1. 1fc6e5b Add CDNA3 MFMA BF16 intrinsics. (#18892) by Benoit Jacob · 5 months ago
  2. 4ad834b Support F8E5M2FNUZ MFMA on CDNA3 (#18887) by Benoit Jacob · 5 months ago
  3. a400cde [ROCM][NFC] Add option to control SLP vectorization in llvm optimizations (#18865) by Nirvedh Meshram · 5 months ago
  4. c08362a GPU target parameters for data tiling. (#18839) by Benoit Jacob · 5 months ago
  5. 2dffc9e Fix MSVC compilation following #18682 (#18830) by Benoit Jacob · 5 months ago
  6. 4f33005 Skip ROCM/test/opt_pass_plugin on Windows while broken. (#18823) by Scott Todd · 5 months ago
  7. 0c6a151 Warn when --iree-llvmcpu-target-cpu defaults to "generic". (#18682) by Benoit Jacob · 5 months ago
  8. 8568efa [GPU] Adding support for opt pass plugins during AMDGPU executable serialization (#18347) by Jose Manuel Monsalve Diaz · 5 months ago
  9. a3d8ad6 [ROCM] Fix feature flags for gfx1100 and improve flag handling (#18781) by Kunwar Grover · 5 months ago
  10. 164a60e [ROCM] Disable mixed precision fma instructions that cause numeric issues (#18753) by Nirvedh Meshram · 5 months ago
  11. a7d84f9 [ROCm] Fix known target info for MI300A (#18648) by Jakub Kuderski · 6 months ago
  12. 20a7638 [ROCm] Always require `--iree-hip-target` (#18645) by Jakub Kuderski · 6 months ago
  13. eef4623 [LLVMGPU][ROCm] Move kernel annotation before serialization (#18573) by Jakub Kuderski · 6 months ago
  14. 0f15c8d [LLVMGPU][ROCm] Add validation on finalized llvm bitcode (#18552) by Jakub Kuderski · 6 months ago
  15. 861695b Integrates/llvm 20240910 (#18480) by Nirvedh Meshram · 6 months ago
  16. 1acbedc Bumping HAL API version to 0.5. by Ben Vanik · 7 months ago
  17. e2a2b2b Removing descriptor set layouts from HAL IR and simplifying bindings. by Ben Vanik · 7 months ago
  18. 758ef19 Dropping WGSLReplacePushConstantsPass. by Ben Vanik · 7 months ago
  19. e044271 Converting Metal target to support executable-create2. by Ben Vanik · 7 months ago
  20. 9bbc926 Removing legacy pipeline layout and dispatch binding model. by Ben Vanik · 7 months ago
  21. 387b772 Converting Vulkan target to support executable-create2. by Ben Vanik · 7 months ago
  22. 21946fe Converting local CPU target to support executable-create2. by Ben Vanik · 7 months ago
  23. f51d4ea Converting CUDA target to support executable-create2. by Ben Vanik · 7 months ago
  24. 894dfbe Converting HIP target to support executable-create2. by Ben Vanik · 7 months ago
  25. f43f59b Renaming [spirv|wgsl]_executable_def to [vulkan|webgpu]. by Ben Vanik · 7 months ago
  26. 4349df7 Renaming rocm executable -> hip executable. by Ben Vanik · 7 months ago
  27. 5c06d4b Factoring out common debug info from GPU executable flatbuffers. by Ben Vanik · 7 months ago
  28. cd5037e Fixing ROCM/CUDA BlockSizeDef -> BlockSize. (tables have Def, structs don't) by Ben Vanik · 7 months ago
  29. c90a885 Fixing some incorrect TODOs referencing #18154. by Ben Vanik · 7 months ago
  30. 7050033 [Codegen][GPU] Add support for WMMA_I32_16x16x16_I8 (#18372) by Quinn Dawkins · 7 months ago
  31. 10aa470 [LLVMGPU][ROCm] Disable kernarg preloading on pre-CDNA3 targets (#18343) by Jakub Kuderski · 7 months ago
  32. c44d29b [compiler] Make cuda/hip/vulkan target cl options consistent (#17710) by Lei Zhang · 7 months ago
  33. 300af39 [codegen] Add max_workgroup_counts to TargetWgpAttr (#17771) by Krzysztof Drewniak · 7 months ago
  34. 8dc6820 Adding simplified HAL dispatch methods. (#18189) by Ben Vanik · 7 months ago
  35. 50f18f1 [NFC][Encoding] Outline encodings in lit tests (#18165) by Max191 · 8 months ago
