[Mlir-commits] [mlir] [mlir][vector] Add verification for incorrect vector.extract (PR #115824)

[Andrzej Warzyński]

================
@@ -1339,6 +1339,50 @@ bool ExtractOp::isCompatibleReturnTypes(TypeRange l, TypeRange r) {
   return l == r;
 }
 
+// Common verification rules for `InsertOp` and `ExtractOp` involving indices.
+// `indexedType` is the vector type being indexed in the operation, i.e., the
+// destination type in InsertOp and the source type in ExtractOp.
+// `vecOrScalarType` is the type that is not indexed in the op and can be
+// either a scalar or a vector, i.e., the source type in InsertOp and the
+// return type in ExtractOp.
+static LogicalResult verifyInsertExtractIndices(Operation *op,
+                                                VectorType indexedType,
+                                                int64_t numIndices,
+                                                Type vecOrScalarType) {
+  int64_t indexedRank = indexedType.getRank();
+  if (numIndices > indexedRank)
+    return op->emitOpError(
+        "expected a number of indices no greater than the indexed vector rank");
+
+  if (auto nonIndexedVecType = dyn_cast<VectorType>(vecOrScalarType)) {
+    // Vector case, including:
+    //  * 0-D vector:
+    //    * vector.extract %src[2]: vector<f32> from vector<8xf32)
----------------
banach-space wrote:

> Honestly, I wouldn't mind going back to that state.

We're on the same page here.

> By making it stricter, even if too much, intentionally, we can have discussions on actual use cases on a case by case basis.

Now I see what you are aiming for, makes sense.

> One thing I really want to avoid is allowing cases that lead to the introduction of canonicalization patterns to turn vector.extract/insert into vector.bitcast

Between e.g. `vector<f32>` and `vector<1xf32>`? Or more general?

https://github.com/llvm/llvm-project/pull/115824