Skip to content
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.

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

[CRT][FLAKY] Flaky failures on 32bit platforms #5298

Closed
tqchen opened this issue Apr 9, 2020 · 6 comments
Closed

[CRT][FLAKY] Flaky failures on 32bit platforms #5298

tqchen opened this issue Apr 9, 2020 · 6 comments

Comments

@tqchen
Copy link
Member

tqchen commented Apr 9, 2020

We have some flaky tests for the CRT under the 32bit platforms(even after #5293) , see example log here:

https://ci.tvm.ai/blue/rest/organizations/jenkins/pipelines/tvm/branches/PR-5282/runs/9/nodes/152/log/?start=0

#5297 temporary disabled the tests. It would be great if we look into the potential cause of the problem. One interesting thing is that it only fails in i386, which suggests that it could be a 32bit 64 bit issue.

cc @liangfu @mehrdadhe @tmoreau89

@tqchen tqchen changed the title [CRT][FLAKY] Look into flaky failures on 32bit platforms [CRT][FLAKY] Flaky failures on 32bit platforms Apr 21, 2020
@tqchen
Copy link
Member Author

tqchen commented Jul 21, 2020

cc @liangfu @areusch can you comment on the status? given the recent significant refactor of CRT, shall we conclude that the flaky error no longer exist?

@liangfu
Copy link
Member

liangfu commented Jul 29, 2020

https://github.com/apache/incubator-tvm/blob/44ff1f3b5ed0751fee39537a0e6e3870a74c930b/tests/scripts/task_python_integration.sh#L39

This wouldn't be successful unless #6145 is merged, and let's see whether the PR works on i386.

@tqchen
Copy link
Member Author

tqchen commented Aug 20, 2020

@liangfu @areusch would be great to get your comment about the status of this issue

@areusch
Copy link
Contributor

areusch commented Aug 20, 2020

thoretically I think this should be fixed now. at least, the test is enabled on i386 and CPU regressions, so I think we should close this issue and reopen if we see it again.

@tqchen tqchen closed this as completed Aug 20, 2020
@tqchen
Copy link
Member Author

tqchen commented Aug 20, 2020

Thanks @areusch

@areusch
Copy link
Contributor

areusch commented Sep 29, 2020

Possibly another instance spotted in #6395, but can't view the old jenkins logs to check:

https://ci.tlcpack.ai/blue/organizations/jenkins/tvm/detail/PR-6395/26/pipeline/

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants