-
Notifications
You must be signed in to change notification settings - Fork 112
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
Allow CubeGenerator to use multiple MPI processes #1105
Conversation
Hi, thanks for the PR! Would it be possible to upgrade to the current master branch? There may still be some file conflicts/file moves there. |
I merged the upstream changes into the current master branch. After fixing the merge-related issues, I ran into the following error when I tried to compile SeisSol:
I fixed it locally by adding the |
Hi, I'm gonna embed the namespace fixes also into #1112—if that's good for you. (it appeared now in other contexts as well, cf. #1113 ; I'll probably just namespace the main SeisSol class) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Hi, thank you for your PR.
I've added some minor comments regarding the code.
Fixing boundary condition 6 may be a bit more involved in the end—so it's not required for this PR.
Thank you for the suggestions, let me know if I missed something! |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Ok, thanks for the changes! I've mirrored the branch to run it through the tests (we currently need to do that for external branches). And it passed.
Extended the
CubeGenerator
class to allow for using more than 1 MPI process.Boundary condition 7 works as expected, however using boundary condition 6 leads to exponential growth in the elastic energy output when using more than 1 MPI process. This also happens with the
NetcdfReader
and meshes generated with thecubeGenerator
located inpreprocessing/meshing/cube_c/
. I added a warning to theCubeGenerator
class which warns the user of this issue when using at least 2 MPI processes and boundary condition 6.