CPP directive around use of ISNAN() intrinsic function

[davegill]

TYPE: enhancement

KEYWORDS: ISNAN, PGI compiler

SOURCE: Found by Prasanth, fixed by internal

DESCRIPTION OF CHANGES:
Problem:
Some PGI compilers do not support the 2003 intrinsic ISNAN() function.

Solution:
Put in an automatic test to include a CPP -Dflag if the ISNAN() function is not part of the compiler.

LIST OF MODIFIED FILES:
modified: configure
modified: phys/module_sf_noahmplsm.F
new file: tools/fortran_2003_isnan_test.F

TESTS CONDUCTED:

1. Going to see if this addresses the user's problem
2. Jenkins testing is OK (expected, as no real changes to the compiled code, except for a no opt CPP directive).

[davegill]

jenkins

Please find result of the WRF regression test cases in the attachment. This build is for Commit ID: b973eb63b9427d24e0c714c65d519981206501f0, requested by: davegill for PR: https://github.com/wrf-model/WRF/pull/1460. For any query please send e-mail to David Gill.

    Test Type              | Expected  | Received |  Failed
    = = = = = = = = = = = = = = = = = = = = = = = =  = = = =
    Number of Tests        : 19           18
    Number of Builds       : 48           46
    Number of Simulations  : 163           161        0
    Number of Comparisons  : 103           102        0

    Failed Simulations are: 
    None
    Which comparisons are not bit-for-bit: 
    None

[davegill]

@prasanthvkrishna
Would you test this PR out to see if the WRF code compiles with your PRG compiler?

[prasanthvkrishna]

@prasanthvkrishna
Would you test this PR out to see if the WRF code compiles with your PRG compiler?

@davegill Sure, Dave. I can do a test.

[prasanthvkrishna]

@prasanthvkrishna
Would you test this PR out to see if the WRF code compiles with your PRG compiler?

@davegill Could you please add a line after your ISNAN check?
"IF ( (IRRLOSS .GT. 100.0) .OR. (IRRLOSS .LT. 0.0) ) IRRLOSS=4.0 ! In case if IRRLOSS is out of range"
This is required as @Cenlin_He suggested for out-of-range values.

[cenlinhe]

@prasanthvkrishna
Would you test this PR out to see if the WRF code compiles with your PRG compiler?

@davegill Could you please add a line after your ISNAN check?
"IF ( (IRRLOSS .GT. 100.0) .OR. (IRRLOSS .LT. 0.0) ) IRRLOSS=4.0 ! In case if IRRLOSS is out of range"
This is required as @Cenlin_He suggested for out-of-range values.

@prasanthvkrishna This range limit code can be added in your PR instead of Dave's PR here. I think it is better to have Dave' PR focusing on resolving the ISNAN issue.

[davegill]

#1456 "Noah-MP Dynamic Irrigation Packaged 2d fields" handles the intrinsic ISNAN() with a method already in place in the WRF model (P3 and TEMF).

We can address all of these at one time, LATER.