This reverts commit 6b56e0b4f0.
commit 6b56e0b4f0 ("linux: disable -Werror") said we needed to disable
CONFIG_WERROR in the kernel because it failed to build the esp-hosted
kernel module. This was very much the hammer approach because a) it'd be
better to fix the warnings in the kernel module or b) disable the
kernel's CONFIG_WERROR *only* if esp-hosted is to be built, via the
ESP_HOSTED_LINUX_CONFIG_FIXUPS variable in
package/esp-hosted/esp-hosted.mk.
This is a simple revert without making any change to any kernel module
(including esp-hosted) because all upstream Buildroot defconfigs with a
kernel module enabled build fine today (albeit with warnings). The two
defconfigs that were pointed at by the now-reverted commit all have been
(updated for and) run on master without failing the build.
Suggested-by: Arnout Vandecappelle <arnout@rnout.be>
Signed-off-by: Quentin Schulz <quentin.schulz@cherry.de>
Signed-off-by: Arnout Vandecappelle <arnout@rnout.be>
The Linux package was not providing any license file when a custom
Linux version was selected.
Fix this by adding a Kconfig option to specify the license file,
with a default value set to the commonly used license paths.
Signed-off-by: Kory Maincent <kory.maincent@bootlin.com>
[Arnout: use a single Kconfig option with conditional prompt]
Signed-off-by: Arnout Vandecappelle <arnout@rnout.be>
vmlinux.efi and vmlinuz.efi are standard file names for LoongArch
EFISTUB kernel.
Introduce them to build system.
Signed-off-by: Jiaxun Yang <jiaxun.yang@flygoat.com>
Reviewed-by: Romain Naour <romain.naour@smile.fr>
Signed-off-by: Julien Olivain <ju.o@free.fr>
Since Linux 6.12, the Buildroot option BR2_LINUX_KERNEL_CUSTOM_DTS_PATH
does not work as expected on arm, arm64, mips and riscv[1]. These are
the architectures that store the in-tree DTS files in vendor-specific
subdirectories of arch/$ARCH/boot/dts/.
BR2_LINUX_KERNEL_CUSTOM_DTS_PATH was introduced in Buildroot 2012.08
(commit 69fc497df0 "Rework support for the device tree"). At the time,
the kernel kept all in-tree DTS files directly in arch/$ARCH/boot/dts/,
and this is where Buildroot drops the user's custom DTS. Vendor-specific
subdirectories appeared in Linux v3.19 for the arm64 architecture, and
this scheme was later adopted by mips, riscv and arm.
For these architectures, Linux 6.12 (commit e7e2941300d2, "kbuild: split
device tree build rules into scripts/Makefile.dtbs") made the DTB build
infrastructure incompatible with the way
BR2_LINUX_KERNEL_CUSTOM_DTS_PATH is implemented. This infrastructure now
expects all DTS files to be in vendor-specific subdirectories on the
architectures that use this scheme.
We can't update easily the current behavior of
BR2_LINUX_KERNEL_CUSTOM_DTS_PATH since it expect a list of files but
can also be used "unexpectedly" with directories [2].
BR2_LINUX_KERNEL_INTREE_DTS_NAME="st/stm32mp135f-dk-mx"
BR2_LINUX_KERNEL_CUSTOM_DTS_PATH="$(BR2_EXTERNAL_ST_PATH)/[...]/linux-dts/st"
In this case, the st directory is copied into the arch/$ARCH/boot/dts/
and BR2_LINUX_KERNEL_INTREE_DTS_NAME is used to build stm32mp135f-dk-mx dtb
as if it was intree.
Introduce BR2_LINUX_KERNEL_CUSTOM_DTS_DIR configuration
parameter to specify a list of directories that are copied as-is over
the arch/<arch>/boot/dts/ directory before building the device tree
blob:
board/acmesystems/acqua-a5/dts/
└── microchip
└── at91-sama5d3_acqua.dts
defconfig:
BR2_LINUX_KERNEL_CUSTOM_DTS_DIR="board/acmesystems/acqua-a5/dts"
Each dts file found is automatically added to the list of devicetree
to build.
