[mpich-discuss] integrating mpich into MS Visual Studio fordebugging FORTRAN

[Jayesh Krishna]

Hi,
 You can use debugging level 0 when running smpd (smpd -d 0) in debug mode
to avoid smpd dumping the messages to stdout.
 
Regards,
Jayesh

  _____  

From: mpich-discuss-bounces at mcs.anl.gov
[mailto:mpich-discuss-bounces at mcs.anl.gov] On Behalf Of michael
Sent: Wednesday, July 08, 2009 8:38 AM
To: mpich-discuss at mcs.anl.gov
Subject: [mpich-discuss] integrating mpich into MS Visual Studio
fordebugging FORTRAN


With thanks to Jayesh, and tips I've garnered from elsewhere (and learning
by mistake!), I've pulled together a short example on using MS Visual
Studio for debugging MPI under MS Windows and would welcome any comments
re mistakes or omissions:

http://wiki.rcs.manchester.ac.uk/community/MPI/VisualStudio_mpich2_howto

Many thanks, Michael, Research Computing Services, University of
Manchester, United Kingdom

 > Comment (by Jayesh Krishna):
 >
 > {{{
 >
 > Hi,
 >  Take a look at the section 9.8 in the window's developer's guide
 >
 
(http://www.mcs.anl.gov/research/projects/mpich2/documentation/index.php?s
 > =docs) on debugging MPI programs.

 Had seen above but didn't seem to meet the requirements of doing it
 within the V.S. IDE... however:

 >
 >
 > # You can also attach to MPI programs that are running using the VS
 > debugger ("Attach to process").

 ah, didn't try the above... but also:

 >
 > # If all the MPI processes are launched on the same machine/node
 > then you
 > can start smpd in debug mode (smpd -stop; smpd -d;) and use multiple
 > instances of VS to start your MPI program (mpiexec -n 2 devenv.exe
 > fpi.exe
 > ==> will start two VS windows and you can step individually through
 > them).

 yes! this seems similar to what I was looking for (but prefer to be
 able to kick off from the VS IDE itself!)

 >
 > (PS: The debug versions of MPICH2 libraries, eg: fmpich2d.lib, is only
 > required if you are debugging the MPICH2 library.)

 good point and many thanks for pointing that out ;)

 M

 >
 > Regards,
 > Jayesh 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/mpich-discuss/attachments/20090708/9b555119/attachment.htm>