Hi Slava, This regression may not be a real regression but a latent bug.
The attached test.f90 is a simplified reproducer. If I run the following command using LLVM built with assertion enabled, flang -c -target aarch64-unknown-linux-gnu -O2 -ffast-math -march=armv8.2-a+sve -msve-vector-bits=256 -flto test.f90 -o test.bc flang -target aarch64-unknown-linux-gnu -O2 -ffast-math -march=armv8.2-a+sve -msve-vector-bits=256 -flto -fuse-ld=lld test.bc I get the following assertion failure. ---------------------------------------------------------------- ld.lld: llvm-project/llvm/lib/CodeGen/SelectionDAG/LegalizeVectorTypes.cpp:3722: llvm::SDValue llvm::DAGTypeLegalizer::SplitVecOp_EXTRACT_SUBVECTOR(llvm::SDNode*): Assertion `IdxVal + SubVT.getVectorMinNumElements() <= LoEltsMin && "Extracted subvector crosses vector split!"' failed. PLEASE submit a bug report to https://github.com/llvm/llvm-project/issues/ and include the crash backtrace. Stack dump: 0. Program arguments: llvm/bin/ld.lld -EL -z relro --hash-style=gnu --eh-frame-hdr -m aarch64linux -pie -dynamic-linker /lib/ld-linux-aarch64.so.1 -o a.out /lib/aarch64-linux-gnu/Scrt1.o /lib/aarch64-linux-gnu/crti.o /usr/lib/gcc/aarch64-linux-gnu/13/crtbeginS.o /usr/lib/gcc/aarch64-linux-gnu/13/crtfastmath.o -Lllvm/bin/../lib/aarch64-unknown-linux-gnu -Lllvm/lib/clang/21/lib/aarch64-unknown-linux-gnu -L/usr/lib/gcc/aarch64-linux-gnu/13 -L/lib/aarch64-linux-gnu -L/usr/lib/aarch64-linux-gnu -L/lib -L/usr/lib -plugin-opt=mcpu=generic -plugin-opt=O2 -plugin-opt=-enable-tlsdesc /tmp/test-074bef.o -Lllvm/lib -lflang_rt.runtime -lm -lgcc --as-needed -lgcc_s --no-as-needed -lc -lgcc --as-needed -lgcc_s --no-as-needed /usr/lib/gcc/aarch64-linux-gnu/13/crtendS.o /lib/aarch64-linux-gnu/crtn.o 1. Running pass 'Function Pass Manager' on module 'ld-temp.o'. 2. Running pass 'AArch64 Instruction Selection' on function '@main' #0 0x0000ffff9c9cdbd8 llvm::sys::PrintStackTrace(llvm::raw_ostream&, int) (llvm/lib/libLLVMSupport.so.21.0git+0x25dbd8) #1 0x0000ffff9c9cb2c0 llvm::sys::RunSignalHandlers() (llvm/lib/libLLVMSupport.so.21.0git+0x25b2c0) #2 0x0000ffff9c9cb428 SignalHandler(int, siginfo_t*, void*) Signals.cpp:0:0 #3 0x0000ffff9d2c07f0 (linux-vdso.so.1+0x7f0) #4 0x0000ffff9c3a7608 (/lib/aarch64-linux-gnu/libc.so.6+0x87608) #5 0x0000ffff9c35cb3c raise (/lib/aarch64-linux-gnu/libc.so.6+0x3cb3c) #6 0x0000ffff9c347e00 abort (/lib/aarch64-linux-gnu/libc.so.6+0x27e00) #7 0x0000ffff9c355cc0 (/lib/aarch64-linux-gnu/libc.so.6+0x35cc0) #8 0x0000ffff9c355d30 __assert_perror_fail (/lib/aarch64-linux-gnu/libc.so.6+0x35d30) #9 0x0000ffff97a88cd4 llvm::DAGTypeLegalizer::SplitVecOp_EXTRACT_SUBVECTOR(llvm::SDNode*) LegalizeVectorTypes.cpp:0:0 #10 0x0000ffff97a8fb88 llvm::DAGTypeLegalizer::SplitVectorOperand(llvm::SDNode*, unsigned int) LegalizeVectorTypes.cpp:0:0 #11 0x0000ffff97a58df4 llvm::DAGTypeLegalizer::run() LegalizeTypes.cpp:0:0 #12 0x0000ffff97a59ad4 llvm::SelectionDAG::LegalizeTypes() (llvm/lib/libLLVMSelectionDAG.so.21.0git+0x299ad4) #13 0x0000ffff97bccbfc llvm::SelectionDAGISel::CodeGenAndEmitDAG() (llvm/lib/libLLVMSelectionDAG.so.21.0git+0x40cbfc) #14 0x0000ffff97bd00a4 llvm::SelectionDAGISel::SelectAllBasicBlocks(llvm::Function