Advanced search

Message boards : Graphics cards (GPUs) : Computation Error - Linux - 185.18.08

Author Message
JG
Send message
Joined: 28 Jun 09
Posts: 7
Credit: 0
RAC: 0
Level

Scientific publications
wat
Message 11246 - Posted: 22 Jul 2009 | 12:41:43 UTC

Hello,

I am having a problem with running GPUGRID on my Linux box. Whenever the program runs after downloading the required files the there is immediately a computation error. I am running drivers 185.18.08. I have no idea what is causing it.

GPUGRID finds my CUDA cards


    Tue 21 Jul 2009 13:39:31 BST||Processor: 4 GenuineIntel Intel(R) Core(TM)2 Quad CPU Q6700 @ 2.66GHz [Family 6 Model 15 Stepping 11]
    Tue 21 Jul 2009 13:39:31 BST||Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx lm constant_tsc arch_perfmon pebs bts rep_good pni monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr lahf_lm
    Tue 21 Jul 2009 13:39:31 BST||OS: Linux: 2.6.24-22-generic
    Tue 21 Jul 2009 13:39:31 BST||Memory: 3.86 GB physical, 7.45 GB virtual
    Tue 21 Jul 2009 13:39:31 BST||Disk: 73.92 GB total, 6.35 GB free
    Tue 21 Jul 2009 13:39:31 BST||Local time is UTC +1 hours
    Tue 21 Jul 2009 13:39:31 BST||Not using a proxy
    Tue 21 Jul 2009 13:39:31 BST||CUDA devices found
    Tue 21 Jul 2009 13:39:31 BST||Coprocessor: GeForce 9800 GTX/9800 GTX+ (2)
    Tue 21 Jul 2009 13:39:31 BST|GPUGRID|URL: http://www.gpugrid.net/; Computer ID: 44559; location: (none); project prefs: default



GPUGRID has read/write access to both GPUs and I know that environment does work as I can run both nvidia CUDA demos and my own CUDA projects without a problem.

Is there any way to get a more detailed verbose from GPUGRID? The message log gives this output from one task been downloaded and run:


