)]}' { "log": [ { "commit": "a9bdc5ff3a92d811c198955747f1ab9d131fb308", "tree": "186958fa1c2c9565c22ab0ab769dafb2812a6fdf", "parents": [ "85b0ac81b41c2f2fc05cdd395353d8fb93878861" ], "author": { "name": "Miguel A. Arroyo", "email": "miguel.arroyo@rockstargames.com", "time": "Wed Jun 03 20:36:37 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed Jun 03 20:40:28 2026" }, "message": "[compiler-rt][cmake][msvc] Install PDBs for shared libraries (#201224)\n\nFollowing up on similar PRs in the past to address PDBs not being copied\nfor install targets:\n* https://github.com/llvm/llvm-project/pull/126961\n* https://github.com/llvm/llvm-project/pull/126680\n* https://github.com/llvm/llvm-project/pull/120683\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: fb649b41b0db8cd33a5be613345be845a41d0d98\n" }, { "commit": "85b0ac81b41c2f2fc05cdd395353d8fb93878861", "tree": "bc238f2ceb2e62ed3a1359e9e4fce5f8b7b9bce9", "parents": [ "5d18abbff30b3713c209e9b49959a01363cb37e3" ], "author": { "name": "George Burgess IV", "email": "george.burgess.iv@gmail.com", "time": "Wed Jun 03 18:55:02 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed Jun 03 19:01:33 2026" }, "message": "[compiler-rt] fix out-of-bounds access in scudo test (#201390)\n\nWhen `I \u003d 0`, we can\u0027t do `Array[I-1]`.\n\nCaught by libcxx hardening.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 5708e7fd9eca368b1b519c13a58360a15a263e4f\n" }, { "commit": "5d18abbff30b3713c209e9b49959a01363cb37e3", "tree": "74a218dd2b226d1ce26090c139f9a40b83bfb837", "parents": [ "45a5d9099951a5665d673912ed2198daa3800b85" ], "author": { "name": "Hans Wennborg", "email": "hans@hanshq.net", "time": "Wed Jun 03 18:05:47 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed Jun 03 18:10:58 2026" }, "message": "Revert \"[PGO][AMDGPU] Add basic HIP offload PGO support (#177665)\" (#201416)\n\nThis broke profiling builds on Windows by switching the profile library\nto link against the dynamic CRT; see discussion on the PR.\n\nThere were already a number of issues reported and fixed after this PR.\nRather than piling on the fixes (and this one may need some work),\nrevert back to green for now to let the project recover.\n\nThis reverts commit 5db13643f4b7038db0ca304d9f8900122502935c.\n\nAdditionally, this reverts the followup PRs in\n635e120fb87304924508a7a204574727e3c37363,\n2766733764f4bdf8399d48c0225e9c64bdfa95f7,\n4c33844b4b560f24b2a3a0cc689d73510cb01ad5, and\n5eca8b67ff3c5f371141d54b6e4544a0ebe77fdb:\n\n\"[PGO][HIP] Stop pulling ROCm.o into every PGO host link (#200101)\"\n\"[compiler-rt][profile] Add COMPILER_RT_BUILD_PROFILE_ROCM option\n(#200127)\"\n\"[PGO][HIP] Skip ROCm interceptor in profile-only compiler-rt builds\n(#200111)\"\n\"[PGO][HIP] Fix profile-only Windows link by gating ROCm interceptor\nmacro (#200859)\"\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 6cfa1a01a0f737ed1d54963810057fb1bd67a274\n" }, { "commit": "45a5d9099951a5665d673912ed2198daa3800b85", "tree": "14f1d02de7dd9dd90bb09ec58491dea29b99d189", "parents": [ "0bbf712d62680b1cebb7f411ab93fa5fe1e87646" ], "author": { "name": "Amit Kumar Pandey", "email": "pandey.kumaramit2023@gmail.com", "time": "Tue Jun 02 09:52:49 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 09:55:44 2026" }, "message": "[compiler-rt][sanitizer_common] Generalize CheckNoDeepBind as OnDlOpen (#200748)\n\nRename the dlopen pre-check hook to OnDlOpen so platform-specific dlopen\nhandling can be extended beyond the RTLD_DEEPBIND guard on Linux.\nWindows and macOS keep no-op implementations. All dlopen interceptors\ncall the shared hook.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: bae4566a422cba33fb41a072320b36fb51cb67ee\n" }, { "commit": "0bbf712d62680b1cebb7f411ab93fa5fe1e87646", "tree": "e55ba105e4dc201576350d587ebd9aac34220053", "parents": [ "a4abe1ba3af2db298c0194f8767b97afb5c21b94" ], "author": { "name": "Christopher Ferris", "email": "cferris1000@users.noreply.github.com", "time": "Tue Jun 02 04:49:18 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 04:50:31 2026" }, "message": "[scudo] Return nullptr if a remap fails on linux. (#200537)\n\nAdd a check if a fixed address mmap doesn\u0027t return the expected address.\n\nAllow a remap call to fail if the mmap fails and returns a nullptr to\nthe caller.\n\nFix a place where if remap fails in the secondary, it didn\u0027t do\nanything. Now it will unmap the original entry on failure.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: adcd285fc4047759e4014d6568d80bb64589aded\n" }, { "commit": "a4abe1ba3af2db298c0194f8767b97afb5c21b94", "tree": "49611c37302209926fb86edf5ca4bb5355511bce", "parents": [ "9889c7e9db4a970fe4e8e16abf86ee602626e02a" ], "author": { "name": "Wael Yehia", "email": "wmyehia2001@yahoo.com", "time": "Mon Jun 01 18:45:34 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 18:50:50 2026" }, "message": "[AIX] Set the ifunc constructor\u0027s priority to 100, to run before any user code (#200893)\n\nCo-authored-by: Wael Yehia \u003cwyehia@ca.ibm.com\u003e\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 9433fd7666ca6e0e38c05e3dc911b0685ffa0ab3\n" }, { "commit": "9889c7e9db4a970fe4e8e16abf86ee602626e02a", "tree": "253471636af3d81752d08d1aaddd2134959298d9", "parents": [ "b899dc508d1da3eb172ba8a65ae6d1176b0358cf" ], "author": { "name": "Yaxun (Sam) Liu", "email": "yaxun.liu@amd.com", "time": "Mon Jun 01 16:55:15 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 17:00:24 2026" }, "message": "[PGO][HIP] Fix profile-only Windows link by gating ROCm interceptor macro (#200859)\n\nPR #200111 stops compiling InstrProfilingPlatformROCm.cpp (which defines\nthe\nHIP GPU helper __llvm_profile_hip_collect_device_data) in profile-only\nbuilds.\nBut the compile define -DCOMPILER_RT_BUILD_PROFILE_ROCM\u003d1 was still\nadded\nwhenever the COMPILER_RT_BUILD_PROFILE_ROCM option was on (the default),\nso\nInstrProfilingFile.c still referenced the helper from\n__llvm_profile_write_file\neven though it was never built.\n\nOn ELF the declaration is weak, so the undefined symbol folds to null\nand the\naddress-guarded call is skipped. COFF/Windows has no such fallback:\n\n error LNK2019: unresolved external symbol\n __llvm_profile_hip_collect_device_data referenced in function\n __llvm_profile_write_file\n\nAdd the define only when PROFILE_HAS_HIP_INTERCEPTOR is true, i.e. the\nsame\ncondition that keeps InstrProfilingPlatformROCm.cpp in the archive, so\nthe\nmacro is defined iff the helper is actually compiled in.\n\nReported by zmodem:\nhttps://github.com/llvm/llvm-project/pull/200111#issuecomment-4593893230\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 5eca8b67ff3c5f371141d54b6e4544a0ebe77fdb\n" }, { "commit": "b899dc508d1da3eb172ba8a65ae6d1176b0358cf", "tree": "6d79713aff4cef91fa6943fd18c81794a89705d0", "parents": [ "03641f7a5b05e48e318d64369057db577cafc594" ], "author": { "name": "Justin T. Gibbs", "email": "gibbs@scsiguy.com", "time": "Mon Jun 01 06:05:59 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 06:10:29 2026" }, "message": "[asan] Allow OOM Policy Injection in Allocator::Allocate API (#196413)\n\nGroundwork for #196388.\n\nAllocator::Allocate() previously unconditionally applied the global\nOOM policy as dictated by the AllocatorMayReturnNull() flag: return\nnullptr or abort via Report*+Die(). Introduce Allocator::AllocateImpl()\nwhich delegates this decision to the caller via its \u0027may_return_null\u0027\nboolean parameter, allowing future changes to the operator new\nimplementation to fully control this behavior.\n\nImplementation:\n\n * Rename Allocator::Allocate(...) to Allocator::AllocateImpl(...) and\n add a may_return_null parameter that replaces the three internal\n AllocatorMayReturnNull() calls (RSS limit, oversize, OOM).\n\n * Reintroduce Allocator::Allocate(...) as a thin wrapper that calls\n AllocateImpl(..., AllocatorMayReturnNull()), preserving the\n flag-deferring semantics for every existing caller.\n\nNFC.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 56f28d538f737bcd50081f75fbb8c01d0bc80c38\n" }, { "commit": "03641f7a5b05e48e318d64369057db577cafc594", "tree": "9deb4fb6b9fecbbca9507ebca4e6d88ab09642b8", "parents": [ "01e6d90b99ee51de2d0ee7eaa91fac9682ae6ebc" ], "author": { "name": "Justin T. Gibbs", "email": "gibbs@scsiguy.com", "time": "Sun May 31 01:57:01 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Sun May 31 02:00:34 2026" }, "message": "[asan] NFC: clang-format allocator-related files (#200478)\n\nDepends on #200615\nGroundwork for #196413.\n\nMechanical cleanup of allocator related files in preparation of\nfunctional changes. clang-format (v21.1.2) applied whole-file to:\n\n compiler-rt/lib/asan/asan_allocator.cpp\n compiler-rt/lib/asan/asan_allocator.h\n compiler-rt/lib/asan/asan_malloc_linux.cpp\n compiler-rt/lib/asan/asan_malloc_mac.cpp\n compiler-rt/lib/asan/asan_new_delete.cpp\n compiler-rt/lib/asan/tests/asan_noinst_test.cpp\n\nBoth compiler-rt/lib/asan/.clang-format and\ncompiler-rt/lib/sanitizer_common/.clang-format use\n\"BasedOnStyle: Google\", so pointer alignment becomes \"Type* name\"\nthroughout.\n\nNFC.\n\nAssisted by: Claude Opus 4.7\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: baa079610087a765edf4edada9e3f155c8ef0b16\n" }, { "commit": "01e6d90b99ee51de2d0ee7eaa91fac9682ae6ebc", "tree": "9c35d8f0e0dbf4630aac9c78c6a0d33f99b42467", "parents": [ "e5b48a92941d94112ceb48a4dc6241dffd32f046" ], "author": { "name": "Justin T. Gibbs", "email": "gibbs@scsiguy.com", "time": "Sun May 31 00:48:42 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Sun May 31 00:50:33 2026" }, "message": "[asan] NFC: tag ambiguous literal args at Allocate call sites (#200615)\n\nGroundwork for #200478\n\nAdd /*name\u003d*/ comments to ambiguous literal call-site arguments in\ncompiler-rt/lib/asan/asan_allocator.cpp so the parameter name is visible\nat the call site for the Allocate / instance.Allocate uses. Covers the\ncan_fill flag and the bare alignment literals (8, 16, 0) used by:\n\n asan_malloc / asan_vec_malloc / asan_realloc / asan_valloc /\n asan_pvalloc / asan_memalign / asan_aligned_alloc /\n asan_posix_memalign / asan_new / asan_new_aligned\n\nand the internal Reallocate / Calloc paths.\n\nNFC.\n\nAssisted by: Claude Opus 4.7\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: b51054818b78dc395cd4d33f17cfb6e98a36a76d\n" }, { "commit": "e5b48a92941d94112ceb48a4dc6241dffd32f046", "tree": "b664885e93675e458c9a5651a16b97101c893787", "parents": [ "130ef2d26cad9a134356a038bdf337f8f726b3f9" ], "author": { "name": "Christopher Ferris", "email": "cferris1000@users.noreply.github.com", "time": "Fri May 29 19:18:47 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 29 19:20:50 2026" }, "message": "[scudo] Improve performance of pushBlocks sort. (#200297)\n\nRan this on an Android device using both algorithms, the new algorithm\nis on average 10% faster, but gets to be 15% faster in some cases. This\nis an example of the speed-ups.