On 30-Apr-2019 11:39, George Bosilca wrote:
Depending on the alignment of the different types there might be small
holes in the low-level headers we exchange between processes It should
not
be a concern for users.
valgrind should not stop on the first detected issue except
if --exit-on-first-error has been provided (the default value should be
no), so the SIGTERM might be generated for some other reason. What is
at jackhmmer.c:1597 ?
Ah, my bad, I didn't notice that the exit was a different line number
than the error message.
1597 is a coded exit on an error condition (triggered by the omission in
the test of
a parameter needed by the modified jackhmmer.) So you are right, it did
not in fact fail
at the first error.
This suppression file:
cat >/usr/common/tmp <<EOD
{
mpi_init_1
Memcheck:Addr4
fun:opal_os_dirpath_create
fun:orte_session_dir
fun:orte_ess_base_app_setup
fun:rte_init
fun:orte_init
fun:ompi_mpi_init
fun:PMPI_Init
fun:main
}
{
mpi_probe_1
Memcheck:Param
socketcall.sendto(msg)
fun:send
fun:mca_btl_tcp_send_blocking
fun:mca_btl_tcp_endpoint_complete_accept
fun:opal_libevent2022_event_base_loop
fun:opal_progress
fun:mca_pml_ob1_probe
fun:PMPI_Probe
fun:mpi_master
fun:main
}
{
mpi_send_1
Memcheck:Param
socketcall.sendto(msg)
fun:send
fun:mca_btl_tcp_send_blocking
fun:mca_btl_tcp_endpoint_send_handler
fun:event_persist_closure
fun:event_process_active_single_queue
fun:event_process_active
fun:opal_libevent2022_event_base_loop
fun:opal_progress
fun:ompi_request_wait_completion
fun:mca_pml_ob1_send
fun:PMPI_Send
fun:mpi_worker
fun:main
}
EOD
removes the mpi_send and some other messages when valgrind is run with
a second suppressions line
--suppressions=/usr/common/tmp/mpi_more_supp.txt \
Thanks,
David Mathog
mat...@caltech.edu
Manager, Sequence Analysis Facility, Biology Division, Caltech
_______________________________________________
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users