Krzysztof Drewniak 149fa17adf

[mlir][AMDGPU] Update gather_to_lds with explicit-async support (#181082) ...

This commit takes advantage of the new `load.async.to.lds` intrinsic in
order to add an `async` mode to `gather_to_lds`. In this mode,
completion of the load needs to be managed with `asyncmark` and
`wait.asyncmark` intrinsics instead of being implicitly derived by alias
analysis.

This commit adds the flag, a lowering for it, and updates tests.

Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>

2026-02-16 20:52:35 +00:00

..

Affine

[MLIR][Affine] Remove restriction in slice validity check on symbols (#180709)

2026-02-11 12:14:44 +05:30

AMDGPU

[mlir][AMDGPU] Update gather_to_lds with explicit-async support (#181082)

2026-02-16 20:52:35 +00:00

AMX

[mlir][amx] Optional stride for tile load and store (#159569)

2025-10-13 14:22:44 +02:00

Arith

[MLIR][Arith] FastMath extf conversion without NaN checks (#180926)

2026-02-11 12:46:07 +00:00

ArmNeon

[MLIR][AArch64] Lower vector.contract to Neon FEAT_BF16 operations (#148198)

2025-07-23 10:15:47 +01:00

ArmSME

[mlir][ArmSME] Remove ConvertIllegalShapeCastOpsToTransposes (#139706)

2025-06-25 16:40:14 +01:00

ArmSVE

[mlir][vector] Avoid setting padding by default to 0 in vector.transfer_read prefer ub.poison (#146088)

2025-06-30 15:20:42 -04:00

Async

[mlir][async]: Make async.execute operation with RecursiveMemoryEffects trait (#116544)

2025-07-31 16:42:10 +03:00

Bufferization

Fix ownership based deallocation pass crash (#179357)

2026-02-12 12:35:37 +00:00

Builtin

…

Complex

[MLIR][Complex] DivOp check for NaN in the folder (#169724)

2026-01-14 19:47:33 +01:00

ControlFlow

[mlir][CF] Add ub.unreachable canonicalization (#169873)

2025-11-30 02:36:57 +00:00

DLTI

…

EmitC

[mlir][emitc] Update the WrapFuncInClassPass pass (#179184)

2026-02-04 16:30:59 +00:00

Func

[MLIR] Fix FileCheck annotations and add 32-bit coverage (#167149)

2025-11-09 00:39:08 +00:00

GPU

Revert "[mlir] Fix the order of operation attribute verification"

2026-02-10 07:39:53 +00:00

Index

[mlir] Add inliner interface to the index dialect (#170459)

2025-12-05 14:43:51 +00:00

IRDL

[MLIR][IRDL] Support camelCase segment size attributes in IRDL verifier (#168836)

2025-11-24 12:04:35 +08:00

Linalg

[mlir][linalg] Remove abandoned Detensorize pass (#177579)

2026-02-10 13:29:19 +01:00

LLVM

…

LLVMIR

[mlir][ROCDL] Add async variants of pre-gfx12 LDS load intrinsics (#181072)

2026-02-16 12:40:54 -08:00

Math

[mlir][Math] Fix math-expand-ops crash on math.ctlz with index type (#181539)

2026-02-15 23:56:44 +00:00

MemRef

[mlir] disable folding collapse expand to cast (#179209)

2026-02-05 22:31:43 +02:00

MLProgram

[ml_program] fix bufferizesToMemoryRead for ml_program.global_store (#177387)

2026-01-23 13:20:30 +01:00

MPI

[mlir][shard,mpi] Fixing lowering allgather shard->mpi->llvm (#178870)

2026-02-04 18:36:03 +01:00

NVGPU

[mlir][nvgpu] Fix a division by zero crash in OptimizeSharedMemoryPass (#174931)

2026-01-16 10:42:08 +08:00

OpenACC

[OpenACC][MLIR] clone private operands during ACCIfClauseLowering (#177458)

2026-01-23 07:57:56 -06:00

OpenMP

[Flang][mlir][OpenMP] Add affinity clause to omp.task and Flang lowering (#179003)

2026-02-04 10:30:35 -06:00

PDL

…

PDLInterp

…

Ptr

[mlir][ptr] Add ptr.ptr_diff operation (#157354)

2025-09-24 13:09:31 -04:00

Quant

Extending UniformQuantizedType with interface-based support for new storage types in Quant dialect (#152966)

2026-02-10 09:56:24 +00:00

SCF

[MLIR][Utils] Fix overflow in constantTripCount for narrow types (#179985)

2026-02-09 14:58:47 -08:00

Shape

[mlir][shape] Fix crash in ShapeOfOpToConstShapeOp (#180737)

2026-02-11 09:09:39 +08:00

Shard

[mlir][shard, mpi] Allow more than one last axis to be "unsplit" (#180754)

2026-02-11 16:31:29 +01:00

SMT

[mlir][SMT] upstream SMT dialect (#131480)

2025-04-11 17:10:09 -04:00

SparseTensor

[mlir][sparse] Fix a crash if block not have terminator (#180741)

2026-02-11 09:08:45 +08:00

SPIRV

[mlir][spirv] Verify vector types when parsing (#180430)

2026-02-09 07:53:36 -05:00

Tensor

[mlir][tensor]-Handle Dynamic Offset in BubbleUpSliceOpThroughCollapse (#178921)

2026-02-13 01:18:33 +02:00

Tosa

[mlir][tosa] Fix validation of dim op when reliant on datatype extension (#180915)

2026-02-11 12:15:14 +00:00

Transform

Revert "[mlir] Fix the order of operation attribute verification"

2026-02-10 07:39:53 +00:00

UB

[mlir][UB] Add ub.unreachable operation (#169872)

2025-11-28 10:35:18 +00:00

Vector

[mlir][vector] Add finer grained populate methods for multi_reduction (NFC). (#180750)

2026-02-10 21:27:23 +00:00

WasmSSA

[MLIR][WASM] Control flow, conversion and comparison in Wasm importer (#154674)

2025-10-17 13:43:30 +02:00

X86Vector

[mlir][x86vector] AVX10 I8 Dot Op (#178807)

2026-02-02 16:49:00 +05:30

XeGPU

[mlir][xegpu] Add support for vector.reduction and vector.multi_reduction subgroup to work-item distribution. (#180308)

2026-02-13 11:49:55 -08:00

common_folders.mlir

[MLIR] Allow constFoldBinaryOp to fold (T1, T1) -> T2 (#151410)

2025-08-07 17:52:03 +02:00

traits.mlir

…