CUDA won't recognize card

Accelerated Computing CUDA CUDA Programming and Performance
1

Hi there,

I’ve been attempting to install CUDA in openSUSE 11.1 (64 bit) for some time now, but to no avail and would really appreciate any help, I have the following set up:

Desktop: KDE4
Card: GeForce 9500 GT
Driver: 185.18.14

I’m using the default kernel for the driver installed through Yast; I attempted installing the driver manually, but sax2 ended up not being able to start and I was without any X, so I reverted to the former. I’ve noticed that at least a few people on this site have used a work around by adding uppermem, vmalloc and pci to the grub boot file (menu.lst); I tried this too:

root (hd0,0)
uppermem 524288
kernel /vmlinuz-2.6.27.23-0.1-default root=/dev/md0 resume=/dev/disk/by-id/ata-ST3500320AS_9QM4X7MA-part3 splash=silent showopts vga=0x31a vmalloc=256MB pci=nommconf
initrd /initrd-2.6.27.23-0.1-default

But my deviceQuery still gives the following output:

CUDA Device Query (Runtime API) version (CUDART static linking)
There is no device supporting CUDA.

Device 0: "Device Emulation (CPU)"
  CUDA Capability Major revision number:         9999
  CUDA Capability Minor revision number:         9999
  Total amount of global memory:                 4294967295 bytes
  Number of multiprocessors:                     16
  Number of cores:                               128
  Total amount of constant memory:               65536 bytes
  Total amount of shared memory per block:       16384 bytes
  Total number of registers available per block: 8192
  Warp size:                                     1
  Maximum number of threads per block:           512
  Maximum sizes of each dimension of a block:    512 x 512 x 64
  Maximum sizes of each dimension of a grid:     65535 x 65535 x 1
  Maximum memory pitch:                          262144 bytes
  Texture alignment:                             256 bytes
  Clock rate:                                    1.35 GHz
  Concurrent copy and execution:                 No
  Run time limit on kernels:                     No
  Integrated:                                    Yes
  Support host page-locked memory mapping:       Yes
  Compute mode:                                  Default (multiple host threads can use this device simultaneously)

Test PASSED

I had CUDA working on XP before I decided to switch to Linux, so this is fairly frustrating… any help would be wonderful, thanks!

2

Do you have nvidia devices?

ls /dev/nv*

Mine are not created automatically, I need to start X first. There are mknod commands somewhere on the forums to do this properly, but starting/stoping X has never been enough hassle for me to actually bother. :)

3

My devices are there and my permissions should be okay (I think at least):

ls -l /dev/nv*:

crw-rw---- 1 root video 195, 0 2009-06-30 11:25 /dev/nvidia0

crw-rw---- 1 root video 195, 255 2009-06-30 11:25 /dev/nvidiactl

crw-r----- 1 root kmem 10, 144 2009-06-30 07:24 /dev/nvram

4

You could also get this message if libcuda.so were not installed. Does Yast break up the nvidia drivers into different pieces like nvidia-drivers, nvidia-drivers-devel, … Maybe the CUDA libs were not installed with the default package.

5

searched for the above…

find / -iname libcuda*:

/usr/local/cuda/lib/libcudart.so.2.2

/usr/local/cuda/lib/libcudart.so

/usr/local/cuda/lib/libcudart.so.2

/usr/programs/VMD/lib/vmd/libcudart.so.2

/usr/lib64/libcuda.so.1

/usr/lib64/libcuda.so.185.18.14

I’m assuming the last 2 are the driver libraries that you were talking about? Thanks for your help though, everything seems to be in place, it just doesn’t want to work.

6

Try making a /usr/lib64/libcuda.so symbolic link which points to /usr/lib64/libcuda.so.1

N.

7

That worked, thanks Nico :) Probably never would have thought of something that simple.