[all-commits] [llvm/llvm-project] f6ddf3: [mlir][ArmNeon] Reject masked contractions in `I8M...

[Federico Bruzzone via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: f6ddf3965910afb16285418fe7bbb9ca36a8b18e
      https://github.com/llvm/llvm-project/commit/f6ddf3965910afb16285418fe7bbb9ca36a8b18e
  Author: Federico Bruzzone <federico.bruzzone.i at gmail.com>
  Date:   2026-08-06 (Thu, 06 Aug 2026)

  Changed paths:
    M mlir/lib/Dialect/ArmNeon/Transforms/LowerContractToNeonPatterns.cpp
    M mlir/test/Dialect/ArmNeon/lower-to-arm-neon.mlir

  Log Message:
  -----------
  [mlir][ArmNeon] Reject masked contractions in `I8MM/BFMMLA` lowering patterns (#213698)

`LowerContractionToNeonI8MMPattern` and
`LowerContractionToNeonBFMMLAPattern` rewrite a `vector.contract` into
several ops (multiple `arm_neon.intr.smmla`/`bfmmla` calls plus
`extract/insert/shape_cast` scaffolding).

When the matched `vector.contract` sits inside a `vector.mask` (e.g.,
from vectorizing a matmul with dynamically-shaped operands), this
in-place multi-op rewrite violates `vector.mask`'s invariant that its
region contain exactly **one** operation, and the verifier rejects the
result: _error: 'vector.mask' op expects only one operation to mask_.

Neither pattern checked for masking, unlike the generic
contraction-lowering
patterns. To fix the problem, we bail out via `notifyMatchFailure` when
the contraction is masked, so it falls through to a different lowering
path instead of producing invalid IR.

Found while adding neon `linalg.matmul` e2e tests in #212809.
AI was used to investigate the problem.

---------

Signed-off-by: Federico Bruzzone <federico.bruzzone.i at gmail.com>



To unsubscribe from these emails, change your notification settings at https://github.com/llvm/llvm-project/settings/notifications