[llvm] AMDGPU: Add round-to-odd rounding during f64 to bf16 conversion (PR #133995)

Changpeng Fang via llvm-commits llvm-commits at lists.llvm.org
Thu Apr 3 08:42:46 PDT 2025


changpeng wrote:

Ping. Is there any further suggestions or comment?
This is the patch from the downstream branches, and the logic is needed to fix the double rounding issue. 
Thanks. 

https://github.com/llvm/llvm-project/pull/133995


More information about the llvm-commits mailing list