(From yocto-docs rev: a587eb476219cb1d6a9b104f4d7470c0d2da3c90)
Signed-off-by: Lee Chee Yang <chee.yang.lee@intel.com>
Reviewed-by: Michael Opdenacker <michael.opdenacker@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
numa is an existing machine feature, add it to the list so that users
are aware of it.
(From yocto-docs rev: 6022d15b8ee5e038d2f1c0f061a871dba40e5213)
Signed-off-by: Jon Mason <jdmason@kudzu.us>
Reviewed-by: Quentin Schulz <foss+yocto@0leil.net>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Pull in some minor fixes:
pseudo_util: Silence symlink errors and fix resolution bug
ports/linux: Remove build dependency on libattr
Minor build fixes
pseudo_util: Fix resolving relative paths from "/"
(From OE-Core rev: c57d0c57d00cdef622dab3bf783a10d52f8d9ffb)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
crossbeam-utils tries to use the triplet to look up whether the target
supports various forms of atomics. We use TARGET_VENDOR and not "-unknown"
in the target case which means this fails and breaks platforms like mips
and powerpc 32 bit. Add a patch to handle TARGET_VENDOR in this case.
(From OE-Core rev: 5d8c01dfed4c82fcc5d504c728a51fb98a262300)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The CVE check database needs to have a shared lock acquired on it before
it is accessed. This to prevent cve-update-db-native from deleting the
database file out from underneath it.
[YOCTO #14899]
(From OE-Core rev: 20a9911b73df62a0d0d1884e57085f13ac5016dd)
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The new classes scope implemented on commit 7bd328f9d
made testimage.bbclass (and perhaps others) stop working
for baremetal-images, the expected way to run testimage
is no longer to use INHERIT but to use IMAGE_CLASSES
instead, however this functionality was not implemented in the
baremetal-image class until now.
Emulate image.bbclass allowing the baremetal-image class to
use IMAGE_CLASSES to fix this issue.
Set defaults for IMAGE_FEATURES to allow bitbake checks to
pass properly.
(From OE-Core rev: bf083929ecd69a144fedeef5a8725775f632a16c)
Signed-off-by: Alejandro Enedino Hernandez Samaniego <alejandro@enedino.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Latest rust has started using ENOTSUP define, which is not available in
the older libc that current release of compiler is using therefore
backport the needed patch. Eventually when vendored version of libc
bumps to 1.33+ we should not need this patch.
(From OE-Core rev: 097e38f28990229c3f54f4e6cb74ca8be0b806e6)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Release notes:
https://blog.rust-lang.org/2022/08/11/Rust-1.63.0.html
This is a standard upgrade aside from the path for the
stage2 tools binaries (clippy, et.al.) changing.
(From OE-Core rev: 9f390accf5fd174c430928cf841728d0456fc1b7)
Signed-off-by: Randy MacLeod <Randy.MacLeod@windriver.com>
Signed-off-by: Alexander Kanavin <alex@linutronix.de>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We see a lot of warnings about incorrect processor types on qemuppc, drowning
out anything else. Fix the option.
(From OE-Core rev: 0371e429d9e127983ddfaec366ce1c38c99158e4)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
meson in mesa no longer uses WANT_LLVM_RELEASE and LLVM_CONFIG so drop
those. Instead, copy the llvm-config binary into the target sysroot which
allows the paths to be relocated to the target in question. This should allow
the llvm patch for YOCTO_ALTERNATE_EXE_PATH to be dropped assuming nothing
else relies upon it.
(From OE-Core rev: 1dfb29a66f0db673c9d6faffed9b5bd855b206e1)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Add a wrapper for llvm-config which provides flags from the current envrionment
instead of the values hardcoded into llvm-native at compile time. Inspiration
taken from the wrapper in meta-clang but I had to totally rewrite it as:
* the TARGET_* prefixes weren't in our environment
* meson uses --libs --ldflags XXX which didn't work
(From OE-Core rev: 193ee1973f613b72f7f99660aecf652b07652d18)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Pass additional arguments in the fileslocked() context manager to the
underlying lockfile() function. This allows the context manager to be
used for any types of locks (non-blocking, shared, etc.) that the
lockfile() function supports.
(Bitbake rev: 7a8eb8da8e8495051e174721062da08e06168024)
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Instead of installing pieces of the build system, we can install the
test driver (which can also be used to list the files needed) and run
the tests directly.
(From OE-Core rev: 5e07e6c376cf46d2788dcef53e9feba890c0236d)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This patch was added in 2017 to fix an ICE in GCC. GCC no longer ICEs,
so the patch isn't needed anymore. Of note is that the random failures
in ptest are in the test being patched, so maybe this is causing subtle
breakage.
(From OE-Core rev: a6cd529ea05e7407a6ef9c6203471bb35e3cc8f9)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
libgcrypt.pc has been built and installed by libgcrypt since 1.9.0[1],
so the manual install can be removed.