BR2_LINUX_KERNEL_CUSTOM_DTS_DIR can also be used for external
devicetree overlays files:
board/ti/am574x-idk/dts/
└── ti
└── omap
└── am57xx-evm.dtso
With this new option, BR2_LINUX_KERNEL_CUSTOM_DTS_PATH is now deprecated.
Note: We want to create a list of dts files (LINUX_DTS_LIST) present in
diectrories listed by BR2_LINUX_KERNEL_CUSTOM_DTS_DIR. But
LINUX_DTS_LIST must not contain BR2_LINUX_KERNEL_CUSTOM_DTS_DIR
paths. Use GNU 'find' print format %P to print each dts file path
without their respective dts overlay directory path.
Do the same for LINUX_DTSO_LIST.
Thanks to Edgar Bonet for the initial contribution [3].
[1] https://lists.buildroot.org/pipermail/buildroot/2024-October/765463.html
[2] 541ba7d963/configs/st_stm32mp135f_dk_demo_defconfig (L21)
[3] https://lore.kernel.org/buildroot/93f83afb-6987-441c-8e06-dab4d43b828f@grenoble.cnrs.fr/
Cc: Michael Walle <michael@walle.cc>
Cc: Gaël PORTAY <gael.portay+rtone@gmail.com>
Reported-by: Chris Packham <judge.packham@gmail.com>
Signed-off-by: Edgar Bonet <bonet@grenoble.cnrs.fr>
[Romain:
- Rework the initial contribution by Edgar Bonet by
BR2_LINUX_KERNEL_CUSTOM_DTS_DIR following
Michael Walle comments and Gaël PORTAY review.
- Reword the commit log accordingly.
]
Signed-off-by: Romain Naour <romain.naour@smile.fr>
Reviewed-by: Niklas Cassel <niklas.cassel@wdc.com>
Signed-off-by: Julien Olivain <ju.o@free.fr>
The linux kernel can build device tree overlays (.dtbo) itself. Add
support to build and copy them along with the actual device trees.
These can either be in-tree device tree overlays
(BR2_LINUX_KERNEL_INTREE_DTSO_NAMES) or they can be provided outside of
the kernel (BR2_LINUX_KERNEL_CUSTOM_DTS_PATH). In the latter case, the
overlay source files will be copied into the kernel tree first.
Signed-off-by: Michael Walle <michael@walle.cc>
Reviewed-by: Niklas Cassel <niklas.cassel@wdc.com>
Signed-off-by: Romain Naour <romain.naour@smile.fr>
Signed-off-by: Julien Olivain <ju.o@free.fr>
The handling of BR2_LINUX_KERNEL_USE_ARCH_DEFAULT_CONFIG is currently
not doing a proper job: it is selecting ppc64le_defconfig if
BR2_powerpc64le, and using the default of "defconfig" for everything
else.
However:
- Since upstream commit 22f17b02f88b48c01d3ac38d40d2b0b695ab2d10,
which landed in Linux 6.8, the default defconfig is
ppc64le_defconfig and no longer ppc64_defconfig. This means that
despite the condition in linux.mk, we are in fact now always
building ppc64le_defconfig.
- It doesn't handle the 32-bit case, as a 64-bit defconfig gets used
by default. This causes build failures in the autobuilders.
To fix this we explicitly handle BR2_powerpc64le, BR2_powerpc64 and
BR2_powerpc, and use appropriate defconfigs for each case.
Fixes:
http://autobuild.buildroot.net/results/c15eaf2e7455aa265cc045e6d8be7cac5348d925/ (powerpc)
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Julien Olivain <ju.o@free.fr>
Commit abce4a2b3 "linux: add BR2_LINUX_KERNEL_NEEDS_HOST_PYTHON3" [1]
introduced a Kernel dependency on host-python3.
Since the Kernel does not have any specific requirements on host
Python modules, or recent host Python version, this commit replaces
the host-python3 dependency with BR2_PYTHON3_HOST_DEPENDENCY. This
will skip the host-python3 compilation if a sufficient version (3.4 or
greater at the time of this commit) is already present on host. This
will save build time.
