[libc-commits] [PATCH] D152486: [libc] Begin implementing a 'libmgpu.a' for math on the GPU

Joseph Huber via Phabricator via libc-commits libc-commits at lists.llvm.org
Fri Jun 9 15:21:14 PDT 2023


jhuber6 added a comment.

It's also worth noting that the vendor libraries are not strictly conformant to what a standard `libm` would contain, for example there is no `errno` handling nor raising of floating point exceptions. Additionally, I think a good long-term goal would be to have GPU math implementations directly in-tree here. But I am not a math person. The ROCm device libraries are open source so inspiration could be drawn there if we were to implement our own version https://github.com/RadeonOpenCompute/ROCm-Device-Libs/tree/amd-stg-open/ocml/src. However that's not the immediate value of having a `libm.a` for the GPU so I'm not planning on working on that front for the time being.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D152486/new/

https://reviews.llvm.org/D152486



More information about the libc-commits mailing list