[clang] [llvm] [RFC][WIP][AMDGPU] Use `bf16` instead of `i16` for bfloat (PR #80908)
Shilei Tian via llvm-commits
llvm-commits at lists.llvm.org
Wed Feb 7 19:58:03 PST 2024
================
@@ -2819,11 +2819,11 @@ def int_amdgcn_fdot2_f16_f16 :
def int_amdgcn_fdot2_bf16_bf16 :
ClangBuiltin<"__builtin_amdgcn_fdot2_bf16_bf16">,
DefaultAttrsIntrinsic<
- [llvm_i16_ty], // %r
+ [llvm_bfloat_ty], // %r
----------------
shiltian wrote:
Does it make sense if we also update both the Clang builtins and LLVM builtins here, and use IR auto upgrade to rewrite existing uses?
https://github.com/llvm/llvm-project/pull/80908
More information about the llvm-commits
mailing list