This optimization was suggested by Peter, in [2].
Note 1: this commit was checked to ensure that Kernel v6.10.9 arm64
defconfig (which does require a python3 interpreter to build) is
working with Python 3.4.
Note 2: BR2_PYTHON3_HOST_DEPENDENCY was introduced in commit b60729784
"support/dependencies: add a check for python3" [3].
[1] abce4a2b36
[2] https://lists.buildroot.org/pipermail/buildroot/2024-September/763967.html
[3] b60729784a
Signed-off-by: Julien Olivain <ju.o@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Kernel commit [1] introduced MSM GPU driver header generation with a
Python script. This commit was first included in v6.10. This driver is
also enabled as a module in the arm64 architecture default
configuration. See [2]. This is a common situation.
This missing dependency is not detected in the Buildroot CI, because
the reference docker image contains a python3 interpreter. See [3].
For information, the Linux latest kernel version was updated in
Buildroot commit 2b6dba00b "linux: bump latest version to 6.10".
See [4].
The issue can be observed by running the following commands on a host
without the python3 interpreter installed:
cat <<EOF >.config
BR2_aarch64=y
BR2_LINUX_KERNEL=y
BR2_LINUX_KERNEL_USE_ARCH_DEFAULT_CONFIG=y
BR2_LINUX_KERNEL_NEEDS_HOST_OPENSSL=y
BR2_TOOLCHAIN_EXTERNAL=y
EOF
make olddefconfig
make linux
The Kernel build fails with output:
GENHDR drivers/gpu/drm/msm/generated/a2xx.xml.h
/bin/sh: 1: python3: not found
make[7]: *** [drivers/gpu/drm/msm/Makefile:176: drivers/gpu/drm/msm/generated/a2xx.xml.h] Error 127
This commit fixes this issue by introducing a new
BR2_LINUX_KERNEL_NEEDS_HOST_PYTHON3 configuration that will need to be
selected in relevant situations.
[1] https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=0fddd045f88e34d6160785a3a5e506d374566454
[2] https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/arch/arm64/configs/defconfig?h=v6.10.8#n868
[3] https://gitlab.com/buildroot.org/buildroot/-/blob/2024.08-rc3/support/docker/Dockerfile?ref_type=tags#L40
[4] 2b6dba00be
Reported-by: Vincent Stehlé <vincent.stehle@arm.com>
Signed-off-by: Julien Olivain <ju.o@free.fr>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Some kernel users find it useful to store submodules in the kernel
source tree for cross source trees definitions. Add option to download
these submodules.
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Disable -Werror to fix the following build failure with esp-hosted
raised at least since commit a382a7d554:
In function ‘destroy_cmd_wq’,
inlined from ‘esp_commands_teardown’ at /home/autobuild/autobuild/instance-4/output-1/build/esp-hosted-ce3c50a33fa4bc562a1b6cbcee292c1ae0b0a404/esp_hosted_ng/host/esp_cmd.c:1467:2:
./include/linux/workqueue.h:639:9: error: call to ‘__warn_flushing_systemwide_wq’ declared with attribute warning: Please avoid flushing system-wide workqueues. [-Werror=attribute-warning]
639 | __warn_flushing_systemwide_wq(); \
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
/home/autobuild/autobuild/instance-4/output-1/build/esp-hosted-ce3c50a33fa4bc562a1b6cbcee292c1ae0b0a404/esp_hosted_ng/host/esp_cmd.c:408:17: note: in expansion of macro ‘flush_scheduled_work’
408 | flush_scheduled_work();
| ^~~~~~~~~~~~~~~~~~~~
Fixes:
- http://autobuild.buildroot.org/results/6ac7a4601938d3296ed1657c06f8cdf433757d73
- http://autobuild.buildroot.org/results/7997cc8a67645a6e1cf4e24d172c6feae459dcfb
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The BCM2712 of the RaspberryPi 5 supports for 16KB page size.
This adds support for 16 KB on ARM64.
