linux/poky.git/meta/conf, branch 5.3_M2

libclc: split out of clang

2025-07-21T22:00:18+00:00

Split the libclc subproject out of the clang recipe and into a dedicated libclc recipe. This is useful because libclc is the OpenCL runtime library and as such isn't target-specific and needs a native clang to build, not a target libllvm. Verified that nothing is dropped by adding clang and libclc to an image and verifying that the file list is the same before and after this change. We need to patch the libclc CMakeLists to allow it to use an out-of-tree prepare_builtins binary, discussion is ongoing with upstream to resolve this properly. (From OE-Core rev: 33a8742a1280b4c6779a7aa487c2dd4a713babe6) Signed-off-by: Ross Burton Signed-off-by: Richard Purdie

tunes: set valid clang tune for big.LITTLE Arm SoCs

2025-07-21T22:00:18+00:00

clang doesn't have Arm big.LITTLE specific tune options - when such an option is used, the compilation fails with an error like this: aarch64-poky-linux-clang: error: unsupported argument 'cortex-a72.cortex-a53+crc+crypto' to option '-mcpu=' To avoid this, in case a big.LITTLE SoC is the target and the toolchain is clang, select the tune corresponding to the LITTLE core. (From OE-Core rev: 12c820f624730a04a17ae2cdbecc4e6987c45cf7) Signed-off-by: Gyorgy Sarvari Signed-off-by: Mathieu Dubois-Briand Signed-off-by: Richard Purdie

icu: Add ptest support

2025-07-21T22:00:18+00:00

Add following patches. - 0001-test-Add-support-ptest.patch - Some default paths in test code are invalid at runtime and cause - test failures. So add a patch to adjust path to test data for ptest - environment. - 0001-ICU-23120-Mask-UnicodeStringTest-TestLargeMemory-on-.patch - Since ICU-77.1, a test case (TestLargeMemory) that fails to build - in a 32-bit environment. So add a patch to skip this test case. - This bug has been reported to upstream. See https://unicode-org.atlassian.net/browse/ICU-23120. Install icu test-suite to run it as a ptest. Add icu to PTESTS_FAST because it takes 27sec (less than 30sec) to complete on qemux86-64 with kvm enabled. root@qemux86-64:~# ptest-runner icu START: ptest-runner 2025-07-06T00:46 BEGIN: /usr/lib/icu/ptest ___(snip)___ -------------------------------------- Elapsed Time: 00:00:23.070 PASS: ./intltest DURATION: 27 END: /usr/lib/icu/ptest 2025-07-06T00:47 STOP: ptest-runner TOTAL: 1 FAIL: 0 (From OE-Core rev: 4a729a529067a5ba7036a224cf330e31b8a5f838) Signed-off-by: Daisuke Yamane CC: Alexander Kanavin Signed-off-by: Antonin Godard Signed-off-by: Richard Purdie

clang: Upgrade to 20.1.8 release

2025-07-17T09:41:18+00:00

Brings following fixes * 87f0227cb601 [InstCombine] Avoid folding `select(umin(X, Y), X)` with min/max values in false arm (#143020) * df43f93388b7 [PhaseOrdering] Add test for #139050 (NFC) * 25bcf1145fd7 [RISCV] Fix assertion failure when using -fstack-clash-protection (#135248) * 6fb913d3e2ec [RelLookupTableConverter] Drop unnamed_addr for GVs in entries to avoid generating GOTPCREL relocations (#146068) * 0c9f909b7976 [AArch64][SME] Fix restoring callee-saves from FP with hazard padding (#143371) * fa792cd4c630 [AsmPrinter] Always emit global equivalents if there is non-global uses (#145648) * ce455b382c08 [objcopy][MachO] Revert special handling of encryptable binaries (#144058) * 0de59a293f7a [X86] Ignore NSW when DstSVT is i32 (#131755) * 9af763f038f7 [gtest] Fix building on OpenBSD/sparc64 (#145225) * 1daceb20611f [LoongArch] Pass OptLevel to LoongArchDAGToDAGISel correctly * b21155f97a0a [LoongArch] Precommit test case to show bug in LoongArchISelDagToDag * da18fb9f04ce [LoongArch] Fix xvshuf instructions lowering (#145868) * 65ce78f338cf [LoongArch] Pre-commit test for fixing xvshuf instructions. NFC * 5532d5b745e4 [AArch64] Ensure the LR is preserved if we must call __arm_get_current_vg (#145760) * 5ac3ce819688 [WebAssembly] Fix inline assembly with vector types (#146574) * b83658b7e2c8 Bump version to 20.1.8 (From OE-Core rev: 4c6e132d6e5e49426c911d0fa9215957eb43f186) Signed-off-by: Khem Raj Signed-off-by: Mathieu Dubois-Briand Signed-off-by: Richard Purdie

dosfstools: Add ptests

2025-07-17T09:41:18+00:00

(From OE-Core rev: 875b8961221875e6a809d15b7d3b83ea00da0c0e) Signed-off-by: Ricardo Simoes Signed-off-by: Mark Jonas Signed-off-by: Mathieu Dubois-Briand Signed-off-by: Richard Purdie

