Can you give me the following:
(1) Does "mpiexec -n 2 ls" work ? any output ?
(2) What hardware do you have ? in particular, any GPU cards, etc ?
(3) What does "ompi_info" report ?
(4) re-run your test program, with the env. variable PMIX_DEBUG=100
(pmix is a new component of openmpi that starts subprocesses; if this
breaks, then it would lead to mpi_init hanging).
thanks
Alastair
On 28/07/2018 22:18, Giacomo Mulas wrote:
(1) How did you execute it ? mpiexec command line , etc?
any way of a number, does not seem to make any difference. Always on
the command line. I tried mpiexec, mpirun, just running the executable
without wrappers (used to work before, if one wanted to run one single
process). I also tried using a different number of processes, from 1
to 8,
no difference. In any case, if I run the code under gdb (or a graphical
wrapper to gdb, like ddd), I can see that the code hangs when trying to
execute the MPI_Init() call. Or, if I try some of the scalapack example
codes, the first call that implicitly tries to initialise MPI.
(2) Can you give the output of :
$ dpkg -l | grep pmix
gmulas@capitanata:~/transfer/debsrc$ dpkg -l | grep pmix
ii libpmix2:amd64 3.0.0-1 amd64 Process Management Interface
(Exascale) library
$ dpkg -l | grep psm
gmulas@capitanata:~/transfer/debsrc$ dpkg -l | grep psm
ii gir1.2-osmgpsmap-1.0 1.1.0-3 amd64 GTK+ library to embed
OpenStreetMap maps - Python bindings
ii libcupsmime1:amd64 2.2.8-5 amd64 Common UNIX Printing
System(tm) - MIME library
ii libosmgpsmap-1.0-1:amd64 1.1.0-3 amd64 GTK+ library to embed
OpenStreetMap maps
ii libpsm-infinipath1 3.3+20.604758e7-5 amd64 PSM Messaging
library for Intel Truescale adapters
ii libpsm2-2 10.3.58-1 amd64 Intel PSM2 library
ii psmisc 23.1-1+b1 amd64 utilities that use the proc file system
Any changes to /etc/openmpi/* files or environmental variables
No changes to /etc/openmpi/*. Here is my complete environment:
gmulas@capitanata:~/transfer/debsrc$ env
LC_ALL=it_IT.utf8
LC_MEASUREMENT=it_IT.UTF-8
LC_PAPER=it_IT.UTF-8
LC_MONETARY=it_IT.UTF-8
XDG_MENU_PREFIX=gnome-
LANG=it_IT.utf8
GDM_LANG=it_IT.utf8
LESS=-eiM
DISPLAY=:0
OLDPWD=/home/gmulas/transfer/debsrc/scalapack-2.0.2
DUSTEM_FILTER_DIR=/home/gmulas/PAHmodels/DUSTEM/Data/FILTERS/
EDITOR=joe
COLORTERM=truecolor
USERNAME=gmulas
NO_AT_BRIDGE=1
XDG_VTNR=2
SSH_AUTH_SOCK=/run/user/1000/keyring/ssh
PARA_ARCH=MPI
CLASSPATH=/usr/local/JMF-2.1.1e/lib/jmf.jar:.:/usr/local/JMF-2.1.1e/lib/jmf.jar:.
XDG_SESSION_ID=20
USER=gmulas
DUSTEM_DAT=/tmp/
DUSTEM_SOFT_DIR=/home/gmulas/PAHmodels/DUSTEM
PAGER=/usr/bin/less -eiMs
DESKTOP_SESSION=gnome-xorg
JMFHOME=/usr/local/JMF-2.1.1e
GNOME_TERMINAL_SCREEN=/org/gnome/Terminal/screen/7aa5ce9c_afbf_4166_9720_8b3d3187d254
PWD=/home/gmulas/transfer/debsrc
HOME=/home/gmulas
TMP=/tmp/user/1000
DUSTEM_DRAINE_DATA_DIR=/home/gmulas/PAHmodels/DUSTEM/Data/DRAINE/
SSH_AGENT_PID=13026
UPARM=/home/gmulas/uparm
QT_ACCESSIBILITY=1
MONKEYSPHERE_VALIDATION_AGENT_SOCKET=http://127.0.0.1:35619
XDG_SESSION_TYPE=x11
MIDASHOME=/usr/local/midas
XDG_DATA_DIRS=/usr/share/gnome:/home/gmulas/.local/share/flatpak/exports/share/:/var/lib/flatpak/exports/share/:/usr/local/share/:/usr/share/:
DUSTEM_RES=/tmp/
XDG_SESSION_DESKTOP=gnome-xorg
