Fix knint

[aliabdolali]

Pull Request Summary

replace KNINT with NINT to be compatible with all compiler options.

Description

The knint in model/src/ww3_grib.F90 routine is an Intel extension and does not work with GNU compilers.
Replacing it with nint makes it generic, compatible with GNU and INTEL.
Is a change of answers expected from this PR? NO

Issue(s) addressed

• Is there an issue associated with this development (bug fix, enhancement, new feature)?

• fixes KNINT is an Intel extension and not portable #531
• fixes Regtest tp2.17 failing to compile since #350 #443

Check list

• Is your feature branch up to date with the authoritative repository (NOAA/develop)? Yes

• Make sure you have checked the checklist for a developer submitting to develop and updating version number

• Please list appropriate labels code managers should add for this PR:
  bug

• Reviewers: @kgerheiser

Commit Message

• Portability fix for all compiler options in ww3_grib (KNINT > NINT)

Testing

• How were these changes tested? regression test
• Are the changes covered by regression tests? yes
• If a new feature was added, was a new regression test added? No
• Have regression tests been run? Yes
• Which compiler / HPC you used to run the regression tests in the PR? Intel/Hera
• Please provide the summary output of matrix.comp (matrix.Diff.txt, matrixCompFull.txt and matrixCompSummary.txt):

**********************************************************************
********************* non-identical cases ****************************
**********************************************************************
mww3_test_03/./work_PR2_UQ_MPI_d2                     (8 files differ)
mww3_test_03/./work_PR2_UNO_MPI_d2                     (6 files differ)
mww3_test_03/./work_PR3_UQ_MPI_d2_c                     (8 files differ)
mww3_test_03/./work_PR1_MPI_d2                     (7 files differ)
mww3_test_03/./work_PR3_UNO_MPI_d2_c                     (6 files differ)
mww3_test_03/./work_PR3_UNO_MPI_d2                     (6 files differ)
mww3_test_03/./work_PR3_UQ_MPI_d2                     (8 files differ)
mww3_test_07/./work_PR3_UQ                     (3 files differ)
ww3_tp2.10/./work_MPI_OMPH                     (7 files differ)
ww3_tp2.16/./work_MPI_OMPH                     (5 files differ)
ww3_ufs1.3/./work_a                     (2 files differ)

matrixCompFull.txt
matrixCompSummary.txt
matrixDiff.txt

[aliabdolali]

Hi @kgerheiser
Have you been able to comple ww3_grib with gnu with the fix? Could you tell me which machine you did it?

[kgerheiser]

Yep, I was able to compile with this change using GCC 9.4.0 on my local macOS machine.

[kgerheiser]

I don't really know how WW3 tests work, but is your test log showing differences in output? I wouldn't expect this change to affect results.

[aliabdolali]

Hi @ukmo-ccbunney
could you check ./bin/run_test -c ukmo_cray_gnu -S -s MPI -s NO_PDLIB -w work_a -g a -f -p mpiexec -n 16 -o all ../model ww3_tp2.17 with GNU compiler to make sure it fixes the problem?

[ukmo-ccbunney]

Hi @ukmo-ccbunney could you check ./bin/run_test -c ukmo_cray_gnu -S -s MPI -s NO_PDLIB -w work_a -g a -f -p mpiexec -n 16 -o all ../model ww3_tp2.17 with GNU compiler to make sure it fixes the problem?

Hi @aliabdolali - so, this now compiles for me (the KNINT -> NINT fix works for me), but it fails to link.
This is because this test now enables the NCEP2 switch which hardcodes the following compiler options in the link file

opt="$opt -convert big_endian -assume byterecl -prec-div -prec-sqrt -ip"

I believe these options are Intel IFORT specific, so I needed to modify these for the GNU compiler. Additionally, I need to compile the relevant grib/bacio/g2lib libraries to get this regtest to actually compile (link) for me (we don't routinely use grib with the wave model).

However, I am going to approve this PR as the fix for KNINT looks to be working OK.

[ukmo-ccbunney]

KNINT fix worked for me.
I am struggling to test the ww3_tp2.17 regtest at the moment though due to other issues described in this PR's comments.