1
0
mirror of https://git.yoctoproject.org/meta-arm synced 2026-07-18 04:27:08 +00:00

Compare commits

..

12 Commits

Author SHA1 Message Date
Tushar Khandelwal 44d3f88ed9 arm: trusted-firmware-a: fix the upstream revision
Signed-off-by: Tushar Khandelwal <tushar.khandelwal@arm.com>
2019-09-19 11:02:15 +01:00
Rui Miguel Silva 1fee2e71d8 arm: trusted-firmware-a: fix compilation order
Swap the compile order and make the fip to build at the end when there
is all the needed targets compiled.

Signed-off-by: Rui Miguel Silva <rui.silva@linaro.org>
2019-09-16 10:58:18 +01:00
Tushar Khandelwal 912967dcd7 iota: use custom kernel instead of the one from poky tiny
Change-Id: I6cf83071d8dfde2d41d4ea3d8ad006ba4bbabda5
Signed-off-by: Tushar Khandelwal <tushar.khandelwal@arm.com>
2019-09-02 15:50:40 +01:00
Rui Miguel Silva f43c35f07c arm: add cramfs class and recipes
Add cramfs-xip to image types and cramfs tools.

Change-Id: I801788c8c73fe354aa526032ae20927a48672cac
Signed-off-by: Rui Miguel Silva <rui.silva@linaro.org>
2019-09-02 15:50:40 +01:00
Rui Miguel Silva ebb261e64b iota: add distro related cofiguration
Add distro configuration and images recipes for tiny and core image for
iota distribution.

Change-Id: I31df5009dbd67b03016150833cc1a6905caa9ca2
Signed-off-by: Rui Miguel Silva <rui.silva@linaro.org>
2019-09-02 15:50:40 +01:00
Rui Miguel Silva f815620455 designstart: add support to a5ds
Add first configuration, layers and recipes to support build of a5ds
images.

Change-Id: Ie6273d0e3feb55471cfc784e62168ce3e2e1f1b6
Signed-off-by: Rui Miguel Silva <rui.silva@linaro.org>
2019-09-02 15:50:40 +01:00
Rui Miguel Silva 9f34cbb321 arm: add arm gcc 7.3.1 native support
Add native gcc version 7.3.1 to support build
of boot-firmware.

Change-Id: I0c64c6217f88e667f24155e1a587badfcc097917
Signed-off-by: Rui Miguel Silva <rui.silva@linaro.org>
2019-09-02 15:50:40 +01:00
Rui Miguel Silva 7e362ece84 arm: add kernel common recipes
add kernel and dtc tool append recipes.

Change-Id: Ie29a4e086de83140f09e7a17b5a6278090c610b0
Signed-off-by: Rui Miguel Silva <rui.silva@linaro.org>
2019-09-02 15:50:40 +01:00
Rui Miguel Silva 96137096b8 arm: add bsp recipes to common layer
Add recipes for common layer for the bsp area, includes
trustedfirmware-a, boot-firmware and u-boot recipe append.

Change-Id: I447a52750afb02cc170f38d5148fa4b5f5777abc
Signed-off-by: Rui Miguel Silva <rui.silva@linaro.org>
2019-09-02 15:50:40 +01:00
Rui Miguel Silva 0d2da45a4f arm: add configuration for common layer
Add initial configuration for arm common layerj

Change-Id: I59cd6b5d4a1fa31eec4472900dbb0fc52f18c4c4
Signed-off-by: Rui Miguel Silva <rui.silva@linaro.org>
2019-09-02 15:50:40 +01:00
Tushar Khandelwal eafd591407 add license file
Signed-off-by: Tushar Khandelwal <tushar.khandelwal@arm.com>
2019-09-02 15:50:40 +01:00
Rui Miguel Silva cc3c00065b Initial commit 2019-05-14 14:52:01 +01:00
608 changed files with 853 additions and 43895 deletions
View File
+12 -12
View File
@@ -1,17 +1,17 @@
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in
The above copyright notice and this permission notice shall be included in
all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
THE SOFTWARE.
-50
View File
@@ -1,50 +0,0 @@
Introduction
------------
This repository contains Arm layers for OpenEmbedded
meta-arm:
This layer provides support for general recipes for the Arm
architecture. Anything that's not needed explicitly for BSPs, the IOTA
distribution, or destined to be upstreamed belongs here.
meta-arm-bsp:
This layer provides support for Arm reference platforms
meta-arm-iota:
This layer provides support for Arm's IOTA Linux Distribution
meta-arm-toolchain:
This layer provides support for Arm's GNU-A toolset releases
meta-arm-autonomy:
This layer provides a reference stack for autonomous systems.
Contributing
------------
Currently, we only accept patches from the meta-arm mailing list. For general
information on how to submit a patch, please read
https://www.openembedded.org/wiki/How_to_submit_a_patch_to_OpenEmbedded
E-mail meta-arm@lists.yoctoproject.org with patches created using this process. You can configure git-send-email to automatically use this address for the meta-arm repository with the following git command:
$ git config --local --add sendemail.to meta-arm@lists.yoctoproject.org
Commits and patches added should follow the OpenEmbedded patch guidelines:
https://www.openembedded.org/wiki/Commit_Patch_Message_Guidelines
The component being changed in the shortlog should be prefixed with the layer name (without the meta- prefix), for example:
arm-bsp/trusted-firmware-a: decrease frobbing level
arm-toolchain/gcc: enable foobar v2
Reporting bugs
--------------
E-mail meta-arm@lists.yoctoproject.org with the error encountered and the steps
to reproduce the issue.
Maintainer(s)
-------------
* Jon Mason <jon.mason@arm.com>
* Ross Burton <ross.burton@arm.com>
+15
View File
@@ -0,0 +1,15 @@
# SPDX-License-Identifier: MIT
#
# Copyright (c) 2019 Arm Limited
#
SUMMARY = "Cramfs XIP image type"
DESCRIPTION = "Cramfs XIP type image to add to IMAGE_TYPES"
LICENSE = "MIT"
inherit image_types
do_image_cramfs_xip[depends] += "cramfs-tools-native:do_populate_sysroot"
IMAGE_TYPES += " cramfs-xip"
IMAGE_CMD_cramfs-xip = "mkcramfs -n ${IMAGE_NAME} -X ${IMAGE_ROOTFS} ${IMGDEPLOYDIR}/${IMAGE_NAME}${IMAGE_NAME_SUFFIX}.cramfs-xip ${EXTRA_IMAGECMD}"
+21
View File
@@ -0,0 +1,21 @@
# SPDX-License-Identifier: MIT
#
# Copyright (c) 2019 Arm Limited
#
# arm has a conf and classes directory, add to BBPATH
BBPATH .= ":${LAYERDIR}"
BBFILES += " \
${LAYERDIR}/recipes-*/*/*.bb \
${LAYERDIR}/recipes-*/*/*.bbappend"
BBFILES += "${@' '.join('${LAYERDIR}/%s/recipes*/*/*.bbappend' % layer \
for layer in BBFILE_COLLECTIONS.split())}"
BBFILE_COLLECTIONS += "arm"
BBFILE_PATTERN_arm := "^${LAYERDIR}/"
BBFILE_PRIORITY_arm = "10"
LAYERSERIES_COMPAT_arm = "warrior"
LAYERDEPENDS_arm = "core"
@@ -0,0 +1,18 @@
# SPDX-License-Identifier: MIT
#
# Copyright (c) 2019 Arm Limited
#
IMAGE_FSTYPES += "tar.bz2 wic cpio.gz"
EXTRA_IMAGEDEPENDS_append = " trusted-firmware-a"
PREFERRED_PROVIDER_u-boot ?= "u-boot"
PREFERRED_PROVIDER_virtual/bootloader ?= "u-boot"
UBOOT_SUFFIX ?= "bin"
UBOOT_MACHINE ?= "vexpress_aemv8a_aarch32_defconfig"
PREFERRED_PROVIDER_virtual/kernel ?= "linux-arm"
SERIAL_CONSOLES = "115200;ttyAMA0"
+8
View File
@@ -0,0 +1,8 @@
# SPDX-License-Identifier: MIT
#
# Copyright (c) 2019 Arm Limited
#
TUNE_FEATURES = "aarch32"
require conf/machine/include/arm/arch-armv8a.inc
+11
View File
@@ -0,0 +1,11 @@
# SPDX-License-Identifier: MIT
#
# Copyright (c) 2019 Arm Limited
#
DEFAULTTUNE="cortexa5"
require conf/machine/include/tune-cortexa5.inc
TF-A_EXTRA_OPTIONS = "ARM_ARCH_MAJOR=7 ARM_CORTEX_A5=yes \
ARM_XLAT_TABLES_LIB_V1=1"
@@ -0,0 +1,61 @@
# SPDX-License-Identifier: MIT
#
# Copyright (c) 2019 Arm Limited
#
SUMMARY = "Trusted Firmware for Cortex-A"
DESCRIPTION = "Trusted Firmware-A"
HOMEPAGE = "https://github.com/ARM-software/arm-trusted-firmware"
LICENSE = "BSD-3-Clause"
TF-A_DEPENDS ?= ""
DEPENDS += " dtc-native coreutils-native"
DEPENDS += " ${TF-A_DEPENDS} "
LIC_FILES_CHKSUM = "file://license.rst;md5=90153916317c204fade8b8df15739cde"
FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:"
SRCREV = "${AUTOREV}"
SRC_URI = "\
git://github.com/ARM-software/arm-trusted-firmware;protocol=https;nobranch=1 \
"
S = "${WORKDIR}/git"
B = "${WORKDIR}/build"
COMPATIBLE_MACHINE = "(vexpress-a5|vexpress-a32|corstone700|a5ds)"
LDFLAGS[unexport] = "1"
TF-A_PLATFORM ?= "fvp"
TF-A_ARCH ?= "aarch32"
TF-A_DEBUG ?= "1"
TF-A_AARCH32_SP ?= "sp_min"
TF-A_BL33 ?= ""
TF-A_TARGET_IMAGES ?= "all fip"
TF-A_EXTRA_OPTIONS ?= ""
do_compile() {
mkdir -p ${B}
oe_runmake -C ${S} BUILD_BASE=${B} \
BUILD_PLAT=${B}/${TF-A_PLATFORM}/ \
PLAT=${TF-A_PLATFORM} \
DEBUG=${TF-A_DEBUG} \
ARCH=${TF-A_ARCH} \
CROSS_COMPILE=${TARGET_PREFIX} \
AARCH32_SP=${TF-A_AARCH32_SP} \
BL33=${TF-A_BL33} \
${TF-A_EXTRA_OPTIONS} \
${TF-A_TARGET_IMAGES}
}
do_install() {
[ -f ${B}/${TF-A_PLATFORM}/bl1.bin ] && install -D -p -m 0644 ${B}/${TF-A_PLATFORM}/bl1.bin ${DEPLOY_DIR_IMAGE}/bl1.bin
[ -f ${B}/${TF-A_PLATFORM}/bl1/bl1.elf ] && install -D -p -m 0644 ${B}/${TF-A_PLATFORM}/bl1/bl1.elf ${DEPLOY_DIR_IMAGE}/bl1.elf
install -D -p -m 0644 ${B}/${TF-A_PLATFORM}/fip.bin ${DEPLOY_DIR_IMAGE}/${TF-A_PLATFORM}.fip
}
+30
View File
@@ -0,0 +1,30 @@
# SPDX-License-Identifier: MIT
#
# Copyright (c) 2019 Arm Limited
#
SRCREV = "9bed08a2ed8d4915e694e5b4a030bd361b829a52"
LIC_FILES_CHKSUM = "file://Licenses/README;md5=a2c678cfd4a4d97135585cad908541c6"
SRC_URI = "\
git://git.linaro.org/landing-teams/working/arm/u-boot;protocol=https;nobranch=1 \
"
S = "${WORKDIR}/git"
B = "${WORKDIR}/build"
U-BOOT_PLATFORM = "fvp_ve"
FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:"
COMPATIBLE_MACHINE = "(vexpress-a5|vexpress-a32|a5ds)"
do_compile() {
mkdir -p ${B}
oe_runmake -C ${S} BUILD_BASE=${B} \
BUILD_PLAT=${B}/${PLATFORM}/ \
SUPPORT_ARCH_TIMER=no \
O=${B}
}
@@ -0,0 +1,23 @@
SUMMARY = "Cramfs tools"
DESCRIPTION = "Cramfs tools to create cramfs images with support to XIP"
HOMEPAGE = "https://github.com/npitre/cramfs-tools"
LICENSE = "GPLv2"
LIC_FILES_CHKSUM = "file://COPYING;md5=393a5ca445f6965873eca0259a17f833"
SRC_URI = "git://github.com/npitre/cramfs-tools;protocol=https;nobranch=1"
SRCREV = "23d032e6e0a973810c6aedf165441592641b60f4"
BBCLASSEXTEND = "native"
S = "${WORKDIR}/git"
do_compile () {
oe_runmake
}
do_install() {
install -d ${D}${bindir}
install ${S}/mkcramfs ${D}${bindir}
install ${S}/cramfsck ${D}${bindir}
}
@@ -0,0 +1,36 @@
# SPDX-License-Identifier: MIT
#
# Copyright (c) 2019 Arm Limited
#
# Recipe for fetching and installing the prebuilt gcc-arm toolchain for
# bare-metal targets.
# Recipe based on the meta-ti layer:
# http://git.yoctoproject.org/cgit/cgit.cgi/meta-ti/tree/recipes-ti/devtools/gcc-arm-none-eabi-native_4.9.2015q3.bb?h=master
#
# Whilst most of the tools built in relies on the linaro-provided
# toolchain, some recipes require the baremetal Arm GCC toolchain. The baremetal
# Arm GCC toolchain is distributed as precompiled binaries. This recipe fetches
# the toolchain and installs it in sysroot.
# Recipes may then 'DEPENDS' on this recipe to fetch the toolchain, making the
# toolchain available in the staging directory of the depending recipe.
GCC_ARM_VERSION = "gcc-arm-none-eabi-7-2018-q2-update"
LICENSE="GPLv3"
LIC_FILES_CHKSUM = "file://share/doc/gcc-arm-none-eabi/license.txt;md5=f77466c63f5787f4bd669c402aabe061"
inherit native
SRC_URI = "https://developer.arm.com/-/media/Files/downloads/gnu-rm/7-2018q2/${GCC_ARM_VERSION}-linux.tar.bz2;name=tc32"
SRC_URI[tc32.md5sum] = "299ebd3f1c2c90930d28ab82e5d8d6c0"
S = "${WORKDIR}/${GCC_ARM_VERSION}"
do_install() {
install -d ${D}${INSTALL_GCC_ARM_NONE_EABI}
cp -r ${S}/. ${D}${INSTALL_GCC_ARM_NONE_EABI}
}
FILES_${PN} = "${INSTALL_GCC_ARM_NONE_EABI}/"
INSANE_SKIP_${PN} = "already-stripped"
+10
View File
@@ -0,0 +1,10 @@
# SPDX-License-Identifier: MIT
#
# Copyright (c) 2019 Arm Limited
#
SRC_URI = "git://git.kernel.org/pub/scm/utils/dtc/dtc.git"
SRCREV = "65cc4d2748a2c2e6f27f1cf39e07a5dbabd80ebf"
LIC_FILES_CHKSUM = "file://GPL;md5=94d55d512a9ba36caa9b7df079bae19f"
@@ -0,0 +1,35 @@
# SPDX-License-Identifier: MIT
#
# Copyright (c) 2019 Arm Limited
#
DESCRIPTION = "Linux Kernel provided and supported by Arm and Linaro with \
focus on vexpress and other Arm models and hardware. Trying to be as much \
as possible close to upstream tree."
SECTION = "kernel"
LICENSE = "GPLv2"
SRCREV_FORMAT = "kernel"
require recipes-kernel/linux/linux-yocto.inc
KERNEL_VERSION_SANITY_SKIP = "1"
S = "${WORKDIR}/git"
COMPATIBLE_MACHINE = "(vexpress-a5|vexpress-a32|corstone700|a5ds)"
# Override do_kernel_configme to avoid kernel being assembled into a linux-yocto style kernel
# https://www.yoctoproject.org/docs/1.8/ref-manual/ref-manual.html#ref-tasks-kernel_configme
do_kernel_configme() {
}
KBRANCH = ""
LINUX_VERSION ?= "5.2"
LIC_FILES_CHKSUM = "file://COPYING;md5=bbea815ee2795b2f4230826c0c6b8814"
SRC_URI = "\
git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux;protocol=https;nobranch=1 \
"
SRCREV = "v5.2"
+8
View File
@@ -0,0 +1,8 @@
# SPDX-License-Identifier: MIT
#
# Copyright (c) 2019 Arm Limited
#
SUMMARY = "Linux Kernel provided and supported by Arm/Linaro"
require recipes-kernel/linux/linux-arm-common.inc
-92
View File
@@ -1,92 +0,0 @@
meta-arm-autonomy Yocto Layer
=============================
Introduction
------------
This layer provides an hypervisor based solution (currently based on Xen) for
autonomous system. It contains recipes and classes to build host and guests
systems.
To start using this layer, please check the
[Quick Start Guide](documentation/arm-autonomy-quickstart.md).
Dependencies
------------
This layer depends on several other Yocto layers:
* meta-openembedded (https://git.openembedded.org/meta-openembedded)
* poky (https://git.yoctoproject.org/poky)
* meta-virtualization (https://git.yoctoproject.org/meta-virtualization)
Distribution Features
---------------------
This layer is adding the following Yocto DISTRO_FEATURES:
* arm-autonomy-host: this feature activates functionalities required to build
an autonomy host system. It is doing the following:
- add 'xen' and 'ipv4' to DISTRO_FEATURES.
- add xen backend drivers to linux kernel configuration.
- To reduce the root filesystem image size the kernel image is not installed.
* arm-autonomy-guest: this feature activates functionalities to run as guest
of an autonomy system. It is doing the following:
- add 'ipv4' to DISTRO_FEATURES.
- add xen frontend drivers to linux kernel configuration.
- add console on hvc0 during init.
Bitbake variables
-----------------
Some recipes and classes in this layer are introducing variables which can be
modified by the user in local.conf.
Each recipe introducing such variables has a chapter "Bitbake parameters" in
its documentation.
Those documentation files should be checked for variables:
- [xen-devicetree](documentation/xen-devicetree.md)
- [xenguest-manager](documentation/xenguest-manager.md)
- [xenguest-network-bridge](documentation/xenguest-network-bridge.md)
BSPs
----
This layer is adding the following machines:
* arm64-autonomy-guest: this machines creates a minimal BSP suitable to be used
as an autonomy guest. It is in fact only activating ARM64 architecture and
SMP in the linux kernel and is enabling the DISTRO_FEATURE
arm-autonomy-guest.
Images
------
This layer is adding the following images:
* arm-autonomy-host-image-minimal: this image includes all elements required
to create a minimal arm-autonomy-host system. This includes xen and tools to
manage xen guests. This image depends on 'arm-autonomy-host' distribution
feature.
Recipes and classes
-------------------
This layer is adding the following recipes and classes:
* [xen-devicetree](documentation/xen-devicetree.md): this is a recipe to modify
a device tree blob to add information required to boot xen and a Dom0 linux.
* [xenguest-mkimage](documentation/xenguest-mkimage.md): this is a tool to
create and modify images to be used as Xen guests.
* [xenguest-manager](documentation/xenguest-manager.md): this is a tool to
create/remove/start/stop xen guest generated using xenguest-mkimage.
* [xenguest-network-bridge](documentation/xenguest-network-bridge.md): this
recipe add tools and init scripts to create a bridge connected to the
external network on the host and allow guests to be connected to it.
Contributing
------------
This project has not put in place a process for contributions currently. If you
would like to contribute, please contact the maintainers
Maintainer(s)
-------------
* Diego Sueiro <diego.sueiro@arm.com>
* Bertrand Marquis <bertrand.marquis@arm.com>
@@ -1,5 +0,0 @@
# Include arm-autonomy distro config files if the distro features are set
require ${@bb.utils.contains('DISTRO_FEATURES', 'arm-autonomy-host', '${ARM_AUTONOMY_DISTRO_CFGDIR}/arm-autonomy-host.inc', '', d)}
require ${@bb.utils.contains('DISTRO_FEATURES', 'arm-autonomy-guest', '${ARM_AUTONOMY_DISTRO_CFGDIR}/arm-autonomy-guest.inc', '', d)}
@@ -1,108 +0,0 @@
# Create a xenguest image with kernel and filesystem produced by Yocto
# This will create a .xenguest file that the xenguest-manager can use.
inherit xenguest-image
# We are creating our guest in a local subdirectory
# force the value so that we are not impacted if the user is changing it
XENGUEST_IMAGE_DEPLOY_DIR = "${WORKDIR}/tmp-xenguest"
# Name of deployed file (keep standard image name and add .xenguest)
XENGUEST_IMAGE_DEPLOY ??= "${IMAGE_NAME}"
# Add kernel XENGUEST_IMAGE_KERNEL from DEPLOY_DIR_IMAGE to image
xenguest_image_add_kernel() {
srcfile="${1:-}"
if [ -z "${srcfile}" ]; then
srcfile="${DEPLOY_DIR_IMAGE}/${XENGUEST_IMAGE_KERNEL}"
fi
call_xenguest_mkimage partial --xen-kernel=$srcfile
}
# Add rootfs file to the image
xenguest_image_add_rootfs() {
call_xenguest_mkimage partial \
--disk-add-file=${IMGDEPLOYDIR}/${IMAGE_LINK_NAME}.${IMAGE_TYPEDEP_xenguest}:rootfs.${IMAGE_TYPEDEP_xenguest}
}
# Pack xenguest image
xenguest_image_pack() {
mkdir -p ${IMGDEPLOYDIR}
rm -f ${IMGDEPLOYDIR}/${XENGUEST_IMAGE_DEPLOY}.xenguest
call_xenguest_mkimage pack \
${IMGDEPLOYDIR}/${XENGUEST_IMAGE_DEPLOY}.xenguest
}
#
# Task finishing the bootimg
# We need this task to actually create the symlinks
#
python do_bootimg_xenguest() {
subtasks = d.getVarFlag('do_bootimg_xenguest', 'subtasks')
bb.build.exec_func('xenguest_image_clone', d)
if subtasks:
for tk in subtasks.split():
bb.build.exec_func(tk, d)
bb.build.exec_func('xenguest_image_pack', d)
bb.build.exec_func('create_symlinks', d)
}
# This is used to add sub-tasks to do_bootimg_xenguest
do_bootimg_xenguest[subtasks] = ""
# Those are required by create_symlinks to find our image
do_bootimg_xenguest[subimages] = "xenguest"
do_bootimg_xenguest[imgsuffix] = "."
