Google GitGoogle Git
Sign in
llvm/llvm-project/mlir/HEAD
  1. 0348b0f [mlir][Diagnostics] Set parallel diagnostic order ids without a lock (#232520) by Bruno Cardoso Lopes · 3 hours ago main master
  2. a83bed5 [MLIR][AsmPrinter] Add alwaysUseAliases printing flag (#232518) by Bruno Cardoso Lopes · 4 hours ago
  3. 1b75ece [APFloat][NFCI] Move x87DoubleExtended next to the other native IEEE types (#223193) by Eric Ross · 5 hours ago
  4. 5fbbb35 [mlir][acc] Verify no nested acc kernel environment (#232503) by Razvan Lupusoru · 7 hours ago
  5. 31dc4fd [mlir][acc] Add compute kernel launch codegen to ACCToLLVM (#232382) by Razvan Lupusoru · 9 hours ago
  6. 62d5a1e [mlir][Arith] Move ceil/floor div expansion to Conversion/ArithCommon. NFC (#231064) by Fangrui Song · 11 hours ago
  7. 34427d4 [mlir][vector] Preserve dynamic mask bounds in shape cast folding (#229699) by purnima shrivastava · 12 hours ago
  8. 0c7b8ab [mlir][scf] Preserve forall shared output order in promote (#222144) by hanbeom · 13 hours ago
  9. 8ce4607 [mlir][Vector] Add a dedicated Transform Dialect op for `UnrollContractPattern` (#230543) by Andrzej Warzyński · 13 hours ago
  10. b7ab091 [mlir][tosa] Remove unnecessary conv verifier checks (#231825) by Luke Hutton · 16 hours ago
  11. c07bbee [MLIR][Remark] Report why the LLVM remark streamer could not be created (#230130) by Henrich Lauko · 18 hours ago
  12. 84a8b07 [mlir][OpenMP] Model memory effects of omp.critical (#230455) by JRP-arm · 19 hours ago
  13. 5f2fdef [mlir][OpenMP] Model recursive memory effects of omp.master (#230190) by JRP-arm · 20 hours ago
  14. 83edf07 [MLIR][Linalg] Add support for folding reduce_or and reduce_and of a broadcast (#227013) by Chibuoyim (Wilson) Ogbonna · 20 hours ago
  15. b640771 [MLIR][NVVM] Add support for pzo in FP8/FP6/FP4 convert Ops (#229660) by Dharuni R Acharya · 21 hours ago
  16. 0364b43 TableGen: Store classes and defs in a StringMap (#231573) by Matt Arsenault · 21 hours ago
  17. 28f0709 [mlir][math] Expose math patterns as transforms (#230042) by Adam Siemieniuk · 22 hours ago
  18. ac3a042 [mlir][nvgpu] Preserve unsigned types in MMA lowering to NVVM (#224837) by Vaisman · 22 hours ago
  19. 3ada166 [mlir][SPIR-V] Support CPacked decorations in binary round trips (#232192) by Alexander Shaposhnikov · 22 hours ago
  20. 4f9575f [mlir][bufferization] Skip read collection when no writes exist (#231771) by Akun · 26 hours ago
  21. da7f3a9 [mlir][NFC] Simplify OpInterfaceRewritePattern using-decl to use alias (#232056) by Benjamin Wu · 28 hours ago
  22. 99ee33a [mlir] `HasParallelRegion` means region may be parallel (#231372) by Matthias Springer · 28 hours ago
  23. 7f97c6f [mlir][memref] Fix the allocation interface model for realloc (#231608) by Matteo Di Iorio · 28 hours ago
  24. 1d533f1 [mlir][scf] Fix front peeling skipped due to float rounding in trip count (#231100) by Prashanth · 28 hours ago
  25. f67d7ed [mlir][scf] Fix front-peel unsigned loops with at most one iteration (#231542) by Prashanth · 29 hours ago
  26. 8e085a0 [mlir][bufferization] Fix bug when moving dealloc operation (#229296) by Pranav Alavandi · 29 hours ago
  27. 4c21662 [mlir][acc] Add map flag for mappings that need no device copy of its descriptor (#232064) by Vijay Kandiah · 32 hours ago
  28. 56601c8 [mlir][acc] Add data clause deduplication pattern for kernel env (#231965) by Razvan Lupusoru · 33 hours ago
  29. 308b69f [mlir] Fix abort on string token in integer dense tensor parsing (#231876) by Wijkqwe · 34 hours ago
  30. af77f79 [MLIR][XeGPU] Rename Xe3 uArch to Xe3p and add 2D block memory restrictions (#229936) by Sang Ik Lee · 35 hours ago
  31. 4c86b15 [mlir][Vector] Promote buffers accessed via dynamic subviews (#217793) by Jianhui Li · 35 hours ago
  32. accee2f [mlir][OpenMP] Model recursive memory effects of omp.masked (#230493) by JRP-arm · 2 days ago
  33. 5135fe8 [mlir][Arith] Rename Conversion/ArithCommon to ArithAttrToLLVM. NFC (#231743) by Fangrui Song · 2 days ago
