Intel® Fortran Compiler
Build applications that can scale for the future with optimized code designed for Intel® Xeon® and compatible processors.
Announcements
FPGA community forums and blogs on community.intel.com are migrating to the new Altera Community and are read-only. For urgent support needs during this transition, please visit the FPGA Design Resources page or contact an Altera Authorized Distributor.
29285 Discussions

MPI: reorder myrank in MPI_CART_CREATE

antony_muthu__Anto_J
Beginner
‎05-05-2010 09:01 AM
1,079 Views
Hi,

I have a question regarding the reordering of the processors rank when using ifort. I had a code that compiled and executed with no problems in GNU compilers. I tried the same code using ifort and it compiled without problem but is crashed during execution. I've found the problem was that just before the instruction MPI_CART_CREATE I had "reorder=.true." Replacing it by "reorder=.false." the code is executed. It also works (compile and execute) in GNU compilers.

Can anyone explain me why it fails in ifort with "reorder=.true." but works with "reorder=.false."???

Does anyone knows a way to make it work with "reorder=.true."???

Thanks a lot,

JoSo
0 Kudos

Link Copied

4 Replies
Ron_Green
Moderator
‎05-05-2010 01:10 PM
1,079 Views
Did you build your MPI library with Intel ifort? the representation of .true. and .false. may vary by compiler. If you built MPI with gfortran it may not work when linked with an app built with ifort (and vice versa).
0 Kudos
Copy link
antony_muthu__Anto_J
Beginner
‎05-06-2010 01:08 AM
1,079 Views
Did you build your MPI library with Intel ifort? the representation of .true. and .false. may vary by compiler. If you built MPI with gfortran it may not work when linked with an app built with ifort (and vice versa).

Hi Ronald,

Thanks for your reply. I buiklt the MPI library using a GNU compiler insted of ifort. I am also getting a segmentation fault when using ifort that I'm not getting using GNU compilers. I'll try to build MPI library using ifort.

Thanks again,

JoSo
0 Kudos
Copy link
Ron_Green
Moderator
‎05-06-2010 06:58 AM
1,079 Views
Rebuilding the MPI library with ifort should fix that issue. Now for the other seg fault, there is an article linked at the top of the main page on this forum. Read through that and try the -heap-arrays compiler option. That fixes 90% of those issues.

ron
0 Kudos
Copy link
antony_muthu__Anto_J
Beginner
‎05-06-2010 10:45 AM
1,079 Views
I have rebuilt mpi with ifort and all previous problems are solved. I have a new one... which is that the results using ifort are completelly different than those obtained using GNU fortran. I'll have to find the appropriate options for ifort... wish me luck!

Thanks.

JoSo
0 Kudos
Copy link
Reply

Community support is provided Monday to Friday. Other contact methods are available here.

Intel does not verify all solutions, including but not limited to any file transfers that may appear in this community. Accordingly, Intel disclaims all express and implied warranties, including without limitation, the implied warranties of merchantability, fitness for a particular purpose, and non-infringement, as well as any warranty arising from course of performance, course of dealing, or usage in trade.

For more complete information about compiler optimizations, see our Optimization Notice.