do_bootimg_xenguest[depends] += "xenguest-base-image:do_deploy"
# Need to have rootfs so all recipes have deployed their content
do_bootimg_xenguest[depends] += "${PN}:do_rootfs"
# This set in python anonymous after, just set a default value here
IMAGE_TYPEDEP_xenguest ?= "tar"
# We must not be built at rootfs build time because we need the kernel
IMAGE_TYPES_MASKED += "xenguest"
IMAGE_TYPES += "xenguest"
python __anonymous() {
# Do not do anything if we are not in the want FSTYPES
if bb.utils.contains_any('IMAGE_FSTYPES', 'xenguest', '1', '0', d):
# Check the coherency of the configuration
rootfs_needed = False
rootfs_file = ''
kernel_needed = False
rootfs_file = xenguest_image_rootfs_file(d)
if rootfs_file:
rootfs_needed = True
if d.getVar('XENGUEST_IMAGE_KERNEL') and not d.getVar('INITRAMFS_IMAGE'):
# If INITRAMFS_IMAGE is set, even if INITRAMFS_IMAGE_BUNDLE is not
# set to 1 to bundle the initramfs with the kernel, kernel.bbclass
# is setting a dependency on ${PN}:do_image_complete. We cannot
# in this case depend on do_deploy as it would create a circular
# dependency:
# do_image_complete would depend on kernel:do_deploy which would
# depend on ${PN}:do_image_complete
# In the case INITRAMFS_IMAGE_BUNDLE = 1, the kernel-xenguest class
# will handle the creation of a xenguest image with the kernel.
# In the other case the kernel can be added manually to the image.
kernel_needed = True
bb.build.addtask('do_bootimg_xenguest', 'do_image_complete', None, d)
if rootfs_needed:
# tell do_bootimg_xenguest to call add_rootfs
d.appendVarFlag('do_bootimg_xenguest', 'subtasks', ' xenguest_image_add_rootfs')
# do_bootimg_xenguest will need the tar file
d.appendVarFlag('do_bootimg_xenguest', 'depends', ' %s:do_image_tar' % (d.getVar('PN')))
# set our TYPEDEP to the proper compression
d.setVar('IMAGE_TYPEDEP_xenguest', 'tar' + (rootfs_file.split('.tar', 1)[1] or ''))
if kernel_needed:
# Tell do_bootimg_xenguest to call xenguest_image_add_kernel
d.appendVarFlag('do_bootimg_xenguest', 'subtasks', ' xenguest_image_add_kernel')
# we will need kernel do_deploy
d.appendVarFlag('do_bootimg_xenguest', 'depends', ' virtual/kernel:do_deploy')
}
@@ -1,34 +0,0 @@
# Create a xenguest image containing the kernel with initramfs when
# initramfs is activated
# This is done using kernel-fitimage as model
# To activate this, kernel-xenguest must be added to KERNEL_CLASSES
inherit xenguest-image
# use a local copy to pack all together
XENGUEST_IMAGE_DEPLOY_DIR = "${WORKDIR}/tmp-xenguest"
python __anonymous () {
# only if xenguest image type is present
if bb.utils.contains('IMAGE_FSTYPES', 'xenguest', '1', '0', d):
# only if initramfs bundle is activated
if d.getVar('INITRAMFS_IMAGE') and d.getVar('INITRAMFS_IMAGE_BUNDLE') == "1":
if not bb.utils.contains('KERNEL_IMAGETYPES', 'Image', '1', '0', d):
bb.fatal("xenguest image type with initramfs require Image kernel image type")
bb.build.addtask('do_assemble_xenguest_initramfs', 'do_deploy', 'do_bundle_initramfs', d)
}
do_assemble_xenguest_initramfs() {
xenguest_image_clone
call_xenguest_mkimage partial --xen-kernel=${B}/${KERNEL_OUTPUT_DIR}/Image.initramfs
rm -f ${B}/${KERNEL_OUTPUT_DIR}/Image-initramfs.xenguest
call_xenguest_mkimage pack ${B}/${KERNEL_OUTPUT_DIR}/Image-initramfs.xenguest
}
do_assemble_xenguest_initramfs[depends] += "${INITRAMFS_IMAGE}:do_image_complete"
kernel_do_deploy_append() {
if [ -f "${B}/${KERNEL_OUTPUT_DIR}/Image-initramfs.xenguest" ]; then
install -m 0644 ${B}/${KERNEL_OUTPUT_DIR}/Image-initramfs.xenguest "$deployDir/Image-${INITRAMFS_NAME}.xenguest"
ln -snf Image-${INITRAMFS_NAME}.xenguest $deployDir/Image-${INITRAMFS_LINK_NAME}.xenguest
fi
}
@@ -1,114 +0,0 @@
# This class must be used to extend the xenguest image
# It provides variables to add init scripts, a dtb, xen files or disk files.
#
# The class is extending deploy function so you recipe must inherit deploy and
# have a do_deploy function (even if it is empty)
# Use standard xenguest-image
inherit xenguest-image
# Add a DTB file for the guest
# Only one file should be added, if this is set multiple times or in several
# recipes, the last recipe setting it will prevail.
XENGUEST_EXTRA_DTB ??= ""
# Add a ramdisk file for the guest
# Only one file should be added, if this is set multiple times or in several
# recipes, the last recipe setting it will prevail.
XENGUEST_EXTRA_RAMDISK ??= ""
# Append something to the guest xen configuration
# All files here will be merged together in the final xen configuration
# This can contain several files or be used in several recipes
XENGUEST_EXTRA_XENCONFIG ??= ""
# Add a xenguest init, init-pre or init-post script
XENGUEST_EXTRA_INIT_PRE ??= ""
XENGUEST_EXTRA_INIT ??= ""
XENGUEST_EXTRA_INIT_POST ??= ""
# Add xenguest files, (to be used in extra xen config for example)
# several files may be added, space separated, the path will be kept on the
# generated xenguest image (if dir1/file1 is added, it can be used as
# dir1/file1 file in the xen configuration).
XENGUEST_EXTRA_FILES ??= ""
# Add xenguest disk files (to be used as disk partition content)
# several files may be added, space separated, the path will be kept on the
# generated xenguest image (if dir1/file1 is added, it can be used as
# dir1/file1 file in the disk content parameters).
XENGUEST_EXTRA_DISK_FILES ??= ""
do_deploy_append() {
if [ -z "${XENGUEST_IMAGE_DEPLOY_DIR}" -o \
-z "${XENGUEST_IMAGE_DEPLOY_SUBDIR}" ]; then
die "Configuration error: XENGUEST_IMAGE_DEPLOY_DIR or XENGUEST_IMAGE_DEPLOY_SUBDIR is empty"
fi
rm -rf ${XENGUEST_IMAGE_DEPLOY_DIR}/${XENGUEST_IMAGE_DEPLOY_SUBDIR}
mkdir -p ${XENGUEST_IMAGE_DEPLOY_DIR}/${XENGUEST_IMAGE_DEPLOY_SUBDIR}
if [ -n "${XENGUEST_EXTRA_DTB}" ]; then
if [ ! -f ${XENGUEST_EXTRA_DTB} ]; then
die "xenguest-image: DTB file ${XENGUEST_EXTRA_DTB} does not exist"
fi
call_xenguest_mkimage partial --xen-device-tree=${XENGUEST_EXTRA_DTB}
fi
if [ -n "${XENGUEST_EXTRA_RAMDISK}" ]; then
if [ ! -f ${XENGUEST_EXTRA_RAMDISK} ]; then
die "xenguest-image: DTB file ${XENGUEST_EXTRA_RAMDISK} does not exist"
fi
call_xenguest_mkimage partial --xen-ramdisk=${XENGUEST_EXTRA_RAMDISK}
fi
if [ -n "${XENGUEST_EXTRA_XENCONFIG}" ]; then
for f in ${XENGUEST_EXTRA_XENCONFIG}; do
if [ ! -f $f ]; then
die "xenguest-image: Xen config $f does not exist"
fi
call_xenguest_mkimage partial --xen-append=$f
done
fi
if [ -n "${XENGUEST_EXTRA_INIT_PRE}" ]; then
if [ ! -f ${XENGUEST_EXTRA_INIT_PRE} ]; then
die "xenguest-image: Init script ${XENGUEST_EXTRA_INIT_PRE} does not exist"
fi
call_xenguest_mkimage partial --init-pre=${XENGUEST_EXTRA_INIT_PRE}
fi
if [ -n "${XENGUEST_EXTRA_INIT}" ]; then
if [ ! -f ${XENGUEST_EXTRA_INIT} ]; then
die "xenguest-image: Init script ${XENGUEST_EXTRA_INIT} does not exist"
fi
call_xenguest_mkimage partial --init-script=${XENGUEST_EXTRA_INIT}
fi
if [ -n "${XENGUEST_EXTRA_INIT_POST}" ]; then
if [ ! -f ${XENGUEST_EXTRA_INIT_POST} ]; then
die "xenguest-image: Init script ${XENGUEST_EXTRA_INIT_POST} does not exist"
fi
call_xenguest_mkimage partial --init-post=${XENGUEST_EXTRA_INIT_POST}
fi
if [ -n "${XENGUEST_EXTRA_FILES}" ]; then
for f in ${XENGUEST_EXTRA_FILES}; do
if [ ! -f $f ]; then
die "xenguest-image: Xen file $f does not exist"
fi
call_xenguest_mkimage partial --xen-add-file=$f
done
fi
if [ -n "${XENGUEST_EXTRA_DISK_FILES}" ]; then
for f in ${XENGUEST_EXTRA_DISK_FILES}; do
if [ ! -f $f ]; then
die "xenguest-image: Disk file $f does not exist"
fi
call_xenguest_mkimage partial --disk-add-file=$f
done
fi
}
# Need to have xenguest-image tool
do_deploy[depends] += "xenguest-base-image:do_deploy"
@@ -1,199 +0,0 @@
# This class must be used to create, extend or pack a xenguest image.
# It is using xenguest-mkimage tool to do operations
DEPENDS += "xenguest-mkimage-native"
#
# Xenguest image parameters
# All the following parameters can be modified in local.conf or on recipes
# inheriting this class
#
# Guest memory size in MB
XENGUEST_IMAGE_MEMORY_SIZE ??= "512"
# Guest number of vcpus
XENGUEST_IMAGE_NUM_VCPUS ??= "1"
# Guest auto boot during init, set to 1 to have guest started during init or
# to 0 if the guest should not be auto started
XENGUEST_IMAGE_AUTOBOOT ??= "1"
# Partition containing the root file system
# Xen will actually add root=${XENGUEST_IMAGE_ROOT} to your guest kernel
# command line
# You can let this empty if the root filesystem is specified in an other way
# and have root= option added to the command line for example or if you don't
# need a root filesystem mounted for your guest (initrd for example)
XENGUEST_IMAGE_ROOT ??= "/dev/xvda1"
# Guest kernel command line arguments
XENGUEST_IMAGE_CMDLINE ??= "earlyprintk=xenboot console=hvc0 rw"
# Extra commands to add to xenguest-image when creating the image
XENGUEST_IMAGE_EXTRA_CMD ??= ""
# Kernel binary
# This value is used by the xenguest image type but is declared here to have
# all variables in the same place
# If this value is empty no kernel will be added to the image
XENGUEST_IMAGE_KERNEL ??= "Image"
# Size of the disk to create (if 0 no disk will be created and rootfs will not
# be included in the xenguest image)
XENGUEST_IMAGE_DISK_SIZE ??= "${@ '4' if not d.getVar('INITRAMFS_IMAGE') else '0'}"
#
# XENGUEST_IMAGE_DISK PARTITIONS is used to describe the partitions to setup
# and their content.
# It must be set to a space separated list of entries with each entry having
# the format num:sz:fs:[file] where:
# - num is a partition number
# - sz is the partition size in Gigabit
# - fs is the filesystem to use for the partition
# - file is optionally pointing to a file to use as content of the partition
# Please check image_types_xenguest.bbclass for rootfs handling of files
#
# Default value creates a partition 1 using the full disk, formated with ext4
# and containing the root filesystem produced by Yocto
XENGUEST_IMAGE_DISK_PARTITIONS ??= "1:${XENGUEST_IMAGE_DISK_SIZE}:ext4:rootfs.tar.gz"
# XENGUEST_IMAGE_NETWORK_TYPE can be set to "bridge", "nat" or "none".
# The "bridge" type will share the physical eth interface from dom0 with the
# domU. This will allow the domU to have access to the external network.
# The "nat" type will setup a virtual network between dom0 and domU and also
# configure and run the kea dhcp4 server on dom0 to serve the domU.
# The "none" type will not affect any networking setting between on dom0 and
# domU.
XENGUEST_IMAGE_NETWORK_TYPE ??= "bridge"
# Sub-directory in wich the guest is created. This is create in deploy as a
# subdirectory and must be coherent between all components using this class so
# it must only be modified from local.conf if needed
XENGUEST_IMAGE_DEPLOY_SUBDIR ?= "xenguest"
# Directory in which the xenguest should be deployed
# a sub-directory named ${XENGUEST_IMAGE_DEPLOY_SUBDIR} will be created there.
# This should be set to:
# - ${DEPLOYDIR} (default) if creating or extending the xenguest for a normal
# recipe.
# - something in ${WORKDIR} if you need to clone and manipulate an image
XENGUEST_IMAGE_DEPLOY_DIR ??= "${DEPLOYDIR}"
#
# Wrapper to call xenguest-mkimage
# It is using XENGUEST_IMAGE_DEPLOY_DIR and XENGUEST_IMAGE_DEPLOY_SUBDIR
# to find the image to operate on
#
# Usage: call_xenguest_mkimage [operation] [args]
call_xenguest_mkimage() {
local cmd="${1}"
local img="${XENGUEST_IMAGE_DEPLOY_DIR}/${XENGUEST_IMAGE_DEPLOY_SUBDIR}"
shift
echo "xenguest-mkimage $cmd $img $@"
xenguest-mkimage $cmd $img $@
}
#
# Create an initial xenguest image.
# This is a task which must be added in a recipe inheriting deploy
# It is using XENGUEST_IMAGE_MEMORY_SIZE, XENGUEST_IMAGE_NUM_VCPUS,
# XENGUEST_IMAGE_AUTOBOOT, XENGUEST_IMAGE_ROOT, XENGUEST_IMAGE_EXTRA_CMD,
# XENGUEST_IMAGE_CMDLINE, XENGUEST_IMAGE_DISK_SIZE and
# XENGUEST_IMAGE_DISK_PARTITIONS to customize the initial guest
#
xenguest_image_create() {
if [ -z "${XENGUEST_IMAGE_DEPLOY_DIR}" -o \
-z "${XENGUEST_IMAGE_DEPLOY_SUBDIR}" ]; then
die "Configuration error: XENGUEST_IMAGE_DEPLOY_DIR or XENGUEST_IMAGE_DEPLOY_SUBDIR is empty"
fi
rm -rf ${XENGUEST_IMAGE_DEPLOY_DIR}/${XENGUEST_IMAGE_DEPLOY_SUBDIR}
mkdir -p ${XENGUEST_IMAGE_DEPLOY_DIR}/${XENGUEST_IMAGE_DEPLOY_SUBDIR}
# Create the image
call_xenguest_mkimage create --xen-memory=${XENGUEST_IMAGE_MEMORY_SIZE} \
--xen-vcpus=${XENGUEST_IMAGE_NUM_VCPUS} \
--xen-root=${XENGUEST_IMAGE_ROOT} \
${XENGUEST_IMAGE_EXTRA_CMD}
# add command line
if [ -n "${XENGUEST_IMAGE_CMDLINE}" ]; then
call_xenguest_mkimage update --xen-clean-extra
for arg in ${XENGUEST_IMAGE_CMDLINE}; do
call_xenguest_mkimage update --xen-extra=$arg
done
fi
# create disk if needed
disksize="${XENGUEST_IMAGE_DISK_SIZE}"
if [ -z "$disksize" ]; then
disksize="0"
fi
if [ $disksize -gt 0 ]; then
# setup disk size
call_xenguest_mkimage update --disk-reset-config --disk-size=$disksize
diskparts="${XENGUEST_IMAGE_DISK_PARTITIONS}"
if [ -n "$diskparts" ]; then
for arg in $diskparts; do
call_xenguest_mkimage update --disk-add-part=$arg
partnum="$(expr $partnum + 1)"
done
fi
fi
if [ "${XENGUEST_IMAGE_AUTOBOOT}" = "1" ]; then
call_xenguest_mkimage update --set-param=GUEST_AUTOBOOT=1
else
call_xenguest_mkimage update --set-param=GUEST_AUTOBOOT=0
fi
if [ -n "${XENGUEST_IMAGE_NETWORK_TYPE}" ]; then
call_xenguest_mkimage update --set-param=XENGUEST_NETWORK_TYPE="${XENGUEST_IMAGE_NETWORK_TYPE}"
else
call_xenguest_mkimage update --set-param=XENGUEST_NETWORK_TYPE="none"
fi
}
#
# Clone the current xenguest from deploy to manipulate it locally
# This is required if you need to change things before packing an image
# To set the local directory where to clone you must set
# XENGUEST_IMAGE_DEPLOY_DIR if you don't want to use do_deploy to modify the
# image
#
xenguest_image_clone() {
if [ -z "${XENGUEST_IMAGE_DEPLOY_DIR}" -o \
-z "${XENGUEST_IMAGE_DEPLOY_SUBDIR}" ]; then
die "Configuration error: XENGUEST_IMAGE_DEPLOY_DIR or XENGUEST_IMAGE_DEPLOY_SUBDIR is empty"
fi
if [ ! -f ${DEPLOY_DIR_IMAGE}/${XENGUEST_IMAGE_DEPLOY_SUBDIR}/guest.cfg ]; then
die "xenguest-image: ${DEPLOY_DIR_IMAGE}/${XENGUEST_IMAGE_DEPLOY_SUBDIR} does not contain a valid guest"
fi
rm -rf ${XENGUEST_IMAGE_DEPLOY_DIR}/${XENGUEST_IMAGE_DEPLOY_SUBDIR}
mkdir -p ${XENGUEST_IMAGE_DEPLOY_DIR}
cp -rf ${DEPLOY_DIR_IMAGE}/${XENGUEST_IMAGE_DEPLOY_SUBDIR} \
${XENGUEST_IMAGE_DEPLOY_DIR}/${XENGUEST_IMAGE_DEPLOY_SUBDIR}
}
# Helper function to retrieve rootfs file if present in one partition
# This can return an empty string or rootfs.tar[.COMP]
def xenguest_image_rootfs_file(d):
disksize = d.getVar('XENGUEST_IMAGE_DISK_SIZE')
# if disksize is 0, we don't create anything
if not disksize or disksize == '0':
return ""
# Find first partition with file=rootfs.tar*
partlist = d.getVar('XENGUEST_IMAGE_DISK_PARTITIONS')
if partlist:
for partdesc in partlist.split():
partelems = partdesc.split(':', 3)
if partelems[3]:
if partelems[3].startswith('rootfs.tar'):
return partelems[3]
return ""
@@ -1,15 +0,0 @@
# This files is added when DISTRO_FEATURES contains arm-autonomy-guest
# We need to have ipv4 activated
DISTRO_FEATURES_append = " ipv4"
# Build a xenguest type image
IMAGE_CLASSES += "image_types_xenguest"
IMAGE_FSTYPES += "xenguest"
# xenguest kernel extension to handle initramfs
KERNEL_CLASSES += "kernel-xenguest"
IMAGE_INSTALL_append = "${@bb.utils.contains('DISTRO_FEATURES', 'docker', \
' packagegroup-docker-runtime-minimal', \
'', d)}"
@@ -1,21 +0,0 @@
# This files is added when DISTRO_FEATURES contains arm-autonomy-host
# We need to have xen and ipv4 activated
DISTRO_FEATURES_append = " xen ipv4"
DISTRO_FEATURES_NATIVE_append = " arm-autonomy-host"
# Don't include kernels in standard images when building arm-autonomy-host
# If the kernel image is needed in the rootfs the following should be set from
# a bbappend: RDEPENDS_${KERNEL_PACKAGE_NAME}-base = "${KERNEL_PACKAGE_NAME}-image"
RDEPENDS_${KERNEL_PACKAGE_NAME}-base ?= ""
# Require extra machine specific settings from meta-arm-bsp dynamic-layers only
# if meta-arm-bsp is in the bblayers.conf
# Directory for meta-arm-autonomy/dynamic-layers/meta-arm-bsp machine extra settings
ARM_AUTONOMY_ARM_BSP_DYNAMIC_EXTRA_CFGDIR = "${ARM_AUTONOMY_ARM_BSP_DYNAMIC_DIR}/conf/machine"
ARM_AUTONOMY_MACHINE_EXTRA_REQUIRE ?= \
"${ARM_AUTONOMY_ARM_BSP_DYNAMIC_EXTRA_CFGDIR}/arm-autonomy-machine-extra-settings.inc"
require ${@bb.utils.contains('BBFILE_COLLECTIONS', 'meta-arm-bsp', \
'${ARM_AUTONOMY_MACHINE_EXTRA_REQUIRE}' , \
'', d)}
-38
View File
@@ -1,38 +0,0 @@
# Add layer to BBPATH
BBPATH .= ":${LAYERDIR}"
# Add recipes-* directories to BBFILES
BBFILES += " \
${LAYERDIR}/recipes-*/*/*.bb \
${LAYERDIR}/recipes-*/*/*.bbappend \
"
BBFILE_COLLECTIONS += "meta-arm-autonomy"
BBFILE_PATTERN_meta-arm-autonomy = "^${LAYERDIR}/"
BBFILE_PRIORITY_meta-arm-autonomy = "6"
LAYERDEPENDS_meta-arm-autonomy = " \
core \
yocto \
openembedded-layer \
virtualization-layer \
"
LAYERSERIES_COMPAT_meta-arm-autonomy = "gatesgarth"
# We don't activate virtualization feature from meta-virtualization as it
# brings in lots of stuff we don't need. We need to disable the sanity check
# otherwise the user will see a warning on each build.
SKIP_META_VIRT_SANITY_CHECK = "1"
ARM_AUTONOMY_LAYERDIR := "${LAYERDIR}"
# Directory of our distro config files
ARM_AUTONOMY_DISTRO_CFGDIR = "${ARM_AUTONOMY_LAYERDIR}/conf/distro/include/"
# Add class to handle arm-autonomy distro extensions
USER_CLASSES_append = " arm-autonomy-features"
BBFILES_DYNAMIC += " \
meta-arm-bsp:${LAYERDIR}/dynamic-layers/meta-arm-bsp/*/*/*.bbappend \
meta-gem5:${LAYERDIR}/dynamic-layers/meta-gem5/*/*/*.bbappend \
"
# Root directory for the meta-arm-autonomy/dynamic-layers/meta-arm-bsp
ARM_AUTONOMY_ARM_BSP_DYNAMIC_DIR = "${ARM_AUTONOMY_LAYERDIR}/dynamic-layers/meta-arm-bsp"
@@ -1,17 +0,0 @@
#@TYPE: Machine
#@NAME: Autonomy Guest ARM64 machine
#@DESCRIPTION: Machine configuration for ARM64 Autonomy Guest
TUNE_FEATURES = "aarch64"
require conf/machine/include/arm/arch-armv8a.inc
KERNEL_IMAGETYPE = "Image"
IMAGE_FSTYPES += "tar.bz2"
PREFERRED_PROVIDER_virtual/kernel = "linux-yocto"
PREFERRED_VERSION_linux-yocto ?= "5.4%"
DISTRO_FEATURES += "arm-autonomy-guest"
@@ -1,232 +0,0 @@
arm-autonomy Quick Start
==================
This documentation is explaining how to quickly start with arm-autonomy layer
and the main features provided.
You will find in the documentation directory some more detailed documentation
for each of the functionalites provided by this layer.
What to use this layer for?
---------------------------
Using this layer, you can easily and rapidly create a system based on Xen with
one or more guests created using Yocto.
For this you will need to create at least 2 Yocto projects:
- a host project: This one will compile Xen and create a Linux system to be
used as Xen Dom0. The Linux system will contain all functionalities required
to start and manage guests.
- one or several guest projects: Those will create Linux systems with the
required Linux kernel configuration to run as Xen DomU.
Prepare your system
-------------------
First you must download the Yocto layers needed:
- [meta-openembedded](https://git.openembedded.org/meta-openembedded)
- [poky](https://git.yoctoproject.org/poky)
- [meta-virtualization](https://git.yoctoproject.org/meta-virtualization)
- [meta-arm](https://git.yoctoproject.org/meta-arm)
- [meta-kernel](https://gitlab.com/openembedded/community/meta-kernel.git)
- all other layers you might want to use
For each of the downloaded layer make sure you checkout the release of Yocto
you want to use (for example zeus using `git checkout zeus`).
Please follow [Yocto documentation](https://www.yoctoproject.org/docs/latest/brief-yoctoprojectqs/brief-yoctoprojectqs.html)
in order to have the required dependencies.
Create a project
----------------
Here are the main steps to create an arm-autonomy project:
1. create a new Yocto project using `oe-init-build-env` in a new directory:
```
oe-init-build-env my-project
```
2. Add `meta-arm/meta-arm-autonomy` layer to the list of layers of your
project in the `conf/bblayers.conf`. Also add any other layers you
might need (for example `meta-arm/meta-arm-bsp` and `meta-arm/meta-arm` to
use Arm boards like Juno or FVP emulator). You can achieve this by using
the `bitbake-layers add-layer layerdir [layerdir ...]` command.
For example:
```
export LAYERDIR_BASE="/home/user/arm-autonomy/"
bitbake-layers add-layer $LAYERDIR_BASE/meta-poky $LAYERDIR_BASE/meta-yocto-bsp \
$LAYERDIR_BASE/meta-openembedded/meta-oe $LAYERDIR_BASE/meta-openembedded/meta-python \
$LAYERDIR_BASE/meta-openembedded/meta-filesystems $LAYERDIR_BASE/meta-openembedded/meta-networking \
$LAYERDIR_BASE/meta-virtualization $LAYERDIR_BASE/meta-kernel \
$LAYERDIR_BASE/meta-arm/meta-arm $LAYERDIR_BASE/meta-arm/meta-arm-toolchain \
$LAYERDIR_BASE/meta-arm/meta-arm-bsp $LAYERDIR_BASE/meta-arm/meta-arm-autonomy \
```
Example of a `conf/bblayers.conf`:
```
BBLAYERS ?= " \
/home/user/arm-autonomy/poky/meta \
/home/user/arm-autonomy/poky/meta-poky \
/home/user/arm-autonomy/poky/meta-yocto-bsp \
/home/user/arm-autonomy/meta-openembedded/meta-oe \
/home/user/arm-autonomy/meta-openembedded/meta-python \
/home/user/arm-autonomy/meta-openembedded/meta-filesystems \
/home/user/arm-autonomy/meta-openembedded/meta-networking \
/home/user/arm-autonomy/meta-virtualization \
/home/user/arm-autonomy/meta-kernel \
/home/user/arm-autonomy/meta-arm/meta-arm \
/home/user/arm-autonomy/meta-arm/meta-arm-toolchain \
/home/user/arm-autonomy/meta-arm/meta-arm-bsp \
/home/user/arm-autonomy/meta-arm/meta-arm-autonomy \
"
```
Be aware that changing the order may break some dependencies if editing the
config file manually.
Those steps will have to be done for each project you will have to create.
Host project
------------
The host project will build Xen and the Dom0 Linux. It will be the only project
that will be specific to the board (MACHINE) you will be running on.
To create a host project:
1. Follow the steps of "Create a project"
2. Add the layers in `bblayers.conf` required to build a Yocto project for the
board you want to use.
For example to use Arm FVP Base emulator, add `meta-arm/meta-arm` and
`meta-arm/meta-arm-bsp`.
3. edit conf/local.conf to add `arm-autonomy-host` to the DISTRO_FEATURES and
set MACHINE to the board you want to use.
For example, add the following lines:
```
MACHINE = "fvp-base"
DISTRO_FEATURES += "arm-autonomy-host"
```
4. build the image using `bitbake arm-autonomy-host-image-minimal`
The project will generate a Linux kernel, a root filesystem, a Xen binary and
a DTB modified to include the required entries to boot Xen and Linux as Dom0
(this DTB has the extension `-xen.dtb`).
To boot the system using an u-boot base board you will need to:
- Load the kernel (by default at 0x80080000 unless you modify
XEN_DEVICETREE_DOM0_ADDR value)
- Load the xen device tree (for example at 0x83000000)
- Load the xen-efi binary (for example at 0x84000000)
- run using `booti 0x84000000 - 0x83000000`
In this example the addresses might need to be adapted depending on your board.
Guest project
-------------
The guest projects are not target specific and will use a Yocto MACHINE defined
in meta-arm-autonomy to include only the Linux configuration required to run
a xen guest.
To create a guest project:
1. Follow the steps of "Create a project"
2. Optionaly add layers required to build the image and features you need.
3. edit conf/local.conf to add `arm-autonomy-guest` to the DISTRO_FEATURES and
set MACHINE to `arm64-autonomy-guest`:
```
MACHINE = "arm64-autonomy-guest"
DISTRO_FEATURES += "arm-autonomy-guest"
```
4. build the image you want.
For example `bitbake core-image-minimal`
The build will create a ".xenguest" image that can be use on an host project
with the xenguest-manager.
Include guests directly in the host image
-----------------------------------------
The layer provides a way to directly include in the host project one or several
images generated by guest projects.
To use this feature, you must edit your host project `local.conf` file and
add set ARM_AUTONOMY_HOST_IMAGE_EXTERN_GUESTS to the list of xenguest images
you want to include in your host. Each xenguest image must be given using a
full path to it.
