1. 7047cc3 Rollup of minor runtime fixes/cleanup from the AMDGPU branch. (#19621) by Ben Vanik · 5 weeks ago
  2. aa06523 [NFC] Comment fixes in iree_bitcode_library. by Ben Vanik · 5 weeks ago
  3. 66723e4 Cleaning up null HAL driver options. by Ben Vanik · 5 weeks ago
  4. 2199c1d Adding iree_arena_block_pool_preallocate. by Ben Vanik · 5 weeks ago
  5. ea462c8 Removing some IREE_RETURN_AND_END_ZONE_IF_ERROR usage that was ugly. by Ben Vanik · 5 weeks ago
  6. 4a04c0a Adding minor iree/base/ time, string view, and memory utilities. by Ben Vanik · 5 weeks ago
  7. a8f7a32 Adding iree_hal_queue_affinity_* utilities. by Ben Vanik · 5 weeks ago
  8. c9fb739 Fixing HAL driver CTS test to not assume numerical indices exist. by Ben Vanik · 5 weeks ago
  9. 1ccabe5 Adding COMPILER_TARGET_DEVICE to iree_hal_cts_test_suite. by Ben Vanik · 5 weeks ago
  10. d517661 [runtime][python] Add debug sink to bindings (#19013) by Boian Petkantchin · 5 weeks ago
  11. c97b084 Including the .kd symbol suffix in AMDGPU executables. by Ben Vanik · 5 weeks ago
  12. d224220 Bump LLVM to llvm/llvm-project@21edac25f09faee23015c6a69d95fcbda287efe2 (#19616) by MaheshRavishankar · 5 weeks ago
  13. b245e6b Delete test_models job using SHARK-TestSuite/iree_tests. (#19614) by Scott Todd · 5 weeks ago
  14. 1445cef Set MLIR_LINK_MLIR_DYLIB to not link shared libMLIR (#19613) by Marius Brehler · 5 weeks ago
  15. 340ffbb [LinalgExt] Drop the unit dims on scatter ops 2/3 (#19450) by Ian Wood · 5 weeks ago
  16. 0820f10 [hip] Don't join the status in dispatch_thread. (#19583) by Andrew Woloszyn · 5 weeks ago
  17. cdf24b9 [Dispatch] Two fixes for CollapseDimensionsPass (#19598) by Ian Wood · 5 weeks ago
  18. 763406f [Codegen][Tuner] skip linking based on the default entry point attribute (#19603) by Bangtian Liu · 5 weeks ago
  19. c992d29 [runtime][hip] Fix format errors and conflicting types. (#19607) by Han-Chung Wang · 5 weeks ago
  20. e7d4fec Bump to LLVM with nanobind reverts (#19605) by Marius Brehler · 5 weeks ago
  21. 71d6de7 [Flow] Improve DeduplicateExecutables bucketing (#19601) by Ian Wood · 5 weeks ago
  22. 6b64fca [LinalgExt] Fusion support for LinalgExt ScatterOp 1/3 (#19560) by Ian Wood · 5 weeks ago
  23. 9cb984f [Codegen] Fix use after erase bug in optimize-tensor-extract-slices (#19460) by Kunwar Grover · 5 weeks ago
  24. 89afd4e Add missing definition of `build` method in `Flow::CallOp` (#19552) by Twice · 5 weeks ago
  25. 0c6d267 Fix typo in serializeExecutable of CUDA target (#19609) by Twice · 5 weeks ago
  26. c7086cf Bump LLVM to llvm/llvm-project@9f5cefe (#19600) by MaheshRavishankar · 5 weeks ago
  27. b5272df [docs] Update copyright year to 2025 in the website footer. (#19599) by Scott Todd · 5 weeks ago
  28. 9df82fd [LinalgExt] Add support for fusing scatter with producers (#19584) by Quinn Dawkins · 5 weeks ago
  29. 5a97523 [GPU] Disable insert/extract slice lowering from pack/unpack by default (#19590) by Zhuoran Yin · 5 weeks ago
  30. 1e935c4 [PJRT] Fix ninja not found error while re-building the plugin (#19553) by Twice · 5 weeks ago
  31. c2d408f Update IREE third-party/benchmark for RISC-V Compatibility (#19538) by Phoebe Chen · 5 weeks ago
  32. e6ac016 [Codegen][Tuner] verifier for the default tuning spec (#19525) by Bangtian Liu · 5 weeks ago
  33. 26b24f2 Bump LLVM to llvm/llvm-project@cbff02b (#19589) by MaheshRavishankar · 5 weeks ago
  34. fc6c518 [docs] Fix diagram rendering on C API reference page. (#19586) by Scott Todd · 6 weeks ago
