]> git.ipfire.org Git - thirdparty/openembedded/openembedded-core-contrib.git/log
thirdparty/openembedded/openembedded-core-contrib.git
7 weeks agospdx: add option to include only compiled sources
Daniel Turull [Tue, 10 Jun 2025 15:24:42 +0000 (17:24 +0200)] 
spdx: add option to include only compiled sources

When SPDX_INCLUDE_COMPILED_SOURCES is enabled, only include the
source code files that are used during compilation.

It uses debugsource information generated during do_package.

This enables an external tool to use the SPDX information to disregard
vulnerabilities that are not compiled.

As example, when used with the default config with linux-yocto, the spdx size is
reduced from 156MB to 61MB.

Tested with bitbake world on oe-core.

CC: Quentin Schulz <quentin.schulz@cherry.de>
CC: Joshua Watt <JPEWhacker@gmail.com>
CC: Peter Marko <peter.marko@siemens.com>
Signed-off-by: Daniel Turull <daniel.turull@ericsson.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
7 weeks agocross/crosssdk: Allow deferring of these classes too
Richard Purdie [Fri, 6 Jun 2025 16:51:23 +0000 (17:51 +0100)] 
cross/crosssdk: Allow deferring of these classes too

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agobitbake.conf: Automatically defer native and nativesdk inherits
Richard Purdie [Fri, 6 Jun 2025 16:45:27 +0000 (17:45 +0100)] 
bitbake.conf: Automatically defer native and nativesdk inherits

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agotoolchain/gcc/clang: Update PREFERRED_PROVIDER to work with multilibs
Richard Purdie [Fri, 6 Jun 2025 13:31:57 +0000 (14:31 +0100)] 
toolchain/gcc/clang: Update PREFERRED_PROVIDER to work with multilibs

As currently written, the PREFERRED_PROVIDER entries don't work with multilib
since they are added after the manipulations by multilib_global at ConfigParsed
time. We therefore need to spell out the configs correctly for multilib.

To avoid variable overwritten warnings from the expansion, we then have to
expand out the nativesdk providers usng class overrides.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agotoolchain: Provide abstraction for recipe specific toolchain selection
Richard Purdie [Fri, 23 May 2025 03:52:35 +0000 (20:52 -0700)] 
toolchain: Provide abstraction for recipe specific toolchain selection

This change implements a toolchain selection mechanism. Selection is
made using a set of variables, primarily PREFERRED_TOOLCHAIN_TARGET which
defaults to gcc.

It uses the familiar name for toolchain e.g. "gcc" which selects GNU
compiler + binutils as default C/C++ toolchain or "clang" which will
use LLVM/Clang Compiler. Layers an add their own toolchain definitions
too.

There are also PREFERRED_TOOLCHAIN_NATIVE and PREFERRED_TOOLCHAIN_SDK
which will ulitmately allow selection of the toolchain used for the
native/cross and nativesdk/crosssdk compilers. This currently isn't
functional but is essential to the patch to ensure things are set
to the existing gcc support in those cases.

Users would most commonly want to set:

PREFERRED_TOOLCHAIN_TARGET ?= "clang"

in local.conf or other distro specific global configuration metadata.

It is also selectable at recipe scope, since not all packages are
buildable with either clang or gcc, a recipe can explicitly require
a given toolchain using the TOOLCAHIN variable, e.g. glibc can not
be built with clang therefore glibc recipe sets:

TOOLCHAIN = "gcc"

The TOOLCHAIN variable is distinct from the user preference so recipes
with specific requirements can be identified. This also allows different
polcies to be be specified for native/SDK cases in the future.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agobase: Add deferred class event handler
Richard Purdie [Fri, 6 Jun 2025 10:46:46 +0000 (11:46 +0100)] 
base: Add deferred class event handler

Use the new deferred class event to set the class overrides earlier.
This improves interaction of the override with PACKAGECONFIG values
that control conditional inherits (such as python support).

This also allows toolchain configuration in an easier and more user
friendly way.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agosanity.conf: Bump bitbake requirement to 2.15.0
Richard Purdie [Mon, 16 Jun 2025 21:25:35 +0000 (22:25 +0100)] 
sanity.conf: Bump bitbake requirement to 2.15.0

This brings in the features we need to toolchain selection.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agowic: Avoid problems with "-" characters in plugin names
Richard Purdie [Fri, 13 Jun 2025 16:20:52 +0000 (17:20 +0100)] 
wic: Avoid problems with "-" characters in plugin names

Remap "-" characters in plugin names to "_" so wic plugins
can be extended using standard python class inheritance.

This change means wic files can be incrementally updated over time
to the correct name rather than breaking everything. Actual plugin
module files will need to be renamed as done in previous patches.

Also remove a double call to get_plugins() which isn't needed.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agowic: Update after plugin name changes
Richard Purdie [Mon, 26 May 2025 11:56:59 +0000 (12:56 +0100)] 
wic: Update after plugin name changes

Update the plugin names to account for the "-" to "_" plugin name change.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agowic: plugins source bootimage/isoimage rename to allow be imported
Anibal Limon [Sat, 23 Mar 2024 19:53:04 +0000 (13:53 -0600)] 
wic: plugins source bootimage/isoimage rename to allow be imported

Python not support importing modules with - so change to _.

Signed-off-by: Anibal Limon <anibal@limonsoftware.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agowic: pluginbase ensure layer order when load plugins
Anibal Limon [Wed, 11 Jun 2025 15:33:18 +0000 (15:33 +0000)] 
wic: pluginbase ensure layer order when load plugins

To support extensions on wic plugins, the load order needs
to be grauntee matching BBLAYERS variable.

Fix cases when try to import a plugin from another layer,
example of the case to fix,

```
Traceback (most recent call last):
  File "/workspaces/ls/linux/layers/openembedded-core/scripts/wic",line
547, in <module>
    sys.exit(main(sys.argv[1:]))
             ^^^^^^^^^^^^^^^^^^
...
  File "/workspaces/ls/linux/layers/openembedded-core/scripts/lib/wic/
engine.py", line 137, in list_source_plugins
    plugins = PluginMgr.get_plugins('source')
              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/workspaces/ls/linux/layers/openembedded-core/scripts/lib/wic/
pluginbase.py", line 73, in get_plugins
    spec.loader.exec_module(module)
  File "<frozen importlib._bootstrap_external>", line 940, in exec_module
  File "<frozen importlib._bootstrap>", line 241, in _call_with_frames_removed
  File "/workspaces/ls/linux/limonsoftware/bsp/scripts/lib/wic/plugins/
source/bootimg_rpi_autoboot_partition.py", line 1, in <module>
    from wic.plugins.source.bootimg_partition import BootimgPartitionPlugin
ModuleNotFoundError: No module named 'wic.plugins.source.bootimg_partition'
```

Signed-off-by: Anibal Limon <anibal@limonsoftware.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agowayland-protocols: upgrade 1.44 -> 1.45
Wang Mingyu [Mon, 16 Jun 2025 09:16:42 +0000 (17:16 +0800)] 
wayland-protocols: upgrade 1.44 -> 1.45

Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agorepo: upgrade 2.55 -> 2.55.2
Wang Mingyu [Mon, 16 Jun 2025 09:16:41 +0000 (17:16 +0800)] 
repo: upgrade 2.55 -> 2.55.2

Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agopython3-typing-extensions: upgrade 4.13.2 -> 4.14.0
Wang Mingyu [Mon, 16 Jun 2025 09:16:40 +0000 (17:16 +0800)] 
python3-typing-extensions: upgrade 4.13.2 -> 4.14.0

Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agopython3-ruamel-yaml: upgrade 0.18.12 -> 0.18.14
Wang Mingyu [Mon, 16 Jun 2025 09:16:39 +0000 (17:16 +0800)] 
python3-ruamel-yaml: upgrade 0.18.12 -> 0.18.14

Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agopython3-pdm: upgrade 2.24.2 -> 2.25.1
Wang Mingyu [Mon, 16 Jun 2025 09:16:36 +0000 (17:16 +0800)] 
python3-pdm: upgrade 2.24.2 -> 2.25.1

Changelog:
=============
- Fix duplicated dependencies added to the lock file when the same dependency with extras is requested.
- Stabilize order of the extras and dependency-groups fields in pylock output.
- Fix Windows 11 install pdm error, which is because of msgpack install failure.
- Change the return type of array_of_inline_tables to list[dict] from list[str]
- Ensure uv resolver to include hash for package files.
- Avoid infinite recursion when reading pyproject.toml with circular file dependencies.
- Support pylock as alternative lock format and make it opt-in by config.
- Search for package metadata in lock file first when reuse strategy is used.

Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agopython3-numpy: upgrade 2.2.6 -> 2.3.0
Wang Mingyu [Mon, 16 Jun 2025 09:16:35 +0000 (17:16 +0800)] 
python3-numpy: upgrade 2.2.6 -> 2.3.0

License-Update: Copyright year updated to 2025

Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agopython3-hypothesis: upgrade 6.132.0 -> 6.135.9
Wang Mingyu [Mon, 16 Jun 2025 09:16:34 +0000 (17:16 +0800)] 
python3-hypothesis: upgrade 6.132.0 -> 6.135.9

Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agopython3-dtschema: upgrade 2025.2 -> 2025.6.1
Wang Mingyu [Mon, 16 Jun 2025 09:16:33 +0000 (17:16 +0800)] 
python3-dtschema: upgrade 2025.2 -> 2025.6.1

Changelog:
===========
- Fix a longstanding ordering issue with extracting type information
  from properties which have a reference to another property. 'mac-mode'
  is the one in the Linux kernel.

- Fix a false positive warning about missing
  unevaluatedProperties/additionalProperties

Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agopython3-cython: upgrade 3.1.1 -> 3.1.2
Wang Mingyu [Mon, 16 Jun 2025 09:16:32 +0000 (17:16 +0800)] 
python3-cython: upgrade 3.1.1 -> 3.1.2

Bugs fixed
----------
* Attribute lookups failed on the "bool" builtin type.
* Type checks on or-ed union types could incorrectly return false.
* Negative list indexing could accidentally wrap around twice in PyPy and the Limited API.
* Iterating over literal sequences with starred (unpacked) items could infer a wrong
  type for the loop variable and fail to assign the values.
* Calls to C functions taking exception types failed to check for a 'None' argument.
* Fused functions had an incorrect "__module__" attribute.
* The type of Cython implemented functions had an incorrect "__module__" attribute.
* Errors while indexing into "bytearray" or "str" in "nogil" sections could crash.
* "bytearray.append()" could silently accept some invalid character numbers.
* The C++11 "<type_traits>" header was included regardless of the C++ version.
* "PyDict_GetItemStringRef()" was accidentally used in older Limited API versions.
* "abort()" was used but not always available in the Limited API.
* Some dependencies were missing from the "depfile".
* Embedded function signatures were not always separated from the existing docstring.
* "numpy.math" was missing from "Cython/Includes/" and could not be cimported.
* Some tests were adapted for NumPy 2.x.
* Some C compiler warnings were fixed.
* "Cython.Build" was not officially exposing the "cythonize" function.

Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agopixman: upgrade 0.46.0 -> 0.46.2
Wang Mingyu [Mon, 16 Jun 2025 09:16:31 +0000 (17:16 +0800)] 
pixman: upgrade 0.46.0 -> 0.46.2

Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agompg123: upgrade 1.32.10 -> 1.33.0
Wang Mingyu [Mon, 16 Jun 2025 09:16:30 +0000 (17:16 +0800)] 
mpg123: upgrade 1.32.10 -> 1.33.0

Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agomeson: upgrade 1.8.1 -> 1.8.2
Wang Mingyu [Mon, 16 Jun 2025 09:16:29 +0000 (17:16 +0800)] 
meson: upgrade 1.8.1 -> 1.8.2

Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agolttng-modules: upgrade 2.13.18 -> 2.13.19
Wang Mingyu [Mon, 16 Jun 2025 09:16:28 +0000 (17:16 +0800)] 
lttng-modules: upgrade 2.13.18 -> 2.13.19

Changelog:
===========
* Fix missing override when CONFIG_COMPAT_OLD_SIGACTION is not defined
* fix: writeback: 'balance_dirty_pages' Respect 'CONFIG_CGROUP_WRITEBACK'
* Fix: scsi: RESERVE and RELEASE renamed in Linux v6.15-rc1
* Fix: del_timer[_sync] deleted in linux v6.15-rc1
* Fix: Use 'nonseekable_open' for proc files
* Fix: trace_balance_dirty_pages in Linux v6.14.2
* fix: version constraint for building lttng-probe-9p

Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agolibsolv: upgrade 0.7.32 -> 0.7.33
Wang Mingyu [Mon, 16 Jun 2025 09:16:27 +0000 (17:16 +0800)] 
libsolv: upgrade 0.7.32 -> 0.7.33

Changelog:
===========
- selected bug fixes:
  * improve transaction ordering by allowing more uninst->uninst
    edges
  * implement color filtering when adding update targets
- new features:
  * support orderwithrequires dependencies in susedata.xml

Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agolibsdl2: upgrade 2.32.6 -> 2.32.8
Wang Mingyu [Mon, 16 Jun 2025 09:16:26 +0000 (17:16 +0800)] 
libsdl2: upgrade 2.32.6 -> 2.32.8

Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agolibjpeg-turbo: upgrade 3.1.0 -> 3.1.1
Wang Mingyu [Mon, 16 Jun 2025 09:16:25 +0000 (17:16 +0800)] 
libjpeg-turbo: upgrade 3.1.0 -> 3.1.1

Changelog:
- Hardened the libjpeg API against hypothetical calling applications that may
  erroneously change the value of the 'data_precision' field in
  'jpeg_compress_struct' or 'jpeg_decompress_struct' after calling
  'jpeg_start_compress()' or 'jpeg_start_decompress()'.

Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agolibffi: upgrade 3.4.8 -> 3.5.1
Wang Mingyu [Mon, 16 Jun 2025 09:16:24 +0000 (17:16 +0800)] 
libffi: upgrade 3.4.8 -> 3.5.1

License-Update: Copyright year updated to 2025

not-win32.patch
refreshed for 3.5.1

Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agolibdrm: upgrade 2.4.124 -> 2.4.125
Wang Mingyu [Mon, 16 Jun 2025 09:16:23 +0000 (17:16 +0800)] 
libdrm: upgrade 2.4.124 -> 2.4.125

Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agohwdata: upgrade 0.395 -> 0.396
Wang Mingyu [Mon, 16 Jun 2025 09:16:22 +0000 (17:16 +0800)] 
hwdata: upgrade 0.395 -> 0.396

Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agogtk4: upgrade 4.18.5 -> 4.18.6
Wang Mingyu [Mon, 16 Jun 2025 09:16:21 +0000 (17:16 +0800)] 
gtk4: upgrade 4.18.5 -> 4.18.6

Bugs fixed:
- Wrong behavior of GdkKeymap on macOS
- cups: NULL-terminate array of choices
- Fix the android build
- icontheme: Load the missing image icon from the theme

Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agodiffoscope: upgrade 297 -> 298
Wang Mingyu [Mon, 16 Jun 2025 09:16:20 +0000 (17:16 +0800)] 
diffoscope: upgrade 297 -> 298

Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agodebianutils: upgrade 5.22 -> 5.23.1
Wang Mingyu [Mon, 16 Jun 2025 09:16:19 +0000 (17:16 +0800)] 
debianutils: upgrade 5.22 -> 5.23.1

Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agolibxml2: fix CVE-2025-6021
hongxu [Mon, 16 Jun 2025 05:00:53 +0000 (13:00 +0800)] 
libxml2: fix CVE-2025-6021

According to [1]

A flaw was found in libxml2's xmlBuildQName function, where integer overflows
in buffer size calculations can lead to a stack-based buffer overflow. This
issue can result in memory corruption or a denial of service when processing
crafted input.

Refer debian [2], backport a fix [3] from upstream

[1] https://nvd.nist.gov/vuln/detail/CVE-2025-6021
[2] https://security-tracker.debian.org/tracker/CVE-2025-6021
[3] https://gitlab.gnome.org/GNOME/libxml2/-/commit/acbbeef9f5dcdcc901c5f3fa14d583ef8cfd22f0

Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agooe/license_finder: support extra hashes being passed to find_licenses
Ross Burton [Fri, 13 Jun 2025 13:16:19 +0000 (14:16 +0100)] 
oe/license_finder: support extra hashes being passed to find_licenses

When using the license finder the caller might know some more license
hashes, for example if it is updating existing metadata.

Allow the caller to pass more hashes that can be used when identifying
licenses.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agooe/license_finder: rewrite license checksum loading, scan more licenses
Ross Burton [Fri, 13 Jun 2025 13:16:18 +0000 (14:16 +0100)] 
oe/license_finder: rewrite license checksum loading, scan more licenses

Rewrite the license checksum generation and loading of CSV files to be
clearer.

This also expands the scan of COMMON_LICENSE_DIR to include LICENSE_PATH,
which can be extended by layers to provide more license texts.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agooe/license_finder: don't return the "crunched" license text in crunch_license
Ross Burton [Fri, 13 Jun 2025 13:16:17 +0000 (14:16 +0100)] 
oe/license_finder: don't return the "crunched" license text in crunch_license

crunch_license() will perform some basic text manipulation to try and
canonicalise the license texts. It also returns the new license text but
none of the callers use this, and as a slightly mangled version of the
original it has no real purpose.

Remove this return value and clean up the callers.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agooe/license_finder: remove unused arguments in get_license_md5sums
Ross Burton [Fri, 13 Jun 2025 13:16:16 +0000 (14:16 +0100)] 
oe/license_finder: remove unused arguments in get_license_md5sums

get_license_md5sums() has two optional arguments:

- static_only: if set, don't checksum the licenses in COMMON_LICENSE_DIR
- linenumbers: if set, the CSV file can contain begin/end/md5 values as
  used in LIC_FILES_CHKSUM.

Neither of these are used and complicate the logic, so remove them.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agooe/license_finder: consolidate hash->license maps
Ross Burton [Fri, 13 Jun 2025 13:16:15 +0000 (14:16 +0100)] 
oe/license_finder: consolidate hash->license maps

There are two locations where mappings of checksums to license names
are: the license-hashes.csv file and a hard-coded set of assignments in
the code.

There's no need for two, so remove the assignments and move the hashes
into the CSV file.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agooe/license_finder: add first_only argument to find_licenses()
Ross Burton [Fri, 13 Jun 2025 13:16:14 +0000 (14:16 +0100)] 
oe/license_finder: add first_only argument to find_licenses()

It may be desired to find only the "top-level" license file instead of
every potential candidate, so add a first_only argument (defaulting to
False to preserve existing behaviour) to return just the first license
found.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agooe/license_finder: skip .sh files when looking for licenses
Ross Burton [Fri, 13 Jun 2025 13:16:13 +0000 (14:16 +0100)] 
oe/license_finder: skip .sh files when looking for licenses

Shell scripts are not licenses, so skip them.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agorecipetool: use oe.license_finder
Ross Burton [Fri, 13 Jun 2025 13:16:12 +0000 (14:16 +0100)] 
recipetool: use oe.license_finder

Delete the now redundant code, and import oe.license_finder instead.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agolib/oe/license_finder: extract license finding code from recipetool
Ross Burton [Fri, 13 Jun 2025 13:16:11 +0000 (14:16 +0100)] 
lib/oe/license_finder: extract license finding code from recipetool

This code is 99% identical to the original code in recipetool/create.py,
but with two minor changes:
- The implicit recipetool logger is changed to an explicit logger
- The CSV of license hashes is moved to meta/files/

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agodefault-distrovars: set an empty default for LICENSE_PATH
Ross Burton [Fri, 13 Jun 2025 13:16:10 +0000 (14:16 +0100)] 
default-distrovars: set an empty default for LICENSE_PATH

This variable is a list of paths that contain extra license texts. It
doesn't have a default so can be unset.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agooe-selftest: fitimage: replace cleansstate with compile -f
Adrian Freihofer [Sat, 14 Jun 2025 15:58:12 +0000 (17:58 +0200)] 
oe-selftest: fitimage: replace cleansstate with compile -f

Avoid using "cleansstate" in tests, as it can remove files from
SSTATE_DIR and disrupt parallel builds on autobuilders. Use
"bitbake kernel-signing-keys-native -c compile -f" to force key
regeneration without affecting shared state.

This issue was introduced in:
oe-selftest: fitimage: cleanup FIT_GENERATE_KEYS
OE-Core rev: 97e58d7c2bc1943f0696fc72984788f459f7f7c4

Signed-off-by: Adrian Freihofer <adrian.freihofer@siemens.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agodevtool: ide-sdk code remove twxs.cmake
Adrian Freihofer [Sat, 14 Jun 2025 15:41:34 +0000 (17:41 +0200)] 
devtool: ide-sdk code remove twxs.cmake

Do no longer recommend the twxs.cmake VSCode plugin. There is now a
language server built into the ms-vscode.cmake-tools plugin as well.

>From Release notes 1.20.53
https://marketplace.visualstudio.com/items/ms-vscode.cmake-tools/changelog
Add notification suggesting users to uninstall twxs.cmake now that we
have built-in Language Services. Follow this advice.

Signed-off-by: Adrian Freihofer <adrian.freihofer@siemens.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agosystemd: upgrade 257.5 -> 257.6
Peter Marko [Fri, 13 Jun 2025 21:46:54 +0000 (23:46 +0200)] 
systemd: upgrade 257.5 -> 257.6

Handles CVE-2025-4598

Rebase patches

Signed-off-by: Peter Marko <peter.marko@siemens.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agoscripts/install-buildtools: Update to 5.2.1
Aleksandar Nikolic [Fri, 13 Jun 2025 20:59:33 +0000 (22:59 +0200)] 
scripts/install-buildtools: Update to 5.2.1

Update to the 5.2.1 release of the 5.2.1 series for buildtools

Signed-off-by: Aleksandar Nikolic <aleksandar.nikolic@zeiss.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agooe-selftest: fitimage: fix new parser compatibility
Adrian Freihofer [Sun, 15 Jun 2025 19:16:01 +0000 (21:16 +0200)] 
oe-selftest: fitimage: fix new parser compatibility

Recent parser changes throw a warning if there is no space around the
= operator.

