[Mlir-commits] [mlir] [mlir][Linalg] Add a pattern to fold concats of fill. (PR #98995)

[llvmlistbot at llvm.org]

================
@@ -879,15 +879,66 @@ struct FoldFillWithTranspose : OpRewritePattern<linalg::TransposeOp> {
   }
 };
 
+// Fold a concat with all elements being fills of the same value
+// into a fill of the concat result shape.
+struct FoldConcatsOfFill : public OpRewritePattern<tensor::ConcatOp> {
+  using OpRewritePattern::OpRewritePattern;
+
+  LogicalResult matchAndRewrite(tensor::ConcatOp concatOp,
+                                PatternRewriter &rewriter) const override {
+    auto concatOperands = concatOp.getInputs();
+    if (concatOperands.empty()) {
----------------
MaheshRavishankar wrote:

I could. So this is
```
%0 = linalg.fill ins (%cst : f32) outs(%init : tensor<...xf32>) -> tensor<...xf32>
%1 = tensor.concat dims(...) %0 : ...
```

I would expect to canonicalize to just
```
%1 = linalg.fill ...
```

but irrespective of that, this pattern would

```
%0 = tensor.concat dim(...) %init : 
%1 = linalg.fill ins(%cst : f32) outs(%0 : tensor<...xf32>) -> tensor<...xf32>
```

which would still a good thing to do and would give what would be best if that canonicalization pattern existed. So I am saying it might be worth even handling the corner case of 1 operand concat even though that shouldnt happen.

https://github.com/llvm/llvm-project/pull/98995