[Mlir-commits] [mlir] [MLIR][Linalg] Scalable Vectorization of Reduction (PR #97788)

[Zhaoshi Zheng]

================
@@ -1947,13 +1956,30 @@ vectorizeScalableVectorPrecondition(Operation *op,
   if (inputVectorSizes.empty())
     return success();
 
+  auto linalgOp = dyn_cast<LinalgOp>(op);
+  if (linalgOp && isLinalgReduction(linalgOp)) {
+    LDBG("Checking reduce op dims for scalable vectorization\n");
+    auto iteratorTypes = linalgOp.getIteratorTypesArray();
+    assert(iteratorTypes.size() == inputScalableVecDims.size() &&
+           "Number of iterator types and input scalable dims mismatch");
----------------
zhaoshiz wrote:

Oh, I was unsure about size of iterator types array.. and both `inputVectorSizes` and `inputScalableVecDims` are set by user. I guess we can assume the user will ensure correct size of them rather than intentionally break the compiler.

https://github.com/llvm/llvm-project/pull/97788