Signed-off-by: Adrian Freihofer <adrian.freihofer@siemens.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agopackagegroup-base: fix override syntax
Adrian Freihofer [Sun, 15 Jun 2025 19:15:35 +0000 (21:15 +0200)] 
packagegroup-base: fix override syntax

Looks like one more left over from the override syntax change.

Signed-off-by: Adrian Freihofer <adrian.freihofer@siemens.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agolib/oe: Move vardepexclude entries alongside functions
Richard Purdie [Thu, 12 Jun 2025 15:46:37 +0000 (16:46 +0100)] 
lib/oe: Move vardepexclude entries alongside functions

Now we have decorators that can do this, move the variable dependencies
exclusions alongside the code that needs them for maintainability.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agoweston: upgrade 14.0.1 -> 14.0.2
Gyorgy Sarvari [Thu, 12 Jun 2025 14:20:34 +0000 (16:20 +0200)] 
weston: upgrade 14.0.1 -> 14.0.2

While at it, also add libpam as a dependency to the vnc pkgconf, to
avoid the following error:

| Run-time dependency pam found: NO (tried pkgconfig and cmake)
|
| ../weston-14.0.2/libweston/meson.build:95:15: ERROR: C shared or static library 'pam' not found

Changelog:
 - frontend: Fix crash in output resize handler
 - libweston: fix bitshift in weston_idalloc_get_id()
 - libweston: fix realloc in weston_idalloc
 - libweston: refactor update_lowest_free_bucket()
 - libweston-desktop: don't try to move child surfaces to not existing layer
 - libweston/input.c: Fix weston crash with the mouse event
 - windowed-output-api.h: Provide ARRAY_LENGH() for windowed-output-api
 - drm: Fix underlay test
 - compositor: Mark pnode accordingly when buffer type is direct
 - compositor: re-order paint node placeholder checks
 - gl-renderer: Take direct-display into consideration
 - shared: fix binding-modifier none
 - vnc: Allow neatvnc in version 0.9.0
 - compositor: Prevent startup crash when hdcp mode is set on display
 - libweston/desktop: Avoid a potential crash on invalid resource
 - libweston/desktop: Don't destroy the xdg_surface

Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agoinsane: Fix debug-deps check
Ryan Eatmon [Thu, 12 Jun 2025 18:28:38 +0000 (13:28 -0500)] 
insane: Fix debug-deps check

Fixes bug: https://bugzilla.yoctoproject.org/show_bug.cgi?id=15901

The cheeck for debug-deps is looking for the mere existence of the
substring "-dbg" inside of the name of an RDEPENDS package, but it should be
an endswith check.  This helps with some eroneous errors in kernel
module names like:

ERROR: linux-xxx do_package_qa: QA Issue: kernel-modules rdepends on kernel-module-g-dbgp-6.12.22-ti [debug-deps]

and

ERROR: QA Issue: kernel-module-mtk-vcodec-dec-6.16.0-rc1-next-20250610-dirty rdepends on kernel-module-mtk-vcodec-dbgfs-6.16.0-rc1-next-20250610-dirty [debug-deps]

Signed-off-by: Ryan Eatmon <reatmon@ti.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agonet-tools: patch CVE-2025-46836
Peter Marko [Thu, 12 Jun 2025 21:14:03 +0000 (23:14 +0200)] 
net-tools: patch CVE-2025-46836

Backport patch for this CVE and also patch for its regression.

Signed-off-by: Peter Marko <peter.marko@siemens.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agolibtheora: disable all 32-bit arm assembler
Ross Burton [Thu, 12 Jun 2025 20:30:01 +0000 (21:30 +0100)] 
libtheora: disable all 32-bit arm assembler

The 32-bit Arm assembler is mostly broken. In 1.1.0 it was never used,
and 1.2.0 tries to enable it and there are a number of different ways it
can fail (some gcc/architecture combinations, and all clang builds).

Until this is fixed upstream, simply disable assembler entirely.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agopython3-requests: upgrade 2.32.3 -> 2.32.4
Jiaying Song [Fri, 13 Jun 2025 02:39:40 +0000 (10:39 +0800)] 
python3-requests: upgrade 2.32.3 -> 2.32.4

Changelog:
https://requests.readthedocs.io/en/latest/community/updates/#release-history

Signed-off-by: Jiaying Song <jiaying.song.cn@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agomesa: do not build nouveau NVIDIA gallium-llvm driver if not requested
Quentin Schulz [Thu, 12 Jun 2025 12:14:32 +0000 (14:14 +0200)] 
mesa: do not build nouveau NVIDIA gallium-llvm driver if not requested

The Nouveau driver is used for NVIDIA GPUs and thus doesn't really need
to be always enabled when building gallium-llvm drivers.

So let's guard it with a nouveau PACKAGECONFIG.

The only intended change is nouveau not being build on target mesa on
non-x86/i686 machines when building gallium-llvm drivers as I assume
NVIDIA GPUs aren't that common on other CPU architectures (tegra already
bringing in the nouveau driver should handle the NVIDIA SoCs that would
benefit from nouveau driver and thus are not impacted by this change).

Signed-off-by: Quentin Schulz <quentin.schulz@cherry.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agomesa: do not build r300 AMD gallium-llvm driver if not requested
Quentin Schulz [Thu, 12 Jun 2025 12:14:31 +0000 (14:14 +0200)] 
mesa: do not build r300 AMD gallium-llvm driver if not requested

It doesn't make much sense to always build the r300 AMD gallium LLVM
driver as that's HW-specific.

Instead, let's guard it with the amd PACKAGECONFIG.

This will only disable R300 for target mesa on non-x86/i686 machines
where gallium-llvm PACKAGECONFIG is selected but not amd, otherwise
behavior is left unchanged.

Signed-off-by: Quentin Schulz <quentin.schulz@cherry.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 weeks agolibcheck: use cmake instead of autotools
Markus Volk [Tue, 10 Jun 2025 05:58:10 +0000 (07:58 +0200)] 
libcheck: use cmake instead of autotools

- this fixes a build issue seen with current master-next branch:
| check.texi:1610: warning: node prev pointer for `Testing Signal Handling and Exit Values' is `Selective Running of Tests' but prev is `Selecting Tests Based on Arbitrary Tags' in menu
| check.texi:2051: warning: node prev pointer for `TAP Logging' is `Test Logging' but prev is `XML Logging' in menu
| make[2]: Leaving directory '/home/flk/poky/build/tmp/work/corei7-64-poky-linux/libcheck/0.15.2/build/doc'
| make[2]: *** [Makefile:452: ../../check-0.15.2/doc/check.info] Error 1
| make[1]: *** [Makefile:574: all-recursive] Error 1
| make[1]: Leaving directory '/home/flk/poky/build/tmp/work/corei7-64-poky-linux/libcheck/0.15.2/build'
| make: *** [Makefile:420: all] Error 2
| ERROR: oe_runmake failed
| WARNING: /home/flk/poky/build/

- checkmk adds a reproducibility issue. @AWK_PATH@ is  unique, because awk is
  in hosttools. We dont want it that way for target. Use a '-D' assignment to
  fix this (Thanks to Ross.Burton@arm.com).

