[Mlir-commits] [mlir] [mlir][linalg] Add pass to transpose A matrix of matmul op (PR #89075)

[Andrzej Warzyński]

================
@@ -141,4 +141,13 @@ def LinalgDetensorizePass : InterfacePass<"linalg-detensorize", "FunctionOpInter
   ];
 }
 
+def LinalgMatmulToMatmulTransposeAPass
+    : Pass<"linalg-matmul-to-matmul-transpose-a"> {
+  let summary = "Converts `linalg.matmul` to `linalg.matmul_transpose_a`.";
+  let dependentDialects = ["linalg::LinalgDialect"];
+  let description = [{
+    Transposes the A matrix of a `linalg.matmul` for contiguous access.
----------------
banach-space wrote:

[nit] It's more like "Replace linalg.matmul with `linalg.transpose` + `linalg.matmul_transpose_a`" :)

Also:
>  for contiguous access

This is a fairly low-level concept that might be irrelevant in the case of "linalg on tensors". And even for "linalg on memrefs", it might not matter for some hardware. It's obviously very beneficial for CPUs :) 

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