Thanks Jed, this is the kind of answer i really needed to hear(I hope it works), I was sure that something in the shadows of mpich2 didn't act acording to my expectations but honestly I didn't even knew where to start looking. I'll try this and then get back to you.
To Barry: Regarding the output from -log_summary for the debug and opt versions, I already sent them a few weeks back in a mail to 'petsc-users', your answer was that the time increase came from the factorization of the matrix. If it will be necessary in my next post I'll attache the two files, but first I want to try Jed's idea. Thanks, Bogdan > On Wed, Dec 7, 2011 at 06:31, Bogdan Dita <bogdan at lmn.pub.ro> wrote: > >> A few weeks ago i posted a question regarding a problem with the >> optimised version of PETSc, what i mean is that I obtained a higher >> solving time with the optimised version then with the debug >> version(~10-15% higher). >> > > What do you get from 'mpicc -show'? > > MPICH2 has a tendency to put optimization flags into the wrappers. It's > possible that the "optimized" build of PETSc is effectively using -O2 > (from > MPICH2) and -g3 (from PETSc with-debugging=1), which might be faster than > -O2 (from MPICH2) -O (from PETSc with-debugging=0, taking precedence) that > you get when you. You can use PETSc's --COPTFLAGS to set your own > optimization flags for PETSc. > > Let us know if this works. > If this doesn't work produce the results from running the same code >using the >-log_summary option with and without debug. Then send the two outputs to >petsc-maint at mcs.anl.gov this will tell us what parts of the code (all?) >are slower >and faster in the different versions. > > Barry
