[llvm] [clang] [RISCV] Support Xsfvfwmaccqqq extensions (PR #68296)

Shao-Ce SUN via llvm-commits llvm-commits at lists.llvm.org
Thu Nov 2 21:18:58 PDT 2023


================
@@ -103,3 +103,15 @@ let SupportOverloading = false in {
     defm sf_vc_v_fvw : RVVVCIXBuiltinSet<["si"],  "UwKzUwUvFe", [-1, 0, 2, 3], UseGPR=0>;
   }
 }
+
+multiclass RVVVFWMACCBuiltinSet<list<list<string>> suffixes_prototypes> {
+  let OverloadedName = NAME,
+      Name = NAME,
+      HasMasked = false,
+      Log2LMUL = [-2, -1, 0, 1, 2] in
+    defm NAME : RVVOutOp1Op2BuiltinSet<NAME, "x", suffixes_prototypes>;
----------------
sunshaoce wrote:

> The SiFive custom vector extension Xsfvfwmaccqqq introduces the Bfloat16 matrix multiply accumulate instruction.


This is wrong! It should be `bfloat16` instead of `float16`. Need to be based on: https://reviews.llvm.org/D152498

https://github.com/llvm/llvm-project/pull/68296


More information about the llvm-commits mailing list