Signed-off-by: Markus Volk <f_l_k@t-online.de>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
7 weeks agoovmf: fix CVE-2024-38797
Hongxu Jia [Fri, 13 Jun 2025 03:43:17 +0000 (20:43 -0700)] 
ovmf: fix CVE-2024-38797

According to [1]:

EDK2 contains a vulnerability in the HashPeImageByType(). A user may cause a read out of
bounds when a corrupted data pointer and length are sent via an adjecent network.
A successful exploit of this vulnerability may lead to a loss of Integrity and/or
Availability.

Backport fixes from upstream edk2 [2][3]

[1] https://nvd.nist.gov/vuln/detail/CVE-2024-38797
[2] https://github.com/tianocore/edk2/security/advisories/GHSA-4wjw-6xmf-44xf
[3] https://github.com/tianocore/edk2/pull/10928

Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com>
7 weeks agotoaster.bbclass: fix toaster error caused by tabs in BBLAYERS
Osose Itua [Wed, 11 Jun 2025 15:24:22 +0000 (11:24 -0400)] 
toaster.bbclass: fix toaster error caused by tabs in BBLAYERS

Users may unknowingly put tabs in BBLAYERS instead of spaces, and this is
interpreted as a literal "\t" at the start of the filepath which causes
_get_layer_dict() function to fail at finding the filepath.

Instead of using split(" "), which restricts it to split on just spaces
replace with split() as this handles spaces, tabs and newlines.

Min steps to reproduce:
- Clone the poky repo:
    git clone git://git.yoctoproject.org/poky
    cd poky
    source oe-init-build-env
- Insert tabs in the BBLAYERS variable in bblayers.conf
  - Note: tab needs to be in the recipe that is being built for the error to be observed
  - Ex: ` /home/<user>/src/poky/meta-skeleton \`
- Start toaster
  source toaster start
  bitbake hello

Error message:
FileNotFoundError: [Errno 2] No such file or directory: '\t/home/<user>/src/poky/meta-skeleton’

Fix by using split() instead of split(" ").

Suggested-by: Anakin Childerhose <anakin.childerhose@savoirfairelinux.com>
CC: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Osose Itua <osose.itua@savoirfairelinux.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agosysfsutils: fix my_strncat function
Hongxu Jia [Wed, 4 Jun 2025 08:17:44 +0000 (16:17 +0800)] 
sysfsutils: fix my_strncat function

The bug was introduced by upstream commit [1] where strncat was replaced with
internal my_strncat function, such as:

  char dest[32] = "/sys/devices/platform/axi";
  my_strncat(dest, "/", sizeof(dest) - strlen(dest) - 1);

Will result in dest string being:

  /sys/

and not the expected:

  /sys/devices/platform/axi/

The meaning of the "len" parameter in the my_strncat function is the size limit for
copying characters from "from", not the size limit for "to" after copying. Also,
the "#define safestrcat(to, from) my_strncat(to, from, sizeof(to) - strlen(to) - 1)"
has already imposed a limit on max based on the size of "to". Modify the function
to prevent truncation of content when too many bytes are passed to the my_strcat function.

[1] https://github.com/linux-ras/sysfsutils/commit/0719881cad85f837f039ecb378b823306640902a

Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
8 weeks agorust: Upgrade 1.85.1->1.86.0
Yash Shinde [Tue, 10 Jun 2025 06:19:54 +0000 (23:19 -0700)] 
rust: Upgrade 1.85.1->1.86.0

Rust stable version updated to 1.86.0.
https://blog.rust-lang.org/2025/04/03/Rust-1.86.0.html

* Add pkgconfig-native and openssl to resolve openssl-sys crate
dependency on pkg-config. As per rust document this is a required dependency.

Fixes:
| error: failed to run custom build command for `openssl-sys v0.9.106`
| Could not find openssl via pkg-config:
|   The pkg-config command could not be found.
|
|   Most likely, you need to install a pkg-config package for your OS.
|   Try `apt install pkg-config`, or `yum install pkg-config`,
|   or `pkg install pkg-config`, or `apk add pkgconfig` depending on your distribution

https://crates.io/crates/openssl-sys/0.9.108/dependencies
https://github.com/rust-lang/rust/blob/master/INSTALL.md#dependencies

* Add Ninja as a dependency for building Rust to prevent bootstrap
  build regression.

Fixes:
| Building LLD for x86_64-unknown-linux-gnu
|
| Couldn't find required command: ninja (or ninja-build)
|
| You should install ninja as described at
| <https://github.com/ninja-build/ninja/wiki/Pre-built-Ninja-packages>,
| or set `ninja = false` in the `[llvm]` section of `config.toml`.
| Alternatively, set `download-ci-llvm = true` in that `[llvm]` section
| to download LLVM rather than building it.

* Add bash to DEPENDS to resolve missing dependency for subtree-sync.sh
Fixes:
ERROR: rust-1.86.0-r0 do_package_qa: QA Issue: /usr/lib/rustlib/src/rust/library/portable-simd/subtree-sync.sh
contained in package rust requires /bin/bash, but no providers found in RDEPENDS:rust? [file-rdeps]

* Add do_install:append() task to remove cargo bin from rust native builds.
This resolves the following conflict:

Fixes:
ERROR: libstd-rs-1.86.0-r0 do_prepare_recipe_sysroot: The file /usr/bin/cargo is
installed by both rust-native and cargo-native, aborting

* Update Unicode-3.0 license checksums.
License-Update: Copyright and license files to distributions are updated.

https://github.com/rust-lang/rust/commit/f9c16997dc016a3ef1456f56df2ab564a1c48cb2
  It adds copyright and license files (including HTML versions) to distributions,
  aligns with license compliance tools like reuse, and ensures all required
  license texts are properly included and formatted.

* Disable building of extended Rust tools to reduce build time and filesystem usage.
  Update config.toml to disable building of extended Rust tools that are not required.
  This helps minimize unnecessary build time and filesystem usage.

* The "remote-test-server" bin is now generated in stage2-tools-bin dir
  rather than stage1. Update the test suite accordingly.

* Fix do_package QA issue by packing missing zsh files and directories:

Fixes:
do_package: QA Issue: rust: Files/directories were installed but not shipped in any package:
/usr/share/zsh
/usr/share/zsh/site-functions
/usr/share/zsh/site-functions/_cargo
Please set FILES such that these items are packaged. Alternatively if they are unneeded,
avoid installing them or delete them within do_install

* From v1.86.0, a "self-contained" LLD is built as part of rust
 bootstrap build. This results in additional build time and
 installations. Disable rust-lld in config.toml to prevent it.

 References: https://github.com/rust-lang/rust/pull/135001
             https://github.com/rust-lang/rust/commit/8744b44e6bde053f7656a3c727c968177ee8e4b6

* Drop Zdual-proc-macros-additional-check.patch patch
 since it's merged with v1.86.0
 https://github.com/rust-lang/rust/commit/139d6ba054a1a4cc5fe64981ad46fd5547bd4916

* LTO config is applied to rustdoc from v1.86.0.
 Rebase 0001-src-core-build_steps-tool.rs-switch-off-lto-for-rust.patch
 which disables it to avoid suffixes in binaries causing non-reproducibility.
 https://github.com/rust-lang/rust/commit/1fe351b

* Restrict tests using "//@only <target_arch>" to avoid failures on riscv64,
  which is now part of default AB testing. Since riscv64 is Tier 2
  with no automated testing, some tests may fail. This approach ensures tests continue
  running on supported architectures while skipping them on riscv64.
  https://doc.rust-lang.org/rustc/platform-support.html#tier-2-with-host-tools

Signed-off-by: Yash Shinde <Yash.Shinde@windriver.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
8 weeks agopython3-ndg-httpsclient: remove unused recipe
Ross Burton [Wed, 11 Jun 2025 12:37:37 +0000 (13:37 +0100)] 
python3-ndg-httpsclient: remove unused recipe

The last dependency in core on this recipe was removed in May 2024[1],
and there don't appear to be any other users that I can find.  The last
upstream release was in 2018 so this is now obsolete.

