When enabling multilib with lib32, the radiusd will use etc file for lib32 as default
#systemctl status radiusd
......
/usr/sbin/radiusd -d /etc/lib32-raddb
It should be lib64 as default.
Signed-off-by: Liu Yiding <liuyd.fnst@fujitsu.com>
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Apparently the git repo in the SRC_URI stopped supporting git
protocol. Switch to https to be able to fetch the source successfully.
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
bpftrace need line up with kernel architecture type,
so add it to NON_MULTILIB_RECIPES.
bpftrace must be built with the same architecture (32-bit/64-bit) as the kernel
to ensure compatibility when tracing kernel functions and using eBPF features.
(cherry-picked from f8498f2699 in master )
Signed-off-by: Sunil Dora <sunilkumar.dora@windriver.com>
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Changelog:
============
- Fix bolt exits when opening an empty version file
- Fix failure on integration test due to the change of the latest pygobject
- Several CI improvements and fixes.
- Fixed memory leak in boltctl
- The unused codes were removed from daemon
- Fixed a NULL syspath variable causes SIGSEGV
- A new NHI for REMBRANDT.
- CI fixes.
- Don't install an empty DB directory.
- Fixed: Determine the string length before writing file.
- Fixed: Free on error to prevent resource leak.
- Add a 'nopcie' security level since some devices report 'nopcie' when Thunderbolt
is disabled through BIOS setting.
- Markdown lint styling is used for documents.
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
(cherry picked from commit 98969e3b44)
Signed-off-by: Ankur Tyagi <ankur.tyagi85@gmail.com>
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Gutenprint install hooks run in parallel but depend on each other. This
is a race condition and might trigger a build failure (e.g on AB [0]):
| chmod 700 $WORKDIR/image/usr/libexec/cups/backend/backend_gutenprint
| chmod: cannot access '$WORKDIR/image/usr/libexec/cups/backend/backend_gutenprint': Not a directory
| make[5]: *** [Makefile:2166: install-exec-hook] Error 1
Fixes this by adding an explicit dependency between the dependent
targets.
[0]: https://autobuilder.yoctoproject.org/valkyrie/#/builders/87/builds/46/steps/33/logs/stdio
Signed-off-by: Yoann Congal <yoann.congal@smile.fr>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
* According to latest comment [1] and the mentioned pull request [2],
build an ENABLE(WEBASSEMBLY) && !ENABLE(JIT) configuration is
supported, so original issue already fixed in current version, the
EXTRA_OECMAKE setting is not needed anymore.
* This EXTRA_OECMAKE setting causes following configure error on
beaglebone-yocto, remove the setting to let the configure process
decide the configuration:
CMake Error at Source/cmake/WebKitFeatures.cmake:312 (message):
ENABLE_JIT conflicts with ENABLE_C_LOOP.You must disable one or the
other.
[YOCTO #15254]
[1] https://github.com/WebKit/WebKit/pull/17447
[2] https://github.com/WebKit/WebKit/pull/17688
Signed-off-by: Jiaying Song <jiaying.song.cn@windriver.com>
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
It fails to compile webkitgtk when debug build enabled:
| /path_to//tmp/work/core2-64-poky-linux/webkitgtk3/2.48.1/webkitgtk-2.48.1/Source/bmalloc/libpas/src/libpas/bmalloc_heap_config.h:82:36:
error: inlining failed in call to 'always_inline' 'pas_page_base* bmalloc_heap_config_page_header(uintptr_t)':
function not considered for inlining
| 82 | PAS_BASIC_HEAP_CONFIG_DECLARATIONS(bmalloc, BMALLOC);
| |
Add a cmake variable WEBKIT_NO_INLINE_HINTS to disable gcc function attribute `always_inline` when debug build is enabled.
And adjust indent as well.
Signed-off-by: Jiaying Song <jiaying.song.cn@windriver.com>
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Building the recipe on x86 platform fails with the following error:
| ../../git/src/utils.c: In function 'get_uint64':
| ../../git/src/utils.c:105:18: error: passing argument 1 of 'str_toul' from incompatible pointer type [-Wincompatible-pointer-types]
| 105 | str_toul(&defval, p, NULL, 16);
Upstream has already changed this function to avoid overflow due to the
size difference in the pointer - this change backports that patch.
