[Mlir-commits] [mlir] [mlir][linalg] Generic to category specialization (PR #184624)

[Andrzej Warzyński]

================
@@ -17,6 +18,10 @@ func.func @unary_op_exp(%A: tensor<?x?x?xf32>, %Out: tensor<?x?x?xf32>) -> tenso
 // CHECK-NOT: linalg.generic
 // CHECK: linalg.exp ins(%[[A]] : tensor<?x?x?xf32>) outs(%[[Out]] : tensor<?x?x?xf32>) -> tensor<?x?x?xf32>
 
+// Not supported yet.
+// CATEGORY-LABEL: unary_op_exp
----------------
banach-space wrote:

Here's an example: https://github.com/llvm/llvm-project/blob/ae363d50ad292b4d8c33bf795ad16c79eb640b19/mlir/test/Dialect/Vector/vector-multi-reduction-unrolling.mlir?plain=1#L1-L4

For this particular test, you would use it like this:
```mlir
// RUN: mlir-opt %s -split-input-file -linalg-morph-ops=generic-to-named | FileCheck %s --check-prefixes=NAMED,ALL
// RUN: mlir-opt %s -split-input-file -linalg-morph-ops=generic-to-category | FileCheck %s --check-prefixes=CATEGORY,ALL

// ALL-LABEL:  unary_op_exp
```

https://github.com/llvm/llvm-project/pull/184624