[Mlir-commits] [mlir] [mlir][AMDGPU] Allow non-contiguous destination memrefs for gather_to_lds (PR #152559)

[Quinn Dawkins]

================
@@ -518,9 +518,6 @@ LogicalResult GatherToLDSOp::verify() {
   MemRefType srcType = cast<MemRefType>(getSrc().getType());
   MemRefType dstType = cast<MemRefType>(getDst().getType());
 
-  if (!dstType.areTrailingDimsContiguous(dstType.getRank()))
-    return emitOpError("destination types must be contiguous");
-
----------------
qedawkins wrote:

Hmmm, yeah inner dim contiguous is probably fine to check. Technically if someone was generating code with a `memref<...x i8192>` that'd still be a false positive, but seems like a ridiculous case to plan for.

https://github.com/llvm/llvm-project/pull/152559