Upstream has closed the pull request due to submitter's inability to sign
the CLA, and the failure was seen once very long time ago on a AB worker
no longer in use.
Signed-off-by: Alexander Kanavin <alex@linutronix.de>
There's a couple of assumptions in devtool that sources
are always in a directory under WORKDIR; this is no longer
the case since introduction of UNPACKDIR: some recipes
are starting to use
S = "${UNPACKDIR}/git"
or similar, and so the logic to determine source locations
needs to be tweaked accordingly.
The issue is that oe-core has no concept of 'top level source
path' (S points to where the build is started from, inside that
top level location), and yet devtool needs to know that in order
to move the complete source tree correctly to a workspace (and possibly
other uses).
And so devtool performs convoluted path calculations from WORKDIR and S;
now this has been extended to include UNPACKDIR and became more convoluted
but hopefully it won't get any worse.
Signed-off-by: Alexander Kanavin <alex@linutronix.de>
This also requires using -fpermissive for gcc 14.x compatibility,
and dropping some of the existing patches (or their portions) which
are either parts of the 9.x port, or rendered obsolete by it.
Signed-off-by: Alexander Kanavin <alex@linutronix.de>
'devtool check-upgrade-status' floods the console with all of
the wget invocations. This is due to the behaviour of the standard
logger/progresshandler used by the wget fetcher which prints to stdout
everything wget executable outputs.
Let's use the logger only when downloading, and not when checking upstream versions.
Also, there is no need to pass the quiet option to runfetchcmd()
as runfetchcmd() has been ignoring that since 2012:
https://git.yoctoproject.org/poky/commit/?id=2b918b623bf8add454c3726dd4ee189230e6fcbc
The 'quiet' argument has since been superseded by the 'log' argument, and ideally
the quiet option should be removed altogether from the runfetchcmd() arguments,
but that would break the fetcher API.
Signed-off-by: Alexander Kanavin <alex@linutronix.de>
mesa build currently emits CLANG_RESOURCE_DIR into compiled objects and
meson calculates it from llvm cmake files from sysroot and it points to
absolute paths in target sysroot. To fix this backport a patch that does
not rely on CLANG_RESOURCE_DIR, however, this patch still leaves it in
code as fallback via FALLBACK_CLANG_RESOURCE_DIR, we are on LLVM 20.x
which will not use this variable, lets just remove detection so it does
not encode hardcoded paths.
Fixes
ERROR: mesa-2_25.0.2-r0 do_package_qa: QA Issue: File /usr/lib/libMesaOpenCL.so.1.0.0 in package libopencl-mesa contains reference to TMPDIR [buildpaths]
ERROR: mesa-2_25.0.2-r0 do_package_qa: Fatal QA errors were found, failing task.
(From OE-Core rev: afcde8eb575684fb514e1012b31bc0da04f4cb28)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Cc: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Brings these changes
- build both static and dynamic libraries on new suse distros
- require cmake >= 3.5
- new features:
* support the apk package and repository format (both v2 and v3)
* new dataiterator_final_{repo,solvable} functions
- Enable static libs for OE as well, this fixes a crash/signsegv seen
in stringpool_strn2id () especially when multithreading is in play,
this is brought forward with clang merge patchset when using ipk
backend
(From OE-Core rev: 50d5ac2199af7b35e18c0015fcc14078e4d1fd2a)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We are using clang-native to build the native versions of these recipes
but we are missing the needed compiler flags to let it c/c++ runtime
appropriately. This also ensures that meta-clang does not have to worry
about the compiler settings for nativesdk/native pieces separately.
(From OE-Core rev: b3b97ff790f42c5844c9f6aaeab9e23d355febfa)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Brings following fixes
* 923a5c4f83d2 Revert "[ARM][ConstantIslands] Correct MinNoSplitDisp calculation (#114590)"
* 86f5891c5986 [llvm][Hexagon] Promote operand v2i1 to v2i32 (#135409)
* d55c3c20520a [libc++] Fix misplaced _LIBCPP_POP_MACROS (#134874)
* 9420327ad768 [Clang] Fix a lambda pattern comparison mismatch after ecc7e6ce4 (#133863)
* 4da7285e636e Silence -Wcast-function-type warnings on idiomatic Windows code (#135660)
* c5109be53b7e [LV] Disable epilogue vectorization for FindLastIV if start is poison.
