llvm-project.git/mlir/lib/Dialect/ArmSVE/TransformOps/ArmSVEVectorTransformOps.cpp, branch main Unnamed repository; edit this file 'description' to name the repository. [MLIR][AArch64] Lower vector.contract to Neon FEAT_BF16 operations (#148198) 2025-07-23T09:15:47+00:00 Momchil Velikov momchil.velikov@arm.com 2025-07-23T09:15:47+00:00 61ce6d70a2fe309d65fe3b7db5b94c17067b9628 This builds upon the framework established by https://github.com/llvm/llvm-project/pull/149810 to add lowering to `bfmmla`. 
This builds upon the framework established by
https://github.com/llvm/llvm-project/pull/149810
to add lowering to `bfmmla`.
 [MLIR][AArch64] Lower `vector.contract` to SVE FEAT_BF16 operations (#147052) 2025-07-22T09:09:28+00:00 Momchil Velikov momchil.velikov@arm.com 2025-07-22T09:09:28+00:00 73b85f87e4771d04e02912a8c2f39ab67910a04d This patch adds lowering of Bfloat16 widening matrix multiply and accumulate `vector.contract`, by parametrising and refactoring the pattern for 8-bit integers. 
This patch adds lowering of Bfloat16 widening matrix multiply and
accumulate `vector.contract`, by parametrising and refactoring the
pattern for 8-bit integers.
 [MLIR] Add apply_patterns.arm_sve.vector_contract_to_i8mm TD Op (#140572) 2025-06-06T09:54:14+00:00 Momchil Velikov momchil.velikov@arm.com 2025-06-06T09:54:14+00:00 b9d3a644c2716e651b388f9fff660b12fdba577c