[Mlir-commits] [mlir] [MLIR][XeGPU] Update layout for simple_mxfp_gemm_dequantizeB_F4.mlir (PR #203636)

[Sang Ik Lee]

================
@@ -8,9 +8,9 @@
 
 // XFAIL: *
 // Note: layouts used by dpas_mx need to match HW constaint. Otherwise dpas_mx is not unrolled.
-#a = #xegpu.layout<sg_layout = [2, 2], sg_data = [16, 1024], inst_data = [8, 64], lane_layout = [1, 16], lane_data = [1, 1]>
-#b_packed = #xegpu.layout<sg_layout = [2, 2], sg_data = [512, 16], inst_data = [32, 16], lane_layout = [1, 16], lane_data = [4, 1]>
-#b = #xegpu.layout<sg_layout = [2, 2], sg_data = [1024, 16], inst_data = [64, 16], lane_layout = [1, 16], lane_data = [8, 1]>
+#a = #xegpu.layout<sg_layout = [2, 2], sg_data = [16, 1024], inst_data = [8, 16], lane_layout = [1, 16], lane_data = [1, 1]>
----------------
silee2 wrote:

"#a" is used for loading A and feed to plain "dpas". so inst_data = [8, 16], lane_layout =  [1, 16], lane_data = [1,1]  

https://github.com/llvm/llvm-project/pull/203636