  36. e9e24f8 [GPU] Follow the official naming convention for WMMA attributes. (#18147) by Han-Chung Wang · 8 months ago
  37. de679c9 Creating reusable command buffers in stream->hal lowering. (#18100) by Ben Vanik · 8 months ago
  38. 82012e6 [GPU][NFC] Follow the official convention to define mfma/wmma attributes (#18127) by Han-Chung Wang · 8 months ago
  39. e9ee5fa [VectorExt] Move VectorExt from llvm-external-projects to Codegen/Dialect (#18082) by Kunwar Grover · 8 months ago
  40. 6c45bef [runtime][HIP] Retire ROCm HAL backend (#17029) by Nithin Meganathan · 8 months ago
  41. 9aaae34 [GlobalOpt] Transition SetEncoding to use round_dims_to and stop creating tensor.pad (#17931) by Max191 · 8 months ago
  42. e900692 Updating various tests to the latest changes. by Ben Vanik · 8 months ago
  43. c05323f New AssignTargetDevices pass to replace the legacy one. by Ben Vanik · 10 months ago
  44. ddad4ec Adding #hal.device.select and related attributes. by Ben Vanik · 1 year, 1 month ago
  45. ae00c4f Revert "Revert "[LLVMGPU][ROCm] Add MFMA_F32_16x16x4_F32 instruction"… (#17921) by Prashant Kumar · 8 months ago
  46. 30e2c20 Integrate llvm-project @266a5a9cb9daa96c1eeaebc18e10f5a37d638734 (#17911) by Avinash Sharma · 8 months ago
  47. 6a82eb5 Add F8_16x16x32_F32 support for MFMA (#17792) by Stanley Winata · 8 months ago
  48. 02c2000 Revert "[LLVMGPU][ROCm] Add MFMA_F32_16x16x4_F32 instruction" (#17894) by Scott Todd · 8 months ago
  49. d65c6d4 [LLVMGPU][ROCm] Add MFMA_F32_16x16x4_F32 instruction (#17847) by Prashant Kumar · 8 months ago
  50. 429aafd [Codegen] Improve ROCm-specific LLVM translations (#17742) by Krzysztof Drewniak · 8 months ago
  51. dd7d4a1 Integrate llvm-project @31015240d366e4bf6f114856caa6e9ce90742b7f (#17799) by Quinn Dawkins · 9 months ago
  52. 4cdc8e4 [rocm] Add --iree-rocm-legacy-sync flag (default true). (#17786) by Stella Laurenzo · 9 months ago
  53. dcba7c5 [LLVMGPU][ROCm] Plumb through i8, i8 -> i32 MFMA intrinsics (#17764) by Jakub Kuderski · 9 months ago
  54. 0d2c780 Ensure IREE GPU dialect is registered for all GPU targets (fixes #17736) (#17737) by Andrea 🦈 · 9 months ago
  55. 90f29a6 Reland "[spirv] Switch to use common target description" (#17699) by Lei Zhang · 9 months ago
  56. d792d24 Revert "[spirv] Switch to use common target description" (#17698) by Scott Todd · 9 months ago
  57. 7b9fb12 [spirv] Switch to use common target description (#17623) by Lei Zhang · 9 months ago
  58. 3428231 [LLVMCPU] Populate ArmSVE to LLVM conversion patterns (#17665) by Benjamin Maxwell · 9 months ago
  59. 5404ad7 Add versioned, automatically installed 'buildifier' pre-commit hook. (#17589) by Scott Todd · 10 months ago
  60. 5b243a8 [Backend][ROCM] Add gfx1150 support. (#17508) by Stanley Winata · 10 months ago
  61. 006af5d [GPU] Support specifying LLVMGPU backend target features (#17451) by Lei Zhang · 10 months ago
  62. f6a38ac [GPU] Thread through a common target description (#17217) by Lei Zhang · 10 months ago
  63. b4fc0b4 Implementing the f64 VM extension and flipping the flag by default. (#17416) by Ben Vanik · 10 months ago
  64. b716704 Update git-clang-format ref and clang-format version. (#16792) by Scott Todd · 10 months ago
  65. a78cee1 Add support for serializing the textual representation of LLVM IR. (#17193) by Phoebe Chen · 10 months ago
  66. 4f27e64 Generalize overriding llvm func attr flags in translation info (#17365) by Kunwar Grover · 10 months ago
  67. a3b7e12 Integrate both llvm-project@2083e97e (+1 :leftwards_arrow_with_hook:, +1 :cherries:) and torch-mlir@bce800a3 (#17330) by Benoit Jacob · 11 months ago
