Changelog:
===========
- Fix spurious readiness flags in some of the wait functions
- Fix high CPU usage using the wait_c function on Windows
- Fix bad data on error in binary copy
- Respect Copy.set_types() in TEXT copy in C version, consistently with the Python version
- Don't raise warning, and don’t leak resources, if a builtin function is used as JSON dumper/loader function
- Improve performance of Python conversion on results loading
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
* Device database changes:
- Support iPad Pro M5 family devices
- Support Apple Vision Pro M5
- Support MacBook Pro 14-inch M5
* Bug Fixes:
Fix: array initialization compatibility with MSVC
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
The project's readme indicates that some parts of the code is covered
by Apache license - add it to the RECIPE accordingly.
Reported-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com>
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
The below reference clearly states that GPL-2.0-with-OpenSSL-exception
is to be used with GPL 2.0 or GPL3.0 and not as a standalone license.
Therefore, update the correct license.
Reference:
569d72e13e/docs/openssl-exception-gpl-2.0.yml (L7)
Signed-off-by: Sana Kazi <Sana.Kazi@bmwtechworks.in>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
postmaster binary doesn't seem to be installed by default, making the
sysv init script fail to start postgres.
Adjust it to use pg_ctl to start the service, just like its systemd
service counterpart.
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Remove a number of ignored tests from the ignore-list. They were missing
some files that weren't installed - after installing them, they pass.
The remaining mongodb test on the list hasn't been tested.
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
The /etc/machine-id file with some value is required for the integration tests,
otherwise it fails with the following error:
terminate called after throwing an instance of 'sdbus::Error'
what(): [org.freedesktop.DBus.Error.FileNotFound] Failed to process bus requests (No such file or directory)
Aborted
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
The recipe inherits the ptest class, however installs no tests nor
run-ptest script.
This change rectifies this.
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
The original content of the ptest package, manette-test, is a helper demo
application (like evtest), and not a test suite. Also, the recipe did not
provide a run-ptest script.
Fix it by installing the actual tests, and adding a run-ptest script.
Note that the test folder structure looks like a gnome desktop test suite
(and the application is under the gnome umbrella), however the project
doesn't provide all necessary scaffolding for gnome-desktop-test to work, so
the tests are executed directly from the run-ptest script.
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Since commit 91e428d4cd ("opencv: Support building for native"),
opencv can be built on a native target. However, not all of their
PACKAGECONFIG dependencies (ade, libgphoto2, ffmpeg, gstreamer1.0 and
gstreamer1.0-plugins-base) can be built in this context due to the
missing native BBCLASSEXTEND.
These recipes would need to be extended for native build, but specific
work is required to ensure they work properly. To make opencv-native work
for now, remove the above features for the native build.
Effectively, PACKAGECONFIG for native is target minus gapi, gstreamer,
gphoto2 and libav.
Fixes this warning from AB[0]:
WARNING: Nothing PROVIDES 'gstreamer1.0-plugins-base-native' (but virtual:native:[...]/meta-openembedded/meta-oe/recipes-support/opencv/opencv_4.12.0.bb DEPENDS on or otherwise requires it). Close matches:
WARNING: Nothing PROVIDES 'ffmpeg-native' (but virtual:native:[...]/meta-openembedded/meta-oe/recipes-support/opencv/opencv_4.12.0.bb DEPENDS on or otherwise requires it). Close matches:
WARNING: Nothing PROVIDES 'libgphoto2-native' (but virtual:native:[...]/meta-openembedded/meta-oe/recipes-support/opencv/opencv_4.12.0.bb DEPENDS on or otherwise requires it). Close matches:
WARNING: Nothing PROVIDES 'gstreamer1.0-native' (but virtual:native:[...]/meta-openembedded/meta-oe/recipes-support/opencv/opencv_4.12.0.bb DEPENDS on or otherwise requires it). Close matches:
WARNING: Nothing PROVIDES 'ade-native' (but virtual:native:[...]/meta-openembedded/meta-oe/recipes-support/opencv/opencv_4.12.0.bb DEPENDS on or otherwise requires it). Close matches:
WARNING: Nothing RPROVIDES 'opencv-dev-native' (but virtual:native:[...]/meta-openembedded/meta-oe/recipes-support/opencv/opencv_4.12.0.bb RDEPENDS on or otherwise requires it)
WARNING: Nothing RPROVIDES 'opencv-native' (but virtual:native:[...]/meta-openembedded/meta-oe/recipes-support/opencv/opencv_4.12.0.bb RDEPENDS on or otherwise requires it)
[0]: https://autobuilder.yoctoproject.org/valkyrie/#/builders/82/builds/469/steps/13/logs/warnings
Signed-off-by: Corentin Guillevic <corentin.guillevic@smile.fr>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
After commit 392f0f0ea7 ("linux-serial-test: Bump SRCREV to allow CMake 4+ compatibility"),
an infinite while loop can occur even when the timeout has been reached.
This patch fixes that regression.
Signed-off-by: Emanuele Ghidoli <emanuele.ghidoli@toradex.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
After commit 392f0f0ea7 ("linux-serial-test: Bump SRCREV to allow CMake 4+ compatibility"),
the behavior of linux-serial-test changed: it now returns 125 instead of 0
when the number of read characters differs from the written ones.
This patch fixes that regression.
Signed-off-by: Emanuele Ghidoli <emanuele.ghidoli@toradex.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Dropped two patches which got incorporated in this version.
Changelogs:
2.0.16:
Add a 'search' feature to fwupdtool and fwupdmgr
Fix missing release locations when loading from artifact
Fix remaining issues to make updates on FreeBSD work
2.0.15:
Allow child devices to use the parent name as a prefix
Add newer commands and options for Fish completion
Allow installing archives named as .CAB rather than .cab
Erase Firehose modem devices correctly
Fix Goodix enumeration issues
Fix sending firmware reports without --force
Fix the FreeBSD build
Fix version number of BnR MTD devices
Require additional requirements for the default PS5512 devboard
Require a full system shutdown for all Micron NVMe updates
Use a better name for Elan touchpad and Intel PCH SPI devices
New supported hw:
Foxconn SDX61 Modem
Jabra Evolve2 child devices
NVIDIA ConnectX-6, ConnectX-7 and ConnectX-8 NICs
2.0.14:
Add support for ignoring the network connectivity requirement
Allow building on RHEL-9 and RHEL-10
Allow plugins to know the firmware version during update
Allow UEFI capsule devices to opt-out of Capsule-on-Disk
Allow unsetting HwID plugin context flags
Allow upgrading from a zero "empty" UEFI dbx
Add an automatic firehose counterpart to the QCDM modem device
Disable signature time checks when verifying firmware
Do not add a vendor ID of UNKNOWN when the signature has no vendor
Do not discover ThunderBolt retimer devices when run in single-shot mode
Do not use deprecated libflashrom API
Enhance firmware metadata generation in firmware_packager
Ensure Lexar NVMe drives use a proper version number
Fix parsing and writing UF2 extension sections
Fix Synaptics RMI initialization for new devices
Fix updating DFOTA and MBIM modem devices
Move some vendor name fixups to the quirk file
Remove CapsuleOnDisk HwID match for Dell
Return a sensible error when using build-cabinet wrong
Set the firehose loader filename in a more permissive way
Update the mapping for TPM vendor names
Verify the checksum of the serialized data in tests
Work around a libmbim bug when detaching
New supported hw:
Egis MoC devices
Framework QMK devices
ILITEK touch controllers
SteelSeries Arctis Nova 3P
2.0.13:
Add a daemon config option to ignore efivars free space
Add support for glob-aware version comparison requirements
Allow targeting specific regions in FMAP when using flashrom
Detect static variables and magic numbers during code review
Remove the unused hailuck and rts54hid plugins
Align MTD erase up to the erasesize as necessary
Allow parsing IGSC OptionROM when using fwupdtool
Allow removing private flags from UEFI capsule devices in quirks
Do not copy the vendor for Intel reference ME firmware
Do not use an interactive console if stdout is redirected
Fix the UEFI self-test when the capsule splash is disabled
Get better device information when using PCI-backed MTD devices
Get the Intel GPU SKU and SVN when using BMG hardware
Make MBIM modem devices emulatable
Make sure fwupdtool.exe is available in the Windows PATH
Only show the 'Full Disk Encryption Detected' warning when required
Set all QCDM modem devices to raw mode when updating
Show all devices for fwupdtool get-devices --show-all --force
Show correct dbx version if non-Microsoft entries are present
Show KEK device attributes in fwupdmgr
Use an alternate GUID when the Intel GPU is in recovery mode
Use the kernel netlink hotplug socket when there is no Udev
Various small changes to speed up startup by 60% and lower RSS by 40%
New supported hw:
HP USB-C 100W G6 Dock
Logitech Bulk Controller pheripherals
More MediaTek scaler devices
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
The ptest package requires dbus as a runtime dependency, otherwise the tests fail
