AW: Linux-x86_64-CUDA version 2.8 on CentOS-5 x86_64 non local user issue?

From: Norman Geist (norman.geist_at_uni-greifswald.de)
Date: Wed Mar 21 2012 - 02:02:41 CDT

> -----Ursprüngliche Nachricht-----
> Von: owner-namd-l_at_ks.uiuc.edu [mailto:owner-namd-l_at_ks.uiuc.edu] Im
> Auftrag von Tru Huynh
> Gesendet: Dienstag, 20. März 2012 14:33
> An: Norman Geist
> Cc: Namd Mailing List
> Betreff: Re: namd-l: Linux-x86_64-CUDA version 2.8 on CentOS-5 x86_64
> non local user issue?
>
> On Tue, Mar 20, 2012 at 07:36:34AM +0100, Norman Geist wrote:
> > Hi,
> >
> > from what you wrote it don't look like a namd problem, but a
> permissions
> > problem. To use the cuda devices (I think it's "/dev/nvidia" or
> similar)
> That was my initial ideas, but that's not the case:
> /dev/nvidia* permissions are 666 as listed in the initial mail.
>
> > If the above is the problem, you shouldn’t be able to run any cuda
> binary
> > with that users, not only namd.
> as stated: nvidia-smi works fine as any user (w/wo begin listed in

Tru,

nvidia-smi is not a cuda program, it's just a driver utility. Please check
if you can run other cuda programs, maybe one example from the cuda sdk,
because I don't think it's a namd problem. You can also look to the namd
source where the message about non available gpus is generated and see what
command fails (I will check that also). It looks like your user has no
permission to list the available devices. So check what is the difference
between local users and non-local (bashrc,LD_LIBRARY_PATH...,cuda-toolkit).
Check with "ldd namd2" if local and non-local users use the same shared
librarys.

> /etc/passwd).
>
> Tru
> --
> Dr Tru Huynh | http://www.pasteur.fr/recherche/unites/Binfs/
> mailto:tru_at_pasteur.fr | tel/fax +33 1 45 68 87 37/19
> Institut Pasteur, 25-28 rue du Docteur Roux, 75724 Paris CEDEX 15
> France

This archive was generated by hypermail 2.1.6 : Tue Dec 31 2013 - 23:21:47 CST