[Mlir-commits] [mlir] [mlir][ArmSME] Support 4-way widening outer products (PR #79288)

[Andrzej Warzyński]

================
@@ -213,6 +213,581 @@ func.func @outerproduct_sub_widening_2way_unsigned_i16i16i32(
   return %1 : vector<[4]x[4]xi32>
 }
 
+// -----
+
+// CHECK-LABEL: @outerproduct_add_widening_4way_signed_i8i8i32
+// CHECK-SAME:    %[[A0:.*]]: vector<[4]xi8>, %[[B0:.*]]: vector<[4]xi8>, %[[A1:.*]]: vector<[4]xi8>, %[[B1:.*]]: vector<[4]xi8>, %[[A2:.*]]: vector<[4]xi8>, %[[B2:.*]]: vector<[4]xi8>, %[[A3:.*]]: vector<[4]xi8>, %[[B3:.*]]: vector<[4]xi8>,
+// CHECK-SAME:    %[[A0_MASK:.*]]: vector<[4]xi1>, %[[B0_MASK:.*]]: vector<[4]xi1>, %[[A1_MASK:.*]]: vector<[4]xi1>, %[[B1_MASK:.*]]: vector<[4]xi1>, %[[A2_MASK:.*]]: vector<[4]xi1>, %[[B2_MASK:.*]]: vector<[4]xi1>, %[[A3_MASK:.*]]: vector<[4]xi1>, %[[B3_MASK:.*]]: vector<[4]xi1>
----------------
banach-space wrote:

[nit] I'd be tempted to split these across multiple lines

https://github.com/llvm/llvm-project/pull/79288