\n\nAverage Operation Time Maximum Operation Time Name\n 326.9(ns) 80770(ns) PushBlocks New\n 365.9(ns) 108032(ns) PushBlocks Old\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: c9cdbc4a30b83b8a8d100c19c4aba16c848b9dfb\n" }, { "commit": "130ef2d26cad9a134356a038bdf337f8f726b3f9", "tree": "81d0a2e56c38ec0142e5a749803ce5334b023b5b", "parents": [ "b845f03aafff16fbc0a94483688f350fbdd65ea1" ], "author": { "name": "eleviant", "email": "eleviant@accesssoftek.com", "time": "Fri May 29 13:22:31 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 29 13:25:32 2026" }, "message": "Honor two\u0027s complement signed overflow with -fms-compatibility (#198538)\n\nThis matches MSVC behavior, where signed integer overflow follows\ntwo\u0027s-complement semantics\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 0fd11e4a83e862fcb48690555015f7da44131075\n" }, { "commit": "b845f03aafff16fbc0a94483688f350fbdd65ea1", "tree": "9397c2281a89985345a38767ea2a0ba538966e9e", "parents": [ "3e60c00fff9bc8bfed48c1e2d1fbb76af6857dd4" ], "author": { "name": "Martin Storsjö", "email": "martin@martin.st", "time": "Fri May 29 10:33:46 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 29 10:35:51 2026" }, "message": "[compiler-rt] [ARM] Restore Windows specific names for fp-\u003eint functions (#200360)\n\nThese Windows specific names were added in\n8c59680ac26cf3d772f51aa5ebd64be8e59ed384 (and refactored later in\ne6407356ba008e474322d52b7d031f65fa4913fc).\n\nWhen ARM specific implementations of them were added in\n174eb79572b3433aa338ede31815fa8fa6a13b29, the ARM specific files replace\nthe generic ones - while the ARM specific file lacked the Windows/ARM\nrenaming of the functions.\n\nRestore the Windows naming of these functions.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 354cdd40be8715c41c1d68faab31c330e026b369\n" }, { "commit": "3e60c00fff9bc8bfed48c1e2d1fbb76af6857dd4", "tree": "b9906719c066d04657f92f5da166915b9ca90636", "parents": [ "a6bba33014dc2fb0533d20ed985b7b4faee6850f" ], "author": { "name": "Dan Liew", "email": "dan@su-root.co.uk", "time": "Fri May 29 08:53:34 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 29 08:55:46 2026" }, "message": "[Compiler-rt][CMake][BoundsSafety] Add `COMPILER_RT_ENABLE_TEST_SUITES` to control which test suites are enabled (#196385)\n\nThis change introduces a new CMake cache variable\n`COMPILER_RT_ENABLE_TEST_SUITES` that allows users to have fine-grained\ncontrol over which test suites are included and therefore run when\nrunning `ninja check-all`.\n\nFor example building with `COMPILER_RT_ENABLE_TEST_SUITES\u003dbuiltins;asan`\nwill only enable the builtins and asan tests under `compiler-rt/tests`.\n\nThe default value of `COMPILER_RT_INCLUDE_TESTS` is `all` which is a\nspecial value that includes everything and it being the default means\nthis patch preserves existing behavior. Note this new option only has\naffect when `COMPILER_RT_INCLUDE_TESTS` is On.\n\nThe primary use case for this change is it allows controlling tests that\ndon\u0027t have a corresponding runtime to be enabled/disabled. An example of\nthis is the `shadowcallstack` test suite. `-fbounds-safety` also has a\ndownstream test suite in compiler-rt like this. Previously there was no\nway to disable including these tests when `COMPILER_RT_INCLUDE_TESTS`\nwas on.\n\nThe `-fbounds-safety` test suite will be upstreamed once the necessary\nclang pieces exist upstream.\n\nA side benefit of this change is it allows building runtimes but not\ntesting some of them which can be useful in some testing environments\n(e.g. smoke-test). It is not the primary motivation though.\n\nWhile the existing `COMPILER_RT_BUILD_\u003cNAME\u003e` options can be used to\nprevent running tests (by not building corresponding runtimes) that\nisn\u0027t a good solution because it:\n\n* Doesn\u0027t work for tests that have no corresponding runtime to be built\n like `shadowcallstack` because there\u0027s no CMake option for it.\n* Requires specifying multiple CMake options to disable building each\n runtime that we don\u0027t want tested. This approach disallows \"building\n but not testing certain runtimes\" and also is annoying to maintain\n because its effectively a subtractive list (you say what you don\u0027t\n want) that has to be updated everytime a new runtime is introduced\n whereas COMPILER_RT_ENABLE_TEST_SUITES is an additive list (you say\n what you want) that doesn\u0027t need modifying everytime a new runtime is\n introduced.\n\nTo implement this all test directories in `compiler-rt/test` are now\nadded through the `compiler_rt_test_runtime()` function which has been\nadapted to accommodate all use cases.\n\nAssisted-by: Claude Code\n\nrdar://176477660\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 0d8316a53da5ef7b46dd54ea5cf69c89129bb46d\n" }, { "commit": "a6bba33014dc2fb0533d20ed985b7b4faee6850f", "tree": "8199170e49120fb645356d38883a7883338aa346", "parents": [ "c82b1c59526104f4b736e90e2e418ceb51e03445" ], "author": { "name": "Yaxun (Sam) Liu", "email": "yaxun.liu@amd.com", "time": "Fri May 29 02:39:20 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 29 02:40:20 2026" }, "message": "[PGO][HIP] Skip ROCm interceptor in profile-only compiler-rt builds (#200111)\n\nPR #177665 made `libclang_rt.profile.a` merge `RTInterception` and\n`sanitizer_common` object libs to support the `hipModuleLoad*` host\ninterceptor added in `InstrProfilingPlatformROCm.cpp`. Those object-lib\ntargets are only created when `COMPILER_RT_BUILD_SANITIZERS / _MEMPROF /\n_XRAY / _CTX_PROFILE` is enabled (see `compiler-rt/lib/CMakeLists.txt`),\nso a profile-only configuration fails at configure time:\n\n```\nCMake Error at cmake/Modules/AddCompilerRT.cmake:365 (add_library):\n Error evaluating generator expression:\n $\u003cTARGET_OBJECTS:RTInterception.x86_64\u003e\n Objects of target \"RTInterception.x86_64\" referenced but no such target\n exists.\nCall Stack (most recent call first):\n lib/profile/CMakeLists.txt:233 (add_compiler_rt_runtime)\n```\n\nFix: gate the object-lib merge on the `RTInterception.\u003carch\u003e` /\n`RTSanitizerCommon*.\u003carch\u003e` targets actually existing, and drop\n`InstrProfilingPlatformROCm.cpp` from `PROFILE_SOURCES` in that case so\nthe static archive stays self-contained. The host-side hook\n`__llvm_profile_hip_collect_device_data` is already declared weak in\n`InstrProfilingFile.c` (PR #200101), so its absence is fine at link\ntime.\n\nVerified:\n- Profile-only standalone compiler-rt build\n(`-DCOMPILER_RT_BUILD_SANITIZERS\u003dOFF -DCOMPILER_RT_BUILD_MEMPROF\u003dOFF\n-DCOMPILER_RT_BUILD_XRAY\u003dOFF -DCOMPILER_RT_BUILD_CTX_PROFILE\u003dOFF\n-DCOMPILER_RT_BUILD_PROFILE\u003dON`) now configures and links cleanly.\n- Default runtimes build still merges the object libs and produces the\ninterceptor; `check-profile` 134/134 pass.\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 4c33844b4b560f24b2a3a0cc689d73510cb01ad5\n" }, { "commit": "c82b1c59526104f4b736e90e2e418ceb51e03445", "tree": "5ce8ed841b7961064862aa211113a9239362dc5d", "parents": [ "9757ce6df3c6285ccbbd8a6ea89d0df19f8a6675" ], "author": { "name": "Christopher Ferris", "email": "cferris1000@users.noreply.github.com", "time": "Thu May 28 18:44:17 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu May 28 18:45:39 2026" }, "message": "[scudo] Create a non-static getErrorInfo function. (#199770)\n\nCreate a getErrorInfo function that operates on the Allocator and\ndoesn\u0027t require passing in all of the extra information.\n\nAdd interface function __scudo_get_fault_error_info that calls this new\nfunction.\n\nAdd all needed functions to support the new getErrorInfo.\n\nThis is being added to replace the static version from Android that\nrequired linking in a copy of libscudo to use. This new version will be\nused directly from libc.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: a0ac7521342567573da5b1e47c94f6199faada39\n" }, { "commit": "9757ce6df3c6285ccbbd8a6ea89d0df19f8a6675", "tree": "c8d910fb63ea98bdb23c9ce90912e616b5a2a182", "parents": [ "42e5d8756b5eb058b0e69f535aa1d863324ff1a5" ], "author": { "name": "Guy David", "email": "guyda96@gmail.com", "time": "Thu May 28 13:55:48 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu May 28 14:01:19 2026" }, "message": "[compiler-rt][profile] Add COMPILER_RT_BUILD_PROFILE_ROCM option (#200127)\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 2766733764f4bdf8399d48c0225e9c64bdfa95f7\n" }, { "commit": "42e5d8756b5eb058b0e69f535aa1d863324ff1a5", "tree": "0edc503b04d4109a6d0561e8bf235bb53ff9ad8d", "parents": [ "b7f9fa6b211b362d3ca07ab2043419ebaf75d1d0" ], "author": { "name": "Matt Turner", "email": "mattst88@gmail.com", "time": "Thu May 28 13:52:52 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu May 28 13:56:01 2026" }, "message": "Add support for Alpha to libsanitizer (#194161)\n\nAdd support for [Alpha](https://en.wikipedia.org/wiki/DEC_Alpha) to\nlibsanitizer.\n\n- [sanitizer] Add SANITIZER_ALPHA platform macro\n- [sanitizer] Add Linux/Alpha syscall implementation\n- [sanitizer] Skip __old_kernel_stat check on Alpha\n- [sanitizer] Use statx for stat syscalls on Linux/Alpha\n- [sanitizer] Add GetPcSpBp signal context support for Alpha\n- [sanitizer] Add Alpha struct size constants\n- [sanitizer] Add Alpha ioctl encoding constants\n- [compiler-rt] sanitizer: Alpha struct sigaction has no sa_restorer\n- [compiler-rt] sanitizer: Disable interceptor trampoline on Alpha\n- [compiler-rt] sanitizer: Add Alpha __sanitizer_dirent layout\n- [compiler-rt] sanitizer: Add Alpha ThreadDescriptorSizeFallback()\n- [compiler-rt] Enable Alpha Linux target\n- [compiler-rt] Use fixed shadow offset 0x10000000000 (1T) for Alpha\n- [compiler-rt][ASan] Limit allocator to 512G on Alpha\n\nTested by porting the changes to gcc\u0027s downstream libsanitizer and\nbuilding and running the libunwind test suite with UBSan / ASan enabled.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 992ee5934c1142c00dc09589380b8277cc43ef09\n" }, { "commit": "b7f9fa6b211b362d3ca07ab2043419ebaf75d1d0", "tree": "b185fe87a523bd8695708de6b1c523ebbf39965c", "parents": [ "759a5cfacd51469570f50a5874895d1e5d9589e7" ], "author": { "name": "Matthew Nagy", "email": "matthew.nagy@sony.com", "time": "Thu May 28 10:02:36 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu May 28 10:05:06 2026" }, "message": "[TySan] Expose __tysan_set_type_unknown interface (#198800)\n\nThis can help work around issues like\n[#143587](https://github.com/llvm/llvm-project/issues/143587)\nThe function is renamed with two trailing underscores to match the\nnaming scheme of the other sanitizers.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 51d823197cb40a57f25d00882546374d460c649e\n" }, { "commit": "759a5cfacd51469570f50a5874895d1e5d9589e7", "tree": "3351a4e8f6abc3507676e21e6d16109ec4477a24", "parents": [ "08d6f615e79eed9ac02f0ae90a08c46a5021f01e" ], "author": { "name": "Simon Tatham", "email": "simon.tatham@arm.com", "time": "Thu May 28 09:42:05 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu May 28 10:04:48 2026" }, "message": "[compiler-rt][ARM] Optimized FP -\u003e integer conversions (#179927)\n\nThis commit adds a total of 8 new functions, all converting a\nfloating-point number to an integer, varying in 3 independent choices:\n\n* input float format (32-bit or 64-bit)\n* output integer size (32-bit or 64-bit)\n* output integer type (signed or unsigned)\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 174eb79572b3433aa338ede31815fa8fa6a13b29\n" }, { "commit": "08d6f615e79eed9ac02f0ae90a08c46a5021f01e", "tree": "ea9ba7fd18238457d8245852ce5e1bc75eacad22", "parents": [ "99238eeee4d28447e509f43427925dae40d89b7d" ], "author": { "name": "Yaxun (Sam) Liu", "email": "yaxun.liu@amd.com", "time": "Thu May 28 02:36:21 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu May 28 02:40:43 2026" }, "message": "[PGO][HIP] Stop pulling ROCm.o into every PGO host link (#200101)\n\nPR #177665 added an unconditional `extern` reference to\n`__llvm_profile_hip_collect_device_data` from `InstrProfilingFile.c`,\nwhich forces `InstrProfilingPlatformROCm.o` (and its sanitizer_common /\ninterception dependencies) out of `libclang_rt.profile.a` in every PGO\nbinary. That breaks bots without `-lpthread` and races dlsym/PLT state\nin non-HIP programs via the interceptor constructor.