[Mlir-commits] [mlir] [mlir][linalg] Add patterns to convert matmul to transposed variants (PR #89075)

Oleksandr Alex Zinenko llvmlistbot at llvm.org
Mon Apr 22 01:57:57 PDT 2024


================
@@ -73,6 +73,21 @@ def ApplyTilingCanonicalizationPatternsOp : Op<Transform_Dialect,
   let assemblyFormat = "attr-dict";
 }
 
+def ApplyTransposeMatmulPatternsOp : Op<Transform_Dialect,
+    "apply_patterns.linalg.transpose_matmul",
+    [DeclareOpInterfaceMethods<PatternDescriptorOpInterface>]> {
+  let description = [{
+    Collects patterns to convert Linalg matmul ops to transposed variants.
+
+    By default the A matrix is transposed, set `transpose_a=false` to transpose
+    B matrix.
----------------
ftynse wrote:

Nit: I would much rather call this "LHS" and "RHS" than A and B.

https://github.com/llvm/llvm-project/pull/89075


More information about the Mlir-commits mailing list