linaro-toolchain

linaro-toolchain@lists.linaro.org

6 participants
5813 discussions

Re: [Linaro-TCWG-CI] gcc-17-3926-g93ac5163b5d: 3 regressions on arm

by Richard Earnshaw (foss)

On 05/09/2026 09:06, ci_notify(a)linaro.org wrote: > Dear contributor, > > Our automatic CI has detected problems related to your patch(es). Please find some details below. > > In armv8l-unknown-linux-gnueabihf native, after: > | commit gcc-17-3926-g93ac5163b5d > | Author: Dominic P <gcc(a)gcc.dp11.uk> > | Date: Sun Aug 2 11:56:05 2026 +0100 > | > | arm: fix vshll_n intrinsics with boundary shift [PR111609] > | > | The NEON widening-shift-left-long intrinsics (vshll_n_s8/s16/s32 and > | vshll_n_u8/u16/u32) accept a shift amount in the range [0, element_size]. > | The bounds check in neon_vshll<sup>_n<mode> already permits a shift of > | ... 32 lines of the commit log omitted. > > Produces 3 regressions: > | > | regressions.sum: > | Running gcc:gcc.target/arm/arm.exp ... > | FAIL: gcc.target/arm/vfp-shift-a2t2.c scan-assembler vshll.u16[ \t]+[qQ][0-9]+, [dD][0-9]+, #[0-9]+!?([ \t]+@[a-zA-Z0-9 ]+)?\n > | FAIL: gcc.target/arm/vfp-shift-a2t2.c scan-assembler vshll.u32[ \t]+[qQ][0-9]+, [dD][0-9]+, #[0-9]+!?([ \t]+@[a-zA-Z0-9 ]+)?\n > | FAIL: gcc.target/arm/vfp-shift-a2t2.c scan-assembler vshll.u8[ \t]+[qQ][0-9]+, [dD][0-9]+, #[0-9]+!?([ \t]+@[a-zA-Z0-9 ]+)?\n > Fixed. R. > Used configuration : > *CI config* tcwg_gcc_check armv8l-unknown-linux-gnueabihf > *configure and test flags:* none, autodetected on armv8l-unknown-linux-gnueabihf--disable-multilib --with-float=hard --with-mode=thumb --with-tune=cortex-a9 --with-arch=armv7-a --with-fpu=neon-fp-armv8 > > We track this bug report under https://linaro.atlassian.net/browse/GNU-2020. (Require Linaro login) Please let us know if you have a fix. > > If you have any questions regarding this report, please ask on linaro-toolchain(a)lists.linaro.org mailing list. > > -----------------8<--------------------------8<--------------------------8<-------------------------- > > The information below contains the details of the failures, and the ways to reproduce a debug environment: > > You can find the failure logs in *.log.1.xz files in > * http://54.172.246.49:9090/jobs/tcwg_gcc_check--master-arm-build/builds/1062… > The full lists of regressions and improvements as well as configure and make commands are in > * http://54.172.246.49:9090/jobs/tcwg_gcc_check--master-arm-build/builds/1062… > The list of [ignored] baseline and flaky failures are in > * http://54.172.246.49:9090/jobs/tcwg_gcc_check--master-arm-build/builds/1062… > > Current build : http://54.172.246.49:9090/jobs/tcwg_gcc_check--master-arm-build/builds/1062… > Reference build : http://54.172.246.49:9090/jobs/tcwg_gcc_check--master-arm-build/builds/1062… > > Instruction to reproduce the build : https://gitlab.com/LinaroLtd/tcwg/ci/interesting-commits/-/raw/master/gcc/s… > > Full commit : See in git+ssh://linaroci@gcc.gnu.org/git/gcc.git

1 month, 2 weeks

Re: [Linaro-TCWG-CI] llvmorg-24-init-6625-gedbf358d0149: 4 regressions on aarch64