bitbake.conf: Remove PR glob part from STAMPCLEAN

2025-07-14T16:51:42+00:00

Since OE-Core rev: cc83e45484656a6b577ff84817131735023daad4 the STAMP value and STAMPCLEAN glob have been mismatched. The issue is present since the PR part was removed from the STAMP variable in that comit. An example use case that I found was broken due to this: 1. Have recipes foo_A.bb and foo_B.bb 2. Build foo-native with PREFERRED_VERSION_foo-native = "A" 3. ${COMPONENTS_DIR}/x86_64-linux/foo-native has version A 4. Build foo-native with PREFERRED_VERSION_foo-native = "B" 5. ${COMPONENTS_DIR}/x86_64-linux/foo-native has version B 6. Build foo-native with PREFERRED_VERSION_foo-native = "A" 7. ${COMPONENTS_DIR}/x86_64-linux/foo-native still has version B In my case the PREFERRED_VERSION comes from different machines. The issue showed itself when a bar-native compiled against foo-native version A was pulled from sstate-cache and foo-native in version B was kept in ${COMPONENTS_DIR} after previous build for a different machine. The two variables should be in sync and this patch corrects that. [RP: Tweak commit message] (From OE-Core rev: 932be19f48735d72a72de2771911119433956f4f) Signed-off-by: Michal Sieron Signed-off-by: Richard Purdie

lldb: add new recipe, split out of clang

2025-07-14T16:49:48+00:00

LLDB takes a reasonable amount of time to compile, but also isn't an essential component of the LLVM suite. Instead of always building it when we build clang, split it out into a separate recipe. On my build machine where clang takes 21 minutes to build with lldb, it takes 19 minutes without lldb. (From OE-Core rev: 9da4900aa5a37718bd42f277d5a1805ec897b1b4) Signed-off-by: Ross Burton Signed-off-by: Richard Purdie

llvm-tblgen-native: add new recipe

2025-07-14T16:49:48+00:00

Building the LLVM projects often means using the TableGen tools (llvm-tblgen etc). We currently build them as part of clang-native, but I am teasing the clang recipe into its component parts and having to build llvm-native or lldb-native simply for one tool isn't ideal. Instead, add a native recipe that simply builds the tablegen binaries for llvm, clang, and lldb (From OE-Core rev: fbf63e03fe09ca74022c9d06442b4f1021b71d57) Signed-off-by: Richard Purdie

bitbake.conf: fix pseudo for devtool deploy-target --strip

2025-07-10T09:47:31+00:00

Without this fix: devtool deploy-target cmake-example qemu1 --strip ... cp: failed to preserve ownership for '.../cmake-example/1.0/devtool-deploy-target-stripped/usr/lib/libcmake-example-lib.so.1.0.0': Operation not permitted cp: failed to preserve ownership for ...cmake-example/1.0/devtool-deploy-target-stripped/usr/lib/libcmake-example-lib.so.1: Operation not permitted cp: failed to preserve ownership for ...cmake-example/1.0/devtool-deploy-target-stripped/usr/lib/libcmake-example-lib.so: Operation not permitted cp: failed to preserve ownership for '.../cmake-example/1.0/devtool-deploy-target-stripped/usr/lib': Operation not permitted cp: failed to preserve ownership for '.../cmake-example/1.0/devtool-deploy-target-stripped/usr/bin/cmake-example': Operation not permitted cp: failed to preserve ownership for '.../cmake-example/1.0/devtool-deploy-target-stripped/usr/bin/test-cmake-example': Operation not permitted cp: failed to preserve ownership for '.../cmake-example/1.0/devtool-deploy-target-stripped/usr/bin': Operation not permitted cp: failed to preserve ownership for '.../cmake-example/1.0/devtool-deploy-target-stripped/usr': Operation not permitted cp: failed to preserve ownership for '.../cmake-example/1.0/devtool-deploy-target-stripped': Operation not permitted tar: ./usr/lib/libcmake-example-lib.so.1.0.0: time stamp 2025-07-06 16:46:06 is 0.527890738 s in the future tar: ./usr/lib/libcmake-example-lib.so.1: time stamp 2025-07-06 16:46:06 is 0.527462566 s in the future tar: ./usr/lib/libcmake-example-lib.so: time stamp 2025-07-06 16:46:06 is 0.526732779 s in the future tar: ./usr/lib: time stamp 2025-07-06 16:46:06 is 0.526415655 s in the future tar: ./usr/bin/cmake-example: time stamp 2025-07-06 16:46:06 is 0.52568721 s in the future tar: ./usr/bin/test-cmake-example: time stamp 2025-07-06 16:46:06 is 0.525054415 s in the future tar: ./usr/bin: time stamp 2025-07-06 16:46:06 is 0.524821739 s in the future INFO: Successfully deployed .../cmake-example/1.0/devtool-deploy-target-stripped With this fix: devtool deploy-target cmake-example qemu1 --strip ... INFO: Successfully deployed .../cmake-example/1.0/devtool-deploy-target-stripped This is most probably related to the commit: 2502da81709f25de499277b28d33c915638c45f6 bitbake.conf/pseudo: Switch from exclusion list to inclusion list (From OE-Core rev: 8b42b698ea13e1951f7bac4ddd89e13c9b0cb3a6) Signed-off-by: Adrian Freihofer Signed-off-by: Mathieu Dubois-Briand Signed-off-by: Richard Purdie

spdx30: Allow VEX Justification to be configurable

2025-07-03T09:40:17+00:00

Instead of hard coding the VEX justifications for "Ignored" CVE status, add a map that configures what justification should be used for each status. This allows other justifications to be easily added, and also ensures that status fields added externally (by downstream) can set an appropriate justification if necessary. (From OE-Core rev: c0fa3d92cefa74fa57c6c48c94acc64aa454e781) Signed-off-by: Joshua Watt Signed-off-by: Richard Purdie