[Mlir-commits] [mlir] [MLIR] Add `InParallelOpInterface` for parallel combining operations (PR #157736)

[Alan Li]

================
@@ -1911,8 +1917,9 @@ struct FoldTensorCastOfOutputIntoForallOp
     auto terminator = newForallOp.getTerminator();
     for (auto [yieldingOp, outputBlockArg] : llvm::zip(
              terminator.getYieldingOps(), newForallOp.getRegionIterArgs())) {
-      auto insertSliceOp = cast<tensor::ParallelInsertSliceOp>(yieldingOp);
-      insertSliceOp.getDestMutable().assign(outputBlockArg);
+      auto insertSliceOp = dyn_cast<tensor::ParallelInsertSliceOp>(yieldingOp);
+      if (insertSliceOp)
+        insertSliceOp.getDestMutable().assign(outputBlockArg);
----------------
lialan wrote:

Should be applicable to all ops implementing the interface. Updated.


https://github.com/llvm/llvm-project/pull/157736