Signed-off-by: Gaël PORTAY <gael.portay@rtone.fr>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Take the basename of the appended DTBs to workaround the following problem:
(cd .../arch/arm/boot; for dtb in cirrus/ep93xx-edb9302
do if test -e ${dtb}.dtb ; then dtbpath=${dtb}.dtb
else dtbpath=dts/${dtb}.dtb ; fi
cat zImage ${dtbpath} > zImage.${dtb} || exit 1; done)
/bin/sh: line 1: zImage.cirrus/ep93xx-edb9302: No such file or directory
Necessary to support ARM Linux starting from commit 724ba6751532
("ARM: dts: Move .dts files to vendor sub-directories"), i.e. Linux v6.4+.
Signed-off-by: Alexander Sverdlin <alexander.sverdlin@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Commit 0b9efc991f ("linux: use BR2_MAKE") switched LINUX_MAKE to
$(BR2_MAKE). However, this also implicitly sets LINUX_KCONFIG_MAKE.
Thus, when host-make is being used in a build that has
PER_PACKAGE_DIRECTORIES enabled, the dotconfig step will try to use the
make instance from the host directory, but since it is not listed in
LINUX_KCONFIG_DEPENDENCIES, it won't be available yet at that point in
time.
Add an explicit dependency to LINUX_KCONFIG_DEPENDENCIES to have it
copied over early enough.
Signed-off-by: Florian Larysch <fl@n621.de>
Signed-off-by: Arnout Vandecappelle <arnout@mind.be>
Commit d06dca38bb (linux: cleanup
kconfig fixup processing) moved the initramfs setup out of an
enclosing statement but kept the closing bracket.
Remove the extra bracket to fix a syntax error in sed processing.
Fixes:
>>> linux 6.4-rc1 Updating kernel config with fixups
/bin/bash: -c: line 1: syntax error near unexpected token `)'
/bin/bash: -c: line 1: `/usr/bin/sed -i -e '/^\(# \)\?CONFIG_INITRAMFS_ROOT_GID\>/d' /home/data/buildroot.experimental/build/linux-6.4-rc1//.config && echo 'CONFIG_INITRAMFS_ROOT_GID=0' >> /home/data/buildroot.experimental/build/linux-6.4-rc1//.config)'
Signed-off-by: Andreas Ziegler <br015@umbiko.net>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Commit d06dca38bb (linux: cleanup kconfig fixup processing) introduced
the conditional LINUX_KCONFIG_FIXUP_CMDS_ROOTFS_CPIO, using a negative
logic condition.
Switch that to use positive logic.
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
The following is an attempt o cleanup the output generated when fixing
up a Linux's Kconfig:
- Quiet the conditional check host-pahole to avoid having the message
used to describe a failure generate in the output in its script form.
- Move the initramfs fixups to their own conditional macros, and move
the comments out of the macro (mainly for developers managing the
implementation).
- Adding a generic "Updating kernel config..." message, to help inform
builders that the framework is actively processing Kconfig's (while
users may be pondering why the closing of a `linux-menuconfig` is
doing).
Signed-off-by: James Knight <james.d.knight@live.com>
[yann.morin.1998@free.fr:
- introduce LINUX_KCONFIG_FIXUP_CMDS_ROOTFS_CPIO
- use MESSAGE to display... a message...
]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Since kernel version 6.2, the minimum GNU Make version is 3.82 [1]. We
have an optional host-make 4.0 minimum dependency, so we can use it as
is. It's a bit unfortunate that we have to apply this even to older
kernel versions, but make itself builds fairly fast compared to the
kernel.
Use BR2_MAKE and BR2_MAKE1 for linux, and depend on
BR2_MAKE_HOST_DEPENDENCY. In addition, we need to set LINUX_MAKE to
BR2_MAKE for use in the kconfig infrastructure.
Signed-off-by: Oleg Lyovin <ovlevin@sberdevices.ru>
Signed-off-by: Arnout Vandecappelle <arnout@mind.be>
A package might install headers that are incompatible with the kernel's
header. One example is the most recent version of pahole (1.24).
HOST_CC includes -I$(HOST_DIR)/include which comes before any include
logic the kernel might have thus forcing the kernel to prefer headers in
HOST_DIR.
