[Mlir-commits] [mlir] [MLIR][XeGPU] Extend unrolling support for scatter ops with chunk_size (PR #144447)

Charitha Saumya llvmlistbot at llvm.org
Mon Jun 16 17:02:08 PDT 2025


================
@@ -256,47 +256,112 @@ gpu.module @test {
   }
 
 //-----
+  // CHECK-LABEL: test_create_tdesc_step_chunk
+  // CHECK-SAME: [[arg0:%.+]]: ui64
+  // CHECK-COUNT-2: xegpu.create_tdesc [[arg0]], {{.*}} : ui64, vector<16xindex> -> !xegpu.tensor_desc<16x4xf32, #xegpu.scatter_tdesc_attr<chunk_size = 4 : i64>>
+  gpu.func @test_create_tdesc_step_chunk(%src: ui64) -> !xegpu.tensor_desc<32x4xf32, #xegpu.scatter_tdesc_attr<chunk_size=4>, #xegpu.layout<inst_data = [16, 4]>> {
----------------
charithaintc wrote:

nit: `test_` is not needed. check tests from other dialects. 

https://github.com/llvm/llvm-project/pull/144447


More information about the Mlir-commits mailing list