Age | Commit message (Collapse) | Author |
|
Backport a patch to fix the kernel hang issue when the uImage is built
by gcc-5.2.
The details can be found in
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=5e95235
Signed-off-by: Zhenhua Luo <zhenhua.luo@freescale.com>
(cherry picked from commit 95d9f9aba81ccdae53011ac8c6fbbc2605b9cedc)
|
|
Fix the below error:
DEBUG: Executing shell function do_configure
NOTE: make -e MAKEFLAGS= clean
make -C /lib/modules/3.16.0-4-amd64/build SUBDIRS=`pwd` clean
make[1]: *** /lib/modules/3.16.0-4-amd64/build: No such file or directory. Stop.
Makefile:39: recipe for target 'clean' failed
make: *** [clean] Error 2
ERROR: oe_runmake failed
This aligns with the cryptodev recipe in Poky.
Signed-off-by: Ting Liu <ting.liu@freescale.com>
Acked-by: Otavio Salvador <otavio@ossystems.com.br>
|
|
To avoid build issue when Poky upgrades cryptodev, keep a copy
of cryptodev-{linux,modules,tests) recipe in this repository.
Based on the upstream, this includes the following FSL patches:
0001-add-support-for-composite-TLS10-SHA1-AES-algorithm-o.patch
0002-add-support-for-COMPAT_CIOCAUTHCRYPT-ioctl.patch
0003-PKC-support-added-in-cryptodev-module.patch
0004-Compat-versions-of-PKC-IOCTLs.patch
0005-Asynchronous-interface-changes-in-cryptodev.patch
0006-ECC_KEYGEN-and-DLC_KEYGEN-supported-in-cryptodev-mod.patch
0007-RCU-stall-fixed-in-PKC-asynchronous-interface.patch
0008-Add-RSA-Key-generation-offloading.patch
0009-Fixed-compilation-error-of-openssl-with-fsl-cryptode.patch
0010-add-support-for-composite-TLS10-SHA1-3DES-algorithm-.patch
0011-add-support-for-TLSv1.1-record-offload.patch
0012-add-support-for-TLSv1.2-record-offload.patch
0013-clean-up-code-layout.patch
0014-remove-redundant-data-copy-for-pkc-operations.patch
0015-fix-pkc-request-deallocation.patch
0016-add-basic-detection-of-asym-features.patch
0017-remove-dead-code.patch
0018-fix-compat-warnings.patch
0019-fix-size_t-print-format.patch
0020-fix-uninitialized-variable-compiler-warning.patch
Signed-off-by: Cristian Stoica <cristian.stoica@freescale.com>
Signed-off-by: Ting Liu <ting.liu@freescale.com>
|
|
Minor version update to 3.12.37-rt51 with new features:
* e6500 hugepage TLB miss performance improvement
* T1023RDB support
* T1040D4RDB and T1042D4RDB support
* DIU [T1042]
* DPAA Ethernet: loadable module
* eMMC: DDR mode [T2080]
* eTSEC: Gianfar upstream updates and fixes
* fmlib: table statistics, stats extension
* IEEE802.1AE (MACSEC) and IEEE802.1X (port-based network access control) [T104x, T102x]
* IEEE1588 ptpd open source stack includes more DPAA processors: P1023, P2041, P3041, P5020, P5040, T4240, T1023
* LAG SGMII 2.5G ports support - IPv4 traffics forwarding on aggregated 2 x 2.5Gb L2 Switch FMAN ports [1040]
* LAG support of IPv6 traffics forwarding and TCP/UDP traffics over IPv6 forwarding (2 x 2.5Gb L2 Switch WAN) [1040]
* LAG support of IPv6 traffics forwarding and TCP/UDP traffics over IPv6 forwarding on both 1 G RGMII port and 1G SGMII port [1040]
* Power Management: Power off feature for all QDS boards except B9132QDS and B4860QDS
* SEC: QI Driver IPSec performance improvement
* SGMII 2.5G fixed link [T1024]
* USB: Dual UTMI
For detailed history, see
http://git.freescale.com/git/cgit.cgi/ppc/sdk/linux.git/tag/?id=fsl-sdk-v1.8
Also remove the patches which already merged in 3.12.37-rt51
Signed-off-by: Ting Liu <ting.liu@freescale.com>
Acked-by: Otavio Salvador <otavio@ossystems.com.br>
|
|
In poky commit 46cdaf1, the kernel build output was put into
kernel-build-artifacts, and kernel-source is kept "pristine".
KERNEL_PATH points to kernel-source, while KBUILD_OUTPUT points
to build-artifacts which can be used for external module build.
asf Makefile tries to include the .config, use KBUILD_OUTPUT to
avoid build error.
It is needed to process all the Makefile with new revision.
Signed-off-by: Ting Liu <ting.liu@freescale.com>
Acked-by: Otavio Salvador <otavio@ossystems.com.br>
|
|
This imports following changes:
62ba10a asf_fmc: Patch to update the name of xml files.
8d744dd asf: Patch to fix ipv6 issue on dpaa.
ddf9678 asf: Patch to return jumbo packet back.
dcda291 asf: Patch to include caam_jr_strstatus in ASF_IPSEC_DEBUG mode.
49f591f asf: Patch to change function signature caam_jr_strstatus
f860eff ipsecproc: Patch to fix display of ipsec stats.
461586d ASFIPSEC: fix compilation for SEC 3.x when ASF_IPSEC_DEBUG is defined
03b9bad asf: Patch to fix sha256 issue.
a21b2d7 Removing Ethernet padding in abuf to skb conversion
a03283d4 Aligning function ASFFFPProcessAndSendFD and ASFFFPProcessAndSendPkt while validating flows
Signed-off-by: Ting Liu <ting.liu@freescale.com>
Acked-by: Otavio Salvador <otavio@ossystems.com.br>
|
|
The license information is moved to COPYING file, use it.
Signed-off-by: Ting Liu <ting.liu@freescale.com>
Acked-by: Otavio Salvador <otavio@ossystems.com.br>
|
|
The changes include:
- bc is added as a runtime dependency for performance tests
- mini_calc is dropped and replaced by bc
- use 'oe_runmake modules_install' to install kernel modules
There are 138 new git commits. For detailed info, see:
http://git.freescale.com/git/cgit.cgi/ppc/sdk/pkc-host.git/tag/?id=fsl-sdk-v1.8
Signed-off-by: Ting Liu <ting.liu@freescale.com>
Acked-by: Otavio Salvador <otavio@ossystems.com.br>
|
|
The right license is BSP & GPLv2+, fix it.
And the licensing information is separated to COPYING file, use it.
Signed-off-by: Ting Liu <ting.liu@freescale.com>
Acked-by: Otavio Salvador <otavio@ossystems.com.br>
|
|
This imports the following changes:
a655c57 rename perf script to avoid conflicts with pkc driver
3e40e67 make perf script executable and install it to /usr/bin
1668005 add gitignore
7abf553 update perf script with the fixes from pkc-host driver
ec39361 extract licensing information to a separate file
e821c4f merge Makefile with config.mk
Signed-off-by: Ting Liu <ting.liu@freescale.com>
Acked-by: Otavio Salvador <otavio@ossystems.com.br>
|
|
This imports the following changes:
80de832 Fix Copyright issues
0c9921f Add support for interrupts on PHYs connected to seville's ports
53de571 fix: Avoid memory leak
6cf2d52 Refractoring uio_seville_main.c file
7b2ccba Fix: Use Seville register when flushing, not remapper's register
d347b98 Beautify: Replace big switch() with a 2-line for()
f8c8249 Beautify: Fix for() alignment
fdde45d Fix: Pad undersized injected CPU control frames
3b9cff6 Fix: Typo for "reciving"
052d776 Fix: Assure that there can't be 2 processes that open the uio device
3c29b27 Fix: Interrupts might occur when /dev/npi is closing
59e3d2e Fix: Assure that only one thread is allowed to inject a frame at a time
b96827c Fix: Assure that only one thread may extract a CPU frame at a time
8c9c47e Fix: Remove unecessary check for data in the polling function
a202daf Fix: L2 Switch interrupt handler is called twice for each packet
985d4c2 Fix: dev/npi might be opened by two applications
3ecb8f3 Fix: Fix initialization of valiables for struct dev_npi
dd74172 Beautify: Add ':' at the end of npi_read_q in description
aceb6be Beautify: function do_control_frame_extr_dev() has duplicated code
262876b Beautify: Remove unnecessary macro
d737cd8 Fix: Treat interrupts only from group 0
181ae15 Beautify: Rename variable leftovers_start to leftovers_begin
Signed-off-by: Ting Liu <ting.liu@freescale.com>
Acked-by: Otavio Salvador <otavio@ossystems.com.br>
|
|
When building external module, $(PWD) is the root of the kernel tree.
