[Wrf-users] wrf 3.1 runtime problem, smpar build on OS X [SEC=UNCLASSIFIED]

John Michalakes john at michalakes.us
Thu Jun 4 16:53:09 MDT 2009


Thanks a bunch to Roman Dubtsov at Intel who's provided tremendous
assistance with this.  I think we have a workaround for the problem with
seg-faults, illegal instructions, & general misbehavior of WRFV3.1 running
OpenMP (either smpar or dm+sm) on the Intel Macs, at least for the latest
version 11 of the Intel Fortran compiler.  I'm putting a couple of files up
on the web at:

 

   http://www.mmm.ucar.edu/wrf/WG2/ifort_mac

 

along with a README.  If you've been having problems (including those
described in this discussion http://forum.wrfforum.com/viewtopic.php?f=6
<http://forum.wrfforum.com/viewtopic.php?f=6&t=609> &t=609 on the WRF
Forum), maybe give this a try.  Note the compiler times are still
surprisingly long for a few routines (module_comm_dm.F and the four nl_get
and nl_set_*_routines.F files).  I don't have a fix for that yet.

 

John

 

-----------------------------------------------------------
John Michalakes, Software Engineer        michalak at ucar.edu
NCAR, MMM Division                   voice: +1 303 497 8199
3450 Mitchell Lane                     fax: +1 303 497 8181
Boulder, Colorado 80301 U.S.A.        cell: +1 720 209 2320
         <http://www.mmm.ucar.edu/individual/michalakes>
http://www.mmm.ucar.edu/individual/michalakes
----------------------------------------------------------- 

 

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/wrf-users/attachments/20090604/e606f945/attachment.html 


More information about the Wrf-users mailing list