Fixing gas washout test for Frontier.

[odiazib]

Fixing the 524 - validate_gas_washout_merged (Failed) and 492 - validate_chm_diags_ts_355 (Failed) tests in Frontier.

[codecov]

Codecov Report

Attention: Patch coverage is 61.53846% with 30 lines in your changes missing coverage. Please review.

Project coverage is 93.45%. Comparing base (60c6c3c) to head (9ae3393).
Report is 12 commits behind head on main.

Files with missing lines	Patch %	Lines
src/mam4xx/mo_chm_diags.hpp	59.45%	30 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #442      +/-   ##
==========================================
+ Coverage   93.43%   93.45%   +0.01%     
==========================================
  Files         303      303              
  Lines       25160    25180      +20     
  Branches     2786     2763      -23     
==========================================
+ Hits        23509    23531      +22     
+ Misses       1651     1649       -2     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[odiazib]

@singhbalwinder @jaelynlitz @mjschmidt271 , can we merge this PR? Most of these changes will not affect eamxx. Please see this PR: E3SM-Project/E3SM#7335.

[mjschmidt271]

Looks good to me!

[singhbalwinder]

nice work! thanks @odiazib !

[odiazib]

@mjschmidt271 @jeff-cohere
The AT is encountering this error in the CUDA build.

/home/runner/_work/mam4xx/mam4xx/haero_install/include/kokkos/Cuda/Kokkos_Cuda_Half_Conversion.hpp(415): error: identifier "__ll2bfloat16_rn" is undefined
  bhalf_t cast_to_bhalf(long long val) { return __ll2bfloat16_rn(val); }

Do you have any thoughts on this?

[jeff-cohere]

When I've encountered errors like this in the past, it has sometimes been due to a mis- or un-specified CUDA architecture or an incorrect version of the library. I'm not able to dig up anything more specific than that at the moment.

[mjschmidt271]

@mjschmidt271 @jeff-cohere The AT is encountering this error in the CUDA build.

/home/runner/_work/mam4xx/mam4xx/haero_install/include/kokkos/Cuda/Kokkos_Cuda_Half_Conversion.hpp(415): error: identifier "__ll2bfloat16_rn" is undefined
  bhalf_t cast_to_bhalf(long long val) { return __ll2bfloat16_rn(val); }

Do you have any thoughts on this?

I know I've urn into the same before--let me give it a look!

[odiazib]

One note: I also see this error in my other PR: #444. I believe this error is happening after the latest merge in haero.

[odiazib]

@singhbalwinder @jeff-cohere can we merge this PR?