For example:
```
ARM_AUTONOMY_HOST_IMAGE_EXTERN_GUESTS = "/home/user/guest-project/tmp/deploy/images/arm64-autonomy-guest/core-image-minimal-arm64-autonomy-guest.xenguest;guestname=myguest"
```
This will add the guest and name it `myguest` on the host project image and
the xenguest-manager will automatically boot it during startup.
Add support for your board
--------------------------
Most of arm-autonomy layer is board independent but some functionalities
might need to be customized for your board:
### Add the kernel configuration for the host
The layer is using KERNEL_FEATURES to add drivers required to be a Xen Dom0
system.
Depending on the kernel used by your BSP and how it is configured you might
need to add the required drivers to your kernel configuration:
- if KERNEL_FEATURES system is supported by your kernel, make sure that the
file `recipes-kernel/linux/linux-arm-autonomy.inc` from the layer is included
by your kernel recipe.
- if it is not supported, you must add the proper drivers inside your kernel
(modules are possible but they must be loaded before xenguest-manager is
started). You can find the complete list of the kernel configuration elements
required in `recipes-kernel/linux/arm-autonomy-kmeta/features/arm-autonomy/xen-host.cfg`.
### Define the drive and partition to use for the LVM volume
The xenguest-manager is creating disk hard drive using LVM on an empty
partition. The default value is set to use /dev/sda2.
You can change this for your board by setting XENGUEST_MANAGER_VOLUME_DEVICE.
Check `recipes-extended/xenguest/xenguest-manager.bbappend` for examples.
Please also read xenguest-manager.md.
### Define the interface to add to xenguest network bridge
xenguest-network bridge is creating a bridge on the host and adds network
interfaces to it so that guest connected to it have access to external network.
By default `eth0` is set as the list of interfaces to be added to the bridge.
Depending on your board or use case you might want to use an other interface
or use multiple interfaces.
You can change this for your board by setting XENGUEST_NETWORK_BRIDGE_MEMBERS.
Check `recipes-extended/xenguest/xenguest-network-bridge.bbappend` for
exmaples.
Please also read xenguest-network-bridge.md.
### Define the network configuration of the xenguest network bridge
xenguest network bridge is putting the host network interfaces in a bridge
and is configuring it by default to use dhcp.
If you need a different type of configuration you can set
XENGUEST_NETWORK_BRIDGE_CONFIG in a xenguest-network-bridge.bbappend to use
a different file.
The recipe will look for the file in ${WORKDIR} so you will need to add it to
SRC_URI in your bbappend.
The recipe will also substitute `###BRIDGE_NAME###` with the bridge name
configured in ${XENGUEST_NETWORK_BRIDGE_NAME}.
You can find an example configuration file in
`recipes-extended/xenguest/files/xenguest-network-bridge-dhcp.cfg.in`.
Please also read xenguest-network-bridge.md.
### Customize Dom0 and Xen boot arguments for you board
xen-devicetree is writting inside the generated DTB Xen and Linux boot
arguments as long as the address where Dom0 Linux kernel can be found.
You might need to have different values for your board or depending on your
use case.
You can find examples to customize this in
`recipes-extended/xen-devicetree/xen-devicetree.bbappend`.
Please also read xen-devicetree.md.
@@ -1,29 +0,0 @@
Customizing Arm Autonomy Host image layout for N1SDP
====================================================
When buiding with `DISTRO_FEATURES += "arm-autonomy-host"` the user can
perform a couple of customizations in the generated wic image:
1. Set the guest partition size (default: 4iG) via `GUEST_PART_SIZE` and
`GUEST_PART_SIZE_UNIT` (M or G) variables to be set in any conf file. The
value of these variables should be aligned with the sum of all
XENGUEST_IMAGE_DISK_SIZE set for the guests. By default, LVM2 metadata is
1 MiB per physical volume, hence it needs to be taken into account when
setting GUEST_PART_SIZE.
2. The wic image partition layout and contents with a custom wks file via
`ARM_AUTONOMY_WKS_FILE` variable (default:
arm-autonomy-n1sdp-efidisk.wks.in which is affected by GUEST_PART_SIZE,
GUEST_PART_SIZE_UNIT and GRUB_CFG_FILE variables).
3. Custom grub.cfg file via `GRUB_CFG_FILE` (default:
arm-autonomy-n1sdp-grub.cfg) variable to be set in any conf file. The full
path or relative to `ARM_AUTONOMY_WKS_FILE` should be set.
The `arm-autonomy-n1sdp-efidisk.wks.in` and `arm-autonomy-n1sdp-grub.cfg` files
are located at `meta-arm-autonomy/dynamic-layers/meta-arm-bsp/wic`.
Other variables can also be custmized to set what files need to be included
in the wic image boot partition. Please refer to
`meta-arm-autonomy/dynamic-layers/meta-arm-bsp/conf/machine/n1sdp-extra-settings.inc`
for more details.
@@ -1,104 +0,0 @@
Xen device tree
===============
Introduction
------------
xen-devicetree recipe can be used to modify an existing Device Tree Blob,
produced by Linux kernel or another recipe, to include in it entries required
to start Xen and a Dom0 Linux on top of it.
The recipe will do the following processing:
- generate a xen.dtsi file with entries required for xen based on parameters
- turn DTBs in ${XEN_DEVICETREE_DTBS} back to dts
- include in xen.dtsi and all other dtsi from
${XEN_DEVICETREE_DTSI_MERGE} (check Bitbake parameters for more information
on this).
- use dtc to create a new DTB file
The recipe will generate new files in deploy/images and name them with the
extension "-xen.dtb".
For example **machine.dtb** will generate **machine-xen.dtb**.
Entries added to the DTBs
-------------------------
When a DTB is processed by the recipe, the following block is added:
```
chosen {
xen,dom0-bootargs = "VAL";
xen,xen-bootargs = "VAL";
modules {
#size-cells = <0x00000001>;
#address-cells = <0x00000001>;
module@0 {
reg = <VAL VAL>;
compatible = "multiboot,module";
};
};
};
```
Each occurence of **VAL** is replaced by the content of the variables listed in
this documentation.
Bitbake parameters
------------------
Several parameters are available to configure the xen-devicetree during Yocto
project compilation (those can be set in your project local.conf, for exmple).
The following parameters are available:
- XEN_DEVICETREE_DEPEND: This variable can be used to indicate which recipe
task is generating the DTBs that xen-devicetree will modify. This makes sure
the xen devicetrees are properly regenerated if the source DTBs are changed.
This variable is set by default to "virtual/kernel:do_deploy" to use DTBs
generated during the compilation of the Linux kernel. This must be changed
if the machine your are using is not using a DTB listed in KERNEL_DEVICETREE.
- XEN_DEVICETREE_DTBS: This should be set to the list of DTBs you want to be
modified by xen-devicetree. Those must be files that xen-devicetree can find
in the ${DEPLOY_DIR_IMAGE} directory using only the basename of the entries.
For example "mydir/board.dtb" will make the recipe look for
${DEPLOY_DIR_IMAGE}/board.dtb.
This variable is set by default to "${KERNEL_DEVICETREE}" to process the DTBs
generated by the Linux kernel.
- XEN_DEVICETREE_DOM0_MEM: Memory size to allocate to Dom0.
This variable is only used if XEN_DEVICETREE_XEN_BOOTARGS has a value
containing "dom0_mem=${XEN_DEVICETREE_DOM0_MEM}" as the memory assigned to
dom0 is defined using Xen boot arguments.
This variable is set by default to "1024M".
- XEN_DEVICETREE_DOM0_BOOTARGS: Boot arguments to pass to Dom0 Linux when
booting it.
This variable is set by default to "console=hvc0 earlycon=xen".
- XEN_DEVICETREE_XEN_BOOTARGS: this variable should be set with the boot
arguments to be passed to Xen on boot.
This variable is set by default to
"noreboot dom0_mem=${XEN_DEVICETREE_DOM0_MEM}".
- XEN_DEVICETREE_DOM0_ADDR: This is the address from which the Linux kernel to
be used for Dom0 will be copied. When using u-boot, this is the address at
which you will load the kernel Image before starting Xen.
This variable is set by default to "0x80080000".
- XEN_DEVICETREE_DOM0_SIZE: This is the size of the kernel loaded at
${XEN_DEVICETREE_DOM0_ADDR}. Xen will copy this amount of data inside the
guest before starting it so the size must be at least equal to the kernel
size but can be bigger. You must be careful not to have a value too big as it
could slow down boot or copy other parts with it (like the DTB).
You might need to increase this if you use a kernel with a bundled initramfs.
This variable is set by default to "0x01000000".
- XEN_DEVICETREE_DTSI_MERGE: This variable contains the list of dtsi files that
must be included inside the generated DTB file. By default the only one
include is the "xen.dtsi" generated by the recipe.
If your board or project needs to include more fixes or entries in the DTB,
this variable can be appended from a bbappend file to include other dtsi. The
files must be inside the recipe workspace during Yocto compilation.
You can check xen-devicetree.bbappend for an example.
@@ -1,101 +0,0 @@
Xenguest Manager
================
Introduction
------------
xenguest-manager is a tool to manage Xenguest images generated by
[xenguest-mkimage](xenguest-mkimage.md).
On a Xen Dom0 system it will:
- create a xen guest from a xenguest image: extract its components, create a
disk for the guest using LVM volumes.
- start/stop a xen guest (during init or using xenguest-manager directly).
- check guest status
xenguest-manager is composed of 2 shell scripts:
- xenguest-manager which can be used from command line to start/stop/check
guests and create or remove guest using xenguest images.
- xenguest-init which is called during init to automatically create and start
some guests as part of the host init process.
Usage
-----
xenguest-manager must be called like this:
`xenguest-manager OPERATION [OPTIONS]`
The following operations are available:
- create XENGUEST_IMAGE [GUESTNAME]: create a guest from a xenguest image file
as guest GUESTNAME. If GUESTNAME is not given the image file name is used
without the xenguest extension.
- remove GUESTNAME: remove the guest GUESTNAME.
- start GUESTNAME: start the guest GUESTNAME.
- stop GUESTNAME: stop the guest GUESTNAME (this is using `xl stop` which is
sending a stop signal to the running guest).
- kill GUESTNAME: force stopping the guest GUESTNAME.
- list: list the available guests.
- status [GUESTNAME]: print the current status of GUESTNAME. If GUESTNAME is
not given, print the status of all guests.
For a detailed help on available options please use:
`xenguest-manager --help`
Bitbake parameters
------------------
Several parameters are available to configure the xenguest manager during Yocto
project compilation (those can be set in your project local.conf, for example).
The following parameters are available:
- XENGUEST_MANAGER_VOLUME_DEVICE: This is the device path used by the
xenguest-manager on the device to create LVM disks when guests have a disk
configuration.
This is set by default to "/dev/sda2".
- XENGUEST_MANAGER_VOLUME_NAME: This is the LVM volume name that the
xenguest-manager will create and use to create guest LVM disks.
This is set by default to "vg-xen".
- XENGUEST_MANAGER_GUEST_DIR: This is the directory on Dom0 where the
xenguest-manager will look for xenguest images to create during init. That's
the place where xenguest images can be added to have them automatically
created during next Dom0 boot. The xenguests found there will only be created
if they were not already before (the basename of the files is used as guest
name).
This is set by default to "/usr/share/guests".
Init scripts
------------
Shell scripts can be executed on the host when a guest is started. Depending on
when the script should be executed it should be installed in a different
directory on the target:
- /etc/xenguest/init.pre : Executed first, prior to guest creation
- /etc/xenguest/init.d : Executed after guest creation, but before it is started
- /etc/xenguest/init.post : Executed after starting the guest
Inside the directory, scripts will be executed in alphabetical order.
Since these scripts are sourced by xenguest-manager they can acccess functions
and variables from the parent file's scope, including:
- ${guestname} : The name of the guest being created
- ${guestdir} : The path to the guest directory
- ${guestcfgfile} : The name of the config file for the starting guest
- ${LOGFILE} : The file to append any logging to, e.g.
echo "Hello, World" >> ${LOGFILE}
Sourcing also allows the script to access params.cfg.
An example of how to create the directory and install an init shell script can
be found in:
recipes-extended/xenguest/xenguest-network.bb
Where network-bridge.sh is installed from network-bridge.sh.in
@@ -1,121 +0,0 @@
Xenguest mkimage
================
Introduction
------------
xenguest-mkimage is a tool to create and modify images to be used as Guest with
Xen. It defines a format to store completely defined guests as a file or as
a directory and provides options to create and modify those images.
A xenguest image contains all elements required to create a xen guest.
This is the base elements like a Xen configuration and a Linux kernel binary
but also some more advanced elements like init scripts or a disk definition.
The format is made to be deployable easily by storing everything in a single
file and provide tools to easily manipulate the images. It can also easily be
extended to have features like encryption or signature of images, updates or
complex configurations by providing features to have init script that will be
executed on the host embedded inside the image.
Xenguest images content
-----------------------
### params.cfg
This file contains parameters that can be used by tools to configure some
functionalities on the host. This can be used by init scripts to have
configurable parameters as it is sourced before calling init scripts.
### guest.cfg and guest.d
guest.cfg is the main xen configuration and guest.d contains optional
configuration parts. All those will be merged into one final xen configuration
before starting the guest.
### files
This directory contains files that can be used by the xen configuration, for
example the binary of the kernel referenced in xen configuration).
This is where the kernel binary, the dtb or a ramdisk will be stored.
### init.pre, init.d and init.post
These directories contain init scripts that will be executed on the host
during the guest startup. Those must be shell scripts and each directory
contains scripts called at a different time:
- init.pre: scripts executed before the guest is created. This can be used
to prepare some features required to create the guest in xen or to
generate part of the xen configuration dynamically.
- init.d: scripts executed when the guest has been created but before it is
started. This can be used to do some xenstore operations or configure the
guest behaviour using xl, for example.
- init.post: scripts executed just after starting the guest. This can be
used to configure things created by xen for the guest like network
network interfaces.
When a directory contains several scripts, those will be called in alphabetical
order.
### disk.cfg and disk-files
disk.cfg contains the guest disk description (disk size and disk partitions).
The file contains the following entries:
- `DISK_SIZE=X`: size of the disk to create in GB
- `DISK_PARTX=SIZE:FS:CONTENT`: create a partition number X (1 to 4) with a
size of SIZE GB, format it with filesystem FS (can be ext2, ext3, ext4, vfat
or swap) and extract CONTENT as initial partition content
(.tar[.gz|.xz|.bz2] file or img[.gz|.bz2] file to be dumped in the partition). FS and
CONTENT can be empty.
The disk-files contain files to be used for initializing the disk partitions
content. Those should be used to create a LVM or a physical disk and initialize
it (create partitions, format them and put the initial content).
Usage
-----
xenguest-mkimage is a shell script which must be called like this:
`xenguest-mkimage OPERATION XENGUEST [OPTIONS]`
### Operations
- create: create a xenguest image. If XENGUEST is an existing empty directory,
the image is created as a directory otherwise it will be created as a file.
- check: verify that XENGUEST is a valid xenguest image.
- update: modify a xenguest image (see --help for a list of operations).
- pack: pack a xenguest image directory into a xenguest image file. The file to
be created must be given as 3rd argument.
- extract: extract a xenguest image file into a directory. The destination
directory must be given as 3rd argument.
- dump-xenconfig: dump xenguest image xen configuration.
- dump-diskconfig: dump xenguest image disk configuration.
- dump-paramsconfig: dump xenguest image parameters configuration.
For a detailed help on available operations, please use:
`xenguest-mkimage --help`
### Options
- --kernel=FILE: add kernel FILE as guest kernel. This is both adding the file
to the image and modifying the xen configuration to use it.
- --xen-memory=SIZE: set the guest memory size in MB.
- --xen-extra: add a kernel command line argument. This can be called several
times to add several command line options.
- --xen-device-tree=FILE: add dtb FILE as device tree. This both adding the
file to the image and modifying the xen configuration to use it.
- --xen-ramdisk=FILE: add ramdisk FILE as guest ramdisk. This both adding the
file to the image and modifying the xen configuration to use it.
- --init-script=FILE: add guest init script. The script is embedded inside the
image file. Several script can be added and the basename of FILE is used to
distinguish them (calling the option twice with the same file will update the
script in the image with the second one).
--disk-size=SIZE: set the guest disk size to SIZE in GB. Calling this with 0
disable the guest disk.
- --disk-add-part=NUM:SIZE:FS:CONTENT: This is adding a partition to the
xenguest image disk. The partition is described with the arguments:
- NUM: partition number.
- SIZE: partition size in GB.
- FS: filesystem to format the partition with. This can be ext2, ext3, ext4,
vfat of swap. If empty the partition is not formated.
- CONTENT: tar of img file to use to initialize the partition. The file must
be added to the image using --disk-add-file=FILE:CONTENT.
For a detailed help on available options, please use:
`xenguest-mkimage OPERATION --help`
@@ -1,80 +0,0 @@
Xenguest Network
================
Introduction
------------
The xenguest-network package is primarly creating a network bridge to share
the host eth physical interfaces with the guests virtual interfaces (vif).
This way the guests can have access to the external network.
At the moment 3 types of network arrangements are provided:
- Bridge: where the guest vif is added to the created bridge interface;
- NAT: where a private subnet is created for the guest,
a kea dhcp4 server is started on the host to serve the guest
and the proper iptables rules are created
to allow the guest to access the external network;
- None: the guest vif is not connected to the bridge.
Usage
-----
On the host project the package xenguest-network must be included in your
image, and on the guest project the XENGUEST_NETWORK_TYPE needs to be set to
"bridge", "nat" or "none".
Bitbake parameters
------------------
Several parameters are available to configure the xenguest network bridge
during Yocto project compilation (those can be set in your project local.conf
or xenguest-network.bbappend, for example).
The following parameters are available:
- XENGUEST_NETWORK_BRIDGE_NAME: This variable defines the name of the network
bridge that is created on the host during init.
This is set by default to "xenbr0".
- XENGUEST_NETWORK_BRIDGE_MEMBERS: This variable defines the list of the
physical network interfaces that are added to the bridge when it is created
on the host during init.
By default no physical interfaces are added.
- XENGUEST_NETWORK_BRIDGE_CONFIG: This variable defines the configuration file
to use to configure the bridge network. By default it points to have file
configuring the network using dhcp.
You can provide a different file using a bbappend and make this variable
point to it if you want to customize your network configuration.
- XENGUEST_IMAGE_NETWORK_TYPE: This variable can be set to "bridge" (default),
"nat" or "none".
The **bridge** type will add the domU vif interface to a bridge which also
contains the dom0 physical interface giving the guest direct access to the
external network.
The **nat** type will setup a private network between dom0 and domU, setup
the appropriate routing table, configure and run the kea dhcp4 server
on dom0 to serve the domU and apply the iptables rules to allow the guest
to acess the external network. The kea dhcp4 server configuration for
the guest can be customised by replacing the
"meta-arm-autonomy/recipes-extended/xenguest/files/kea-subnet4.json" file
in a xenguest-network.bbappend. The kea-subnet4.json file is installed in
the xenguest image and copied to
"/etc/xenguest/guests/${guestname}/files/kea-subnet4.json" when the guest
image is created. It will be consumed by the
"/etc/xen/scripts/vif-post.d/00-vif-xenguest.hook" script which is called by
"/etc/xen/scripts/vif-nat" script when starting/stopping the xenguest.
After guest start, "/etc/xenguest/init.post/xenguest-network-init-post.sh"
script is called to reload kea dhcp4 server with updated configuration,
after virtual network interface is ready.
In the guest project, the NAT port forward can be customised by changing
the XENGUEST_IMAGE_HOST_PORT (default: "1000 + ${domid}") and
XENGUEST_IMAGE_GUEST_PORT (default: "22") variables in local.conf or
xenguest-base-image.bbappend. This configuration is implemented and installed
in "/etc/xenguest/guests/${guestname}/files/00-xenguest-nat-port-forward.hook"
script which is called by "/etc/xen/scripts/vif-post.d/00-vif-xenguest.hook".
The **none** type will not affect any networking setting between on dom0 and
domU.
@@ -1,7 +0,0 @@
# Require extra machine specific settings
ARM_BSP_DYN_MACHINE_EXTRA_REQUIRE ?= ""
ARM_BSP_DYN_MACHINE_EXTRA_REQUIRE_n1sdp = "n1sdp-extra-settings.inc"
ARM_BSP_DYN_MACHINE_EXTRA_REQUIRE_fvp-base = "fvp-base-extra-settings.inc"
ARM_BSP_DYN_MACHINE_EXTRA_REQUIRE_juno = "juno-extra-settings.inc"
require ${ARM_BSP_DYN_MACHINE_EXTRA_REQUIRE}
@@ -1,7 +0,0 @@
# Extra machine settings for fvp-base
# FVP uses vda as hard drive and partition 2 is the
# default rootfs, so use vda3 for guest lvm
XENGUEST_MANAGER_VOLUME_DEVICE ?= "/dev/vda3"
XENGUEST_NETWORK_BRIDGE_MEMBERS ?= "eth0"
@@ -1,4 +0,0 @@
# Extra machine settings for juno
# Juno board has 2 network interfaces, add both of them to the bridge
XENGUEST_NETWORK_BRIDGE_MEMBERS ?= "eth0 eth1"
@@ -1,40 +0,0 @@
# Extra machine settings for n1sdp
# We need to extent the wks search path to be able to find the wks file set in
# ARM_AUTONOMY_WKS_FILE.
WKS_SEARCH_PATH_prepend := "${ARM_AUTONOMY_ARM_BSP_DYNAMIC_DIR}/wic:"
ARM_AUTONOMY_WKS_FILE ?= "arm-autonomy-n1sdp-efidisk.wks.in"
WKS_FILE = "${ARM_AUTONOMY_WKS_FILE}"
# Set the wks guest partition size and unit. It must be aligned with the sum of
# all XENGUEST_IMAGE_DISK_SIZE set for the guests. By default, LVM2 metadata is
# 1 MiB per physical volume, hence it needs to be taken into account when
# setting GUEST_PART_SIZE. The XENGUEST_IMAGE_DISK_SIZE default value is 4GiB.
GUEST_PART_SIZE ?= "4097"
GUEST_PART_SIZE_UNIT ?= "M"
# The GRUB_CFG_FILE affects arm-autonomy-n1sdp-efidisk.wks.in file
GRUB_CFG_FILE ?= "${ARM_AUTONOMY_ARM_BSP_DYNAMIC_DIR}/wic/arm-autonomy-n1sdp-grub.cfg"
# From arm-autonomy-n1sdp-efidisk.wks.in, the /boot partition is /dev/sda1, and
# the "/" partition is /dev/sda2.
XENGUEST_MANAGER_VOLUME_DEVICE ?= "/dev/sda3"
XENGUEST_NETWORK_BRIDGE_MEMBERS ?= "eth0"
# The XEN_DEVICETREE_DEPEND and XEN_DEVICETREE_DTBS variables aftect the
# xen-devicetree.bb recipe
XEN_DEVICETREE_DEPEND = "virtual/trusted-firmware-a:do_deploy"
XEN_DEVICETREE_DTBS ?= "n1sdp-single-chip.dtb"
# XEN_MOD_DEVICETREE_DTBS are the generated devicetrees for Xen. By default the
# xen-devicetree.bb recipe adds '-xen' suffix to it
XEN_MOD_DEVICETREE_DTBS ?= "n1sdp-single-chip-xen.dtb"
# When generating the wic image we need to have the xen deployed
do_image_wic[depends] += "xen:do_deploy"
# Select the extra files to be included in the boot partition
IMAGE_EFI_BOOT_FILES += "xen-n1sdp.efi;xen.efi"
IMAGE_EFI_BOOT_FILES += "${XEN_MOD_DEVICETREE_DTBS}"
