Commits

Branch selector

 main

User selector

All users

All time

Commit History

Commits on Oct 16, 2025

• [Codegen] Allow pre-padding other dims of a conv except the input channel (#22296)

  Show description for fecdfeb
  yzhang93

  authored
  fecdfeb

  Copy full SHA for fecdfeb

Commits on Oct 15, 2025

• Drop revert of llvm/llvm-project#159083 (#22298)

  Show description for d639bd8
  MaheshRavishankar

  authored
  d639bd8

  Copy full SHA for d639bd8

• Fix non-deterministic hoisting (#22319)

  Show description for 188c42c
  IanWood1

  authored
  188c42c

  Copy full SHA for 188c42c

• [Stream] Replicate globals per affinity before Stream conversion. (#22117)

  Show description for 271e7f6
  hanhanW

  authored
  271e7f6

  Copy full SHA for 271e7f6

• [codegen][gpu] Make transfer_write conditional when not fully distributed (#22198)

  Show description for a93cdb0
  

  newling

  and

  fabianmcg

  authored
  a93cdb0

  Copy full SHA for a93cdb0

• [GPU] Update K Tile Size for multiple K dims (#22310)

  Show description for 8d950fd
  Muzammiluddin-Syed-ECE

  authored
  8d950fd

  Copy full SHA for 8d950fd

• [DT][NFC] Trim IRs in encoding materialization tests for GPU and RISCV backends. (#22313)

  Show description for dad4b2d
  hanhanW

  authored
  dad4b2d

  Copy full SHA for dad4b2d

• [python] Set up python binding for matcher convolution and attention op (#22311)

  Show description for 9fdd3dd
  bangtianliu

  authored
  9fdd3dd

  Copy full SHA for 9fdd3dd

• [GPU] Enabling Gather-like ops to go through GPUTileAndFuse pipeline (#22251)

  Show description for 6fbb919
  Abhishek-Varma

  authored
  6fbb919

  Copy full SHA for 6fbb919

Commits on Oct 14, 2025

• Bump version to 3.9.0 after 3.8.0 release. (#22308)

  Show description for f4ae0ce
  sa-faizal

  authored
  f4ae0ce

  Copy full SHA for f4ae0ce

• Promote iree-opt-data-tiling to pipeline options. (#22295)

  Show description for d909d36
  hanhanW

  authored
  d909d36

  Copy full SHA for d909d36

• [CI][TorchModels] Update flags used for LLaMa 8b f8/fp16. (#22297)

  Show description for 9b54341
  MaheshRavishankar

  authored
  9b54341

  Copy full SHA for 9b54341

• [LLVMCPU] Propagate target features and CPU name to individual LLVMFuncOp (#22036)

  Show description for 71b4396
  

  mshockwave

  and

  kuhar

  authored
  71b4396

  Copy full SHA for 71b4396

• Let MLIR ukernels provide their matching and data-tiled-layout info. (#22254)

  Show description for d1e43f8
  bjacob

  authored
  d1e43f8

  Copy full SHA for d1e43f8

• Bump LLVM to llvm/llvm-project@5a636c6 (#22290)

  Show description for d49f6f5
  MaheshRavishankar

  authored
  d49f6f5

  Copy full SHA for d49f6f5

Commits on Oct 13, 2025

• [Codegen] Adding an optional `dma_sizes` field in GPU attributes (#22281)

  Show description for 0eee7d8
  lialan

  authored
  0eee7d8

  Copy full SHA for 0eee7d8

• [Docs] Read from first line of `rocm_agent_enumerator` output (#22283)

  Show description for 35a2fa4
  sjain-stanford

  authored
  35a2fa4

  Copy full SHA for 35a2fa4

• [GPU] Support iree_tensor_ext.dispatch.tensor.store for broadcast producer (#22291)

  Show description for 26c70db
  nirvedhmeshram

  authored
  26c70db

  Copy full SHA for 26c70db

• [GPU] Move convolution check out of unrelated function (#22287)

  Show description for 99b898b
  Groverkss

  authored
  99b898b

  Copy full SHA for 99b898b

• [NFC][GPU] Move reduction configuration to gpu utilities (#22286)

  Show description for 4efac5b
  Groverkss

  authored
  4efac5b

  Copy full SHA for 4efac5b

• Use llvm wrappers for accumulate. NFC. (#22279)

  Show description for 3efb376
  kuhar

  authored
  3efb376

  Copy full SHA for 3efb376

• Support skinny scaled matmul in kernel config (#22042)

  Show description for d1b4076
  jtuyls

  authored
  d1b4076

  Copy full SHA for d1b4076

Commits on Oct 10, 2025

• Integrate llvm/llvm-project@b92483c (#22274)

  Show description for de81737
  newling

  authored
  de81737

  Copy full SHA for de81737

• Revert "Re-enable lds_barrier on RDNA4" (#22278)

  Show description for cdce159
  kuhar

  authored
  cdce159

  Copy full SHA for cdce159

• [Codegen] Update the td spec using the attention matcher op (#22266)

  Show description for 99abbc8
  bangtianliu

  authored
  99abbc8

  Copy full SHA for 99abbc8

• [Codegen] Update the td spec using the contraction matcher op (#22249)

  Show description for ed0973a
  bangtianliu

  authored
  ed0973a

  Copy full SHA for ed0973a

• [debugging][gpu] Add --iree-hip-emit-debug-info flag (#22216)

  Show description for d7db886
  willghatch

  authored
  d7db886

  Copy full SHA for d7db886

• Reapply "[Dispatch Creation] Rework dispatch formation logic (#21854)" (#22065)

  Show description for 941241b
  IanWood1

  authored
  941241b

  Copy full SHA for 941241b

• Reapply "[GPU] Allow multi result and indexing compute generic ops in TilleAndFuse pipeline" (#22205)" (#22223)

  Show description for 6701262
  nirvedhmeshram

  authored
  6701262

  Copy full SHA for 6701262

• [CI][iree-test-suites] Try to make torch_models benchmarks more stable (#22271)

  Show description for 0b0a17a
  Groverkss

  authored
  0b0a17a

  Copy full SHA for 0b0a17a

• Re-enable lds_barrier on RDNA4 (#21922)

  Show description for 6c07196
  krzysz00

  authored
  6c07196

  Copy full SHA for 6c07196

• [python] Set up binding for preprocessing transform ops (#22227)

  Show description for aea660c
  bangtianliu

  authored
  aea660c

  Copy full SHA for aea660c

• [LinalgExt] Don't vectorize map_scatter in non-contiguous sub-byte access (#22242)

  Show description for e2bd9ad
  jtuyls

  authored
  e2bd9ad

  Copy full SHA for e2bd9ad

• [Codegen] Update the assembly formats and corresponding tests for matcher ops (#22270)

  Show description for afe01bd
  bangtianliu

  authored
  afe01bd

  Copy full SHA for afe01bd

• [build flags] prepare to enable more warnings in compile flags (#21996) (#22252)

  Show description for b09740d
  schuermans-roofline

  authored
  b09740d

  Copy full SHA for b09740d