<div dir="ltr"><div dir="ltr">On Wed, Nov 6, 2019 at 4:40 AM hg <<a href="mailto:hgbk2008@gmail.com">hgbk2008@gmail.com</a>> wrote:<br></div><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr"><div>Look into arch-linux2-cxx-opt/externalpackages/pastix_5.2.3/src/sopalin/src/sopalin_thread.c I saw something like:</div><div><br></div><div>#ifdef HAVE_OLD_SCHED_SETAFFINITY<br> if(sched_setaffinity(0,&mask) < 0)<br>#else /* HAVE_OLD_SCHED_SETAFFINITY */<br> if(sched_setaffinity(0,sizeof(mask),&mask) < 0)<br>#endif /* HAVE_OLD_SCHED_SETAFFINITY */<br> {<br> perror("sched_setaffinity");<br> EXIT(MOD_SOPALIN, INTERNAL_ERR);<br> }<br></div><div><br></div><div>Is there possibility that Petsc turn on HAVE_OLD_SCHED_SETAFFINITY during compilation?</div><div><br></div><div>May I know how to trigger re-compilation of external packages with petsc? I may go in there and check what's going on.</div></div></blockquote><div><br></div><div>If we built it during configure, then you can just go to</div><div><br></div><div><div> $PETSC_DIR/$PETSC_ARCH/externalpackages/*pastix*/</div><div><br></div><div>and rebuild/install it to test. If you want configure to do it, you have to delete</div><div><br></div><div></div></div><div> $PETSC_DIR/$PETSC_ARCH/lib/petsc/conf/pkg.conf.pastix</div><div><br></div><div>and reconfigure.</div><div><br></div><div> Thanks,</div><div><br></div><div> Matt</div><div> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr"><div><div dir="ltr"><div dir="ltr">Giang</div></div></div><br></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Wed, Nov 6, 2019 at 10:12 AM hg <<a href="mailto:hgbk2008@gmail.com" target="_blank">hgbk2008@gmail.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr"><span style="color:rgb(69,69,69);font-family:Lato,sans-serif;font-size:12.8px">sched_setaffinity: Invalid argument</span> only happens when I launch the job with sbatch. Running without scheduler is fine. I think this has something to do with pastix.<div><br clear="all"><div><div dir="ltr"><div dir="ltr">Giang</div></div></div><br></div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Wed, Nov 6, 2019 at 4:37 AM Smith, Barry F. <<a href="mailto:bsmith@mcs.anl.gov" target="_blank">bsmith@mcs.anl.gov</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><br>
Google finds this <a href="https://gforge.inria.fr/forum/forum.php?thread_id=32824&forum_id=599&group_id=186" rel="noreferrer" target="_blank">https://gforge.inria.fr/forum/forum.php?thread_id=32824&forum_id=599&group_id=186</a><br>
<br>
<br>
<br>
> On Nov 5, 2019, at 7:01 PM, Matthew Knepley via petsc-users <<a href="mailto:petsc-users@mcs.anl.gov" target="_blank">petsc-users@mcs.anl.gov</a>> wrote:<br>
> <br>
> I have no idea. That is a good question for the PasTix list.<br>
> <br>
> Thanks,<br>
> <br>
> Matt<br>
> <br>
> On Tue, Nov 5, 2019 at 5:32 PM hg <<a href="mailto:hgbk2008@gmail.com" target="_blank">hgbk2008@gmail.com</a>> wrote:<br>
> Should thread affinity be invoked? I set -mat_pastix_threadnbr 1 and also OMP_NUM_THREADS to 1<br>
> <br>
> Giang<br>
> <br>
> <br>
> On Tue, Nov 5, 2019 at 10:50 PM Matthew Knepley <<a href="mailto:knepley@gmail.com" target="_blank">knepley@gmail.com</a>> wrote:<br>
> On Tue, Nov 5, 2019 at 4:11 PM hg via petsc-users <<a href="mailto:petsc-users@mcs.anl.gov" target="_blank">petsc-users@mcs.anl.gov</a>> wrote:<br>
> Hello<br>
> <br>
> I got crashed when using Pastix as solver for KSP. The error message looks like:<br>
> <br>
> ....<br>
> NUMBER of BUBBLE 1<br>
> COEFMAX 1735566 CPFTMAX 0 BPFTMAX 0 NBFTMAX 0 ARFTMAX 0<br>
> ** End of Partition & Distribution phase **<br>
> Time to analyze 0.225 s<br>
> Number of nonzeros in factorized matrix 708784076<br>
> Fill-in 12.2337<br>
> Number of operations (LU) 2.80185e+12<br>
> Prediction Time to factorize (AMD 6180 MKL) 394 s<br>
> 0 : SolverMatrix size (without coefficients) 32.4 MB<br>