* 91a3f14d9497 [LV] Add tests with FindLastIV and epilogue vectorization.
* 2131242240f7 [LLVM][MemCpyOpt] Unify alias tags if we optimize allocas (#129537)
* 86c98536380b [libc++] Fix deployment targets that were incorrectly bumped (#134278)
* dfd6f123362a [libc++] Guard additional headers with _LIBCPP_HAS_LOCALIZATION (#131921)
* dc9d4f9a7008 [lldb] Respect LaunchInfo::SetExecutable in ProcessLauncherPosixFork (#133093)
* 9c7d72869876 [LoongArch] Move fix-tle-le-sym-type test to test/MC. NFC (#133839)
* 0c30835a63db [X86][AVX10] Remove VAES and VPCLMULQDQ feature from AVX10.1 (#135489)
* 2e0966408283 [X86] Backport saturate-convert intrinsics renaming & YMM rounding intrinsics removal in AVX10.2
* 0dd4235473d4 [SCEV] Use ashr to adjust constant multipliers (#135534)
* a141e58685fd [llvm][CodeGen] avoid repeated interval calculation in window scheduler (#132352)
* d88cd35023b4 [llvm][CodeGen] Fix the empty interval issue in Window Scheduler (#129204)
* 73d1e8598eda [CodeGen] Avoid repeated hash lookups (NFC) (#130237)
* 7034995f1029 [clang] Handle Binary StingLiteral kind in one more place (#132201)
* 2e7710eaffdd [clang] Introduce "binary" StringLiteral for #embed data (#127629)
* e0db588f3db4 [IR] Fix assertion error in User new/delete edge case (#129914)
* d5bb7b866e59 Avoid a race condition in opt-viewer/optrecord (#131214)
* d15fef4209f1 [IndVarSimplify] Handle the case where both operands are the same when widening IV (#135207)
* 91647ae0dffe [X86][SSE] Don't emit SSE2 load instructions in SSE1-only mode (#134547)
* d05543ed0796 [clang-format] Keep the space between `not` and a unary operator (#135035)
* 81220e68a496 [fatlto] Add coroutine passes when using FatLTO with ThinLTO (#134434)
* edb54a7821fe Release/20.x: [clang-format] Set C11 instead of C17 for LK_C
* 4181e829d1db [LLDB][LoongArch] Fix build errors about NT_LOONGARCH_HW_{BREAK,WATCH} (#126020)
* 7436329bfee9 Revert "[clang] [ARM] Explicitly enable NEON for Windows/Darwin targets (#122095)"
* a0c8959cc880 [X86] When expanding LCMPXCHG16B_SAVE_RBX, substitute RBX in base (#134109)
* a8b5fe017a5e [libc++] Add missing release note for LLVM 20 about zip_view (#134144)
* 41aefdbebe64 cuda clang: Fix argument order for __reduce_max_sync (#132881)
* 19c2e1c12d47 [clang-tidy] Fix broken HeaderFilterRegex when read from config file (#133582)
* ac31db0463c0 [Sanitizers][Darwin][Test] XFAIL malloc_zone.cpp
* 53141e4e3c65 [clang] Do not infer lifetimebound for functions with void return type (#131997)
* cf7bb13f0c7f [TailDuplicator] Determine if computed gotos using `blockaddress` (#132536)
* 656289ffa0a6 Bump version to 20.1.3 (#134187)
(From OE-Core rev: 2199334354162a6304c4b0d4f35efb67e8e7d36b)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This is needed for clang to find the c++ headers when invoked via
bindgen
Fixes
| /mnt/b/yoe/master/sources/poky/build/tmp/work/cortexa57-poky-linux/mesa/25.0.2/recipe-sysroot/usr/include/llvm/ADT/DenseMapInfo.h:17:10: fatal error: 'cassert' file not found
| Unable to generate bindings: clang diagnosed error: /mnt/b/yoe/master/sources/poky/build/tmp/work/cortexa57-poky-linux/mesa/25.0.2/recipe-sysroot/usr/include/llvm/ADT/DenseMapInfo.h:17:10: fatal error: 'cassert' file not found
|
(From OE-Core rev: 6d9f95a25dd32871e8c07ec3295a216236b3a7e0)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Cc: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This paves the way to spin out these packages out of clang in future
specifically spirv-llvm-translator which is not yet part of llvm monorepo
(From OE-Core rev: fc58a1ee75737851d6e97be8a0b7ac739819efc8)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This can ensure that meta-clang does not carry them and can rely on core layer
while mesa can use them too.