  35. 88848b4 [runtime][hip] Do not error when peered memory access already enabled (#19582) by Boian Petkantchin · 6 weeks ago
  36. c203e6b [LLVMGPU] Prefer non-scatter ops as the configuration root (#19581) by Quinn Dawkins · 6 weeks ago
  37. fa325c5 [infra] Add `--force` flag to `git submodule update` in script. (#19550) by Scott Todd · 6 weeks ago
  38. f27feff Update LLVM to llvm/llvm-project@ac8bb735 (#19566) by Stanley Winata · 6 weeks ago
  39. a43d893 [Dispatch] Disable scatter fusion with producers (#19565) by Ian Wood · 6 weeks ago
  40. d746a57 Remove revert for https://github.com/llvm/llvm-project/pull/120115 (#19567) by Kunwar Grover · 6 weeks ago
  41. f1e1866 Update LLVM to llvm/llvm-project@b13592219c421820b (#19554) by Stanley Winata · 6 weeks ago
  42. 1f19761 Enable peering among all devices on the system. (#19555) by Andrew Woloszyn · 7 weeks ago
  43. 0184eee [Codegen][RoCDL] Add patterns for lowering bit-width emulation operations to LLVM (#19551) by MaheshRavishankar · 7 weeks ago
  44. 76a7b89 Revert "[hip] Fixed a busy wait in event_semaphore." (#19548) by Andrew Woloszyn · 7 weeks ago
  45. 0a0483e [hip] Add trace zones to copy/fill/update buffer commands. (#19544) by Andrew Woloszyn · 7 weeks ago
  46. 47ccd93 [hip] Implement asynchronous file reads in hip. (#19545) by Andrew Woloszyn · 7 weeks ago
  47. d917e7d [hip] Fixes a race in allocator_free_async. (#19541) by Andrew Woloszyn · 7 weeks ago
  48. 9b8bba8 [hip] Fixed a busy wait in event_semaphore. (#19540) by Andrew Woloszyn · 7 weeks ago
  49. 604cba8 Fix incorrect offset in fd_file.c (#19543) by Andrew Woloszyn · 7 weeks ago
  50. 7ff83ea [hip][cuda] Increase the size of the query pool. (#19542) by Andrew Woloszyn · 7 weeks ago
  51. 07f81f0 Revert "Enable scatter fusion with index operand. (#19198)" (#19535) by MaheshRavishankar · 7 weeks ago
  52. 83af679 Bump to llvm/torch-mlir@061bbc5e1bc4f7880bb565e404a6709f97396818 (#19531) by Rob Suderman · 8 weeks ago
  53. fb4d094 Ukernel lowering for data-tiled `multi_mma` with `mfma_i32_16x16x32_i8` (#19522) by Benoit Jacob · 8 weeks ago
  54. 5c4bc67 Trigger presubmit ci workflows from `ci.yml` via `workflow_call`. (#19445) by Scott Todd · 8 weeks ago
  55. 01f0900 [NFC] Fixing typo (mutli -> multi). (#19526) by Han-Chung Wang · 8 weeks ago
  56. ed9a028 GPU Data-tiled multi-mma: subgroup dimensions should be outer (#19521) by Benoit Jacob · 8 weeks ago
  57. 16097c1 Remove the operand promotion for LHS and RHS. (#19516) by Prashant Kumar · 8 weeks ago
  58. e553425 [Codegen][Tuner] attr verifier for tuning specs (#19486) by Bangtian Liu · 8 weeks ago
  59. 3614f69 Update LLVM to llvm/llvm-project@b07e7b76c5d532a61 (#19500) by Stanley Winata · 8 weeks ago
  60. ce65948 Adding iree_io_file_handle_create/iree_io_file_handle_open. (#19510) by Ben Vanik · 8 weeks ago
  61. 078c3ec [runtime][python] Add IRPA entry conversion to/from numpy (#19492) by Boian Petkantchin · 8 weeks ago
  62. 4e29bbb Bump Sharktank forward to bypass failing test flag (#19519) by Rob Suderman · 8 weeks ago
  63. 700572c Adding experimental iree_io_file_map_view API. (#19515) by Ben Vanik · 8 weeks ago
  64. 101f55c Adding fd-based file handle support to the HAL. (#19514) by Ben Vanik · 8 weeks ago
  65. 4c00a22 Enable scatter fusion with index operand. (#19198) by MaheshRavishankar · 8 weeks ago
  66. cbdcdd0 Adding iree_hal_file_from_handle factory for common file impls. (#19513) by Ben Vanik · 8 weeks ago
  67. 5b67943 Moving synchronous HAL file APIs to the public API. (#19512) by Ben Vanik · 8 weeks ago