[1] oe-core dfa482f1998 ("python3-requests: cleanup RDEPENDS")

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agogo: upgrade 1.24.3 -> 1.24.4
Peter Marko [Tue, 10 Jun 2025 22:07:38 +0000 (00:07 +0200)] 
go: upgrade 1.24.3 -> 1.24.4

Upgrade to latest 1.24.x release [1]:

$ git --no-pager log --oneline go1.24.3..go1.24.4
6796ebb2cb [release-branch.go1.24] go1.24.4
85897ca220 [release-branch.go1.24] net/http: strip sensitive proxy headers from redirect requests
9f9cf28f8f [release-branch.go1.24] os: don't follow symlinks on Windows when O_CREATE|O_EXCL
a31c931adf [release-branch.go1.24] cmd/link: allow linkname reference to a TEXT symbol regardless of size
03811ab1b3 [release-branch.go1.24] crypto/x509: decouple key usage and policy validation
04a9473847 [release-branch.go1.24] lib/fips140: set inprocess.txt to v1.0.0
db8f1dc948 [release-branch.go1.24] hash/maphash: hash channels in purego version of maphash.Comparable
664cf832ec [release-branch.go1.24] runtime/debug: document DefaultGODEBUG as a BuildSetting
431f75a0b9 [release-branch.go1.24] os: fix Root.Mkdir permission bits on OpenBSD

Fixes CVE-2025-4673, CVE-2025-0913 and CVE-2025-22874 [2].

[1] https://github.com/golang/go/compare/go1.24.3...go1.24.4
[2] https://groups.google.com/g/golang-announce/c/ufZ8WpEsA3A

Signed-off-by: Peter Marko <peter.marko@siemens.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agoseatd: fix packaging error with systemd DISTRO_FEATURES
Gyorgy Sarvari [Tue, 10 Jun 2025 09:49:37 +0000 (11:49 +0200)] 
seatd: fix packaging error with systemd DISTRO_FEATURES

The systemd unit file for this recipe is only installed in case
the target system's init system is systemd. It is achieved by
inspecting the VIRTUAL-RUNTIME_init_manager variable.

However the systemd class was inherited unconditionally. This caused
a failure in do_package task, in case systemd is present in the
DISTRO_FEATURES but the system's init manager is sysvinit: in this case
the systemd unit file is not installed, however systemd.bbclass is
still trying to register is as a startup service. At this point
it failed:

ERROR: seatd-0.9.1-r0 do_package: Didn't find service unit 'seatd.service', specified in SYSTEMD_SERVICE:seatd.

To avoid this, install the systemd unit file unconditionally, regardless
of the used init system.

Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agopackage: export debugsources in PKGDESTWORK as json
Daniel Turull [Tue, 10 Jun 2025 15:24:41 +0000 (17:24 +0200)] 
package: export debugsources in PKGDESTWORK as json

The source information used during packaging can be use from other tasks to
have more detailed information on the files used during the compilation and
improve SPDX accuracy.

Source files used during compilation are store as compressed zstd json in
pkgdata/debugsources/$PN-debugsources.json.zstd
Format:
{ binary1: [src1, src2, ...], binary2: [src1, src2, ...] }

I checked the sstate size, and it slightly increases using core-image-full-cmdline:
without patch: 2456792 KB sstate-cache/
with patch:    2460028 KB sstate-cache/
(4236 KB or 0.17%)

CC: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Daniel Turull <daniel.turull@ericsson.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agobinutils: Fix for CVE-2025-3198
Harish Sadineni [Tue, 10 Jun 2025 08:30:59 +0000 (01:30 -0700)] 
binutils: Fix for CVE-2025-3198

Upstream-Status: Backport [https://sourceware.org/git/?p=binutils-gdb.git;a=patch;h=ba6ad3a18cb26b79e0e3b84c39f707535bbc344d]
CVE: CVE-2025-3198

Signed-off-by: Harish Sadineni <Harish.Sadineni@windriver.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agobinutils: Fix CVE-2025-5244
Deepesh Varatharajan [Tue, 10 Jun 2025 04:36:54 +0000 (21:36 -0700)] 
binutils: Fix CVE-2025-5244

PR32858 ld segfault on fuzzed object
We missed one place where it is necessary to check for empty groups.

Backport a patch from upstream to fix CVE-2025-5244
Upstream-Status: Backport [https://sourceware.org/git/?p=binutils-gdb.git;a=patch;h=d1458933830456e54223d9fc61f0d9b3a19256f5]

Signed-off-by: Deepesh Varatharajan <Deepesh.Varatharajan@windriver.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agotune-cortexr52: Remove aarch64 for ARM Cortex-R52
Sandeep Gundlupet Raju [Mon, 9 Jun 2025 14:36:39 +0000 (08:36 -0600)] 
tune-cortexr52: Remove aarch64 for ARM Cortex-R52

Remove aarch64 for ARM Cortex-R52 processor as it supports only 32-bit
ISA but not 64-bit ISA. Also update ARMPKGARCH for cortexr52hf.

Signed-off-by: Sandeep Gundlupet Raju <sandeep.gundlupet-raju@amd.com>
Signed-off-by: Sandeep Gundlupet Raju <grsandeep85@gmail.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agokea: upgrade 2.6.1 -> 2.6.3
Yi Zhao [Mon, 9 Jun 2025 14:28:47 +0000 (22:28 +0800)] 
kea: upgrade 2.6.1 -> 2.6.3

ReleaseNotes:
https://downloads.isc.org/isc/kea/2.6.2/Kea-2.6.2-ReleaseNotes.txt
https://downloads.isc.org/isc/kea/2.6.3/Kea-2.6.3-ReleaseNotes.txt

Security fixes:
CVE-2025-32801
CVE-2025-32802
CVE-2025-32803

License-Update: Update copyright years

* Drop backport patches.

Signed-off-by: Yi Zhao <yi.zhao@windriver.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agokernel-module-split: fix conf file generation when KERNEL_SPLIT_MODULES=0
Dixit Parmar [Mon, 9 Jun 2025 13:45:35 +0000 (19:15 +0530)] 
kernel-module-split: fix conf file generation when KERNEL_SPLIT_MODULES=0

KERNEL_MODULE_AUTOLOAD defines the list of the kernel modules to be autoloaded
on boot. kernel-module-split.bbclass generates the required modules.load.d and
conf files for each kernel module. This conf files inturn read by system service
to perform module loading and configuration. When a kernel module is added to
KERNEL_MODULE_AUTOLOAD the conf files must be generated in all cases.
When KERNEL_SPLIT_MODULES=0 modprobe and autoload conf files are not
getting generated for the kernel modules.
To fix that enhanced the class implementation by separating out conf
file handling mechanism in two functions, generate_conf_files() and
frob_metadata(). generate_conf_files() handles no-split case where as
frob_metadata() keeps handling the existing case for spliting the modules.
Splitted common handling/generation of conf files stuff in to handle_conf_files()
function which gets invoked by both frob_metadata() and generate_conf_files()
on top of the scenario specific handling done in respective functions.
This implementation covers generation of the conf files for in-tree kernel
modules as well as standalone kernel module built as seperate package/recipe.