\n\nFix:\n- Declare the hook `COMPILER_RT_WEAK` and gate the call on its address.\nNo `COMPILER_RT_VISIBILITY`: a hidden weak-undef function would be\nnon-preemptible and the address test would fold to true.\n- Gate `installHipModuleInterceptors` on `dlsym(hipModuleLoad)` so the\nconstructor is a no-op if `ROCm.o` is still pulled in.\n\nFixes:\n- https://lab.llvm.org/buildbot/#/builders/66/builds/31311\n- https://lab.llvm.org/buildbot/#/builders/174/builds/36180\n\nVerified:\n- `check-profile` 134/134 pass.\n- `nm` on a non-HIP `clang -fprofile-generate` binary: zero\n`installHip`/`ROCm`/`sanitizer`/`hip_collect` symbols.\n- HIP offload PGO end-to-end on gfx1101 (compile → run → `llvm-profdata\nmerge` → `llvm-cov`) still works; interceptor installs, device profile\ncollected via shared API.\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 635e120fb87304924508a7a204574727e3c37363\n" }, { "commit": "99238eeee4d28447e509f43427925dae40d89b7d", "tree": "64c2e62ee31a7d7d0cbbbeb9ff509e6b6a933465", "parents": [ "c1ddb046eb137333d9d1395cdfb0cde2745dd9ae" ], "author": { "name": "Yaxun (Sam) Liu", "email": "yaxun.liu@amd.com", "time": "Wed May 27 23:47:30 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed May 27 23:50:39 2026" }, "message": "[PGO][AMDGPU] Add basic HIP offload PGO support (#177665)\n\nProvide the minimum HIP/offload path for device profile collection and\nmerge on HIP before layering profile-format and uniformity-specific\nchanges separately.\n\nThis adds the ROCm collection runtime, hooks device collection into the\nhost write-file path, lowers AMDGPU instrumentation to\n__llvm_profile_instrument_gpu with regular counters, and disables GPU\nindirect-call value profiling.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 5db13643f4b7038db0ca304d9f8900122502935c\n" }, { "commit": "c1ddb046eb137333d9d1395cdfb0cde2745dd9ae", "tree": "dbac9cfaa2198d3f03d47ddfd724cdc5280a5402", "parents": [ "8c1626a36680fc5c3f34f8ef0fbe56cd04b94f69" ], "author": { "name": "Simon Tatham", "email": "simon.tatham@arm.com", "time": "Wed May 27 10:55:37 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed May 27 11:00:40 2026" }, "message": "[compiler-rt][ARM] Optimized FP double \u003c-\u003e single conversion (#179926)\n\nThis commit provides assembly versions of the conversions both ways\nbetween double and float.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: ed11d7a52a505ff53523c2c635f114d865273f24\n" }, { "commit": "8c1626a36680fc5c3f34f8ef0fbe56cd04b94f69", "tree": "545387bd8ee82723f8aa0674c8376908d3ebdd50", "parents": [ "54ac4478b2cb4c783ace86b8180de1fcef80c194" ], "author": { "name": "Garvit Gupta", "email": "garvgupt@qti.qualcomm.com", "time": "Wed May 27 05:02:40 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed May 27 05:05:18 2026" }, "message": "[compiler-rt][ARM] Add missing SUPERSEDES for optimized FP comparison sources (#199604)\n\nPR #179924 and #179925 added optimized assembly implementations for ARM\ndouble-precision and single-precision FP comparisons (arm/cmpdf2.S,\narm/gedf2.S, arm/unorddf2.S, arm/cmpsf2.S, arm/gesf2.S, arm/unordsf2.S)\nbut only added SUPERSEDES annotations for the thumb1 variants. The arm\nvariants were missing these annotations, causing both the generic and\noptimized implementations to be included in libclang_rt.builtins.a.\n\nFor double-precision, the archive contains:\n - comparedf2.c.obj (pos 28): defines __unorddf2, __aeabi_dcmpun, ...\n - divdc3.c.obj (pos 32): defines __divdc3; refs __aeabi_dcmpun\n - unorddf2.S.obj (pos 126): defines __unorddf2, __aeabi_dcmpun\n - aeabi_dcmp.S.obj (pos 158): defines __aeabi_dcmpeq; refs __eqdf2\n\nWhen linking divdc3_test.c, the linker loads divdc3.c.obj which\nintroduces __aeabi_dcmpun as undefined. BFD-like linkers (GNU ld, ELD)\ncontinue scanning forward and resolve __aeabi_dcmpun from unorddf2.S.obj\n(pos 126). Later, aeabi_dcmp.S.obj introduces __eqdf2 as undefined,\nwhich is resolved by comparedf2.c.obj (pos 28) on the next pass. Since\nboth comparedf2.c.obj and unorddf2.S.obj define __unorddf2, the linker\nreports a duplicate symbol error.\n\nlld does not encounter this because of the difference in the way it\nresolves symbol references. This causes comparedf2.c.obj (pos 28) to be\nselected first for __aeabi_dcmpun, making unorddf2.S.obj unnecessary.\n\nThe same pattern exists for single-precision where arm/comparesf2.S and\narm/unordsf2.S both define __unordsf2 and __aeabi_fcmpun.\n\nThe fix adds SUPERSEDES annotations so that the generic implementations\n(comparedf2.c for double-precision and arm/comparesf2.S for single-\nprecision) are removed from the source list when the optimized assembly\nreplacements are present. The assembly files together provide all\nsymbols that the generic implementations define.\n\nThe surrounding code was reviewed, and this PR was developed with the\nassistance of AI.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: c85c8af5426eb1affa14388deb78a4731737bf04\n" }, { "commit": "54ac4478b2cb4c783ace86b8180de1fcef80c194", "tree": "5913eef5da1e00fb600a5e09a42ebc9be4afa3d9", "parents": [ "0f2d540c3176ca95cb7919301f1939919e50c58b" ], "author": { "name": "YunQiang Su", "email": "syq@debian.org", "time": "Tue May 26 05:57:22 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue May 26 06:00:55 2026" }, "message": "sanitizer_common: Fix build on MIPS with _TIME_BITS\u003d64 (#199590)\n\nWhen we build sanitizer_common with -D_TIME_BITS\u003d64, the assert of\nstruct_kernel_stat_sz fails due to the size of struct stat get different\nsize.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 44d60bad70c7832c2917724d183aa6c208fe589b\n" }, { "commit": "0f2d540c3176ca95cb7919301f1939919e50c58b", "tree": "35bb7087e41eb196bb6438e321add44f3a20989e", "parents": [ "d9c3f909cbe17ef17fab93212eec8c46571af19e" ], "author": { "name": "Aiden Grossman", "email": "aidengrossman@google.com", "time": "Mon May 25 23:54:49 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon May 25 23:55:43 2026" }, "message": "Revert \"[compiler-rt][ASan] Add function copying annotations (#91702)\" (#194204)\n\nThis reverts commit c76045d9bf3bd1c7a381dc85d1db63a38fd69aa4.\n\nIt does not look like this has been used anywhere since it was\nimplemented. I see no uses of it in LLVM, anywhere in our internal\nmonorepo, or across the entirety of Github outside of other copies of\nLLVM tests. Given that, remove it. The intended use case around SSO ASan\nstring annotations is also likely to be significantly reworked soon.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 269719b1f63c68fffbca43f54cebaad23cf0246a\n" }, { "commit": "d9c3f909cbe17ef17fab93212eec8c46571af19e", "tree": "a37405073ea95ab182c12b8aec63ee278d5ed184", "parents": [ "2b1936227ea38ef85e5b18579b69545251752d4f" ], "author": { "name": "Mark Zhuang", "email": "mark.zhuang@spacemit.com", "time": "Mon May 25 09:46:25 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon May 25 09:50:55 2026" }, "message": "[compiler-rt] Inline __NR_riscv_flush_icache instead of including linux/unistd.h (#193645)\n\nAvoid the dependency on kernel headers to get __NR_riscv_flush_icache,\nmirroring the existing approach used for __ARM_NR_cacheflush.\n\nAssisted-by: Claude Sonnet 4.6\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 48cec8344ab4edf284c80909fa00aff06fc8c806\n" }, { "commit": "2b1936227ea38ef85e5b18579b69545251752d4f", "tree": "48c4b9915b2182264d9727a2a89034b415e1286b", "parents": [ "1ab266094fd3f0ab59f2f624d12f9b1bc028c44f" ], "author": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Sun May 24 05:35:55 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Sun May 24 05:41:04 2026" }, "message": "[compiler-rt] Use `size_t` rather than `int` for first argument to `__atomic_load_c` et al. (#197519)\n\nI noticed this discrepancy in emscripten when trying to test 128 bit\natomics under wasm64:\nhttps://github.com/emscripten-core/emscripten/pull/26937\n\nThe LLVM CodeGen appears to use `size_t` in this position when it\ngenerates calls to these functions.\n\nThis doesn\u0027t effect other platforms I imagine because they don\u0027t require\nsignature checking at the linker level.\n\nThis doesn\u0027t effect wasm32 where size_t and int are the same size.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 78cbf39d3bc990e4562c98dbddd26cd3de4eca96\n" }, { "commit": "1ab266094fd3f0ab59f2f624d12f9b1bc028c44f", "tree": "fba37f0271c0b5515bd0636fe08bcaf2a20d0c6f", "parents": [ "a25e5fee87da31818643536b6cee168a2879f9b1" ], "author": { "name": "Alison Zhang", "email": "alisonzhang@ibm.com", "time": "Thu May 21 18:58:44 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu May 21 19:00:28 2026" }, "message": "[AIX][clang][compiler_rt] rename libatomic archive to libclang_rt (#197485)\n\nThis PR implements the following on AIX to avoid conflicts between LLVM\nlibatomic and the GNU libatomic in the AIX toolbox as they share the\nsame library name:\n\n- Updates the clang driver to use `-lcompiler_rt` instead of `-latomic`\n- Renames the compiler-rt archive from `libatomic.a` to\n`libcompiler_rt.a`\n\nOnly the archive and not the shared object (`libatomic.so.1`) is renamed\nbecause renaming one component is enough to distinguish between the LLVM\nand GNU libatomic libraries. This also allows us to add additional\nshared objects to the `libcompiler_rt.a` archive in the future if\nneeded.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: d5ba663c28ec6374f02e630f79459d4ef639f51c\n" }, { "commit": "a25e5fee87da31818643536b6cee168a2879f9b1", "tree": "747fe043a7ed6c1e6919725234195293980c5f0f", "parents": [ "1afe45ce59d3ea56d3f1bd4329907119b9891b6d" ], "author": { "name": "Brad Smith", "email": "brad@comstyle.com", "time": "Thu May 21 04:14:58 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu May 21 04:16:25 2026" }, "message": "[compiler-rt][builtins] A few fixes cpu_model files (#198957)\n\n- Fix typo in include guard with the word features\n- Correct header in cpu_model.h header file and include guard after\n the file has been renamed\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 1229385ace03c6120b2268a7a4bf80d3f56b2135\n" }, { "commit": "1afe45ce59d3ea56d3f1bd4329907119b9891b6d", "tree": "89056f1a4d095a720f3097c472f12fb3c9795290", "parents": [ "ae0a786ae297df4ef0fe418c058ac0a46e9685f9" ], "author": { "name": "Tom Stellard", "email": "tstellar@redhat.com", "time": "Wed May 20 18:45:42 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed May 20 18:50:41 2026" }, "message": "[test][msan] Update rss_limit for release_origin.c (#196565)\n\nThis fixes test failures on aarch64.\n\nFixes #171209\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: fd3e453aed4fbf655d7d37589572d82bd666e273\n" }, { "commit": "ae0a786ae297df4ef0fe418c058ac0a46e9685f9", "tree": "d4cf1d59fe46f5916f6fa4287b1ed08f0197ac67", "parents": [ "62796adebbb9ede615d573bac1d734f0229714f4" ], "author": { "name": "Chris Apple", "email": "cja-private@pm.me", "time": "Wed May 20 18:17:14 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed May 20 18:21:23 2026" }, "message": "[rtsan] Fix REAL call for preadv64 (#198563)\n\nSeemed to just be a typo matching the name of the intercepted function\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 064e529cd78e93bf50e0855daceabdb3a7d34953\n" }, { "commit": "62796adebbb9ede615d573bac1d734f0229714f4", "tree": "a9b9d5470d60a09c6f966752740695ecf24d5c34", "parents": [ "7331edcb52fdf91deb036eeca97a581c6fe3d4e0" ], "author": { "name": "Hans Wennborg", "email": "hans@hanshq.net", "time": "Tue May 19 13:15:39 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue May 19 13:20:53 2026" }, "message": "Revert \"[MC/DC][Coverage] Enable profile correlation for MC/DC\" (#198520)\n\nThe instrprof-mcdc-correlation.c test doesn\u0027t pass on Mac, see\ndiscussion on the PR.