Floating point exception mpiexec.hydra "$@" 0<&0

[Iaquinta__Jean]

Hi everyone,

I struggle to make Intel® Parallel Studio XE 2019 to run this simple hello_mpi.f90:

program hello_mpi
   implicit none
   include 'mpif.h'
   integer  ::  rank, size, ierror, tag
   integer  ::  status(MPI_STATUS_SIZE)

   call MPI_INIT(ierror)
   call MPI_COMM_SIZE(MPI_COMM_WORLD, size, ierror)
   call MPI_COMM_RANK(MPI_COMM_WORLD, rank, ierror)
   print*, 'node', rank, ': Hello world'
   call MPI_FINALIZE(ierror)
end program hello_mpi

I tried first with Update3 and now Update4, and although the code compiles I get this error when trying to run it:

mpirun -np 4 ./hello_mpi
.../intel/compilers_and_libraries_2019.4.243/linux/mpi/intel64/bin/mpirun: line 103: 13574 Floating point exceptionmpiexec.hydra "$@" 0<&0

Here is what gdb --args mpiexec.hydra -n 2 hostname returns:

GNU gdb (GDB) Red Hat Enterprise Linux 7.6.1-114.el7
Copyright (C) 2013 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "x86_64-redhat-linux-gnu".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
Reading symbols from /opt/uio/intel/compilers_and_libraries_2019.4.243/linux/mpi/intel64/bin/mpiexec.hydra...done.
(gdb) run
Starting program: .../intel/compilers_and_libraries_2019.4.243/linux/mpi/intel64/bin/mpiexec.hydra -n 2 hostname
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib64/libthread_db.so.1".

Program received signal SIGFPE, Arithmetic exception.
IPL_MAX_CORE_per_package () at ../../../../../src/pm/i_hydra/../../intel/ipl/include/../src/ipl_processor.c:336
336     ../../../../../src/pm/i_hydra/../../intel/ipl/include/../src/ipl_processor.c: No such file or directory.
Missing separate debuginfos, use: debuginfo-install intel-mpi-rt-2019.4-243-2019.4-243.x86_64
 

Can someone help?

Thanks in advance,

Jean

[Anatoliy_R_Intel]

Hi Jean,

Could you run `lscpu` command to show CPU architecture? This issue should be fixed in Intel MPI 2019 Update 5 if you have AMD processor. 

--

Best regards, Anatoliy.

[Iaquinta__Jean]

Hi Anatoliy,

This is what the command returns:

lscpu
Architecture:          x86_64
CPU op-mode(s):        32-bit, 64-bit
Byte Order:            Little Endian
CPU(s):                32
On-line CPU(s) list:   0-31
Thread(s) per core:    1
Core(s) per socket:    1
Socket(s):             32
NUMA node(s):          1
Vendor ID:             AuthenticAMD
CPU family:            23
Model:                 1
Model name:            AMD EPYC Processor
Stepping:              2
CPU MHz:               1996.241
BogoMIPS:              3992.48
Hypervisor vendor:     KVM
Virtualization type:   full
L1d cache:             32K
L1i cache:             64K
L2 cache:              512K
L3 cache:              8192K
NUMA node0 CPU(s):     0-31
Flags:                 fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm art rep_good nopl xtopology extd_apicid eagerfpu pni pclmulqdq ssse3 fma cx16 sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw topoext retpoline_amd vmmcall fsgsbase bmi1 avx2 smep bmi2 rdseed adx smap clflushopt sha_ni xsaveopt xsavec xgetbv1 arat

When is the 5th update meant to be released?

​Regards,

Jean

[Anatoliy_R_Intel]

Unfortunately, I cannot say when it will be released. It should be soon. 

Now as a workaround you can try to set legacy hydra process manager:

PATH=${I_MPI_ROOT}/intel64/bin/legacy:$PATH mpiexec.hydra ...

And please try Update 5 when it is released.

--

Best regards, Anatoliy.

[Steve_B]

Interesting, just ran across the same issue using AMD EPYC 7551 processors.