TSMP-Intel on Stages/2024 #252
Closed
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Contributor
|
Excellent! Have runtime tests been successful in juwels and jureca? |
Contributor
Author
|
I was a little too enthusiastic yesterday evening after the module environment was easily transferable and the environment loaded - sadly all my test-compilations on JUWELS ( CLM-PFL-PDAFOASIS error, from CLM-PDAFCLM5-PDAF |
jjokella
commented
Dec 6, 2024
| #load the default values | ||
| platform=$def_platform | ||
| if [[ $platform == "" ]] then ; platform="JUWELS" ; fi #We need a hard default here | ||
| if [[ $platform == "" ]] then ; platform="JEDI" ; fi #We need a hard default here |
Contributor
Author
There was a problem hiding this comment.
I think JEDI should not be default in a PR to master. As I understand, it is a testing platform. I think we should leave JUWELS until JUPITER may become a default.
| compileClm(){ | ||
| route "${cyellow}> c_compileClm${cnormal}" | ||
| comment " source clm interface script" | ||
| if echo "$compiler" | grep -qE 'Gnu'; then |
Contributor
Author
There was a problem hiding this comment.
Is this needed here? Should be moved to common_build_interface like for OASIS below, I guess.
Contributor
Author
There was a problem hiding this comment.
Should this really be deleted? Or could it be a copy of loadenvs.Gnu.2023 or loadenvs.Gnu.2024 depending on which is currently the default. This is how I understood loadenvs.Gnu and loadenvs.Intel so far: Copy of the default stage. Also there are some codeparts, where loadenvs.$compiler is used and this may cause problems, when loadenvs.Gnu and loadenvs.Intel are gone.
Add JEDI as supported platform
GCC with a version =>12, rmm needs a patch. Muhammad repo includes this patch already on the rmm. -DCMAKE_EXE_LINKER_FLAGS=\"-lcurand -lcusparse -lcublas\ are not read correctly with "ksh" shell.
I added to the cmake the flags of -DCMAKE_EXE_LINKER_FLAGS that weren't read in file build_interface_parflow.ksh.
Update dev-stages-2024 for GPU-rmm on Juwels and Jedi
Contributor
Author
|
further build discussion moved to https://github.com/HPSCTerrSys/TSMP2 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
loadenvs.Intel: module environment forStages/2024