[Wrf-users] [Re:]So close! Compile Error using gcc and gfortran

[jcrosby at scotiaweather.com]

[quote]

I appear to be so close to a successful compilation of WRF NMM.  But 
there also appears to be one more error.  (Let's hope it's the last).

   I'm attempting a compile an installation of WRF NMM on a dual 
processor Linux box running Centos 5.1, Kernel 2.6.19xxx.  There are two 
Opteron 248 processors and 2 Gb installed RAM.  I'm using gcc and 
gfortran.  The version of WRF is 3.0.1.1.  A significant error in the 
compiler logs occurs at lines 1455:

1455   In file module_NONHY_DYNAM.f90: 333
1456
1457   !$OMP& private(b1_ik,b2_ik,b3_ik,c0_ik,chi_ik,coff_ik,cwm_ik,   
     &
1458                                                                     
                                    1
1459   Syntax error in OpenMP variable list at (1)

   From references in the OpenMP manuals and users e-mails, it's clear 
the form of the variable statement is causing an error.  Due to my own 
inexperience, however, I don't know which file(s) need to be changed to 
eliminate this syntax error (and go on to others!)

   I'm attempting to compile the SMpar version of WRF (with 0 nesting).

   Any help or suggestions would be much appreciated.

Michael Allen
Industrial Weather
Plymouth, MN

[/quote]

I ran into this problem as well a few months ago.  I fixed it by editing 
the source file dyn_nmm/module_NONHY_DYNAM.F and on lines 319 to 323 put 
a space between ! and $omp&
That got past the error and generated the exes for me, but I imagine it 
may cause something, somewhere to not work correctly, but I haven't 
encountered anything yet.

It seemed that was the only place it complained about !$omp being an issue.

Hope that helps.

-- 
Joseph Crosby
Network Administrator
Scotia Weather Service Inc.
netadmin at scotiaweather.com
-----------
A byte walks into a bar and orders a pint. Bartender asks him "What's wrong?" Byte says "Parity error." Bartender nods and says "Yeah, I thought you looked a bit off."