James Foucar
James Foucar
| AppName | Area | Institution | Website | Status | Uses KokkosKernels | Contact Name | Contact Email | --- | --- | --- | --- | --- |...
Upon a second look, `nlev%pack_size !=0` is not necessary to demonstrate the error but it does make the errors more frequent. When I switched to `ExeSpaceUtils::parallel_reduce` on scalarized views, I...
I should also note that the error only occurs when team_size > 1, which is what you get when MIMIC_GPU is On (team size 7) which is On by default...
I believe the problems with `ExeSpaceUtils::parallel_reduce` were not fixed. @bartgol , correct me if I'm wrong.
@ambrad , sure. Here's a change I had to make in my dev branch: ``` --- a/components/scream/src/physics/shoc/atmosphere_macrophysics.hpp +++ b/components/scream/src/physics/shoc/atmosphere_macrophysics.hpp @@ -298,10 +298,12 @@ public: inv_qc_relvar(i,k) = 1; const auto condition...
@ambrad , I am about to make a PR; my branch is `jgfouca/fix_fpe_testing`
@ambrad , i think maybe i picked one that was more complicated than necessary. Most of the cases are very simple where you have a zero if check in fortran...
@bartgol , the other couple places where vector_disabled is used has a comment indicating that Intel18 has a problem. I'm pretty sure @ambrad put those there, so he may know...
1 seems easiest.
@sarats , just a heads up that I have not forgotten about this. There's a PR for CIME that moves most model-specific stuff, including provenance, out of CIME. Once we...