The logic to substituting -I with -isystem is taken from
boot/uboot/uboot.mk.
Signed-off-by: Daniel Lang <d.lang@abatec.at>
Reviewed-by: Francis Laniel <flaniel@linux.microsoft.com>
Tested-by: Francis Laniel <flaniel@linux.microsoft.com>
Signed-off-by: Arnout Vandecappelle <arnout@mind.be>
Commit 666084f494 ("linux:linux.mk: Add
"firmware-imx" dependency if needed") introduced a dependency from
linux to firmware-imx, but based on the incorrect BR2 option.
This commit fixes this mistake.
Signed-off-by: Francois Perrad <francois.perrad@gadz.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
This commit is based on earlier work from Łukasz Stelmach
<l.stelmach@samsung.com> to add support for different page sizes on
ARM64.
In his initial submission, Łukasz took an approach similar to this
one, i.e make it ARM64-specific. Following the feedback on the mailing
list, his second version [1] tried to generalize the logic to
configure the page size between architectures. But the general
consensus during the review process was that there wasn't much to
generalize in the end.
So, this new iteration is back to a simpler approach:
* We have new options in Config.in.arm to configure the page
size. Only 4 KB and 64 KB are supported, because our testing in
Qemu and real hardware has not allowed to get a successful setup
for 16 KB pages. We can always re-add support for 16 KB later if
that is resolved.
* The logic to define the ARCH_TOOLCHAIN_WRAPPER_OPTS options is
moved from the ARC-specific file to arch/arch.mk, and extended to
cover ARM64.
* The appropriate logic in uclibc.mk and linux.mk is added to tweak
the relevant configuration options.
* A test case is added in the runtime test infrastructure to test
building and booting under Qemu a 64 KB configuration, with all 3 C
libraries.
For the regular configuration of 4 KB pages, this commit makes one
functional change: on ARM64, -Wl,-z,max-page-size=4096 is now passed in
the compiler flags of the wrapper.
[1] https://patchwork.ozlabs.org/project/buildroot/list/?series=275452
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
gcc-12 is starting to trickle down to some distros, like Archlinux.
gcc-12 has new warnings, and detects more cases of issues, like new
UAF cases, which is causing build issues in code that was previously
building fine, as reported in #14826:
In file included from sigchain.c:3:
In function 'xrealloc',
inlined from 'sigchain_push.isra' at sigchain.c:26:2:
subcmd-util.h:56:23: error: pointer may be used after 'realloc' [-Werror=use-after-free]
56 | ret = realloc(ptr, size);
| ^~~~~~~~~~~~~~~~~~
subcmd-util.h:52:21: note: call to 'realloc' here
52 | void *ret = realloc(ptr, size);
| ^~~~~~~~~~~~~~~~~~
subcmd-util.h:58:31: error: pointer may be used after 'realloc' [-Werror=use-after-free]
58 | ret = realloc(ptr, 1);
| ^~~~~~~~~~~~~~~
subcmd-util.h:52:21: note: call to 'realloc' here
52 | void *ret = realloc(ptr, size);
| ^~~~~~~~~~~~~~~~~~
In that case, the kernel has already fixed their code, which is part of
5.17:
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=52a9dab6d892763b2a8334a568bd4e2c1a6fde66
However, we can't easily carry that patch, because we don't know
whether the kernel the user uses already has the fix or not.
Instead, we can just tell the kernel to disable use of -Werror when
building host tools.
As a consequence, we can drop it from the perf-specific setting.
Fixes: #14826
Reported-by: Anders Pitman <buildroot@apitman.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
The default defconfig target for the 64 bit powerpc kernel is
ppc64_defconfig, the big endian configuration.
When building for powerpc64le users want the little endian kernel as
they can't boot LE userspace on a BE kernel.
