You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Using ctsm5.4.060, when a build fails (e.g. due to a syntax error), build_status is set to 1 in env_build. Trying to rebuild after correcting build error is then blocked with the following message:
BUILD_STATUS' has changed from 0 to 1
ERROR: For your changes to take effect, run:
./case.build --clean-all
./case.build
This is undesirable because it forces a complete clean and rebuild, instead of just rebuilding necessary files. A work-around is to manually set build_status to 0 with ./xmlchange BUILD_STATUS=0; then ./case.build can proceed.
General bug information
CTSM version you are using: ctsm5.4.060
Does this bug cause significantly incorrect results in the model's science? No
Configurations affected: anytime the build fails and you have to rerun, because there was a problem in one of the Fortran source files
Details of bug
In the chat discussion, Erik suggests that likely it's the changes that came in with cime6.5.4 that was brought in with ctsm5.4.054. Erik, replicated this in the latest dev version. It only happens when the build fails. But, that's the time that you want it to just rebuild the few files that failed.
@ekluzek rather than marking next, I assigned this to you since you offered to open a corresponding CIME issue. Please link the CIME issue to this one for easy cross referencing.
@swensosc we have a fix in CIME for this now. So we just need to update to the latest cime to get the fix in for CTSM. They didn't add a test that verifies that it'll work going forward, which is unfortunate. But, at least there is a version with a fix in place.
Activity
slevis-lmwg commentedon Oct 7, 2026
@ekluzek rather than marking next, I assigned this to you since you offered to open a corresponding CIME issue. Please link the CIME issue to this one for easy cross referencing.
ekluzek commentedon Oct 7, 2026
CIME issue is here:
ESMCI/cime#5077
ekluzek commentedon Oct 16, 2026
@swensosc we have a fix in CIME for this now. So we just need to update to the latest cime to get the fix in for CTSM. They didn't add a test that verifies that it'll work going forward, which is unfortunate. But, at least there is a version with a fix in place.
wwieder commentedon Oct 16, 2026
Will this automatically close when we update CIME, or should we close it now?
ekluzek commentedon Oct 18, 2026
It won't auto close. But we should leave it open until we have a PR with an updated cime. In that we'll note that it fixes this issue.