From: John Stone (johns_at_ks.uiuc.edu)
Date: Wed Dec 03 2014 - 14:34:31 CST

Vlad,
  Here is further discussion of the XOpenIM() stuff in FLTK, pertaining
to some potential performance problems with certain locales and the
most recent versions of FLTK:
  https://groups.google.com/forum/#!topic/fltkgeneral/X1heD2dWaOw

Do you get the same XOpenIM() error/warning if you run some of the
FLTK example/demo applications, or only when running VMD?

Cheers,
  John

On Wed, Dec 03, 2014 at 02:31:20PM -0600, John Stone wrote:
> Vlad,
> I did some googling, and it appears that FLTK 1.3.x might be the source of
> the "XOpenIM() failed" message:
> https://www.mail-archive.com/fltk-bugs@easysw.com/msg04795.html
>
> That email chain indicates they fixed the problem, but perhaps they didn't
> fix it in all cases, or some of the changes that were made were either
> missing for FLTK 1.3.2, or maybe they were reverted or lost during some
> other updates. If you can try a different FLTK version, you might be
> able to prove whether or not FLTK is the code that's generating the warning.
>
> Cheers,
> John Stone
> vmd_at_ks.uiuc.edu
>
> On Wed, Dec 03, 2014 at 09:23:30PM +0100, Vlad Cojocaru wrote:
> > Hi John,
> >
> > I dont remember to have had any issue with the on-the-fly recognition of
> > CUDA devices before on openSUSE-12.X ... I dont remember exactly now if
> > users were automatically added to the "video" group which may explain why
> > I did not have issues (and dont have machines with earlier OS anymore)
> >
> > As for the XOpenIM issue, I initially thought it is related to the CUDA
> > stuff ... But other than that, I did not see any issue with VMD
> > functionality so far. So, I'll ignore it for now ... If I notice anything
> > again, I will get back to you and the list
> >
> > Best,
> > Vlad
> >
> > On December 3, 2014 9:04:03 PM CET, John Stone <johns_at_ks.uiuc.edu> wrote:
> >
> > Vlad,
> > Is the device node permission setting behavior different between the
> > old revs of OpenSuse you used before and the new one? If so, this is an
> > interesting thing to note. I still have no idea what to tell you about
> > the "XOpenIM()" error you see, since that presumably has nothing to do
> > with CUDA or the GPU device permissions and is more likely an issue of
> > compatibility between your X11 and the FLTK/Tcl/Tk libraries you're
> > compiling against.
> >
> > Cheers,
> > John
> >
> > On Wed, Dec 03, 2014 at 09:00:49PM +0100, Vlad Cojocaru wrote:
> >
> > Thanks John and Axel, ...
> >
> > Changing permissions on /dev/nvidia* (chmod o+rw /dev/nvidia*) solved
> > the issue. The permissions were rw only for the user "root" and group
> > "video" .... An alternative solution maybe to add user "vlad" to
> > group
> > "video" but did not test yet.
> >
> > Best
> > Vlad
> >
> >
> > On 12/03/2014 08:21 PM, John Stone wrote:
> >
> > If you get no output when running as vlad, that may explain
> > what is going wrong with VMD. Are the permissions set correctly
> > on the GPU device nodes when you login as vlad? If you can't
> > run 'nvidia-smi' as 'vlad', then you won't be able to get CUDA working
> > as 'vlad' either. You need to figure out what's going wrong there and
> > correct file permissions or whatever it is that's preventing it from
> > working.
> >
> > Cheers,
> > John Stone
> > vmd_at_ks.uiuc.edu
> >
> > On Wed, Dec 03, 2014 at 07:50:06PM +0100, Vlad Cojocaru wrote:
> >
> > Sorry John,
> >
> > I was running as
