Commits

Branch selector

 main

User selector

All users

All time

Commit History

Commits on Mar 30, 2026

• [compiler][nfc] Move general passes from Codegen/Common/ to compiler/Transforms/ (#23961)

  Show description for 1279411
  hanhanW

  authored
  1279411

  Copy full SHA for 1279411

• [GPU] Implement block intrinsics introduced in CDNA3 (1/6) (#23949)

  Show description for 17b18d6
  

  nirvedhmeshram

  and

  claude

  authored
  17b18d6

  Copy full SHA for 17b18d6

• [LLVMCPU] Modernize CPU pipeline tests. (#23965)

  Show description for 262b4cc
  hanhanW

  authored
  262b4cc

  Copy full SHA for 262b4cc

• [Tuner] Pass iree emitReset option to mlir smt (#23966)

  Show description for b82a886
  RattataKing

  authored
  b82a886

  Copy full SHA for b82a886

• [LLVMGPU] Defer HoistExecutableObjectPass to the end of pipeline. (#23962)

  Show description for 03c158e
  hanhanW

  authored
  03c158e

  Copy full SHA for 03c158e

• [Codegen][CPU] Switch LLVMCPU to module-scope pipeline. (#23953)

  Show description for a867e93
  hanhanW

  authored
  a867e93

  Copy full SHA for a867e93

• [Codegen][LLVMGPU] Add layout support for ArgCompare operations (#23693)

  Show description for 4d476ec
  bangtianliu

  authored
  4d476ec

  Copy full SHA for 4d476ec

• [Codegen] Plumb through dispatch_config for workgroup count computation. (#23899)

  Show description for e4b4da7
  hanhanW

  authored
  e4b4da7

  Copy full SHA for e4b4da7

• Integrates/llvm@8a4f210 (#23957)

  Show description for 4c7022b
  sommerlukas

  authored
  4c7022b

  Copy full SHA for 4c7022b

• [AArch64] Disable breaking up transfer+extract into multiple scalar loads for AArch64 (#23172)

  Show description for 4bc58ba
  momchil-velikov

  authored
  4bc58ba

  Copy full SHA for 4bc58ba

Commits on Mar 27, 2026

• Integrates/llvm@23eec12 (#23946)

  Show description for cdf0367
  bangtianliu

  authored
  cdf0367

  Copy full SHA for cdf0367

• [Codegen] Introduce PropagateDispatchConfig pass. (#23895)

  Show description for 53263de
  hanhanW

  authored
  53263de

  Copy full SHA for 53263de

• [Codegen] Introduce CreateDispatchConfig pass. (#23894)

  Show description for 3d50bd4
  hanhanW

  authored
  3d50bd4

  Copy full SHA for 3d50bd4

• Integrate openxla/stablehlo@b744df4fa70f (#23940)

  Show description for c233b32
  

  jschuhmacher

  and

  hanhanW

  authored
  c233b32

  Copy full SHA for c233b32

• [LLVMGPU] Support minor identity permutation maps in ROCDLBufferInstructionsOptimization (#23947)

  Show description for c5f9a92
  

  Max191

  and

  claude

  authored
  c5f9a92

  Copy full SHA for c5f9a92

• Integrates/llvm 20260326 (#23938)

  Show description for ffcbc3b
  bangtianliu

  authored
  ffcbc3b

  Copy full SHA for ffcbc3b

Commits on Mar 26, 2026

• Revert "[CPU] Sink down reshapes across packing ops." (#23943)

  Show description for ba73bed
  hanhanW

  authored
  ba73bed

  Copy full SHA for ba73bed

• [Codegen] Skip bf16 expansion for targets with native conversion (#23904)

  Show description for 15697bb
  Muzammiluddin-Syed-ECE

  authored
  15697bb

  Copy full SHA for 15697bb

• [GPU][Codegen] Support unique per-lane load option when prod(threads) < subgroupsize (#23657)

  Show description for 9f0b79d
  

  efric

  and

  claude

  authored
  9f0b79d

  Copy full SHA for 9f0b79d

• [Tuner] Expose IntKnobAttr to python bindings (#23931)

  Show description for a3b5201
  RattataKing

  authored
  a3b5201

  Copy full SHA for a3b5201

• [GPU] Add in-tree e2e fixed intrinisic tests with TileAndFuse (#23903)

  Show description for c4ace7f
  

  nirvedhmeshram

  and

  claude

  authored
  c4ace7f

  Copy full SHA for c4ace7f

• [CPU] Add vector level tiling sizes to LinalgExt::FftOp. (#23889)

  Show description for 0f8d3ff
  hanhanW

  authored
  0f8d3ff

  Copy full SHA for 0f8d3ff

• [CPU] Sink down reshapes across packing ops. (#23922)

  Show description for 0e0ca91
  hanhanW

  authored
  0e0ca91

  Copy full SHA for 0e0ca91

Commits on Mar 25, 2026

• [Codegen] Add pattern to hoist expand_shape out of forall (#23930)

  Show description for a6722ae
  

  Max191

  and

  claude

  authored
  a6722ae

  Copy full SHA for a6722ae

• [Util] Add integer divisibility inference for arith.select (#23929)

  Show description for c490027
  Max191

  authored
  c490027

  Copy full SHA for c490027

• [CI] Move torch tests to presubmit (#23913)

  Show description for 65f8952
  bangtianliu

  authored
  65f8952

  Copy full SHA for 65f8952

• Integrates/llvm 20260325 (#23923)

  Show description for 6a99588
  bangtianliu

  authored
  6a99588

  Copy full SHA for 6a99588

• [Tokenizer] Fix deferred emission dropping back-to-back special tokens (#23852)

  Show description for c7775e4
  

  jtuyls

  and

  claude

  authored
  c7775e4

  Copy full SHA for c7775e4

• [GPU][NFC] Document useDirectLoad and prefetchNumStages in deduceMMASchedule (#23754)

  Show description for e3df65e
  Yu-Zhewen

  authored
  e3df65e

  Copy full SHA for e3df65e

Commits on Mar 24, 2026

• Add byte_offset to hal.tensor.import for subspan alignment (#23915)

  Show description for 7703cd6
  

  stellaraccident

  and

  claude

  authored
  7703cd6

  Copy full SHA for 7703cd6

• Remove unused cmake target: iree-run-module-test-deps (#23921)

  Show description for 819c1ef
  HanKuanChen

  authored
  819c1ef

  Copy full SHA for 819c1ef

• [VectorDistribute] Pack partial reduction instead of using expand_dims (#23871)

  Show description for ae98b5f
  

  Groverkss

  and

  claude

  authored
  ae98b5f

  Copy full SHA for ae98b5f

• [Codegen] Define DispatchConfigOp and YieldOp to IREE::Codegen. (#23893)

  Show description for 42d82a6
  hanhanW

  authored
  42d82a6

  Copy full SHA for 42d82a6

• Integrates/llvm 20260324 (#23917)

  Show description for 1470782
  bangtianliu

  authored
  1470782

  Copy full SHA for 1470782

• Reland "Add F16 support to contraction input demotion pass (#23621)" (#23850)

  Show description for 0a13120
  HanKuanChen

  authored
  0a13120

  Copy full SHA for 0a13120