Fix up the defconfig used in this case. This will avoid the following
autobuilder failure:
VDSO32A arch/powerpc/kernel/vdso32/sigtramp.o
cc1: error: ‘-m32’ not supported in this configuratioin
make[4]: *** [arch/powerpc/kernel/vdso32/Makefile:49: arch/powerpc/kernel/vdso32/sigtramp.o] Error 1
http://autobuild.buildroot.net/results/dd76d53bab56470c0b83e296872d7bb90f9e8296/
Note that the failure indicates the toolchain is configured to disable
the 32 bit target, causing the kernel to fail when building the 32 bit
VDSO. This is only a problem on the BE kernel as the LE kernel disables
CONFIG_COMPAT, aka 32 bit userspace support, by default.
Signed-off-by: Joel Stanley <joel@jms.id.au>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Since SDMA firmwares for imx[6,7,8] are now provided only by
firmware-imx package and not linux-firmware package [1]. If
CONFIG_EXTRA_FIRMWARE is set in the kernel config, the build will fail
if the imx firmware is not available.
[1] http://lists.busybox.net/pipermail/buildroot/2021-January/603807.html
Signed-off-by: Leger Charlie <c.leger@borea-dental.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Some older versions of linux, or custom versions (like forks for some
boards), fail to build with host-gcc 10+, because of redefined symbols:
HOSTLD scripts/dtc/dtc
/usr/bin/ld: scripts/dtc/dtc-parser.tab.o:(.bss+0x10): multiple definition
of `yylloc'; scripts/dtc/dtc-lexer.lex.o:(.bss+0x0): first defined here
collect2: error: ld returned 1 exit status
Since this has been fixed in recent-ish versions, we can't use an
unconditionaly patch, so we must have a conditional patch. However, a
patch may not always apply to arbitrary Linux versions or forks.
Upstream just dropped that line altogether:
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=e33a814e772cdc36436c8c188d8c42d019fda639
So, we use a little sed-grep combo to do the exact same change.
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Cc: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Reviewed-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
Tested-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
To support building in the wireless regulatory database files (regulatory.db*)
into the kernel using the CONFIG_EXTRA_FIRMWARE option, we need to ensure that
the database files are installed before the Linux kernel is built.
The dependency is harmless if CONFIG_EXTRA_FIRMWARE isn't actually set.
Signed-off-by: Jörg Krause <joerg.krause@embedded.rocks>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Buildroot uses variable LINUX_ARCH_PATH to refer to the arch-specific
directory in the Linux tree, which may not necessarily be arch/$(KERNEL_ARCH).
Buildroot already accounts for the case of KERNEL_ARCH=i386 and
KERNEL_ARCH=x86_64, but does not for KERNEL_ARCH=sparc64, in which case the
correct directory is arch/sparc.
Reported-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas De Schampheleire <thomas.de_schampheleire@nokia.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
CONFIG_DEBUG_BTF_INFO relies on pahole, so kernel DWARF are converted to BTF.
If CONFIG_DEBUG_BTF_INFO is set and BR2_LINUX_KERNEL_NEEDS_HOST_PAHOLE not,
an error message is shown and .config is not written.
Signed-off-by: Francis Laniel <flaniel@linux.microsoft.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
As Thomas put it:
The <pkg>_HELP_CMDS variable allows packages using the
kconfig-package infrastructure to display their specific
targets related to the handling of their configuration.
However, it was not consistently used and handled by the
different packages.
So, this commit switches all the kconfig-based package to use the
generic help helper.
As a consequence:
- all kconfig packages now advetise their kconfig-related actions,
where some were previously missing: at91bootstrap3, linux-backports,
swupdate, xvisor;
- busybox advertises it does not support defconfig files;
- the 'foo-savedfconfig' action is no longer advertised: it is to be
considered an internal implementation detail.
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
To support building in (a subset of) the linux-firmware files into the
kernel using the CONFIG_EXTRA_FIRMWARE option, we need to ensure that the
firmware files are installed before the Linux kernel is built, similar to
how it is done for intel-microcode.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
arm64 / riscv supports building a gzip compressed 'Image' format kernel,
which is sometimes useful. From arch/arm64/Makefile:
all: Image.gz
Image: vmlinux
$(Q)$(MAKE) $(build)=$(boot) $(boot)/$@
Image.%: Image
$(Q)$(MAKE) $(build)=$(boot) $(boot)/$@
(and similar logic for riscv)
Future architectures may or may not copy this logic, so for robustness add
an explicit Image.gz format rather than copying both Image and Image.gz when
the Image format is used.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
As we discussed on the mailing list, using $(<pkg>_NAME) when defining
CPE ID variables feels a bit odd and needlessly complicated. Just use
the package name directly.
Cc: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
If the modules directory that corresponds to the version of the kernel
being built has been deleted, don't try to run depmod, which will
obviously fail.
This can happen for instance when the modules are stripped from the main
root filesystem, and placed into a separate filesystem image, so that
the root filesystem and the kernel can be updated separately.
Signed-off-by: Paul Cercueil <paul@crapouillou.net>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
On Ubuntu 18.04, make-4.1 emits spurious, incorrect "entering/leaving"
messages, which end up in the LINUX_VERSION_PROBED variable:
printf 'probed linux version: "%s"\n' "$(LINUX_VERSION_PROBED)"
probed linux version: "make[1]: Entering directory '/home/buildroot'
4.19.78-linux4sam-6.2
make[1]: Leaving directory '/home/buildroot/output/build/linux-linux4sam_6.2'"
First, the messages are displayed even though we do explicitly pass
--no-print-directory -s.
Second, the entering and leaving messages are not about the same
directory!
This *only* occurs in the following conditions:
- the user has the correct 0022 umask,
- top-level parallel is used (with or without PPD),
- initial -C is specified as well.
$ umask 0022
$ make -j16 -C $(pwd)
[...]
depmod: ERROR: Bad version passed make[1]:
[...]
(yes, 'make[1]:' is the string depmod is trying, and fails, to parse as
a version string).
If any of the three conditions above is removed, the problem no longer
occurs. Here's a table of the MAKEFLAGS:
| 0002 | 0022 |
----+-------+------------------------------------------------+--------------------------+
| no-j | --no-print-directory -- | |
noC | +------------------------------------------------+--------------------------+
| -j16 | -j --jobserver-fds=3,4 --no-print-directory -- | -j --jobserver-fds=3,4 |
----+-------+------------------------------------------------+--------------------------+
| no-j | --no-print-directory -- | w |
-C | +------------------------------------------------+--------------------------+
| -j16 | -j --jobserver-fds=3,4 --no-print-directory -- | w -j --jobserver-fds=3,4 |
----+-------+------------------------------------------------+--------------------------+
0002: umask == 0002
0022: umask == 0022
no-j: no -j flag
-j16: -j16 flag
noC: no -C flag
-C : -C /path/of/buildroot/
Only the bottom-right-most case fails...
This behaviour goes against what is documented:
https://www.gnu.org/software/make/manual/make.html#g_t_002dw-Option
5.7.4 The ‘--print-directory’ Option
[...]
you do not need to specify this option because ‘make’ does it for
you: ‘-w’ is turned on automatically when you use the ‘-C’ option,
and in sub-makes. make will not automatically turn on ‘-w’ if you
also use ‘-s’, which says to be silent, or if you use
‘--no-print-directory’ to explicitly disable it.
So this exactly describes our situation; yet 'w' is added to MAKEFLAGS.
Getting rid of the 'w' flag makes the build succeed again, so that's
what we do here (bleark, icky)...
Furthermore, the documented way to override MAKEFLAGS is to do so as a
make parameter:
https://www.gnu.org/software/make/manual/make.html#Options_002fRecursion
5.7.3 Communicating Options to a Sub-make
[...]
If you do not want to pass the other flags down, you must change the
value of MAKEFLAGS, like this:
subsystem:
cd subdir && $(MAKE) MAKEFLAGS=
However, doing so does not fix the issue. So we resort to pass the
modified MAKEFLAGS via the environment (bleark, icky)...
Fixes: #13141
Reported-by: Laurent <laurent@neko-labs.eu>
Reported-by: Asaf Kahlon <asafka7@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Commit a4eef9a395 (linux: introduce BR2_KERNEL_DTB_KEEP_DIRNAME)
introduced a new config option, but its name was not matching the
naming-scheme of the other config options.
Rename it.
We don't need legacy handling, because that config option was added very
recently and was never part of a release.
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
