llvm-project

History

[MLIR][Vector] Enhance vector.multi_reduction unrolling to handle scalar result (#188633 )

Previously, UnrollMultiReductionPattern bailed out when all the
dimensions were reduced to a scalar. This PR adds support for this case
by tiling the source vector and chaining partial reductions through the
accumulator operand.

2026-04-01 14:59:08 -07:00

BufferizableOpInterfaceImpl.cpp

[MLIR] Extend vector.scatter to accept tensor as base (#165548 )

2025-11-14 19:56:24 +00:00

CMakeLists.txt

[mlir] Move vector.{to_elements,from_elements} unrolling to VectorUnroll.cpp (#159118 )

2025-09-18 12:03:54 -04:00

LowerVectorBitCast.cpp

…

LowerVectorBroadcast.cpp

[MLIR][Vector] Move scalable dims check before IR creation in BroadcastOpLowering (#188953 )

2026-03-27 12:24:57 +01:00

LowerVectorContract.cpp

Fix outdated docs with vector.reduce instead of vector.reduction (#178111 )

2026-02-02 17:52:04 +00:00

LowerVectorGather.cpp

[mlir][vector] Lower vector.gather with delinearization approach (#184706 )

2026-03-13 14:58:40 -07:00

LowerVectorInterleave.cpp

[mlir] Apply clang-tidy check llvm-use-vector-utils. NFC. (#178526 )

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

LowerVectorMask.cpp

[mlir][vector] Simplify op rewrite pattern inheriting constructors. NFC. (#161670 )

2025-10-02 19:07:25 -04:00

LowerVectorMultiReduction.cpp

[mlir][vector] Replace OneDimMultiReductionToTwoDim with OneDimMultiReductionToReduction (#184241 )

2026-03-04 16:13:11 +00:00

LowerVectorScan.cpp

[MLIR][Vector] Move scalable dims check before IR creation in ScanToArithOps (#188954 )

2026-03-27 12:25:14 +01:00

LowerVectorShapeCast.cpp

[mlir] Use llvm accumulate wrappers. NFCI. (#162957 )

2025-10-11 11:33:18 -04:00

LowerVectorShuffle.cpp

[mlir] Switch uses of deprecated .create methods to free function. NFC. (#164635 )

2025-10-22 14:51:03 +00:00

LowerVectorStep.cpp

[mlir][vector] Simplify op rewrite pattern inheriting constructors. NFC. (#161670 )

2025-10-02 19:07:25 -04:00

LowerVectorToFromElementsToShuffleTree.cpp

[mlir][vector] Simplify op rewrite pattern inheriting constructors. NFC. (#161670 )

2025-10-02 19:07:25 -04:00

LowerVectorTransfer.cpp

…

LowerVectorTranspose.cpp

[mlir][vector] Simplify op rewrite pattern inheriting constructors. NFC. (#161670 )

2025-10-02 19:07:25 -04:00

SubsetOpInterfaceImpl.cpp

…

VectorDistribute.cpp

[MLIR][Vector] Fix direct operand.set() bypassing rewriter in WarpOpScfIfOp/ForOp (#188948 )

2026-03-30 12:28:07 +02:00

VectorDropLeadUnitDim.cpp

[mlir] Use llvm accumulate wrappers. NFCI. (#162957 )

2025-10-11 11:33:18 -04:00

VectorEmulateMaskedLoadStore.cpp

[mlir][vector] Fix masked load/store emulation for rank-0 memrefs (#173325 )

2026-01-05 17:56:09 +00:00

VectorEmulateNarrowType.cpp

[MLIR][Vector] Fix RewriteAlignedSubByteIntExt/Trunc producing invalid IR when source is already i8 (#188941 )

2026-03-27 12:34:47 +01:00

VectorInsertExtractStridedSliceRewritePatterns.cpp

[mlir][vector] Simplify op rewrite pattern inheriting constructors. NFC. (#161670 )

2025-10-02 19:07:25 -04:00

VectorLinearize.cpp

[mlir] Fix new clang-tidy warning llvm-type-switch-case-types. NFC. (#178487 )

2026-01-28 19:13:47 +00:00

VectorMaskElimination.cpp

[mlir][vector] Skip vector mask elimination for funcs without body (#173330 )

2025-12-29 22:37:16 +00:00

VectorTransferOpTransforms.cpp

[mlir][vector] Drop trailing 1-dims from constant_mask (#187383 )

2026-04-01 09:05:25 -04:00

VectorTransferSplitRewritePatterns.cpp

[mlir] Apply clang-tidy check llvm-use-vector-utils. NFC. (#178526 )

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

VectorTransforms.cpp

[mlir][vector] Add support for dropping inner unit dims for transfer_read/write with masks. (#188841 )

2026-03-27 10:21:20 -07:00

VectorUnroll.cpp

[MLIR][Vector] Enhance vector.multi_reduction unrolling to handle scalar result (#188633 )

2026-04-01 14:59:08 -07:00