[Mlir-commits] [mlir] [mlir][xegpu] Add SIMT distribution support for GEMM transpose B case. (PR #155517)

[Adam Siemieniuk]

================
@@ -1276,6 +1305,142 @@ struct VectorShapeCastDistribution : public gpu::WarpDistributionPattern {
   }
 };
 
+/// Sink a memref::ExtractAlignedPointerAsIndex op feeding into yield op of an
+/// enclosing `gpu.warp_execute_on_lane_0` region. This will simply move the op
+/// outside of the warp op.
+struct MemrefExtractAlignedPointerAsIndexDistribution final
+    : public gpu::WarpDistributionPattern {
+  using gpu::WarpDistributionPattern::WarpDistributionPattern;
+  LogicalResult matchAndRewrite(gpu::WarpExecuteOnLane0Op warpOp,
+                                PatternRewriter &rewriter) const override {
+    OpOperand *operand = getWarpResult(
+        warpOp, llvm::IsaPred<memref::ExtractAlignedPointerAsIndexOp>);
+    if (!operand)
+      return rewriter.notifyMatchFailure(
+          warpOp,
+          "warp result is not a xegpu::MemrefExtractAlignedPointerAsIndex op");
----------------
adam-smnk wrote:

typo in the op's name

https://github.com/llvm/llvm-project/pull/155517