[YOCTO #15145]

Signed-off-by: Dixit Parmar <dixitparmar19@gmail.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agolibtheora: upgrade 1.1.1 -> 1.2.0
Gyorgy Sarvari [Mon, 9 Jun 2025 10:47:32 +0000 (12:47 +0200)] 
libtheora: upgrade 1.1.1 -> 1.2.0

Drop no-docs.patch, and use "--disable-doc" configuration instead.
Drop autoreconf.patch, because it is included in the release.

Add 0001-add-missing-files.patch to mitigate a release issue, which
caused some files to be missing from the tarball.

Major changes:
  - New 'ptalarbvorm' encoder
  - New th_encode_ctl option for copying configuration from an existing
    setup header, useful for splicing streams.
  - Added support for RISC OS.
  - Improved ARM support.
  - Various speed, bug fixes and code quality improvements.

See CHANGES file for full changelog.

Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agopiglit: drop python3-six dependency
Guðni Már Gilbert [Mon, 9 Jun 2025 08:50:26 +0000 (08:50 +0000)] 
piglit: drop python3-six dependency

six was dropped as a dependency five years ago:
https://gitlab.freedesktop.org/mesa/piglit/-/commit/288e7b3b74a88e5cfa6a7fe26d52bf093e16365f

Signed-off-by: Guðni Már Gilbert <gudni.m.g@gmail.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agoRevert "coreutils: split out coreutils-getlimits"
Etienne Cordonnier [Sun, 8 Jun 2025 21:00:32 +0000 (23:00 +0200)] 
Revert "coreutils: split out coreutils-getlimits"

This reverts commit 149584beef1e1b28a4e51f40515204e7edaa9d1f.

The coreutils-getlimits package is no longer needed since findutils-ptest
now uses its own getlimits implementation. This improves modularity and
makes it easier to use alternative coreutils implementations via PREFERRED_PROVIDER
(e.g. uutils-coreutils recipe in meta-openembedded).

Signed-off-by: Etienne Cordonnier <ecordonnier@snap.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agofindutils: use getlimits from findutils
Etienne Cordonnier [Sun, 8 Jun 2025 21:00:31 +0000 (23:00 +0200)] 
findutils: use getlimits from findutils

Having findutils depending on coreutils-getlimits makes it harder to replace GNU coreutils with uutils-coreutils
using PREFERRED_PROVIDER.
Also, the findutils repository contains a version of getlimits used for the tests, so there is no need to depend
on coreutils's version of getlimits.

Signed-off-by: Etienne Cordonnier <ecordonnier@snap.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agosystemtap: correctly set include location for the python module
Alexander Kanavin [Fri, 6 Jun 2025 20:58:06 +0000 (22:58 +0200)] 
systemtap: correctly set include location for the python module

This issue was as well exposed by setting S to be in UNPACKDIR.

Signed-off-by: Alexander Kanavin <alex@linutronix.de>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agoutil-linux: Improve the package licenses
Peter Kjellerstedt [Thu, 5 Jun 2025 21:31:21 +0000 (23:31 +0200)] 
util-linux: Improve the package licenses

The default license for util-linux is GPL-2.0-or-later.

Most of the applications and the libraries are also linked with
libcommon.la, which uses these licenses:
LGPL-2.1-or-later & BSD-2-Clause & BSD-3-Clause & MIT

Set the above licenses for all dynamic packages, unless explicitly
specified.

In particular, this avoids setting BSD-4-Clause-UC and EUPL-1.2 on
packages where they do not belong.

License-Update: See above
Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agoutil-linux: Move the license information to the recipe
Peter Kjellerstedt [Thu, 5 Jun 2025 21:31:20 +0000 (23:31 +0200)] 
util-linux: Move the license information to the recipe

util-linux does not share the license information with
util-linux-libuuid so it makes no sense to have it in the util-linux.inc
file.

This also drops libuuid/COPYING from LIC_FILES_CHKSUM as libuuid is not
built by this recipe.

License-Update: See above
Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agoutil-linux-libuuid: Set LIC_FILES_CHKSUM
Peter Kjellerstedt [Thu, 5 Jun 2025 21:31:19 +0000 (23:31 +0200)] 
util-linux-libuuid: Set LIC_FILES_CHKSUM

The license information is not the same as for util-linux, so set it in
the recipe rather than expecting util-linux.inc to set it.

License-Update: See above
Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agopiglit: upgrade to latest revision
Gyorgy Sarvari [Thu, 5 Jun 2025 05:42:15 +0000 (07:42 +0200)] 
piglit: upgrade to latest revision

Changelog:
  tests: Add test for GL_EXT_shader_clock
  tests: Add test for GL_EXT_shader_realtime_clock
  vulkan: Add cmat tests using constant as accumulator
  tests: test UBO and SSBO backed by the same buffer
  arb_direct_state_access: fix internaformat query behavior
  khr_texture_compression_astc: normalize error-color
  tests/util: provide simple array for cleanup handlers
  ext_external_objects: use piglit_set_destroy_func instead of atexit
  ext_external_objects: fix timeline semaphore validation error
  ext_external_objects: fix validation error with barriers
  ext_external_objects: fix validation error on cmdbuf usage
  ext_external_objects: fix validation error releasing resources
  ext_external_objects: fix validation error for cmdbuffer usage
  ext_external_objects: fix validation error with semaphore value
  arb_shading_language_420pack: test uniform block matching
  gl: test push pop attribute with color material
  GL_EXT_texture_integer: basic validation tests
  cmake: Fix mis-matching arguments warning
  fix invalid typecast

Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agopuzzles: upgrade to latest revision
Gyorgy Sarvari [Thu, 5 Jun 2025 05:28:40 +0000 (07:28 +0200)] 
puzzles: upgrade to latest revision

Changelog:
  Describe front-end pixel scaling in devel.but
  js: handle most of the device pixel ratio in the front end
  js: replace blitters array with a Map
  js: index blitters Map by C address of blitter structure
  js: trust CanvasRenderingContext2D.drawImage() to DTRT
  js: unwrap most drawing functions
  GTK: avoid explicitly using Cairo image surfaces
  Document getenv_bool()
  Update reference Git commit in devel.but
  Remove BLITTER_FROMSAVED from drawing API
  GTK: correct offset when copying backing to window
  GTK/Cairo: simplify do_blitter_load()
  GTK: add support for using a GdkFrameClock where available
  Mosaic: make solver notice when it gets stuck.
  Mines: put a warning on the web page about bug reports.
  Palisade: preference to clear complete regions.

Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agoutil-linux: fix ptest failure for musl
Chen Qi [Thu, 5 Jun 2025 03:42:10 +0000 (11:42 +0800)] 
util-linux: fix ptest failure for musl

The kill/decode test case fails for musl. The root cause is the test
case only considers glibc and uses 34 as SIGRTMIN while musl uses 35.
Add patches to fix this issue.

Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agobluez5: upgrade 5.82 -> 5.83
Guðni Már Gilbert [Wed, 4 Jun 2025 18:32:38 +0000 (18:32 +0000)] 
bluez5: upgrade 5.82 -> 5.83

Changelog:
Fix issue with handling BAP and removal of PAC.
Fix issue with handling SID for broadcast receiver.
Fix issue with handling HSP/HFP reconnection policy.
Fix issue with handling cable pairing and Sixaxis controllers.
Fix issue with handling virtual cable unplug for HID devices.
Fix issue with handling service records for HID devices.
Add support for AVDTP and TX timestamps.

Full changelog:
https://github.com/bluez/bluez/compare/5.82...5.83

Signed-off-by: Guðni Már Gilbert <gudni.m.g@gmail.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agosquashfs-tools: clean up PACKAGECONFIG
Ross Burton [Mon, 9 Jun 2025 11:19:42 +0000 (12:19 +0100)] 
squashfs-tools: clean up PACKAGECONFIG

LZMA isn't supported by the mainline kernel driver, so there's no real
point in enabling it.

Disable LZO by default, as LZO is very obsolete (last released in 2017)
and there are better algorithms available.

Move xattr to the end so it's separate from the compression algorithms.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agosquashfs-tools: update to 4.7
Ross Burton [Mon, 9 Jun 2025 11:19:41 +0000 (12:19 +0100)] 
squashfs-tools: update to 4.7

The REPRODUCIBLE_DEFAULT option was removed in 4.7[1].

Backport a patch to fix missing pthread typedefs, and a patch to fix
missing pid_t typedefs.

Also clean up the tasks to be neater and more concise.

[1] squash-tools ae9914d15293af7c596148aedd997b3058e35d9e

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agosquashfs-tools: put version in recipe filename
Ross Burton [Mon, 9 Jun 2025 11:19:40 +0000 (12:19 +0100)] 
squashfs-tools: put version in recipe filename

A recipe that fetches from git but is otherwise a released version
should not be using _git.bb recipe names.

Put the version in the filename, drop the explicit PV, and set the tag
name in the SRC_URI to verify the SHA.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 weeks agoelfutils: Upgrade 0.192 -> 0.193
Soumya Sambu [Fri, 6 Jun 2025 06:44:03 +0000 (12:14 +0530)] 
elfutils: Upgrade 0.192 -> 0.193

Refreshed patches:
0001-libelf-Add-libeu-objects-to-libelf.a-static-archive.patch
ptest.patch

Changelog:
Version 0.193 "Bugs erased, stacks traced"

debuginfod: Add CORS (webapp access) support to webapi and --cors option.

            Add --listen-address option for binding the HTTP listen socket
            to a specific IPv4 or IPv6 address.

            debuginfod client now caches x-debuginfod-* HTTP headers
            alongside downloaded files.

libdw: Add dwarf_language and dwarf_language_lower_bound functions.

       Improved support for DWARF6 language metadata as well as DWARF
       language constants for Nim, Dylan, Algol68, V and Mojo.

       dwarf_srclang is now forward-compatible with DWARF6 language
       constants.

libdwfl_stacktrace: Experimental new library interface for unwinding
                    stack samples into call chains, and tracking and
                    caching Elf data for multiple processes, building
                    on libdwfl.  Initially supports perf_events stack
                    sample data.

libelf: elf_scnshndx has been rewritten to be more robust, particularily
        for ELF files with more than 64K sections.

readelf: Improved handling of corrupt ELF data.

         --section-headers output now includes a "Key to Flags" explaining
         section flag meanings.

Signed-off-by: Soumya Sambu <soumya.sambu@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
2 months agosystemtap: put version in recipe filename
Ross Burton [Mon, 9 Jun 2025 11:19:47 +0000 (12:19 +0100)] 
systemtap: put version in recipe filename

A recipe that fetches from git but is otherwise a released version
should not be using _git.bb recipe names.

Put the version in the filename and drop the explicit PV.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
2 months agotcf-agent: put version in recipe filename
Ross Burton [Mon, 9 Jun 2025 11:19:46 +0000 (12:19 +0100)] 
tcf-agent: put version in recipe filename

A recipe that fetches from git but is otherwise a released version
should not be using _git.bb recipe names.

Put the version in the filename and drop the explicit PV.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
2 months agounfs3: upgrade to 0.11
Ross Burton [Mon, 9 Jun 2025 11:19:45 +0000 (12:19 +0100)] 
unfs3: upgrade to 0.11

Drop all of the patches that have been merged upstream.

The build no longer needs the flex runtime library so remove flex from
DEPENDS.

License-Update: contributor list updated.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
2 months agounfs3: clean up DEPENDS
Ross Burton [Mon, 9 Jun 2025 11:19:44 +0000 (12:19 +0100)] 
unfs3: clean up DEPENDS

flex-nativesdk is both the wrong name (should be nativesdk-flex), and in
nativesdk builds it will be a dependency anyway thanks to the DEPENDS
remapping.

Merge multiple assignments into a single line.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
2 months agounfs3: put version in recipe filename
Ross Burton [Mon, 9 Jun 2025 11:19:43 +0000 (12:19 +0100)] 
unfs3: put version in recipe filename

A recipe that fetches from git but is otherwise a released version
should not be using _git.bb recipe names.

Put the version in the filename, and drop the explicit PV.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
2 months agomobile-broadband-provider-info: put version in recipe filename
Ross Burton [Mon, 9 Jun 2025 11:19:39 +0000 (12:19 +0100)] 
mobile-broadband-provider-info: put version in recipe filename

A recipe that fetches from git but is otherwise a released version
should not be using _git.bb recipe names.

Put the version in the filename and drop the explicit PV.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
2 months agoblktrace: put version in recipe filename
Ross Burton [Mon, 9 Jun 2025 11:19:38 +0000 (12:19 +0100)] 
blktrace: put version in recipe filename

A recipe that fetches from git but is otherwise a released version
should not be using _git.bb recipe names.

Put the version in the filename and drop the explicit PV.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
2 months agosassc: put version in recipe filename
Ross Burton [Mon, 9 Jun 2025 11:19:37 +0000 (12:19 +0100)] 
sassc: put version in recipe filename

A recipe that fetches from git but is otherwise a released version
should not be using _git.bb recipe names.

Put the version in the filename and drop the explicit PV.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
2 months agolibsass: put version in recipe filename
Ross Burton [Mon, 9 Jun 2025 11:19:36 +0000 (12:19 +0100)] 
libsass: put version in recipe filename

A recipe that fetches from git but is otherwise a released version
should not be using _git.bb recipe names.

Put the version in the filename and drop the explicit PV.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
2 months agolibmodulemd: add manpages
Ross Burton [Mon, 9 Jun 2025 11:19:35 +0000 (12:19 +0100)] 
libmodulemd: add manpages

Inherit manpages and optionally enable the manpages. They're in the
repository so don't need more tools to build.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
2 months agolibmodulemd: remove libmagic option
Ross Burton [Mon, 9 Jun 2025 11:19:34 +0000 (12:19 +0100)] 
libmodulemd: remove libmagic option

Meson shows this warning in do_configure:

  WARNING: libmagic option is obsolete. libmodulemd can detect
  compression formats without a magic library now. Please stop using
  this option. It will be removed in the future and will cause a meson
  failure.

Remove the option as it's now redundant.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
2 months agolibmodulemd: put version in recipe filename
Ross Burton [Mon, 9 Jun 2025 11:19:33 +0000 (12:19 +0100)] 
libmodulemd: put version in recipe filename

A recipe that fetches from git but is otherwise a released version
should not be using _git.bb recipe names.

Put the version in the filename, and drop the explicit PV.

Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>