[Mlir-commits] [mlir] [mlir][VectorOps][nfc] Add result pretty printing to `vector.vscale` (PR #83439)

Cullen Rhodes llvmlistbot at llvm.org
Thu Feb 29 23:26:27 PST 2024


================
@@ -19,7 +19,7 @@ func.func @matmul(%A: tensor<1024x512xf32>,
 // CHECK-DAG:  %[[C16:.*]] = arith.constant 16 : index
 // CHECK:      %[[VSCALE:.*]] = vector.vscale
 // CHECK:      %[[STEP:.*]] = arith.muli %[[VSCALE]], %[[C16]] : index
-// CHECK:      %2 = scf.for {{.*}} %[[C0]] to %[[C1024]] step %[[C8]] iter_args(%arg4 = %arg2) -> (tensor<1024x2000xf32>) {
----------------
c-rhodes wrote:

so is `%2` now `vscale`? Not obvious why this has changed

https://github.com/llvm/llvm-project/pull/83439


More information about the Mlir-commits mailing list