[MLIR][Linalg] Remove matmul_transpose variants (#147961)

Removes the `(batch_)matmul_transpose_{a|b}` variants from OpDSL and
replace it with `matmul affine_maps [...]` whenever appropriate. This is
in line with the
[plan](https://discourse.llvm.org/t/rfc-op-explosion-in-linalg/82863),
and can be done since #104783 merged.

See:
https://discourse.llvm.org/t/deprecate-batch-matmul-transpose-a-b-linalg-operations/87245

Issues investigated:
* pad transform tests that could use `matmul` instead, so change to
that.
* ArmSME test using transpose actually needed it, so changed to `matmul`
+ affine maps.

Arm tests validated by @banach-space (thanks!!).

1 files changed, 1 insertions, 7 deletions

diff --git a/mlir/lib/Dialect/Linalg/Transforms/Vectorization.cpp b/mlir/lib/Dialect/Linalg/Transforms/Vectorization.cpp
index cf65e673a5c4..80fbe3cb9ff1 100644
--- a/mlir/lib/Dialect/Linalg/Transforms/Vectorization.cpp
+++ b/mlir/lib/Dialect/Linalg/Transforms/Vectorization.cpp
@@ -2563,7 +2563,7 @@ vectorizeScalableVectorPrecondition(Operation *op,
                 "vectorization";
       return failure();
     }
-    if (isa<linalg::MatmulOp>(op) || isa<linalg::MatmulTransposeAOp>(op)) {
+    if (isa<linalg::MatmulOp>(op)) {
       LDBG()
           << "Scalable vectorization of the reduction dim in Matmul-like ops "
              "is not supported";
@@ -2604,15 +2604,9 @@ vectorizeScalableVectorPrecondition(Operation *op,
       return failure();
   }
 
-  // Check to not let go the matmul with extended semantic, through this
-  // transform.
-  if (linalgOp.hasUserDefinedMaps())
-    return failure();
-
   // Cond 4: Only the following ops are supported in the
   // presence of scalable vectors
   return success(isElementwise(linalgOp) || isa<linalg::MatmulOp>(op) ||
-                 isa<linalg::MatmulTransposeAOp>(op) ||
                  isa<linalg::DepthwiseConv1DNwcWcOp>(op) ||
                  isa<linalg::MatvecOp>(op) || isa<linalg::Mmt4DOp>(op) ||
                  hasReductionIterator(linalgOp));