Omit time stamps in openmp from generated files to improve
reproducibility
fix the issue that:
| file /usr/include/llvm/Config/llvm-config.h conflicts between attempted installs of lib32-llvm-dev-20.1.2-r0.core2_32 and llvm-dev-20.1.2-r0.core2_64
(From OE-Core rev: d77e398095228b34851762858a76640e3c2cb0ab)
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Overview of Changes in 4.18.5, 03-05-2025
=========================================
* Bugs fixed:
- #5844 `gdk_surface_get_device_position: assertion 'GDK_IS_SURFACE (surface)'
failed` on popover close (Alessandro Astone)
- #7451 Cairo device scale changes lead to significant performance regression
in cairo blur (Sebastian Keller)
- #7479 gtkmm programs fail to start on Windows with AccessKit support enabled
(Matthias Clasen)
- #7482 Epiphany Crash in GTK file chooser when last used directory has
insufficient filesystem permissions: "GFileInfo created without
standard::type" (Philip Withnall)
- #7495 Failed to link with the mold linker (Matthias Clasen)
- #7513 Broken boxed-list ListBox (Sergey Bugaev)
- #7520 XCompose preventing GTK4 applications from launching (again) (Matthias
Clasen)
- #7522 critical when selecting label with use-underline but mnemonic with
screenreader enabled (Markus Göllnitz)
- !8432 GTK 4.18.4
- !8461 Some cherry-picked fixes for 4.18 (Sebastian Keller, Benjamin Otte,
Florian "sp1rit")
- !8487 early exit if the clipped bound rectangle's area is zero after
snapping (Benjamin Otte)
- !8496 macos: fix crash when unfullscreening windows (4.18 backport) (Arjan
Molenaar)
- !8505 Backport: update macOS job for new CI runner (René de Hesselle)
- !8526 build: pin the msvc build to meson 1.7 (Matthias Clasen)
- !8527 build: pin the msvc build to meson 1.7 (Alessandro Astone, Sébastien
Wilmet, Sergey Bugaev, Vladimir Vaskov, Markus Göllnitz, Michael Weghorn,
Philip Withnall, Lukáš Tyrychtr, Benjamin Otte, Matthias Clasen, Liu
Jinchang, kolunmi, FeRD (Frank Dana))
* Translation updates:
- Bulgarian (Alexander Shopov)
- Catalan (Jordi Mas i Hernandez)
- Chinese (China) (Luming Zh)
- Czech (Daniel Rusek)
- Georgian (Ekaterine Papava)
- Hungarian (Balázs Úr)
- Lithuanian (Aurimas Černius)
- Persian (Danial Behzadi)
- Polish (Piotr Drąg)
- Portuguese (Hugo Carvalho)
- Portuguese (Brazil) (Álvaro Burns)
- Romanian (Antonio Marin)
- Russian (Artur S0)
- Slovenian (Martin)
- Spanish (Daniel Mustieles)
- Swedish (Anders Jonsson)
- Turkish (Emin Tufan Çetin)
- Ukrainian (Yuri Chornoivan)
(From OE-Core rev: 96c0dc8f5c3a4a270f040a3cbcbfe20b132c1afc)
Signed-off-by: Markus Volk <f_l_k@t-online.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
- remove backport patch
Overview of changes in GLib 2.84.1, 2025-04-03
==============================================
* Fix test failure when building against gobject-introspection ≥1.83.4 (#3634,
work by Philip Withnall)
* Bugs fixed:
- #3630 2.84.0 build failure on Linux: ../gio/gnetworkmonitornetlink.c:47:10:
fatal error: netlink/netlink_route.h: No such file or directory (Philip
Withnall)
- #3634 test failure with gobject-introspection 1.83.4: warning: element
doc:format from state 3 is unknown, ignoring (Philip Withnall)
- #3636 gio/trash does not handle special characters well
- #3642 `g_cancellable_connect()` documentation incorrect (Marco Trevisan
(Treviño))
- #3643 g_cancellable_connect(): is it safe to unref cancellable from
callback? (Marco Trevisan (Treviño))
- #3649 Crash with some registry key values in GWin32AppInfo (Philip Withnall)
