[Mlir-commits] [mlir] [MLIR] Allowing unsupported conv2d op to fail gracefully from vectorization (PR #130181)

[Han-Chung Wang]

================
@@ -1939,6 +1939,127 @@ vectorizeInsertSliceOpPrecondition(tensor::InsertSliceOp sliceOp,
   return success();
 }
 
+namespace {
+bool isCastOfBlockArgument(Operation *op) {
+  return isa<CastOpInterface>(op) && op->getNumOperands() == 1 &&
+         isa<BlockArgument>(op->getOperand(0));
+}
+
+// Returns true iff it is a valid conv/pooling op.
+// If (region has 2 ops (reduction + yield) or 3 ops (extension + reduction
+// + yield) and rhs is not used) then it is the body of a pooling
+// If conv, check for single `mul` predecessor. The `mul` operands must be
+// block arguments or extension of block arguments.
+// Otherwise, check for one or zero `ext` predecessor. The `ext` operands
+// must be block arguments or extension of block arguments.
+enum OperKind { Conv, Pool };
+bool getOperKind(Operation *reduceOp, OperKind &oper) {
----------------
hanhanW wrote:

I think it is easier if you return `std::optional<OperKind>`. It is better in practice because callers do not need to declare a variable, and the API looks cleaner.

https://github.com/llvm/llvm-project/pull/130181