GJS_DEBUG_OUTPUT=stderr
LC_NUMERIC=it_IT.UTF-8
DUSTEM_WHICH=DESERT
GTK_MODULES=gtk-vector-screenshot:gail:atk-bridge
WINDOWPATH=2
TURBOMOLE_SYSNAME=em64t-unknown-linux-gnu
TERM=xterm-256color
SHELL=/bin/bash
VTE_VERSION=5202
TEMPDIR=/tmp/user/1000
TURBODIR=/usr/local/tmoleX/TURBOMOLE
XDG_CURRENT_DESKTOP=GNOME
GPG_AGENT_INFO=/run/user/1000/gnupg/S.gpg-agent:0:1
GNOME_TERMINAL_SERVICE=:1.319
SHLVL=1
XDG_SEAT=seat0
DUSTEM_PAHUV=DESERT
TEMP=/tmp/user/1000
GDMSESSION=gnome-xorg
LESSCHARSET=utf-8
GNOME_DESKTOP_SESSION_ID=this-is-deprecated
LOGNAME=gmulas
DBUS_SESSION_BUS_ADDRESS=unix:path=/run/user/1000/bus
XDG_RUNTIME_DIR=/run/user/1000
XAUTHORITY=/run/user/1000/gdm/Xauthority
UDEdir=
JACK_START_SERVER=1
PATH=/usr/local/tmoleX/TURBOMOLE/bin/em64t-unknown-linux-gnu_mpi:/usr/local/tmoleX/TURBOMOLE/scripts:/usr/local/bin:/usr/bin:/bin:/usr/local/games:/usr/games:/usr/local/gOpenMol-3.00/bin
MIDVERS=13SEP
GJS_DEBUG_TOPICS=JS ERROR;JS LOG
SESSION_MANAGER=local/capitanata:@/tmp/.ICE-unix/13067,unix/capitanata:/tmp/.ICE-unix/13067
PIPE_HOME=/usr/local/midas/13SEP/pipeline
LC_TIME=it_IT.UTF-8
_=/usr/bin/env
And, in case it may help, here goes the output of ldd on the simple
"hello
world" MPI code I sent you, that hangs here:
gmulas@capitanata:~/PAHmodels/anharmonica-scalapack$ ldd sample_printf
linux-vdso.so.1 (0x00007ffd034e3000)
libmpi.so.40 => /usr/lib/x86_64-linux-gnu/libmpi.so.40
(0x00007f94cc309000)
libpthread.so.0 => /lib/x86_64-linux-gnu/libpthread.so.0
(0x00007f94cc2e8000)
libc.so.6 => /lib/x86_64-linux-gnu/libc.so.6 (0x00007f94cc12b000)
libdl.so.2 => /lib/x86_64-linux-gnu/libdl.so.2 (0x00007f94cc126000)
libopen-rte.so.40 => /usr/lib/x86_64-linux-gnu/libopen-rte.so.40
(0x00007f94cc06f000)
libopen-pal.so.40 => /usr/lib/x86_64-linux-gnu/libopen-pal.so.40
(0x00007f94cbfc2000)
librt.so.1 => /lib/x86_64-linux-gnu/librt.so.1 (0x00007f94cbfb6000)
libm.so.6 => /lib/x86_64-linux-gnu/libm.so.6 (0x00007f94cbe22000)
libutil.so.1 => /lib/x86_64-linux-gnu/libutil.so.1
(0x00007f94cbe1d000)
libhwloc.so.5 => /usr/lib/x86_64-linux-gnu/libhwloc.so.5
(0x00007f94cbddc000)
libevent-2.1.so.6 => /usr/lib/x86_64-linux-gnu/libevent-2.1.so.6
(0x00007f94cbb86000)
libevent_pthreads-2.1.so.6 =>
/usr/lib/x86_64-linux-gnu/libevent_pthreads-2.1.so.6 (0x00007f94cb983000)
/lib64/ld-linux-x86-64.so.2 (0x00007f94cc4a8000)
libnuma.so.1 => /usr/lib/x86_64-linux-gnu/libnuma.so.1
(0x00007f94cb776000)
libltdl.so.7 => /usr/lib/x86_64-linux-gnu/libltdl.so.7
(0x00007f94cb56c000)
Please let me know whatever I can do to help pinpoint this. I need to be
able to use my computer to develop my MPI codes, and I cannot
understand at
all why it abruptly stopped working. Also, if you tell me it does work
properly on another current sid system, I'd like to find out what
makes the
difference.
Thanks in advance
Giacomo Mulas
--
Alastair McKinstry, <alast...@sceal.ie>, <mckins...@debian.org>,
https://diaspora.sceal.ie/u/amckinstry
Commander Vimes didn’t like the phrase “The innocent have nothing to fear,”
believing the innocent had everything to fear, mostly from the guilty but in
the longer term
even more from those who say things like “The innocent have nothing to fear.”
- T. Pratchett, Snuff
