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

cmake: clarify test parallelization #67

Merged
merged 2 commits into from
Oct 24, 2024
Merged

Conversation

junghans
Copy link
Contributor

@junghans junghans commented Oct 23, 2024

To fix over subscribing (see #59 (comment))

@junghans
Copy link
Contributor Author

There is also a way to specify the number of GPUs a test needs: https://cmake.org/cmake/help/latest/prop_test/RESOURCE_GROUPS.html

@mkstoyanov
Copy link
Collaborator

I've looked into the groups but it seems a bit too complicated for heFFTe. Testing does not take long.

How does this PR help? The PROCESSORS property indicates the number of ranks that will be used by the test, provably redundant if RUN_SERIAL is set, but how is it that removing the variable it reduces concurrency?

@junghans
Copy link
Contributor Author

If you run ctest -jN, it knows you have N cores and with the PROCESSORS property ctest can then figure out how many jobs to run in parallel.

@mkstoyanov
Copy link
Collaborator

I misread the PR thinking you are removing the PROCESSORS count.

If you want to skip parallel testing completely, try Heffte_SEQUENTIAL_TESTING we had to do that a while back due to spack setting env variable forcing parallel tests which in turn lead to running out of RAM on a 90+GB GPU.

@mkstoyanov mkstoyanov merged commit 192e67b into icl-utk-edu:master Oct 24, 2024
10 checks passed
@junghans junghans deleted the patch-2 branch October 24, 2024 16:34
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

Successfully merging this pull request may close these issues.

2 participants