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
A fix was sent in:
https://lists.openembedded.org/g/openembedded-devel/message/116537
but it causes issues with clang as well as older gcc as reported on
ML and also in upstream PR:
https://github.com/openlink/iODBC/pull/111
Use gnu17 until this is properly resolved to avoid:
http://errors.yoctoproject.org/Errors/Details/852861/
In file included from ../../libiodbc-3.52.16/iodbc/execute.c:94:
../../libiodbc-3.52.16/iodbc/execute.c: In function '_ReBindParam':
../../libiodbc-3.52.16/iodbc/execute.c:643:49: error: too many arguments to function 'hproc3'; expected 0, have 8
643 | CALL_DRIVER (pstmt->hdbc, pstmt, retcode, hproc3,
| ^~~~~~
http://errors.yoctoproject.org/Errors/Details/853276/
../../libiodbc-3.52.16/iodbcinst/unicode.c: In function 'dm_AtoUW':
../../libiodbc-3.52.16/iodbcinst/unicode.c:1565:16: error: initialization of 'ucs4_t *' {aka 'unsigned int *'} from incompatible pointer type 'wchar_t *' {aka 'long int *'} [-Wincompatible-pointer-types]
1565 | ucs4_t *us = dest;
| ^~~~
Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
The recipe can't be compiled for 32-bit targets currently, due to
GTimeVal usage - see bug report in the project[1].
Remove the 32-bit targets from the compatibility list until it is
fixed.
[1]: https://gitlab.gnome.org/GNOME/libchamplain/-/issues/61
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
When compiling for 32-bit targets, compilation fails with the following error:
| ../libgdata-0.18.1/demos/calendar/calendar-cli.c:47:22: error: passing argument 1 of 'gmtime' from incompatible pointer type [-Wincompatible-pointer-types]
| 47 | tm = gmtime (&tv->tv_sec);
Upstream meanwhile has refactored the failing part in an untagged commit:
they have removed the usage of GTimeVal, since it has been deprecated.
Since it also solves the compilation issue, backport that patch.
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
When compiling for 32-bit targets, two issues came to surface:
1. gcc was complaining that math.h is not included in snprintf, and some
calls were implicitly defined. Added a patch that includes the required
headers in snprintf.c file:
| snprintf.c: In function 'fmtfloat':
| snprintf.c:1232:13: error: implicit declaration of function 'isnan' [-Wimplicit-function-declaration]
| 1232 | if (isnan(value))
| | ^~~~~
| snprintf.c:50:1: note: include '<math.h>' or provide a declaration of 'isnan'
2. The code passes a time_t argument to a function that expects a long. This works for
64-bit targets, because on those usually time_t is long.
However on 32-bit systems time_t is usually long long, which makes compilation fail
with the following error:
| wd_json_data.c:540:66: error: passing argument 3 of 'json_get_long_value_for_key' from incompatible pointer type [-Wincompatible-pointer-types]
| 540 | if (json_get_long_value_for_key(root, "StartupTimeSecs", &wdNode->startup_time.tv_sec))
| | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~
To account for this, add a new helper method in a patch that returns the required json value
as a time_t value.
The patches are in pending state, because the mailing list of the project is sufferring from
technical problems - when the site loads, sign up attempts throw internal server errors.
It is planned to submit the patches and to update the status once their infrastructure is back.
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
On 32-bit targets the compilation fails with the following error:
| ../../metacity-3.52.0/src/core/screen.c:2431:55: error: passing argument 2 of 'sn_startup_sequence_get_last_active_time' from incompatible pointer type [-Wincompatible-pointer-types]
| 2431 | sn_startup_sequence_get_last_active_time (sequence, &tv_sec, &tv_usec);
| | ^~~~~~~
| | |
| | long int *
Backport patch from upstream to address this problem.
