Remove groups entries, as that has been removed in favor of the coreutils implementation.
Add a patch to support older host compilers.
(From OE-Core rev: 8d1041c02c600229b3487bc37f929c2b57f0b3ce)
Signed-off-by: Alexander Kanavin <alex@linutronix.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Backport a kea patch to address 1.87.0 compatibility.
(From OE-Core rev: ac328183d4592ad146c41e48c2c92d1dbb53a894)
Signed-off-by: Alexander Kanavin <alex@linutronix.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Upstream now ships pre-generated libmath.h, so we don't need to do so.
Drop 0001-dc-fix-exit-code-of-q-command.patch
(merged upstream).
License-update: copyright years.
(From OE-Core rev: ab7c33ba324af89a0f96d9eb480ce422f3b92fb5)
Signed-off-by: Alexander Kanavin <alex@linutronix.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
License-Update: BSD items clarified to be 2-clause:
d4ec06ec58
passprompt plugin removed upstream.
Add an enable a multilink option, as the build breaks without it at the linking
step (upstream presumably didn't test that it works): https://github.com/ppp-project/ppp/issues/541
(From OE-Core rev: e6ea3ade8f3977ec93c68229d2660bbc335a80a6)
Signed-off-by: Alexander Kanavin <alex@linutronix.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The transaction model was only introduced in git 2.27 whereas Ubuntu focal
(20.04) has 2.25. This causes failures. We don't need the transations here
so simply drop the commit piece, fixing on older git versions.
Credit to Nick Owens <nick.owens@eero.com> for working out how to fix it.
(Bitbake rev: 0723ec9d4cd7c9b2d46904c3a038be123feea374)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
License-Update: renamed file
Convert from setuptools to flit.
(From OE-Core rev: 33dfffac6f65145b6e8671a8988a62f1f0e41f72)
Signed-off-by: Alexander Kanavin <alex@linutronix.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Some tightly coupled packages still need updating (we are
waiting on upstream), but there's no reason to hold the dev
kernel update on those packages (core-image-minimal is working)
(From OE-Core rev: c791afb0df94cf44dcb371dcd62dee2f2533de53)
Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Otherwise fragments with 'require' statements will not parse,
as the parser will not be able to find files referred to by
those statements.
Add such a statement to the test fragment so that the scenario
is tested.
[YOCTO #15707]
(From OE-Core rev: d88e63d1a7867b3a4ebfc57d472900136b5361d8)
Signed-off-by: Alexander Kanavin <alex@linutronix.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
It does not compile anymore with 2.46+ as we dropped the patch
https://bugs.webkit.org/show_bug.cgi?id=250681
This patch it seems is still needed but has not yet been
accepted upstream yet.
(From OE-Core rev: be5dd5630a0fa53d16b661b088d6b4fb31e8e9b2)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Changelog:
===========
- Never add a vendored smmap directory to sys.path
- Revise and update the readme
- Set Dependabot submodule update cadence to weekly
- Add missing asserts in test_base.py
- Use contextlib.suppress instead of except: pass
- Add support for Python 3.13
- Potential Race Condition Fix - OS Rename & Chmod - PermissionError
- Bump gitdb/ext/smmap from 256c5a2 to 04dd210
- Bump actions/setup-python from 4 to 5
(From OE-Core rev: d899b50547e319cca8bea41301a4758e7635b6cf)
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Changelog:
===========
- remove unneeded use of Import::Into
- fix tests that invoke Test::Warnings->import directly
- avoid redefinition warnings when $^W is enabled in tests
- allow tests to pass even when Import::Into is not installed
- add backcompat shim for cases where Test::More is loaded in tests
after Test::Warnings
- fix Test2 compability for done_testing()
- Test2::Warnings added as a simple wrapper (for now)
(From OE-Core rev: f203ec55e1fd12e4e6822839a4979b3054617d59)
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
License-Update: copyright updated to 2025
Changelog:
===========
- Fixed a crash if the controller product name is NULL
- Fixed the PS3 controller mapping on Windows
- Allow rendering during the modal resize loop on macOS
- Corrected CoreAudio surround sound channel layouts
- Enabled high refresh rates on iOS
- Fixed SDL_OpenURL() on iOS 18.2 and newer
- Implemented SDL_OpenURL() on tvOS
- Fixed Chinese locales on PSP
(From OE-Core rev: f59f1ef898993a09cdb243d80b1bab624472ea10)
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Changelog:
==========
* GtkFileChooser:
- Fix a focus problem with the server entry
* GtkEmojiChooser:
- Update to CLDR 46 / Unicode 16
* GtkAboutDialog:
- Small improvements
* Wayland:
- Fix some Vulkan sizing corner cases
- Fix problems with rotated, offloaded images
- Fix a few crashes
* Demos:
- Don't create extra windows for D-Bus activation
(From OE-Core rev: ed161f0a0cbcef50c43820d58989e5a020763dbf)
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Changelog:
===========
- Fix unusual PAX header handling for tarerofs;
- Fix LIBARCHIVE.xattr decoding;
- Speed up handling of incompressible data;
- Add a '-E^fragdedupe' option to explicitly disable fragment deduplication;
- Fixes around '-Eall-fragments' and '-Ededupe';
- Use external xxhash library if possible.
