = "grub-efi" +DEPENDS:remove = "grub-efi" inherit uefi-comboapp WKS_FILE = "generic-bootdisk.wks.in" -do_uefiapp_deploy_append() { +do_uefiapp_deploy:append() { for i in ${DEPLOY_DIR_IMAGE}/${IMAGE_LINK_NAME}.boot*.efi; do target=`basename $i` target=`echo $target | sed -e 's/${IMAGE_LINK_NAME}.//'` diff --git a/recipes-support/isa-l/isa-l_2.30.0.bb b/recipes-support/isa-l/isa-l_2.31.0.bb index f06fd0b7..d5d09002 100644 --- a/recipes-support/isa-l/isa-l_2.30.0.bb +++ b/recipes-support/isa-l/isa-l_2.31.0.bb @@ -2,16 +2,18 @@ SUMMARY = "Intelligent Storage Acceleration Library" DESCRIPTION = "ISA-L is a collection of optimized low-level functions \ targeting storage applications." LICENSE = "BSD-3-Clause" -LIC_FILES_CHKSUM = "file://LICENSE;md5=e6c7d82ad7dd512687b0991a159a1ca9" +LIC_FILES_CHKSUM = "file://LICENSE;md5=2476688d0948d4a3f7c69473a603540d" SECTION = "lib" inherit autotools pkgconfig S = "${WORKDIR}/git" -SRCREV = "2df39cf5f1b9ccaa2973f6ef273857e4dc46f0cf" -SRC_URI = "git://github.com/intel/isa-l.git" +SRCREV = "bd226375027899087bd48f3e59b910430615cc0a" +SRC_URI = "git://github.com/intel/isa-l.git;branch=master;protocol=https" DEPENDS = "nasm-native" AS[unexport] = "1" +TARGET_CC_ARCH += "${LDFLAGS}" + COMPATIBLE_HOST = '(x86_64).*-linux' diff --git a/recipes-support/jhi/jhi_1.41.bb b/recipes-support/jhi/jhi_1.41.bb deleted file mode 100644 index badb928f..00000000 --- a/recipes-support/jhi/jhi_1.41.bb +++ /dev/null @@ -1,72 +0,0 @@ -SUMMARY = "Intel Dynamic Application Loader (DAL) Host Interface" -DESCRIPTION = "A daemon and libraries which allow user space applications \ -to install Java applets on DAL FW and communicate with them" -LICENSE = "Apache-2.0" -LIC_FILES_CHKSUM = "file://LICENSE;md5=833126f14614a3276708a4d7c9734645" - -SRC_URI = "git://github.com/intel/dynamic-application-loader-host-interface.git;protocol=https" - -inherit cmake useradd systemd pkgconfig update-rc.d - -SRCREV = "4c9ef52bcd55fa03b30ac323e553a05be858d41e" - -S = "${WORKDIR}/git" - -DEPENDS += "util-linux libxml2" -DEPENDS += "${@bb.utils.contains('DISTRO_FEATURES', 'systemd', 'systemd', '', d)}" -RDEPENDS_${PN} += "bash" - -USERADD_PACKAGES = "${PN}" -USERADD_PARAM_${PN} = "--system --no-create-home --shell /bin/false -g mei jhi" - -GROUPADD_PARAM_${PN} = "-g 880 mei" - -COMPATIBLE_HOST_libc-musl = 'null' - -SYSTEMD_SERVICE_${PN} = "jhi.service" - -INITSCRIPT_PACKAGES = "${PN}" -INITSCRIPT_NAME_${PN} = "${PN}" -INITSCRIPT_PARAMS_${PN} = "defaults" - -# systemd is the default so they are installed when sysvinit is not selected as INIT_SYSTEM -EXTRA_OECMAKE = "-DCMAKE_SKIP_RPATH=ON \ - -DINIT_SYSTEM=${@bb.utils.contains('DISTRO_FEATURES', 'sysvinit', 'SysVinit', 'systemd', d)} \ - -DAPPLETS_DIR=${libdir}/dal/applets \ - -DAPP_REPO_DIR=${localstatedir}/cache/dal/applet_repository \ - " - -do_install_append () { - install -d ${D}${localstatedir}/cache/dal/applet_repository - - chown -R jhi ${D}${localstatedir}/cache/dal/applet_repository - chgrp -R mei ${D}${localstatedir}/cache/dal/applet_repository - - install -d ${D}${bindir} - install -m 755 ${B}/bin_linux/smoketest ${D}${bindir} - install -m 755 ${B}/bin_linux/bist ${D}${bindir} - - install -d ${D}${libdir}/dal/applets - cp -r ${S}/test/smoketest/applets/* ${D}${libdir}/dal/applets/ -} - -PACKAGES += "${PN}-test" - -FILES_${PN}-dev = "" - -FILES_${PN} = "\ - ${sbindir} \ - ${sysconfdir} \ - ${libdir}/lib*${SOLIBSDEV} \ - ${libdir}/dal/applets/SpoolerApplet.dalp \ - ${nonarch_libdir}/tmpfiles.d \ - ${systemd_system_unitdir} \ - ${localstatedir}/cache/dal/applet_repository \ - " - -FILES_${PN}-test = "\ - ${bindir} \ - ${libdir}/dal/applets \ - " - -UPSTREAM_CHECK_COMMITS = "1" diff --git a/recipes-support/libipt/libipt_2.0.4.bb b/recipes-support/libipt/libipt_2.1.1.bb index a0554f44..205967c1 100644 --- a/recipes-support/libipt/libipt_2.0.4.bb +++ b/recipes-support/libipt/libipt_2.1.1.bb @@ -5,23 +5,24 @@ library or it can be partially or fully integrated into your tool." HOMEPAGE = "https://github.com/intel/libipt" LICENSE = "BSD-3-Clause" -LIC_FILES_CHKSUM = "file://LICENSE;md5=aea4783603fc6bd8dff5157996ccb9d9" +LIC_FILES_CHKSUM = "file://LICENSE;md5=a429afa59f273b5d12778eda69d10313" inherit pkgconfig cmake S = "${WORKDIR}/git" -SRC_URI = "git://github.com/intel/libipt.git;protocol=https;branch=stable/v2.0" +SRC_URI = "git://github.com/intel/libipt.git;protocol=https;branch=stable/v2.1" -SRCREV = "63813186694487a2bc005f32be57eb526e16714b" +SRCREV = "1c9bc700f4b9a71fd2d1cf1742de7e2351ddb281" EXTRA_OECMAKE += " \ -DPTDUMP=ON \ -DPTTC=ON \ + -DPTSEG=ON \ -DCMAKE_SKIP_RPATH=ON \ " -do_install_append() { +do_install:append() { install -d ${D}${bindir}/libipt install -d ${D}${bindir}/libipt/tests @@ -29,5 +30,5 @@ do_install_append() { cp -r ${WORKDIR}/git/test/src/* ${D}${bindir}/libipt/tests } -FILES_${PN}-test = "${bindir}" +FILES:${PN}-test = "${bindir}" PACKAGES =+ "${PN}-test" diff --git a/recipes-support/sbsigntool/sbsigntool-native_git.bb b/recipes-support/sbsigntool/sbsigntool-native_git.bb index f6094a5c..5a9f5b4d 100644 --- a/recipes-support/sbsigntool/sbsigntool-native_git.bb +++ b/recipes-support/sbsigntool/sbsigntool-native_git.bb @@ -1,5 +1,5 @@ DESCRIPTION = "Utility for signing and verifying files for UEFI Secure Boot" -LICENSE = "GPLv3 & LGPL-2.1 & LGPL-3.0 & MIT" +LICENSE = "GPL-3.0-only & LGPL-2.1-only & LGPL-3.0-only & MIT" # sbsigntool statically links to libccan.a which is built with modules # passed to "create-ccan-tree" (and their dependencies). Therefore, @@ -19,25 +19,25 @@ LIC_FILES_CHKSUM = "file://LICENSE.GPLv3;md5=9eef91148a9b14ec7f9df333daebc746 \ # The original upstream is git://kernel.ubuntu.com/jk/sbsigntool but it has # not been maintained and many patches have been backported in this repo. -SRC_URI = "git://git.kernel.org/pub/scm/linux/kernel/git/jejb/sbsigntools.git;protocol=https;name=sbsigntools \ - git://github.com/rustyrussell/ccan.git;protocol=https;destsuffix=git/lib/ccan.git;name=ccan \ +SRC_URI = "git://git.kernel.org/pub/scm/linux/kernel/git/jejb/sbsigntools.git;protocol=https;name=sbsigntools;branch=master \ + git://github.com/rustyrussell/ccan.git;protocol=https;destsuffix=git/lib/ccan.git;name=ccan;branch=master \ file://0001-configure-Fixup-build-dependencies-for-cross-compili.patch \ - " + " -SRCREV_sbsigntools ?= "f12484869c9590682ac3253d583bf59b890bb826" +SRCREV_sbsigntools ?= "9cfca9fe7aa7a8e29b92fe33ce8433e212c9a8ba" SRCREV_ccan ?= "b1f28e17227f2320d07fe052a8a48942fe17caa5" SRCREV_FORMAT = "sbsigntools_ccan" DEPENDS = "binutils-native gnu-efi-native help2man-native openssl-native util-linux-native" -PV = "0.9.4-git${SRCPV}" +PV = "0.9.5" S = "${WORKDIR}/git" inherit autotools pkgconfig inherit native -do_configure_prepend() { +do_configure:prepend() { cd ${S} sed -i s#RECIPE_SYSROOT#${RECIPE_SYSROOT_NATIVE}#g configure.ac @@ -79,3 +79,5 @@ EXTRA_OEMAKE = "\ -I${STAGING_INCDIR_NATIVE} \ -I${STAGING_INCDIR_NATIVE}/efi/${@efi_arch(d)}' \ " + +CFLAGS:append = " -Wno-error" diff --git a/recipes-support/sbsigntool/sbsigntool/0001-configure-Fixup-build-dependencies-for-cross-compili.patch b/recipes-support/sbsigntool/sbsigntool/0001-configure-Fixup-build-dependencies-for-cross-compili.patch index ef686511..ea7bee29 100644 --- a/recipes-support/sbsigntool/sbsigntool/0001-configure-Fixup-build-dependencies-for-cross-compili.patch +++ b/recipes-support/sbsigntool/sbsigntool/0001-configure-Fixup-build-dependencies-for-cross-compili.patch @@ -10,7 +10,7 @@ under /usr/include and /usr/lib. Prepend these paths with a placeholder that can be replaced with the actual paths once they are resolved. -Upstream status: inappropriate [OE specific] +Upstream-Status: Inappropriate [OE specific] Signed-off-by: Ricardo Neri <ricardo.neri-calderon@linux.intel.com> |