Error stack: OFI endpoint open failed (ofi_init.c:982:create_vni_context:Cannot allocate memory)

[sagitter]

Hi all again.

I don't know what is happening with MPICH-4.1.2, Sundials's tests are always failing in x86_64 architectures like reported here.
(And i don't even know if these others errors are related to that reported here)

      Start 80: test_sunlinsol_spgmr_parallel_100_2_1_50_1e-3_0
80/98 Test #80: test_sunlinsol_spgmr_parallel_100_2_1_50_1e-3_0 .....***Failed    0.50 sec
03c9b3c8ba754b1c9335e6cc38cd5087:rank3.test_sunlinsol_spgmr_parallel: Failed to get eth0 (unit 0) cpu set
03c9b3c8ba754b1c9335e6cc38cd5087:rank3.test_sunlinsol_spgmr_parallel: Failed to get eth1 (unit 1) cpu set
03c9b3c8ba754b1c9335e6cc38cd5087:rank3: PSM3 can't open nic unit: -1 (err=23)
Abort(606197647): Fatal error in internal_Init: Other MPI error, error stack:
internal_Init(66)........: MPI_Init(argc=0x7ffed119a5ec, argv=0x7ffed119a5e0) failed
MPII_Init_thread(234)....: 
MPID_Init(513)...........: 
MPIDI_OFI_init_local(604): 
create_vni_context(982)..: OFI endpoint open failed (ofi_init.c:982:create_vni_context:Cannot allocate memory)

===================================================================================
=   BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
=   PID 99975 RUNNING AT 03c9b3c8ba754b1c9335e6cc38cd5087
=   EXIT CODE: 9
=   CLEANING UP REMAINING PROCESSES
=   YOU CAN IGNORE THE BELOW CLEANUP MESSAGES
===================================================================================
YOUR APPLICATION TERMINATED WITH THE EXIT STRING: Killed (signal 9)
This typically refers to a problem with your application.
Please see the FAQ page for debugging suggestions

Full build log attached
sundials_OFI_endpoint_errors.gz

[balos1]

I don't think this is related to #312 since in the test_sunlinsol_spgmr_parallel example we do call MPI_Init at the beginning.

[balos1]

This seems like an error in the MPI installation. Can you verify that it works for something other than SUNDIALS?

[sagitter]

This seems like an error in the MPI installation. Can you verify that it works for something other than SUNDIALS?

Other libraries like hypre or MUMPS are tested with MPICH-4.1.2
superlu_dist fails

[balos1]

Since we never could reproduce this, I am closing it. Please reopen it if its still a problem with the current sundials release.