by Newcombe, Caroline

Thanks for your report. I think you’d need to add 0640_0214, 0640_0215, 0640_0223 and 0640_0224 to Fujitsu/Fortran/0640/fpconfig.cmake with TEST_SUITE_FP_TOLERANCE_FP32. Alternatively, applying -fno-reciprocal-math to these tests should also take care of the discrepancy in results. All four print REAL*4 values but are currently graded against the directory default FP_TOLERANCE, which is derived from the FP64 constants, so 1.0e-14 with TEST_SUITE_FUJITSU_WITH_FAST_MATH=ON and 0x1.0p-51 without. Testing on aarch64 with the fast-math configuration, I see a single REAL*4 value change in each of them, from .9 to .90000003. That is one ULP of binary32 arising from reciprocal substitution on 9.0/10.0: the correctly rounded quotient is 0x3F666666, or 0.899999976, while 9.0f * (1.0f/10.0f) rounds to 0x3F666667, or 0.900000036. The relative difference is 6.6e-8, roughly six orders of magnitude above the tolerance being applied, so the comparison fails even though the result is within one ULP. Some of the outputs also show REAL*8 values shifting by one ULP, for instance .6 to .6000000000000001, but those are around 1.8e-16 and pass comfortably. Since -freciprocal-math is implied by -ffast-math, which this configuration uses, the substitution is permitted and I don't think the new value should be treated as incorrect. TEST_SUITE_FP_TOLERANCE_FP32 is 2.38e-7, which covers the new test results. It appears the tests have been passing just because the reciprocal substitution happened not to fire. My recent change to Flang enables the FIR loop invariant code motion pass by default. The pass perturbs the IR enough that the cost model's reciprocal decision flips, resulting in the slight difference to reference output. Best, Caroline From: ci_notify(a)linaro.org <ci_notify(a)linaro.org> Date: Thursday, September 3, 2026 at 5:25 PM To: contact-ci-compiler-test-suite(a)cs.jp.fujitsu.com <contact-ci-compiler-test-suite(a)cs.jp.fujitsu.com> Cc: Newcombe, Caroline <caroline.newcombe(a)hpe.com>; maxim.kuvyrkov(a)linaro.org <maxim.kuvyrkov(a)linaro.org> Subject: [Linaro-TCWG-CI] llvmorg-24-init-6625-gedbf358d0149: 4 regressions on aarch64 Dear contributor, Our automatic CI has detected problems related to your patch(es). Please find some details below. In tcwg_flang_test/main-aarch64-Ofast-sve_vla, after: | commit llvmorg-24-init-6625-gedbf358d0149 | Author: Caroline Newcombe <caroline.newcombe(a)hpe.com> | Date: Mon Aug 31 11:20:03 2026 -0400 | | [flang] Enable FIR loop invariant code motion by default (#218703) | | The FIR loop invariant code motion pass was added in #173438, where it | was disabled by default. This enables it for optimization levels above | `-O0`. The option `-disable-fir-licm` has been added in place of | ... 5 lines of the commit log omitted. Produces 4 regressions: | | regressions.sum: | Running test-suite:Fujitsu/Fortran/0640 ... | FAIL: test-suite :: Fujitsu/Fortran/0640/Fujitsu-Fortran-0640_0214.test | FAIL: test-suite :: Fujitsu/Fortran/0640/Fujitsu-Fortran-0640_0215.test | FAIL: test-suite :: Fujitsu/Fortran/0640/Fujitsu-Fortran-0640_0223.test | FAIL: test-suite :: Fujitsu/Fortran/0640/Fujitsu-Fortran-0640_0224.test | # "FAIL" means : the execution of the compiled binary failed / output of the binary differs from the expected one Used configuration : * Toolchain : cmake -G Ninja ../llvm/llvm "-DLLVM_ENABLE_PROJECTS=clang;lld;flang;clang-tools-extra" "-DLLVM_ENABLE_RUNTIMES=openmp" -DCMAKE_BUILD_TYPE=Release -DLLVM_ENABLE_ASSERTIONS=True -DCMAKE_INSTALL_PREFIX=../llvm-install "-DLLVM_TARGETS_TO_BUILD=AArch64" -DCLANG_DEFAULT_LINKER=lld * Testsuite : export LD_LIBRARY_PATH=$\WORKSPACE/llvm-install/lib/aarch64-unknown-linux-gnu$\{LD_LIBRARY_PATH:+:$\LD_LIBRARY_PATH} cmake -GNinja -DCMAKE_C_COMPILER="$\WORKSPACE/llvm-install/bin2/clang" -DCMAKE_CXX_COMPILER="$\WORKSPACE/llvm-install/bin2/clang++" -DCMAKE_Fortran_COMPILER="$\WORKSPACE/llvm-install/bin2/flang-new" -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_FLAGS= -DCMAKE_CXX_FLAGS= -DCMAKE_Fortran_FLAGS= -DCMAKE_C_FLAGS_RELEASE="-O3 -ffast-math -march=armv8.4-a+sve -msve-vector-bits=scalable -mllvm -scalable-vectorization=preferred -DNDEBUG" -DCMAKE_CXX_FLAGS_RELEASE="-O3 -ffast-math -march=armv8.4-a+sve -msve-vector-bits=scalable -mllvm -scalable-vectorization=preferred -DNDEBUG" -DCMAKE_Fortran_FLAGS_RELEASE="-O3 -ffast-math -march=armv8.4-a+sve -msve-vector-bits=scalable -mllvm -scalable-vectorization=preferred -DNDEBUG" -DTEST_SUITE_FORTRAN=ON -DTEST_SUITE_SUBDIRS=Fujitsu -DTEST_SUITE_FUJITSU_WITH_FAST_MATH=ON "$\WORKSPACE/test/test-suite" We track this bug report under https://urldefense.com/v3/__https://linaro.atlassian.net/browse/LLVM-2272__… . (Require Linaro login) Please let us know if you have a fix. If you have any questions regarding this report, please ask on linaro-toolchain(a)lists.linaro.org mailing list. -----------------8<--------------------------8<--------------------------8<-------------------------- The information below contains the details of the failures, and the ways to reproduce a debug environment: You can find the failure logs in *.log.1.xz files in * https://urldefense.com/v3/__http://54.172.246.49:9090/jobs/tcwg_flang_test-… The full lists of regressions and improvements as well as configure and make commands are in * https://urldefense.com/v3/__http://54.172.246.49:9090/jobs/tcwg_flang_test-… The list of [ignored] baseline and flaky failures are in * https://urldefense.com/v3/__http://54.172.246.49:9090/jobs/tcwg_flang_test-… Fujitsu testsuite : https://github.com/fujitsu/compiler-test-suite/ Current build : https://urldefense.com/v3/__http://54.172.246.49:9090/jobs/tcwg_flang_test-… Reference build : https://urldefense.com/v3/__http://54.172.246.49:9090/jobs/tcwg_flang_test-… Instruction to reproduce the build : https://urldefense.com/v3/__https://gitlab.com/LinaroLtd/tcwg/ci/interestin… Full commit : https://github.com/llvm/llvm-project/commit/edbf358d014942a17953c3ccfcf9f20…

1 month, 2 weeks

Re: [Linaro-TCWG-CI] gcc-17-3835-g8858cb3a6af5: Failure on arm

by Ian Lance Taylor

Should be fixed by GCC commit 6401cc2efae456806f3376810098d99e518577fe. Ian On Thu, Sep 3, 2026 at 1:30 AM <ci_notify(a)linaro.org> wrote: > > Dear contributor, > > Our automatic CI has detected problems related to your patch(es). Please find some details below. > > In bootstrap_build master-arm-bootstrap_ubsan, after: > | commit gcc-17-3835-g8858cb3a6af5 > | Author: Ian Lance Taylor <iant(a)golang.org> > | Date: Tue Sep 1 15:08:41 2026 -0700 > | > | libbacktrace: handle MOREDATA correctly without debug info > | > | Based on patch from Andi Kleen. > | > | * elf.c (elf_nodebug): Handle state->moredata. > | ... 14 lines of the commit log omitted. > > Produces Failure: > | Results changed to > | # reset_artifacts: > | -10 > | # true: > | 0 > | # build_abe bootstrap_ubsan: > | # FAILED > | # First few build errors in logs: > | # 00:52:21 collect2: error: ld returned 1 exit status > | # 00:52:21 make[3]: *** [/home/tcwg-buildslave/workspace/tcwg_gnu_0/abe/snapshots/gcc.git~master/gcc/c/Make-lang.in:88: cc1] Error 1 > | # 00:52:21 make[2]: *** [Makefile:5747: all-stage2-gcc] Error 2 > | # 00:52:21 make[1]: *** [Makefile:27721: stage2-bubble] Error 2 > | # 00:52:21 make: *** [Makefile:1140: all] Error 2 > | > | From > | # reset_artifacts: > | -10 > | # true: > | 0 > | # build_abe bootstrap_ubsan: > | 1 > > Used configuration : > *CI config* tcwg_bootstrap_build master-arm-bootstrap_ubsan > *configure and test flags:* none, autodetected on armv8l-unknown-linux-gnueabihf > > We track this bug report under https://linaro.atlassian.net/browse/GNU-2019. (Require Linaro login) Please let us know if you have a fix. > > If you have any questions regarding this report, please ask on linaro-toolchain(a)lists.linaro.org mailing list. > > -----------------8<--------------------------8<--------------------------8<-------------------------- > > The information below contains the details of the failures, and the ways to reproduce a debug environment: > > You can find the failure logs in > * http://54.172.246.49:9090/jobs/tcwg_bootstrap_build--master-arm-bootstrap_u… > The full lists of regressions and improvements as well as configure and make commands are in > * http://54.172.246.49:9090/jobs/tcwg_bootstrap_build--master-arm-bootstrap_u… > > Current build : http://54.172.246.49:9090/jobs/tcwg_bootstrap_build--master-arm-bootstrap_u… > Reference build : http://54.172.246.49:9090/jobs/tcwg_bootstrap_build--master-arm-bootstrap_u… > > Instruction to reproduce the build : https://gitlab.com/LinaroLtd/tcwg/ci/interesting-commits/-/raw/master/gcc/s… > > Full commit : See in git+ssh://linaroci@gcc.gnu.org/git/gcc.git

1 month, 3 weeks

Re: [Linaro-TCWG-CI] llvmorg-24-init-6563-ge448a895252d: 1 regression on aarch64

by Yasuyuki Ohno (Fujitsu)

Hi, We investigated the failure and confirmed that it is caused by an overly strict floating-point tolerance in the test. After the LoopInterchange-related optimization changes, the order of some floating-point operations differs, producing small rounding differences typical of single-precision arithmetic. The observed relative error is approximately 2.8e-7, which is reasonable for REAL*4 calculations. The test currently uses a tolerance equivalent to double-precision accuracy, so we plan to update the test tolerance to an appropriate single-precision value (1.0e-6). Please ignore the mail. We'll update the test. We apologize for the false detection. Best regards, Yasuyuki Ohno Software Development Div., Advanced Technology Development Unit, Fujitsu Research, Fujitsu Limited, JAPAN. ________________________________ 差出人: ci_notify(a)linaro.org <ci_notify(a)linaro.org> 送信: 2026 年 9 月 1 日 (火曜日) 20:11 宛先: contact-ci-compiler-test-suite(a)cs.jp.fujitsu.com <contact-ci-compiler-test-suite(a)cs.jp.fujitsu.com> Cc: maxim.kuvyrkov(a)linaro.org <maxim.kuvyrkov(a)linaro.org>; madhura(a)nvidia.com <madhura(a)nvidia.com> 件名: [Linaro-TCWG-CI] llvmorg-24-init-6563-ge448a895252d: 1 regression on aarch64 Dear contributor, Our automatic CI has detected problems related to your patch(es). Please find some details below. In tcwg_flang_test/main-aarch64-Ofast-sve_vla, after: | commit llvmorg-24-init-6563-ge448a895252d | Author: Madhur Amilkanthwar <madhura(a)nvidia.com> | Date: Mon Aug 31 09:56:48 2026 +0530 | | [Clang][Flang] Enable loop interchange by default at >O1 opt levels (#216920) | | PR #124911 enabled the LoopInterchange pass by default in the LLVM | optimization pipeline (the PipelineTuningOptions default used by opt), | but the clang and flang frontends override that default and keep it off. | ... 15 lines of the commit log omitted. Produces 1 regression: | | regressions.sum: | Running test-suite:Fujitsu/Fortran/0344 ... | FAIL: test-suite :: Fujitsu/Fortran/0344/Fujitsu-Fortran-0344_0317.test | # "FAIL" means : the execution of the compiled binary failed / output of the binary differs from the expected one Used configuration : * Toolchain : cmake -G Ninja ../llvm/llvm "-DLLVM_ENABLE_PROJECTS=clang;lld;flang;clang-tools-extra" "-DLLVM_ENABLE_RUNTIMES=openmp" -DCMAKE_BUILD_TYPE=Release -DLLVM_ENABLE_ASSERTIONS=True -DCMAKE_INSTALL_PREFIX=../llvm-install "-DLLVM_TARGETS_TO_BUILD=AArch64" -DCLANG_DEFAULT_LINKER=lld * Testsuite : export LD_LIBRARY_PATH=$\WORKSPACE/llvm-install/lib/aarch64-unknown-linux-gnu$\{LD_LIBRARY_PATH:+:$\LD_LIBRARY_PATH} cmake -GNinja -DCMAKE_C_COMPILER="$\WORKSPACE/llvm-install/bin2/clang" -DCMAKE_CXX_COMPILER="$\WORKSPACE/llvm-install/bin2/clang++" -DCMAKE_Fortran_COMPILER="$\WORKSPACE/llvm-install/bin2/flang-new" -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_FLAGS= -DCMAKE_CXX_FLAGS= -DCMAKE_Fortran_FLAGS= -DCMAKE_C_FLAGS_RELEASE="-O3 -ffast-math -march=armv8.4-a+sve -msve-vector-bits=scalable -mllvm -scalable-vectorization=preferred -DNDEBUG" -DCMAKE_CXX_FLAGS_RELEASE="-O3 -ffast-math -march=armv8.4-a+sve -msve-vector-bits=scalable -mllvm -scalable-vectorization=preferred -DNDEBUG" -DCMAKE_Fortran_FLAGS_RELEASE="-O3 -ffast-math -march=armv8.4-a+sve -msve-vector-bits=scalable -mllvm -scalable-vectorization=preferred -DNDEBUG" -DTEST_SUITE_FORTRAN=ON -DTEST_SUITE_SUBDIRS=Fujitsu -DTEST_SUITE_FUJITSU_WITH_FAST_MATH=ON "$\WORKSPACE/test/test-suite" We track this bug report under https://linaro.atlassian.net/browse/LLVM-2269. (Require Linaro login) Please let us know if you have a fix. If you have any questions regarding this report, please ask on linaro-toolchain(a)lists.linaro.org mailing list. -----------------8<--------------------------8<--------------------------8<-------------------------- The information below contains the details of the failures, and the ways to reproduce a debug environment: You can find the failure logs in *.log.1.xz files in * http://54.172.246.49:9090/jobs/tcwg_flang_test--main-aarch64-Ofast-sve_vla-… The full lists of regressions and improvements as well as configure and make commands are in * http://54.172.246.49:9090/jobs/tcwg_flang_test--main-aarch64-Ofast-sve_vla-… The list of [ignored] baseline and flaky failures are in * http://54.172.246.49:9090/jobs/tcwg_flang_test--main-aarch64-Ofast-sve_vla-… Fujitsu testsuite : https://github.com/fujitsu/compiler-test-suite/ Current build : http://54.172.246.49:9090/jobs/tcwg_flang_test--main-aarch64-Ofast-sve_vla-… Reference build : http://54.172.246.49:9090/jobs/tcwg_flang_test--main-aarch64-Ofast-sve_vla-… Instruction to reproduce the build : https://gitlab.com/LinaroLtd/tcwg/ci/interesting-commits/-/raw/master/llvm/… Full commit : https://github.com/llvm/llvm-project/commit/e448a895252d5dd74f85450a2b1438e…

