[Mlir-commits] [mlir] [mlir][vector] Fix vector.gather lowering for strided memrefs. (PR #184706)

[Andrzej Warzyński]

================
@@ -183,15 +190,21 @@ struct Gather1DToConditionalLoads : OpRewritePattern<vector::GatherOp> {
     Value condMask = op.getMask();
     Value base = op.getBase();
 
-    // vector.load requires the most minor memref dim to have unit stride
-    // (unless reading exactly 1 element)
+    // For multi-dimensional memrefs, use linearize+delinearize to compute
+    // correct N-D load indices from the 1-D gather offset.
+    bool useDelinearization = false;
     if (auto memType = dyn_cast<MemRefType>(base.getType())) {
+      // vector.load requires the most minor memref dim to have unit stride
+      // (unless reading exactly 1 element).
       if (auto stridesAttr =
               dyn_cast_if_present<StridedLayoutAttr>(memType.getLayout())) {
         if (stridesAttr.getStrides().back() != 1 &&
             resultTy.getNumElements() != 1)
           return failure();
----------------
banach-space wrote:

[nit] Since you are adding comments, would you mind replacing this with `notifyMatchFailure`?

https://github.com/llvm/llvm-project/pull/184706