diff options
Diffstat (limited to 'meta-seattle')
| -rw-r--r-- | meta-seattle/README.md (renamed from meta-seattle/README) | 68 |
1 files changed, 16 insertions, 52 deletions
diff --git a/meta-seattle/README b/meta-seattle/README.md index fd4d4610..d2e7ed88 100644 --- a/meta-seattle/README +++ b/meta-seattle/README.md @@ -1,59 +1,27 @@ -meta-seattle -============ +# meta-seattle -This is the location for AMD Seattle maintained BSP. +This is the location for AMD Seattle BSP. - -Overview of AMD 64-bit ARM-based processor ------------------------------------------- +## Overview of AMD 64-bit ARM-based processor The AMD Opteron A1100-Series features AMDs first 64-bit ARM-based processor, codenamed "Seattle". - -Yocto Project Compatible -======================== +## Yocto Project Compatible This BSP is compatible with the Yocto Project as per the requirements listed here: https://www.yoctoproject.org/webform/yocto-project-compatible-registration -Dependencies -============ +## Dependencies This layer depends on: - URI: git://git.openembedded.org/bitbake - branch: master - - URI: git://git.openembedded.org/openembedded-core - layers: meta - branch: master - - URI: git://git.yoctoproject.org/meta-amd - layers: common - branch: master - -Patches -======= - -Please submit any patches against this BSP to the maintainer and cc -the meta-amd mailing list (meta-amd@yoctoproject.org): +[bitbake](https://github.com/openembedded/bitbake) layer, +[oe-core](https://github.com/openembedded/openembedded-core) layer, +meta-amd/common layer -Maintainer: Adrian Calianu <adrian.calianu@enea.com> - - -Table of Contents -================= - - I. Building the meta-seattle BSP layer - II. Booting the images -III. How to Run 32-bit Applications on aarch64 - IV. Limitations - - -I. Building the meta-seattle BSP layer -======================================= +## Building the meta-seattle BSP layer The following instructions require a Poky installation (or equivalent). @@ -83,8 +51,7 @@ Once complete the images for the target machine will be available in the output directory 'tmp/deploy/images'. -II. Booting the images -===================================== +## Booting the images Booting the images using UEFI firmware @@ -119,8 +86,7 @@ References: 1) https://www.kernel.org/doc/Documentation/efi-stub.txt 2) http://www.uefi.org/specifications -III. How to Run 32-bit Applications on aarch64 -============================================== +## How to Run 32-bit Applications on aarch64 Since multilib is not yet available for aarch64, this BSP offers some alternatives in order to run 32-bit legacy applications on an aarch64. @@ -132,8 +98,7 @@ chroot and another one qemu installation. meta-seattle BSP includes two scripts that help you to setup environment for 32bit applications that can be found under following path meta-seattle/recipes-support/32b-env/ -III.1 Using chroot tool to run 32-bit applications on aarch64 -============================================================= +### Using chroot tool to run 32-bit applications on aarch64 This solution requires the chroot tool on the target, a 32b-built rootfs, and a build for the seattle machine. @@ -144,15 +109,14 @@ root@seattle:~# tar -zxf rootfs32.tar.gz root@seattle:~# set_32b_env_chroot.sh -r ./rootfs32 -# 32b environment started +- 32b environment started @seattle:~# ./hello32d Starting from this point, any 32-bit application can be executed, but not 64-bit applications which will fail due to wrong path to 32-bit libraries. To run a 64-bit application, chroot mode must be quit by using the "exit" command. -III.2 Using qemu to run 32-bit applications on aarch64 -====================================================== +### Using qemu to run 32-bit applications on aarch64 This example requires an existing 32b-built rootfs and a build for the seattle machine. It is also expected that qemu-arm is already installed!!! @@ -163,8 +127,8 @@ root@seattle:~# source /usr/bin/set_32b_env_qemu.sh -r ./rootfs32 Starting from this point, any 32-bit or 64-bit application can be executed. -IV. Limitations -=============== +## Limitations + In order to enable 32-bit support in the aarch64 kernel, COMPAT mode is enabled (CONFIG_COMPAT=y), page size set to 4K, and VirtualAddress set to 48-bit. |