1 month, 3 weeks

Re: [Linaro-TCWG-CI] 2025-09-68-g2d608e140: 1 regression on aarch64

by Yasuyuki Ohno (Fujitsu)

Hi, This is the test-suite issue rather than the compiler defect. The source code contains an inconsistent declaration for src2, and the failure became visible after removal of the option that suppressed incompatible pointer type diagnostics. Please ignore the mail. We'll update the test. We apologize for the false detection. Best regards, Yasuyuki Ohno Software Development Div., Advanced Technology Development Unit, Fujitsu Research, Fujitsu Limited, JAPAN. ________________________________ 差出人: ci_notify(a)linaro.org <ci_notify(a)linaro.org> 送信日時: 2026年9月2日 11:26 宛先: contact-ci-compiler-test-suite(a)cs.jp.fujitsu.com <contact-ci-compiler-test-suite(a)cs.jp.fujitsu.com> CC: 114733434+ohno-fj(a)users.noreply.github.com <114733434+ohno-fj(a)users.noreply.github.com>; maxim.kuvyrkov(a)linaro.org <maxim.kuvyrkov(a)linaro.org> 件名: [Linaro-TCWG-CI] 2025-09-68-g2d608e140: 1 regression on aarch64 Dear contributor, Our automatic CI has detected problems related to your patch(es). Please find some details below. In tcwg_flang_test/main-aarch64-O0-debug, after: | commit 2025-09-68-g2d608e140 | Author: ohno-fj <114733434+ohno-fj(a)users.noreply.github.com> | Date: Tue Sep 1 08:01:57 2026 +0900 | | Workitem 16933_0002 (for GCC) (#131) | | Co-authored-by: ohno <ohno.yasuyuki(a)jp.fujitsu.com> Produces 1 regression: | | regressions.sum: | Running test-suite:Fujitsu/C/0002 ... | NOEXE: test-suite :: Fujitsu/C/0002/Fujitsu-C-0002_0050.test | # "NOEXE" means : the test program cannot be compiled Used configuration : * Toolchain : cmake -G Ninja ../llvm/llvm "-DLLVM_ENABLE_PROJECTS=clang;lld;flang;clang-tools-extra" "-DLLVM_ENABLE_RUNTIMES=openmp" -DCMAKE_BUILD_TYPE=Release -DLLVM_ENABLE_ASSERTIONS=True -DCMAKE_INSTALL_PREFIX=../llvm-install "-DLLVM_TARGETS_TO_BUILD=AArch64" -DCLANG_DEFAULT_LINKER=lld * Testsuite : export LD_LIBRARY_PATH=$\WORKSPACE/llvm-install/lib/aarch64-unknown-linux-gnu$\{LD_LIBRARY_PATH:+:$\LD_LIBRARY_PATH} cmake -GNinja -DCMAKE_C_COMPILER="$\WORKSPACE/llvm-install/bin2/clang" -DCMAKE_CXX_COMPILER="$\WORKSPACE/llvm-install/bin2/clang++" -DCMAKE_Fortran_COMPILER="$\WORKSPACE/llvm-install/bin2/flang-new" -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_FLAGS= -DCMAKE_CXX_FLAGS= -DCMAKE_Fortran_FLAGS= -DCMAKE_C_FLAGS_RELEASE="-O0 -g" -DCMAKE_CXX_FLAGS_RELEASE="-O0 -g" -DCMAKE_Fortran_FLAGS_RELEASE="-O0 -g" -DTEST_SUITE_FORTRAN=ON -DTEST_SUITE_SUBDIRS=Fujitsu "$\WORKSPACE/test/test-suite" We track this bug report under https://linaro.atlassian.net/browse/LLVM-2271. (Require Linaro login) Please let us know if you have a fix. If you have any questions regarding this report, please ask on linaro-toolchain(a)lists.linaro.org mailing list. -----------------8<--------------------------8<--------------------------8<-------------------------- The information below contains the details of the failures, and the ways to reproduce a debug environment: You can find the failure logs in *.log.1.xz files in * http://54.172.246.49:9090/jobs/tcwg_flang_test--main-aarch64-O0-debug-build… The full lists of regressions and improvements as well as configure and make commands are in * http://54.172.246.49:9090/jobs/tcwg_flang_test--main-aarch64-O0-debug-build… The list of [ignored] baseline and flaky failures are in * http://54.172.246.49:9090/jobs/tcwg_flang_test--main-aarch64-O0-debug-build… Fujitsu testsuite : https://github.com/fujitsu/compiler-test-suite/ Current build : http://54.172.246.49:9090/jobs/tcwg_flang_test--main-aarch64-O0-debug-build… Reference build : http://54.172.246.49:9090/jobs/tcwg_flang_test--main-aarch64-O0-debug-build… Instruction to reproduce the build : https://gitlab.com/LinaroLtd/tcwg/ci/interesting-commits/-/raw/master/fujit… Full commit : https://github.com/fujitsu/compiler-test-suite/commit/2d608e140687b2b0de363…