const&) (llvm/lib/libLLVMSelectionDAG.so.21.0git+0x4100a4) #15 0x0000ffff97bd16d4 llvm::SelectionDAGISel::runOnMachineFunction(llvm::MachineFunction&) (llvm/lib/libLLVMSelectionDAG.so.21.0git+0x4116d4) #16 0x0000ffff97bbe20c llvm::SelectionDAGISelLegacy::runOnMachineFunction(llvm::MachineFunction&) (llvm/lib/libLLVMSelectionDAG.so.21.0git+0x3fe20c) #17 0x0000ffff989d6eb0 llvm::MachineFunctionPass::runOnFunction(llvm::Function&) (.part.0) MachineFunctionPass.cpp:0:0 #18 0x0000ffff992eec18 llvm::FPPassManager::runOnFunction(llvm::Function&) (llvm/lib/libLLVMCore.so.21.0git+0x49ec18) #19 0x0000ffff992eee1c llvm::FPPassManager::runOnModule(llvm::Module&) (llvm/lib/libLLVMCore.so.21.0git+0x49ee1c) #20 0x0000ffff992efa80 llvm::legacy::PassManagerImpl::run(llvm::Module&) (llvm/lib/libLLVMCore.so.21.0git+0x49fa80) #21 0x0000ffff9ae3dcb8 codegen(llvm::lto::Config const&, llvm::TargetMachine*, std::function<llvm::Expected<std::unique_ptr<llvm::CachedFileStream, std::default_delete<llvm::CachedFileStream>>> (unsigned int, llvm::Twine const&)>, unsigned int, llvm::Module&, llvm::ModuleSummaryIndex const&) LTOBackend.cpp:0:0 #22 0x0000ffff9ae3f1ac llvm::lto::backend(llvm::lto::Config const&, std::function<llvm::Expected<std::unique_ptr<llvm::CachedFileStream, std::default_delete<llvm::CachedFileStream>>> (unsigned int, llvm::Twine const&)>, unsigned int, llvm::Module&, llvm::ModuleSummaryIndex&) (llvm/lib/libLLVMLTO.so.21.0git+0x6f1ac) #23 0x0000ffff9ae30958 llvm::lto::LTO::runRegularLTO(std::function<llvm::Expected<std::unique_ptr<llvm::CachedFileStream, std::default_delete<llvm::CachedFileStream>>> (unsigned int, llvm::Twine const&)>) (llvm/lib/libLLVMLTO.so.21.0git+0x60958) #24 0x0000ffff9ae341e4 llvm::lto::LTO::run(std::function<llvm::Expected<std::unique_ptr<llvm::CachedFileStream, std::default_delete<llvm::CachedFileStream>>> (unsigned int, llvm::Twine const&)>, llvm::FileCache) (llvm/lib/libLLVMLTO.so.21.0git+0x641e4) #25 0x0000ffff9d110f74 lld::elf::BitcodeCompiler::compile() (llvm/lib/liblldELF.so.21.0git+0x220f74) #26 0x0000ffff9d06fb50 void lld::elf::LinkerDriver::compileBitcodeFiles<llvm::object::ELFType<(llvm::endianness)1, true>>(bool) (llvm/lib/liblldELF.so.21.0git+0x17fb50) #27 0x0000ffff9d08f0e8 void lld::elf::LinkerDriver::link<llvm::object::ELFType<(llvm::endianness)1, true>>(llvm::opt::InputArgList&) (llvm/lib/liblldELF.so.21.0git+0x19f0e8) #28 0x0000ffff9d093694 lld::elf::LinkerDriver::linkerMain(llvm::ArrayRef<char const*>) (llvm/lib/liblldELF.so.21.0git+0x1a3694) #29 0x0000ffff9d0949a4 lld::elf::link(llvm::ArrayRef<char const*>, llvm::raw_ostream&, llvm::raw_ostream&, bool, bool) (llvm/lib/liblldELF.so.21.0git+0x1a49a4) #30 0x0000ffff9ca674bc lld::unsafeLldMain(llvm::ArrayRef<char const*>, llvm::raw_ostream&, llvm::raw_ostream&, llvm::ArrayRef<lld::DriverDef>, bool) (llvm/lib/liblldCommon.so.21.0git+0x274bc) #31 0x0000aaaae10c2734 lld_main(int, char**, llvm::ToolContext const&) (llvm/bin/ld.lld+0x12734) #32 0x0000aaaae10c2a00 main (llvm/bin/ld.lld+0x12a00) #33 0x0000ffff9c3484c4 (/lib/aarch64-linux-gnu/libc.so.6+0x284c4) #34 0x0000ffff9c348598 __libc_start_main (/lib/aarch64-linux-gnu/libc.so.6+0x28598) #35 0x0000aaaae10c2070 _start (llvm/bin/ld.lld+0x12070) flang-21: error: unable to execute command: Aborted (core dumped) flang-21: error: linker command failed due to signal (use -v to see invocation) ---------------------------------------------------------------- I've confirmed this assertion failure using this simplified reproducer occurs with earlier LLVM revisions. > Hi Slava, > > I looked through bisection logs and could not find any issues with the > bisection. > > The commit history is this: > > 2d12d31f44acac54d7b2858624cb8a1db5a0a8ce - FAIL > 3aad7d7a3ce57bc06b3cd08df8030064da97bf4e - PASS > 6dc6ca3302edadad7d4b195a05038995b84606d0 - UNTESTED > > These tests don't fail for commit 3aad7d7a3ce57bc06b3cd08df8030064da97bf4e > (also yours) and fail for 2d12d31f44acac54d7b2858624cb8a1db5a0a8ce. See > results in [1] and [2]. > > I don't have ready results for 6dc6ca3302edadad7d4b195a05038995b84606d0. > Could you please double check that these tests FAIL for you for the immediate > parent of llvmorg-21-init-11733-g2d12d31f44ac -- > 3aad7d7a3ce57bc06b3cd08df8030064da97bf4e ? > > The only other data point I have is that these tests PASSed in > cf3242f3b0f6eec47786426777f6e0b18363caa1, which is 8 commits before yours. > > Finally, are you using below flags in your testing? In particular "-flto > -fuse-ld=lld"? Error in [3] happens during LTO processing. > > > -O3 -ffast-math -march=armv8.4-a+sve -msve-vector-bits=256 -mllvm > > -treat-scalable-fixed-error-as-warning=false -flto -fuse-ld=lld -DNDEBUG > > > [1] > https://ci.linaro.org/job/tcwg_flang_test--main-aarch64-Ofast-sve_vls-lto-lld-bisect/176/artifact/artifacts/build-last_good/00-sumfiles/ > [2] > https://ci.linaro.org/job/tcwg_flang_test--main-aarch64-Ofast-sve_vls-lto-lld-bisect/176/artifact/artifacts/build-first_bad/00-sumfiles/ > [3] > https://ci.linaro.org/job/tcwg_flang_test--main-aarch64-Ofast-sve_vls-lto-lld-bisect/176/artifact/artifacts/build-first_bad/00-sumfiles/ninja.log.xz > > Thanks! > > -- > Maxim Kuvyrkov > https://www.linaro.org > > > On May 24, 2025, at 02:20, Slava Zakharin <szakha...@nvidia.com> wrote: > > > > Hello, > > > > I think there was an error in bisecting these regressions. I can reproduce > > the same failure with commit 6dc6ca3302edadad7d4b195a05038995b84606d0, > > which happened before mine. > > > > Thanks, > > Slava > > > > -----Original Message----- > > From: ci_not...@linaro.org <ci_not...@linaro.org> > > Sent: Wednesday, May 21, 2025 9:10 PM > > To: ohno.yasuy...@fujitsu.com; itou.tets...@fujitsu.com; > > t-kawash...@fujitsu.com > > Cc: Slava Zakharin <szakha...@nvidia.com>; maxim.kuvyr...@linaro.org > > Subject: [Linaro-TCWG-CI] llvmorg-21-init-11733-g2d12d31f44ac: 3 > > regressions on aarch64 > > > > [You don't often get email from ci_not...@linaro.org. Learn why this is > > important at https://aka.ms/LearnAboutSenderIdentification ] > > > > External email: Use caution opening links or attachments > > > > > > 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-21-init-11733-g2d12d31f44ac > > | Author: Slava Zakharin <szakha...@nvidia.com> > > | Date: Mon May 12 18:33:47 2025 -0700 > > | > > | [flang] Propagate contiguous attribute through HLFIR. (#138797) > > | > > | This change allows marking more designators producing an opaque > > | box with 'contiguous' attribute, e.g. like in test1 case > > | in flang/test/HLFIR/propagate-contiguous-attribute.fir. > > | ... 