> > user "vlad" and that did not produce any output ..
> > Running as root I get this output below (nothing that I notice to be
> > wrong):
> >
> > Best
> > Vlad
> >
> > Wed Dec 3 19:48:41 2014
> > +
> >
> > ---------------------------------------------------------------------------
> >
> > +
> > | NVIDIA-SMI 340.46 Driver Version: 340.46 |
> > |
> >
> > ---------------------------------------------------------------------------
> >
> > +----------------------+----------------------+
> > | GPU Name Persistence-M| Bus-Id Disp.A | Volatile
> > Uncorr. ECC |
> > | Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util
> > Compute M. |
> > |
> >
> > ---------------------------------------------------------------------------
> >
> > +======================+======================|
> > | 0 Quadro K4000 Off | 0000:02:00.0 On
> > | N/A |
> > | 30% 46C P8 23W / 87W | 1448MiB / 3071MiB | 3% Default |
> > +
> >
> > ---------------------------------------------------------------------------
> >
> > +----------------------+----------------------+
> >
> > +
> >
> > ---------------------------------------------------------------------------
> >
> > +
> > | Compute processes: GPU Memory |
> > | GPU PID Process name Usage |
> > |
> >
> > ---------------------------------------------------------------------------
> >
> > |
> > | No running compute processes
> > found
> > |
> > +
> >
> > ---------------------------------------------------------------------------
> >
> > +
> >
> > On 12/03/2014 07:47 PM, Vlad Cojocaru wrote:
> >
> > Hi John,
> >
> > NVIDIA driver: 340.46
> >
> > Weirdly, "nvidia-smi" does not produce any output whatsoever ...
> > There shouldn't be any special compute nodes. This is my own
> > workstation ...
> >
> > Best
> > Vlad
> >
> > On 12/03/2014 07:34 PM, John Stone wrote:
> >
> > Vlad,
> > Thanks. What NVIDIA driver version is installed on this machine?
> > If you run "nvidia-smi", does it print any warnings or errors, and
> > do you see all of your GPUs listed? Are any set for "prohibited"
> > or "exclusive" compute mode?
> >
> > Cheers,
> > John Stone
> > vmd_at_ks.uiuc.edu
> >
> > On Wed, Dec 03, 2014 at 07:27:55PM
> > +0100, Vlad Cojocaru wrote:
> >
> > Hi John,
> >
> > These are the options I used (same as for all my previous compilations
> > that worked just fine on openSUSE 12.X):
> >
> > LINUXAMD64 OPENGL FLTK TK ACTC CUDA OPENCL IMD XINERAMA XINPUT
> > LIBTACHYON NETCDF COLVARS TCL PYTHON PTHREADS NUMPY LP64
> >
> > My system has a standard openSUSE 13.1 with tcl 8.6 and Fltk 1.3.2, and
> > CUDA 6.5.14 .. I am not aware of any internationalized input scheme
> > setup ....
> >
> > Best
> > Vlad
> >
> > On 12/03/2014 05:26 PM, John Stone wrote:
> >
> > Vlad,
> > What do you get when you run this command:
> > vmdinfo options
> >
> > Cheers,
> > John Stone
> >
> > On Wed, Dec 03, 2014 at 01:47:01PM +0100, Vlad Cojocaru wrote:
> >
> > Dear all,
> >
> > I am experiencing some issues with running VMD with CUDA
> > acceleration on
> > my machine. I compiled from source on linux x86_64 (openSUSE 13.1)
> > against CUDA 6.5.14 ...NVIDIA driver 340.46 ... At compilation, no
> > issues whatsoever. However, every time I try to run VMD I get the
> > error
> > below. The program runs fine but CUDA acceleration is not available.
> >
> > Has anybody encountered a similar issue ? If yes, any idea how to
> > fix it ?
> >
> > Thanks in advance
> >
> > Best wishes
> > Vlad
> >
> > Info) VMD for LINUXAMD64, version 1.9.2beta1 (October 23, 2014)