1 month, 3 weeks

[Linaro-TCWG-CI] glibc-2.44.9000-140-ged0c137b97e: 1 regression on arm

by ci_notify＠linaro.org

Dear contributor, Our automatic CI has detected problems related to your patch(es). Please find some details below. In glibc_check master-arm, after: | commit glibc-2.44.9000-140-ged0c137b97e | Author: Adhemerval Zanella <adhemerval.zanella(a)linaro.org> | Date: Mon Aug 17 15:15:32 2026 -0300 | | elf: Open the normalized $\ORIGIN rpath in AT_SECURE programs (BZ 34360) | | For AT_SECURE programs the loader honors $\ORIGIN in DT_RPATH only when the | expansion is rooted in a trusted directory, but it validated the lexically | normalized path while opening the raw expansion. As "a/b/../c" only names | ... 13 lines of the commit log omitted. Produces 1 regression: | | regressions.sum: | Running glibc:elf ... | FAIL: elf/tst-origin-secure Used configuration : *CI config* tcwg_glibc_check master-arm *configure and test flags:* none, autodetected on armv8l-unknown-linux-gnueabihf We track this bug report under https://linaro.atlassian.net/browse/GNU-2016. (Require Linaro login) Please let us know if you have a fix. If you have any questions regarding this report, please ask on linaro-toolchain(a)lists.linaro.org mailing list. -----------------8<--------------------------8<--------------------------8<-------------------------- The information below contains the details of the failures, and the ways to reproduce a debug environment: You can find the failure logs in *.log.1.xz files in * http://54.172.246.49:9090/jobs/tcwg_glibc_check--master-arm-build/builds/10… The full lists of regressions and improvements as well as configure and make commands are in * http://54.172.246.49:9090/jobs/tcwg_glibc_check--master-arm-build/builds/10… The list of [ignored] baseline and flaky failures are in * http://54.172.246.49:9090/jobs/tcwg_glibc_check--master-arm-build/builds/10… Current build : http://54.172.246.49:9090/jobs/tcwg_glibc_check--master-arm-build/builds/10… Reference build : http://54.172.246.49:9090/jobs/tcwg_glibc_check--master-arm-build/builds/10… Instruction to reproduce the build : https://gitlab.com/LinaroLtd/tcwg/ci/interesting-commits/-/raw/master/glibc… Full commit : https://sourceware.org/git/?p=glibc.git;a=commitdiff;h=ed0c137b97eb940b4b64…

1 month, 3 weeks

[Linaro-TCWG-CI] glibc-2.44.9000-126-gd179e02998c: 1 regression on aarch64

by ci_notify＠linaro.org

Dear contributor, Our automatic CI has detected problems related to your patch(es). Please find some details below. In aarch64 native, after: | commit glibc-2.44.9000-126-gd179e02998c | Author: Adhemerval Zanella <adhemerval.zanella(a)linaro.org> | Date: Tue Jul 28 18:13:22 2026 -0300 | | elf: Allow RPATH/RUNPATH for static-pie (BZ 33326) | | The initial static-pie support (commit 9d7a3741c9e59eba87fb) reused | ld.so ELF parsing logic, even though RPATH/RUNPATH should not appear | in the static-pie bootstrap. With static PIE, RPATH/RUNPATH on the | ... 15 lines of the commit log omitted. Produces 1 regression: | | regressions.sum: | Running glibc:elf ... | FAIL: elf/tst-pie-rpath-static Used configuration : *CI config* tcwg_gnu_native_check_glibc aarch64-linux-gnu *configure and test flags:* none, autodetected on aarch64-unknown-linux-gnu--disable-multilib --enable-fix-cortex-a53-835769 --enable-fix-cortex-a53-843419 --with-arch=armv8-a We track this bug report under https://linaro.atlassian.net/browse/GNU-2015. (Require Linaro login) Please let us know if you have a fix. If you have any questions regarding this report, please ask on linaro-toolchain(a)lists.linaro.org mailing list. -----------------8<--------------------------8<--------------------------8<-------------------------- The information below contains the details of the failures, and the ways to reproduce a debug environment: You can find the failure logs in *.log.1.xz files in * http://54.172.246.49:9090/jobs/tcwg_gnu_native_check_glibc--master-aarch64-… The full lists of regressions and improvements as well as configure and make commands are in * http://54.172.246.49:9090/jobs/tcwg_gnu_native_check_glibc--master-aarch64-… The list of [ignored] baseline and flaky failures are in * http://54.172.246.49:9090/jobs/tcwg_gnu_native_check_glibc--master-aarch64-… Current build : http://54.172.246.49:9090/jobs/tcwg_gnu_native_check_glibc--master-aarch64-… Reference build : http://54.172.246.49:9090/jobs/tcwg_gnu_native_check_glibc--master-aarch64-… Instruction to reproduce the build : https://gitlab.com/LinaroLtd/tcwg/ci/interesting-commits/-/raw/master/glibc… Full commit : https://sourceware.org/git/?p=glibc.git;a=commitdiff;h=d179e02998ce31c93436…