(From OE-Core rev: 441abb84142eae10155fbcc3c5a99091afa5dfd9)
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Upgrade to U-Boot 2025.01.
Remove the patch as it has been upstreamed.
(From OE-Core rev: e16a24f04b56ca3fec6509a3286e2e3bb5c21c1e)
Signed-off-by: Fabio Estevam <festevam@denx.de>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
In the same fashion as the previous commit ("ref-manual/packages: move
ptest section to the test-manual"), move the runtime testing section of
the development tasks manual to the test environment manual.
Add a link to it from the test-manual/intro document.
(From yocto-docs rev: 6b44257874858db3aa426d3e84a79c41cb4937a3)
Signed-off-by: Antonin Godard <antonin.godard@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
[ YOCTO #15106 ]
It makes more sense to document ptests in the test-manual. Since ptests
are still related to packages, keep a link to ptests from packages.rst
to the test-manual.
Reported-by: Yoann Congal <yoann.congal@smile.fr>
(From yocto-docs rev: b389c06b709e4791e1cce5e8a5b58f6b0cd03a14)
Signed-off-by: Antonin Godard <antonin.godard@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
[ YOCTO #15218 ]
There are different ways of configuring the build host when
authentication to a mirror (SSTATE_MIRRORS or SOURCE_MIRROR_URL) is
required.
Document these methods, and begin with the preferred approach which is
to edit ~/.netrc appropriately. When that is not possible, specifying
the credentials from the URL is also possible, so document it as well.
Reported-by: peter.zsifkovits@at.bosch.com
Tested-by: Antonin Godard <antonin.godard@bootlin.com> # Debian 12 container
(From yocto-docs rev: 4c303cce0627e0f2ce52e72597541e220585aaed)
Signed-off-by: Antonin Godard <antonin.godard@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The command `devtool deploy-target` does not handle any dependency of
the application, even when properly listed in DEPENDS or RDEPENDS. While
the current documentation was clear on that, it felt needed to add a
note on the fact that build-time dependencies may also result in
runtime dependencies, especially for people new to the project.
Add a warning block below the existing deploy-target documentation.
Also, link to the Overview and Concepts page, which explains it
thoroughly.
[ YOCTO #14377 ]
Reported-By: Robert Berger <pokylinux@reliableembeddedsystems.com>
(From yocto-docs rev: 56c84004daad5f8a5db53e36852331435d57c8af)
Signed-off-by: Antonin Godard <antonin.godard@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
By default when sending patches Git adds the git version at the bottom
of the patch, which is not necessary in most cases. Suggest to remove it
when configuring Git when sending patches.
Suggested-By: Jaekyu Lee <jaekyu.lee@lge.com>
(From yocto-docs rev: d3465691231f10466703198a831a681803ffdfad)
Signed-off-by: Antonin Godard <antonin.godard@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Being able to set maintainers is really a "per layer" configuraiton item. Use
include_all so that such maintainer information would be added for all configured
layers, if the layer chooses to have such information in a
conf/distro/include/maintainers.inc file in a given layer.
(From OE-Core rev: d6b9dc4629642a4bf0d8f34a90bd8e3291c87501)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The cve_check functionality to parse CVE IDs from the patch filename and
patch contents have been reworked to improve parsing and also utilize
tests. This ensures that the parsing works as intended.
Additionally, the new patched_cves dict has a few issues I tried to fix
as well. If multiple patch files exist for a single CVE ID, only the
last one will show up with the "resource" key. The value for the
"resource" key has been updated to hold a list and return all patch
files associated with a given CVE ID. Also, at the end of
get_patch_cves, CVE_STATUS can overwrite an existing entry in the dict.
This could cause an issue, for example, if a CVE has been addressed via
a patch, but a CVE_STATUS line also exists that ignores the given CVE
ID. A warning has been added if this ever happens.
(From OE-Core rev: 87c6da681609b4f8e048eca2a27ae8e068c724e1)
Signed-off-by: Colin McAllister <colinmca242@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Two warning lines in cve_check.py reference a variable that doesn't
exist. These would cause a runtime error if the conditions they are
hidden in were to be entered. The log lines have been updated to no
longer reference an undefined variable.