- !4484 Memory sanitizer fixes
- !4489 gobject: Be consistent in using atomic logic to handle the
GParamSpecPool
- !4541 gsettings: Port docs to gi-docgen format, add missing annotations and
make various improvements
- !4544 tests: Don't install runner scripts without installed_tests
- !4545 Update French translation
- !4547 Update Catalan translation
- !4548 Update Turkish translation
- !4551 Updated Danish translation
- !4552 Update Persian translation
- !4553 docs: Document GSignalFlags members added after 2.0
- !4554 Update Indonesian translation
- !4555 tests: Add a test for g_object_freeze_notify() being called too often
- !4557 gfileinfo: Slightly expand docs for
g_file_info_get_attribute_as_string()
- !4558 gi: Dynamically set doc-format
- !4561 tests: Various fixes to create temporary files in /tmp rather than the
build directory
- !4562 gdbusnameowning: Convert docs to gi-docgen linking syntax
- !4563 giounix-private: Fix macro for checking for epoll_create1()
- !4565 Fix LGPL in header
- !4567 gutils: make documentation of g_set_prgname() clearer
- !4568 docs: Add some detail
- !4569 Update Romanian translation
- !4570 gspawn-win32: Fix potential integer overflows in argv handling
- !4571 gvarianttype: Improve docs on type validation
* Translation updates:
- Catalan (Jordi Mas)
- Danish (Ask Hjorth Larsen)
- French (Vincent Chatelain)
- Indonesian (Andika Triwidada)
- Persian (Danial Behzadi)
- Romanian (Antonio Marin)
- Turkish (Sabri Ünal)
(From OE-Core rev: 676b9acbe94f055a351da3bdcfbe457411e1877c)
Signed-off-by: Markus Volk <f_l_k@t-online.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We had this CVE patched but the patch was removed with last 2.44 branch
updates as it is now included.
Since there is no new version which could be set in NVD DB, this needs
to be explicitly handled.
(From OE-Core rev: 32f18145dee54f61203506daef339cd132908287)
Signed-off-by: Peter Marko <peter.marko@siemens.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
NVD has CVE-2023-25584 listed as < 2.40, so we don't need to ignore it
for version 2.44 anymore.
(From OE-Core rev: eaf80096f96e5bebed53076c1dfe7e35e539f383)
Signed-off-by: Peter Marko <peter.marko@siemens.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
In version 10, emulation of 64 bit targets on 32 bit
systems is no longer supported. The target list is tweaked accordingly.
acce728cbc
Running qemu on *any* 32 bit host is also deprecated and going to be
removed upstream in a later version soon:
6d701c9bac
The pixman configuration option now needs to be specified so add a
PACKAGECONFIG entry, defautling to enabled. This is a required dependency
of some of the graphical backends like gtk.
(From OE-Core rev: 811bf03dc511b3dbe30f6c65112a5eadd6a88f81)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Add a new test that checks that no tasks between do_fetch (exclusive)
and do_build (inclusive) are allowed to use the network, with rare
exceptions.
The only exception currently is build-appliance-image's do_image task,
as that currently usese pip to install the required Toaster dependencies.
Note that this will mean layers that have Go-based recipes will fail
unless they're using the gomod fetcher and have a complete list of
modules in the SRC_URI.
(From OE-Core rev: e95b3bd194e294412bc0419c9c74abfc2f37406f)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Rearrange the class so that the check is a separate function, to make it
neater to add further tests in the future.
