remove ESMF_GridCompGetInternalState from use statement, not working… #449
Conversation
…with nag and apparently not needed
| @@ -1601,10 +1601,8 @@ subroutine set_aoflux_in_pointers(fldbun_a, fldbun_o, aoflux_in, lsize, xgrid, r | |||
| if (chkerr(rc,__LINE__,u_FILE_u)) return | |||
| call fldbun_getfldptr(fldbun_a, 'Sa_shum', aoflux_in%shum, xgrid=xgrid, rc=rc) | |||
| if (chkerr(rc,__LINE__,u_FILE_u)) return | |||
| if (add_gusts) then | |||
There was a problem hiding this comment.
@ekluzek I wish to point out this change required for nag because it doesn't like the unallocated pointer in the subroutine interface even though the pointer is never used. This is a case where the nag compiler is requiring changes we don't otherwise need to make.
There was a problem hiding this comment.
I see arguments both ways here, but I lean slightly towards saying that this is an argument for keeping nag: I believe this is enforcing an aspect of the Fortran standard, so by fixing this we (1) make it more likely that our code won't be rejected by various other compilers that we don't currently test with, and (2) avoid the potential for undefined behavior in compilers that accept the code.
There was a problem hiding this comment.
My main argument for nag itself is that it's found legit subscript overflow problems that for whatever reasons other compilers haven't caught. If we can use different options to catch those problems I'm happy to move away from nag.
There was a problem hiding this comment.
In terms of the code here, I see Jim's point that it's annoying here. But, if Bill is right that it's enforcing the standard I do side towards that. But, it also might be something that other compilers could find by using stricter options to enforce the standard?
|
I ran and passed ERC_D_Ln9.T5_T5_mg37.QPC4.izumi_nag.cam-outfrq3s_usecase, due to the urgency of the request this code change was not reviewed prior to merging. |
There was a problem hiding this comment.
Approving after the fact... thank you very much @jedwards4b !
… with nag and apparently not needed
Description of changes
Remove the use ESMF_GridCompGetInternalState it is not needed and fails for the nag compiler.
Specific notes
Contributors other than yourself, if any:
CMEPS Issues Fixed (include github issue #):
Are changes expected to change answers? (specify if bfb, different at roundoff, more substantial)
Any User Interface Changes (namelist or namelist defaults changes)?
Testing performed
Please describe the tests along with the target model and machine(s)
If possible, please also added hashes that were used in the testing