> 0 : Number of nonzeros (local block structure) 365309391<br>
> Numerical Factorization (LU) :<br>
> 0 : Internal CSC size 1.08 GB<br>
> Time to fill internal csc 6.66 s<br>
> --- Sopalin : Allocation de la structure globale ---<br>
> --- Fin Sopalin Init ---<br>
> --- Initialisation des tableaux globaux ---<br>
> sched_setaffinity: Invalid argument<br>
> [node083:165071] *** Process received signal ***<br>
> [node083:165071] Signal: Aborted (6)<br>
> [node083:165071] Signal code: (-6)<br>
> [node083:165071] [ 0] /lib64/libpthread.so.0(+0xf680)[0x2b8081845680]<br>
> [node083:165071] [ 1] /lib64/libc.so.6(gsignal+0x37)[0x2b8082191207]<br>
> [node083:165071] [ 2] /lib64/libc.so.6(abort+0x148)[0x2b80821928f8]<br>
> [node083:165071] [ 3] /sdhome/bui/opt/petsc-3.11.0_ompi-3.0.0/lib/libpetsc.so.3.11(sopalin_launch_comm+0x0)[0x2b80a4124c9d]<br>
> [node083:165071] [ 4] Launching 1 threads (1 commputation, 0 communication, 0 out-of-core)<br>
> --- Sopalin : Local structure allocation ---<br>
> /sdhome/bui/opt/petsc-3.11.0_ompi-3.0.0/lib/libpetsc.so.3.11(D_sopalin_init_smp+0x29b)[0x2b80a40c39d2]<br>
> [node083:165071] [ 5] /sdhome/bui/opt/petsc-3.11.0_ompi-3.0.0/lib/libpetsc.so.3.11(D_ge_sopalin_smp+0x68)[0x2b80a40cf4c2]<br>
> [node083:165071] [ 6] /sdhome/bui/opt/petsc-3.11.0_ompi-3.0.0/lib/libpetsc.so.3.11(sopalin_launch_thread+0x4ba)[0x2b80a4124a31]<br>
> [node083:165071] [ 7] /sdhome/bui/opt/petsc-3.11.0_ompi-3.0.0/lib/libpetsc.so.3.11(D_ge_sopalin_thread+0x94)[0x2b80a40d6170]<br>
> [node083:165071] [ 8] /sdhome/bui/opt/petsc-3.11.0_ompi-3.0.0/lib/libpetsc.so.3.11(D_pastix_task_sopalin+0x5ad)[0x2b80a40b09a2]<br>
> [node083:165071] [ 9] /sdhome/bui/opt/petsc-3.11.0_ompi-3.0.0/lib/libpetsc.so.3.11(d_pastix+0xa8a)[0x2b80a40b2325]<br>
> [node083:165071] [10] /sdhome/bui/opt/petsc-3.11.0_ompi-3.0.0/lib/libpetsc.so.3.11(+0x63927b)[0x2b80a35bf27b]<br>
> [node083:165071] [11] /sdhome/bui/opt/petsc-3.11.0_ompi-3.0.0/lib/libpetsc.so.3.11(MatLUFactorNumeric+0x19a)[0x2b80a32c7552]<br>
> [node083:165071] [12] /sdhome/bui/opt/petsc-3.11.0_ompi-3.0.0/lib/libpetsc.so.3.11(+0xa46c09)[0x2b80a39ccc09]<br>
> [node083:165071] [13] /sdhome/bui/opt/petsc-3.11.0_ompi-3.0.0/lib/libpetsc.so.3.11(PCSetUp+0x311)[0x2b80a3a8f1a9]<br>
> [node083:165071] [14] /sdhome/bui/opt/petsc-3.11.0_ompi-3.0.0/lib/libpetsc.so.3.11(KSPSetUp+0xbf7)[0x2b80a3b46e81]<br>
> [node083:165071] [15] /sdhome/bui/opt/petsc-3.11.0_ompi-3.0.0/lib/libpetsc.so.3.11(KSPSolve+0x210)[0x2b80a3b4746e]<br>
> <br>
> Does anyone have an idea what is the problem and how to fix it? The PETSc parameters I used are as below:<br>
> <br>
> It looks like PasTix is having trouble setting the thread affinity:<br>
> <br>
> sched_setaffinity: Invalid argument<br>
> <br>
> so it may be your build of PasTix.<br>
> <br>
> Thanks,<br>
> <br>
> Matt<br>
> <br>
> -pc_type lu<br>
> -pc_factor_mat_solver_package pastix<br>
> -mat_pastix_verbose 2<br>
> -mat_pastix_threadnbr 1<br>
> <br>
> Giang<br>
> <br>
> <br>
> <br>
> -- <br>
> What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead.<br>
> -- Norbert Wiener<br>
> <br>
> <a href="https://www.cse.buffalo.edu/~knepley/" rel="noreferrer" target="_blank">https://www.cse.buffalo.edu/~knepley/</a><br>
> <br>
> <br>
> -- <br>
> What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead.<br>
> -- Norbert Wiener<br>
> <br>
> <a href="https://www.cse.buffalo.edu/~knepley/" rel="noreferrer" target="_blank">https://www.cse.buffalo.edu/~knepley/</a><br>
<br>
</blockquote></div>
</blockquote></div>
</blockquote></div><br clear="all"><div><br></div>-- <br><div dir="ltr" class="gmail_signature"><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div>What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead.<br>-- Norbert Wiener</div><div><br></div><div><a href="http://www.cse.buffalo.edu/~knepley/" target="_blank">https://www.cse.buffalo.edu/~knepley/</a><br></div></div></div></div></div></div></div></div>