KEMBAR78
cast return of cudaGetLastError() to void when discarding by jeffdaily · Pull Request #62518 · pytorch/pytorch · GitHub
Skip to content

cast return of cudaGetLastError() to void when discarding #62518

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Closed
wants to merge 1 commit into from
Closed

cast return of cudaGetLastError() to void when discarding #62518

wants to merge 1 commit into from

Conversation

@jeffdaily
Copy link
Collaborator

Fixes #62511.

@jeffdaily jeffdaily requested a review from ngimel July 31, 2021 00:10
@facebook-github-bot
Copy link
Contributor

facebook-github-bot commented Jul 31, 2021 •
edited
Loading

🔗 Helpful links

💊 CI failures summary and remediations

As of commit 517fe87 (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 to the (internal) Dr. CI Users group.

Click here to manually regenerate this comment.

@facebook-github-bot
Copy link
Contributor

@ngimel has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

@malfet
Copy link
Contributor

malfet commented Aug 2, 2021

Unfortunately it does not fix all the issues, digging further...

@jeffdaily
Copy link
Collaborator Author

Let me know how I can help, if there is any sort of build log I could access.

@malfet
Copy link
Contributor

malfet commented Aug 3, 2021 •
edited
Loading

@jeffdaily do you know if older version of hipcc could generate following error:

caffe2/aten/src/ATen/core/TensorAccessor.h:160:5: error: throw is prohibited in AMP-restricted functions
    TORCH_CHECK_INDEX(
    ^

Can workaround it by disabling this check for HIP code, but wonder if better fix would be to just make this function CPU-only.

Edit, replacing C10_HOST_DEVICE with C10_HOST indeed fixes the issue, testing if it passes the CI here #62628

@malfet
Copy link
Contributor

malfet commented Aug 3, 2021

Reduced the PR to change in only 4 files: c10/cuda/impl/CUDAGuardImpl.h, c10/cuda/CUDAStream.h, caffe2/core/context_gpu.h and caffe2/core/event_gpu.cc, landing it now

@facebook-github-bot
Copy link
Contributor

@malfet merged this pull request in b7391f4.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@ngimel ngimel Awaiting requested review from ngimel

Assignees

No one assigned

Labels

cla signed Merged open source

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

Internal builds are failing with HIP errors

4 participants

@jeffdaily @facebook-github-bot @malfet @pytorchbot