[1] 97194b422bc89a6137f4e218d4cdee118c63e96e
(From OE-Core rev: 117e3c7f56246da39971d5eacc3d780eb9d25c25)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Currently, there's no way for the user's site.conf, local.conf or
similar to set BB_DEFAULT_UMASK, because those files are included by
bitbake.conf prior to the unconditional assignment of
BB_DEFAULT_UMASK. To make that possible, use a weak default assignment
instead. This is also consistent with most other variable assignments
in the lower half of bitbake.conf.
I believe the risk of a regression is very small; it would require
something like somebody having a definition of BB_DEFAULT_UMASK in a
local configuration file, and having been relying on that _not_ taking
effect.
(From OE-Core rev: e3dbded499f0bd1e71abb0650ae98fd9ade94250)
Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
AC_TRY_RUN macro means the test needs to run to find the result and we
are cross compiling so this will always get wrong results, this results
in miscompiling apache2 on musl because it disables rlimit
(ac_cv_struct_rlimit) wrongly.
All these variables are determined with AC_TRY_RUN checks
(From OE-Core rev: 504eb0ff1cae200ee85ec18ebae564cae9bf9c8c)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Fixes build errors seen with clang/musl like on x86
error: the 'sse' unit is not supported with this instruction set
(From OE-Core rev: f99bc43c4564b8a7b2047c5c1009a6eb74b393ea)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
With the current regex expression, a machine that is not part of the
compatible could match the regex expression.
For example, consider the following COMPATIBLE_MACHINE:
COMPATIBLE_MACHINE = "qemuarm|qemuarm64"
A machine definition bringing in "qemuarm-foo" would match against the
COMPATIBLE_MACHINE pattern above (see base.bbclass for implementation
details).
Fix this by matching the start and the end of the string.
(From OE-Core rev: 63db2c6baadb4b9aa0c8ae82a497f30840b3b347)
Signed-off-by: Andrei Gherzan <andrei.gherzan@huawei.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Add a new selftest to exercise the debuginfod support, by starting a
debuginfod on DEPLOY_DIR and verifying that an image can fetch the
symbols for a binary.
(From OE-Core rev: d035fd394fd2747ab4b75867af6123f3efb1990f)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This option doesn't seem to exist any more and causes lots of warnings.
Remove it.
(From OE-Core rev: 8e1614a906086fb46c5dd7b7f2dffab91194165c)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Alex Stewart assumed maintainership of the yocto project's opkg fork,
and opkg recipes, from Alejandro Del Castilo back in Q1 of 2020.
Update maintainership of the opkg recipes.
(From OE-Core rev: fd0511080fb5744b4b58df43184fa2561cc37134)
Signed-off-by: Alex Stewart <alex.stewart@ni.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Solves the mount.h conflict with linux kernel headers with needed
backports from trunk
These are the complete list of changes this brings
* 3bd3c612e9 Linux: Fix enum fsconfig_command detection in <sys/mount.h>
* bb1e8b0ca9 linux: Fix sys/mount.h usage with kernel headers
* d48813227b linux: Use compile_c_snippet to check linux/mount.h availability
* 4dad97e2a2 linux: Mimic kernel defition for BLOCK_SIZE
* 1cc5513114 linux: Use compile_c_snippet to check linux/pidfd.h availability
* 0062e7dd1c glibcextract.py: Add compile_c_snippet
* 5c62874f42 NEWS: Add entry for bug 28846
* d13a7a6f10 socket: Check lengths before advancing pointer in CMSG_NXTHDR
* 8b139cd4f1 alpha: Fix generic brk system call emulation in __brk_call (bug 29490)
* e982657073 Linux: Terminate subprocess on late failure in tst-pidfd (bug 29485)
(From OE-Core rev: 344378cc811a9a2630b178ca58b946ecd10b53bc)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The full path of patch may contain '/meta/' but not in oe-core, skip
patches by checking it starts with oe-core full path or not.
(From OE-Core rev: d8a525afdfb5d371e76b09301c8b2741d23d1d10)
Signed-off-by: Yang Xu <yang.xu@mediatek.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
GCC 12.2 is the first bug-fix release from the GCC 12 branch containing
important fixes for regressions and serious bugs in GCC 12.1 with more
than 71 bugs fixed since the previous release comprising of a number
of libstdc++ and C++ and fortran fixes [1]
Remove backported patch to fix libsanitizers with glibc 2.36, its
already present in 12.2
[1] https://gcc.gnu.org/bugzilla/buglist.cgi?bug_status=RESOLVED&resolution=FIXED&target_milestone=12.2
(From OE-Core rev: b73f5c0a7b94d9d04dd69fe5a5b871eab05714a3)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Changes from 1.9.2 to 1.9.3:
----------------------------
* Fix a bunch of minor code issues pointed out using Clang static analyzer.
* New API: pkgconf_solution_free(), which frees a compiled solution graph.
* Fix behavior when overriding global variables with '--define-variable'.