5 lines of the commit log omitted. > > > > Produces 3 regressions: > > | > > | regressions.sum: > > | Running test-suite:Fujitsu/Fortran/0105 ... > > | NOEXE: test-suite :: Fujitsu/Fortran/0105/Fujitsu-Fortran-0105_0089.test > > | NOEXE: test-suite :: Fujitsu/Fortran/0105/Fujitsu-Fortran-0105_0204.test > > | NOEXE: test-suite :: Fujitsu/Fortran/0105/Fujitsu-Fortran-0105_0205.test > > | # "NOEXE" means : the test program cannot be compiled > > > > Used configuration : > > * Toolchain : cmake -G Ninja ../llvm/llvm > > "-DLLVM_ENABLE_PROJECTS=clang;lld;flang;openmp;clang-tools-extra" > > -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/bin/clang" > > -DCMAKE_CXX_COMPILER="$\WORKSPACE/llvm-install/bin/clang++" > > -DCMAKE_Fortran_COMPILER="$\WORKSPACE/llvm-install/bin/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 -mllvm > > -treat-scalable-fixed-error-as-warning=false -flto -fuse-ld=lld -DNDEBUG" > > -DCMAKE_CXX_FLAGS_RELEASE="-O3 -ffast-math -march=armv8.4-a+sve > > -msve-vector-bits=256 -mllvm -treat-scalable-fixed-error-as-warning=false > > -flto -fuse-ld=lld -DNDEBUG" -DCMAKE_Fortran_FLAGS_RELEASE="-O3 > > -ffast-math -march=armv8.4-a+sve -msve-vector-bits=256 -mllvm > > -treat-scalable-fixed-error-as-warning=false -flto -fuse-ld=lld -DNDEBUG" > > -DTEST_SUITE_FORTRAN=ON -DTEST_SUITE_SUBDIRS=Fujitsu > > "$\WORKSPACE/test/test-suite" > > > > We track this bug report under > > https://linaro.atlassian.net/browse/LLVM-1686. Please let us know if you > > have a fix. > > > > If you have any questions regarding this report, please ask on > > linaro-toolchain@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://ci.linaro.org/job/tcwg_flang_test--main-aarch64-Ofast-sve_vls-lto-lld-build/2038/artifact/artifacts/00-sumfiles/ > > The full lists of regressions and improvements as well as configure and > > make commands are in > > * > > https://ci.linaro.org/job/tcwg_flang_test--main-aarch64-Ofast-sve_vls-lto-lld-build/2038/artifact/artifacts/notify/ > > The list of [ignored] baseline and flaky failures are in > > * > > https://ci.linaro.org/job/tcwg_flang_test--main-aarch64-Ofast-sve_vls-lto-lld-build/2038/artifact/artifacts/sumfiles/xfails.xfail > > > > Fujitsu testsuite : https://github.com/fujitsu/compiler-test-suite/ > > > > Current build : > > https://ci.linaro.org/job/tcwg_flang_test--main-aarch64-Ofast-sve_vls-lto-lld-build/2038/artifact/artifacts > > Reference build : > > https://ci.linaro.org/job/tcwg_flang_test--main-aarch64-Ofast-sve_vls-lto-lld-build/2036/artifact/artifacts > > > > Instruction to reproduce the build : > > https://git-us.linaro.org/toolchain/ci/interesting-commits.git/plain/llvm/sha1/2d12d31f44acac54d7b2858624cb8a1db5a0a8ce/tcwg_flang_test/main-aarch64-Ofast-sve_vls-lto-lld/reproduction_instructions.txt > > > > Full commit : > > https://github.com/llvm/llvm-project/commit/2d12d31f44acac54d7b2858624cb8a1db5a0a8ce _______________________________________________ linaro-toolchain mailing list -- linaro-toolchain@lists.linaro.org To unsubscribe send an email to linaro-toolchain-le...@lists.linaro.org
