[Mlir-commits] [mlir] [mlir][tosa] Harden folds/canonicalizations for unranked and dynamic shapes (PR #188188)

[Hocky Yudhiono]

================
@@ -1496,7 +1513,7 @@ OpFoldResult SubOp::fold(FoldAdaptor adaptor) {
 }
 
 OpFoldResult GreaterOp::fold(FoldAdaptor adaptor) {
-  auto resultTy = llvm::dyn_cast<RankedTensorType>(getType());
+  auto resultTy = llvm::dyn_cast<ShapedType>(getType());
----------------
hockyy wrote:

also changed for all ops on top, addressed https://github.com/llvm/llvm-project/commit/44184f56a452d508a668a73714334eb7395ea65a

https://github.com/llvm/llvm-project/pull/188188