Migrate CUDA unary complex kernel to c10::complex #37647
Conversation
zasdfgbnm
changed the title
|
@anjali411 This is ready |
💊 Build failures summary and remediationsAs of commit d668228 (more details on the Dr. CI page): 💚 💚 Looks good so far! There are no failures yet. 💚 💚 This comment was automatically generated by Dr. CI (expand for details).Follow this link to opt-out of these comments for your Pull Requests.Please report bugs/suggestions on the GitHub issue tracker. This comment has been revised 3 times. |
There was a problem hiding this comment.
@anjali411 has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
|
@anjali411 The fix for ROCm is #ifdef __HIP_PLATFORM_HCC__
#define ROCm_Bug(x)
#else
#define ROCm_Bug(x) x
#endif
template<typename T>
C10_HOST_DEVICE T arg(const c10::complex<T>& z) {
return ROCm_Bug(std)::atan2(std::imag(z), std::real(z));
}
#undef ROCm_BugI am being very explicit on all the compiler bugs I found so that people can have a good idea of why the code is written that way. That's why there were a |
There was a problem hiding this comment.
@anjali411 has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
|
@anjali411 merged this pull request in d7ccb4b. |
Summary: Pull Request resolved: pytorch#37647 Differential Revision: D21351018 Pulled By: anjali411 fbshipit-source-id: 51e4a4a3bdc9b3f8b9f7a5e0d65c06f209c55401
5 participants
No description provided.