(From OE-Core rev: 36658a113811de98249bc1e1b79cfadd405e5391)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* Nios2 has been removed and aarch64/ilp32 is deprecated
* Default C dialect is switched to C23
* {0} initializer in C or C++ for unions no longer guarantees clearing of the whole union
* Compile speed improvements with LTO
* Vectorizer can support loops with early exists but it is limited to loops
with fixed vector lengths
This is major release of gcc, the changes are noted [1]
[1] https://gcc.gnu.org/gcc-15/changes.html
(From OE-Core rev: 0ddda88b928a8b4e5ca3a4cc6112b8331cdff544)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
They do not pass with GCC-15, ignore them for now
[RP: Fix formatting of exclude to be whitespace separated]
(From OE-Core rev: dc989569d3528d999f2a1f4363a9ef8325e09600)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Includes a fix for the library version to match the tagged version in git.
(From OE-Core rev: 3ba4b22ef7e50e017d25ba974666f2fdf190a8fd)
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
GCC-15 has _CHKFEAT_GCS defined in arm_acle.h to indicate gcs intrinsics
support, this trips llvm libunwind gcs feature detection logic to set
gcs feature on. However the contructs used in unwindlib are assuming clang
and the needed target attribute is not available in gcc it should be +gcs
to work with both clang and gcc
(From OE-Core rev: aa5b42b32093fdd10e0a8f8a124322610c84f60b)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
CVE-2025-32911:
A use-after-free type vulnerability was found in libsoup, in the
soup_message_headers_get_content_disposition() function. This flaw
allows a malicious HTTP client to cause memory corruption in the libsoup
server.
Backport patches to fix it
[1] https://nvd.nist.gov/vuln/detail/CVE-2025-32911
[2] https://gitlab.gnome.org/GNOME/libsoup/-/issues/433
(From OE-Core rev: 839d93bbb1ca7a51b659b8cb9def9b354a99518f)
Signed-off-by: Changqing Li <changqing.li@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The current MULTILIB_SCRIPTS entry incorrectly assigns
g-ir-annotation-tool and g-ir-scanner to the ${PN} package. However,
these scripts are actually installed in the gobject-introspection-tools
package. This leads to incorrect symlinks being created when multilib is
enabled. Therefore, this patch correctly moves them to the ${PN}-tools
package.
Error log:
root@intel-x86-64:~# ls -l /usr/bin/g-ir-scanner
lrwxrwxrwx 1 root root 27 Mar 9 2018 /usr/bin/g-ir-scanner -> /usr/bin/g-ir-scanner-lib64
root@intel-x86-64:~# ls -l /usr/bin/g-ir-scanner-lib64
ls: cannot access '/usr/bin/g-ir-scanner-lib64': No such file or directory
(From OE-Core rev: 94aa798624dd3e1d0bfeccc046936a4a4f87b8ca)
Signed-off-by: Jiaying Song <jiaying.song.cn@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
All qemu machines have ext4 in IMAGE_FSTYPES which makes
native mkfs.ext4 available at wic image build time.
genericarm64 defaults to only wic in IMAGE_FSTYPES
and this test was failing:
| DEBUG: Python function extend_recipe_sysroot finished
| DEBUG: Executing python function set_image_size
| DEBUG: 148548.400000 = 114268 * 1.300000
| DEBUG: 652644.400000 = max(148548.400000, 8192)[148548.400000] + 504096
| DEBUG: 652645.000000 = int(652644.400000)
| DEBUG: 652645 = aligned(652645)
| DEBUG: returning 652645
| DEBUG: Python function set_image_size finished
| DEBUG: Executing shell function do_image_wic
| INFO: Creating image(s)...
|
| ERROR: A native program mkfs.ext4 required to build the image was not found (see details above).
|
| Please make sure wic-tools have e2fsprogs-native in its DEPENDS, build it with 'bitbake wic-tools' and try again.
|
| WARNING: exit code 1 from a shell command.
NOTE: recipe core-image-minimal-1.0-r0: task do_image_wic: Failed
(From OE-Core rev: 59b5ec3d53058237dc6f3e85385851d1a8a97f97)
Signed-off-by: Mikko Rapeli <mikko.rapeli@linaro.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This test is failing on genericarm64 which includes all kernel
modules on rootfs by default. Recently added kernel modules
triggered the size limits to be exceeded. Fixes:
ERROR: Actual rootfs size (112884 kB) is larger than allowed size 102400 kB
Doubling the size to avoid issues like this on machines which
may include a lot of packages on default images. Size of the
rootfs does not matter for the test, only that the offsets are
correct in various usecases.