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
The time_t type varies for 64bit and 32bit. So instead of using
long long, we should just use time_t. In this way, things build
for both 64bit and 32bit targets.
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Fix the following compile error:
../../sources/libwnck-2.31.0/libwnck/tasklist.c:4355:49: error: passing argument 2 of 'sn_startup_sequence_get_last_active_time' from incompatible pointer type [-Wincompatible-pointer-types]
4355 | &tv_sec, &tv_usec);
| ^~~~~~~
| |
| long int *
In file included from /build/tmp/work/core2-32-wrsmllib32-linux/lib32-libwnck/2.31.0/lib32-recipe-sysroot/usr/include/startup-notification-1.0/libsn/sn.h:32,
from ../../sources/libwnck-2.31.0/libwnck/private.h:36,
from ../../sources/libwnck-2.31.0/libwnck/tasklist.c:40:
/build/tmp/work/core2-32-wrsmllib32-linux/lib32-libwnck/2.31.0/lib32-recipe-sysroot/usr/include/startup-notification-1.0/libsn/sn-monitor.h:84:79: note: expected 'time_t *' {aka 'long long int *'} but argument is of type 'long int *'
84 | time_t *tv_sec,
| ~~~~~~~~~~~~~~~~~~~^~~~~~
Signed-off-by: Changqing Li <changqing.li@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
This fixes the following build issue:
wnck-3.so.0.3.0.p/pager.c.o.d -o libwnck/libwnck-3.so.0.3.0.p/pager.c.o -c ../libwnck-43.2/libwnck/pager.c
| In file included from ../libwnck-43.2/libwnck/pager.c:34:
| ../libwnck-43.2/libwnck/pager-accessible-factory.h:19: warning: header guard '__WNCK_PAGER_ACCESSIBLE_FACTORY_H__' followed by '#define' of a different macro [-Wheader-guard]
| 19 | #ifndef __WNCK_PAGER_ACCESSIBLE_FACTORY_H__
| ../libwnck-43.2/libwnck/pager-accessible-factory.h:20: note: '__WBCK_PAGER_ACCESSIBLE_FACTORY_H__' is defined here; did you mean '__WNCK_PAGER_ACCESSIBLE_FACTORY_H__'?
| 20 | #define __WBCK_PAGER_ACCESSIBLE_FACTORY_H__
| In file included from ../libwnck-43.2/libwnck/pager.c:35:
| ../libwnck-43.2/libwnck/workspace-accessible-factory.h:19: warning: header guard '__WNCK_WORKSPACE_ACCESSIBLE_FACTORY_H__' followed by '#define' of a different macro [-Wheader-guard]
| 19 | #ifndef __WNCK_WORKSPACE_ACCESSIBLE_FACTORY_H__
| ../libwnck-43.2/libwnck/workspace-accessible-factory.h:20: note: '__WBCK_WORKSPACE_ACCESSIBLE_FACTORY_H__' is defined here; did you mean '__WNCK_WORKSPACE_ACCESSIBLE_FACTORY_H__'?
| 20 | #define __WBCK_WORKSPACE_ACCESSIBLE_FACTORY_H__
And the subsequent build error caused by an OE-core patch to
startup-notification.
Signed-off-by: Koen Kooi <koen@dominion.thruhere.net>
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Without uinput.h header, the evdev module can't be imported because it
lacks the UI_FF* constants.
Signed-off-by: Yoann Congal <yoann.congal@smile.fr>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
* Remove a merged reproducibility patch
* Adapt the --reproducible option
Changelog (from [0]):
1.9.2 (May 01, 2025)
====================
- Add the "--reproducible" build option which removes the build date and used headers from the
generated ``ecodes.c``. Example usage::
python -m build --config-setting=--build-option='build_ecodes --reproducible' -n
- Use ``Generic`` to set precise type for ``InputDevice.path``.