  34. b7ec026 [mlir][DataLayout] Add getTypeAllocSize (#231262) by Adam Smith · 2 days ago
  35. 9d75d42 [mlir][flang][OpenMP] Select iterators per locator and allow multi-block regions (#227454) by Chi-Chun, Chen · 2 days ago
  36. c6ecde9 [mlir][vector] Expose vector linearize as transforms (#231412) by Adam Siemieniuk · 2 days ago
  37. ca9d51d [mlir][SPIRV] Support Alignment and FuncParamAttr on parameters (#231069) by Alexander Shaposhnikov · 2 days ago
  38. d4a05fb [mlir][tosa] Lower MATMUL_T to Linalg (#222661) by Jack Frankland · 2 days ago
  39. 6c967bd [mlir][SPIR-V] Add ComplexToSPIRV lowering for complex.sign (#216722) by Arseniy Obolenskiy · 2 days ago
  40. 3adf85e [mlir][vector] Document dimension kinds in `vector.contract` (NFC) (#230458) by Federico Bruzzone · 2 days ago
  41. 477caff [mlir][SPIR-V] Derive Aligned memory operand for PhysicalStorageBuffer in vector.load/store (#214200) by Arseniy Obolenskiy · 2 days ago
  42. 748aa3d [mlir][SPIR-V] Add SPIRVToLLVM conversions for SMulExtended and UMulExtended (#228052) by Arseniy Obolenskiy · 2 days ago
  43. c8d5979 [mlir][OpenMP] Keep loop allocas in the entry block after nested splits (#231270) by Sairudra More · 2 days ago
  44. d139559 [mlir][SPIR-V] Add saturating signed/unsigned integer conversions (#231642) by Alexander Shaposhnikov · 2 days ago
  45. 9cbdc69 [mlir][SPIR-V] Add OpenCL ctz, ilogb, and sincos operations (#231713) by Alexander Shaposhnikov · 2 days ago
  46. 30df7ff [mlir][SPIR-V] Support SaturatedConversion serialization (#231622) by Alexander Shaposhnikov · 2 days ago
  47. 8016a8b [mlir][vector] Do not unroll ops masked by vector.mask (#230461) by Vaisman · 2 days ago
  48. e83987d [mlir][SCF] Avoid linear scans when querying forall buffer aliases (#231187) by Akun · 2 days ago
  49. 6a0b6f6 [mlir][bufferization] Speed up DPS aliasing operand lookup (#231362) by Akun · 2 days ago
  50. f844e03 [mlir][python] Handle rank-zero MemRefs in NumPy conversion (#231403) by Anutosh Bhat · 2 days ago
  51. d05cebf [mlir][arith] Don't fold to negative values of unsigned float types (#230974) by SomnusGG · 2 days ago
  52. 15c2fc3 [MLIR][XeGPU] set_anchor_layout transform op supports packed arguments (#230135) by Tuomas Kärnä · 2 days ago
  53. 9b98b7f [MLIR][LLVM] Add range metadata to llvm.load (#228571) by Arseniy Obolenskiy · 3 days ago
  54. e9d8b67 [mlir][ptr] Lower vector pointer addition to LLVM GEP (#231080) by Alexander Shaposhnikov · 3 days ago
  55. 2ef38ba [mlir][EmitC] lower roundeven to __builtin_roundeven (#227822) by Colin Cross · 3 days ago
  56. 1df1388 [mlir][vector] Handle rank mismatch when unrolling broadcasts (#227351) by purnima shrivastava · 3 days ago
  57. a7a094c [mlir][ROCDL][AMDGPU] Support string waves_per_eu for min,max occupancy range (#225663) by jli · 3 days ago
  58. 8980222 [mlir][sparse][gpu] Add libgen support for transposed SpMV and SpMM (#230412) by Vito Secona · 3 days ago
  59. 7091c7c [mlir][LLVM] Put personality pointer in program address space (#230848) by Arseniy Obolenskiy · 3 days ago
  60. d40bccf [mlir][toy] Add missing spacing to tensor literal closing bracket (#223275) by rdong8 · 3 days ago
  61. 9bf3f72 TableGen: Move value reference locations into Record (#231411) by Matt Arsenault · 3 days ago
  62. e4af4db [mlir][bufferization] handle exit aliases in buffer-loop-hoisting (#226951) by Matteo Di Iorio · 3 days ago
  63. fa095f6 [mlir][ArithToLLVM] Do not lower sitofp to unsupported float types (#230975) by SomnusGG · 3 days ago
  64. 3547ca7 [mlir][x86] Fix AMX loop carried accumulation (#231170) by Adam Siemieniuk · 3 days ago
  65. 2387cb1 [mlir][tosa] Add MXFP support for CONV3D (#230148) by Deeptanshu Sekhri · 3 days ago
  66. 29394b2 [mlir][tosa] remove duplicate CONV2D compliance entries (#231152) by Deeptanshu Sekhri · 3 days ago
  67. 18c2f1f [mlir][SCFToControlFlow] Mark the scf.for increment as non-wrapping (#230891) by Slava Zakharin · 3 days ago