(From OE-Core rev: 1ff542318ea66514ef550d4817ba9b2ef688134c)
Signed-off-by: Mikko Rapeli <mikko.rapeli@linaro.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The wic test_exclude_path_with_extra_space does not succeed on debian-based
AB workers since they, by default, do not install the 'parted' utility. This
test installs the 'wic-tools' package, which includes the 'parted' utility,
but it is not being found/used for some unknown reason.
In the previous patch it was believed that doing some extra PATH handling
(as performed by some other tests with a python try...finally block) would
solve the issue. That turned out to not be the case. This patch starts by
reverting that change, since it has no benefit.
In order to use the native tools from the 'wic-tools' package, wic's
'--native-sysroot' can be used (as demonstrated in other tests). In fact
the wic help message explaining the purpose of this flag states that it is
for passing "...the path to the native sysroot containing the tools(parted
and mtools) to use"[sic].
Removing the host's 'parted' utility better simulates the situation found
on the debian workers for testing. This patch is seen to make this test pass
with no host 'parted' utility in place.
(From OE-Core rev: 306e4831f07d4a8db6ab880487813a6abd0ffa08)
Signed-off-by: Trevor Woerner <twoerner@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Declare riscv64 as a COMPATIBLE_HOST since upstream now support it.
For that target 'valgrind /bin/ls' warns with:
--253-- WARNING: unhandled riscv64-linux syscall: 258
--253-- You may be able to write your own handler.
--253-- Read the file README_MISSING_SYSCALL_OR_IOCTL.
--253-- Nevertheless we consider this a bug. Please report
--253-- it at http://valgrind.org/support/bug_reports.html.
This is a known issue:
https://bugs.kde.org/show_bug.cgi?id=503253
but the tool does work, despite this warning with glibc.
For riscv64 with musl, the build is fine but running:
# valgrind /bin/ls
produces the fatal error:
==306== Process terminating with default action of signal 4 (SIGILL)
==306== Illegal opcode at address 0x1002015952
==306== at 0x40E5F26: map_library (dynlink.c:845)
==306== by 0x40E6C97: load_library (dynlink.c:1183)
==306== by 0x40E7B63: load_preload (dynlink.c:1365)
==306== by 0x40E7B63: __dls3 (dynlink.c:1963)
==306== by 0x40E75D9: __dls2 (dynlink.c:1767)
==306== by 0xFFFFFFFFFFFFFFFF: ???
which is not yet reported.
Patches reviewed, manually fix-up one word change in:
0001-none-tests-fdleak_cmsg.stderr.exp-adjust-tmp-paths.patch
and refresh patches with devtool.
* ==================== CORE CHANGES ===================
* The valgrind gdbserver now supports the GDB remote protocol packet
'x addr,len' (available in GDB release >= 16).
The x packet can reduce the time taken by GDB to read memory from valgrind.
* Valgrind now supports zstd compressed debug sections.
* The Linux Test Project (ltp) is integrated in the testsuite try
'make ltpchecks' (this will take a while and will point out various
missing syscalls and valgrind crashes!)
* ================== PLATFORM CHANGES =================
* Added RISCV64 support for Linux. Specifically for the RV64GC
instruction set.
* Numerous bug fixes for Illumos, in particular fixed a Valgrind crash
whenever a signal handler was called.
* On FreeBSD, a change to the libc code that runs atexit handlers was
causing Helgrind to produce an extra error about exiting threads
still holding locks for. This applied to every multithreaded application.
The extra error is now filtered out. A syscall wrapper had been added
for getrlimitusage.
* On Linux various new syscalls are supported (landlock*, io_pgetevents,
open_tree, move_mount, fsopen, fsconfig, fsmount, fspick, userfaultfd).
* s390x has support for various new instructions (BPP, BPRP, PPA and NIAI).
* ==================== TOOL CHANGES ===================
* The --track-fds=yes and --track-fds=all options now treat all
inherited file descriptors the same as 0, 1, 2 (stdin/out/err).
And when the stdin/out/err descriptors are reassigned they are
now treated as normal (non-inherited) file descriptors.
* A new option --modify-fds=high can be used together with
--track-fds=yes to create new file descriptors with the highest
possible number (and then decreasing) instead of always using the
lowest possible number (which is required by POSIX). This will help
catch issues where a file descriptor number might normally be reused
between a close and another open call.
* Helgrind:
There is a change to warnings about calls to pthread_cond_signal and
pthread_cond_broadcast when the associated mutex is unlocked. Previously
Helgrind would always warn about this. Now this error is controlled by
a command line option, --check-cond-signal-mutex=yes|no. The default is
no. This change has been made because some C and C++ standard libraries
use pthread_cond_signal/pthread_cond_broadcast in this way. Users are
obliged to use suppressions if they wish to avoid this noise.
The full 3.25.0 bug fix list can be found here:
https://valgrind.org/docs/manual/dist.news.html
(From OE-Core rev: 04b3e0c1384469167ddfb4d22e2b513f9a59aca9)
Signed-off-by: Randy MacLeod <Randy.MacLeod@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Debug level was not added as a parameter, causing a warning.
(From OE-Core rev: 40157fcbd9066f261812ba665ec963b2e496aa53)
Signed-off-by: Daniel Turull <daniel.turull@ericsson.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
ChangeLog:
https://github.com/makedumpfile/makedumpfile/releases/tag/1.7.7
This version supports the following new kernels:
- 6.12, 6.13, 6.14
Drop backport patch.
(From OE-Core rev: 364e15aadcbb3ac754c2c484f7920420b4cb3138)
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>
- New argument android_exe_type for executables
- Changes to the b_sanitize option
- New C standard c2y (and gnu2y)
- i18n module xgettext
- version_compare now accept multiple compare strings
- Improvements to Objective-C and Objective-C++
- Per project subproject options rewrite
- objects added correctly to Rust executables
- rust.test now supports link_whole
- Meson can run "rustdoc" on Rust projects
- The Wayland module is stable
- New swift_std compiler option
- New option to execute a slice of tests
- Valgrind now fails tests if errors are found
Full release notes: https://mesonbuild.com/Release-notes-for-1-8-0.html.
(From OE-Core rev: 84ded53ea7e4db1e3d7ea60527374040ec64b4fd)
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>
libbfd is enabled and it invokes libtool relink step to get rid of
rpaths during install step. libtool 2.4 patches rightly prepend
lt_sysroot to target libdir which relink step would use, such a patch
is also upstreamed into libtool [1]
However, when we do not specify --with-libtool-sysroot, lt_sysoot prefix
turns out to be empty and relink step adds -L/usr/lib to linker command
line. it cause all sort of problems e.g. when the build host has zlib
installed then libbfd tries to link with /usr/lib/libz.so instead of the
one in target sysroot, fortunately OE linker reports host path contamination
and fails the build. Second problem seen when building target binutils on
musl, it links with libc.so correctly but libtool relink step resolves
libc.so in /usr/lib due to this path and expands it into libc.so.6 and
ld-linux-x86-64.so.2 ( glibc host's libc.so is a linker stub ) in
DT_NEEDED ELF section which does not work with
musl linker and build QA detects this anamoly as well.
ERROR: binutils-2.44-r0 do_package_qa: QA Issue: /usr/lib/libbfd-2.44.so contained in package libbfd requires libc.so.6(GLIBC_2.34)(64bit), but no providers found in RDEPENDS:libbfd? [file-rdeps]
This inconsistency may cause produced binaries mistakenly linked against
libraries from the host distro.
Add libtool sysroot to point libtool to correct sysroot location during
build.
[1] https://git.savannah.gnu.org/cgit/libtool.git/commit/?id=3221f9f0fb98d5740ab5d0e8db6a731302520644
(From OE-Core rev: 89e984bd7edb9a2013b65d0215fd6af9aadd59e8)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This recipe takes longer time >20min when bitbake for package
write stage. When cross-verified for longer time duration, found
that do_check() stage taking 20min while other stages completes
before 6min.
This recipe gives only below two test binaries in the packages to
test (ptest: glibc-y2038-tests):
io/ftwtest
io/ftwtest-time64
The above test binaries are already included for testing in recipe
glibc-testsuite_2.41.bb.
