[Mlir-commits] [mlir] [MLIR][affine] Fix for #115849 Illegal affine loop fusion with vector types (PR #117617)

[llvmlistbot at llvm.org]

================
@@ -285,3 +285,50 @@ module {
     spirv.ReturnValue %3 : !spirv.array<8192 x f32>
   }
 }
+
+// -----
+
+// Basic test for not fusing loops where a vector load depends on 
+// the entire result of a previous loop. store shape < load shape
+
+// CHECK-LABEL: func @should_not_fuse_across_memref_store_load_bounds() {
+func.func @should_not_fuse_across_memref_store_load_bounds() {
+  %a = memref.alloc() : memref<64x512xf32> 
+  %b = memref.alloc() : memref<64x512xf32>
+  %c = memref.alloc() : memref<64x512xf32> 
+  %d = memref.alloc() : memref<64x4096xf32>
+  %e = memref.alloc() : memref<64x4096xf32>
----------------
bane0101 wrote:

%e is unused

https://github.com/llvm/llvm-project/pull/117617