  68. 036ef61 [mlir][ptr] Preserve broadcasts when folding zero pointer offsets (#231077) by Alexander Shaposhnikov · 3 days ago
  69. a03ebf1 [MLIR][LLVM] Add noundef metadata to llvm.load (#228517) by Arseniy Obolenskiy · 3 days ago
  70. 3ce557a [MLIR] Support scalars in ValueBounds of linalg.index (#230962) by Alexey Kreshchuk · 4 days ago
  71. 7f278e9 [mlir][x86] Bail out on AMX accumulation without result write (#231144) by Adam Siemieniuk · 4 days ago
  72. ec3b87b [ViewOpGraph] Re-use `AsmState` to significantly speed-up printing. (#208240) by Philipp Schilk · 4 days ago
  73. 8c8805d [mlir][tosa] Fix no op clamp canonicalization when nan_mode = ignore (#229887) by Luke Hutton · 4 days ago
  74. 1e09589 [mlir][LLVM] Add ignore_denormal_mode UnitAttr to LLVM::AtomicRMWOp (#217588) by Christian Sigg · 4 days ago
  75. 2779f9e [mlir][acc] Report independent loops not parallelized due to unstructured control flow (#230066) by Kareem Ergawy · 4 days ago
  76. 2605640 [mlir][RemoveDeadValues] Replace dead function argument uses with poison (#208881) by Jinpeng Wang · 4 days ago
  77. 501bf96 [MLIR] Guard against overflow when simplifying affine expressions. (#228010) by Jiaqi He · 4 days ago
  78. f9ff926 [mlir][NFC] Simplify using OpRewritePattern::OpRewritePattern to use Base alias (#230717) by Benjamin Wu · 5 days ago
  79. bc3baad [MLIR][CAPI][Python] Add C API and Python bindings for the remark engine (#229094) by Guray Ozen · 6 days ago
  80. 5f87c63 [xegpu][transform] Add support for general offset caculation in `xegpu.insert_prefetch` (#230660) by Charitha Saumya · 6 days ago
  81. 310ceac [mlir][mem2reg] Skip erasing successor operands missing on dead predecessors (#230330) by Bruno Cardoso Lopes · 7 days ago
  82. ef2cce9 [mlir][inliner] Skip first-round use recompute given no change (#230335) by Bruno Cardoso Lopes · 7 days ago
  83. 2c5301a [mlir][x86] Fix VNNI operand load offsets (#230205) by Adam Siemieniuk · 7 days ago
  84. 18364ee [mlir][x86] Fix result tracing through loops (#230229) by Adam Siemieniuk · 7 days ago
  85. 68266c3 [mlir][ArithToLLVM] Fix index lowering for addui_extended (#223265) by Bruce Xi · 7 days ago
  86. f46787c [mlir][affine] Admit signless integers for affine.parallel maxs/mins/maxu/minu reductions (#230332) by William Moses · 7 days ago
  87. 21256b8 [mlir][acc] Declare the memory effects of acc.atomic.write (#230029) by Kareem Ergawy · 7 days ago
  88. ba3b9cf [mlir][ArmNeon] Add pack/mmt4d path to f32 linalg.matmul integration test (#225063) by Federico Bruzzone · 7 days ago
  89. 83f8c8d [mlir][ArmSME] Make tile allocation order deterministic (#229096) by Federico Bruzzone · 7 days ago
  90. caedb62 [mlir][NFC] Use DefaultUnreachable in TypeSwitch (#230333) by Benjamin Wu · 7 days ago
  91. 4674d87 [nfc][mlir][acc] Improve acc.on_device documentation (#230143) by Razvan Lupusoru · 7 days ago
  92. 8109746 [mlir][LLVM] Verify llvm.invoke callees like llvm.call (#229954) by Bruno Cardoso Lopes · 7 days ago
  93. 1c13530 [mlir][LLVMIR] Bound import diagnostics and share one slot tracker (#229956) by Bruno Cardoso Lopes · 7 days ago
  94. 0c405ae [mlir][Vector] Reject scalable masks in maskedload/maskedstore emulation (#223316) by Cheng-Chieh Liu · 7 days ago
  95. 6d2910c [mlir][affine] Clamp negative static trip counts to zero (#227134) by Slava Zakharin · 7 days ago
  96. 6214307 [mlir][affine] Fix unroll cleanup loop bound for negative trip counts (#227135) by Slava Zakharin · 7 days ago
  97. a1f8aad [mlir][vector] Account for vector.mask in transfer RAW/WAW checks (#228328) by Vaisman · 7 days ago
  98. 6ae10c8 [mlir][tosa] Extend allow-non-finites to max_pool2d (#229869) by Hariprasad Ravishankar · 8 days ago
  99. 8c29add [MLIR][Remark] Look through nested locations for the remark file location (#230128) by Henrich Lauko · 8 days ago
  100. 003002b [mlir][sparse] Collapse dense and compressed iteration ranges (#211154) by qyingwu · 8 days ago