1 month, 3 weeks

Re: [Linaro-TCWG-CI] llvmorg-24-init-6091-g44ea65c07a60: 3 regressions on aarch64

by Yasuyuki Ohno (Fujitsu)

Hi, The failures were caused by re-applying the LLVM change introduced by commit(efd053f8ef306e6fa21ab0cd1bd185ef48f09fc0). Although the change had previously been reverted due to regression failures, it was later re-applied unchanged by commit(44ea65c07a600ca1c32912b5cd7356f6792b2b03). This reintroduced the same failures that had already been reported in LLVM-2266. The issue has already been addressed as part of the resolution for LLVM-2266, and the corresponding fix was committed in: https://dev.azure.com/miratech-soft/b6c3f009-db3a-422a-91be-d5c0809bb862/_g… Please ignore the mail. We apologize for the false detection. Best regards, Yasuyuki Ohno Software Development Div., Advanced Technology Development Unit, Fujitsu Research, Fujitsu Limited, JAPAN. ________________________________ 差出人: ci_notify(a)linaro.org <ci_notify(a)linaro.org> 送信日時: 2026年8月28日 1:43 宛先: contact-ci-compiler-test-suite(a)cs.jp.fujitsu.com <contact-ci-compiler-test-suite(a)cs.jp.fujitsu.com> CC: rupprecht(a)google.com <rupprecht(a)google.com>; maxim.kuvyrkov(a)linaro.org <maxim.kuvyrkov(a)linaro.org> 件名: [Linaro-TCWG-CI] llvmorg-24-init-6091-g44ea65c07a60: 3 regressions on aarch64 Dear contributor, Our automatic CI has detected problems related to your patch(es). Please find some details below. In tcwg_flang_test/main-aarch64-Ofast-sve_vls-lto-lld, after: | commit llvmorg-24-init-6091-g44ea65c07a60 | Author: Jordan Rupprecht <rupprecht(a)google.com> | Date: Wed Aug 26 13:02:20 2026 -0500 | | Re-apply "[PromoteMemToReg] Insert store undef when removing lifetime markers" (#218935) | | Reapplies #191909 / reverts #218804 | | Re-apply as-is. The issues were downstream and have been since fixed. Produces 3 regressions: | | regressions.sum: | Running test-suite:Fujitsu/C++/0041 ... | FAIL: test-suite :: Fujitsu/C++/0041/Fujitsu-C++-0041_0335.test | FAIL: test-suite :: Fujitsu/C++/0041/Fujitsu-C++-0041_0337.test | Running test-suite:Fujitsu/C/0096 ... | FAIL: test-suite :: Fujitsu/C/0096/Fujitsu-C-0096_0358.test | # "FAIL" means : the execution of the compiled binary failed / output of the binary differs from the expected one Used configuration : * Toolchain : cmake -G Ninja ../llvm/llvm "-DLLVM_ENABLE_PROJECTS=clang;lld;flang;clang-tools-extra" "-DLLVM_ENABLE_RUNTIMES=openmp" -DCMAKE_BUILD_TYPE=Release -DLLVM_ENABLE_ASSERTIONS=True -DCMAKE_INSTALL_PREFIX=../llvm-install "-DLLVM_TARGETS_TO_BUILD=AArch64" -DCLANG_DEFAULT_LINKER=lld * Testsuite : export LD_LIBRARY_PATH=$\WORKSPACE/llvm-install/lib/aarch64-unknown-linux-gnu$\{LD_LIBRARY_PATH:+:$\LD_LIBRARY_PATH} cmake -GNinja -DCMAKE_C_COMPILER="$\WORKSPACE/llvm-install/bin2/clang" -DCMAKE_CXX_COMPILER="$\WORKSPACE/llvm-install/bin2/clang++" -DCMAKE_Fortran_COMPILER="$\WORKSPACE/llvm-install/bin2/flang-new" -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_FLAGS= -DCMAKE_CXX_FLAGS= -DCMAKE_Fortran_FLAGS= -DCMAKE_C_FLAGS_RELEASE="-O3 -ffast-math -march=armv8.4-a+sve -msve-vector-bits=256 -flto -fuse-ld=lld -DNDEBUG" -DCMAKE_CXX_FLAGS_RELEASE="-O3 -ffast-math -march=armv8.4-a+sve -msve-vector-bits=256 -flto -fuse-ld=lld -DNDEBUG" -DCMAKE_Fortran_FLAGS_RELEASE="-O3 -ffast-math -march=armv8.4-a+sve -msve-vector-bits=256 -flto -fuse-ld=lld -DNDEBUG" -DTEST_SUITE_FORTRAN=ON -DTEST_SUITE_SUBDIRS=Fujitsu -DTEST_SUITE_FUJITSU_WITH_FAST_MATH=ON "$\WORKSPACE/test/test-suite" We track this bug report under https://linaro.atlassian.net/browse/LLVM-2267. (Require Linaro login) Please let us know if you have a fix. If you have any questions regarding this report, please ask on linaro-toolchain(a)lists.linaro.org mailing list. -----------------8<--------------------------8<--------------------------8<-------------------------- The information below contains the details of the failures, and the ways to reproduce a debug environment: You can find the failure logs in *.log.1.xz files in * http://54.172.246.49:9090/jobs/tcwg_flang_test--main-aarch64-Ofast-sve_vls-… The full lists of regressions and improvements as well as configure and make commands are in * http://54.172.246.49:9090/jobs/tcwg_flang_test--main-aarch64-Ofast-sve_vls-… The list of [ignored] baseline and flaky failures are in * http://54.172.246.49:9090/jobs/tcwg_flang_test--main-aarch64-Ofast-sve_vls-… Fujitsu testsuite : https://github.com/fujitsu/compiler-test-suite/ Current build : http://54.172.246.49:9090/jobs/tcwg_flang_test--main-aarch64-Ofast-sve_vls-… Reference build : http://54.172.246.49:9090/jobs/tcwg_flang_test--main-aarch64-Ofast-sve_vls-… Instruction to reproduce the build : https://gitlab.com/LinaroLtd/tcwg/ci/interesting-commits/-/raw/master/llvm/… Full commit : https://github.com/llvm/llvm-project/commit/44ea65c07a600ca1c32912b5cd7356f…

