[Mlir-commits] [mlir] [mlir][tosa] Allow rank-0 vector operands in tosa.apply_scale (PR #199924)

[Luke Hutton]

================
@@ -1777,3 +1777,10 @@ func.func @test_assert_equal_shape() {
   tosa.assert_equal_shape %0, %1 {allow_broadcast = true} : (!tosa.shape<2>, !tosa.shape<2>) -> ()
   return
 }
+
+// -----
+// CHECK-LABEL: test_apply_scale_rank_0_vector
+func.func @test_apply_scale_rank_0_vector(%arg0: vector<i32>, %arg1: vector<i32>, %arg2: vector<i8>) -> vector<i32> {
----------------
lhutton1 wrote:

nit: I don't think we need this test. This op is more of a utility to aid lowering and should only be produced during conversion to linalg. Perhaps it's worth adding a new test case in tosa-to-linalg to demonstrate apply_scale is correctly constructed for scalar values?

https://github.com/llvm/llvm-project/pull/199924