with the following error:
Failed to connect to daemon: Could not connect: No such file or directory
FAIL: fwupd/fwupd.test (Child process exited with code 1)
SUMMARY: total=1; passed=0; skipped=0; failed=1; user=0.7s; system=0.1s; maxrss=21716
FAIL: fwupd/fwupd.test (Child process exited with code 1)
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
After upgrade to 0.24.0, do_configure failed with error:
| CMake Error at cmake/Embed.cmake:3 (find_program):
| Could not find XXD using the following names: xxd
| Call Stack (most recent call first):
| src/stdlib/CMakeLists.txt:1 (include)
[1] df21d917d9
Signed-off-by: Changqing Li <changqing.li@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
This upgrade includes fix for CVE-2025-52885
Poppler 25.10.0 Changelog:
=========================
core:
* Fix image signature getting lost
* Don't embed substitutions for base14 fonts
* Form font improvements
* Handle signatures padded with random data
* Add feature to Ink annotation to render with multiply blend mode
* Internal code improvements
* Fix crashes in malformed documents
glib:
* Fix signature text
* Add feature to Ink annotation to render with multiply blend mode
cpp:
* Added embedded_file::unicodeName function
Poppler 25.09.1 Changelog:
==========================
core:
* Internal code improvements
build system:
* Fix generated .pc files when using old gpgme
Poppler 25.09.0 Changelog:
=========================
core:
* Speed improvements when reusing the same document with different output devices
* Speed improvements when reading from network file systems
* Internal code improvements
* Fix crashes in malformed documents
glib:
* fix ODR issue with enum
utils:
* pdftohtml: Fix text positioning. (Regressed in 25.07.0)
build system:
* Better pkgconfig support for static builds
Signed-off-by: Yogita Urade <yogita.urade@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Update to latest release and add support for zsh completions.
Signed-off-by: Marc Ferland <marc.ferland@sonatest.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Details: https://nvd.nist.gov/vuln/detail/CVE-2025-6750
Pick the patch that is marked to resolve the issue linked in
the nvd report.
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Details: https://nvd.nist.gov/vuln/detail/CVE-2025-2925
Pick the patch that's marked to resolve the issue linked
in the nvm report.
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Details: https://nvd.nist.gov/vuln/detail/CVE-2025-2924
Pick the patch that is marked to resolve the issue linked in the
nvd report.
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Details: https://nvd.nist.gov/vuln/detail/CVE-2025-2914
Pick the patch that is linked in the issue from the nvd report.
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Details: https://nvd.nist.gov/vuln/detail/CVE-2025-2310
Pick the patch that mentions the CVE in its description.
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Details: https://nvd.nist.gov/vuln/detail/CVE-2025-2153
Pick the patch that resolved the issue from the nvd report.
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
A corresponding patch was removed with the upgrade to 3.4.1 since it had
been integrated upstream. However, it had also been reverted again,
which was not noticed.
This backports a subsequent fix for the problem.
Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
There's no need to inherit python3native as there are no python3- native
dependencies.
Remove RDEPENDS_${PN}, these are all libraries that are linked to so the
relevant RDEPENDS are created. The fact that this recipe is still using
the old syntax suggests that they're not needed.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
The libaio subsystem in MariaDB was rewritten in 11.4.8,
which made the previously used patch called
0001-aio_linux-Check-if-syscall-exists-before-using-it.patch
non-applicable.
This patch was deleted in commit f3cd830c09 ("mariadb: 11.4.8")
but it's still needed, because the original code was simply
moved into a new source file.
Add a new patch applying the same change to tpool/aio_libaio.cc
to fix the build error on riscv32.
Signed-off-by: Zoltán Böszörményi <zboszor@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
