Re: [AMBER] Amber22 CUDA compilation problem

From: Irek Porebski via AMBER <amber.ambermd.org>
Date: Wed, 31 May 2023 01:40:44 +0000

Hi Florian,

Thank you for your help, this solution partially helps. The compilation almost finished, but it failed on:

[ 96%] Building C object AmberTools/src/mdgx/CMakeFiles/mdgx.cuda.dir/AmberNetcdf.c.o
[ 96%] Building C object AmberTools/src/mdgx/CMakeFiles/mdgx.cuda.dir/mdgx.c.o
[ 96%] Linking CXX executable mdgx.cuda
CMakeFiles/mdgx.cuda.dir/mdgx.cuda_generated_ArraySimulator.cu.o: In function `GetGpuSpecs':
tmpxft_00040c81_00000000-6_ArraySimulator.compute_80.cudafe1.cpp:(.text+0x12f): undefined reference to `nvmlDeviceGetComputeRunningProcesses_v3'
collect2: error: ld returned 1 exit status
make[2]: *** [AmberTools/src/mdgx/mdgx.cuda] Error 1
make[1]: *** [AmberTools/src/mdgx/CMakeFiles/mdgx.cuda.dir/all] Error 2
make: *** [all] Error 2

Does anyone have solution for this error?

Thanks,
Irek

From: Pertschy, Florian <florian.pertschy.medunigraz.at>
Sent: Wednesday, May 31, 2023 6:43 AM
To: Irek Porebski <i.porebski.uq.edu.au>; AMBER Mailing List <amber.ambermd.org>
Subject: Re: Amber22 CUDA compilation problem


Hey!

Funny coincidence, I asked more or less the exact same question today, a couple of hours before you. I have since stumbled upon a thread regarding this issue (http://archive.ambermd.org/202105/0103.html) saying
>Please edit your run_cmake script, and add the following flag:
>
> -DFORCE_INTERNAL_LIBS="fftw"
>
>Then re-run "run_cmake" and "make install".


This was related to Amber20 and not Amber22 though, and I haven't yet been able to test if it works

Best
Florian
________________________________
From: Irek Porebski via AMBER <amber.ambermd.org<mailto:amber.ambermd.org>>
Sent: 30 May 2023 15:06:01
To: amber.ambermd.org<mailto:amber.ambermd.org>
Subject: [EXT] [AMBER] Amber22 CUDA compilation problem

Hi All

I am trying to compile Amber22 with CUDA support. I tried many version of CUDA and GCC but sooner or later compilation failed.
My current configuration is

Currently Loaded Modules:
  1) cmake/3.25.3 3) bison-3.0.4-gcc-7.2.0-nl7vmez 5) cuda/11.4.1 7) boost/1.75.0
  2) python-3.7.0-gcc-12.3.0-tvnfvfi 4) anaconda/3.6 6) gnu7/7.3.0 8) fftw-3.3.7-gcc-4.8.5-5igtfs5

And the compilation ends like this:

[ 14%] Building CXX object AmberTools/src/cew/CMakeFiles/libcew.dir/iface.cpp.o
[ 14%] Linking CXX shared library libcew.so
/usr/bin/ld: /usr/local/lib/libfftw3.a(apiplan.o): relocation R_X86_64_32S against `.rodata' can not be used when making a shared object; recompile with -fPIC
/usr/bin/ld: /usr/local/lib/libfftw3.a(mapflags.o): relocation R_X86_64_32 against `.rodata' can not be used when making a shared object; recompile with -fPIC
/usr/bin/ld: /usr/local/lib/libfftw3.a(alloc.o): relocation R_X86_64_32 against `.rodata.str1.1' can not be used when making a shared object; recompile with -fPIC
Could you please help me with this or point where to search for more information?
...
/usr/bin/ld: /usr/local/lib/libfftw3.a(direct-r2r.o): relocation R_X86_64_32 against `.text' can not be used when making a shared object; recompile with -fPIC
/usr/bin/ld: /usr/local/lib/libfftw3.a(direct2.o): relocation R_X86_64_32 against `.text' can not be used when making a shared object; recompile with -fPIC
/usr/bin/ld: /usr/local/lib/libfftw3.a(hc2hc-direct.o): relocation R_X86_64_32 against `.text' can not be used when making a shared object; recompile with -fPIC
/usr/bin/ld: final link failed: Nonrepresentable section on output
collect2: error: ld returned 1 exit status
make[2]: *** [AmberTools/src/cew/libcew.so] Error 1
make[1]: *** [AmberTools/src/cew/CMakeFiles/libcew.dir/all] Error 2
make: *** [all] Error 2

Thanks,
Irek

_______________________________________________
AMBER mailing list
AMBER.ambermd.org<mailto:AMBER.ambermd.org>
http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue May 30 2023 - 19:00:02 PDT
Custom Search