Signed-off-by: Liu Ting-B28495 <ting.liu@freescale.com>
|
|
In poky commit 46cdaf1, the kernel build output was put into
kernel-build-artifacts, and kernel-source is kept "pristine".
KERNEL_PATH points to kernel-source, while KBUILD_OUTPUT points
to build-artifacts which can be used for external module build.
ipc Makefile tries to include the .config, use KBUILD_OUTPUT to
avoid build error.
Signed-off-by: Liu Ting-B28495 <ting.liu@freescale.com>
|
|
Fix the following build error:
* ERROR: QA Issue: skmm-host: Files/directories were installed but not shipped
/usr/bin/mini_calc [installed-vs-shipped]
Signed-off-by: Zhenhua Luo <zhenhua.luo@freescale.com>
|
|
In poky commit 46cdaf1, the kernel build output was put into
kernel-build-artifacts, and kernel-source is kept "pristine".
KERNEL_PATH points to kernel-source, while KBUILD_OUTPUT points
to build-artifacts which can be used for external module build.
ceetm Makefile tries to include the .config, use KBUILD_OUTPUT to
avoid build error.
Signed-off-by: Liu Ting-B28495 <ting.liu@freescale.com>
|
|
In poky commit 46cdaf1, the kernel build output was put into
kernel-build-artifacts, and kernel-source is kept "pristine".
KERNEL_PATH points to kernel-source, while KBUILD_OUTPUT points
to build-artifacts which can be used for external module build.
asf Makefile tries to include the .config, use KBUILD_OUTPUT to
avoid build error.
Signed-off-by: Liu Ting-B28495 <ting.liu@freescale.com>
|
|
Signed-off-by: Zhenhua Luo <zhenhua.luo@freescale.com>
|
|
fix the following error:
|make: *** No rule to make target `/poky/build/tmp/work-shared/p4080ds/kernel-source/.config
(From OE-Core rev: 6a1ff0e7eacef595738f2fed086986fd622ec32a)
Signed-off-by: Chunrong Guo <B40290@freescale.com>
|
|
This fixes a 1-byte NULL write past the end of allocated memory
References
http://seclists.org/oss-sec/2015/q1/582
https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-9683
Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
|
|
Fixes memory corruption or panic during key garbage collection
Reference:
https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-9529
Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
|
|
Insufficient syscall number validation in perf and ftrace subsystems
CVE-2014-7825
Fixes an out-of-bounds memory access flaw, in the syscall tracing
functionality of the Linux kernel's perf subsystem.
CVE-2014-7826
Fixes an out-of-bounds memory access flaw, in the syscall
tracing functionality of the Linux kernel's ftrace subsystem.
References:
http://www.openwall.com/lists/oss-security/2014/11/06/11
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-7825
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-7826
Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
|
|
Fixes a potential privilege escalation flaw when the Linux
kernel is built with the user namespaces(CONFIG_USER_NS)
References:
http://www.openwall.com/lists/oss-security/2014/06/10/4
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-4014
Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
|
|
Fixes a denial of service flaw in the Linux kernel
built with the shared memory suppor
Reference:
http://www.openwall.com/lists/oss-security/2014/06/18/11
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-4171
Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
|
|
try_to_unmap_cluster() should lock_page() before mlocking
Reference:
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-3122
Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
|
|
Explicitly clear ramdisk_mcp backend pages
References:
https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-4027
http://permalink.gmane.org/gmane.linux.scsi.target.devel/6618
Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
|
|
Handle numid overflow
Make sure that id->index does not overflow
References:
http://www.openwall.com/lists/oss-security/2014/06/26/6
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-4656
Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
|
|
NULL pointer dereference in af->from_addr_param on malformed packet
Reference:
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-7841
Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
|
|
sk_ack_backlog wrap-around problem
Reference:
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-4667
Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
|
|
CVE-2014-4652
Protect user controls against concurrent access
CVE-2014-4653
Don't access controls outside of protected regions
Reference:
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-4652
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-4653
Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
|
|
audit_krule mask accesses need bounds checking
Reference:
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-3917
Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
|
|
CVE-2014-3673
skb_over_panic when receiving malformed ASCONF chunks
Fixes: b896b82be4ae ("[SCTP] ADDIP: Support for processing incoming ASCONF_ACK
chunks.")
CVE-2014-3687
panic on duplicate ASCONF chunks
Fixes: 2e3216cd54b1 ("sctp: Follow security requirement of responding with 1
packet")
CVE-2014-3688
remote memory pressure from excessive queueing
Fixes: 2e3216cd54b1 ("sctp: Follow security requirement of responding with 1
packet")
References:
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-3673
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-3687
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-3688
http://www.openwall.com/lists/oss-security/2014/11/13/8
Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
|
|
CVE-2014-3601
Fixes the third parameter of kvm_iommu_put_pages
The third parameter of kvm_iommu_put_pages is wrong,
It should be 'gfn - slot->base_gfn'.
CVE-2014-8369
Fixes excessive pages un-pinning in kvm_iommu_map error path.
(This vulnerability exists because of an incorrect fix for CVE-2014-3601
Reference:
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-3601
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-8369
https://bugzilla.redhat.com/show_bug.cgi?id=1156518
https://lkml.org/lkml/2014/10/24/460
Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
|
|
CVE-2014-3181 Kernel: HID: OOB write in magicmouse driver
CVE-2014-3182 Kernel: HID: logitech-dj OOB array access
CVE-2014-3184 Kernel: HID: off by one error in various _report_fixup routine
CVE-2014-3185 Kernel: USB serial: memory corruption flaw
References:
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-3181
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-3182
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-3184
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-3185
Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
|
|
Fixes null pointer dereference when processing authenticated cookie_echo chunk
Reference:
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-0101
https://bugzilla.redhat.com/show_bug.cgi?id=1070705
Introduced by:
http://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=bbd0d59809f9
Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
|
|
Avoid infinite loop when processing indirect ICBs
References:
http://seclists.org/oss-sec/2014/q3/600
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-6410
Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
|
|
This patches fixes mount flags handling during remount issue.
The patches come from:
https://www.kernel.org (remotes/origin/linux-3.12.y branch)
References:
http://seclists.org/oss-sec/2014/q3/357
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-5206
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-5207
Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
|
|
Signed-off-by: Ting Liu <b28495@freescale.com>
|
|
This includes following fixes:
35af73f Fix: Copy user-space buffer of injected control frame to kernel
00c8040 Add multiple error labels in the probe function
2835689 Fix: Remove memory leaks when the module is removed or fails on probing
d770a37 Fix: Remove compile warning on 32b
89e29fc Fix: Remove unnecessary checks
adbb47e Beautify: Rename macros into more proper names
d5cac6e Add "poll()" function for NPI device
220cee3 Replace rescheduling with work queues
2a5fe4e Add cacheline support for extraction of control frames
6f14f0b Fix: UIO device might not be removed properly if module fails to initialize
2dcea55 Removed unecessary includes
1077880 Add UIO driver
68ab7bd Initial empty repository
update COMPATIBLE_MACHINE to use soc_family.
Signed-off-by: Zhenhua Luo <zhenhua.luo@freescale.com>
|
|
This update to fsl sdk v1.7 release. The detailed changes
can be found at:
http://git.freescale.com/git/cgit.cgi/ppc/sdk/asf.git/
Other changes to sync with meta-fsl-arm:
* install kernel modules to ${D}/lib/modules/${KERNEL_VERSION}/asf
* install scripts to ${D}/${libexecdir}/
* remove dependency on virtual/kernel as it inherit module.bbclass
Signed-off-by: Zhenhua Luo <zhenhua.luo@freescale.com>
|
|
Currently pkc-host does not support RSA_KEYGEN. When pkc-host
installed, RSA keygen operations should be avoided.
Introduce DISTRO_FEATURE "c29x_pkc". To install pkc-host, this
feature should be enabled. Then cryptodev checks it to disable
RSA_KEYGEN.
this can be done in conf/local.conf:
DISTRO_FEATURES_append = " c29x_pkc"
Signed-off-by: Cristian Stoica <cristian.stoica@freescale.com>
|
|
- Nikos handed over project maintainance to Phil Sutter.
- Several pending patches have been merged upstream so we removed
them from the recipe. The remaining ones have been sorted
Signed-off-by: Cristian Stoica <cristian.stoica@freescale.com>
Change-Id: I0c6160c739d379ba787e72423d1564b9a3d05d8b
Reviewed-on: http://git.am.freescale.net:8181/24177
Reviewed-by: Zhenhua Luo <zhenhua.luo@freescale.com>
Tested-by: Zhenhua Luo <zhenhua.luo@freescale.com>
|
|
This update to fsl sdk v1.7 release.
For detailed changes, see:
http://git.freescale.com/git/cgit.cgi/ppc/sdk/linux.git/
Remove the patches which were applied in the repository.
Signed-off-by: Zhenhua Luo <zhenhua.luo@freescale.com>
|
|
Signed-off-by: Ting Liu <ting.liu@freescale.com>
|
|
Having a number higher than necessary for NR_CPUS wastes memory by
instantiating unnecessary structures in RAM. An example is in the DPAA
where DPAA_ETH_TX_QUEUES is defined based on NR_CPUS and used to create
dozens of extra qman_fq structures. Using the prior value of 24, which
was left over from the T4240 created an additonal 60 frame queue
structures alone.
This has been tested on t1040rdb-64b.
Signed-off-by: Bob Cochran <yocto@mindchasers.com>
|
|
Although the flag existed in Kconfig for FMAN_V3L, it was not set by
default. Also, the T1040 and FMAN V3L only support four O/H ports, so
remove the last two from the DTS files. Otherwise, MAJOR FM Errors
will be reported during FMAN probing / init.
This was tested on t1040rdb-64b
Signed-off-by: Bob Cochran <yocto@mindchasers.com>
|
|
T1040 uses an E5500 processor, and E5500 doesn't have an Altivec unit
This was tested on a t1040rdb
Signed-off-by: Bob Cochran <yocto@mindchasers.com>
|
|
Fix unbounded recursion when processing relocated directories
Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
Signed-off-by: Zhenhua Luo <zhenhua.luo@freescale.com>
|
|
A NULL pointer dereference flaw was found in the way the
Linux kernel's Stream Control Transmission Protocol
(SCTP) implementation handled simultaneous connections
between the same hosts. A remote attacker could use this
flaw to crash the system.
References:
- https://access.redhat.com/security/cve/CVE-2014-5077
- http://patchwork.ozlabs.org/patch/372475/
Signed-off-by: Liviu Gheorghisan <liviu.gheorghisan@enea.com>
|
|
Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
|
|
While other cryptodev-fsl's bbapend to assign FILESEXTRAPATHS,
there was a potential dir split error.
Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com>
|