[Mlir-commits] [mlir] 9adc011 - [mlir][linalg] PadTensorOp vectorization: Avoid redundant FillOps

[Matthias Springer]

Author: Matthias Springer
Date: 2021-09-15T09:28:37+09:00
New Revision: 9adc0114bfeb704ca62d8c369fa52d0530179274

URL: https://github.com/llvm/llvm-project/commit/9adc0114bfeb704ca62d8c369fa52d0530179274
DIFF: https://github.com/llvm/llvm-project/commit/9adc0114bfeb704ca62d8c369fa52d0530179274.diff

LOG: [mlir][linalg] PadTensorOp vectorization: Avoid redundant FillOps

Do not generate FillOps when these would be entirely overwritten.

Differential Revision: https://reviews.llvm.org/D109741

Added: 
    

Modified: 
    mlir/lib/Dialect/Linalg/Transforms/Vectorization.cpp

Removed: 
    


################################################################################
diff  --git a/mlir/lib/Dialect/Linalg/Transforms/Vectorization.cpp b/mlir/lib/Dialect/Linalg/Transforms/Vectorization.cpp
index 7007d8b1101ac..913fa72dc1f2d 100644
--- a/mlir/lib/Dialect/Linalg/Transforms/Vectorization.cpp
+++ b/mlir/lib/Dialect/Linalg/Transforms/Vectorization.cpp
@@ -751,6 +751,13 @@ struct GenericPadTensorOpVectorizationPattern
         padOp.getLoc(), vecType, padOp.source(), readIndices, padValue,
         readInBounds);
 
+    // If `dest` is a FillOp and the TransferWriteOp would overwrite the entire
+    // tensor, write directly to the FillOp's operand.
+    if (llvm::equal(vecShape, resultType.getShape())
+        && llvm::all_of(writeInBounds, [](bool b) { return b; }))
+      if (auto fill = dest.getDefiningOp<FillOp>())
+        dest = fill.output();
+
     // Generate TransferWriteOp.
     auto writeIndices = ofrToIndexValues(
         rewriter, padOp.getLoc(), padOp.getMixedLowPad());