@@ -1,21 +0,0 @@
# Use OVERRIDES to minimize the usage of
# ${@bb.utils.contains('DISTRO_FEATURES', 'xen', ...
OVERRIDES_append = "${@bb.utils.contains('DISTRO_FEATURES', 'xen', ':xen', '', d)}"
FILESEXTRAPATHS_prepend_xen := "${THISDIR}/${PN}:"
DEPLOY_EXTRA_DEPS ??= ""
DEPLOY_EXTRA_DEPS_xen = "xen:do_deploy xen-devicetree:do_deploy"
do_deploy[depends] += "${DEPLOY_EXTRA_DEPS}"
do_deploy_prepend_xen() {
# To avoid dependency loop between firmware-image-juno:do_install,
# xen:do_deploy and xen-devicetree:do_deploy when
# INITRAMFS_IMAGE_BUNDLE = "1", we need to handle the xen and
# xen-devicetree binaries copying in the do_deploy task.
cp ${DEPLOY_DIR_IMAGE}/xen-${COMPATIBLE_MACHINE}.efi \
${D}/${UNPACK_DIR}/SOFTWARE/xen
cp ${DEPLOY_DIR_IMAGE}/*xen.dtb \
${D}/${UNPACK_DIR}/SOFTWARE/
}
@@ -1,78 +0,0 @@
TITLE: Versatile Express Images Configuration File
[IMAGES]
TOTALIMAGES: 11 ;Number of Images (Max: 32)
NOR0UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR0ADDRESS: 0x00000000 ;Image Flash Address
NOR0FILE: \SOFTWARE\fip.bin ;Image File Name
NOR0LOAD: 00000000 ;Image Load Address
NOR0ENTRY: 00000000 ;Image Entry Point
NOR1UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR1ADDRESS: 0x03EC0000 ;Image Flash Address
NOR1FILE: \SOFTWARE\bl1.bin ;Image File Name
NOR1LOAD: 00000000 ;Image Load Address
NOR1ENTRY: 00000000 ;Image Entry Point
NOR2UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR2ADDRESS: 0x00500000 ;Image Flash Address
NOR2FILE: \SOFTWARE\Image ;Image File Name
NOR2NAME: norkern ;Rename kernel to norkern
NOR2LOAD: 00000000 ;Image Load Address
NOR2ENTRY: 00000000 ;Image Entry Point
NOR3UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR3ADDRESS: 0x02700000 ;Image Flash Address
NOR3FILE: \SOFTWARE\juno-xen.dtb ;Image File Name
NOR3NAME: board.dtb ;Specify target filename to preserve file extension
NOR3LOAD: 00000000 ;Image Load Address
NOR3ENTRY: 00000000 ;Image Entry Point
NOR4UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR4ADDRESS: 0x01D00000 ;Image Flash Address
NOR4FILE: \SOFTWARE\xen ;Image File Name
NOR4NAME: xen
NOR4LOAD: 00000000 ;Image Load Address
NOR4ENTRY: 00000000 ;Image Entry Point
NOR5UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR5ADDRESS: 0x025C0000 ;Image Flash Address
NOR5FILE: \SOFTWARE\hdlcdclk.dat ;Image File Name
NOR5LOAD: 00000000 ;Image Load Address
NOR5ENTRY: 00000000 ;Image Entry Point
NOR6UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR6ADDRESS: 0x03E40000 ;Image Flash Address
NOR6FILE: \SOFTWARE\scp_bl1.bin ;Image File Name
NOR6LOAD: 00000000 ;Image Load Address
NOR6ENTRY: 00000000 ;Image Entry Point
NOR7UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR7ADDRESS: 0x0BF00000 ;Image Flash Address
NOR7FILE: \SOFTWARE\startup.nsh ;Image File Name
NOR7NAME: startup.nsh
NOR7LOAD: 00000000 ;Image Load Address
NOR7ENTRY: 00000000 ;Image Entry Point
NOR8UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR8ADDRESS: 0x0BFC0000 ;Image Flash Address
NOR8FILE: \SOFTWARE\blank.img ;Image File Name
NOR8NAME: BOOTENV
NOR8LOAD: 00000000 ;Image Load Address
NOR8ENTRY: 00000000 ;Image Entry Point
NOR9UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR9ADDRESS: 0x02600000 ;Image Flash Address
NOR9FILE: \SOFTWARE\selftest ;Image File Name
NOR9LOAD: 00000000 ;Image Load Address
NOR9ENTRY: 00000000 ;Image Entry Point
NOR10UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR10ADDRESS: 0x02780000 ;Image Flash Address
NOR10NAME: uEnv.txt
NOR10FILE: \SOFTWARE\uEnv.txt ;Image File Name
NOR10LOAD: 00000000 ;Image Load Address
NOR10ENTRY: 00000000 ;Image Entry Point
@@ -1,78 +0,0 @@
TITLE: Versatile Express Images Configuration File
[IMAGES]
TOTALIMAGES: 11 ;Number of Images (Max: 32)
NOR0UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR0ADDRESS: 0x00000000 ;Image Flash Address
NOR0FILE: \SOFTWARE\fip.bin ;Image File Name
NOR0LOAD: 00000000 ;Image Load Address
NOR0ENTRY: 00000000 ;Image Entry Point
NOR1UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR1ADDRESS: 0x03EC0000 ;Image Flash Address
NOR1FILE: \SOFTWARE\bl1.bin ;Image File Name
NOR1LOAD: 00000000 ;Image Load Address
NOR1ENTRY: 00000000 ;Image Entry Point
NOR2UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR2ADDRESS: 0x00500000 ;Image Flash Address
NOR2FILE: \SOFTWARE\Image ;Image File Name
NOR2NAME: norkern ;Rename kernel to norkern
NOR2LOAD: 00000000 ;Image Load Address
NOR2ENTRY: 00000000 ;Image Entry Point
NOR3UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR3ADDRESS: 0x02700000 ;Image Flash Address
NOR3FILE: \SOFTWARE\juno-r1-xen.dtb ;Image File Name
NOR3NAME: board.dtb ;Specify target filename to preserve file extension
NOR3LOAD: 00000000 ;Image Load Address
NOR3ENTRY: 00000000 ;Image Entry Point
NOR4UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR4ADDRESS: 0x01D00000 ;Image Flash Address
NOR4FILE: \SOFTWARE\xen ;Image File Name
NOR4NAME: xen
NOR4LOAD: 00000000 ;Image Load Address
NOR4ENTRY: 00000000 ;Image Entry Point
NOR5UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR5ADDRESS: 0x025C0000 ;Image Flash Address
NOR5FILE: \SOFTWARE\hdlcdclk.dat ;Image File Name
NOR5LOAD: 00000000 ;Image Load Address
NOR5ENTRY: 00000000 ;Image Entry Point
NOR6UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR6ADDRESS: 0x03E40000 ;Image Flash Address
NOR6FILE: \SOFTWARE\scp_bl1.bin ;Image File Name
NOR6LOAD: 00000000 ;Image Load Address
NOR6ENTRY: 00000000 ;Image Entry Point
NOR7UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR7ADDRESS: 0x0BF00000 ;Image Flash Address
NOR7FILE: \SOFTWARE\startup.nsh ;Image File Name
NOR7NAME: startup.nsh
NOR7LOAD: 00000000 ;Image Load Address
NOR7ENTRY: 00000000 ;Image Entry Point
NOR8UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR8ADDRESS: 0x0BFC0000 ;Image Flash Address
NOR8FILE: \SOFTWARE\blank.img ;Image File Name
NOR8NAME: BOOTENV
NOR8LOAD: 00000000 ;Image Load Address
NOR8ENTRY: 00000000 ;Image Entry Point
NOR9UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR9ADDRESS: 0x02600000 ;Image Flash Address
NOR9FILE: \SOFTWARE\selftest ;Image File Name
NOR9LOAD: 00000000 ;Image Load Address
NOR9ENTRY: 00000000 ;Image Entry Point
NOR10UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR10ADDRESS: 0x02780000 ;Image Flash Address
NOR10NAME: uEnv.txt
NOR10FILE: \SOFTWARE\uEnv.txt ;Image File Name
NOR10LOAD: 00000000 ;Image Load Address
NOR10ENTRY: 00000000 ;Image Entry Point
@@ -1,78 +0,0 @@
TITLE: Versatile Express Images Configuration File
[IMAGES]
TOTALIMAGES: 11 ;Number of Images (Max: 32)
NOR0UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR0ADDRESS: 0x00000000 ;Image Flash Address
NOR0FILE: \SOFTWARE\fip.bin ;Image File Name
NOR0LOAD: 00000000 ;Image Load Address
NOR0ENTRY: 00000000 ;Image Entry Point
NOR1UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR1ADDRESS: 0x03EC0000 ;Image Flash Address
NOR1FILE: \SOFTWARE\bl1.bin ;Image File Name
NOR1LOAD: 00000000 ;Image Load Address
NOR1ENTRY: 00000000 ;Image Entry Point
NOR2UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR2ADDRESS: 0x00500000 ;Image Flash Address
NOR2FILE: \SOFTWARE\Image ;Image File Name
NOR2NAME: norkern ;Rename kernel to norkern
NOR2LOAD: 00000000 ;Image Load Address
NOR2ENTRY: 00000000 ;Image Entry Point
NOR3UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR3ADDRESS: 0x02700000 ;Image Flash Address
NOR3FILE: \SOFTWARE\juno-r2-xen.dtb ;Image File Name
NOR3NAME: board.dtb ;Specify target filename to preserve file extension
NOR3LOAD: 00000000 ;Image Load Address
NOR3ENTRY: 00000000 ;Image Entry Point
NOR4UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR4ADDRESS: 0x01D00000 ;Image Flash Address
NOR4FILE: \SOFTWARE\xen ;Image File Name
NOR4NAME: xen
NOR4LOAD: 00000000 ;Image Load Address
NOR4ENTRY: 00000000 ;Image Entry Point
NOR5UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR5ADDRESS: 0x025C0000 ;Image Flash Address
NOR5FILE: \SOFTWARE\hdlcdclk.dat ;Image File Name
NOR5LOAD: 00000000 ;Image Load Address
NOR5ENTRY: 00000000 ;Image Entry Point
NOR6UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR6ADDRESS: 0x03E40000 ;Image Flash Address
NOR6FILE: \SOFTWARE\scp_bl1.bin ;Image File Name
NOR6LOAD: 00000000 ;Image Load Address
NOR6ENTRY: 00000000 ;Image Entry Point
NOR7UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR7ADDRESS: 0x0BF00000 ;Image Flash Address
NOR7FILE: \SOFTWARE\startup.nsh ;Image File Name
NOR7NAME: startup.nsh
NOR7LOAD: 00000000 ;Image Load Address
NOR7ENTRY: 00000000 ;Image Entry Point
NOR8UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR8ADDRESS: 0x0BFC0000 ;Image Flash Address
NOR8FILE: \SOFTWARE\blank.img ;Image File Name
NOR8NAME: BOOTENV
NOR8LOAD: 00000000 ;Image Load Address
NOR8ENTRY: 00000000 ;Image Entry Point
NOR9UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR9ADDRESS: 0x02600000 ;Image Flash Address
NOR9FILE: \SOFTWARE\selftest ;Image File Name
NOR9LOAD: 00000000 ;Image Load Address
NOR9ENTRY: 00000000 ;Image Entry Point
NOR10UPDATE: AUTO ;Image Update:NONE/AUTO/FORCE
NOR10ADDRESS: 0x02780000 ;Image Flash Address
NOR10NAME: uEnv.txt
NOR10FILE: \SOFTWARE\uEnv.txt ;Image File Name
NOR10LOAD: 00000000 ;Image Load Address
NOR10ENTRY: 00000000 ;Image Entry Point
@@ -1,15 +0,0 @@
xen_name=xen
xen_addr=0x84000000
uenvcmd=run mybootcmd
mybootcmd=echo Loading custom boot command; \
echo Loading kernel; \
afs load ${kernel_name} ${kernel_addr} ; \
if test $? -eq 1; then echo Loading ${kernel_alt_name} instead of ${kernel_name}; afs load ${kernel_alt_name} ${kernel_addr}; fi; \
echo Loading device tree; \
afs load ${fdtfile} ${fdt_addr}; \
if test $? -eq 1; then echo Loading ${fdt_alt_name} instead of ${fdtfile}; \
afs load ${fdt_alt_name} ${fdt_addr}; fi; fdt addr ${fdt_addr}; fdt resize; \
echo Loading Xen; \
afs load ${xen_name} ${xen_addr}; \
if test $? -eq 0; then echo Booting Xen; bootefi ${xen_addr} ${fdt_addr}; fi;
@@ -1,11 +0,0 @@
/*
* Disable IOMMU on juno board when Xen is used
*/
/ {
/* turn off iommu */
iommu@2b600000 {
status = "disabled";
};
};
@@ -1,41 +0,0 @@
/*
* Modify N1SDP DTB to work properly when Xen is used
*/
/ {
/*
* pmu is using PPI interrupts which are reserved by xen
* remove pm nodes
*/
/delete-node/ pmu;
/delete-node/ spe-pmu;
soc {
/*
* disable IOMMU until we have a proper support in xen
*/
/delete-node/ iommu@4f000000;
/delete-node/ iommu@4f400000;
/*
* Set extra registers required for PCI quirks to communicate with SCP
* and remove invalid properties due to removal
*/
pcie@68000000 {
reg = <0 0x68000000 0 0x1200000>,
<0 0x06000000 0 0x80000>,
<0 0x62000000 0 0x80000>;
/delete-property/ msi-map;
/delete-property/ iommu-map;
};
pcie@70000000 {
reg = <0 0x70000000 0 0x1200000>,
<0 0x06000000 0 0x80000>,
<0 0x60000000 0 0x80000>;
/delete-property/ msi-map;
/delete-property/ iommu-map;
};
};
};
@@ -1,23 +0,0 @@
# Platform dependent parameters
FILESEXTRAPATHS_prepend := "${THISDIR}/files:"
# Add a dtb snippet to turn off iommu in dom0 on Juno board
SRC_URI_append_juno = " file://xen-juno.dtsi"
XEN_DEVICETREE_DTSI_MERGE_append_juno = " xen-juno.dtsi"
# Add a dtb snippet to remove pmu and iommu in dom0 on N1SDP
SRC_URI_append_n1sdp = " file://xen-n1sdp.dtsi"
XEN_DEVICETREE_DTSI_MERGE_append_n1sdp = " xen-n1sdp.dtsi"
# For N1SDP, the XEN_DEVICETREE_DEPEND and XEN_DEVICETREE_DTBS variables are
# being set in meta-arm-autonomy/dynamic-layers/meta-arm-bsp/conf/machine/n1sdp-extra-settings.inc
# Board specific configs
XEN_DEVICETREE_DOM0_BOOTARGS_append_juno = " root=/dev/sda1 rootwait"
XEN_DEVICETREE_XEN_BOOTARGS_append_juno = " console=dtuart dtuart=serial0 bootscrub=0 iommu=no"
XEN_DEVICETREE_DOM0_BOOTARGS_append_n1sdp = " root=/dev/sda2 rootwait"
XEN_DEVICETREE_XEN_BOOTARGS_append_n1sdp = " console=dtuart dtuart=serial0 bootscrub=0 iommu=no"
XEN_DEVICETREE_DOM0_BOOTARGS_append_fvp-base = " root=/dev/vda2"
XEN_DEVICETREE_XEN_BOOTARGS_append_fvp-base = " console=dtuart dtuart=serial0 bootscrub=0"
@@ -1,2 +0,0 @@
CONFIG_DEBUG=y
CONFIG_EARLY_PRINTK_FASTMODEL=y
@@ -1,2 +0,0 @@
CONFIG_DEBUG=y
CONFIG_EARLY_PRINTK_JUNO=y
@@ -1,3 +0,0 @@
CONFIG_DEBUG=y
CONFIG_EARLY_UART_CHOICE_PL011=y
CONFIG_EARLY_UART_BASE_ADDRESS=0x2a400000
@@ -1,5 +0,0 @@
CONFIG_EXPERT=y
# Enable ACPI support
CONFIG_ACPI=y
# Enable ARM Interrupt Translation Service (ITS) emulation
CONFIG_HAS_ITS=y
@@ -1,15 +0,0 @@
# FVP Base support
COMPATIBLE_MACHINE_fvp-base = "fvp-base"
FILESEXTRAPATHS_prepend_fvp-base := "${THISDIR}/files:"
SRC_URI_append_fvp-base = " file://early-printk.cfg"
# Juno support
COMPATIBLE_MACHINE_juno = "juno"
FILESEXTRAPATHS_prepend_juno := "${THISDIR}/files:"
SRC_URI_append_juno = " file://early-printk.cfg"
# N1SDP support
COMPATIBLE_MACHINE_n1sdp = "n1sdp"
FILESEXTRAPATHS_prepend_n1sdp := "${THISDIR}/files:"
SRC_URI_append_n1sdp = " file://n1sdp.cfg \
file://early-printk.cfg"
@@ -1,4 +0,0 @@
# We need to turn off SVE support in the Linux kernel otherwise Xen is stopping
# Linux kernel with a coredump while trying to access ZEN bit of CPACR1 core
# register.
# CONFIG_ARM64_SVE is not set
@@ -1,3 +0,0 @@
define KFEATURE_DESCRIPTION "Disable SVE support"
kconf non-hardware disable-arm64-sve.cfg
@@ -1,15 +0,0 @@
FILESEXTRAPATHS_prepend := "${THISDIR}:"
#
# arm-autonomy kmeta extra
#
SRC_URI_append = " file://arm-autonomy-kmeta-extra;type=kmeta;name=arm-autonomy-kmeta-extra;destsuffix=arm-autonomy-kmeta-extra"
# We need to turn off SVE support in the Linux kernel otherwise Xen is stopping
# Linux kernel with a coredump while trying to access XEN bit of CPACR1 core
# register.
LINUX_ARM_AUTONOMY_DISABLE_ARM64_SVE = "${@bb.utils.contains_any('DISTRO_FEATURES', \
'arm-autonomy-host arm-autonomy-guest', \
' features/arm-autonomy/disable-arm64-sve.scc','',d)}"
KERNEL_FEATURES_append_fvp-base = "${LINUX_ARM_AUTONOMY_DISABLE_ARM64_SVE}"
@@ -1,14 +0,0 @@
# short-description: Create an EFI disk image
# long-description: Creates a partitioned EFI disk image that the user
# can directly dd to boot media. Uses a custom grub.cfg file to configure the boot.
# First boot partition normally populated as /dev/sda1
part /boot --source bootimg-efi --sourceparams="loader=grub-efi" --ondisk sda --label msdos --active --align 1024
# Second rootfs partition normally populated as /dev/sda2
part / --source rootfs --ondisk sda --fstype=ext4 --label root --align 1024
# Third partition to accomodate guests images normally populated as /dev/sda3 (used by XENGUEST_MANAGER_VOLUME_DEVICE)
part --label guests --source empty --ondisk sda --size="${GUEST_PART_SIZE}${GUEST_PART_SIZE_UNIT}" --system-id 8e --align 1024
bootloader --ptable msdos --configfile="${GRUB_CFG_FILE}"
@@ -1,28 +0,0 @@
set term="vt100"
set default="3"
set timeout="5"
set kernel_cmdline="earlycon=pl011,0x2A400000 console=ttyAMA0,115200 root=/dev/sda2 rootwait"
menuentry 'N1SDP ACPI Boot' {
linux /Image $kernel_cmdline acpi=force
}
menuentry 'N1SDP Single-Chip Boot (Device Tree)' {
devicetree /n1sdp-single-chip.dtb
linux /Image $kernel_cmdline
}
menuentry 'N1SDP Multi-Chip Boot (Device Tree)' {
devicetree /n1sdp-multi-chip.dtb
linux /Image $kernel_cmdline
}
menuentry 'N1SDP Arm Autonomy (Xen) Single-Chip Boot (Device Tree)' {
# no_argumants under here is a workaround for a bug
# also any kernel or dom0 cmdline arguments has to be passed via dtb
# because any arguments put here will be ignored by xen
xen_hypervisor /xen.efi no_arguments=use_dtb_for_xen_or_kernel_cmdline
xen_module /Image
devicetree /n1sdp-single-chip-xen.dtb
}
@@ -1,32 +0,0 @@
# Use OVERRIDES to minimize the usage of
# ${@bb.utils.contains('DISTRO_FEATURES', 'xen', ...
OVERRIDES_append = "${@bb.utils.contains('DISTRO_FEATURES', 'xen', ':xen', '', d)}"
# Xen image to put in the image
# This should point to a file in the deploy image directory
BOOT_WRAPPER_AARCH64_XEN ??= "xen-${MACHINE}"
# Xen command line for the image
BOOT_WRAPPER_AARCH64_XEN_CMDLINE ??= "noreboot dom0_mem=256M"
BOOT_WRAPPER_AARCH64_XEN_CMDLINE_gem5-arm64 = "noreboot dom0_mem=256M console=dtuart \
dtuart=/uart@1c090000 bootscrub=0"
# Fix command line in the axf file for gem5-arm64 when Xen is present
BOOT_WRAPPER_AARCH64_CMDLINE_xen_gem5-arm64 = "console=hvc0 root=/dev/vda rw"
# Image generated by boot wrapper when Xen is present
BOOT_WRAPPER_AARCH64_IMAGE_xen ?= "xen-system.axf"
EXTRA_OECONF_append_xen = " \
--with-xen=${WORKDIR}/kernel/arch/arm64/boot/Image \
--with-xen-cmdline="" \
"
EXTRA_OEMAKE_append_xen = " \
XEN_IMAGE=${DEPLOY_DIR_IMAGE}/${BOOT_WRAPPER_AARCH64_XEN} \
XEN_CMDLINE="${BOOT_WRAPPER_AARCH64_XEN_CMDLINE}" \
"
# We need xen if it is activated
do_deploy[depends] += "${@bb.utils.contains('DISTRO_FEATURES', 'xen', 'xen:do_deploy', '', d)}"
@@ -1,7 +0,0 @@
# When booting gem5-arm64 with Xen we need to set the cpu as Cortex A53 and
# remove support for pointer authentification
GEM5_RUN_EXTRA_append = " \
${@bb.utils.contains('DISTRO_FEATURES_NATIVE', 'arm-autonomy-host', \
'--param=system.cpu_cluster[0].cpus[0].isa[0].midr=0x410fd030 \
--param=system.cpu_cluster[0].cpus[0].isa[0].id_aa64isar1_el1=0x0', \
'', d)}"
@@ -1,2 +0,0 @@
CONFIG_DEBUG=y
CONFIG_EARLY_PRINTK_VEXPRESS=y
@@ -1,4 +0,0 @@
# gem5-arm64 support
COMPATIBLE_MACHINE_gem5-arm64 = "gem5-arm64"
FILESEXTRAPATHS_prepend_gem5-arm64 := "${THISDIR}/files:"
SRC_URI_append_gem5-arm64 = " file://early-printk.cfg"
@@ -1 +0,0 @@
XENGUEST_NETWORK_BRIDGE_MEMBERS_gem5-arm64 ?= "eth0"
@@ -1,4 +0,0 @@
# We need to turn off SVE support in the Linux kernel otherwise Xen is stopping
# Linux kernel with a coredump while trying to access ZEN bit of CPACR1 core
# register.
# CONFIG_ARM64_SVE is not set
@@ -1,3 +0,0 @@
define KFEATURE_DESCRIPTION "Disable SVE support"
kconf non-hardware disable-arm64-sve.cfg
@@ -1,15 +0,0 @@
FILESEXTRAPATHS_prepend_gem5-arm64 := "${THISDIR}:"
#
# arm-autonomy kmeta extra
#
SRC_URI_append_gem5-arm64 = " file://arm-autonomy-kmeta-extra-gem5;type=kmeta;name=arm-autonomy-kmeta-extra-gem5;destsuffix=arm-autonomy-kmeta-extra-gem5"
# We need to turn off SVE support in the Linux kernel otherwise Xen is stopping
# Linux kernel with a coredump while trying to access XEN bit of CPACR1 core
# register.
LINUX_ARM_AUTONOMY_DISABLE_ARM64_SVE_gem5-arm64 = "${@bb.utils.contains_any('DISTRO_FEATURES', \
'arm-autonomy-host arm-autonomy-guest', \
' features/arm-autonomy/disable-arm64-sve.scc','',d)}"
KERNEL_FEATURES_append_gem5-arm64 = "${LINUX_ARM_AUTONOMY_DISABLE_ARM64_SVE}"
@@ -1 +0,0 @@
GRUB_BUILDIN += "${@bb.utils.contains('DISTRO_FEATURES', 'xen', 'xen_boot', '', d)}"
@@ -1,19 +0,0 @@
# SPDX-License-Identifier: MIT
#
# Copyright (c) 2020 Arm Limited
#
SUMMARY = "Docker runtime minimal requirements"
DESCRIPTION = "The minimal set of packages required for running Docker"
inherit packagegroup
RDEPENDS_${PN} = "\
docker-ce \
docker-ce-contrib \
kernel-module-xt-nat \
kernel-module-xt-masquerade \
kernel-module-xt-addrtype \
kernel-module-xt-conntrack \
kernel-module-xt-ipvs \
"
@@ -1,127 +0,0 @@
# Recipe to create a minimal Arm Autonomy stack host image
DESCRIPTION = "Arm Autonomy stack host minimal image"
inherit core-image
LICENSE = "MIT"
LIC_FILES_CHKSUM = "file://${COMMON_LICENSE_DIR}/MIT;md5=0835ade698e0bcf8506ecda2f7b4f302"
# The ARM_AUTONOMY_HOST_IMAGE_EXTERN_GUESTS variable can be used to include in the
# image one or several xenguest images.
# The list must be space separated and each entry must have the following
# format: URL[;guestname=NAME]
# - URL can be the full path to a file or a Yocto compatible SRC_URI url
# - guestname=NAME can be used to specify the name of the guest. If not
# specified the basename of the file (without .xenguest extension) is used.
# Here are examples of values:
# /home/mydir/myguest.xenguest;guestname=guest1
# http://www.url.com/testguest.xenguest
#
# If you are using the output of an other Yocto project, you should use the
# full path syntax instead of the Yocto SRC_URI to be able to use the
# symlink version of your image (as the real file has a new name on each
# build as it includes the date). You must not use SRC_URI type file:// as
# it will try to include the symlink and not the destination file which will
# be detected by the recipe and output an error 'Guest file is a symlink'.
#
# Guests can also be added using a bbapend to this recipe by adding entries
# to SRC_URI with parameter ;guestname=NAME to specify the destination
# guestname. The parameter guestname must be present as it is used to detect
# guests to be added
ARM_AUTONOMY_HOST_IMAGE_EXTERN_GUESTS ??= ""
