Hello!
I got something error when I tried to make a simuation with QUANTUM-ESPRESSO.
When I input command: pw.x,the output will be following:
pw.x: error while loading shared libraries: libacccuda.so: cannot open shared object file: No such file or directory
When I input command: mpirun, the output is following:
Sorry! You were supposed to get help about:
orterun:nothing-to-do
But I couldn’t open the help file:
/proj/nv/libraries/Linux_x86_64/24.11/openmpi4-12/264112-rel-1/comm_libs/12.6/openmpi4/openmpi-4.1.5/share/openmpi/help-orterun.txt: No such file or directory. Sorry!
But I can find the help-orterun.txt is in /opt/nvidia/hpc_sdk/Linux_x86_64/24.11/comm_libs/12.6/openmpi4/openmpi-4.1.5/share/openmpi/help-orterun.txt
And when I print in the command: mpirun -np 2 pw.x, the error is following:
Caught signal 11 (Segmentation fault: address not mapped to object at address (nil))
==== backtrace (tid: 107197) ====
0 /opt/nvidia/hpc_sdk/Linux_x86_64/24.11/comm_libs/12.6/openmpi4/openmpi-4.1.5/lib/libucs.so.0(ucs_handle_error+0x124) [0xc634ac32084]
1 /opt/nvidia/hpc_sdk/Linux_x86_64/24.11/comm_libs/12.6/openmpi4/openmpi-4.1.5/lib/libucs.so.0(+0x323ac) [0xc634ac323ac]
2 /opt/nvidia/hpc_sdk/Linux_x86_64/24.11/comm_libs/12.6/openmpi4/openmpi-4.1.5/lib/libucs.so.0(+0x3261b) [0xc634ac3261b]
3 /lib/x86_64-linux-gnu/libc.so.6(+0x45320) [0xc634be45320]
Segmentation fault (core dumped)
My enviroment is with two RTX A4000, driver version is 560, cuda version is 12.6, hpc_sdk version is 24.11.
I also have tried to reinstall the driver, cuda toolkit, HPC_SDK and quantume-espresso, but error is still there.
And I also found one thing. When I looked at files in openmpi/share, I found :
linker_flags=-L/proj/nv/libraries/Linux_x86_64/24.11/openmpi4-12/264112-rel-1/comm_libs/12.6/openmpi4/openmpi-4.1.5/lib -Wl,-rpath,$ORIGIN:$ORIGIN/…/…/lib:$ORIGIN/…/…/…/…/…/compilers/lib:$ORIGIN/…/…/…/…/…/…/compilers/lib -Wl,-rpath -Wl,@{libdir} -Wl,-rpath -Wl,/proj/nv/libraries/Linux_x86_64/24.11/openmpi4-12/264112-rel-1/comm_libs/12.6/openmpi4/openmpi-4.1.5/lib
All these links is not the correct directory for my nvidia openmpi.
Could anyone help me solve this problem?