\n\nReverts llvm/llvm-project#136437\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 4a32cf0b6d90f6bec3b5f2ad50b11ca1dba62954\n" }, { "commit": "7331edcb52fdf91deb036eeca97a581c6fe3d4e0", "tree": "c70b65635b300167d8dba7eaf573ba3ea88900e8", "parents": [ "4bcd976da113a97fa2d55355a5fba046efd7b427" ], "author": { "name": "Sadaf Ebrahimi", "email": "sadafebrahimi@google.com", "time": "Mon May 18 19:38:53 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon May 18 19:40:52 2026" }, "message": "[scudo] Add unit test for StackDepot edge cases\n\nThis increases test coverage for stack_depot.h to 100%.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 0b4a877e0e37f28d5b4be589bed055d149ac88dd\n" }, { "commit": "4bcd976da113a97fa2d55355a5fba046efd7b427", "tree": "9ae2975554dd81fb6a160a7c1155e60f729ddf99", "parents": [ "6d7caef3875e666fd33a490ee02f59e4dc92fac9" ], "author": { "name": "Roman Beliaev", "email": "r.beliaev@ispras.ru", "time": "Mon May 18 15:22:44 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon May 18 15:26:02 2026" }, "message": "[MC/DC][Coverage] Enable profile correlation for MC/DC (#136437)\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 2021782464a016b645bdac0fe723979f9f7afdd2\n" }, { "commit": "6d7caef3875e666fd33a490ee02f59e4dc92fac9", "tree": "4a15f0354370a60b0fd0f099c71655888469162e", "parents": [ "337f935365241535b355ae2fe7b4f20e13027d27" ], "author": { "name": "Andrei Sabalenka", "email": "mechakotik@gmail.com", "time": "Sun May 17 22:19:05 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Sun May 17 22:21:05 2026" }, "message": "[compiler-rt] Fix StackDepot benchmark thread barrier (#197633)\n\nUse Param.Threads (number of worker threads) as barrier threshold\ninstead of Param.UniqueThreads (boolean that controls input generation).\n\nThis also silences\n[-Wbool-integral-comparison](https://github.com/llvm/llvm-project/pull/194180)\nwarning I\u0027m working on.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 3d3f4be0f13d69d46914c330895c9d5eff48a8d1\n" }, { "commit": "337f935365241535b355ae2fe7b4f20e13027d27", "tree": "ae15c2b0184e55641ff069a16559449f87223ad4", "parents": [ "c2e0f4a937ba36366bbc942088b3bd60cfae9dbf" ], "author": { "name": "Rainer Orth", "email": "ro@gcc.gnu.org", "time": "Sun May 17 19:23:22 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Sun May 17 19:26:13 2026" }, "message": "[sanitizer_common] Fix sanitizer_platform_limits_solaris.cpp compilation (#198158)\n\nWhen switching `clang++` to the default Solaris 11.4 compilation\nenvironment, XPG7 + extensions, `sanitizer_platform_limits_solaris.cpp`\nfails to compile:\n\n```\ncompiler-rt/lib/sanitizer_common/sanitizer_platform_limits_solaris.cpp:93:53: error: use of undeclared identifier \u0027ucontext_t\u0027; did you mean \u0027ucontext_t_sz\u0027?\n 93 | unsigned ucontext_t_sz(void *ctx) { return sizeof(ucontext_t); }\n | ^~~~~~~~~~\n | ucontext_t_sz\ncompiler-rt/lib/sanitizer_common/sanitizer_platform_limits_solaris.cpp:93:12: note: \u0027ucontext_t_sz\u0027 declared here\n 93 | unsigned ucontext_t_sz(void *ctx) { return sizeof(ucontext_t); }\n | ^\ncompiler-rt/lib/sanitizer_common/sanitizer_platform_limits_solaris.cpp:93:52: error: invalid application of \u0027sizeof\u0027 to a function type\n 93 | unsigned ucontext_t_sz(void *ctx) { return sizeof(ucontext_t); }\n | ^~~~~~~~~~~~\n\n```\n\nPreviously, `\u003csys/signal.h\u003e` would include `\u003csys/ucontext.h\u003e` due to a\nXPG4v2 requirement. Now the latter needs to be included directly.\n\nTested on `amd64-pc-solaris2.11` and `sparcv9-sun-solaris2.11`.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 0e92b557a78eddc7493ade4cfc0a3b6b93ee61cf\n" }, { "commit": "c2e0f4a937ba36366bbc942088b3bd60cfae9dbf", "tree": "5720dada904914a94c338221ee082e1a53b934a1", "parents": [ "c58d91bc9a5e62770dfb144ce852dfff2efdd682" ], "author": { "name": "Rainer Orth", "email": "ro@gcc.gnu.org", "time": "Sun May 17 19:22:35 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Sun May 17 19:26:01 2026" }, "message": "[sanitizer][test] Fix coverage-module-unloaded.cpp etc. on Solaris (#198164)\n\nWhen switching `clang++` to the default Solaris 11.4 compilation\nenvironment, XPG7 + extensions, two tests `FAIL`:\n\n```\n AddressSanitizer-i386-sunos :: TestCases/Posix/coverage-module-unloaded.cpp\n AddressSanitizer-i386-sunos-dynamic :: TestCases/Posix/coverage-module-unloaded.cpp\n\n SanitizerCommon-asan-i386-SunOS :: get_module_and_offset_for_pc.cpp\n SanitizerCommon-ubsan-i386-SunOS :: get_module_and_offset_for_pc.cpp\n SanitizerCommon-ubsan-x86_64-SunOS :: get_module_and_offset_for_pc.cpp\n```\n\nThe failure mode is the same in both cases: the tests fail to link with\n`main` undefined. This happens because `\u003csys/mman.h\u003e` defines\n\n```\n#define\tSHARED\t\t0x10\n```\n\nfor the benefit of the Solaris-specific `memcntl(2)`. This was\npreviously hidden since `clang++` defined `_XOPEN_SOURCE\u003d600`. This\ninterferes with the use of `SHARED` in the tests.\n\nWhile I\u0027m the first to admit that such a generic identifier is highly\nunfortunate, this is also true for the tests. The interface goes all the\nway back to Solaris 1.0/SunOS 4, thus isn\u0027t going to change.\n\nThis patch fixes this by following the lead of other tests that use\n`SHARED_LIB` instead of `SHARED` for this purpose.\n\nTested on `amd64-pc-solaris2.11` and `sparcv9-sun-solaris2.11`.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 598acfb82d27b583e7adb1526b3cfd27f2a84041\n" }, { "commit": "c58d91bc9a5e62770dfb144ce852dfff2efdd682", "tree": "6246a27254d5f3fcc2f9091ecb0c9bd10dc67807", "parents": [ "502748f263557173830714a3c56563324ca0048d" ], "author": { "name": "Rainer Orth", "email": "ro@gcc.gnu.org", "time": "Fri May 15 18:41:07 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 15 18:45:11 2026" }, "message": "[safestack][test] Fix sigaltstack.c on Solaris (#197967)\n\nThe `SafeStack-Standalone-*:: sigaltstack.c` test `FAIL`s on Solaris. It\nuses `MAP_STACK` which is unportable and on glibc systems just a no-op.\nTherefore this patch provides a fallback definition.\n\nTested on `x86_64-pc-solaris2.11` and `x86_64-pc-linux-gnu`.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 172480354ab7ba2078318b3a3d0087043d1cee97\n" }, { "commit": "502748f263557173830714a3c56563324ca0048d", "tree": "5359a4c7afa735d1a17f0c692c3d6102571d511c", "parents": [ "c4bd1bf7f7507e950d9601c31e38ac9d6d69801a" ], "author": { "name": "Brian Cain", "email": "brian.cain@oss.qualcomm.com", "time": "Fri May 15 15:22:26 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 15 15:25:40 2026" }, "message": "[Hexagon] Add XRay custom and typed event support (#191749)\n\nAdd support for XRay custom events (llvm.xray.customevent) and typed\nevents (llvm.xray.typedevent) for Hexagon.\n\nLLVM:\n* Add Hexagon to the architecture gate in SelectionDAGBuilder for\nxray_customevent and xray_typedevent intrinsic lowering\n* Implement EmitInstrWithCustomInserter for PATCHABLE_EVENT_CALL and\nPATCHABLE_TYPED_EVENT_CALL pseudo instructions\n* Implement LowerPATCHABLE_EVENT_CALL in HexagonAsmPrinter that emits\ninline sleds with jump-over, allocframe/deallocframe for LR:FP save,\nargument register save/restore, and call to the event handler\n* Add event pseudo dispatch in HexagonMCInstLower\n* Prevent event pseudos from being packetized (solo instructions)\n\ncompiler-rt:\n* Implement patchCustomEvent and patchTypedEvent in xray_hexagon.cpp to\npatch the sled jump to nop (enable) or back (disable)\n* Add __xray_CustomEvent and __xray_TypedEvent trampolines in the\nHexagon XRay trampoline assembly\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 8af0fc60c05bdc29e8175d3d7fc1bda175533e95\n" }, { "commit": "c4bd1bf7f7507e950d9601c31e38ac9d6d69801a", "tree": "332c4758707e253be5cc2894202488b10616c496", "parents": [ "48874b5bf8edccf1c2158534dfdbd4e8c0399ad3" ], "author": { "name": "Martin Storsjö", "email": "martin@martin.st", "time": "Fri May 15 08:45:53 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 15 08:50:57 2026" }, "message": "[compiler-rt] [arm] Fix compilation on non-ELF targets (#197880)\n\n0c539fc057b2e838dc30c7e3229110d0c0e168c8 added new assembly files,\nwith a literal \u0027.rodata\u0027 directive. This doesn\u0027t build for non-ELF\ntargets. Switch to the preexisting CONST_SECTION macro instead, which\nexpands to either \u0027.const\u0027, \u0027.section .rdata,\"rd\"\u0027 or \u0027.section .rodata\u0027\ndepending on platform.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: fd9aeed6aa8c15a2c0123fe6e034651443144019\n" }, { "commit": "48874b5bf8edccf1c2158534dfdbd4e8c0399ad3", "tree": "c5cca67546701589e0478f18a758ca61413fdd9e", "parents": [ "afc316755cd2b296f2bfc1fede07814b7956625f" ], "author": { "name": "Simon Tatham", "email": "simon.tatham@arm.com", "time": "Thu May 14 09:28:46 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu May 14 09:30:50 2026" }, "message": "[compiler-rt][ARM] Optimized single-precision FP comparisons (#179925)\n\nThese comparison functions follow the same structure as the\ndouble-precision ones in a prior commit, of a header file containing the\nmain logic and some entry points varying the construction of the return\nvalue.\n\nIn this case, we have provided versions for Thumb1 as well as\nArm/Thumb2.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 1cb92d817468c6fbe1b9c6905bcf84f712de742c\n" }, { "commit": "afc316755cd2b296f2bfc1fede07814b7956625f", "tree": "ccf5abf92c86f6c38823d93037d00dcf086c65fe", "parents": [ "6815b37391e693de9f132c772e5dca88f12d75d3" ], "author": { "name": "Simon Tatham", "email": "simon.tatham@arm.com", "time": "Thu May 14 09:03:32 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu May 14 09:05:33 2026" }, "message": "[compiler-rt][ARM] Optimized double-precision FP comparisons (#179924)\n\nThe structure of these comparison functions consists of a header file\ncontaining the main code, and several `.S` files that include that\nheader with different macro definitions, so that they can use the same\nprocedure to determine the logical comparison result and then just\ntranslate it into a return value in different ways.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: d2af73c9fe5ebf5c2671c127dfdd86877465740c\n" }, { "commit": "6815b37391e693de9f132c772e5dca88f12d75d3", "tree": "ee5b222feed91cf85e245f7dc50707a79bde056e", "parents": [ "d053136d3241b879d72be315c3ee00a9622b194b" ], "author": { "name": "Simon Tatham", "email": "simon.tatham@arm.com", "time": "Thu May 14 08:24:33 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu May 14 08:30:15 2026" }, "message": "[compiler-rt][ARM] Optimized double-precision FP mul/div (#179923)\n\nOptimized AArch32 implementations of `muldf3` and `divdf3` are provided.\nThe division function is particularly tricky because its Newton-Raphson\napproximation strategy requires a rigorous error bound. In this version\nof the commit I\u0027ve left out the full supporting machinery that validates\nthe error bound via Gappa and Rocq, but full details are provided via\nlinks to the upstream version of this code in the Arm Optimized Routines\nrepository, and to a pair of Arm Community blog posts.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 0c539fc057b2e838dc30c7e3229110d0c0e168c8\n" }, { "commit": "d053136d3241b879d72be315c3ee00a9622b194b", "tree": "15c44791c1185e2eed0257fe346910c7897f0d5d", "parents": [ "5995512911fffb29e307ee14277892ad8c91aba8" ], "author": { "name": "Yaxun (Sam) Liu", "email": "yaxun.liu@amd.com", "time": "Thu May 14 03:54:58 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu May 14 04:00:39 2026" }, "message": "[compiler-rt][profile][test] Match clang_rt.profile CRT model on MSVC (#197474)\n\nOn MSVC, Profile-* tests must link with the same CRT model as the\nclang_rt.profile static archive they exercise. When that archive pulls\nin RTInterception / RTSanitizerCommon object libraries, those are built\nwith MultiThreadedDLL (/MD), so the .objs reference `__imp_*` symbols.\nThe test binary defaults to /MT and fails to link with LNK2019\n(`__imp__stricmp` from `interception_win.cpp`) and LNK4098 default-lib\nconflicts.\n\nMatch the DLL CRT on the test side so test executables and the static\narchive use the same runtime. The change is gated on\n`COMPILER_RT_HAS_INTERCEPTION` and `!COMPILER_RT_PROFILE_BAREMETAL`, so\nconfigurations that don\u0027t pull interception into profile are unaffected.\n\nSplit out as NFC from #177665 per review feedback.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 923a29a1b4ba855c7504128ed3cfe177d709d73d\n" }, { "commit": "5995512911fffb29e307ee14277892ad8c91aba8", "tree": "f1cc19ab3a3e9b034ec42582a01e618d3414014d", "parents": [ "5aa06da7100d79855e44e368874b80ebf1cb0146" ], "author": { "name": "Sadaf Ebrahimi", "email": "sadafebrahimi@google.com", "time": "Thu May 14 00:11:01 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu May 14 00:16:21 2026" }, "message": "[scudo] Add test for initFlags()\n\nAdd a test case to verify that initFlags() correctly reads the\nSCUDO_ALLOCATION_RING_BUFFER_SIZE environment variable and updates the\ncorresponding flag. This increases line coverage for flags.cpp to 100%.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: c5f84148725e69bbc43831aaa88158db50f3722d\n" }, { "commit": "5aa06da7100d79855e44e368874b80ebf1cb0146", "tree": "2456b44266888f3b29d5374806935192e5644fad", "parents": [ "95a12cfab972ca8dbe0f3d5083becf89080379ca" ], "author": { "name": "Jordan R AW", "email": "ajordanr@google.com", "time": "Wed May 13 23:48:31 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed May 13 23:51:18 2026" }, "message": "[compiler-rt][cmake] Fix check_cxx_compiler_flag calls (#197529)\n\ncheck_cxx_compiler_flag, when passing multiple flags, we must separate\nthem using a SEMICOLON-separated list. Not spaces. These checks\nsucceed incorrectly sometimes because \"-Werror -mcrc\" has a different\nreturn value than \"-Werror\" \"-mcrc\" on some systems.\n\nThis issue was verified with LLVM_ENABLE_PROJECTS\u003dllvm;compiler-rt,\nand I\u0027m uncertain whether it exists in runtime CMake builds.\nNonetheless, it\u0027s still a bug.\n\nSee:\nhttps://cmake.org/cmake/help/latest/module/CheckCXXCompilerFlag.html\n\nThis issue was identified downstream in ChromiumOS.\n\nChromiumOS Bug:\nhttps://issuetracker.google.com/507177988\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 7e735ea180fe6d199dd71dacf4396bcbbb3bbb3c\n" }, { "commit": "95a12cfab972ca8dbe0f3d5083becf89080379ca", "tree": "83e3fbd83b2dec46163f2da931e7e7802450e015", "parents": [ "44d05ea87cf30c33d1774dbbe3bdc9872fa9fa4e" ], "author": { "name": "Sadaf Ebrahimi", "email": "sadafebrahimi@google.com", "time": "Wed May 13 22:57:09 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed May 13 23:00:52 2026" }, "message": "[scudo] Add unit tests for optional config flags\n\nThis patch adds checks for all optional configuration flags defined in\nallocator_config.def to allocator_config_test.cpp. This increases the\ncode coverage of allocator_config_wrapper.h by exercising the template\ninstantiations for these optional flags.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 8536f609854abb96d02a22a20a2772c8ee3337bc\n" }, { "commit": "44d05ea87cf30c33d1774dbbe3bdc9872fa9fa4e", "tree": "9340338cbb3cdc62647e9e9e6a47831bd6bdfcd0", "parents": [ "4d26592f71a9a0fb5c86f3d270e425232a94c65d" ], "author": { "name": "Jakob Koschel", "email": "jakobkoschel@google.com", "time": "Wed May 13 16:54:35 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed May 13 16:55:57 2026" }, "message": "[SafeStack] Add sigaction interceptor (#196968)\n\nCurrently SafeStack does not support sigaltstack in the sense that\nwhile the \"normal\"/safe stack is getting replaced by the kernel, the\nunsafe stack used during signal handling is unchanged.\n\nThis commit adds the interceptor for sigaction so we can intercept the\nfunction being called for signal handling to add the logic to swap the\nunsafe stack before entering the actual signal handler.\n\nPR Stack:\n\n* ➤ https://github.com/llvm/llvm-project/pull/196968\n* https://github.com/llvm/llvm-project/pull/196969\n* https://github.com/llvm/llvm-project/pull/196970\n* https://github.com/llvm/llvm-project/pull/196971\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 0e81f935151e7921b333b63bbb0ca65e605fe902\n" }, { "commit": "4d26592f71a9a0fb5c86f3d270e425232a94c65d", "tree": "538c29ffdc0b34ca8119d9010df9cb6a775371bb", "parents": [ "5aa86ebdad6aa6c6d74d003be3dd96a0102ed689" ], "author": { "name": "Christopher Ferris", "email": "cferris1000@users.noreply.github.com", "time": "Tue May 12 23:40:59 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue May 12 23:46:12 2026" }, "message": "[scudo] Add a generic allocation fuzzer. (#197317)\n\nThe fuzzer disabled memory tagging if supported since there is an\nunknown problem with trying to create an allocator instance and doing\nraw allocate/deallocate calls.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: d3831ef1eb178caea5298ad945369ee9eb7670f0\n" }, { "commit": "5aa86ebdad6aa6c6d74d003be3dd96a0102ed689", "tree": "3547607855fd7f4727fc6b5f886e6acb1454d4ea", "parents": [ "333ee9e0184057b6cf01b7e4c286e286f7db5609" ], "author": { "name": "Sadaf Ebrahimi", "email": "sadafebrahimi@google.com", "time": "Tue May 12 20:40:36 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue May 12 20:46:14 2026" }, "message": "[scudo] Move MemMap tests from to map_test.cpp\n\nThe tests VerifyGetResidentPages, VerifyReleasePagesToOS, and Zeros test\nMemMapT functionality and fit better in map_test.cpp where other MemMapT\ntests reside.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 939912e66b8e36f1564ecef21ed9fa4b9440fe67\n" }, { "commit": "333ee9e0184057b6cf01b7e4c286e286f7db5609", "tree": "61a2a1a900655472b1af45b5db5d091dae0eb029", "parents": [ "2b78843fddff01238e5d397395a081702c9e3407" ], "author": { "name": "Igor Kudrin", "email": "ikudrin@accesssoftek.com", "time": "Tue May 12 19:18:14 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue May 12 19:20:59 2026" }, "message": "Reland \"[Clang][CodeGen] Report when an alias points to an incompatible target\" (#195550)\n\nThis relands #192397, which was reverted in #194106. The new version\nincludes the following fixes:\n- Set an explicit triple in the `attr-alias.m` test because aliases are\nnot supported on Darwin.\n- Relax the check to only diagnose mismatches in return types and\nparameter lists, while ignoring exception specifications and other\nattributes.\n\nOriginal description follows:\n\nAdd checks to ensure that an alias and its target have compatible types:\n- Generate an error if a function alias points to a variable or vice\nversa.\n- Issue a warning for mismatches in function types.\n- Ignore type discrepancies for variables.\n\nThis behavior aligns with similar diagnostics in GCC.\n\nResolves: #47301\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 40da6920a0d71d49dfa2392b09153600b0759f5e\n" }, { "commit": "2b78843fddff01238e5d397395a081702c9e3407", "tree": "0cdfb4581fd73cb46aea2bdcf871acf6b97eed40", "parents": [ "0e61a368495bc3622422979a0f83ec1187b51c2c" ], "author": { "name": "Jake Egan", "email": "Jake.egan@ibm.com", "time": "Tue May 12 17:41:35 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue May 12 17:46:01 2026" }, "message": "[sanitizer_common] Implement address sanitizer on AIX: platform specific support (#131866)\n\nAdd recognition of AIX and some platform specific changes. This lays the\ngroundwork to implement AIX in sanitizer_common/asan.\n\nIssue: https://github.com/llvm/llvm-project/issues/138916\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: db436f826bb97a1a36d300eaa8f8f5317fb66686\n" }, { "commit": "0e61a368495bc3622422979a0f83ec1187b51c2c", "tree": "6b55f4f594bcaed398c2e9ddb9435c6aad492bf8", "parents": [ "b5e4fbff518f569a5100f379f52363c38b9c36b5" ], "author": { "name": "Dan Blackwell", "email": "dan_blackwell@apple.com", "time": "Tue May 12 14:41:16 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue May 12 14:46:10 2026" }, "message": "[ASan][Darwin] Make multiple_sigaltstack.cpp test use MINSIGSTKSZ (#197204)\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 9c4ff6e82edc00616d0324b4a437c0e15bf6cebf\n" }, { "commit": "b5e4fbff518f569a5100f379f52363c38b9c36b5", "tree": "af07fd08907fa365dd41ead2529efee7daf857b1", "parents": [ "39c1982a7cd016d75508c5858ee4f830c0fadc6e" ], "author": { "name": "Sadaf Ebrahimi", "email": "sadafebrahimi@google.com", "time": "Mon May 11 22:33:45 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon May 11 22:36:02 2026" }, "message": "[scudo] Add unit tests for common functions\n\nThis patch adds unit tests for isPowerOfTwo, computePercentage, and\nisAlignedSlow in common_test.cpp. These additions increase the test\ncoverage for common.h to 100%.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 0d552b1ffb5ac80bfaadd7d43afef04185888a00\n" }, { "commit": "39c1982a7cd016d75508c5858ee4f830c0fadc6e", "tree": "1eb77e66338c0769ae20b728b456ecca29a1b76b", "parents": [ "a1c21c9589eafe2891f09d997111bfb72946b92c" ], "author": { "name": "Paul Kirth", "email": "paulkirth@google.com", "time": "Mon May 11 21:11:00 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon May 11 21:15:38 2026" }, "message": "[compiler-rt][common] Only unmap stacks the runtime has actually mapped (#179000)\n\nWhen the sanitizer hasn\u0027t mapped the alternate signal stack, but the\nhost program has (like LLVM), the runtime still tries to unilaterally\nunmap the alternate stack. Instead, the runtime should just check if\nit\u0027s actually mmaped the alternate stack, and only unmap it if it has.\n\n---------\n\nCo-authored-by: Vitaly Buka \u003cvitalybuka@google.com\u003e\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 437803e4af2046801a0144404698796d4b0e55d2\n" }, { "commit": "a1c21c9589eafe2891f09d997111bfb72946b92c", "tree": "a8b3da21385bf10d14d06da802f33fb34620448a", "parents": [ "29a5a44edb2294815d2b20e76d3d5934737ec31a" ], "author": { "name": "Yaxun (Sam) Liu", "email": "yaxun.liu@amd.com", "time": "Fri May 08 18:30:56 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 08 18:35:25 2026" }, "message": "[compiler-rt][profile][NFC] Introduce INSTR_PROF_INSTRUMENT_GPU_FUNC macro (#196538)\n\nAdd a macro INSTR_PROF_INSTRUMENT_GPU_FUNC for the name of the GPU\nprofiling function __llvm_profile_instrument_gpu (added in #187136),\nfollowing the same pattern as INSTR_PROF_VALUE_PROF_MEMOP_FUNC. Use the\nmacro in both the declaration in InstrProfiling.h and the definition in\nInstrProfilingPlatformGPU.c.\n\nThis prepares the upcoming HIP/AMDGPU offload PGO patch (#177665) to use\nthe same macro when calling this function.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 0e4477b677f04d4970a413be954f6b6792b560df\n" }, { "commit": "29a5a44edb2294815d2b20e76d3d5934737ec31a", "tree": "0e4a7d9b75c50094fa540a30159c946896482264", "parents": [ "3de4c1e31fa344fb030e4d7345e0e344d5a9ec47" ], "author": { "name": "sujianIBM", "email": "98488060+sujianIBM@users.noreply.github.com", "time": "Fri May 08 12:20:04 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 08 12:25:51 2026" }, "message": "Exclude unsupported compiler-rt tests on z/OS. (#194437)\n\nThis PR excludes unsupported part (NAN, -NAN, INFINITY, -INFINITY) from\nthe following 2 compiler-rt tests on z/OS.\n```\ncompiler-rt/test/builtins/Unit/compiler_rt_scalbnl_test.c\ncompiler-rt/test/builtins/Unit/compiler_rt_logbl_test.c\n```\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 1bba0bfe1c2957ed21b712ffddeb9fd0348ff06a\n" }, { "commit": "3de4c1e31fa344fb030e4d7345e0e344d5a9ec47", "tree": "44587730b606bd9191c3f9fffd18480427817977", "parents": [ "635bbb935123826fd2cc48ba4974142a0813a873" ], "author": { "name": "Wenju He", "email": "wenju.he@intel.com", "time": "Fri May 08 07:31:57 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 08 07:36:12 2026" }, "message": "[Runtimes] Fix /clang: prefix warning for GNU-like clang on Windows (#192041)\n\nlibclc has configure warning on Windows:\nclang: error: no such file or directory:\n\u0027/clang:--target\u003damdgcn-amd-amdhsa-llvm\u0027 clang: error: no such file or\ndirectory: \u0027/clang:-print-target-triple\u0027\n CMake Warning at CMakeLists.txt:239 (message):\n Failed to execute `llvm-project/build/bin/clang.exe\n/clang:--target\u003damdgcn-amd-amdhsa-llvm /clang:-print-target-triple` to\n normalize target triple.\n\nSwitch to check CMAKE_C_COMPILER_FRONTEND_VARIANT because\n- CMAKE_C_SIMULATE_ID\u003dMSVC: true for both clang and clang-cl.\n- CMAKE_C_COMPILER_FRONTEND_VARIANT\u003dMSVC: true for clang-cl; false for clang.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 77b7183542f7f6b3b47a271324c2ac93feb8f811\n" }, { "commit": "635bbb935123826fd2cc48ba4974142a0813a873", "tree": "7b815d435fc3a44ebee9bfd2b97216d81f6e9f73", "parents": [ "36d66b1a8d309768c95d32cf877251d88afbfb0e" ], "author": { "name": "Andrew Haberlandt", "email": "ahaberlandt@apple.com", "time": "Thu May 07 18:23:45 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu May 07 18:26:30 2026" }, "message": "[compiler-rt] [Darwin] Internal lit shell support for Darwin.haswell-symbolication.cpp (#196380)\n\nSomehow, #196152 fixed a bug where the x86_64h feature wasn\u0027t getting\ncorrectly set and so some tests that weren\u0027t running before started\nrunning. One such test is\n[AddressSanitizer-x86_64-darwin.TestCases/Darwin.haswell-symbolication.cpp](https://green.lab.llvm.org/job/llvm.org/job/clang-stage1-RA-cmake-incremental/job/main/872/testReport/AddressSanitizer-x86_64-darwin/TestCases_Darwin/haswell_symbolication_cpp/),\nwhich appears to have never been updated for the internal lit shell.\n\nThe internal lit shell does not support sub-shells, so the typical\npattern appears to be to write results to a file and use\n`%{readfile:%t.whatever}`.\n\nrdar://176390171\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: f8a433157531ecd53043d5427396077d564e0807\n" }, { "commit": "36d66b1a8d309768c95d32cf877251d88afbfb0e", "tree": "b93f6cc43d9be987644d5524e1b6ddef2ae72782", "parents": [ "e3baeb8273f43906c436582e7d114da3f54bba61" ], "author": { "name": "Zack Johnson", "email": "zacklj89@gmail.com", "time": "Thu May 07 16:29:27 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu May 07 16:31:34 2026" }, "message": "[ASan] MSVC test compatibility fixes (#196319)\n\nFixes a few tests with the MSVC toolchain due to incompatibilities:\n\n1. `asan_and_llvm_coverage_test.cpp`: Adds MSVC-specific linker flags\n(/link /NODEFAULTLIB:libcmt ...) alongside the\nexisting clang-cl -Wl flags. It is probably better to convert these\nlater to substitution changes in `lit.cfg.py`, but that will require\nupstreaming more test changes.\n1. `debug_invalid_pointer_pair.cpp`: Marks the test `UNSUPPORTED: MSVC`\nbecause `-mllvm` and codegen isn\u0027t supported with MSVC.\n1. `debug_memcpy_overlap.cpp` : Adds `/Oi` (enable intrinsics) when\nbuilding with MSVC, so `memcpy` is emitted as a call\nthat ASan can intercept.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 00d43e4af0c976899722fae1e30513e67ac66460\n" }, { "commit": "e3baeb8273f43906c436582e7d114da3f54bba61", "tree": "586ad684cd1d9d1c40f25cea1a34cc6f72c1765b", "parents": [ "94c3caeb79770b3a4c95d527691964ee25eb97b5" ], "author": { "name": "Victor Mustya", "email": "victor.mustya@intel.com", "time": "Thu May 07 12:40:58 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu May 07 12:42:25 2026" }, "message": "[compiler-rt] Enable ubsan_minimal runtime for SPIR-V (#196149)\n\nThe SPIR-V target support for UBSan Minimal Runtime is added to provide\nbasic undefined behavior detection capabilities for SPIR-V based\ndevices.\n\nCurrently, only the `spirv64` target is supported.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: ce16067aab54ae04f98e6ef5edd1bd7a68d33d90\n" }, { "commit": "94c3caeb79770b3a4c95d527691964ee25eb97b5", "tree": "7a4b1c06196743ecb9c828c3526792a7db2415ca", "parents": [ "fc9a115a3bd208a64d7b299f9dbe59c71189fe33" ], "author": { "name": "Andrew Haberlandt", "email": "ahaberlandt@apple.com", "time": "Wed May 06 21:01:56 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed May 06 21:06:05 2026" }, "message": "[reland] [lit] [compiler-rt] Add llvm-lit global command cache to speed up test config (#196152)\n\nRe-lands #195888\n\nFixes two issues:\n- `date -Ins` is not available on older macOS versions (I think\npre-15.4). This caused the new `test_cache` test to fail. Switched to\njust using `date` + a sleep (with a comment explaining why). Even if the\nsleep is too long/short, the test should still pass.\n- `functools.cache` is not available on Python 3.8. I\u0027ve moved the\n`_memoize` helper out of TestRunner.py into util.py, and switched to it\ninstead. I had to make a small change to the memoize helper to support\narbitrary args/kwargs.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: cb4e752a6dcb9700d86f715d9682dd8a47d61376\n" }, { "commit": "fc9a115a3bd208a64d7b299f9dbe59c71189fe33", "tree": "da67bcb7a9132872f590c38e4446247a0c166442", "parents": [ "c4469dd477f8ca255a31d10c68d876e5c97418bd" ], "author": { "name": "Aiden Grossman", "email": "aidengrossman@google.com", "time": "Wed May 06 18:34:44 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed May 06 18:40:29 2026" }, "message": "Revert \"[lit] [compiler-rt] Add llvm-lit global command cache to speed up test config\" (#196139)\n\nReverts llvm/llvm-project#195888\n\n`functools.cache` is not available until Python 3.9 whereas LLVM\u0027s\nminimum python version is 3.8.\n\n`date -Ins` also seems nonportable.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 103d6c9c7a7a4fe9254b9cf3b75a27401b07bcd7\n" }, { "commit": "c4469dd477f8ca255a31d10c68d876e5c97418bd", "tree": "7a4b1c06196743ecb9c828c3526792a7db2415ca", "parents": [ "41c134e9dfd8657ba1ff37de0f09106230eb5d24" ], "author": { "name": "Sadaf Ebrahimi", "email": "sadafebrahimi@google.com", "time": "Wed May 06 17:06:16 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed May 06 17:11:39 2026" }, "message": "[scudo] Dumping allocator config when printStats()\n\nSo far printStats() dumps partial information of the allocator config.\nGiven that user is able to have custom config, we want to dump all the\nconfigurations to reduce variances while reviewing the stats.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: c52158612b4563711ae60a92231a53d79c35545e\n" }, { "commit": "41c134e9dfd8657ba1ff37de0f09106230eb5d24", "tree": "73c4927f6e4931825994be323c95305e8c40c608", "parents": [ "1e4f5d3db18d07a27162641c468189d3b3cc0437" ], "author": { "name": "Andrew Haberlandt", "email": "ahaberlandt@apple.com", "time": "Wed May 06 16:32:27 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed May 06 16:36:27 2026" }, "message": "[lit] [compiler-rt] Add llvm-lit global command cache to speed up test config (#195888)\n\nCompiler-rt lit test discovery takes quite a while on Darwin (i.e. the\ntime from when you launch llvm-lit to when the first test runs can be\nminutes for check-compiler-rt). This appears to be mostly due to\nsubprocess calls during test configuration.\n\nThis adds a memoized command-runner to llvm-lit, so that lit.cfg.py\nscripts can re-use the result of feature-detection commands when running\nmultiple test suites at once. I\u0027ve adopted it for several subprocess\ncalls in compiler-rt.\n\nrdar://175893448\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 4e007c117e00efceacb9cd0560544491d1474106\n" }, { "commit": "1e4f5d3db18d07a27162641c468189d3b3cc0437", "tree": "88e815c2cb194fff4290fdaa6030377350451d12", "parents": [ "b70ed9037cbee75622128bfc683b404aa13a07bc" ], "author": { "name": "David Justo", "email": "david.justo.1996@gmail.com", "time": "Tue May 05 16:36:24 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue May 05 16:40:44 2026" }, "message": "[NFCI] clarify that asan-*linux.cpp files affect *nix OS\u0027es (#195565)\n\n**Prior Work:** Aims to supersede (#132263), which seems inactive,\nspecifically by applying my own comment:\nhttps://github.com/llvm/llvm-project/pull/132263#issuecomment-3051238734\n\n**Context:** It aims to minimally document that the\n`asan_(malloc_)?linux.cpp` files may impact non-linux OS\u0027es (despite the\nname) such as Solaris, BSD, and other *nix OS\u0027es. This is worth\ndocumenting as otherwise we risk breakage due to confusion, as occurred\n[here](https://github.com/llvm/llvm-project/pull/131975#issuecomment-2741097471).\n\nThis is done simply by minimally augmenting the file header comment\nsaying precisely this.\nUnlike the prior PR, this does not rename any files, which should reduce\nthe \u0027git noise\u0027 impact of this change.\n\n_Thanks!_\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: e4e41984bd92d3afb363ce4ae03c3dbf291b10c1\n" }, { "commit": "b70ed9037cbee75622128bfc683b404aa13a07bc", "tree": "0d818ee21536dfb77e8c85966f6de870172a6a90", "parents": [ "99ede8c3848f11105426bb8a136d1e148120df2f" ], "author": { "name": "Yaxun (Sam) Liu", "email": "yaxun.liu@amd.com", "time": "Tue May 05 02:42:33 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue May 05 02:45:39 2026" }, "message": "[compiler-rt] Add MSVC CRT flags to ASan coverage test (#195719)\n\nThe ASan coverage test also links clang_rt.profile because it uses\n-coverage. On Windows MSVC, this can fail if the test is linked with the\nstatic CRT but clang_rt.profile was built with the dynamic CRT.\n\nFor example, a profile runtime that uses /MD may reference DLL CRT\nsymbols\nsuch as __imp_* symbols. Add the same dynamic CRT flags directly to this\none test on Windows MSVC so it keeps linking when clang_rt.profile\nchanges.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: e37ca02de89706cd05e77cba9bd7125f6839a722\n" }, { "commit": "99ede8c3848f11105426bb8a136d1e148120df2f", "tree": "03825f6a905b97b5ae91f6e9f959bddc2aaef626", "parents": [ "6be57d2015982f92996e1d063d8a49d643ce61f2" ], "author": { "name": "Vitaly Buka", "email": "vitalybuka@google.com", "time": "Tue May 05 00:05:38 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue May 05 00:10:58 2026" }, "message": "[asan] Add HINT prefix to poison_history_size suggestions (#195733)\n\nConsistency update to prefix poison tracking suggestions with \"HINT:\"\nto match other ASan diagnostic hints.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 88ec045b5bac8a97b93a3a9b8c81eefb0786a28f\n" }, { "commit": "6be57d2015982f92996e1d063d8a49d643ce61f2", "tree": "a9fa4ab78f10f977d893989cd95cc2b61ab4a7ba", "parents": [ "c2f268526e867e7a44e3d73d1945f68927b72b4e" ], "author": { "name": "Vitaly Buka", "email": "vitalybuka@google.com", "time": "Mon May 04 23:50:43 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon May 04 23:55:37 2026" }, "message": "[asan] Only suggest increasing poison_history_size if the buffer is full (#195732)\n\nI unlikely but possible to setup shadow state, e.g unpoison heap red\nzone just after partial granule.\n\nIf buffer is not full, increasing it will not help.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 0a62c541467eb15ebd39de379b23492e691dc965\n" }, { "commit": "c2f268526e867e7a44e3d73d1945f68927b72b4e", "tree": "9e88bdcb9062bd1bf49d246a9f2670b0c2523831", "parents": [ "7d18573e9eecfb953c8cb8a6784f22fe1588bd22" ], "author": { "name": "Vitaly Buka", "email": "vitalybuka@google.com", "time": "Mon May 04 23:45:35 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon May 04 23:50:46 2026" }, "message": "[NFC][asan] Re-use ErrorGeneric::shadow_val (#195684)\n\nIt\u0027s done to reduce code size, as additional load on reporting is\nnegligible.\n\nThis requires change in ErrorGeneric constructor, to now we skip partial\nshadow only if next one has a better value.\n\nBefore shadow_val was either `0x[a-f].` or 0.\nNow it may be partial granule as well.\n\nHowever it\u0027s NFC as `0 \u003c shadow_val \u003c ASAN_SHADOW_GRANULARITY` does not\naffect reporting outside of `CheckPoisonRecords`.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 11317352614b0009d5bd265ac1d8c0c6b8e207de\n" }, { "commit": "7d18573e9eecfb953c8cb8a6784f22fe1588bd22", "tree": "f87a1ba8d0216ae32e667bd8938268f9fb354236", "parents": [ "46c20847873972cb3e53f2c57b1984917fd34b47" ], "author": { "name": "Alexander Richardson", "email": "alexrichardson@google.com", "time": "Mon May 04 23:34:08 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon May 04 23:36:07 2026" }, "message": "[asan] Remove XFAIL from tests that are passing with debug checks\n\nThese tests have been passing with COMPILER_RT_DEBUG\u003dON since 2022 (at\nleast for x86).\n\nFixes https://github.com/llvm/llvm-project/issues/46206\n\nPull Request: https://github.com/llvm/llvm-project/pull/84614\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: f68a900e8a8ce6e5632346d743545c1a7dd3b5a2\n" }, { "commit": "46c20847873972cb3e53f2c57b1984917fd34b47", "tree": "383d51e763ea65a590b3a00443b184853a86a80b", "parents": [ "8aa63fa5160ff8c8a80fa257393089b7f3c76d22" ], "author": { "name": "Steven Wu", "email": "stevenwu@apple.com", "time": "Mon May 04 23:03:25 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon May 04 23:05:30 2026" }, "message": "[compiler-rt][test] Disable create_thread_loop2 for lsan on Darwin (#195753)\n\ncreate_thread_loop2 occasionally hangs on macOS till hitting timeout.\nDisable the tests for LSAN on macOS.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 53e3ca1612d9aff5fa3e7353b66d69f76332b834\n" }, { "commit": "8aa63fa5160ff8c8a80fa257393089b7f3c76d22", "tree": "8785fc9b1fe6d17675f82d01218396c257ba60d3", "parents": [ "bcbfad0b49947e1ddb078dad24ed337e4fbeee19" ], "author": { "name": "Ian Anderson", "email": "iana@apple.com", "time": "Mon May 04 22:31:07 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon May 04 22:35:35 2026" }, "message": "[cmake][compiler-rt][darwin] builtin libraries don\u0027t build for armv6m in Darwin (#195372)\n\ndarwin_add_builtin_libraries tests for _Float16 and __bf16 for the host\narchitecture rather than the one being built, add -arch to fix that so\nthat armv6m correctly reports that it does not support __bf16.\ncfcmp/cdcmp get \"error: unsupported relocation type\" on their \"Branch to\ntarget address\" to c{f,d}cmple. Switch those to \"Call a subroutine\"\ninstructions on Thumb-1 (e.g. armv6m).\n\nAssisted-by: Claude Code\n\nrdar://167828904\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 6e3ab87e38e770119d9b776b7ab98489bcb8bc46\n" }, { "commit": "bcbfad0b49947e1ddb078dad24ed337e4fbeee19", "tree": "1882307eeb2eaaa8f20fed28e1e7a90b51fac90b", "parents": [ "a03f00a7575f668d573e51859803de5f9c522446" ], "author": { "name": "Vitaly Buka", "email": "vitalybuka@google.com", "time": "Mon May 04 22:28:43 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon May 04 22:30:56 2026" }, "message": "[asan] Record container poisoning in poison history (#195674)\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: cafb64410fd3e19f128ac0cbaeee60e255f767ae\n" }, { "commit": "a03f00a7575f668d573e51859803de5f9c522446", "tree": "fe43827ef53e6e91e817d50719bd475fcc3cfd6b", "parents": [ "95e97c09131d24c5ce422bbdcfe85d668f5d9268" ], "author": { "name": "Vitaly Buka", "email": "vitalybuka@google.com", "time": "Mon May 04 21:53:19 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon May 04 21:55:45 2026" }, "message": "[sanitizer] Capitalize HINT in reports (#195734)\n\nConsistency with other sanitizer reports which use uppercase \"HINT:\".\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 9aa55b291e6da18432f2c502367daa9a3ff4642e\n" }, { "commit": "95e97c09131d24c5ce422bbdcfe85d668f5d9268", "tree": "5b1323758ae2d75c2b52f69ef2054fe46a737e19", "parents": [ "bd4b378e8f127c021343273e2ab56b0fec33a7cd" ], "author": { "name": "Vitaly Buka", "email": "vitalybuka@google.com", "time": "Mon May 04 21:39:59 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon May 04 21:45:42 2026" }, "message": "[asan] Add tests for __sanitizer_annotate_double_ended_contiguous_container (#195673)\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 9792c778d0fc58d84b42c97b7f76f782705d72cd\n" }, { "commit": "bd4b378e8f127c021343273e2ab56b0fec33a7cd", "tree": "ec9ea134d116377e8b5ff133c7018b57cc722204", "parents": [ "dc80f93d9035134a48d564f9e48e2a7a77a10300" ], "author": { "name": "Vitaly Buka", "email": "vitalybuka@google.com", "time": "Mon May 04 21:02:45 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon May 04 21:05:40 2026" }, "message": "[NFC][asan] Use asan_interface.h in contiguous_container_crash.cpp (#195671)\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 9e095580756a0aba8cbaeb51a2f3b5ec2f1c8420\n" }, { "commit": "dc80f93d9035134a48d564f9e48e2a7a77a10300", "tree": "6e8f8dbdfdb7512151e50132dd32c2cd3932cec1", "parents": [ "cc34d83191e1ceeca6d5e983c97cc3e0a17b25c3" ], "author": { "name": "Vitaly Buka", "email": "vitalybuka@google.com", "time": "Mon May 04 19:31:58 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon May 04 19:36:15 2026" }, "message": "[asan] Print standalone partial granules in poison_history_size (#195723)\n\nCurrently, `poison_history_size` only reports the poisoning stack trace\nif the shadow value is exactly `kAsanUserPoisonedMemoryMagic`.\n\nUnlike heap or stack poisoning, where red-zones\nguaranteed, user poisoning more likely end-up with\nsingle standalong granule. Printing history could be helpful.\n\nMerge https://github.com/llvm/llvm-project/pull/195670 into main.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 33a4c4940c66dd2bf34ec24aa0c26c59b4c8eead\n" }, { "commit": "cc34d83191e1ceeca6d5e983c97cc3e0a17b25c3", "tree": "e20ff52ed3c3d12f5ecabf9aa03fe70038ff803b", "parents": [ "d2c40935ee1ab182eb83a5503c534cf799ed0567" ], "author": { "name": "Vitaly Buka", "email": "vitalybuka@google.com", "time": "Mon May 04 18:48:35 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon May 04 18:50:54 2026" }, "message": "[NFC][asan] Extract out RecordPoison helper function (#195672)\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 98419d22153932cceb7808a0d00e347571de500c\n" }, { "commit": "d2c40935ee1ab182eb83a5503c534cf799ed0567", "tree": "4232fe51faac98b1faa682c053fd3c1638e6a43a", "parents": [ "98991ee0bf9f9580dd8fa567874fda0f3c3bce1d" ], "author": { "name": "Vitaly Buka", "email": "vitalybuka@google.com", "time": "Mon May 04 18:26:48 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon May 04 18:31:21 2026" }, "message": "[NFC][asan] Cleanup use-after-poison-history-size-partial-granule.cpp test (#195667)\n\nUpdate the test to use more robust argument passing for offsets and\nsizes.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 8e341130d9c740a17e6321810204438f25963429\n" }, { "commit": "98991ee0bf9f9580dd8fa567874fda0f3c3bce1d", "tree": "b4e911befc6fef131b095448782d2eac902704e7", "parents": [ "81c00b38c7693bcb69e54a22ab7d8b99bbd3aa01" ], "author": { "name": "Vitaly Buka", "email": "vitalybuka@google.com", "time": "Mon May 04 17:51:03 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon May 04 17:55:27 2026" }, "message": "[asan] Change error to note when poison record is not found (#195669)\n\nWhen `CheckPoisonRecords` fails to find a record, it\u0027s often due to the\nhistory buffer being too small rather than a functional error in the\nlogic.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: c738bfa51edb4de09c4aec7eaa8c8e6849c789eb\n" }, { "commit": "81c00b38c7693bcb69e54a22ab7d8b99bbd3aa01", "tree": "c932bac7655ef224127534286042775398cf412d", "parents": [ "fa83127529daf724fd19b9181bbcb6ea66be8407" ], "author": { "name": "Vitaly Buka", "email": "vitalybuka@google.com", "time": "Mon May 04 17:39:01 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon May 04 17:40:42 2026" }, "message": "[asan] Improve manual poison reporting (#195666)\n\nAlways print the thread ID that poisoned the memory, even if the\nstack trace is unavailable.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: b8142ec5020b0c95a7a83f92adc7388adaa2f127\n" }, { "commit": "fa83127529daf724fd19b9181bbcb6ea66be8407", "tree": "72c7bffbeb771e37bd308a2e2e6c4f1bb4a292dc", "parents": [ "0408cce08083f3d81379ed7d9f5bd26c03e1495b" ], "author": { "name": "Nikita Taranov", "email": "nickita.taranov@gmail.com", "time": "Mon May 04 16:51:54 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon May 04 16:55:48 2026" }, "message": "Fix flaky test xray/basic-filtering.cpp (#186611)\n\nIncrease time thresholds and sleep time to decrease the probability of\nfailure.\n\nCloses: #175866\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: f306525759aeb6b9c081e1885849fcb5fca0575f\n" }, { "commit": "0408cce08083f3d81379ed7d9f5bd26c03e1495b", "tree": "5f13a120824ba988647f3d057c0154169a0e71b5", "parents": [ "01e3f87586e066371cee32fa73951733d1416516" ], "author": { "name": "Kleis Auke Wolthuizen", "email": "github@kleisauke.nl", "time": "Sun May 03 18:26:40 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Sun May 03 18:30:52 2026" }, "message": "[compiler-rt][UBSan] Add __ubsan_default_suppressions() hook (#194862)\n\nIn line with commit 5c62af5 and 83566da.\n\nAssisted-by: Gemini\n\n---------\n\nCo-authored-by: Vitaly Buka \u003cvitalybuka@google.com\u003e\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: aec3ce0ac5bdfa5db75db07e91220a1c374022e7\n" }, { "commit": "01e3f87586e066371cee32fa73951733d1416516", "tree": "cc437fc56cf1c6a39c9543e7928cd4d224204cf3", "parents": [ "b87c0f26db1eae15bb17a62166893ae22b50edbb" ], "author": { "name": "Kleis Auke Wolthuizen", "email": "github@kleisauke.nl", "time": "Sat May 02 20:46:35 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Sat May 02 20:50:25 2026" }, "message": "[compiler-rt][UBSan][NFC] Reorder includes in ubsan_diag.cpp (#195435)\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 13371e24812609de8dc65cd9fd36461037fb1581\n" }, { "commit": "b87c0f26db1eae15bb17a62166893ae22b50edbb", "tree": "d06b412d67cea8821646ee7fed7d3d3f8373a620", "parents": [ "ba0ac68c0f2cbd6e5861c743663fff1b1363899e" ], "author": { "name": "Vitaly Buka", "email": "vitalybuka@google.com", "time": "Sat May 02 02:09:55 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Sat May 02 02:12:09 2026" }, "message": "[LowerTypeTests] Add debug info to jump table entries (#194493)\n\n[LowerTypeTests] Add debug info to jump table entries (#192736)\n\nWhen Control Flow Integrity (CFI) is enabled, jump tables are used to\nredirect indirect calls. Previously, these jump table entries lacked\ndebug information, making it difficult for profilers and debuggers to\nattribute execution time correctly.