[0]: https://github.com/gvalkov/python-evdev/blob/v1.9.2/docs/changelog.rst
Signed-off-by: Yoann Congal <yoann.congal@smile.fr>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
It was missing as the recipe is using --with-tirpc
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Adapted for Walnascar
Signed-off-by: Gyorgy Sarvari <gyorgy.sarvari@gmail.com>
The currently generated LibVNCServerTargets.cmake will include the
following 'set_target_properties':
set_target_properties(LibVNCServer::vncclient PROPERTIES
INTERFACE_INCLUDE_DIRECTORIES "${_IMPORT_PREFIX}/include"
INTERFACE_LINK_LIBRARIES "systemd;/usr/lib/libz.so;/usr/lib/liblzo2.so;/usr/lib/libjpeg.so;/usr/lib/libgcrypt.so;/usr/lib/libgnutls.so"
)
INTERFACE_LINK_LIBRARIES here points to absolute paths which hardcodes
the library paths. From CMake doc [1]:
Note that it is not advisable to populate the INTERFACE_LINK_LIBRARIES
of a target with absolute paths to dependencies. That would hard-code
into installed packages the library file paths for dependencies as
found on the machine the package was made on.
This breaks krfb build (kde desktop sharing server) since CMake cannot
find these libraries. Removing the absolute paths solves the issue.
Note: I also added a 'inherit pkgconfig' since libvncserver uses it to
detect libsystemd presence.
1: https://cmake.org/cmake/help/latest/prop_tgt/INTERFACE_LINK_LIBRARIES.html
Signed-off-by: Marc Ferland <marc.ferland@sonatest.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
(cherry picked from commit 2156942867)
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
This is required in order to build krdp (KDE's remote desktop
integration).
The do_configure task for krdp expects both 'winpr-makecert3' and
'winpr-hash3' to be present, failure to do so results in:
| CMake Error at /path/to/krdp/6.4.4/recipe-sysroot/usr/lib/cmake/WinPR3/WinPRTargets.cmake:98 (message):
| The imported target "winpr-makecert" references the file
|
| "/path/to/krdp/6.4.4/recipe-sysroot/usr/bin/winpr-makecert3"
|
| but this file does not exist. Possible reasons include:
|
| * The file was deleted, renamed, or moved to another location.
|
| * An install or uninstall procedure did not complete successfully.
|
| * The installation package was faulty and contained
|
| "/path/to/6.4.4/recipe-sysroot/usr/lib/cmake/WinPR3/WinPRTargets.cmake"
|
| but not all the files it references.
|
| Call Stack (most recent call first):
| /path/to/krdp/6.4.4/recipe-sysroot/usr/lib/cmake/WinPR3/WinPRConfig.cmake:44 (include)
| /path/to/krdp/6.4.4/recipe-sysroot-native/usr/share/cmake-3.31/Modules/CMakeFindDependencyMacro.cmake:76 (find_package)
| /path/to/krdp/6.4.4/recipe-sysroot/usr/lib/cmake/FreeRDP3/FreeRDPConfig.cmake:2 (find_dependency)
| CMakeLists.txt:45 (find_package)
|
|
| -- Configuring incomplete, errors occurred!
Signed-off-by: Marc Ferland <marc.ferland@sonatest.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
(cherry picked from commit 6c30f47645)
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
This is a small version bump. It includes to following two
commits to fix CVE-2025-4565.
05ba1a810 Add recursion depth limits to pure python
1ef3f01c4 Internal pure python fixes
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
When building an SDK with lcov included, gcov isn't included in the SDK
by default. Running lcov to generate coverage fails, because it tries to
use the gcov binary from the host system instead and that cause problems
if the gcc versions do not match.
Signed-off-by: Jef Driesen <jefdriesen@telenet.be>
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Liu Yiding
Gyorgy Sarvari
Archana Polampalli
Sunil Dora
Libo Chen
Ankur Tyagi
Changqing Li
Mikko Rapeli
Wang Mingyu
Praveen Kumar
Zoltán Böszörményi
Yoann Congal
Jiaying Song
Martin Jansa
Jason Schonberg
Chen Qi
Koen Kooi
Khem Raj
Hitendra Prajapati
Marc Ferland
Peter Marko
Jef Driesen
Yogita Urade