-
Notifications
You must be signed in to change notification settings - Fork 364
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
homme SYCL changes #6594
homme SYCL changes #6594
Conversation
|
components/homme/src/theta-l_kokkos/cxx/HyperviscosityFunctorImpl.cpp
Outdated
Show resolved
Hide resolved
As I understand it, the int/bool code is just to work around a compiler bug affecting the F90-C++ interface; essentially, the compiler's treatment of c_bool is broken, or something like that. In particular, using bool in a kernel is not an issue; we do that in several spots. Oksana, is all of that correct? If so, I can add a commit to this branch next week that isolates the workaround to just the F90-C++ interface code. In particular, I'd like to remove the two theta_hydrostatic_mode variables from control_mod.F90. Is that OK? |
Yes, correct about int-bool issue.
yes, thanks! |
@bartgol thank for reducing the number of commits but you could do this one more time using --author on the commit command to give author credit to Oksana? |
Of course. I'll do it as soon as I get back to my computer |
df31c59
to
6c33665
Compare
* Disable timing in prim_main for the first two teps * Add some timers in EulerStepFunctor
6c33665
to
132b3a5
Compare
@rljacob @oksanaguba I edited the commits, to fix the author. @ambrad FYI, in case you pulled and was planning to do the mods you mentioned... |
I think I'll have a commit to push to this branch by tonight. I've made the changes and am running homme_integration on Chrysalis. |
Also fix preqx's use of use_moisture. Remove MOIST-DRY enum to avoid confusion, since it's no longer used.
Done. I also fixed some moisture-related code in preqx so that homme_integration passes on Chrysalis. |
79ebefa
to
f30afb2
Compare
homme suite pass on chrysalis |
Changes needed for spot/aurora: sycl kokkos spaces, temporary TP changes (TP init will be sorted out later), switch booleans to ints in F<-->xx interfaces. see #6569 for the original branch/history. [bfb]
in next |
hommexx ers_d test failed on cdash with an mpi error, but i ran it by hand with no issues. |
These are due to PR E3SM-Project#6594 and break the E3SM-repo EAMxx build. Once the SCREAM and E3SM repos are unified, we can back out these workarounds. The workarounds are isolated to components/homme, keeping the commit separation of components/eamxx (SCREAM repo) and components/homme (E3SM repo) clean.
…6645) Hommexx: Temporarily work around some EAMxx-Hommexx incompatibilities. These are due to PR #6594 and break the E3SM-repo EAMxx build. Once the SCREAM and E3SM repos are unified, we can back out these workarounds. The workarounds are isolated to components/homme, keeping the commit separation of components/eamxx (SCREAM repo) and components/homme (E3SM repo) clean. Fixes #6635. [BFB]
Hommexx: Temporarily work around some EAMxx-Hommexx incompatibilities. These are due to PR #6594 and break the E3SM-repo EAMxx build. Once the SCREAM and E3SM repos are unified, we can back out these workarounds. The workarounds are isolated to components/homme, keeping the commit separation of components/eamxx (SCREAM repo) and components/homme (E3SM repo) clean. Fixes #6635. [BFB]
Changes needed for spot/aurora: sycl kokkos spaces, temporary TP changes (TP init will be sorted out later), switch booleans to ints in F<-->xx interfaces.
see #6569 for the original branch/history.
[bfb]