Hi Sam, I can confirm that if I set OMP_NUM_TRHEADS=1 the test passes.
Best, Gabriele On Thu, Nov 2, 2023 at 2:35 PM Cupp, Samuel D. <[email protected]> wrote: > Hi Gabriele, > I've been experimenting with this, and I've reproduced this issue on a > different machine. In my case, it runs with 1 OMP thread but has the > segfault with 2. Can you run that test parfile > (repos/GRHayLET/GRHayLHDX/test/Balsara0.par) without OMP? If this is the > source, it will help narrow down the problem. I can run the test with > multiple threads on my local machine though, so I'm not sure why OMP only > causes a segfault sometimes. > > Samuel Cupp > Postdoctoral Researcher > Department of Physics > University of Idaho > ------------------------------ > *From:* Gabriele Bozzola <[email protected]> > *Sent:* Sunday, October 29, 2023 6:16 PM > *To:* Cupp, Samuel D. <[email protected]> > *Cc:* Steven R. Brandt <[email protected]>; [email protected] < > [email protected]> > *Subject:* Re: [Users] meeting minutes for 2023-10-19 > > Hi Sam, > > I verified that the testsuite passes if I remove the GRHayL thorns. > Other than that, ET seems to work fine. > > Attached is the cfg file I used to compile ET on Anvil. > > Best, > Gabriele > > On Thu, Oct 26, 2023 at 12:45 PM Cupp, Samuel D. <[email protected]> > wrote: > > Hi Gabriele, > We discussed this in the call this morning, and there's a few things we > can try. First, it would be helpful if you created a ticket so we can track > progress on the issue. This is especially true since the testsuite > shouldn't hang if a test fails. The expected behavior would be for it to > continue with testing, but for some reason it didn't. Also, it might help > to remove GRHayLHDX from the thornlist, recompile, and see if that is the > only test failing. Knowing if this is specific to the thorn or a broader > issue would help diagnose the source of the problem. > > Samuel Cupp > Postdoctoral Researcher > Department of Physics > University of Idaho > ------------------------------ > *From:* Gabriele Bozzola <[email protected]> > *Sent:* Tuesday, October 24, 2023 1:27 PM > *To:* Cupp, Samuel D. <[email protected]> > *Cc:* Steven R. Brandt <[email protected]>; [email protected] < > [email protected]> > *Subject:* Re: [Users] meeting minutes for 2023-10-19 > > Hi Sam, > > This is just CPU. The .out for the entire testsuite was attached to my > first email. > > Best, > Gabriele > > > > On Tue, Oct 24, 2023 at 12:14 PM Cupp, Samuel D. <[email protected]> > wrote: > > Do you know if any other CarpetX tests fail? Also, is this using gpus or > cpus? > > Samuel Cupp > Postdoctoral Researcher > Department of Physics > University of Idaho > ------------------------------ > *From:* Gabriele Bozzola <[email protected]> > *Sent:* Tuesday, October 24, 2023 10:07 AM > *To:* Cupp, Samuel D. <[email protected]> > *Cc:* Steven R. Brandt <[email protected]>; [email protected] < > [email protected]> > *Subject:* Re: [Users] meeting minutes for 2023-10-19 > > Hi Sam, > > I pulled the latest version. Tests are not failing anymore, but looking at > the .out, it seems that the Balsara0 test fails and stalls execution of the > entire testsuite. > > I am attaching the .out for the test. > > Best, > Gabriele > > On Fri, Oct 20, 2023 at 2:32 PM Cupp, Samuel D. <[email protected]> > wrote: > > Hi Gabriele, > It looks like some of the failures are in GRHayLHD. I've been having > trouble getting consistent output with nproc>1. The data doesn't change, it > just gets rearranged in the datafile. I made changes to the tests this > week, so could you rerun the GRHayLHD tests and tell me if they fail? If > they do, I'll change the tests so that they only run for nproc=1. > > I don't know why it would take that long to run, however. Nothing in the > ET CI suggests it should take that long, best I can tell. Do you have an > idea of which test is taking so long? > > Samuel Cupp > Postdoctoral Researcher > Department of Physics > University of Idaho > ------------------------------ > *From:* Users <[email protected]> on behalf of Gabriele > Bozzola <[email protected]> > *Sent:* Thursday, October 19, 2023 5:46 PM > *To:* Steven R. Brandt <[email protected]> > *Cc:* [email protected] <[email protected]> > *Subject:* Re: [Users] meeting minutes for 2023-10-19 > > Hello, > > Is there a tested configuration for anvil? > > I compiled it one last weekend and ran the tests with the master branch. I > found a few failures in the > MPI runs, and the test suite does not complete even with a walltime of 6 > hours. > > I am attaching the output. > > Best, > Gabriele > > On Thu, Oct 19, 2023 at 7:36 AM Steven R. Brandt <[email protected]> > wrote: > > Present: Steve, Peter, Sam, Zach, Leo > > Release > Gallery Examples > - TOV is Peter > - BBH is Steve > - Roland might have students for the other three? > Tests > - Need to be run > - Clusters: > Stampede2 isn't listed, and Delta and Anvil are added. > - Release name: Lise Meitner > https://en.wikipedia.org/wiki/Lise_Meitner > - Autogenerated codes need to be regenerated, Leo will do it > > Mailing list moderation > - Remind Roland to send Peter the password > > Unanswered question on the mailing list > - Need answer for Enzo. Roland was going to answer? > - Ticket 2749: Leo says he fixed it. > - Ticket 2647: One of Zach's students is looking at that in Grail (sp) > - Ticket 2609: Steve thinks he approved PR for Roland > > _______________________________________________ > Users mailing list > [email protected] > http://lists.einsteintoolkit.org/mailman/listinfo/users > >
_______________________________________________ Users mailing list [email protected] http://lists.einsteintoolkit.org/mailman/listinfo/users
