summaryrefslogtreecommitdiff
path: root/mlir/lib/Dialect/ArmSVE/TransformOps/ArmSVEVectorTransformOps.cpp
AgeCommit message (Collapse)Author
2025-07-23[MLIR][AArch64] Lower vector.contract to Neon FEAT_BF16 operations (#148198)Momchil Velikov
This builds upon the framework established by https://github.com/llvm/llvm-project/pull/149810 to add lowering to `bfmmla`.
2025-07-22[MLIR][AArch64] Lower `vector.contract` to SVE FEAT_BF16 operations (#147052)Momchil Velikov
This patch adds lowering of Bfloat16 widening matrix multiply and accumulate `vector.contract`, by parametrising and refactoring the pattern for 8-bit integers.
2025-06-06[MLIR] Add apply_patterns.arm_sve.vector_contract_to_i8mm TD Op (#140572)Momchil Velikov