# Includes minimal set required to start and manage guest. The xen specific
# modules are not explicitly included as they are built as part of the kernel
# image for performance reasons. It doesn't include all kernel modules to
# reduce the image size. If the kernel-modules packages are needed they can
# be appended to IMAGE_INSTALL in a bbappend.
IMAGE_INSTALL += " \
packagegroup-core-boot \
packagegroup-core-ssh-openssh \
qemu-system-i386 \
xenguest-manager \
xenguest-network \
"
# Build xen binary
EXTRA_IMAGEDEPENDS += "xen"
# Build xen-devicetree to produce a xen ready devicetree
EXTRA_IMAGEDEPENDS += "xen-devicetree"
python __anonymous() {
if bb.utils.contains('DISTRO_FEATURES', 'arm-autonomy-host', False, True, d):
raise bb.parse.SkipRecipe("DISTRO_FEATURES does not contain 'arm-autonomy-host'")
if bb.utils.contains('DISTRO_FEATURES', 'xen', False, True, d):
raise bb.parse.SkipRecipe("DISTRO_FEATURES does not contain 'xen'")
# Check in ARM_AUTONOMY_HOST_IMAGE_EXTERN_GUESTS for extra guests and add them
# to SRC_URI with xenguest parameter if not set
guestlist = d.getVar('ARM_AUTONOMY_HOST_IMAGE_EXTERN_GUESTS')
if guestlist:
for guest in guestlist.split():
# If the user just specified a file instead of file://FILE, add
# the file:// prefix
if guest.startswith('/'):
guestfile = ''
guestname = ''
if ';guestname=' in guest:
# user specified a guestname
guestname = guest.split(';guestname=')[1]
guestfile = guest.split(';guestname=')[0]
else:
# no guestname so use the basename
guestname = os.path.basename(guest)
guestfile = guest
# in case we have a link we need the destination
guestfile = os.path.realpath(guestfile)
# make sure the file exist to give a meaningfull error
if not os.path.exists(guestfile):
raise bb.parse.SkipRecipe("ARM_AUTONOMY_HOST_IMAGE_EXTERN_GUESTS entry does not exist: " + guest)
# In case the file is a symlink make sure we use the destination
d.appendVar('SRC_URI', ' file://' + guestfile + ';guestname=' + guestname)
else:
# we have a Yocto URL
try:
_, _, path, _, _, parm = bb.fetch.decodeurl(guest)
# force guestname param in if not already there
if not 'guestname' in parm:
guest += ';guestname=' + os.path.basename(path)
d.appendVar('SRC_URI', ' ' + guest)
except:
raise bb.parse.SkipRecipe("ARM_AUTONOMY_HOST_IMAGE_EXTERN_GUESTS contains an invalid entry: " + guest)
}
python add_extern_guests () {
# Destination directory on the rootfs
guestdir = d.getVar('IMAGE_ROOTFS') + d.getVar('datadir') + '/guests'
# Parse SRC_URI for files with ;guestname= parameter
src_uri = d.getVar('SRC_URI')
for entry in src_uri.split():
_, _, path, _, _, parm = bb.fetch.decodeurl(entry)
if 'guestname' in parm:
if os.path.islink(path):
realpath = os.path.realpath(path)
if not os.path.exists(realpath):
bb.fatal("ARM_AUTONOMY_HOST_IMAGE_EXTERN_GUESTS link does not resolve: " + path)
bb.note("Guest file is a symlink:\n " + path + "\nResolved to:\n " + realpath)
path = realpath
bb.utils.mkdirhier(guestdir)
dstname = parm['guestname']
# Add file extension if not there
if not dstname.endswith('.xenguest'):
dstname += '.xenguest'
if not bb.utils.copyfile(path, guestdir + '/' + dstname):
bb.fatal("Fail to copy Guest file " + path)
}
IMAGE_PREPROCESS_COMMAND += "add_extern_guests; "
@@ -1,18 +0,0 @@
# Add support for interface configuration via interfaces.d directory. This
# feature can be used by other packages to add network interface
# configuration by adding network network interface configuration file under
# interfaces.d directory.
do_install_append() {
# Add scan of interfaces.d to interfaces file
cp -f ${WORKDIR}/interfaces ${WORKDIR}/interfaces.subdir
echo "source-directory ${sysconfdir}/network/interfaces.d/" \
>> ${WORKDIR}/interfaces.subdir
# Create interfaces.d script in case nobody is adding a script
# so that there is no error about non existing directory
install -d -m 755 ${D}${sysconfdir}/network/interfaces.d
# Install our file instead of the original one
install -m 644 interfaces.subdir ${D}${sysconfdir}/network/interfaces
}
@@ -1,16 +0,0 @@
# if arm-autonomy-guest is activated, we are running as a xen guest so we must
# have a console on hvc0
# This is normally done in meta-virtualization if xen is activated but here
# we don't have xen activated.
#
SYSVINIT_ADDHVC0 = "${@bb.utils.contains('DISTRO_FEATURES', \
'arm-autonomy-guest', 'true', 'false', d)}"
do_install_append() {
if ${SYSVINIT_ADDHVC0}; then
echo "" >> ${D}${sysconfdir}/inittab
echo "X0:12345:respawn:/sbin/getty 115200 hvc0" >> \
${D}${sysconfdir}/inittab
fi
}
@@ -1,12 +0,0 @@
# Use OVERRIDES to minimize the usage of
# ${@bb.utils.contains('DISTRO_FEATURES', 'xen', ...
OVERRIDES_append = "${@bb.utils.contains('DISTRO_FEATURES', 'xen', ':xen', '', d)}"
# For Xen we only need the i386 binaries
QEMU_TARGETS_xen = "i386"
PACKAGECONFIG[noaudio] = "--audio-drv-list='',,"
PACKAGECONFIG_append_xen = " noaudio"
PACKAGECONFIG_remove_xen = "fdt sdl kvm"
require ${@bb.utils.contains('DISTRO_FEATURES', 'xen', 'recipes-devtools/qemu/${BPN}-package-split.inc', '', d)}
@@ -1,17 +0,0 @@
/ {
chosen {
xen,dom0-bootargs = "###XEN_DOM0_BOOTARGS###";
xen,xen-bootargs = "###XEN_XEN_BOOTARGS###";
modules {
#size-cells = <0x00000001>;
#address-cells = <0x00000001>;
module@0 {
reg = <###XEN_DOM0_ADDR### ###XEN_DOM0_SIZE###>;
compatible = "multiboot,module";
};
};
};
};
@@ -1,72 +0,0 @@
# This recipe can be used to modify one or several DTBS to add
# entries required to declare and boot Linux as Dom0 from Xen
SRC_URI = "file://xen.dtsi.in"
LICENSE = "MIT"
LIC_FILES_CHKSUM = "\
file://${COMMON_LICENSE_DIR}/MIT;md5=0835ade698e0bcf8506ecda2f7b4f302 \
"
S = "${WORKDIR}"
DESCRIPTION = "Add entries in DTB for Xen and Dom0"
# Please refer to documentation/xen-devicetree.md for documentation on those
# parameters
XEN_DEVICETREE_DEPEND ?= "virtual/kernel:do_deploy"
XEN_DEVICETREE_DTBS ?= "${KERNEL_DEVICETREE}"
XEN_DEVICETREE_XEN_BOOTARGS ?= "noreboot dom0_mem=${XEN_DEVICETREE_DOM0_MEM}"
XEN_DEVICETREE_DOM0_MEM ?= "1024M"
XEN_DEVICETREE_DOM0_BOOTARGS ?= "console=hvc0 earlycon=xen"
XEN_DEVICETREE_DOM0_ADDR ?= "0x80080000"
XEN_DEVICETREE_DOM0_SIZE ?= "0x01000000"
XEN_DEVICETREE_DTSI_MERGE ?= "xen.dtsi"
# Our package does not generate any package for the rootfs but contributes to
# deploy
inherit nopackages deploy
DEPENDS += "dtc-native"
do_configure[noexec] = "1"
do_compile[noexec] = "1"
do_install[noexec] = "1"
do_deploy() {
if [ ! -f ${WORKDIR}/xen.dtsi.in ]; then
die "xen.dtsi.in does not exist"
fi
cat ${WORKDIR}/xen.dtsi.in \
| sed -e "s?###XEN_DOM0_BOOTARGS###?${XEN_DEVICETREE_DOM0_BOOTARGS}?" \
| sed -e "s?###XEN_XEN_BOOTARGS###?${XEN_DEVICETREE_XEN_BOOTARGS}?" \
| sed -e "s?###XEN_DOM0_ADDR###?${XEN_DEVICETREE_DOM0_ADDR}?" \
| sed -e "s?###XEN_DOM0_SIZE###?${XEN_DEVICETREE_DOM0_SIZE}?" \
> ${WORKDIR}/xen.dtsi
# Generate final dtbs
for dtbf in ${XEN_DEVICETREE_DTBS}; do
rdtb=`basename $dtbf`
if [ ! -f ${DEPLOY_DIR_IMAGE}/$rdtb ]; then
die "Wrong file in XEN_DEVICETREE_DTBS: ${DEPLOY_DIR_IMAGE}/$rdtb does not exist"
fi
dtc -I dtb -O dts -o ${WORKDIR}/dom0-linux.dts ${DEPLOY_DIR_IMAGE}/$rdtb
# Add external includes
for inc in ${XEN_DEVICETREE_DTSI_MERGE}; do
if [ ! -f ${WORKDIR}/${inc} ]; then
die "Wrong file in XEN_DEVICETREE_DTSI_MERGE: ${WORKDIR}/${inc} does not exist"
fi
echo "/include/ \"$inc\"" >> ${WORKDIR}/dom0-linux.dts
done
rdtbnoextn=`basename $dtbf ".dtb"`
dtc -I dts -O dtb \
-o ${WORKDIR}/${rdtbnoextn}-xen.dtb ${WORKDIR}/dom0-linux.dts
install -m 644 ${rdtbnoextn}-xen.dtb ${DEPLOYDIR}/.
done
}
do_deploy[depends] += "${XEN_DEVICETREE_DEPEND}"
addtask deploy after do_install
@@ -1,40 +0,0 @@
From 3b418b33265402aab0cb1bf2b745a25724bae2d8 Mon Sep 17 00:00:00 2001
Message-Id: <3b418b33265402aab0cb1bf2b745a25724bae2d8.1602684880.git.diego.sueiro@arm.com>
From: Bertrand Marquis <bertrand.marquis@arm.com>
Date: Tue, 18 Aug 2020 14:47:38 +0100
Subject: [PATCH] arm: Add Neoverse N1 processor identification
Add MIDR and CPU part numbers for Neoverse N1
Signed-off-by: Bertrand Marquis <bertrand.marquis@arm.com>
Acked-by: Julien Grall <jgrall@amazon.com>
Upstream-Status: Backport
Signed-off-by: Diego Sueiro <diego.sueiro@arm.com>
---
xen/include/asm-arm/processor.h | 2 ++
1 file changed, 2 insertions(+)
diff --git a/xen/include/asm-arm/processor.h b/xen/include/asm-arm/processor.h
index aa642e3..3ca67f8 100644
--- a/xen/include/asm-arm/processor.h
+++ b/xen/include/asm-arm/processor.h
@@ -58,6 +58,7 @@
#define ARM_CPU_PART_CORTEX_A73 0xD09
#define ARM_CPU_PART_CORTEX_A75 0xD0A
#define ARM_CPU_PART_CORTEX_A76 0xD0B
+#define ARM_CPU_PART_NEOVERSE_N1 0xD0C
#define MIDR_CORTEX_A12 MIDR_CPU_MODEL(ARM_CPU_IMP_ARM, ARM_CPU_PART_CORTEX_A12)
#define MIDR_CORTEX_A17 MIDR_CPU_MODEL(ARM_CPU_IMP_ARM, ARM_CPU_PART_CORTEX_A17)
@@ -68,6 +69,7 @@
#define MIDR_CORTEX_A73 MIDR_CPU_MODEL(ARM_CPU_IMP_ARM, ARM_CPU_PART_CORTEX_A73)
#define MIDR_CORTEX_A75 MIDR_CPU_MODEL(ARM_CPU_IMP_ARM, ARM_CPU_PART_CORTEX_A75)
#define MIDR_CORTEX_A76 MIDR_CPU_MODEL(ARM_CPU_IMP_ARM, ARM_CPU_PART_CORTEX_A76)
+#define MIDR_NEOVERSE_N1 MIDR_CPU_MODEL(ARM_CPU_IMP_ARM, ARM_CPU_PART_NEOVERSE_N1)
/* MPIDR Multiprocessor Affinity Register */
#define _MPIDR_UP (30)
--
2.7.4
@@ -1,42 +0,0 @@
From 858c0be8c2fa4125a0fa0acaa03ae730e5c7cb3c Mon Sep 17 00:00:00 2001
Message-Id: <858c0be8c2fa4125a0fa0acaa03ae730e5c7cb3c.1602684183.git.diego.sueiro@arm.com>
From: Bertrand Marquis <bertrand.marquis@arm.com>
Date: Tue, 18 Aug 2020 14:47:39 +0100
Subject: [PATCH] xen/arm: Enable CPU Erratum 1165522 for Neoverse
Enable CPU erratum of Speculative AT on the Neoverse N1 processor
versions r0p0 to r2p0.
Also Fix Cortex A76 Erratum string which had a wrong errata number.
Signed-off-by: Bertrand Marquis <bertrand.marquis@arm.com>
Acked-by: Julien Grall <jgrall@amazon.com>
Upstream-Status: Backport
Signed-off-by: Diego Sueiro <diego.sueiro@arm.com>
---
xen/arch/arm/cpuerrata.c | 8 +++++++-
1 file changed, 7 insertions(+), 1 deletion(-)
diff --git a/xen/arch/arm/cpuerrata.c b/xen/arch/arm/cpuerrata.c
index 0248893..6c09017 100644
--- a/xen/arch/arm/cpuerrata.c
+++ b/xen/arch/arm/cpuerrata.c
@@ -477,8 +477,14 @@ static const struct arm_cpu_capabilities arm_errata[] = {
},
#endif
{
+ /* Neoverse r0p0 - r2p0 */
+ .desc = "ARM erratum 1165522",
+ .capability = ARM64_WORKAROUND_AT_SPECULATE,
+ MIDR_RANGE(MIDR_NEOVERSE_N1, 0, 2 << MIDR_VARIANT_SHIFT),
+ },
+ {
/* Cortex-A76 r0p0 - r2p0 */
- .desc = "ARM erratum 116522",
+ .desc = "ARM erratum 1165522",
.capability = ARM64_WORKAROUND_AT_SPECULATE,
MIDR_RANGE(MIDR_CORTEX_A76, 0, 2 << MIDR_VARIANT_SHIFT),
},
--
2.7.4
@@ -1,35 +0,0 @@
From 1814a626fb5811184eda64fe22f0055df4600211 Mon Sep 17 00:00:00 2001
Message-Id: <1814a626fb5811184eda64fe22f0055df4600211.1602684203.git.diego.sueiro@arm.com>
From: Julien Grall <jgrall@amazon.com>
Date: Tue, 25 Aug 2020 18:38:10 +0100
Subject: [PATCH] xen/arm: Update silicon-errata.txt with the Neovers AT
erratum
Commit 858c0be8c2fa "xen/arm: Enable CPU Erratum 1165522 for Neoverse"
added a new erratum but forgot to update silicon-errata.txt.
Update the file accordingly to keep track of errata workaround in Xen.
Signed-off-by: Julien Grall <jgrall@amazon.com>
Reviewed-by: Bertrand Marquis <bertrand.marquis@arm.com>
Acked-by: Stefano Stabellini <sstabellini@kernel.org>
Upstream-Status: Backport
Signed-off-by: Diego Sueiro <diego.sueiro@arm.com>
---
docs/misc/arm/silicon-errata.txt | 1 +
1 file changed, 1 insertion(+)
diff --git a/docs/misc/arm/silicon-errata.txt b/docs/misc/arm/silicon-errata.txt
index 11e5a9d..e15d092 100644
--- a/docs/misc/arm/silicon-errata.txt
+++ b/docs/misc/arm/silicon-errata.txt
@@ -51,4 +51,5 @@ stable hypervisors.
| ARM | Cortex-A57 | #1319537 | N/A |
| ARM | Cortex-A72 | #1319367 | N/A |
| ARM | Cortex-A76 | #1165522 | N/A |
+| ARM | Neoverse-N1 | #1165522 | N/A
| ARM | MMU-500 | #842869 | N/A |
--
2.7.4
@@ -1,51 +0,0 @@
From 968bb86d04913f52d7678a842474f2a674a8b23e Mon Sep 17 00:00:00 2001
Message-Id: <968bb86d04913f52d7678a842474f2a674a8b23e.1602683678.git.diego.sueiro@arm.com>
From: Wei Chen <wei.chen@arm.com>
Date: Fri, 28 Aug 2020 02:34:03 +0000
Subject: [PATCH] xen/arm: Missing N1/A76/A75 FP registers in vCPU context
switch
Xen has cpu_has_fp/cpu_has_simd to detect whether the CPU supports
FP/SIMD or not. But currently, these two MACROs only consider value 0
of ID_AA64PFR0_EL1.FP/SIMD as FP/SIMD features enabled. But for CPUs
that support FP/SIMD and half-precision floating-point arithmetic, the
ID_AA64PFR0_EL1.FP/SIMD are 1 (see Arm ARM DDI0487F.b, D13.2.64).
For these CPUs, xen will treat them as no FP/SIMD support, the
vfp_save/restore_state will not take effect.
From the TRM documents of Cortex-A75/A76/N1, we know these CPUs support
basic Advanced SIMD/FP and half-precision floating-point arithmetic. In
this case, on N1/A76/A75 platforms, Xen will always miss the floating
pointer registers save/restore. If different vCPUs are running on the
same pCPU, the floating pointer registers will be corrupted randomly.
This patch fixes Xen on these new cores.
Signed-off-by: Wei Chen <wei.chen@arm.com>
Reviewed-by: Bertrand Marquis <bertrand.marquis@arm.com>
Reviewed-by: Julien Grall <jgrall@amazon.com>
Upstream-Status: Backport
Signed-off-by: Diego Sueiro <diego.sueiro@arm.com>
---
xen/include/asm-arm/cpufeature.h | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/xen/include/asm-arm/cpufeature.h b/xen/include/asm-arm/cpufeature.h
index 674beb0..10878ea 100644
--- a/xen/include/asm-arm/cpufeature.h
+++ b/xen/include/asm-arm/cpufeature.h
@@ -13,8 +13,8 @@
#define cpu_has_el2_64 (boot_cpu_feature64(el2) >= 1)
#define cpu_has_el3_32 (boot_cpu_feature64(el3) == 2)
#define cpu_has_el3_64 (boot_cpu_feature64(el3) >= 1)
-#define cpu_has_fp (boot_cpu_feature64(fp) == 0)
-#define cpu_has_simd (boot_cpu_feature64(simd) == 0)
+#define cpu_has_fp (boot_cpu_feature64(fp) < 8)
+#define cpu_has_simd (boot_cpu_feature64(simd) < 8)
#define cpu_has_gicv3 (boot_cpu_feature64(gic) == 1)
#endif
--
2.7.4
@@ -1,51 +0,0 @@
From f4c1a541fa351e4f613471bbf397931f9e1ddd27 Mon Sep 17 00:00:00 2001
Message-Id: <f4c1a541fa351e4f613471bbf397931f9e1ddd27.1602683707.git.diego.sueiro@arm.com>
From: Wei Chen <wei.chen@arm.com>
Date: Fri, 28 Aug 2020 02:34:04 +0000
Subject: [PATCH] xen/arm: Throw messages for unknown FP/SIMD implement ID
Arm ID_AA64PFR0_EL1 register provides two fields to describe CPU
FP/SIMD implementations. Currently, we exactly know the meaning of
0x0, 0x1 and 0xf of these fields. Xen treats value < 8 as FP/SIMD
features presented. If there is a value 0x2 bumped in the future,
Xen behaviors for value <= 0x1 can also take effect. But what Xen
done for value <= 0x1 may not always cover new value 0x2 required.
We throw these messages to break the silence when Xen detected
unknown FP/SIMD IDs to notice user to check.
Signed-off-by: Wei Chen <wei.chen@arm.com>
Reviewed-by: Bertrand Marquis <bertrand.marquis@arm.com>
Acked-by: Julien Grall <jgrall@amazon.com>
Upstream-Status: Backport
Signed-off-by: Diego Sueiro <diego.sueiro@arm.com>
---
xen/arch/arm/setup.c | 12 ++++++++++++
1 file changed, 12 insertions(+)
diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c
index 7968cee..f16b33f 100644
--- a/xen/arch/arm/setup.c
+++ b/xen/arch/arm/setup.c
@@ -133,6 +133,18 @@ static void __init processor_id(void)
cpu_has_simd ? " AdvancedSIMD" : "",
cpu_has_gicv3 ? " GICv3-SysReg" : "");
+ /* Warn user if we find unknown floating-point features */
+ if ( cpu_has_fp && (boot_cpu_feature64(fp) >= 2) )
+ printk(XENLOG_WARNING "WARNING: Unknown Floating-point ID:%d, "
+ "this may result in corruption on the platform\n",
+ boot_cpu_feature64(fp));
+
+ /* Warn user if we find unknown AdvancedSIMD features */
+ if ( cpu_has_simd && (boot_cpu_feature64(simd) >= 2) )
+ printk(XENLOG_WARNING "WARNING: Unknown AdvancedSIMD ID:%d, "
+ "this may result in corruption on the platform\n",
+ boot_cpu_feature64(simd));
+
printk(" Debug Features: %016"PRIx64" %016"PRIx64"\n",
boot_cpu_data.dbg64.bits[0], boot_cpu_data.dbg64.bits[1]);
printk(" Auxiliary Features: %016"PRIx64" %016"PRIx64"\n",
--
2.7.4
@@ -1,18 +0,0 @@
Upstream-Status: Pending
Signed-off-by: Diego Sueiro <diego.sueiro@arm.com>
Setting the hostname is failing because the "$XENBUS_PATH/domain"
doesn't exist anymore. To fix this we set it to dom$domid
Index: git/tools/hotplug/Linux/vif-nat
===================================================================
--- git.orig/tools/hotplug/Linux/vif-nat
+++ git/tools/hotplug/Linux/vif-nat
@@ -86,6 +86,7 @@ router_ip=$(routing_ip "$ip")
vif_ip=`echo ${ip} | awk -F/ '{print $1}'`
hostname=$(xenstore_read "$XENBUS_PATH/domain" | tr -- '_.:/+' '-----')
+[ -z "${hostname}" ] && hostname=dom$domid
if [ "$vifid" != "1" ]
then
hostname="$hostname-$vifid"
@@ -1,9 +0,0 @@
FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:"
SRC_URI += "file://0001-vif-nat-fix-hostname.patch \
"
PACKAGECONFIG_remove = "\
${@bb.utils.contains('DISTRO_FEATURES', \
'arm-autonomy-host', \
'sdl', '', d)}"
@@ -1,9 +0,0 @@
# Make Xen machine specific
# This ensures that sstate is properly handled and that each machine can have
# its own configuration
PACKAGE_ARCH = "${MACHINE_ARCH}"
PACKAGECONFIG_remove = "\
${@bb.utils.contains('DISTRO_FEATURES', \
'arm-autonomy-host', \
'sdl', '', d)}"
@@ -1,9 +0,0 @@
FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}-4.14:"
SRC_URI += " \
file://0001-arm-Add-Neoverse-N1-processor-identification.patch \
file://0002-xen-arm-Enable-CPU-Erratum-1165522-for-Neoverse.patch \
file://0003-xen-arm-Update-silicon-errata.txt-with-the-Neovers-A.patch \
file://0004-xen-arm-Missing-N1-A76-A75-FP-registers-in-vCPU-cont.patch \
file://0005-xen-arm-Throw-messages-for-unknown-FP-SIMD-implement.patch \
"
@@ -1,138 +0,0 @@
#============================================================================
# ${XEN_SCRIPT_DIR}/vif-post.d/00-vif-xenguest.hook
#
# Script for performing local configuration of a vif.
# This script will be sourced by, e.g., vif-bridge after the hotplugging
# system calls vif-bridge. The script is here and not simply executed as
# a udev rule because this allows simple access to several environment
# variables set by the calling vif-* script.
#
# Environment vars:
# command (add|remove|online|offline)
# dev vif interface name (required).
# main_ip IP address of Dom0
# ip list of IP networks for the vif, space-separated
# XENBUS_PATH path to this device's details in the XenStore (required).
#============================================================================
domid=$(xenstore_read "${XENBUS_PATH}/frontend-id")
guestname=$(xenstore_read "/local/domain/${domid}/name")
bridge=$(xenstore_read "${XENBUS_PATH}/bridge")
mac=$(xenstore_read "$XENBUS_PATH/mac")
kea_conf_file="/etc/kea/kea-dhcp4.conf"
if [ ! -f /etc/xenguest/guests/${guestname}/params.cfg ]; then
log debug "No /etc/xenguest/guests/${guestname}/params.cfg. Exiting."
return
fi
# Source the params file to get the choosen XENGUEST_NETWORK_TYPE
. /etc/xenguest/guests/${guestname}/params.cfg
# We need to get the xenguest subnet prefix to set the subnet and
# the fixed ip to assing to the guest.
get_subnet_prefix() {
# ${vif_ip} is set in the vif-nat script
echo ${vif_ip} | awk -F. '{print $1"."$2"."$3}'
}
subnetprefix=$(get_subnet_prefix)
kea_remove_conf_entry()
{
log debug "kea_remove_conf_entry"
claim_lock "vif-nat-kea"
# Remove the the xenguest kea config file inclusion in the kea main config
sed -i "/${kea_guest_config//'/'/'\/'}/d" "${kea_conf_file}"
# Remove interface entry 1st case: [ "dev1", "dev_to_remove" ]
sed -i "s/,\ \"$dev\"//" "${kea_conf_file}"
# Remove interface entry 2nd case: [ "dev_to_remove", "dev1" ]
sed -i "s/\"$dev\",\ //" "${kea_conf_file}"
# Remove interface entry 3rd case: [ "dev_to_remove" ]
sed -i "s/\[\ \"$dev\"\ \]/\[\ \]/" "${kea_conf_file}"
# Remove the generated the xenguest kea subnet conf file
rm ${XENGUEST_KEA_SUBNET_CONFIG}
keactrl reload
release_lock "vif-nat-kea"
}
kea_add_conf_entries()
{
log debug "kea_add_conf_entries"
claim_lock "vif-nat-kea"
# Include vif in the interfaces 1st case - no other entries:
search_for='\"interfaces\": \[ '
sed -i "/${search_for}\"*\"\ */ s/\ \]/,\ \"${dev}\"\ \]/g" "${kea_conf_file}"
sed -i "s/${search_for}]/${search_for}\"${dev}\"\ \]/g" "${kea_conf_file}"
# Include the xenguest kea subnet conf file in the kea main config
search_for='\"subnet4\": \['
sed -i "/$search_for/a $kea_guest_config" "${kea_conf_file}"
# Generate the xenguest kea subnet conf file
echo -e "$(eval "echo -e \"$(cat ${XENGUEST_KEA_SUBNET_TEMPLATE})\"")" \
> "${XENGUEST_KEA_SUBNET_CONFIG}"
release_lock "vif-nat-kea"
}
call_extra_hooks() {
for f in /etc/xenguest/guests/${guestname}/files/*.hook; do
if [ -x "$f" ]; then
log info "Executing $f"
. "$f"
if [ $? -ne 0 ]; then
log err "$f failed."
fi
else
log info "$f is not executable. Skipping."
