[Mlir-commits] [mlir] [mlir][linalg] Add runtime verification for linalg ops (PR #89342)

[Matthias Springer]

================
@@ -0,0 +1,298 @@
+// RUN: mlir-opt %s -generate-runtime-verification \
+// RUN: -one-shot-bufferize="bufferize-function-boundaries" \
+// RUN: -convert-linalg-to-loops \
+// RUN: -expand-strided-metadata \
+// RUN: -lower-affine \
+// RUN: -convert-scf-to-cf \
+// RUN: -test-cf-assert \
+// RUN: -convert-index-to-llvm \
+// RUN: -finalize-memref-to-llvm \
+// RUN: -convert-func-to-llvm \
+// RUN: -reconcile-unrealized-casts | \
+// RUN: mlir-cpu-runner -e main -entry-point-result=void \
+// RUN:     -shared-libs=%mlir_runner_utils \
+// RUN:     -shared-libs=%mlir_c_runner_utils 2>&1 | \
+// RUN: FileCheck %s
+
+func.func @main() {
+  %c5x = arith.constant dense<0.0> : tensor<5xf32>
+  %c4x = arith.constant dense<0.0> : tensor<4xf32>
+  %d5x = tensor.cast %c5x : tensor<5xf32> to tensor<?xf32>
+  %d4x = tensor.cast %c4x : tensor<4xf32> to tensor<?xf32>
+
+  // CHECK-NOT: ERROR: Runtime op verification failed
+  func.call @simple_add(%d5x, %d5x) : (tensor<?xf32>, tensor<?xf32>) -> (tensor<?xf32>)
+
+  // CHECK: ERROR: Runtime op verification failed
----------------
matthias-springer wrote:

This was meant to check that the above `call @simple_add` succeeds (no error) and the this `call @simple_add` fails (error).

We should probably move all passing cases to the end and put a `CHECK-NOT` there.

https://github.com/llvm/llvm-project/pull/89342