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

Cray XC case added #71

Open
wants to merge 2 commits into
base: master
Choose a base branch
from

Conversation

ChristopherBignamini
Copy link

Hello, I'm opening this (draft, for the moment) PR in order to add an additional Cray system case. Particularly, the modified code is a workaround to build MUSIC on PizDaint (Cray XC) for a CSCS user. I've used the same linux command of the existing Cray (XE6) case to identify the system in configure.ac, in order to avoid too many changes, but for sure is not the best solution: we could probably use the node interconnect technology to identify the system. If you think this PR is useful for MUSIC user community please let me know, I will try to finalize the system identification issue. Thank you!

@mdjurfeldt
Copy link
Contributor

Yes, thank you. This PR would be useful.

Please fix the system identification. I'll then merge it.

Best regards,
Mikael

@ChristopherBignamini
Copy link
Author

I've updated the system identification, by checking the hostname. The problem in using something more generic to identify the system model, like the interconnect technology which is usually model specific, is that different sites could have different software setups for MPI version, compiler wrappers, etc...

@ChristopherBignamini ChristopherBignamini marked this pull request as ready for review April 23, 2021 09:45
@mdjurfeldt
Copy link
Contributor

I'm reluctant to identify the host system type based on the hostname. Then a better option would be to enable the user to explicitly set system type as an argument to the configure script.

Meanwhile, I'll send an email to Cray and ask them if they are aware of ways to identify a Cray XC,

But: Should we do that at all? Looking at your changes, they seem to rely on the environment variable SLURM_PROCID, so perhaps what we really (or also) should do is to check for Slurm?

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