It is by now well established that glibc itself works as it should,
that all affected 32 bit targets are configured to use 64 bit time_t,
and that any lingering y2038 issues are in components other than the c
library, and usually come from C programming mistakes (e.g. storing
timestamps in long). Maybe we can simply remove the recipe?
Review comments for fixing above longer time duration ended up in
removing this recipe as a proposal is below
https://lists.openembedded.org/g/openembedded-core/topic/112188476#msg214636
Removed lines having reference to glibc-y2038-tests in the files.
(From OE-Core rev: fbe3679ba3c12c52a502511f5dde91fb4de7a6b6)
Signed-off-by: rajmohan r <semc.2042@gmail.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Move the functions in qemu.bbclass to meta/lib/oe/qemu.py as they are
generally useful.
The qemu.bbclass is still kept, and recipes can continue to use functions
from it, though they have become wrapper functions on qemu.py functions.
Note that the QEMU_OPTIONS settings are still kept in qemu.bbclass.
This sets a clear barrier for people to use qemu user mode.
(From OE-Core rev: 7b3563b3b3901c96c3e498799a83ab8cabcf84b4)
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>
Fix to ensure it would work with both libsdl2 and libsdl2-compat
(From OE-Core rev: 21033e98045d03025f6e12dba16428fa484b3057)
Signed-off-by: Markus Volk <f_l_k@t-online.de>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We have long since obsoleted md5sum in favour of sha256sum. Drop the remaining
56 entries (which were showing many recipes hadn't been touched in a long time).
They all do have the corresponding sha256sum entries as is clear from the diff.
(From OE-Core rev: 7e4bfcc9706fa8a09f6a0004174a2c3b21c90df3)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
I previously upgrade autoconf 2.72e (the fifth release candidate) to the
final release 2.72, but unfortunately failed to notice that the version
appears to go backwards as the package managers sort 2.72e _after_ 2.72.
We could resolve this with an epoch (via PE) but epochs are forever and
this mistake is temporary. We could set a convoluted PV but that gets
messy as it then means SRC_URI needs to set manually. Instead, set a
convoluted PKGV to fix the sorting order in a way that only alters the
output packages and nothing else, which can be removed when the next
major release is made.
[1] oe-core 11ff8dba0ce ("autoconf: upgrade to final 2.72 release")
(From OE-Core rev: 58b185b25dff627eecc980919dd44dcce32fbca2)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Fix the numpy ptests by doing the following:
- Add meson to ptest RDEPENDS in the recipe;
- Add python3-unittest-automake-output as a ptest RDEPENDS;
- Convert run-ptest to a shell script that sets PYTEST_DEBUG_TEMPROOT to
a directory inside the same path that contains the script, create that
directory, and then invoke the tests with `pytest --automake` pointed
at the numpy path in site-packages (copying the tests into the normal
PTEST_DIRECTORY seems to cause module import breakages in some tests).
This also includes skipping two problematic tests which require a C
compiler and use up a lot of space, respectively;
- Set 'IMAGE_ROOTFS_EXTRA_SPACE = "3048576"' for python3-numpy in
core-image-ptest.bb;
- Also set 'QB_MEM:virtclass-mcextend-python3-numpy = "-m 4096"' in
core-image-ptest.bb;
- Move python3-numpy from the PTESTS_PROBLEMS list to the PTESTS_SLOW
one.
Results on qemux86-64:
Testsuite summary
DURATION: 87
END: /usr/lib/python3-numpy/ptest
2025-04-29T17:35
STOP: ptest-runner
TOTAL: 1 FAIL: 0
Note that many of the skipped tests are due to the absence of
python3-mypy (which is currently in meta-python) and of a Fortran
compiler.
(From OE-Core rev: 7f0b5e8faa1b246531ac425c99a629eb344b21e2)
Signed-off-by: Trevor Gamblin <tgamblin@baylibre.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
If extra-utils package is not included in the image, you'll always see a
warning that password agent is missing whenever you start/stop a service:
Failed to execute /usr/bin/systemd-tty-ask-password-agent: No such file or directory
(From OE-Core rev: a87d523ab24e4ea87d1b19ea3a0c515cc47db1b6)
Signed-off-by: Vyacheslav Yurkov <uvv.mail@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