> > Info) [1]http://www.ks.uiuc.edu/Research/vmd/
> > Info) Email questions and bug reports to vmd_at_ks.uiuc.edu
> > Info) Please include this reference in published work using VMD:
> > Info)
> > Humphrey, W., Dalke, A. and Schulten, K., `VMD - Visual
> > Info) Molecular Dynamics', J. Molec. Graphics 1996, 14.1, 33-38.
> > Info)
> >
> > ---------------------------------------------------------------------------
> >
> > Info) Multithreading available, 24 CPUs detected.
> > Info) Free system memory: 56003MB (86%)
> > XOpenIM() failed
> > Info) No CUDA accelerator devices available.
> > Info) OpenGL renderer: Quadro K4000/PCIe/SSE2
> > Info) Features: STENCIL STEREO MSAA(4) MDE MTX NPOT PP PS
> > GLSL(OVFG)
> > Info) Full GLSL rendering mode is available.
> > Info) Textures: 2-D (16384x16384), 3-D (4096x4096x4096),
> > Multitexture
> > (4)
> >
> > --
> > Dr. Vlad Cojocaru
> > Computational Structural Biology Laboratory
> > Department of Cell and Developmental Biology
> > Max Planck Institute for Molecular Biomedicine
> > RAPntgenstrasse 20, 48149 MA 1/4nster, Germany
> > Tel: +49-251-70365-324; Fax: +49-251-70365-399
> > Email: vlad.cojocaru[at][2]mpi-muenster.mpg.de
> > [3]http://www.mpi-muenster.mpg.de/research/teams/groups/rgcojocaru
> >
> > --
> > Dr. Vlad Cojocaru
> > Computational Structural Biology Laboratory
> > Department of Cell and Developmental Biology
> > Max Planck Institute for Molecular Biomedicine
> > RAPntgenstrasse 20, 48149 MA 1/4nster, Germany
> > Tel: +49-251-70365-324; Fax: +49-251-70365-399
> > Email: vlad.cojocaru[at][4]mpi-muenster.mpg.de
> > [5]http://www.mpi-muenster.mpg.de/research/teams/groups/rgcojocaru
> >
> > --
> > Dr. Vlad Cojocaru
> > Computational Structural Biology Laboratory
> > Department of Cell and Developmental Biology
> > Max Planck Institute for Molecular Biomedicine
> > RAPntgenstrasse 20, 48149 MA 1/4nster, Germany
> > Tel: +49-251-70365-324; Fax:
> > +49-251-70365-399
> > Email: vlad.cojocaru[at][6]mpi-muenster.mpg.de
> > [7]http://www.mpi-muenster.mpg.de/research/teams/groups/rgcojocaru
> >
> >
> > --
> > Dr. Vlad Cojocaru
> > Max Planck Institute for Molecular Biomedicine
> > Department of Cell and Developmental Biology
> > RAPntgenstrasse 20, 48149 MA 1/4nster, Germany
> > Tel: +49-251-70365-324; Fax: +49-251-70365-399
> > Email: vlad.cojocaru[at][8]mpi-muenster.mpg.de
> >
> > --
> > Sent from my Android device with K-9 Mail. Please excuse my brevity.
> >
> > References
> >
> > Visible links
> > 1. http://www.ks.uiuc.edu/Research/vmd
> > 2. http://mpi-muenster.mpg.de/
> > 3. http://www.mpi-muenster.mpg.de/research/teams/groups/rgcojocaru
> > 4. http://mpi-muenster.mpg.de/
> > 5. http://www.mpi-muenster.mpg.de/research/teams/groups/rgcojocaru
> > 6. http://mpi-muenster.mpg.de/
> > 7. http://www.mpi-muenster.mpg.de/research/teams/groups/rgcojocaru
> > 8. http://mpi-muenster.mpg.de/
>
> --
> NIH Center for Macromolecular Modeling and Bioinformatics
> Beckman Institute for Advanced Science and Technology
> University of Illinois, 405 N. Mathews Ave, Urbana, IL 61801
> http://www.ks.uiuc.edu/~johns/ Phone: 217-244-3349
> http://www.ks.uiuc.edu/Research/vmd/

-- 
NIH Center for Macromolecular Modeling and Bioinformatics
Beckman Institute for Advanced Science and Technology
University of Illinois, 405 N. Mathews Ave, Urbana, IL 61801
http://www.ks.uiuc.edu/~johns/           Phone: 217-244-3349
http://www.ks.uiuc.edu/Research/vmd/