(From OE-Core rev: a48dc992ebf8e664bde4de61ba937a87fc142533)
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Changelog:
==========
New features:
context: Support (NEVRA forms, provides, file provides) including globs in the dnf_context_remove func
Bug fixes:
dnf-context: Disconnect signal handler before dropping file monitor ref
Filter out advisory pkgs with different arch during advisory upgrade, fixes possible problems in dependency resulution
Gracefully handle failure to open repo primary file
Fix listing a repository without cpeid
(From OE-Core rev: 57893d38c2725c07b29d974e184be37b3d265d9b)
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
0001-device-utils.c-Use-linux-mount.h-instead-of-sys-moun.patch
removed since it's included in 5.19
Changelog:
==========
* send: support protocol version 2
* fi show: print all missing devices
* device stats: add tabular output
* replace: add alias to device group (device replace)
* check: validate free space tree items
* fixes:
* convert: support large filesystems (block count > 32bit)
* recognize filesystems with verity enabled
* mkfs and DUP could write out of order, fix it for zoned mode
* build:
* optional support for LZO and ZSTD in receive
* compatibility with glibc 2.36 (mount.h)
* add fallbacks for new GCC builtins
* other:
* corrupt-block: target specific items, offsets
* documentation updates, new pages from wiki
* new tests
(From OE-Core rev: 53ead3fb0e51e71207dfd62f43f0ec4304524dcc)
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This ensures that definitions of functions e.g. getpass() are correctly
sourced from system headers, since it depends on feature test macros
(From OE-Core rev: 0d117e5d17c491e1d26aefb4b919410b07fd5347)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
musl does not implement GNU extention of strerror_r but XSI compliant
version, therefore add it via a packageconfig to set right variables
during configure to cache the value.
configure detection logic depends on runtime test which will always be
wrong on cross compiles therefore backport a patch to make it possible
to cache the needed configure variable.
(From OE-Core rev: ded3d76a844dd1aef9ac610fbe506bf76285369b)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This is defined for glibc automatically with _GNU_SOURCE but not for
musl
(From OE-Core rev: 84375186570af548b49e44e884925f4feaddcc7a)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This patch ensures that a fallback to use XSI compliant strerror_r when
GNU extention is not implemented, which fixes the case when building
with musl
(From OE-Core rev: 095024552cafd2aa926b968a876cecb08e63df18)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We do not get proper function definitions otherwise e.g. fcvt()
(From OE-Core rev: 21dd5bae177b64e314a6423e5ffbd7b28b6b5891)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The libnss configuration file is only installed when glibc is used. The
inexistence of it on a musl-based rootfs, will make shadow complain
about it:
Failed opening /etc/nsswitch.conf
This is because shadow will try to use nsswich when dealing with
subordinate IDs and the message is just a warning as the tool will still
generate them correctly in subuid/subgid files.
We drop this log message for class native to avoid an error when rootfs
logs are checked ('Failed' will match the regex bitbake is using to
check for rootfs generation errors).
(From OE-Core rev: 3000840a271534fa907ce0684b81a6d278e64a44)
Signed-off-by: Andrei Gherzan <andrei.gherzan@huawei.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
'shadow_sort' was renamed to 'tidy_shadowutils_files' in
rootfspostcommands.py so we reflect this in
SORT_PASSWD_POSTPROCESS_COMMAND. This also creates a deprecation
function for 'shadow_sort'.
(From OE-Core rev: b474d7598d0803ba92b01944b961b28beec3bfe2)
Signed-off-by: Andrei Gherzan <andrei.gherzan@huawei.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When creating users, shadow-utils might create backup files for
subordinate ID files (subid, subgid). Make sure we clean them up
similarly to the other backup files shadow-utils creates.
(From OE-Core rev: 4e4ea5adea8a00b4a78ffbe7cc60931deb74c161)
Signed-off-by: Andrei Gherzan <andrei.gherzan@huawei.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This change proposes a restructure of the functions in
rootfspostcommandstests.py to clarify the purpose of each function and
also, make it scalable for other use cases (for example adding support
for removing subid backup files).
The main function of interest here is 'tidy_shadowutils_files' which
brings in the functionality of the old 'sort_passwd' making it clear
that it doesn't only sort the passwd file:
- delete backup files
- it sorts passwd, group and the associated shadow files
The other functions are also renamed for consistency and clarity and
more documentation was added.
(From OE-Core rev: 81a0a4dbfb0313b967a7a98eb7fd1e13edb1a9be)
Signed-off-by: Andrei Gherzan <andrei.gherzan@huawei.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
shadow utils are used when creating users at image creation time. The
useradd/usermod tools will only try to add a default configuration for
subid files if they exist.
(From OE-Core rev: 364a6f408c9feb5b9472ddabbc352d8b432bfffd)
Signed-off-by: Andrei Gherzan <andrei.gherzan@huawei.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Check whether magic.h header exists before defining HAVE_MAGIC.
Despite library availability there still can be missing header.
Current test doesn't cover that possibility which will lead compilation
to fail in case of separate sysroot.
Upstream-Status: Backport
[263381ddd4]
(From OE-Core rev: a93e5c4aba34996fdb9b253a4a8529f33ec114f5)
Signed-off-by: Mateusz Marciniec <mateuszmar2@gmail.com>
Signed-off-by: Tomasz Dziendzielski <tomasz.dziendzielski@gmail.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