  68. 345b1da Revert "[LLVMGPU] Deprecate the matmul simt pipeline (#19335)" (#19508) by Archana Ramalingam · 8 weeks ago
  69. 8ae1b54 [GPU] Use padding in IGEMM pipeline to support unaligned to intrinsic shapes (#19484) by Nirvedh Meshram · 8 weeks ago
  70. 78ea0ad Bump to use the flash attention variant (#19505) by Rob Suderman · 8 weeks ago
  71. 3509ead Cleanup `ConvertToStream` to accomodate llvm/llvm-project@3f136f7 (#19451) by MaheshRavishankar · 8 weeks ago
  72. 72d98bc GPU ukernel lowering config for data-tiled multi_mma, and a simple ukernel. (#19504) by Benoit Jacob · 8 weeks ago
  73. a31da1f Fixing missing trace zone end in iree_io_scope_map. by Ben Vanik · 8 weeks ago
  74. a5cf548 [NFC] GPU ukernels cleanups (#19503) by Benoit Jacob · 8 weeks ago
  75. 1894af3 Update LLVM to llvm/llvm-project@dd6f6a0 (#19489) by Stanley Winata · 8 weeks ago
  76. 362b554 [docs] Refresh `status: new` usage across website pages. (#19495) by Scott Todd · 8 weeks ago
  77. f2690e2 [docs] Expand on instructions for installing torch for CPU. (#19493) by Scott Todd · 8 weeks ago
  78. 8e86bcf Refresh architecture diagram with ONNX and LiteRT inputs. (#19494) by Scott Todd · 8 weeks ago
  79. f90771d Refresh project status in README and website homepage. (#19482) by Scott Todd · 8 weeks ago
  80. 3ab9d4b Revert "Skip test_sharktank job until quota issues are fixed." (#19491) by Scott Todd · 8 weeks ago
  81. 6ff00a8 [LLVMGPU] Deprecate the matmul simt pipeline (#19335) by Prashant Kumar · 8 weeks ago
  82. fdf4ae6 Update emsdk in samples workflow. (#19490) by Scott Todd · 8 weeks ago
  83. 7013101 [PJRT] Fix compile error while tracing is enabled (#19485) by Twice · 8 weeks ago
  84. 25de549 [NFC] Delete outdated e2e encoding tests. (#19487) by Han-Chung Wang · 8 weeks ago
  85. 05ce39f [DT] Unify encoding materialization pass into a single pass. (#19454) by Han-Chung Wang · 8 weeks ago
  86. 67a05a4 [DT][NFC] Internalize transposeNarrowN logic to LayoutAttrInterface Impl (#19453) by Han-Chung Wang · 8 weeks ago
  87. dc29ee7 Move GPU ukernel selection to KernelConfig (#19440) by Benoit Jacob · 8 weeks ago
  88. 63cdc7d Reapply "[Codegen][GPU] Add range information to GPU dispatch IDs" (#19361) (#19372) by Krzysztof Drewniak · 8 weeks ago
  89. 442956c Use LLVMGPUTileandFuse instead of LLVMGPUVectorize for convolutions (#19469) by Nirvedh Meshram · 8 weeks ago
  90. 99b600f [Codegen] Allow padding of dynamic allocas (#19399) by Max191 · 8 weeks ago
  91. 8a7b754 Update website copyright text per Linux Foundation guidance. (#19480) by Scott Todd · 8 weeks ago
  92. ffa0f42 [PJRT] Allow to pass extra compile options via env variables (#19418) by Twice · 8 weeks ago
  93. 0cafee9 [vm] Add support for SI64 to F32 casts (#19455) by Vinayak Dev · 8 weeks ago
  94. ad938ae [DT][NFC] Localize CPU specific encoding materialization logic. (#19452) by Han-Chung Wang · 8 weeks ago
  95. c618134 Calculate storage bytes through interface method for encoding types. (#19413) by Han-Chung Wang · 8 weeks ago
  96. 900ef1d [PJRT] Update README to align with the current status (#19457) by Twice · 9 weeks ago
  97. 7e1804b Update LLVM to llvm/llvm-project@3f136f7 (#19479) by MaheshRavishankar · 9 weeks ago
  98. e562559 Increase all timeouts in pkgci_regression_test.yml. (#19477) by Scott Todd · 9 weeks ago
  99. 9b8595d Revert llvm submodule change that was accidentally added in #18790 (#19476) by Andrew Woloszyn · 9 weeks ago
  100. 27742f6 Deflake some pkgci jobs. (#19472) by Scott Todd · 9 weeks ago