  68. 8f6ecc5 Hide ExecutableVariantOp from TargetBackend pipeline factory methods. (#17255) by Ben Vanik · 11 months ago
  69. 01cdb0c [CodeGen] Clean up header includes and build dependencies (#17209) by Lei Zhang · 11 months ago
  70. 1ac066a [NFC][GPU] Move SPIRVTile to `Codegen/Common/GPU` (#17142) by Max191 · 11 months ago
  71. 2d0e649 [ROCm] Clean up ROCM target code. NFC. (#17168) by Jakub Kuderski · 11 months ago
  72. aa1769e Moving the LocalDevice impl out of LLVM-CPU/VMVX. by Ben Vanik · 11 months ago
  73. d8c59a4 Cleanup of TargetRegistry after #15468. by Ben Vanik · 1 year ago
  74. 0f50ece Moving LLVMLinkerUtils to a Utils/ directory. by Ben Vanik · 1 year ago
  75. eb49f72 Fixing ROCM bitcode deps when the compiler is built static. (#17163) by Ben Vanik · 11 months ago
  76. 330651e [plugins][ROCM] Fix minor loc source resize bug (#17133) by Stanley Winata · 11 months ago
  77. 3f51a55 Replace openxla/iree with iree-org/iree across the project. (#17110) by Scott Todd · 11 months ago
  78. 954cb36 Move Codegen pass pipelines to nest on `FunctionOpInterface`. (#16665) by MaheshRavishankar · 11 months ago
  79. 190d959 Allow users to specify riscv cpu and get hardware features (#16902) by Alex Chiang · 12 months ago
  80. 3fa9fbd Integrate LLVM at llvm/llvm-project@a6d932bca8875198fbf34564cda8a8d1640cdcbc (#16944) by Benoit Jacob · 12 months ago
  81. 2c88e49 [LLVMGPU] Wmma layout for LLVMGPU vector distribute pipeline (#16928) by Stanley Winata · 12 months ago
  82. ed59bed Enabling external HIP HSACO support ala CUDA external PTX support. (#16830) by Ben Vanik · 1 year ago
  83. 2395046 [rocm] Excises (almost) dependence on /opt/rocm from the compiler. (#16803) by Stella Laurenzo · 1 year ago
  84. 15d9039 Embedding executable source contents in binaries for tracing. (#16757) by Ben Vanik · 1 year ago
  85. 05ff3e2 Don't link `opencl.bc` when compiling for ROCm. (#16778) by Scott Todd · 1 year ago
  86. 7da8af6 Cleanup a few docs/references to old paths in HAL/Target. (#16716) by Scott Todd · 1 year ago
  87. c2a3245 Convert LLVMCPU compiler target to a plugin. (#16704) by Scott Todd · 1 year ago
  88. b027da4 Convert VulkanSPIRV compiler target into a plugin. (#16699) by Scott Todd · 1 year, 1 month ago
  89. c344e26 Cleanup compiler plugin directory and include paths. (#16691) by Scott Todd · 1 year, 1 month ago
  90. 890b070 Forking off device methods from TargetBackend->TargetDevice. (#16591) by Ben Vanik · 1 year, 1 month ago
  91. 09deadf [rocdl] Register some MI210 (gfx90a) supported mfma cases (#16592) by Lei Zhang · 1 year, 1 month ago
  92. 4b1a4e2 Typing IREE::HAL::DeviceTargetAttr executable targets. (#16588) by Ben Vanik · 1 year, 1 month ago
  93. eeda5ca Renaming WebGPU to WebGPU-SPIRV (ala Metal-SPIRV). (#16586) by Ben Vanik · 1 year, 1 month ago
  94. c730000 [ROCM] Use translation info to store waves-per-eu (#16573) by Quinn Dawkins · 1 year, 1 month ago
  95. 2fe2975 Collapse LinalgExt into the main source tree (#16407) by Han-Chung Wang · 1 year, 1 month ago
  96. c2afb6e [ROCM] Add supported intrinsics for gfx942 (#16498) by Quinn Dawkins · 1 year, 1 month ago
  97. fadc018 Removing the use of the legacy_sync hack from all but ROCM. (#16493) by Ben Vanik · 1 year, 1 month ago
  98. 0c540db Revert "Turn on SLPVectorizer for ROCM backend (#16412)" (#16417) by harsh-nod · 1 year, 1 month ago
  99. c066ceb Turn on SLPVectorizer for ROCM backend (#16412) by harsh-nod · 1 year, 1 month ago
  100. 2f20165 [ROCM] Create hasco image as string as expected by the schema (#16384) by Nithin Meganathan · 1 year, 1 month ago