[Mlir-commits] [mlir] [MLIR][XeGPU] Add lowering from transfer_read/transfer_write to load_gather/store_scatter (PR #152429)

[Chao Chen]

================
@@ -165,6 +496,20 @@ struct TransferReadLowering : public OpRewritePattern<vector::TransferReadOp> {
     if (failed(transferPreconditions(rewriter, readOp)))
       return failure();
 
+    auto chip = xegpu::getXeGPUChipStr(readOp);
+    if ( chip != "pvc" && chip != "bmg") {
----------------
chencha3 wrote:

So the logical is transferReadOp will be lowered into LoadGather for PVC amd bmg. and to LoadNd for others?  

https://github.com/llvm/llvm-project/pull/152429