PGI User Forum

[mkcolg]

[alechand]

i am trying to install the cuda driver from nvidia website you recomended,
but after installation, it says that the samples could not be installed because missing libraries: Missing required library libglut.so

My question is, the cuda driver which comes with pgi compiler i downloaded from pgi website, is not the appropriate to use ?

thanks

[mkcolg]

[alechand]

Hello.

I decided to erase my machine and install the old ( it was working ) kubuntu 11.04 and the PGI compiler 12.10 ( for which i have bought a license ). I installed the last graphic card driver from NVIDIA website.
I needed to make a copy of crt* files from /usr/lib/i386-linux-gnu/
-> to -> /usr/lib, in order to make the compiler work.

When i compile a code, it seems to be nice:

######################################

alechand@pcsantos2:~/test_openacc$ pgfortran -fast -Minfo=all -o MOL_DYN Mol_Dyn.f90 -ta=nvidia
mol_dyn:
28, Loop unrolled 4 times
33, Loop unrolled 8 times
38, Loop unrolled 16 times
76, Generating present_or_copyin(xold(1:5000))
Generating present_or_copy(x(1:5000))
Generating present_or_copy(v(1:5000))
Generating present_or_copy(f(1:5000))
Generating compute capability 1.3 binary
Generating compute capability 2.0 binary
77, Loop is parallelizable
Accelerator kernel generated
77, !$acc loop gang, vector(128) ! blockidx%x threadidx%x
CC 1.3 : 18 registers; 60 shared, 12 constant, 0 local memory bytes
CC 2.0 : 20 registers; 0 shared, 76 constant, 0 local memory bytes
83, Loop is parallelizable
106, Loop unrolled 16 times

#######################################

When i execute i see the problem:

#######################################

alechand@pcsantos2:~/test_openacc$ ./MOL_DYN
call to cuMemFree returned error 700: Launch failed
CUDA driver version: 5050

#######################################

I tryed to install the cuda-5 from nvidia website, but it did not change nothing.

I really want to put this to work, that is the reason i bought it ...
Please, can you help me ?

PS: the compilator was working properly with the same code. The problem SEEMS to have started after i've tryed to write the result of a code in an output text file, (using WRITE at the fortran code). Do you think the graphic card can be with memory problem from that time ?

I really appreciate your attention.

PS 2: the previous simple program picalc.f90 is also giving a similar memory error :

################################

alechand@pcsantos2:~/test_openacc$ pgfortran -fast -Minfo=all -o MOL_DYN picalc.f90 -ta=nvidia
picalc:
7, Accelerator kernel generated
7, CC 1.3 : 24 registers; 32 shared, 36 constant, 0 local memory bytes
CC 2.0 : 23 registers; 0 shared, 52 constant, 0 local memory bytes
8, !$acc loop gang, vector(256) ! blockidx%x threadidx%x
7, Generating compute capability 1.3 binary
Generating compute capability 2.0 binary
alechand@pcsantos2:~/test_openacc$ ./MOL_DYN
call to cuMemcpyDtoH returned error 700: Launch failed
CUDA driver version: 5050

#################################

Thanks a lot !

[mkcolg]