fi
done
}
case "${XENGUEST_NETWORK_TYPE}" in
nat)
XENGUEST_KEA_SUBNET_TEMPLATE=${XENGUEST_KEA_SUBNET_TEMPLATE:-"/etc/xenguest/guests/${guestname}/files/kea-subnet4.json"}
if [ ! -f ${XENGUEST_KEA_SUBNET_TEMPLATE} ]; then
log debug "No ${XENGUEST_KEA_SUBNET_TEMPLATE} file. Aborting"
return
fi
XENGUEST_KEA_SUBNET_CONFIG="/etc/kea/kea-subnet4.dom$domid.json"
if [ ! -d "$(dirname ${XENGUEST_KEA_SUBNET_CONFIG})" ]; then
log debug "No kea configuration directory. Aborting"
return
fi
kea_guest_config="\ \ \ \ <?include \"${XENGUEST_KEA_SUBNET_CONFIG}\"?>"
case "$command" in
online)
kea_add_conf_entries
# Enable ip forwarding and NAT for the ${bridge} interface
sysctl -w net.ipv4.ip_forward=1
iptables_w -t nat -A POSTROUTING -o ${bridge} -j MASQUERADE -m comment --comment "dom${domid}"
;;
offline)
kea_remove_conf_entry
# Remove the NAT iptables rules created for the dom${domid}
guest_ipt_rule=$(iptables_w -t nat -vL POSTROUTING -n --line-number | grep -w dom${domid} | awk '{print $1}' | tac)
for rule in ${guest_ipt_rule}; do iptables_w -t nat --delete POSTROUTING ${rule}; done
# If there is no more NAT iptables rules we disable ip forwarding
ipt_nat_rules=$(iptables_w -t nat -vL POSTROUTING -n --line-number | grep MASQUERADE | awk '{print $1}')
[ -z "${ipt_nat_rules##*[!0-9]*}" ] && sysctl -w net.ipv4.ip_forward=0
;;
esac
# We might have extra configs to be applied (e.g.: NAT port forward).
call_extra_hooks
;;
esac
@@ -1,48 +0,0 @@
#============================================================================
# /etc/xenguest/guests/${guestname}/files/00-xenguest-nat-port-forward.hook
#
# Script for performing local configuration related to NAT port forwarding of
# a vif.
# This script will be sourced by
# /etc/xen/scripts/vif-post.d/00-vif-xenguest.hook when
# XENGUEST_IMAGE_NETWORK_TYPE="nat".
# The ${bridge} and ${domid} are set in the 00-vif-xenguest.hook context,
# and ${vip_if} in the vif-nat script context.
#
# Environment vars:
# command (add|remove|online|offline)
# dev vif interface name (required).
# main_ip IP address of Dom0
# ip list of IP networks for the vif, space-separated
# XENBUS_PATH path to this device's details in the XenStore (required).
#============================================================================
host_port="###HOST_PORT###"
guest_port="###GUEST_PORT###"
port_num_check() {
if [ ${host_port} -gt 65535 -o ${guest_port} -gt 65535 ]; then
log error "host_port=${host_port} or guest_port=${guest_port} greater than 65535."
return 1
fi
return 0
}
case "${command}" in
online)
port_num_check
if [ $? -eq 0 ]; then
iptables_w -t nat -A PREROUTING -i ${bridge} -p tcp \
--dport ${host_port} -j DNAT \
--to-destination ${vif_ip}:${guest_port} \
-m comment --comment "dom${domid}"
fi
;;
offline)
# Remove the NAT iptables rules created for the dom${domid}
guest_ipt_rule=$(iptables_w -t nat -vL PREROUTING -n --line-number \
| grep -w dom${domid} | awk '{print $1}' | tac)
for rule in ${guest_ipt_rule}; \
do iptables_w -t nat --delete PREROUTING ${rule}; done
;;
esac
@@ -1,30 +0,0 @@
# This file holds the minimal kea dhcp4 server configuration.
# It replaces existing kea dhcp configuration file via
# "/etc/init.d/kea-restore-default-config" on every system boot.
# The "/etc/xen/scripts/vif-post.d/00-vif-xenguest.hook" called in the end of
# the vif-nat script will append each xenguest configuration to this file,
# to generate the final dhcp configuration.
# This file can be replaced with xenguest-network.bbappend.
{
"Dhcp4": {
"interfaces-config": {
"interfaces": [ ]
},
"lease-database": {
"type": "memfile",
"persist": false
},
"subnet4": [
{
"subnet": "192.0.2.0/24"
}
],
"loggers": [
{
"name": "kea-dhcp4",
"output_options": [ { "output": "/var/log/kea-dhcp4.log" } ],
"severity": "INFO"
}
]
}
}
@@ -1,29 +0,0 @@
#!/bin/sh
#
# Xenguest Restore default kea dhcp4 configuration
# This script restores a dhcp4 server configuration, to prevent
# kea startup failures after random board hangs or powercuts.
#
case "$1" in
start)
ret=$(keactrl status | grep -i "dhcpv4 server" | cut -d ' ' -f 3)
if [ "$ret" = "active" ]; then
echo "WARNING: cowardly refusing to replace dhcp4 configuration"
echo "kea dhcp4 server is up and running!"
else
echo "Resetting kea dhcp4 configuration."
cp -f /etc/kea/kea-dhcp4.conf.original /etc/kea/kea-dhcp4.conf
fi
;;
status|stop|reload|force-reload|restart)
true
;;
*)
# do not advertise unreasonable commands that there is no reason
# to use with this device
echo $"Usage: $0 {start|stop|status|restart|force-reload}"
exit 1
esac
exit $?
@@ -1,35 +0,0 @@
# This file holds the guest kea dhcp4 server options running on Dom0.
# The "/etc/xen/scripts/vif-post.d/00-vif-xenguest.hook" called in the end of
# the vif-nat script will use this file to generate the final dhcp
# configuration.
# This file is added in the xenguest image and installed in dom0 under
# /etc/xenguest/guests/${guestname}/files/kea-subnet4.json when the guest
# image is created.
# Any customizations to it should be performed by replacing it via a bbappend.
# The \${vif_ip} and \${router_ip} variables are set in
# the vif-nat script context. The \${subnetprefix} and \${mac} variable are set in the
# 00-vif-xenguest.hook script context.
# The "subnet" configuration node is mandatory in order to have the dhcp
# properly running.
{
\"subnet\": \"${subnetprefix}.0/24\",
\"option-data\": [
{
\"name\": \"routers\",
\"data\": \"${router_ip}\"
},
{
\"name\": \"domain-name-servers\",
\"data\": \"8.8.8.8, 8.8.4.4\"
}
],
\"reservations\": [
{
\"hw-address\": \"${mac}\",
\"ip-address\": \"${vif_ip}\"
}
]
},
@@ -1,21 +0,0 @@
#!/bin/sh
# This script is setting up a virtual network interface connected to the
# xenguest-network-bridge if NETWORK_BRIDGE is set to 1 in the guest params
#
# Since this script is sourced by xenguest-manager, it can access variables
# such as ${guestname} from the parent file's scope, as well as those in
# params.cfg, for example XENGUEST_NETWORK_TYPE
BRIDGE_NAME="###BRIDGE_NAME###"
case "${XENGUEST_NETWORK_TYPE:=}" in
nat)
echo "vif = ['script=vif-nat']" >> ${guestcfgfile}
;;
bridge)
echo "vif = ['script=vif-bridge,bridge=${BRIDGE_NAME}']" >> ${guestcfgfile}
;;
*)
echo "${@}: XENGUEST_NETWORK_TYPE=$XENGUEST_NETWORK_TYPE invalid"
;;
esac
@@ -1,88 +0,0 @@
#!/bin/bash
#
# xenguest Init Script to start and stop xenguests during boot
#
XENGUEST_CONF_BASE="/etc/xenguest"
if [ ! -f ${XENGUEST_CONF_BASE}/xenguest-manager.conf ]; then
echo "No xenguest manager configuration !!"
exit 1
fi
source ${XENGUEST_CONF_BASE}/xenguest-manager.conf
# Make sure we properly initialized the Dom0
/usr/bin/xenguest-manager check-xen || exit $?
# Create guest list
guestlist=$(/usr/bin/xenguest-manager list)
# Unconfigured guests
case "$1" in
start)
echo "Starting xenguest"
# Create unconfigured guests
if [ -d ${XENGUEST_GUEST_DIR} ]; then
for f in $(find ${XENGUEST_GUEST_DIR} -name "*.xenguest" \
-exec basename {} .xenguest \;); do
if [ ! -f ${XENGUEST_CONF_BASE}/guests/${f}/guest.cfg ]; then
# Check if this guest should be auto booted
autoboot=$(/usr/bin/xenguest-mkimage dump-paramsconfig \
${XENGUEST_GUEST_DIR}/${f}.xenguest | \
grep GUEST_AUTOBOOT | sed -e "s,.*=,," | tr -d '"')
if [ "$autoboot" = "0" ]; then
echo "Do not create $f, autoboot disabled"
else
/usr/bin/xenguest-manager create \
${XENGUEST_GUEST_DIR}/${f}.xenguest
# Update guestlist
if [ $? -eq 0 ]; then
guestlist="${guestlist} ${f}"
fi
fi
fi
done
fi
if [ -n "${guestlist}" ]; then
for f in ${guestlist}; do
GUEST_AUTOBOOT="1"
if [ -f ${XENGUEST_CONF_BASE}/guests/${f}/params.cfg ]; then
source ${XENGUEST_CONF_BASE}/guests/${f}/params.cfg
fi
if [ "${GUEST_AUTOBOOT}" = "1" ]; then
/usr/bin/xenguest-manager start ${f}
fi
done
else
echo "No guest configured"
fi
;;
status)
true
;;
stop)
echo "Stopping xenguest"
for f in ${guestlist}; do
/usr/bin/xenguest-manager stop ${f}
done
;;
reload)
echo >&2 'Reload not available; use force-reload'; exit 1
;;
force-reload|restart)
echo "Restarting xenguest"
$0 stop
$0 start
;;
*)
# do not advertise unreasonable commands that there is no reason
# to use with this device
echo $"Usage: $0 {start|stop|status|restart|force-reload}"
exit 1
esac
exit $?
@@ -1,758 +0,0 @@
#!/bin/bash
# This script manages xenguest
#
set -u
this="$0"
XENGUEST_CONF_BASE="/etc/xenguest"
LOGFILE="/var/log/xenguest"
if [ ! -f ${XENGUEST_CONF_BASE}/xenguest-manager.conf ]; then
echo "Cannot find xenguest manager configuration"
exit 1
fi
# Following variables must be set in configuration:
# XENGUEST_VOLUME_DEVICE: device to use for lvm
# XENGUEST_VOLUME_NAME: lvm volume name to create on device
source ${XENGUEST_CONF_BASE}/xenguest-manager.conf
PREF="xenguest:"
function usage() {
cat <<EOF
Usage $this ACTION [OPTIONS]
with ACTION being one of:
help
Display this help
create GUESTFILE [GUESTNAME]
Create a guest using xenguest image GUESTFILE and name it GUESTNAME.
This will extract and configure the guest and will also create the guest
disk if guest has one configured.
GUESTNAME is set to the basename of GUESTFILE if unspecified.
GUESTNAME guest must not exist
remove GUESTNAME
Remove GUESTNAME and destroy its disk (if it has one)
start GUESTNAME
Start guest GUESTNAME
stop|shutdown GUESTNAME
Stop guest GUESTNAME (send stop signal and let it shutdown normally)
kill|destroy GUESTNAME
Kill guest GUESTNAME (stop directly the guest without signaling it)
list
List configured guests
status
List guests and their current status (running or stopped)
EOF
}
function xenguest_volume_init()
{
# Return:
# 0 - success
# 1 - failure
if [ -z "${XENGUEST_VOLUME_DEVICE:-}" -o \
! -b ${XENGUEST_VOLUME_DEVICE:-} ]; then
echo "${PREF} Invalid volume device in configuration: ${XENGUEST_VOLUME_DEVICE:-}"
return 1
fi
if [ -z "${XENGUEST_VOLUME_NAME:-}" ]; then
echo "${PREF} No volume name in configuration, using vg-xen..."
XENGUEST_VOLUME_NAME="vg-xen"
fi
pvs ${XENGUEST_VOLUME_DEVICE} > /dev/null 2>&1
if [ $? -ne 0 ]; then
# Check if there is no filesystem in the block device
echo "lsblk -n -o FSTYPE ${XENGUEST_VOLUME_DEVICE}" >> ${LOGFILE} 2>&1
filesystem=$(lsblk -n -o FSTYPE ${XENGUEST_VOLUME_DEVICE} 2>> ${LOGFILE})
if [[ $? -eq 0 && -z "$filesystem" ]]; then
echo "${PREF} Initialize lvm on ${XENGUEST_VOLUME_DEVICE}"
echo "pvcreate -f ${XENGUEST_VOLUME_DEVICE}" >> ${LOGFILE} 2>&1
pvcreate -f ${XENGUEST_VOLUME_DEVICE} >> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
echo "${PREF} Error: initialing lvm on " \
"${XENGUEST_VOLUME_DEVICE} failed." | tee -a ${LOGFILE}
return 1
fi
else
[ -z "$filesystem" ] || \
echo "${PREF} Error: The ${XENGUEST_VOLUME_DEVICE} is already " \
"formatted as $filesystem." | tee -a ${LOGFILE}
return 1
fi
fi
vgs ${XENGUEST_VOLUME_NAME} > /dev/null 2>&1
if [ $? -ne 0 ]; then
echo "${PREF} Create ${XENGUEST_VOLUME_NAME} volume"
echo "vgcreate ${XENGUEST_VOLUME_NAME} ${XENGUEST_VOLUME_DEVICE}" \
>> ${LOGFILE} 2>&1
vgcreate ${XENGUEST_VOLUME_NAME} ${XENGUEST_VOLUME_DEVICE} \
>> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
echo "${PREF} Error: creating ${XENGUEST_VOLUME_NAME} volume " \
"failed." | tee -a ${LOGFILE}
return 1
fi
fi
return 0
}
# Detach a disk we attached to xen
function xenguest_detach_disk()
{
echo "xl block-detach 0 \$\(xl block-list 0 | " \
"grep \"domain/0\" | awk '{print \$1}'\)" \
>> ${LOGFILE} 2>&1
xl block-detach 0 $(xl block-list 0 | \
grep "domain/0" | awk '{print $1}') \
>> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
echo "${PREF} Error detaching partition ${part}"
return 1
fi
}
function xenguest_disk_init()
{
# Inputs:
# $1 - guestname
# $2 - guestfile
#
# Outputs:
# 0 - success
# 1 - failed at guest disk preparation
# 2 - failed at guest disk creation
guestname="$1"
guestfile="$2"
devname="/dev/${XENGUEST_VOLUME_NAME}/${guestname}"
source ${XENGUEST_CONF_BASE}/guests/${guestname}/disk.cfg
if [ ${DISK_SIZE:-0} -eq 0 ]; then
echo "${PREF} No disk for ${guestname}"
return
fi
echo "${PREF} Create ${guestname} disk."
# Init our volume
xenguest_volume_init ${guestname}
if [ $? -ne 0 ]; then
return 1
fi
echo "${PREF} Create hard drive for ${guestname}." \
"This might take a while..."
# Remove volume if it already exist
echo "lvs ${XENGUEST_VOLUME_NAME}/${guestname}" >> ${LOGFILE} 2>&1
lvs ${XENGUEST_VOLUME_NAME}/${guestname} >> ${LOGFILE} 2>&1
if [ $? -eq 0 ]; then
echo "lvremove -y ${devname}" >> ${LOGFILE} 2>&1
lvremove -y ${devname} >> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
echo "${PREF} Error removing volume ${guestname}"
return 1
fi
fi
# Create volume
echo "lvcreate -y -L ${DISK_SIZE}G -n ${guestname} ${XENGUEST_VOLUME_NAME}" \
>> ${LOGFILE} 2>&1
lvcreate -y -L ${DISK_SIZE}G -n ${guestname} ${XENGUEST_VOLUME_NAME} \
>> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
echo "${PREF} Error creating volume ${guestname}"
return 1
fi
# Add partition table
echo "parted -s ${devname} mklabel msdos" >> ${LOGFILE} 2>&1
parted -s ${devname} mklabel msdos >> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
echo "${PREF} Error creating partition table on ${guestname}"
return 1
fi
# Setup disk name in xen configuration
echo "xenguest-mkimage update ${XENGUEST_CONF_BASE}/guests/${guestname}" \
"--xen-disk=${devname}" >> ${LOGFILE} 2>&1
xenguest-mkimage update ${XENGUEST_CONF_BASE}/guests/${guestname} \
--xen-disk=${devname} >> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
echo "${PREF} Error setting disk in xen configuration"
return 1
fi
# Create partitions
partstart="0"
# For each partition X the disk.cfg file should set a variable DISK_PARTX
# with a : separated list defining the partition:
# DISK_PART3="4:ext4:disk.tgz" means that partition 3 should be 4G formated
# with ext4 and initialized with the content of disk.tgz
for part in $(seq 1 4); do
eval partdesc="\${DISK_PART${part}:-0}"
size=$(echo ${partdesc} | sed -e "s/\(.*\):.*:.*/\1/")
fstype=$(echo ${partdesc} | sed -e "s/.*:\(.*\):.*/\1/")
content=$(echo ${partdesc} | sed -e "s/.*:.*:\(.*\)/\1/")
if [ "${size}" -ne 0 ]; then
# Size is expressed in GB, pass it in MB
size=$(expr ${size} \* 1024)
partend=$(expr ${partstart} + ${size})
# Let first MB of disk free for partition table
if [ ${partstart} -eq 0 ]; then
partstart="1"
fi
# Create partition
echo "parted -s ${devname} unit MB mkpart primary ${partstart}" \
"${partend}" >> ${LOGFILE} 2>&1
parted -s ${devname} unit MB mkpart primary ${partstart} \
${partend} >> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
echo "${PREF} Error adding partition ${part}"
return 1
fi
# Set next partition start to current partition end
partstart="${partend}"
# Sync to see the created partition
echo "sync" >> ${LOGFILE} 2>&1
sync >> ${LOGFILE} 2>&1
# Prepare format command
if [ -n "${fstype}" ]; then
case ${fstype} in
vfat|ext2|ext3|ext4)
formatcmd="mkfs.${fstype} -F"
;;
swap)
formatcmd="mkswap"
;;
*)
echo "${PREF} partition ${part} of ${guestname}" \
"fstype is invalid: ${fstype}"
return 1
;;
esac
else
formatcmd=""
fi
# Attach disk to xen
echo "xl block-attach 0 phy:${devname} xvda w" >> ${LOGFILE} 2>&1
xl block-attach 0 phy:${devname} xvda w >> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
echo "${PREF} Error attaching partition ${part}"
return 1
fi
# Loop for 20s to wait until /dev/xvdaX appears
i=0
while [ ! -b /dev/xvda${part} ]; do
((i++))
if [[ "$i" == '40' ]]; then
break;
fi
sleep 0.5
done
if [ ! -b /dev/xvda${part} ]; then
echo "${PREF} Partition ${part} creation error"
return 2
fi
if [ -n "${formatcmd}" ]; then
echo "${formatcmd} /dev/xvda${part}" >> ${LOGFILE} 2>&1
${formatcmd} /dev/xvda${part}
if [ $? -ne 0 ]; then
echo "${PREF} Cannot create partition ${part} FS"
return 2
fi
fi
case ${content} in
*.img*)
decompress=""
case ${content} in
*.img.gz)
decompress='zcat'
;;
*.img.bz2)
decompress='bzcat'
;;
*.img)
decompress='cat'
;;
*)
# invalid/unknown compression type
echo "${PREF} Invalid file format in disk ${content}"
return 2
;;
esac
# dd into partition
echo "xenguest-mkimage extract-disk-file ${guestfile} " \
"${content} | ${decompress} | dd of=/dev/xvda${part} " >> ${LOGFILE} 2>&1
xenguest-mkimage extract-disk-file ${guestfile} ${content} \
| ${decompress} | dd of=/dev/xvda${part} >> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
echo "${PREF} Cannot populate partition ${part}"
return 2
fi
;;
*.tar*)
tararg=""
case ${content} in
*.tar.gz)
tararg="z"
;;
*.tar.bz2)
tararg="j"
;;
*.tar.xz)
tararg="J"
;;
*.tar)
tararg=""
;;
*)
# invalid/unknown tar type
echo "${PREF} Invalid file format in disk ${content}"
return 2
;;
esac
# must mount the partition and extract
mntdir=$(mktemp -d)
echo "mount /dev/xvda${part} ${mntdir}" >> ${LOGFILE} 2>&1
mount /dev/xvda${part} ${mntdir} >> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
echo "${PREF} Cannot mount partition ${part}"
rm -rf ${mntdir}
return 2
fi
# tar and unmount
echo "xenguest-mkimage extract-disk-file ${guestfile}" \
"${content} | tar -C ${mntdir} -x${tararg}f - " \
>> ${LOGFILE} 2>&1
xenguest-mkimage extract-disk-file ${guestfile} ${content} \
| tar -C ${mntdir} -x${tararg}f - >> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
echo "${PREF} Cannot populate partition ${part}"
umount ${mntdir}
rm -rf ${mntdir}
return 2
fi
echo "umount ${mntdir}" >> ${LOGFILE} 2>&1
umount ${mntdir} >> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
echo "${PREF} Error unmounting ${part}"
rm -rf ${mntdir}
return 2
fi
rm -rf ${mntdir}
;;
*)
#invalid content type
;;
esac
# Detach disk
xenguest_detach_disk
if [ $? -ne 0 ]; then
return 1
fi
fi
done
}
function xenguest_guest_create()
{
guestfile="$1"
guestname="$2"
# extract xenguest tar
# put xen config in etc ?
# if disk config file:
# disk init
# add partititions
echo "${PREF} Create ${guestname} using ${guestfile}"
rm -rf ${XENGUEST_CONF_BASE}/guests/${guestname}
mkdir -p ${XENGUEST_CONF_BASE}/guests/${guestname}
echo "xenguest-mkimage extract-config ${guestfile}" \
"${XENGUEST_CONF_BASE}/guests/${guestname}" >> ${LOGFILE} 2>&1
xenguest-mkimage extract-config ${guestfile} \
${XENGUEST_CONF_BASE}/guests/${guestname} >> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
echo "${PREF} Error extracting guest image"
exit 1
fi
# Set guest name inside config
echo "xenguest-mkimage update ${XENGUEST_CONF_BASE}/guests/${guestname}" \
"--xen-name=${guestname}" >> ${LOGFILE} 2>&1
xenguest-mkimage update ${XENGUEST_CONF_BASE}/guests/${guestname} \
--xen-name=${guestname} >> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
echo "${PREF} Error setting guest name"
xenguest_guest_remove ${guestname}
exit 1
fi
xenguest_disk_init ${guestname} ${guestfile}
disk_init_status=$?
if [ $disk_init_status -ne 0 ]; then
echo "${PREF} Error: ${guestname} disk creation failed."
if [ $disk_init_status -eq 2 ]; then
xenguest_detach_disk
fi
xenguest_guest_remove ${guestname}
exit 1
fi
}
function xenguest_guest_remove()
{
guestname="$1"
devname="/dev/${XENGUEST_VOLUME_NAME}/${guestname}"
# check if guest had a volume
echo "lvs ${XENGUEST_VOLUME_NAME}/${guestname}" >> ${LOGFILE} 2>&1
lvs ${XENGUEST_VOLUME_NAME}/${guestname} >> ${LOGFILE} 2>&1
if [ $? -eq 0 ]; then
# Remove guest volume
echo "${PREF} Removing ${guestname} volume. This might take a while..."
echo "lvremove -y ${devname}" >> ${LOGFILE} 2>&1
lvremove -y ${devname} >> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
echo "${PREF} Error removing volume ${guestname}"
exit 1
fi
fi
# remove guest files
echo "${PREF} Removing ${guestname} configuration files."
rm -rf ${XENGUEST_CONF_BASE}/guests/${guestname}
}
function xenguest_guest_start()
{
guestname="${1}"
guestdir=${XENGUEST_CONF_BASE}/guests/${guestname}
guestcfgfile=$(mktemp -u "${guestname}.XXXXXX" --tmpdir="${guestdir}" --suffix=".cfg")
# Get guest configuration
source ${guestdir}/params.cfg
pushd ${guestdir} > /dev/null 2>&1
# create config by merging all configurations together
cat guest.cfg $(find guest.d -type f 2> /dev/null) > ${guestcfgfile}
# Build init script lists (ignore non existing dirs errors,
# sort alphabetically and run global scripts first)
#
# These scripts are sourced throughout the start operation if they
# are executable
init_pre="$(find ${XENGUEST_CONF_BASE}/init.pre -type f 2> /dev/null | \
sort) $(find ${guestdir}/init.pre -type f 2> /dev/null | sort)"
init_d="$(find ${XENGUEST_CONF_BASE}/init.d -type f 2> /dev/null | \
sort) $(find ${guestdir}/init.d -type f 2> /dev/null | sort)"
init_post="$(find ${XENGUEST_CONF_BASE}/init.post -type f 2> /dev/null | \
sort) $(find ${guestdir}/init.post -type f 2> /dev/null | sort)"
# call pre init scripts
for f in ${init_pre}; do
if [ -x "$f" ]; then
echo "( . $f )" >> ${LOGFILE} 2>&1
( . $f ) >> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
rm -f ${guestcfgfile}
popd > /dev/null 2>&1
echo "Error in init script $f" >> ${LOGFILE} 2>&1
echo "${PREF} Error during pre init script of ${guestname}"
exit 1
fi
else
echo "$f is not executable. Skipping." >> ${LOGFILE}
fi
done
# Create non started guest
echo "xl create -p ${guestcfgfile}" >> ${LOGFILE} 2>&1
xl create -p ${guestcfgfile} >> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
rm -f ${guestcfgfile}
popd > /dev/null 2>&1
echo "${PREF} Error starting ${guestname}"
exit 1
fi
# call init scripts
for f in ${init_d}; do
if [ -x "$f" ]; then
echo "( . $f )" >> ${LOGFILE} 2>&1
( . $f ) >> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
rm -f ${guestcfgfile}
echo "xl destroy ${guestname}" >> ${LOGFILE} 2>&1
xl destroy ${guestname} >> ${LOGFILE} 2>&1
popd > /dev/null 2>&1
echo "Error in init script $f" >> ${LOGFILE} 2>&1
echo "${PREF} Error during init script of ${guestname}"
exit 1
fi
else
echo "$f is not executable. Skipping." >> ${LOGFILE}
fi
done
# Start guest
echo "xl unpause ${guestname}" >> ${LOGFILE} 2>&1
xl unpause ${guestname} >> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
rm -f ${guestcfgfile}
popd > /dev/null 2>&1
echo "${PREF} Error starting ${guestname}"
exit 1
fi
# call post init scripts
for f in ${init_post}; do
if [ -x "$f" ]; then
echo "( . $f )" >> ${LOGFILE} 2>&1
( . $f ) >> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
rm -f ${guestcfgfile}
echo "xl destroy ${guestname}" >> ${LOGFILE} 2>&1
xl destroy ${guestname} >> ${LOGFILE} 2>&1
popd > /dev/null 2>&1
echo "Error in init script $f" >> ${LOGFILE} 2>&1
echo "${PREF} Error during post init script of ${guestname}"
exit 1
fi
else
echo "$f is not executable. Skipping." >> ${LOGFILE}
fi
done
rm -f ${guestcfgfile}
popd > /dev/null 2>&1
}
function xenguest_guest_stop()
{
guestname="${1}"
echo "xl shutdown ${guestname}" >> ${LOGFILE} 2>&1
xl shutdown ${guestname} >> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
echo "${PREF} Error stopping ${guestname}"
exit 1
fi
}
function check_guest_arg()
{
cmd="${1}"
guestname="${2:-}"
if [ -z "${guestname:-}" ]; then
echo "${PREF} Usage ${this} ${cmd} GUESTNAME"
exit 1
fi
}
function check_guest_exist()
{
guestname="${1}"
if [ ! -f ${XENGUEST_CONF_BASE}/guests/${guestname}/guest.cfg -o \
! -f ${XENGUEST_CONF_BASE}/guests/${guestname}/params.cfg ]; then
echo "${PREF} Invalid guest name: ${guestname}"
exit 1
fi
}
function xl_list_contains()
{
guestname="${1}"
# Select first column of xl list, and find guestname exactly using regex
running=$(xl list | awk 'NR > 1 {print $1}' | grep "^${guestname}$" || echo)
if [ "${running}" = "${guestname}" ]; then
return 0
fi
return 1
}
function check_guest_running()
{
guestname="${1}"
if ! xl_list_contains $guestname; then
echo "${PREF} Guest ${guestname} is not running"
exit 1
fi
}
function check_guest_not_running()
{
guestname="${1}"
if xl_list_contains $guestname; then
echo "${PREF} Guest ${guestname} is running"
exit 1
fi
}
cmd="${1:-help}"
arg1="${2:-}"
arg2="${3:-}"
case ${cmd} in
help|--help|-h|-?)
usage
exit 0
;;
esac
# Check if we have a valid Dom0 booted with Xen
ERROR_MSG=$(xl info 2>&1)
if [ $? -ne 0 ]; then
echo "ERROR: Xen environment is not valid!!!" | tee -a ${LOGFILE}
echo "ERROR: Check if Xen has booted and the kernel configuration." \
| tee -a ${LOGFILE}
echo "ERROR: Output from 'xl info' command:" | tee -a ${LOGFILE}
echo "$ERROR_MSG" | tee -a ${LOGFILE}
exit 1
fi
case ${cmd} in
check-xen)
exit 0
;;
create)
guestfile="${arg1}"
guestname="${arg2}"
if [ -z "${guestfile}" -o ! -f "${guestfile}" ]; then
echo "${PREF} Usage ${this} create XENGUEST_FILE [NAME]"
exit 1
fi
if [ -z "${guestname}" ]; then
guestname=$(basename ${guestfile} .xenguest)
fi
if [ -f ${XENGUEST_CONF_BASE}/guests/${guestname}/guest.cfg ]; then
# Guest already exist
echo "${PREF} A guest ${guestname} already exist"
exit 1
fi
xenguest_guest_create ${guestfile} ${guestname}
echo "${PREF} ${guestname} created."
;;
remove)
guestname="${arg1:-}"
check_guest_arg ${cmd} ${guestname}
check_guest_exist ${guestname}
# We need to stop the guest first if it is running
if xl_list_contains $guestname; then
echo "xl destroy ${guestname}" >> ${LOGFILE} 2>&1
xl destroy ${guestname} >> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
echo "${PREF} Error killing ${guestname}"
exit 1
fi
fi
xenguest_guest_remove ${guestname}
echo "${PREF} ${guestname} removed."
;;
start)
guestname="${arg1:-}"
check_guest_arg ${cmd} ${guestname}
check_guest_exist ${guestname}
check_guest_not_running ${guestname}
xenguest_guest_start ${guestname}
;;
stop|shutdown)
guestname="${arg1:-}"
check_guest_arg ${cmd} ${guestname}
check_guest_exist ${guestname}
check_guest_running ${guestname}
xenguest_guest_stop ${guestname}
;;
kill|destroy)
guestname="${arg1:-}"
check_guest_arg ${cmd} ${guestname}
check_guest_exist ${guestname}
check_guest_running ${guestname}
echo "xl destroy ${guestname}" >> ${LOGFILE} 2>&1
xl destroy ${guestname} >> ${LOGFILE} 2>&1
if [ $? -ne 0 ]; then
echo "${PREF} Error killing ${guestname}"
exit 1
fi
;;
list)
if [ -d ${XENGUEST_CONF_BASE}/guests ]; then
for f in $(find ${XENGUEST_CONF_BASE}/guests -mindepth 1 \
-maxdepth 1 -type d -exec basename {} \;); do
if [ -f ${XENGUEST_CONF_BASE}/guests/$f/guest.cfg ]; then
echo "$f"
fi
done
fi
;;
status)
single_status() (
guestname="${1}"
check_guest_exist ${guestname}
if xl_list_contains $guestname; then
echo "${guestname}: Running"
else
echo "${guestname}: Stopped"
fi
)
guestname="${arg1}"
if [ -n "${guestname}" ]; then
single_status ${guestname}
else
guestlist=$($this list)
if [ -n "${guestlist}" ]; then
for f in ${guestlist}; do
single_status $f
done
fi
fi
;;
*)
echo "${PREF} Invalid argument ${cmd}"
exit 1
;;
esac
@@ -1,789 +0,0 @@
#!/bin/bash
# This script must be used to manipulate xenguest images
#
# xenguest image topology:
# params.cfg: guest global configuration file. Only edited using this script.
# guest.cfg: xen main configuration file. Only edited using this script.
# guest.d: directory contains files with custom xen configuration entries
# which are appended to guest.cfg before starting the guest
# files: directory where files used by xen configuration are stored
# disk.cfg: guest disk configuration file. Only edited using this script.
# (dtb, kernel image, etc)
# disk: directory where files for disk creation are stored
# init.[pre,d,post]: directories containing init pre, base and post scripts
set -u
set -e
this="$0"
IMAGE_TMPDIR=""
usage() {
cat <<EOF
Usage $this ACTION XENGUEST [ARGS]
Where XENGUEST is a xenguest image file or a xenguest directory.
The following actions are supported:
help Display this help
create Create a xenguest image
update Update/modify a xenguest image
partial Create partial xenguest image in a directory
pack Pack a xenguest directory into an image
check Check a xenguest image
dump-paramsconfig Display the guest configuration of a xenguest image
dump-xenconfig Display the xen configuration of a xenguest image
dump-diskconfig Display the disk configuration of a xenguest image
dump-init Display init scripts of a xenguest image
extract Extract a xenguest image content
extract-config Extract the guest configuration from a xenguest image
extract-disk-file Extract a disk file from a xenguest image
Use $this ACTION --help to have help on a specific action and its arguments.
EOF
}
usage-check() {
cat <<EOF
Usage $this check XENGUEST
Check a xenguest image or a xenguest directory.
EOF
}
usage-update-create() {
cat <<EOF
All arguments are handled in order.
Global configuration for the guest
--guest-config-reset reset guest global configuration
--set-param=PARAM disable parameter PARAM in guest global configuration
--set-param=PARAM=VAL set parameter PARAM to value VAL in guest global
configuration.
Example of parameters supported are:
GUEST_AUTOBOOT: if set to 1 (default), guest will be
automatically created and started during host init.
Xen configuration for the guest
--xen-reset-config reset xen guest configuration to default
--xen-name= disable name parameter in xen configuration
--xen-name=NAME set guest name in xen configuration
--xen-kernel= disable guest kernel parameter in xen configuration
--xen-kernel=FILE set guest kernel to FILE (file is added and xen
configuration is set to use it)
--xen-memory set guest memory size (in MB)
--xen-vcpus set guest number of virtual cpus
--xen-clean-extra set guest command line (extra) to an empty string
--xen-extra=ARG append ARG to the guest command line (with space)
use this several time to set command line.
To set the command line to "console=ttyS0 rw" do
--xen-extra=console=ttyS0 --xen-extra=rw
--xen-root= disable root parameter in xen configuration
--xen-root=ROOT set guest root in xen configuration
--xen-device-tree= disable device tree parameter in xen configuration
--xen-device-tree=FILE set guest device tree in xen configuration and add
file to xen files
--xen-ramdisk= disable ramdisk parameter in xen configuration
--xen-ramdisk=FILE set guest ramdisk in xen configuration and add
file to xen files
--xen-disk= disable disk parameter in xen configuration
--xen-disk=DEV set guest disk to device DEV (phy:DEV,xvda,w is set)
--xen-append=FILE append FILE content to xen configuration
Xen files
--xen-add-file=SRC:DST add file SRC as file DST in the xenguest image.
If DST already exist in the image, it is overwritten.
DST must be the same as arguments passed to other
options (like --xen-kernel)
--xen-rm-file=DST remove file DST from the xenguest image.
Init configuration
This can be used to add init scripts for the guest. There are 3 possible init
scripts which are called at different time. The pre scripts are called first,
then the xen guest is created in pause and standard init scripts are called.
Finally the xen guest is started then the post init scripts are called.
Each script is called with the name of the guest as first argument and
multiple scripts can be added (they must have different names).
--init-script=FILE add FILE as init script
--init-pre=FILE add FILE as pre init script
--init-post=FILE add FILE as post init script
Disk configuration
--disk-reset-config reset disk guest configuration to default (no disk)
--disk-size=SZ set guest disk size (in GB)
--disk-device=DEV set device to be used to create the guest disk
if unset or set to an empty string, the volume will be
create in the default manager volume group.
--disk-add-part=DEF add a partition to the guest disk with definition DEF
a partition definition must have the following format:
ID:SIZE:FORMAT:CONTENT where:
- ID is the partition numeric ID (1 to 4)
- FORMAT is the filesystem format (supported formats
are none, vfat, swap, ext2, ext3 and ext4) or can be
left empty to not format
- CONTENT can be used to point to a file added using
--disk-add-file to be used as partition initial
content (tar file or img file)
--disk-rm-part=ID remove partition ID from the guest disk
--disk-add-file=SRC:DST add file SRC as disk file DST in the xenguest image.
DST can then be used as a partition CONTENT.
--disk--rm-file=DST remove disk file DST from the xenguest image.
EOF
}
usage-create() {
cat <<EOF
Usage $this create XENGUEST [ARGS]
Create a xenguest image as XENGUEST file.
EOF
usage-update-create
}
usage-update() {
cat <<EOF
Usage $this update XENGUEST [ARGS]
Update or modify a xenguest image or a xenguest directory.
EOF
usage-update-create
}
usage-pack() {
cat <<EOF
Usage $this pack XENGUEST DESTFILE
Pack a xenguest directory in XENGUEST to create a xenguest image DESTFILE.
EOF
}
usage-partial() {
cat <<EOF
Usage $this partial XENGUEST [ARGS]
Update or modify a partial xenguest image.
EOF
usage-update-create
}
usage-dump-paramsconfig() {
cat <<EOF
Usage $this dump-paramsconfig XENGUEST
Dump the guest parameters of a xenguest image or directory
EOF
}
usage-dump-xenconfig() {
cat <<EOF
Usage $this dump-xenconfig XENGUEST
Dump the xen configuration of a xenguest image or directory
EOF
}
usage-dump-diskconfig() {
cat <<EOF
Usage $this dump-diskconfig XENGUEST
Dump the disk configuration of a xenguest image or directory
EOF
}
usage-extract() {
cat <<EOF
Usage $this extract XENGUEST DESTDIR
Extract guest to DESTDIR
EOF
}
usage-extract-config() {
cat <<EOF
Usage $this extract-config XENGUEST DESTDIR
Extract guest configuration to DESTDIR
EOF
}
usage-extract-disk-file() {
cat <<EOF
Usage $this extract-disk-file XENGUEST DISKFILE
Extract disk file DISKFILE to stdout.
EOF
}
check_image() {
local tstfile=${1}
if [ ! -e ${tstfile} -o ! -w ${tstfile} ]; then
echo "Error: File ${tstfile} does not exist or is not writeable"
exit 1
fi
if [ -f ${tstfile} ]; then
# This is a xenguest file
local res=$(tar -tvf ${tstfile} ./guest.cfg ./disk.cfg \
./params.cfg > /dev/null 2>&1 || echo "error")
if [ -n "${res}" ]; then
echo "Error: File ${tstfile} is not a valid xenguest"
exit 1
fi
elif [ -d ${tstfile} ]; then
if [ ! -f ${tstfile}/guest.cfg -o ! -f ${tstfile}/disk.cfg -o \
! ${tstfile}/params.cfg ]; then
echo "Error: Directory ${tstfile} is not a valid xenguest"
exit 1
fi
fi
}
params_config_reset() {
cat <<EOF > ${IMAGE_TMPDIR}/params.cfg
# Xenguest-image guest global configuration
#
# !! This file must not be modified manually !!
#
# You can use xenguest-image to modify parameters.
#
# Guest auto boot during Dom0 init
GUEST_AUTOBOOT="1"
EOF
}
params_config_setparam() {
param="${1}"
shift
value="$@"
if [ -z "$value" ]; then
sed -i "/.*${param}=.*/d" ${IMAGE_TMPDIR}/params.cfg
elif grep -e "^${param}=" ${IMAGE_TMPDIR}/params.cfg > /dev/null; then
sed -i "s/${param}=\".*\"/${param}=\"${value}\"/" \
${IMAGE_TMPDIR}/params.cfg
else
echo "${param}=\"${value}\"" >> ${IMAGE_TMPDIR}/params.cfg
fi
}
xen_config_reset() {
cat <<EOF > ${IMAGE_TMPDIR}/guest.cfg
# Xenguest-image main configuraiton
#
# !! This file must not be modified manually !!
#
# You can use xenguest-image to modify parameters.
#
# You can add custom entries to configuration in the guest.d directory.
# Guest name (set by manager when guest is created)
# name = ""
# Guest memory size in MB
memory = 1024
# Number of VCPUS
vcpus = 1
# Guest command line
extra = "earlyprintk=xenboot console=hvc0 rw"
# Guest root filesystem device (from guest point of view)
# root = "/dev/xvda2"
# Disk that will be used by the guest (set by manager when guest is created)
# disk = ['phy:/dev/vg-xen/guestname,xvda,w']
EOF
}
get_param_file() {
param="${1}"
if grep ${param} ${IMAGE_TMPDIR}/guest.cfg > /dev/null 2>&1; then
echo "${IMAGE_TMPDIR}/guest.cfg"
else
if [ ! -f ${IMAGE_TMPDIR}/guest.d/${param}.cfg ]; then
mkdir -p ${IMAGE_TMPDIR}/guest.d
echo "# ${param} = \"\"" > ${IMAGE_TMPDIR}/guest.d/${param}.cfg
fi
echo "${IMAGE_TMPDIR}/guest.d/${param}.cfg"
fi
}
xen_config_disable_param() {
param="${1}"
dst=$(get_param_file ${param})
sed -i "s@.*\(${param} = .*\)\$@# \1@" ${dst}
}
xen_config_set_number() {
param="${1}"
shift
value="$@"
dst=$(get_param_file ${param})
sed -i "s@.*${param} = .*@${param} = ${value}@" ${dst}
}
xen_config_set_string() {
param="${1}"
shift
value="$@"
dst=$(get_param_file ${param})
sed -i "s@.*${param} = .*@${param} = \"${value}\"@" ${dst}
}
xen_config_append_string() {
param="${1}"
shift
value="$@"
dst=$(get_param_file ${param})
sed -i "s@.*${param} = \"\([^\"]*\)\"@${param} = \"\1 ${value}\"@" ${dst}
}
xen_config_set_list() {
param="${1}"
shift
value=$(echo $@ | tr " " ",")
dst=$(get_param_file ${param})
sed -i "s@.*${param} = .*@${param} = ['${value}']@" ${dst}
}
disk_config_reset() {
echo "DISK_SIZE=\"0\"" > ${IMAGE_TMPDIR}/disk.cfg
echo "DISK_DEVICE=\"\"" >> ${IMAGE_TMPDIR}/disk.cfg
}
disk_config_rm_part() {
partid=$1
sed -i "/DISK_PART${partid}=.*/d" ${IMAGE_TMPDIR}/disk.cfg
}
disk_config_add_part() {
partconf="${1}"
partid=$(echo ${partconf} | sed -e "s/:.*//")
partinfo=$(echo ${partconf} | sed -e "s/[^:]*://")
# Make sure we don't add the same partition twice
disk_config_rm_part ${partid}
echo "DISK_PART${partid}=\"${partinfo}\"" >> \
${IMAGE_TMPDIR}/disk.cfg
}
# We need an action as first argument
action="${1:-}"
if [ -z "${action}" ]; then
echo "Error: No ACTION provided"
usage
exit 1
fi
# Only help does not require a xenguest argument so treat this first
# while there we also check that user is asking for a supported action
case $action in
help|--help|-h|-?)
usage
exit 0
;;
check|create|update|pack|partial)
;;
dump-xenconfig|dump-diskconfig|dump-init|dump-paramsconfig)
;;
extract|extract-config|extract-disk-file)
;;
*)
echo "Error: Invalid action $action"
exit 1
;;
esac
# Second argument should be the file name or directory
guestfile="${2:-}"
# Handle user asking for help on a specific action
case $guestfile in
help|--help|-h|-?)
usage-${action}
exit 0
;;
esac
if [ -z "${guestfile}" ]; then
echo "Error: no GUESTFILE provided"
usage
exit 1
fi
shift 2
case ${action} in
check)
check_image ${guestfile}
echo "Image is OK"
exit 0
;;
dump-paramsconfig)
check_image ${guestfile}
echo "Guest configuration:"
if [ -f ${guestfile} ]; then
tar -xOf ${guestfile} ./params.cfg
else
cat ${guestfile}/params.cfg
fi
exit 0
;;
dump-xenconfig)
check_image ${guestfile}
echo "Xen configuration:"
if [ -f ${guestfile} ]; then
tar -xOf ${guestfile} ./guest.cfg
tar -xOf ${guestfile} ./guest.d 2> /dev/null || true
else
cat ${guestfile}/guest.cfg
cat ${guestfile}/guest.d/* 2> /dev/null || true
fi
echo
exit 0
;;
dump-diskconfig)
check_image ${guestfile}
echo "Disk configuration:"
if [ -f ${guestfile} ]; then
tar -xOf ${guestfile} ./disk.cfg
else
cat ${guestfile}/disk.cfg
fi
echo
exit 0
;;
dump-init)
check_image ${guestfile}
for init in init.d init-pre init-post; do
echo "=== ${init} ==="
if [ -f ${guestfile} ]; then
tar -xOf ${guestfile} ./${init} 2> /dev/null || \
echo "No ${init} scripts."
else
cat ${guestfile}/${init}/* 2> /dev/null || \
echo "No ${init} scripts."
fi
echo "==============="
echo
done
exit 0
;;
pack)
check_image ${guestfile}
if [ ! -d ${guestfile} ]; then
echo "Error: Pack can only be done on a xenguest directory"
exit 1
fi
if [ -z "${1:-}" ] || [ -f ${1} ]; then
echo "Error: No destination file or already existing file"
exit 1
fi
tar -C ${guestfile} -cf ${1} .
exit 0
;;
extract)
check_image ${guestfile}
if [ -d ${guestfile} ]; then
echo "Error: Cannot extract config from xenguest directory"
exit 1
fi
if [ -z "${1:-}" ] || [ ! -d ${1} ]; then
echo "Error: No destination directory for image extract"
exit 1
fi
tar -C ${1} -xf ${guestfile}
exit 0
;;
extract-config)
check_image ${guestfile}
if [ -d ${guestfile} ]; then
echo "Error: Cannot extract config from xenguest directory"
exit 1
fi
if [ -z "${1:-}" ] || [ ! -d ${1} ]; then
echo "Error: No destination directory for config extract"
exit 1
fi
#extract all but disk files
tar -C ${1} --exclude='./disk' -xf ${guestfile}
exit 0
;;
extract-disk-file)
check_image ${guestfile}
if [ -d ${guestfile} ]; then
echo "Error: Cannot extract disk file from xenguest directory" >&2
exit 1
fi
if [ -z "${1:-}" ]; then
echo "Error: No file to extract" >&2
exit 1
fi
tar -xOf ${guestfile} ./disk/${1}
exit 0
;;
create)
if [ -f ${guestfile} ]; then
echo "Error: File ${guestfile} already exist"
exit 1
elif [ -d ${guestfile} ]; then
if [ -n "$(ls -A ${guestfile})" ]; then
echo "Error: Directory ${guestfile} is not empty"
exit 1
fi
IMAGE_TMPDIR=$(realpath -m ${guestfile})
else
IMAGE_TMPDIR=$(mktemp -d)
fi
# Create initial content
params_config_reset
xen_config_reset
disk_config_reset
;;
update)
check_image ${guestfile}
if [ -f ${guestfile} ]; then
# Extract the image to update it
IMAGE_TMPDIR=$(mktemp -d)
tar -C ${IMAGE_TMPDIR} -xf ${guestfile}
else
IMAGE_TMPDIR=$(realpath -m ${guestfile})
fi
;;
partial)
if [ -e ${guestfile} -a ! -d ${guestfile} ]; then
echo "Error: Invalid partial output directory"
exit 1
fi
mkdir -p ${guestfile}
IMAGE_TMPDIR=$(realpath -m ${guestfile})
;;
*)
echo "Invalid action ${action}"
usage
exit 1
;;
esac
# Process command line arguments
for arg in "${@}"; do
case ${arg} in
--*=*)
optarg=$(echo ${arg} | sed -e "s/[^=]*=//")
;;
*)
optarg=""
;;
esac
case ${arg} in
--guest-reset-config)
params_config_reset
;;
--set-param=*=*)
param_name=$(echo $optarg | sed -e "s/=.*//")
param_value=$(echo $optarg | sed -e "s/[^=]*=//")
params_config_setparam "$param_name" "$param_value"
;;
--set-param=*)
params_config_setparam "$optarg"
;;
--xen-reset-config)
xen_config_create
;;
--xen-name=*)
if [ -z "${optarg}" ]; then
xen_config_disable_param "name"
else
xen_config_set_string "name" "${optarg}"
fi
;;
--xen-kernel=*)
if [ -z "${optarg}" ]; then
xen_config_disable_param "kernel"
rm -f ${IMAGE_TMPDIR}/files/kernel
else
if [ ! -f ${optarg} ]; then
echo "Error: invalid kernel file ${optarg}"
exit 1
fi
xen_config_set_string "kernel" "files/kernel"
mkdir -p ${IMAGE_TMPDIR}/files
install -m 644 ${optarg} ${IMAGE_TMPDIR}/files/kernel
fi
;;
--xen-memory=*)
xen_config_set_number "memory" ${optarg}
;;
--xen-vcpus=*)
xen_config_set_number "vcpus" ${optarg}
;;
--xen-clean-extra)
xen_config_set_string "extra" ""
;;
--xen-extra=*)
xen_config_append_string "extra" ${optarg}
;;
--xen-root=*)
if [ -z "${optarg}" ]; then
xen_config_disable_param "root"
else
xen_config_set_string "root" "${optarg}"
fi
;;
--xen-device-tree=*)
if [ -z "${optarg}" ]; then
xen_config_disable_param "device_tree"
rm -f ${IMAGE_TMPDIR}/files/guest.dtb
else
if [ ! -f ${optarg} ]; then
echo "Error: invalid dtb file ${optarg}"
exit 1
fi
xen_config_set_string "device_tree" "files/guest.dtb"
mkdir -p ${IMAGE_TMPDIR}/files
install -m 644 ${optarg} ${IMAGE_TMPDIR}/files/guest.dtb
fi
;;
--xen-ramdisk=*)
if [ -z "${optarg}" ]; then
xen_config_disable_param "ramdisk"
rm -f ${IMAGE_TMPDIR}/files/ramdisk.img
else
if [ ! -f ${optarg} ]; then
echo "Error: invalid ramdisk file ${optarg}"
exit 1
fi
xen_config_set_string "ramdisk" "files/ramdisk.img"
mkdir -p ${IMAGE_TMPDIR}/files
install -m 644 ${optarg} ${IMAGE_TMPDIR}/files/ramdisk.img
fi
;;
--xen-disk=*)
if [ -z "${optarg}" ]; then
xen_config_disable_param "disk"
else
xen_config_set_list "disk" "phy:${optarg}" "xvda" "w"
fi
;;
--xen-append=*)
if [ ! -f ${optarg} ]; then
echo "Error: invalid xen append file ${optarg}"
exit 1
fi
mkdir -p ${IMAGE_TMPDIR}/guest.d
install -m 755 ${optarg} ${IMAGE_TMPDIR}/guest.d/.
;;
--xen-add-file=*)
src=$(echo "${optarg}" | sed -e "s/:.*//")
dst=$(echo "${optarg}" | sed -e "s/.*://")
if [ ! -f ${src} ]; then
echo "Error: Invalid file: ${src}"
rm -rf ${IMAGE_TMPDIR}
exit 1
fi
if [ -z "${dst}" ]; then
dst=$(basename ${src})
fi
mkdir -p ${IMAGE_TMPDIR}/files/$(dirname ${dst})
cp -f ${src} ${IMAGE_TMPDIR}/files/${dst}
;;
--xen-rm-file=*)
rm -f ${IMAGE_TMPDIR}/files/${optarg}
;;
--init-script=*|--init-pre=*|--init-post=*)
dst=""
case $arg in
--init-script=*)
dst="init.d"
;;
--init-pre=*)
dst="init.pre"
;;
--init-post=*)
dst="init.post"
;;
esac
if [ ! -f ${optarg} ]; then
echo "${optarg} does not point to a valid file"
exit 1
else
mkdir -p ${IMAGE_TMPDIR}/${dst}
install -m 755 ${optarg} ${IMAGE_TMPDIR}/${dst}/.
fi
;;
--disk-reset-config)
disk_config_reset
;;
--disk-size=*)
sed -i "s/DISK_SIZE=.*/DISK_SIZE=\"${optarg}\"/" \
${IMAGE_TMPDIR}/disk.cfg
;;
--disk-device=*)
sed -i "s/DISK_DEVICE=.*/DISK_SIZE=\"${optarg}\"/" \
${IMAGE_TMPDIR}/disk.cfg
;;
--disk-add-part=*)
disk_config_add_part ${optarg}
;;
--disk-rm-part=*)
disk_config_rm_part ${optarg}
;;
--disk-add-file=*)
src=$(echo "${optarg}" | sed -e "s/:.*//")
dst=$(echo "${optarg}" | sed -e "s/.*://")
if [ ! -f ${src} ]; then
echo "Error: Invalid disk file: ${src}"
rm -rf ${IMAGE_TMPDIR}
exit 1
fi
if [ -z "${dst}" ]; then
dst=$(basename ${src})
fi
mkdir -p ${IMAGE_TMPDIR}/disk/$(dirname ${dst})
cp -f ${src} ${IMAGE_TMPDIR}/disk/${dst}
;;
--disk-rm-file=*)
rm -f ${IMAGE_TMPDIR}/disk/${optarg}
;;
*)
echo "Unsupported command: ${arg}"
exit 1
;;
esac
done
if [ ! -d ${guestfile} ]; then
# If the original guest was in a file we need to repack the file
# with the changes we did on it in the IMAGE_TMPDIR
rm -f ${guestfile}
tar -C ${IMAGE_TMPDIR} -cf ${guestfile} .
rm -rf ${IMAGE_TMPDIR}
fi
@@ -1,3 +0,0 @@
# Xenguest Network Bridge interface configuration
auto ###BRIDGE_NAME###
iface ###BRIDGE_NAME### inet dhcp
@@ -1,44 +0,0 @@
#!/bin/sh
#
# Xenguest Network Bridge init
# This script creates a network bridge and add host interfaces to it
# It will then be used by xen guests to connect to the external nework
#
INTFS="###BRIDGE_MEMBERS###"
BR_INTF="###BRIDGE_NAME###"
case "$1" in
start)
echo "Starting $BR_INTF"
brctl addbr $BR_INTF
for intf in $INTFS; do
echo "Adding $intf to $BR_INTF"
brctl addif $BR_INTF $intf
done
;;
status)
true
;;
stop)
echo "Stopping $BR_INTF"
ifdown $BR_INTF
brctl delbr $BR_INTF
;;
reload)
echo >&2 'Reload not available; use force-reload'; exit 1
;;
force-reload|restart)
echo "Restarting host-bridge"
$0 stop
$0 start
;;
*)
# do not advertise unreasonable commands that there is no reason
# to use with this device
echo $"Usage: $0 {start|stop|status|restart|force-reload}"
exit 1
esac
exit $?
@@ -1,45 +0,0 @@
#!/bin/sh
# This script is to reload kea dhcp4 server when guest interface will be ready
# include locking functions
. /etc/xen/scripts/locking.sh
set +u
# $1 is vif name, e.g. "vif/15/0"
# returns 0 on success,
# 1 otherwize
#
check_if_vif_is_ready() {
ret=($(xl network-list "${guestname}" | grep "${1}"))
# ${ret[4]} is network interface status value
# 1 means vif is not ready
# 4 means vif is ready
[ "${ret[4]}" = "4" ] && return 0
return 1
}
case "${XENGUEST_NETWORK_TYPE:-}" in
nat)
vif_name="$(xl network-list ${guestname} | grep -o vif.*)"
for try in {1..20}
do
if check_if_vif_is_ready "${vif_name}"; then
claim_lock "vif-nat-kea"
keactrl reload
release_lock "vif-nat-kea"
exit 0
fi
echo "Waiting for ${vif_name} - network interface is not ready..."\
" try #${try}" >> "${LOGFILE}" 2>&1
sleep 1
done
echo "ERROR: Failed to get ${vif_name} "\
"network interface ready!" >> "${LOGFILE}" 2>&1
exit 1
;;
*)
echo "No action needed" >> "${LOGFILE}" 2>&1
;;
esac
@@ -1,149 +0,0 @@
# Create a xenguest base image
#
# This recipe creates a base image that is then extended by other recipes
# through xenguest-image class.
# xenguest image type is using this as base to add a kernel and a disk image
# to create a guest
#
# The recipe is also adding files in those directories to the xenguest image:
# - ${WORKDIR}/extend/disk-files: all files in this directory will be added to
# the guest disk files (using --disk-add-file)
# - ${WORKDIR}/extend/files: all files in this directory will be added to the
# guest xen files (using --xen-add-file)
# - ${WORKDIR}/extend/guest.d: all files in this directory will be added to
# the xen append configuration files (using --xen-append)
# - ${WORKDIR}/extend/init.[pre|post|d]: all files in those directories will
# be added to the corresponding init scripts (using --init-[pre|post|script])
# You can bbappend this recipe and put files in ${WORKDIR}/extend to add
# elements to the image.
#
DESCRIPTION = "Xenguest Base Image"
LICENSE = "MIT"
LIC_FILES_CHKSUM = "file://${COMMON_LICENSE_DIR}/MIT;md5=0835ade698e0bcf8506ecda2f7b4f302"
# When XENGUEST_IMAGE_NETWORK_TYPE="nat", the "00-xenguest-nat-port-forward.hook"
# is called by "/etc/xen/scripts/vif-post.d/00-vif-xenguest.hook" to apply NAT
# port forwarding. Both dom0 and domU ports can be be set by changing the
# XENGUEST_IMAGE_HOST_PORT and XENGUEST_IMAGE_GUEST_PORT variables in local.conf
# or xenguest-base-image.bbappend. The XENGUEST_IMAGE_NAT_PORT_FORWARD_SCRIPT
# can also be replaced in a xenguest-base-image.bbappend
XENGUEST_IMAGE_HOST_PORT ?= "\$( expr 1000 + \${domid} )"
XENGUEST_IMAGE_GUEST_PORT ?= "22"
XENGUEST_IMAGE_NAT_PORT_FORWARD_SCRIPT ?= "00-xenguest-nat-port-forward.hook"
#
# The following variables can contain SRC_URI compatible entries to add
# files to the xenguest image.
# You can set those variable in local.conf to add one or several files
# For example to add a boot.tar.gz file that has to be downloaded to the file
# useable for disk partition initialisation:
# XENGUEST_IMAGE_SRC_URI_DISK_FILES += "https://www.test.com/files/boot.tar.gz"
# Add disk files
XENGUEST_IMAGE_SRC_URI_DISK_FILES ??= ""
# Add xen files
# Any extrafiles files to be added to XENGUEST_IMAGE_SRC_URI_XEN_FILES should
# be performed via XENGUEST_IMAGE_SRC_URI_XEN_FILES_append.
# The kea-subnet4.json holds the kea dhcp4 subnet configuration for Dom0.
# And it is used when XENGUEST_IMAGE_NETWORK_TYPE="nat".
# Any customizations to it should be performed by replacing it
# via a xenguest-network.bbappend.
# The XENGUEST_IMAGE_NAT_PORT_FORWARD_SCRIPT file is only added if the
# variable is set.
XENGUEST_IMAGE_SRC_URI_XEN_FILES = "file://kea-subnet4.json \
${@ "file://" + d.getVar('XENGUEST_IMAGE_NAT_PORT_FORWARD_SCRIPT') \
if d.getVar('XENGUEST_IMAGE_NAT_PORT_FORWARD_SCRIPT') else "" } \
"
# Add xen configuration elements
XENGUEST_IMAGE_SRC_URI_XEN_CONFIG ??= ""
# Add pre init script
XENGUEST_IMAGE_SRC_URI_INIT_PRE ??= ""
# Add init script
XENGUEST_IMAGE_SRC_URI_INIT ??= ""
# Add post init script
XENGUEST_IMAGE_SRC_URI_INIT_POST ??= ""
S = "${WORKDIR}"
inherit deploy xenguest-image
# parse XENGUEST_IMAGE_SRC_URI_ variables and add them to SRC_URI
python __anonymous() {
def parse_extend_variable(d, varname, destdir):
list = d.getVar(varname)
if list:
for entry in list.split():
#Check the URL
try:
decode = bb.fetch.decodeurl(entry)
d.appendVar('SRC_URI', ' ' + entry + ';unpack=0;subdir=extend/' + destdir)
except:
bb.fatal("%s: %s contains an invalid URL: %s" \
% (d.getVar('PF'), varname, entry))
parse_extend_variable(d, 'XENGUEST_IMAGE_SRC_URI_DISK_FILES', 'disk-files')
parse_extend_variable(d, 'XENGUEST_IMAGE_SRC_URI_XEN_FILES', 'files')
parse_extend_variable(d, 'XENGUEST_IMAGE_SRC_URI_XEN_CONFIG', 'guest.d')
parse_extend_variable(d, 'XENGUEST_IMAGE_SRC_URI_INIT_PRE', 'init.pre')
parse_extend_variable(d, 'XENGUEST_IMAGE_SRC_URI_INIT', 'init.d')
parse_extend_variable(d, 'XENGUEST_IMAGE_SRC_URI_INIT_POST', 'init.post')
}
# Make sure we are removing old files before redoing a fetch
do_fetch[cleandirs] += "${WORKDIR}/extend"
do_fetch[vardeps] += "XENGUEST_IMAGE_HOST_PORT XENGUEST_IMAGE_GUEST_PORT"
do_compile[noexec] = "1"
do_install[noexec] = "1"
add_extend_files() {
local subdir="$1"
local cmd="$2"
local stripdest="${3:-n}"
if [ -d ${WORKDIR}/extend/$subdir ]; then
filelist=$(find ${WORKDIR}/extend/$subdir -type f)
if [ -n "$filelist" ]; then
for f in $filelist; do
if [ "$stripdest" = "y" ]; then
call_xenguest_mkimage update --$cmd=$f:$(basename $f)
else
call_xenguest_mkimage update --$cmd=$f
fi
done
fi
fi
}
do_configure() {
if [ -f ${WORKDIR}/extend/files/${XENGUEST_IMAGE_NAT_PORT_FORWARD_SCRIPT} ]; then
sed -i "s,###HOST_PORT###,${XENGUEST_IMAGE_HOST_PORT}," \
${WORKDIR}/extend/files/${XENGUEST_IMAGE_NAT_PORT_FORWARD_SCRIPT}
sed -i "s,###GUEST_PORT###,${XENGUEST_IMAGE_GUEST_PORT}," \
${WORKDIR}/extend/files/${XENGUEST_IMAGE_NAT_PORT_FORWARD_SCRIPT}
fi
}
do_deploy() {
# Create a new image
xenguest_image_create
# Add our extra files if any
add_extend_files "disk-files" "disk-add-file" "y"
add_extend_files "files" "xen-add-file" "y"
add_extend_files "guest.d" "xen-append"
add_extend_files "init.pre" "init-pre"
add_extend_files "init.d" "init-script"
add_extend_files "init.post" "init-post"
}
addtask deploy after do_install before do_build
@@ -1,52 +0,0 @@
# Xenguest manager recipe
#
# xenguest-manager is a tool to control xen guest (e.g. create, start, stop)
#
DESCRIPTION = "Xen Guest Manager"
LICENSE = "MIT"
SRC_URI = " \
file://xenguest-manager \
file://xenguest-init \
"
LIC_FILES_CHKSUM = "file://${COMMON_LICENSE_DIR}/MIT;md5=0835ade698e0bcf8506ecda2f7b4f302"
S = "${WORKDIR}"
# Please refer to documentation/xenguest-manager.md for documentation on those
# parameters
XENGUEST_MANAGER_VOLUME_DEVICE ?= "/dev/sda2"
XENGUEST_MANAGER_VOLUME_NAME ?= "vg-xen"
XENGUEST_MANAGER_GUEST_DIR ?= "${datadir}/guests/"
# We add an init script to create and start guests automatically
INITSCRIPT_NAME = "xenguest"
INITSCRIPT_PARAMS = "defaults 90"
inherit update-rc.d
do_compile() {
echo "XENGUEST_VOLUME_DEVICE=\"${XENGUEST_MANAGER_VOLUME_DEVICE}\"" > \
xenguest-manager.conf
echo "XENGUEST_VOLUME_NAME=\"${XENGUEST_MANAGER_VOLUME_NAME}\"" >> \
xenguest-manager.conf
echo "XENGUEST_GUEST_DIR=\"${XENGUEST_MANAGER_GUEST_DIR}\"" >> \
xenguest-manager.conf
}
do_install() {
install -d -m 755 ${D}${bindir}
install -m 755 xenguest-manager ${D}${bindir}/.
install -d -m 755 ${D}${sysconfdir}/xenguest
install -m 644 xenguest-manager.conf ${D}${sysconfdir}/xenguest/.
install -d -m 755 ${D}${sysconfdir}/init.d
install -m 755 xenguest-init ${D}${sysconfdir}/init.d/${INITSCRIPT_NAME}
install -d -m 755 ${D}${XENGUEST_GUEST_DIR}
}
# Things that we need on the target
RDEPENDS_${PN} += "bash tar xenguest-mkimage lvm2 xen-tools parted e2fsprogs dosfstools"
FILES_${PN} += "${bindir}/xenguest-manager \
${sysconfdir}/xenguest"
@@ -1,31 +0,0 @@
# Xenguest mkimage recipe
#
# xenguest-mkimage is a tool to create/modify images to be used as xen guests
# Produced images contains a xen configuration and several optional components
# (kernel, device-tree, disk definition and files, init scripts) which all
# together fully define a full xen guest
DESCRIPTION = "Xenguest mkimage tool"
LICENSE = "MIT"
SRC_URI = "file://xenguest-mkimage"
LIC_FILES_CHKSUM = "file://${COMMON_LICENSE_DIR}/MIT;md5=0835ade698e0bcf8506ecda2f7b4f302"
S = "${WORKDIR}"
# Can be built native also to produce xenguest images during Yocto build
BBCLASSEXTEND = "native"
do_configure[noexec] = "1"
do_compile[noexec] = "1"
do_install() {
install -d -m 755 ${D}${bindir}
install -m 755 xenguest-mkimage ${D}${bindir}/.
}
# We need bash and tar
RDEPENDS_${PN} = "bash tar"
FILES_${PN} = "${bindir}/xenguest-mkimage"
@@ -1,95 +0,0 @@
# Recipe to handle xenguest network configuration
DESCRIPTION = "Xenguest Network"
LICENSE = "MIT"
LIC_FILES_CHKSUM = "file://${COMMON_LICENSE_DIR}/MIT;md5=0835ade698e0bcf8506ecda2f7b4f302"
S = "${WORKDIR}"
# Please refer to documentation/xenguest-network-bridge.md for documentation on
# those parameters
XENGUEST_NETWORK_BRIDGE_NAME ?= "xenbr0"
# The XENGUEST_NETWORK_BRIDGE_MEMBERS should be set in a machine.conf
# or bbappend file.
#XENGUEST_NETWORK_BRIDGE_MEMBERS ?= "eth0"
XENGUEST_NETWORK_BRIDGE_CONFIG ?= "xenguest-network-bridge-dhcp.cfg.in"
SRC_URI = " \
file://xenguest-network-bridge.in \
file://xenguest-network-bridge-dhcp.cfg.in \
file://network-bridge.sh.in \
file://00-vif-xenguest.hook \
file://xenguest-network-init-post.sh \
file://kea-dhcp4.conf \
file://kea-restore-default-config \
"
PACKAGES =+ "${PN}-kea-dhcp4"
# Bridge configurator needs to run before S01networking init script
# Prefix with a_ to make sure it is executed in runlevel 01 before others
INITSCRIPT_PACKAGES = "${PN} ${PN}-kea-dhcp4"
INITSCRIPT_NAME_${PN} = "a_xenguest-network-bridge"
INITSCRIPT_PARAMS_${PN} = "defaults 01"
# Kea configuration needs to be restored before kea init scripts:
# Kea dhcp4 server is 30, so lets use 20, to have higher priority
INITSCRIPT_NAME_${PN}-kea-dhcp4 = "kea-restore-default-config"
INITSCRIPT_PARAMS_${PN}-kea-dhcp4 = "defaults 20"
inherit update-rc.d
do_install() {
cat ${WORKDIR}/xenguest-network-bridge.in \
| sed -e "s,###BRIDGE_MEMBERS###,${XENGUEST_NETWORK_BRIDGE_MEMBERS}," \
| sed -e "s,###BRIDGE_NAME###,${XENGUEST_NETWORK_BRIDGE_NAME}," \
> ${WORKDIR}/xenguest-network-bridge
cat ${WORKDIR}/${XENGUEST_NETWORK_BRIDGE_CONFIG} \
| sed -e "s,###BRIDGE_NAME###,${XENGUEST_NETWORK_BRIDGE_NAME}," \
> ${WORKDIR}/xenguest-network-bridge.cfg
cat ${WORKDIR}/network-bridge.sh.in \
| sed -e "s,###BRIDGE_NAME###,${XENGUEST_NETWORK_BRIDGE_NAME}," \
> ${WORKDIR}/network-bridge.sh
install -d -m 755 ${D}${sysconfdir}/init.d
install -m 755 ${WORKDIR}/xenguest-network-bridge \
${D}${sysconfdir}/init.d/${INITSCRIPT_NAME_${PN}}
install -d -m 755 ${D}${sysconfdir}/network/interfaces.d
install -m 755 ${WORKDIR}/xenguest-network-bridge.cfg \
${D}${sysconfdir}/network/interfaces.d/.
install -d -m 755 ${D}${sysconfdir}/xenguest/init.pre
install -m 755 ${WORKDIR}/network-bridge.sh \
${D}${sysconfdir}/xenguest/init.pre/.
install -d ${D}${sysconfdir}/xen/scripts/vif-post.d
install -m 755 ${WORKDIR}/00-vif-xenguest.hook \
${D}${sysconfdir}/xen/scripts/vif-post.d/.
install -d -m 755 ${D}${sysconfdir}/xenguest/init.post
install -m 755 ${WORKDIR}/xenguest-network-init-post.sh \
${D}${sysconfdir}/xenguest/init.post/.
install -m 755 ${WORKDIR}/kea-restore-default-config \
${D}${sysconfdir}/init.d/.
install -d -m 755 ${D}${sysconfdir}/kea/
install -m 755 ${WORKDIR}/kea-dhcp4.conf \
${D}${sysconfdir}/kea/kea-dhcp4.conf.original
}
RDEPENDS_${PN} += "bridge-utils \
iptables \
kea \
${PN}-kea-dhcp4 \
kernel-module-xt-tcpudp \
kernel-module-xt-physdev \
kernel-module-xt-comment \
kernel-module-xt-nat \
kernel-module-xt-masquerade \
"
FILES_${PN} += "${sysconfdir}/network/interfaces.d/xenguest-network-bridge.cfg"
FILES_${PN} += "${sysconfdir}/xenguest/init.pre/network-bridge.sh"
FILES_${PN} += "${sysconfdir}/xen/scripts/vif-post.d/00-vif-xenguest.hook"
FILES_${PN}-kea-dhcp4 = "${sysconfdir}/kea/kea-dhcp4.conf.original"
FILES_${PN}-kea-dhcp4 += "${sysconfdir}/init.d/${INITSCRIPT_NAME_${PN}-kea-dhcp4}"
FILES_${PN}-kea-dhcp4 += "${sysconfdir}/xenguest/init.post/xenguest-network-init-post.sh"
@@ -1,38 +0,0 @@
# Create a xenguest image with kernel but no rootfs or an external rootfs
DESCRIPTION = "Xenguest No Disk Image"
LICENSE = "MIT"
LIC_FILES_CHKSUM = "file://${COMMON_LICENSE_DIR}/MIT;md5=0835ade698e0bcf8506ecda2f7b4f302"
S = "${WORKDIR}"
inherit deploy xenguest-image
# Name of the file we create in deploy
XENGUEST_IMAGE_NODISK_DEPLOY = "xenguest-nodisk-image.xenguest"
# use a local copy to pack all together
XENGUEST_IMAGE_DEPLOY_DIR = "${WORKDIR}/tmp-xenguest"
do_configure[noexec] = "1"
do_compile[noexec] = "1"
do_install[noexec] = "1"
do_deploy() {
xenguest_image_clone
# Add kernel to the image
if [ -n "${XENGUEST_IMAGE_KERNEL}" ]; then
call_xenguest_mkimage partial \
--xen-kernel=${DEPLOY_DIR_IMAGE}/${XENGUEST_IMAGE_KERNEL}
fi
# Pack and deploy the final image
rm -f ${DEPLOYDIR}/${XENGUEST_IMAGE_NODISK_DEPLOY}
call_xenguest_mkimage pack ${DEPLOYDIR}/${XENGUEST_IMAGE_NODISK_DEPLOY}
}
do_deploy[depends] += "virtual/kernel:do_deploy"
do_deploy[depends] += "xenguest-base-image:do_deploy"
addtask deploy before do_build after do_install
@@ -1,5 +0,0 @@
define KMACHINE arm64-autonomy-guest
define KTYPE standard
define KARCH arm64
include arm64-autonomy-guest.scc
@@ -1,2 +0,0 @@
CONFIG_ARM64=y
CONFIG_SMP=y
@@ -1,6 +0,0 @@
include ktypes/standard/standard.scc
include features/input/input.scc
include features/net/net.scc
include cfg/timer/no_hz.scc
kconf hardware arm64-autonomy-guest.cfg
@@ -1,83 +0,0 @@
From f88af7229f6f22ce7313c11a0bbb9c88317b7c5e Mon Sep 17 00:00:00 2001
Message-Id: <f88af7229f6f22ce7313c11a0bbb9c88317b7c5e.1602759709.git.diego.sueiro@arm.com>
From: Stefano Stabellini <stefano.stabellini@xilinx.com>
Date: Thu, 24 Sep 2020 16:49:55 -0700
Subject: [PATCH] xen/arm: do not setup the runstate info page if kpti is
enabled
The VCPUOP_register_runstate_memory_area hypercall takes a virtual
address of a buffer as a parameter. The semantics of the hypercall are
such that the virtual address should always be valid.
When KPTI is enabled and we are running userspace code, the virtual
address is not valid, thus, Linux is violating the semantics of
VCPUOP_register_runstate_memory_area.
Do not call VCPUOP_register_runstate_memory_area when KPTI is enabled.
Signed-off-by: Stefano Stabellini <stefano.stabellini@xilinx.com>
CC: Bertrand Marquis <Bertrand.Marquis@arm.com>
CC: boris.ostrovsky@oracle.com
CC: jgross@suse.com
Link: https://lore.kernel.org/r/20200924234955.15455-1-sstabellini@kernel.org
Reviewed-by: Bertrand Marquis <bertrand.marquis@arm.com>
Signed-off-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Upstream-Status: Backport
Signed-off-by: Diego Sueiro <diego.sueiro@arm.com>
---
arch/arm/include/asm/xen/page.h | 5 +++++
arch/arm/xen/enlighten.c | 6 ++++--
arch/arm64/include/asm/xen/page.h | 6 ++++++
3 files changed, 15 insertions(+), 2 deletions(-)
diff --git a/arch/arm/include/asm/xen/page.h b/arch/arm/include/asm/xen/page.h
index 31bbc80..dc7f6e9 100644
--- a/arch/arm/include/asm/xen/page.h
+++ b/arch/arm/include/asm/xen/page.h
@@ -1 +1,6 @@
#include <xen/arm/page.h>
+
+static inline bool xen_kernel_unmapped_at_usr(void)
+{
+ return false;
+}
diff --git a/arch/arm/xen/enlighten.c b/arch/arm/xen/enlighten.c
index e93145d..ea76562 100644
--- a/arch/arm/xen/enlighten.c
+++ b/arch/arm/xen/enlighten.c
@@ -158,7 +158,8 @@ static int xen_starting_cpu(unsigned int cpu)
BUG_ON(err);
per_cpu(xen_vcpu, cpu) = vcpup;
- xen_setup_runstate_info(cpu);
+ if (!xen_kernel_unmapped_at_usr())
+ xen_setup_runstate_info(cpu);
after_register_vcpu_info:
enable_percpu_irq(xen_events_irq, 0);
@@ -387,7 +388,8 @@ static int __init xen_guest_init(void)
return -EINVAL;
}
- xen_time_setup_guest();
+ if (!xen_kernel_unmapped_at_usr())
+ xen_time_setup_guest();
if (xen_initial_domain())
pvclock_gtod_register_notifier(&xen_pvclock_gtod_notifier);
diff --git a/arch/arm64/include/asm/xen/page.h b/arch/arm64/include/asm/xen/page.h
index 31bbc80..dffdc77 100644
--- a/arch/arm64/include/asm/xen/page.h
+++ b/arch/arm64/include/asm/xen/page.h
@@ -1 +1,7 @@
#include <xen/arm/page.h>
+#include <asm/mmu.h>
+
+static inline bool xen_kernel_unmapped_at_usr(void)
+{
+ return arm64_kernel_unmapped_at_el0();
+}
--
2.7.4
@@ -1,27 +0,0 @@
CONFIG_OVERLAY_FS=y
CONFIG_NAMESPACES=y
CONFIG_NET_NS=y
CONFIG_PID_NS=y
CONFIG_IPC_NS=y
CONFIG_UTS_NS=y
CONFIG_CGROUPS=y
CONFIG_CGROUP_CPUACCT=y
CONFIG_CGROUP_DEVICE=y
CONFIG_CGROUP_FREEZER=y
CONFIG_CGROUP_SCHED=y
CONFIG_CPUSETS=y
CONFIG_MEMCG=y
CONFIG_KEYS=y
CONFIG_POSIX_MQUEUE=y
CONFIG_VETH=y
CONFIG_IPV6=y
CONFIG_BRIDGE=y
CONFIG_BRIDGE_NETFILTER=y
CONFIG_IP_NF_FILTER=m
CONFIG_IP_NF_NAT=m
CONFIG_NF_NAT=m
CONFIG_IP_NF_TARGET_MASQUERADE=m
CONFIG_NETFILTER_XT_MATCH_ADDRTYPE=m
CONFIG_NETFILTER_XT_MATCH_CONNTRACK=m
CONFIG_IP_VS=y
CONFIG_NETFILTER_XT_MATCH_IPVS=m

Some files were not shown because too many files have changed in this diff Show More