[Mlir-commits] [mlir] [mlir][affine]make affine-loop-unroll a FunctionOpInterface pass. (PR #126475)

[Kunwar Grover]

================
@@ -240,6 +240,23 @@ func.func @loop_nest_unroll_full() {
   return
 } // UNROLL-FULL }
 
+gpu.module @unroll_full {
+  // UNROLL-FULL-LABEL: func @gpu_loop_nest_simplest() {
+  gpu.func @gpu_loop_nest_simplest() {
+    // UNROLL-FULL: affine.for %arg0 = 0 to 100 step 2 {
+    affine.for %i = 0 to 100 step 2 {
+      // UNROLL-FULL: %c1_i32 = arith.constant 1 : i32
+      // UNROLL-FULL-NEXT: %c1_i32_0 = arith.constant 1 : i32
+      // UNROLL-FULL-NEXT: %c1_i32_1 = arith.constant 1 : i32
+      // UNROLL-FULL-NEXT: %c1_i32_2 = arith.constant 1 : i32
+      affine.for %j = 0 to 4 {
+        %x = arith.constant 1 : i32
+      }
+    }       // UNROLL-FULL:  }
+    gpu.return  // UNROLL-FULL:  return
+  }
+}
+
----------------
Groverkss wrote:

I don't think we need a test enumerated for every gpu.func case... The functionality is same, just add one simple test. Having a minimal set of tests is important.

https://github.com/llvm/llvm-project/pull/126475