Closed miroi closed 3 years ago
Is this with docker? If the answer is yes, what docker image are you using? SIGILL means that the NWChem installation was optimized for computer instructions not present on the host computer you are using.
Hi, this is via singularity, which can pull containers from docker hub. The processor is Intel(R) Xeon(R) CPU E5-2640 0 @ 2.50GHz.
Here is the list of libraries nwchem-in-container is using...How to find out which one is causing problems ?
ilias@login1.kelinux.saske.sk:~/work/qch/projects/open-collection/computer_science/containers/kelinux_saske_sk/singularity/nwchem/.singularity pull $SINGULARITY_CONTAINERS/nwchem.sif docker://nwchemorg/nwchem-702.mpipr.nersc
.
.
2021/08/28 18:07:53 info unpack layer: sha256:cadb02169db62b58185ae002dac457bd0e7ebdf79468ef6c5a13db393b5e87a8
INFO: Creating SIF file...
ilias@login1.kelinux.saske.sk:~/work/qch/projects/open-collection/computer_science/containers/kelinux_saske_sk/singularity/nwchem/.singularity exec $SINGULARITY_CONTAINERS/nwchem.sif /bin/sh
Singularity> ldd /opt/nwchem-7.0.2/bin/LINUX64/nwchem
linux-vdso.so.1 (0x00007fff03375000)
libmpifort.so.12 => /usr/local/lib/libmpifort.so.12 (0x00007f0a84eb5000)
libmpi.so.12 => /usr/local/lib/libmpi.so.12 (0x00007f0a84b5e000)
librt.so.1 => /lib/x86_64-linux-gnu/librt.so.1 (0x00007f0a84b53000)
libpthread.so.0 => /lib/x86_64-linux-gnu/libpthread.so.0 (0x00007f0a84b30000)
libgfortran.so.5 => /lib/x86_64-linux-gnu/libgfortran.so.5 (0x00007f0a84868000)
libm.so.6 => /lib/x86_64-linux-gnu/libm.so.6 (0x00007f0a84719000)
libmvec.so.1 => /lib/x86_64-linux-gnu/libmvec.so.1 (0x00007f0a846eb000)
libpython3.8.so.1.0 => /lib/x86_64-linux-gnu/libpython3.8.so.1.0 (0x00007f0a84195000)
libc.so.6 => /lib/x86_64-linux-gnu/libc.so.6 (0x00007f0a83fa3000)
libgcc_s.so.1 => /lib/x86_64-linux-gnu/libgcc_s.so.1 (0x00007f0a83f88000)
/lib64/ld-linux-x86-64.so.2 (0x00007f0a941e2000)
libquadmath.so.0 => /lib/x86_64-linux-gnu/libquadmath.so.0 (0x00007f0a83f3e000)
libexpat.so.1 => /lib/x86_64-linux-gnu/libexpat.so.1 (0x00007f0a83f0e000)
libz.so.1 => /lib/x86_64-linux-gnu/libz.so.1 (0x00007f0a83ef2000)
libdl.so.2 => /lib/x86_64-linux-gnu/libdl.so.2 (0x00007f0a83eec000)
libutil.so.1 => /lib/x86_64-linux-gnu/libutil.so.1 (0x00007f0a83ee7000)
Singularity>
Singularity> /usr/local/bin/mpirun -np 2 /opt/nwchem-7.0.2/bin/LINUX64/nwchem h2o_scf_6-31g.nw
argument 1 = h2o_scf_6-31g.nw
.
.
C2V symmetry detected
Program received signal SIGILL: Illegal instruction.
Backtrace for this error:
#0 0x7f3f30716d01 in ???
#1 0x7f3f30715ed5 in ???
#2 0x7f3f2fe7720f in ???
#3 0x55576496b5ef in ???
===================================================================================
= BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
= PID 20524 RUNNING AT login1
= EXIT CODE: 4
= CLEANING UP REMAINING PROCESSES
= YOU CAN IGNORE THE BELOW CLEANUP MESSAGES
===================================================================================
YOUR APPLICATION TERMINATED WITH THE EXIT STRING: Illegal instruction (signal 4)
This typically refers to a problem with your application.
Please see the FAQ page for debugging suggestions
Singularity>
Ah, @edoapra , the " BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES" message is connected with Intel compilers.
Looking into containers /usr/local/bin folder, I see
Singularity> mpifort --version
GNU Fortran (Ubuntu 9.3.0-17ubuntu1~20.04) 9.3.0
Copyright (C) 2019 Free Software Foundation, Inc.
Maybe something is missing from Intel libraries in the container ?
~If you don't get me the details of the docker image you are using, there is nothing I can do about it and I will close this issue.~ ~To be more precise:~ ~ did you create the docker image yourself? What Dockerfile did you use?~ ~ did you pull a docker image from hub.docker.com? What image did you pull~
I managed to understand what you have been doing by scrolling the output.
Hi, yes, I did pull your docker image via "singularity pull $SINGULARITY_CONTAINERS/nwchem.sif docker://nwchemorg/nwchem-702.mpipr.nersc" and was using only this image, no other.
As described in https://nwchemgit.github.io/Containers.html Singularity images of NWChem are available https://cloud.sylabs.io/library/edoapra Please keep in mind that they have been optimized for certain CPU instructions
Thanks, closing the ticket.
Just checked containers with NWChem in https://cloud.sylabs.io/library/edoapra , some are running on my machine(s). Thanks !
Thanks for the feedback
Hello, on one cluster, I got this error with nwchem-in-container: