Re: [petsc-users] Error running make check with OneAPI C/C++ and Fortran compilers on Windows

2023-11-17 Thread Barry Smith
OneAPI has two sets of compilers, old icc and new icx. Does it fail in this way for both? Can you try with the old? > On Nov 17, 2023, at 12:59 PM, Maruthi NH wrote: > > Hi Barry, > > It doesn't even start to run, I still get the same error. This is all the > output I get. > >

Re: [petsc-users] VecNorm causes program to hang

2023-11-17 Thread Sreeram R Venkat
Thank you; that fixed the problem. I added an else { PetscCall(VecCUDAReplaceArray(v, NULL)); } Thanks, Sreeram On Fri, Nov 17, 2023 at 12:09 PM Barry Smith wrote: > >So the "bug" is not as ginormous as I originally thought. It will never > produce incorrect results but can result

Re: [petsc-users] Error running make check with OneAPI C/C++ and Fortran compilers on Windows

2023-11-17 Thread Maruthi NH
Hi Barry, It doesn't even start to run, I still get the same error. This is all the output I get. === = BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES = RANK 0 PID 28192 RUNNING AT BLRLAP1521 = EXIT STATUS

Re: [petsc-users] Error running make check with OneAPI C/C++ and Fortran compilers on Windows

2023-11-17 Thread Barry Smith
Please run mpiexec -n 1 ./ex19 -da_refine 3 -pc_type mg -ksp_type fgmres -info and send all the output > On Nov 17, 2023, at 12:14 PM, Maruthi NH wrote: > > Hi Barry, > > I get the following error. > > ngh@ngh1 ~/petsc/src/snes/tutorials > $ make ex19 > /home/ngh/petsc/lib/petsc/bin/win

Re: [petsc-users] Error running make check with OneAPI C/C++ and Fortran compilers on Windows

2023-11-17 Thread Maruthi NH
Hi Barry, I get the following error. ngh@ngh1 ~/petsc/src/snes/tutorials $ make ex19 /home/ngh/petsc/lib/petsc/bin/win32fe/win32fe icl -Qwd10161 -Qstd=c99 -MT -O3 -I/home/ngh/petsc/include -I/home/ngh/petsc/intel-petsc-tag-v3.20.1/include -I/cygdrive/c/PROGRA~2/Intel/oneAPI/mpi/2021.6.0/include

Re: [petsc-users] VecNorm causes program to hang

2023-11-17 Thread Barry Smith
So the "bug" is not as ginormous as I originally thought. It will never produce incorrect results but can result in the errors you received. The problem is if (row_rank == 0) { PetscCall(VecCUDAReplaceArray(v, d_a)); } The place/replacearray routines are actually collect

Re: [petsc-users] VecNorm causes program to hang

2023-11-17 Thread Sreeram R Venkat
I've updated the buildme script to specify the MPI and CUDA compilers. Please make sure those modules are loaded, and let me know if it works. Thanks, Sreeram On Fri, Nov 17, 2023 at 9:32 AM Mark Adams wrote: > I get this error: > > (base) 06:30 2 login10 master= perlmutter:~/petsc-test$ bash -

Re: [petsc-users] Error running make check with OneAPI C/C++ and Fortran compilers on Windows

2023-11-17 Thread Barry Smith
Please do cd src/snes/tutorials make ex19 mpiexec -n 1 ./ex19 -da_refine 3 -pc_type mg -ksp_type fgmres then with -n 2 Cut and paste all the output and include it in your email response Barry > On Nov 17, 2023, at 12:57 AM, Maruthi NH wrote: > > Hi all, > > I could succes

Re: [petsc-users] VecNorm causes program to hang

2023-11-17 Thread Mark Adams
I get this error: (base) 06:30 2 login10 master= perlmutter:~/petsc-test$ bash -x buildme.sh + '[' -z '' ']' + case "$-" in + __lmod_vx=x + '[' -n x ']' + set +x Shell debugging temporarily silenced: export LMOD_SH_DBG_ON=1 for this output (/opt/cray/pe/lmod/lmod/init/bash) Shell debugging restart