1 month, 3 weeks

Re: [Linaro-TCWG-CI] llvmorg-24-init-6257-gc47197af5327: 3 regressions on aarch64

by Yasuyuki Ohno (Fujitsu)

Hi, These tests are incorrect. Therefore, we will correct the tests. Details are as follows. Investigation shows that all failures are caused by assumptions in the test programs rather than compiler defects. Two independent issues were identified. Issue 1: Fujitsu-C-0060_0008.test and Fujitsu-C-0076_0009.test These tests use NaN/Infinity related computations while being compiled with -ffast-math. The source files explicitly enable: - -fhonor-nans but do not enable: - -fhonor-infinities As a result, optimizations assuming that infinities are not significant remain legal under -ffast-math. Fujitsu-C-0060_0008.test generates Inf and -Inf through division by zero and compares them against expected values. The test assumes IEEE infinity semantics are preserved, but the compilation options do not require that behavior. Fujitsu-C-0076_0009.test exhibits the same problem. Issue 2: Fujitsu-Fortran-0311_0026.test This test uses Cray pointers and an allocatable array that shares the same storage. However, the allocatable arrays participating in the aliasing relationship do not have the TARGET attribute. Under Flang's documented aliasing rules for Cray pointers, optimizations may assume that Cray-pointer targets do not alias ordinary variables unless the aliased variables are declared with the TARGET attribute. Consequently, the optimizer may legally treat the two objects as non-aliasing despite referring to the same storage at runtime. The failure therefore originates from the test program not expressing the required aliasing relationship. Please ignore the mail. We'll update the test. We apologize for the false detection. Best regards, Yasuyuki Ohno Software Development Div., Advanced Technology Development Unit, Fujitsu Research, Fujitsu Limited, JAPAN. ________________________________ 差出人: ci_notify(a)linaro.org <ci_notify(a)linaro.org> 送信: 2026 年 8 月 29 日 (土曜日) 16:20 宛先: contact-ci-compiler-test-suite(a)cs.jp.fujitsu.com <contact-ci-compiler-test-suite(a)cs.jp.fujitsu.com> Cc: jfargnoli(a)nvidia.com <jfargnoli(a)nvidia.com>; maxim.kuvyrkov(a)linaro.org <maxim.kuvyrkov(a)linaro.org> 件名: [Linaro-TCWG-CI] llvmorg-24-init-6257-gc47197af5327: 3 regressions on aarch64 Dear contributor, Our automatic CI has detected problems related to your patch(es). Please find some details below. In tcwg_flang_test/main-aarch64-Ofast-sve_vls-lto-lld, after: | commit llvmorg-24-init-6257-gc47197af5327 | Author: Justin Fargnoli <jfargnoli(a)nvidia.com> | Date: Thu Aug 27 13:03:39 2026 -0700 | | [LTO] Enable `PrepareForLTO` for loop unrolling in pre-link pipelines (#192155) | | Currently, the unroller will not unroll loops with inlineable calls. | However, calls that may not appear to be inlineable during the pre-link | phase may become inlineable during the post-link phase. The unroller's | ... 4 lines of the commit log omitted. Produces 3 regressions: | | regressions.sum: | Running test-suite:Fujitsu/C/0060 ... | FAIL: test-suite :: Fujitsu/C/0060/Fujitsu-C-0060_0008.test | Running test-suite:Fujitsu/C/0076 ... | FAIL: test-suite :: Fujitsu/C/0076/Fujitsu-C-0076_0009.test | Running test-suite:Fujitsu/Fortran/0311 ... | ... and 1 more | # "FAIL" means : the execution of the compiled binary failed / output of the binary differs from the expected one Used configuration : * Toolchain : cmake -G Ninja ../llvm/llvm "-DLLVM_ENABLE_PROJECTS=clang;lld;flang;clang-tools-extra" "-DLLVM_ENABLE_RUNTIMES=openmp" -DCMAKE_BUILD_TYPE=Release -DLLVM_ENABLE_ASSERTIONS=True -DCMAKE_INSTALL_PREFIX=../llvm-install "-DLLVM_TARGETS_TO_BUILD=AArch64" -DCLANG_DEFAULT_LINKER=lld * Testsuite : export LD_LIBRARY_PATH=$\WORKSPACE/llvm-install/lib/aarch64-unknown-linux-gnu$\{LD_LIBRARY_PATH:+:$\LD_LIBRARY_PATH} cmake -GNinja -DCMAKE_C_COMPILER="$\WORKSPACE/llvm-install/bin2/clang" -DCMAKE_CXX_COMPILER="$\WORKSPACE/llvm-install/bin2/clang++" -DCMAKE_Fortran_COMPILER="$\WORKSPACE/llvm-install/bin2/flang-new" -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_FLAGS= -DCMAKE_CXX_FLAGS= -DCMAKE_Fortran_FLAGS= -DCMAKE_C_FLAGS_RELEASE="-O3 -ffast-math -march=armv8.4-a+sve -msve-vector-bits=256 -flto -fuse-ld=lld -DNDEBUG" -DCMAKE_CXX_FLAGS_RELEASE="-O3 -ffast-math -march=armv8.4-a+sve -msve-vector-bits=256 -flto -fuse-ld=lld -DNDEBUG" -DCMAKE_Fortran_FLAGS_RELEASE="-O3 -ffast-math -march=armv8.4-a+sve -msve-vector-bits=256 -flto -fuse-ld=lld -DNDEBUG" -DTEST_SUITE_FORTRAN=ON -DTEST_SUITE_SUBDIRS=Fujitsu -DTEST_SUITE_FUJITSU_WITH_FAST_MATH=ON "$\WORKSPACE/test/test-suite" We track this bug report under https://linaro.atlassian.net/browse/LLVM-2268. (Require Linaro login) Please let us know if you have a fix. If you have any questions regarding this report, please ask on linaro-toolchain(a)lists.linaro.org mailing list. -----------------8<--------------------------8<--------------------------8<-------------------------- The information below contains the details of the failures, and the ways to reproduce a debug environment: You can find the failure logs in *.log.1.xz files in * http://54.172.246.49:9090/jobs/tcwg_flang_test--main-aarch64-Ofast-sve_vls-… The full lists of regressions and improvements as well as configure and make commands are in * http://54.172.246.49:9090/jobs/tcwg_flang_test--main-aarch64-Ofast-sve_vls-… The list of [ignored] baseline and flaky failures are in * http://54.172.246.49:9090/jobs/tcwg_flang_test--main-aarch64-Ofast-sve_vls-… Fujitsu testsuite : https://github.com/fujitsu/compiler-test-suite/ Current build : http://54.172.246.49:9090/jobs/tcwg_flang_test--main-aarch64-Ofast-sve_vls-… Reference build : http://54.172.246.49:9090/jobs/tcwg_flang_test--main-aarch64-Ofast-sve_vls-… Instruction to reproduce the build : https://gitlab.com/LinaroLtd/tcwg/ci/interesting-commits/-/raw/master/llvm/… Full commit : https://github.com/llvm/llvm-project/commit/c47197af53270a4795040756c27802f…

1 month, 3 weeks

Re: [Linaro-TCWG-CI] 4 patches in gcc: 2 regressions on aarch64

by Avinash Jayakar

Hi, I believe these may be flaky since I checked with my 4 patches on cfarm185 machine with following configure command configure --prefix=/home/avinashd/opt --enable-languages=c,c++ -- target=aarch64-unknown-linux-gnu --disable-multilib --enable-fix- cortex-a53-835769 --enable-fix-cortex-a53-843419 --with-arch=armv8-a -- disable-bootstrap on top of the commit a5da203dd95900ba8bab16432667989f3d018c76 : xtensa: Make use of CONST.S instruction On Fri, 2026-08-28 at 22:17 +0000, ci_notify(a)linaro.org wrote: > Dear contributor, > > Our automatic CI has detected problems related to your patch(es). > Please find some details below. > > In aarch64 native, after: > | 4 patches in gcc > | Patchwork URL: https://patchwork.sourceware.org/patch/142333 > | 25ff335aad8 [PATCH 4/4] rs6000: Add sha3 builtins and > documentation [RFC02645] > | f47d33c1b4e [PATCH 3/4] rs6000: Add __dmr2048 type (QDOmode) and > map it to DMR pair register > | 1eb45af654c [PATCH 2/4] genmodes: Use wider types for > mode_unit_size array based on max size The only change in middle-end that has happened is with this patch (2/4). And that is to generate the macro MODE_UNIT_SIZE_TYPE instead of unsigned char. This should not have caused the below mentioned regressions. All other patches are rs6000 backend related. Please let me know if these regressions are always reproducible or if I have tested it incorrectly on the cfarm185 machine. I can have a look at what is happening. Thanks, Avinash Jayakar > | 63b651a8d8f [PATCH 1/4] rs6000: Add sha2 and shapad builtins > [RFC02645] > | ... applied on top of baseline commit: > | aaa0f6b1e87 match.pd: fold reductions of vector duplicates and > repeats [PR127039] > > Produces 2 regressions: > | > | regressions.sum: > | Running gcc:gcc.misc-tests/gcov.exp ... > | FAIL: gcc.misc-tests/gcov-39.c gcov: 0 failures in line counts, 0 > in branch percentages, 0 in condition/decision, 1 in prime-paths, 0 > in return percentages, 0 in intermediate format, 0 failed in filters > | FAIL: gcc.misc-tests/gcov-39.c line 175: expected 1/8 covered > paths, was 1/7 > > Used configuration : > *CI config* tcwg_gcc_check aarch64-linux-gnu > *configure and test flags:* none, autodetected on aarch64-unknown- > linux-gnu--disable-multilib --enable-fix-cortex-a53-835769 --enable- > fix-cortex-a53-843419 --with-arch=armv8-a > > If you have any questions regarding this report, please ask on > linaro-toolchain(a)lists.linaro.org mailing list. > > -----------------8<--------------------------8<---------------------- > ----8<-------------------------- > > The information below contains the details of the failures, and the > ways to reproduce a debug environment: > > You can find the failure logs in *.log.1.xz files in > * > http://54.172.246.49:9090/jobs/tcwg_gcc_check--master-aarch64-precommit/bui… > The full lists of regressions and improvements as well as configure > and make commands are in > * > http://54.172.246.49:9090/jobs/tcwg_gcc_check--master-aarch64-precommit/bui… > The list of [ignored] baseline and flaky failures are in > * > http://54.172.246.49:9090/jobs/tcwg_gcc_check--master-aarch64-precommit/bui… > > Current build : > http://54.172.246.49:9090/jobs/tcwg_gcc_check--master-aarch64-precommit/bui… > Reference build : > http://54.172.246.49:9090/jobs/tcwg_gcc_check--master-aarch64-build/builds/… > > Warning: we do not enable maintainer-mode nor automatically update > generated files, which may lead to failures if the patch modifies the > master files.

1 month, 3 weeks

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

linaro-toolchain