[clang] [mlir] [llvm] [AMDGPU] Add GFX12 WMMA and SWMMAC instructions (PR #77795)

[Krzysztof Drewniak via llvm-commits]

================
@@ -253,22 +253,22 @@ def ROCDL_mfma_f32_32x32x16_fp8_fp8 : ROCDL_Mfma_IntrOp<"mfma.f32.32x32x16.fp8.f
 
 //===---------------------------------------------------------------------===//
 // WMMA intrinsics
-class ROCDL_Wmma_IntrOp<string mnemonic, list<Trait> traits = []> :
+class ROCDL_Wmma_IntrOp<string mnemonic, list<int> overloadedOperands> :
   LLVM_IntrOpBase<ROCDL_Dialect, mnemonic,
                   "amdgcn_" # !subst(".","_", mnemonic),
-                  [0], [], traits, 1>,
+                  [0], overloadedOperands, [], 1>,
   Arguments<(ins Variadic<LLVM_Type>:$args)> {
   let assemblyFormat =
     "$args attr-dict `:` functional-type($args, $res)";
 }
 
 // Available on RDNA3
-def ROCDL_wmma_f32_16x16x16_f16 : ROCDL_Wmma_IntrOp<"wmma.f32.16x16x16.f16">;
-def ROCDL_wmma_f32_16x16x16_bf16 : ROCDL_Wmma_IntrOp<"wmma.f32.16x16x16.bf16">;
-def ROCDL_wmma_f16_16x16x16_f16 : ROCDL_Wmma_IntrOp<"wmma.f16.16x16x16.f16">;
-def ROCDL_wmma_bf16_16x16x16_bf16 : ROCDL_Wmma_IntrOp<"wmma.bf16.16x16x16.bf16">;
-def ROCDL_wmma_i32_16x16x16_iu8 : ROCDL_Wmma_IntrOp<"wmma.i32.16x16x16.iu8">;
-def ROCDL_wmma_i32_16x16x16_iu4 : ROCDL_Wmma_IntrOp<"wmma.i32.16x16x16.iu4">;
+def ROCDL_wmma_f32_16x16x16_f16 : ROCDL_Wmma_IntrOp<"wmma.f32.16x16x16.f16", [0]>;
----------------
krzysz00 wrote:

(This isn't PR-critical, we can clean this up later)
(... though if we're here, we might as well add the new WMMA intrinsics to this file)

https://github.com/llvm/llvm-project/pull/77795