(From OE-Core rev: c9d059e8a362b3c9d604f7ebe8fd1dd994f0af6b)
Signed-off-by: Colin McAllister <colinmca242@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
To make use of the hardware watchdog it is sufficient to set
the RuntimeWatchdogSec= option in /etc/systemd/system.conf.
It defaults to 0 (i.e. no hardware watchdog use). Set it to
a value like 20s and the watchdog is enabled.
ref: https://0pointer.de/blog/projects/watchdog.html
(From OE-Core rev: 1c61a1eb9c4faa9ab32b0440bbdd88c22c3cd945)
Signed-off-by: Benjamin Szőke <egyszeregy@freemail.hu>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
After commit [spdx 3.0: Rework how SPDX aliases are linked] applied, it set
license_text_map with SPDX alias other than actual ID
The property of simplelicensing_customIdToUri is ListProp(ObjectProp(DictionaryEntry))),
and class DictionaryEntry has key and value, the property of value is StringProp other
than ObjectProp in which could not support to decode/extract SPDX alias with actual ID
in image jsonld file
-----------
"simplelicensing_customIdToUri": [
{
"type": "DictionaryEntry",
"key": "LicenseRef-PD",
"value": "http://spdxdocs.org/openembedded-alias/by-doc-hash/d53e90e23b12c4ad640809a74a810e86f31c76cdbdf36487712d22a33d53362a/sqlite3-native/UNIHASH/license-text/PD"
}
],
-----------
Add special code in the linking to manually go through all of the
simplelicensing_customIdToUri DictionaryEntry items and resolve
any aliases to actual objects
-----------
"simplelicensing_customIdToUri": [
{
"type": "DictionaryEntry",
"key": "LicenseRef-PD",
"value": "http://spdx.org/spdxdocs/sqlite3-native-e5cc0672-d8dd-57e8-a2df-fe4615831fda/162c62b5b011cd3f82f413b3dae4d6d1542201552c964d5ce69fe170e0720b85/license-text/PD"
}
]
-----------
Suggested-by: Joshua Watt <jpewhacker@gmail.com>
(From OE-Core rev: f16c2eae5a7e1a483e1d537d8711d83fede75691)
Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
So that PARSE_ALL_RECIPES will work.
(From OE-Core rev: c925098dc0faf2b69637de914958557b23615c7a)
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Building weston with core-image-weston SDK fails:
```
../libweston/renderer-gl/gl-shader-config-color-transformation.c:29:10: fatal error: GLES3/gl3.h: No such file or directory
29 | #include <GLES3/gl3.h>
| ^~~~~~~~~~~~~
```
Both GLES2 and GLES3 implementations are contained in libGLESv2.so.2,
which is packaged in libgles2-mesa. However, the headers are split
between libgles2-mesa-dev and libgles3-mesa-dev, which is why the
GLES3 headers end up missing in the SDK sysroot.
Add a dependency so the GLES3 headers are properly associated with
the GLES3 implementation.
(From OE-Core rev: 7e1308ec413e69a8427ac5998431005d9e4b8033)
Signed-off-by: Tom Hochstein <tom.hochstein@oss.nxp.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Extends the 'enable-fragment' and 'disable-fragment' commands so that
they accept multiple fragments at once as a convenience for the user
(From OE-Core rev: 50c3cdb3a3b9c7daa55ff26d302d95e5f350e4d2)
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The current implementation of the config fragments is too aggressive in
checking files; any file in the fragment directory is checked, including
hidden files or files with weird extensions. In particular, if an editor
is creating temporary backup files when editing, these will be checked
and will almost assuredly fail, which prevents the tool from running.
Add a filter so that only non-hidden files that end with .conf are
checked.
(From OE-Core rev: 93edab0e2bccacb808421f0766d587c576c1a12b)
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Error: Transaction test error:
file /usr/bin/find-debuginfo conflicts between attempted installs of debugedit-5.1-r0.core2_64 and lib32-debugedit-5.1-r0.core2_32
The differences of find-debuginfo are as follows:
@@ -360,7 +360,7 @@ add_minidebug()
# symbol and NOBITS sections so cannot use --keep-only because that is
# too aggressive. Field $2 is the section name, $3 is the section type
# and $8 are the section flags.
- local remove_sections=`x86_64-poky-linux-readelf -W -S "$debuginfo" \
+ local remove_sections=`i686-pokymllib32-linux-readelf -W -S "$debuginfo" \
| awk '{ if (index($2,".debug_") != 1 \
&& ($3 == "PROGBITS" || $3 == "NOTE" || $3 == "NOBITS") \
&& index($8,"A") == 0) \
(From OE-Core rev: 2bc0fcacf0d33eea59a33d28f7263c90c2bf0823)
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