\n\nNow stack trace, when stopped on jump table entry will looks like this:\n```\n#0: c::c() (.cfi_jt) at sanitizer/ubsan_interface.h:0:0\n#1: __ubsan_check_cfi_icall_jt at sanitizer/ubsan_interface.h:0\n```\n\nFollowing up on previous attempts #192736 and #193670, this PR is\nessentially #192736 but with the `(.cfi_jt)` and\n`__ubsan_check_cfi_icall_jt`\nframes swapped. While the specific order of `__ubsan_check_cfi_icall_jt`\nisn\u0027t strictly necessary, swapping them helps maintain existing\ndiagnostics\nbehavior.\nAdditionally, the diagnostics must remove `ubsan_interface.h` to allow\nfor a fallback to printing the module name.\nSee \"Commits\" tab for details.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 82bf5e84005f7506f7b98d08203eb990da1c45aa\n" }, { "commit": "ba0ac68c0f2cbd6e5861c743663fff1b1363899e", "tree": "f6a0ac75ab0359b7202f75fbf325c33d0df3ad16", "parents": [ "397865791f8ea02d9bf5758ec2c87162de2b189b" ], "author": { "name": "James Y Knight", "email": "jyknight@google.com", "time": "Fri May 01 21:06:30 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 01 21:11:33 2026" }, "message": "Fix test\u0027s FileCheck match to allow a column number in message. (#195356)\n\nIn some cases, the cfi diagnostic is emitted with a column\nnumber. Allow that in the test.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 1879af132627bbd326ce5cd82e45c6813a4727d8\n" }, { "commit": "397865791f8ea02d9bf5758ec2c87162de2b189b", "tree": "260c3cc4d6a8537b0eaab64876ce92151b68c588", "parents": [ "df2978763050eb86d648018e500b3631661ae79b" ], "author": { "name": "Sadaf Ebrahimi", "email": "sadafebrahimi@google.com", "time": "Fri May 01 19:32:23 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 01 19:35:50 2026" }, "message": "[scudo] Add append overloads for integers and bool\n\nThis patch adds overloads for append to support directly appending s32,\ns64, u32, u64, and bool values without requiring format strings. This\nsimplifies usage and avoids the overhead of parsing format strings for\nsimple type appends.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: f242ad87ff4496283896ae9c492dd67e511c501a\n" }, { "commit": "df2978763050eb86d648018e500b3631661ae79b", "tree": "55327d16ffddc78ebe4e61c25b6e6c2ff6cebb33", "parents": [ "5726037279937bf1e8d302c39fffe0e4d3872a99" ], "author": { "name": "Doug Gregor", "email": "dgregor@apple.com", "time": "Fri May 01 18:31:26 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 01 18:36:47 2026" }, "message": "[compiler-rt] Only build the embedded mach-o libraries when the ARM backend is available (#195313)\n\nFixes builds that only target AArch64, which I broke in #195184\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: ba2f42d5deb3b8392588515bb9b8f48690d6d816\n" }, { "commit": "5726037279937bf1e8d302c39fffe0e4d3872a99", "tree": "3711ec5d2ad27962fc8fe1a131a1594a76f57392", "parents": [ "e994282528d60d62d3c7bd5db590423c27f3ede1" ], "author": { "name": "Ruoyu Zhong", "email": "zhongruoyu@outlook.com", "time": "Fri May 01 15:29:45 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 01 15:35:20 2026" }, "message": "[cmake] Refactor DIA SDK detection into FindDIASDK module (#160354)\n\nThis consolidates the DIA SDK detection logic from\n`{llvm,compiler-rt}/cmake/config-ix.cmake` into a new centralized,\nreusable `FindDIASDK.cmake` module.\n\nIn addition to code deduplication, it also helps to avoid hard-coded\nreferences to the DIA SDK location in `LLVMExports.cmake`, hence\nallowing a pre-built LLVM distribution for Windows to be used on another\nhost without requiring the DIA SDK location to be the same.\n\nFixes https://github.com/llvm/llvm-project/issues/86250.\nFixes https://github.com/llvm/llvm-project/issues/100372.\nFixes https://github.com/llvm/llvm-project/issues/111829.\nFixes https://github.com/llvm/llvm-project/issues/152268.\n\n---------\n\nSigned-off-by: Ruoyu Zhong \u003czhongruoyu@outlook.com\u003e\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 3545244333dfc64de756ad13f7e3792d79989898\n" }, { "commit": "e994282528d60d62d3c7bd5db590423c27f3ede1", "tree": "551e078740a5a322b6a494674f2cee9e52deffdc", "parents": [ "1de0eb7d9b439a9d82ea2a474d7a0cbbb994aa98" ], "author": { "name": "Doug Gregor", "email": "dgregor@apple.com", "time": "Fri May 01 03:02:55 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 01 03:05:29 2026" }, "message": "Add ARMv8-M and ARMv8.1-M support for Darwin (#195184)\n\nExtend Mach-O reading/writing and Clang\u0027s driver with support for\nARMv8-M and ARMv8.1-M. Then, build libclang_rt for armv8m.main and\narmv8.1m.main target triples.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 70b2f06ab493e29bb5dcc627bd016714dd21405f\n" }, { "commit": "1de0eb7d9b439a9d82ea2a474d7a0cbbb994aa98", "tree": "8c82ee417e47400c3997407fff139df548639878", "parents": [ "c415a8008b6cf0a4d6408135e00ef5cd69bd6fba" ], "author": { "name": "Vitaly Buka", "email": "vitalybuka@google.com", "time": "Tue Apr 28 17:55:18 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Apr 28 18:01:04 2026" }, "message": "[NFC][cfi] Update diagnostic tests for better location matching (#194557)\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 6d01ff1c503c8f1de7c86d5b94424695bc85d655\n" }, { "commit": "c415a8008b6cf0a4d6408135e00ef5cd69bd6fba", "tree": "4a51bdae9b09e4e9135657c4412cc8339645d1e0", "parents": [ "cb73d9b5ac5159e008792dcc4c7edb0507f7bdfc" ], "author": { "name": "Mariusz Borsa", "email": "wrotki@msn.com", "time": "Mon Apr 27 23:06:19 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Apr 27 23:12:12 2026" }, "message": "[asan_abi] Skip new __asan_get_report_* from ABI (#194463)\n\nPR #181446 (\"[asan] API for getting multiple pointer ranges\") added five\nnew __asan_get_report_{dealloc,dest,first,second,src}_address entries to\ncompiler-rt/lib/asan/asan_interface.inc without updating\nasan_abi_tbd.txt\nor implementing them in compiler-rt/lib/asan_abi/asan_abi.cpp. This\nbroke\nthe AddressSanitizerABI-arm64-darwin ::\nDarwin/llvm_interface_symbols.cpp\ntest, which diffs asan_interface.inc (minus asan_abi_tbd.txt) against\nthe\nsymbols actually exported by libclang_rt.asan_abi_osx.a.\n\nList the new symbols alongside the existing __asan_get_report_* entries\nso the stable-ABI test passes. The symbols remain unimplemented in the\nstable ABI library; this change only reflects that they are\nintentionally\nnot part of the stable ABI surface.\n\nAssisted-by: Claude Opus 4.7\n\nrdar://175286011\n\nCo-authored-by: Mariusz Borsa \u003cm_borsa@apple.com\u003e\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 6722bb7fe9b5e029c485bd0bc33269d553b7ef1a\n" }, { "commit": "cb73d9b5ac5159e008792dcc4c7edb0507f7bdfc", "tree": "d0ab58e617bc35be1c8dc5751865e1993edfcf9b", "parents": [ "cb9546b4ec1504170f8e5801aa75e6affd6592af" ], "author": { "name": "Vitaly Buka", "email": "vitalybuka@google.com", "time": "Mon Apr 27 00:17:50 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Apr 27 00:20:30 2026" }, "message": "Reland \"[sanitizer] Fix race condition in GetNamedMappingFd with decorate_pro…\"\" (#194273)\n\nReverts llvm/llvm-project#194271\n\nRelands llvm/llvm-project#190981.\n\nThreadID is u64, format must be `%llu`.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 652700b4cb3c87294f2d78cb87df5e394e589984\n" }, { "commit": "cb9546b4ec1504170f8e5801aa75e6affd6592af", "tree": "9796b31411ff7fdb6026e8aa8fe3a848a2c84214", "parents": [ "f91f4ceb9537017d62be83ca8d044264e72fcc2f" ], "author": { "name": "Thurston Dang", "email": "thurston@google.com", "time": "Sun Apr 26 23:01:07 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Sun Apr 26 23:03:13 2026" }, "message": "Revert \"[sanitizer] Fix race condition in GetNamedMappingFd with decorate_pro…\" (#194271)\n\nReverts llvm/llvm-project#190981 due to buildbot failure\n(https://lab.llvm.org/buildbot/#/builders/66/builds/29993):\n```\n SanitizerCommon-asan-i386-Linux :: Linux/decorate_proc_maps.cpp\n```\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 57494cc7b4e8b59714ee9e312812d8421f41d27c\n" }, { "commit": "f91f4ceb9537017d62be83ca8d044264e72fcc2f", "tree": "6c07c0456627f86d7fd2f1a59a0ada1946875d4a", "parents": [ "4c93fec099ec5346fcb96699571cfe8878d7eabd" ], "author": { "name": "martin0413133", "email": "129967631+martin0413133@users.noreply.github.com", "time": "Sun Apr 26 22:26:52 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Sun Apr 26 22:30:37 2026" }, "message": "[sanitizer] Fix race condition in GetNamedMappingFd with decorate_pro… (#190981)\n\n…c_maps\u003d1\n\nMulti-threaded programs crash randomly when\nASAN_OPTIONS\u003ddecorate_proc_maps\u003d1 is enabled due to filename collision\nin /dev/shm.\n\nRoot Cause:\nAll threads use the same filename format \u0027/dev/shm/\u003cPID\u003e [name]\u0027,\ncausing race conditions where one thread deletes a file created by\nanother thread, resulting in ENOENT errors.\n\nSolution:\nAdd thread ID (TID) to the filename to ensure uniqueness:\n- Old format: /dev/shm/\u003cPID\u003e [name]\n- New format: /dev/shm/\u003cPID\u003e.\u003cTID\u003e [name]\n\nThis ensures each thread has a unique filename, eliminating the race\ncondition.\n\nTesting:\n- Original version: 30% crash rate (6 crashes in 20 runs)\n- Fixed version: 0% crash rate (0 crashes in 50 runs)\n\nFixes #190604\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 13cee9be088057f198e08ee7217ed2af08cfd825\n" }, { "commit": "4c93fec099ec5346fcb96699571cfe8878d7eabd", "tree": "9796b31411ff7fdb6026e8aa8fe3a848a2c84214", "parents": [ "6239a4e22cd112ff0446eaab47b8fc2cb5e5e306" ], "author": { "name": "Midhunesh", "email": "midhunesh.p@ibm.com", "time": "Sun Apr 26 21:37:21 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Sun Apr 26 21:40:37 2026" }, "message": "[Asan]Add align argument to Realloc() (#194255)\n\nAdd align argument to the function Realloc() to ensure original\nallocation alignment through realloc\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 28c4c25c0cdf4bb41059fb48927f0f866b72bd5a\n" }, { "commit": "6239a4e22cd112ff0446eaab47b8fc2cb5e5e306", "tree": "03799a491045c66f25deadb1e8443d977b927871", "parents": [ "3fc32f63ba0be4b2ece992bdc55f015688fb9c1c" ], "author": { "name": "Matt Turner", "email": "mattst88@gmail.com", "time": "Sun Apr 26 01:28:22 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Sun Apr 26 01:30:44 2026" }, "message": "[compiler-rt] Use asm .set only for Hexagon (#194160)\n\nTwo incompatible assembler syntaxes exist for symbol assignment:\n```\n sym \u003d val -- accepted by most GNU assembler targets; rejected by\n Hexagon, which interprets it as a mnemonic\n .set sym, val -- accepted by Hexagon; rejected by Alpha, which\n reserves .set for assembler mode flags\n```\nSwitch all to `sym \u003d val`, and opt out Hexagon to `.set sym`.\n\nFixes: dbb03f8f606e (\"[compiler-rt] Replace assignment w/.set directive\n(#107667)\")\n\n---------\n\nCo-authored-by: Vitaly Buka \u003cvitalybuka@google.com\u003e\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 0ccb181514cda5245ed031b210a2d97282bc1ff0\n" }, { "commit": "3fc32f63ba0be4b2ece992bdc55f015688fb9c1c", "tree": "29259395eb29e9a8308b9bec67b878405ab8a794", "parents": [ "8cdae84669bad1f7a093ef3e7edb2ff30d9b2779" ], "author": { "name": "Aiden Grossman", "email": "aidengrossman@google.com", "time": "Sat Apr 25 21:24:40 2026" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Sat Apr 25 21:30:57 2026" }, "message": "[Profile] Reenable instrprof-tmpdir.c (#194181)\n\nenv -u is supported by the internal shell which is now the default\neverywhere.\n\nNOKEYCHECK\u003dTrue\nGitOrigin-RevId: 28de393b6412f4149a25029f4201dc7ebdb3c9f4\n" } ], "next": "8cdae84669bad1f7a093ef3e7edb2ff30d9b2779" }