    Wed 22 Jul 2009 13:16:01 BST||Starting BOINC client version 6.4.5 for x86_64-pc-linux-gnu
    Wed 22 Jul 2009 13:16:01 BST||log flags: task, file_xfer, sched_ops
    Wed 22 Jul 2009 13:16:01 BST||Libraries: libcurl/7.18.0 OpenSSL/0.9.8g zlib/1.2.3.3 c-ares/1.5.1
    Wed 22 Jul 2009 13:16:01 BST||Data directory: /home/james/Downloads/BOINC
    Wed 22 Jul 2009 13:16:01 BST||Processor: 4 GenuineIntel Intel(R) Core(TM)2 Quad CPU Q6700 @ 2.66GHz [Family 6 Model 15 Stepping 11]
    Wed 22 Jul 2009 13:16:01 BST||Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx lm constant_tsc arch_perfmon pebs bts rep_good pni monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr lahf_lm
    Wed 22 Jul 2009 13:16:01 BST||OS: Linux: 2.6.24-22-generic
    Wed 22 Jul 2009 13:16:01 BST||Memory: 3.86 GB physical, 7.45 GB virtual
    Wed 22 Jul 2009 13:16:01 BST||Disk: 73.92 GB total, 5.88 GB free
    Wed 22 Jul 2009 13:16:01 BST||Local time is UTC +1 hours
    Wed 22 Jul 2009 13:16:01 BST||Not using a proxy
    Wed 22 Jul 2009 13:16:01 BST||CUDA devices found
    Wed 22 Jul 2009 13:16:01 BST||Coprocessor: GeForce 9800 GTX/9800 GTX+ (2)
    Wed 22 Jul 2009 13:16:02 BST|GPUGRID|URL: http://www.gpugrid.net/; Computer ID: 44559; location: (none); project prefs: default
    Wed 22 Jul 2009 13:16:02 BST||No general preferences found - using BOINC defaults
    Wed 22 Jul 2009 13:16:02 BST||Preferences limit memory usage when active to 1975.91MB
    Wed 22 Jul 2009 13:16:02 BST||Preferences limit memory usage when idle to 3556.65MB
    Wed 22 Jul 2009 13:16:02 BST||Preferences limit disk usage to 5.79GB
    Wed 22 Jul 2009 13:16:02 BST|GPUGRID|Sending scheduler request: Requested by project. Requesting 120960 seconds of work, reporting 0 completed tasks
    Wed 22 Jul 2009 13:16:07 BST|GPUGRID|Scheduler request completed: got 1 new tasks
    Wed 22 Jul 2009 13:16:09 BST|GPUGRID|Started download of 133-KASHIF_HIVPR_sub_so_ba2-7-LICENSE
    Wed 22 Jul 2009 13:16:09 BST|GPUGRID|Started download of 133-KASHIF_HIVPR_sub_so_ba2-7-COPYRIGHT
    Wed 22 Jul 2009 13:16:10 BST|GPUGRID|Finished download of 133-KASHIF_HIVPR_sub_so_ba2-7-LICENSE
    Wed 22 Jul 2009 13:16:10 BST|GPUGRID|Finished download of 133-KASHIF_HIVPR_sub_so_ba2-7-COPYRIGHT
    Wed 22 Jul 2009 13:16:10 BST|GPUGRID|Started download of 133-KASHIF_HIVPR_sub_so_ba2-7-133-KASHIF_HIVPR_sub_so_ba2-6-100-RND3088_1
    Wed 22 Jul 2009 13:16:10 BST|GPUGRID|Started download of 133-KASHIF_HIVPR_sub_so_ba2-7-133-KASHIF_HIVPR_sub_so_ba2-6-100-RND3088_2
    Wed 22 Jul 2009 13:16:14 BST|GPUGRID|Finished download of 133-KASHIF_HIVPR_sub_so_ba2-7-133-KASHIF_HIVPR_sub_so_ba2-6-100-RND3088_1
    Wed 22 Jul 2009 13:16:14 BST|GPUGRID|Finished download of 133-KASHIF_HIVPR_sub_so_ba2-7-133-KASHIF_HIVPR_sub_so_ba2-6-100-RND3088_2
    Wed 22 Jul 2009 13:16:14 BST|GPUGRID|Started download of 133-KASHIF_HIVPR_sub_so_ba2-7-133-KASHIF_HIVPR_sub_so_ba2-6-100-RND3088_3
    Wed 22 Jul 2009 13:16:14 BST|GPUGRID|Started download of 133-KASHIF_HIVPR_sub_so_ba2-7-pdb_file
    Wed 22 Jul 2009 13:16:16 BST|GPUGRID|Finished download of 133-KASHIF_HIVPR_sub_so_ba2-7-133-KASHIF_HIVPR_sub_so_ba2-6-100-RND3088_3
    Wed 22 Jul 2009 13:16:16 BST|GPUGRID|Started download of 133-KASHIF_HIVPR_sub_so_ba2-7-psf_file
    Wed 22 Jul 2009 13:16:17 BST|GPUGRID|Finished download of 133-KASHIF_HIVPR_sub_so_ba2-7-psf_file
    Wed 22 Jul 2009 13:16:17 BST|GPUGRID|Started download of 133-KASHIF_HIVPR_sub_so_ba2-7-par_file
    Wed 22 Jul 2009 13:16:22 BST|GPUGRID|Finished download of 133-KASHIF_HIVPR_sub_so_ba2-7-pdb_file
    Wed 22 Jul 2009 13:16:22 BST|GPUGRID|Started download of 133-KASHIF_HIVPR_sub_so_ba2-7-133
    Wed 22 Jul 2009 13:16:23 BST|GPUGRID|Finished download of 133-KASHIF_HIVPR_sub_so_ba2-7-133
    Wed 22 Jul 2009 13:16:30 BST|GPUGRID|Finished download of 133-KASHIF_HIVPR_sub_so_ba2-7-par_file
    Wed 22 Jul 2009 13:16:31 BST|GPUGRID|Starting 133-KASHIF_HIVPR_sub_so_ba2-7-100-RND3088_0
    Wed 22 Jul 2009 13:16:31 BST|GPUGRID|Starting task 133-KASHIF_HIVPR_sub_so_ba2-7-100-RND3088_0 using acemd version 664
    Wed 22 Jul 2009 13:16:32 BST|GPUGRID|Computation for task 133-KASHIF_HIVPR_sub_so_ba2-7-100-RND3088_0 finished
    Wed 22 Jul 2009 13:16:32 BST|GPUGRID|Output file 133-KASHIF_HIVPR_sub_so_ba2-7-100-RND3088_0_0 for task 133-KASHIF_HIVPR_sub_so_ba2-7-100-RND3088_0 absent
    Wed 22 Jul 2009 13:16:32 BST|GPUGRID|Output file 133-KASHIF_HIVPR_sub_so_ba2-7-100-RND3088_0_1 for task 133-KASHIF_HIVPR_sub_so_ba2-7-100-RND3088_0 absent
    Wed 22 Jul 2009 13:16:32 BST|GPUGRID|Output file 133-KASHIF_HIVPR_sub_so_ba2-7-100-RND3088_0_2 for task 133-KASHIF_HIVPR_sub_so_ba2-7-100-RND3088_0 absent
    Wed 22 Jul 2009 13:16:32 BST|GPUGRID|Output file 133-KASHIF_HIVPR_sub_so_ba2-7-100-RND3088_0_3 for task 133-KASHIF_HIVPR_sub_so_ba2-7-100-RND3088_0 absent



As you can see the project finishes immediately, there are missing output files according to the log.
The task status comes up as computation error.

Under the gpugrid folder in the boinc folder I found an executable which I presume is called by boinc for the computation. Running ldd show no missing links.


    ldd acemd_6.64_x86_64-pc-linux-gnu__cuda
    linux-vdso.so.1 => (0x00007fff3d3fe000)
    libcufft.so.2 => /usr/local/cuda/lib/libcufft.so.2 (0x00007f8b34dbb000)
    libcuda.so.1 => /usr/lib/libcuda.so.1 (0x00007f8b348ee000)
    libcudart.so.2 => /usr/local/cuda/lib/libcudart.so.2 (0x00007f8b346a7000)
    libstdc++.so.6 => /usr/lib/libstdc++.so.6 (0x00007f8b3439c000)
    libm.so.6 => /lib/libm.so.6 (0x00007f8b3411b000)
    libgcc_s.so.1 => /lib/libgcc_s.so.1 (0x00007f8b33f0d000)
    libc.so.6 => /lib/libc.so.6 (0x00007f8b33bab000)
    libdl.so.2 => /lib/libdl.so.2 (0x00007f8b339a7000)
    libpthread.so.0 => /lib/libpthread.so.0 (0x00007f8b3378b000)
    libz.so.1 => /usr/lib/libz.so.1 (0x00007f8b33574000)
    librt.so.1 => /lib/librt.so.1 (0x00007f8b3336b000)
    /lib64/ld-linux-x86-64.so.2 (0x00007f8b350d6000)



Any suggestions for how to solve this problem or how I could some further information out the program to try locate the problem would be greatly appreciated.

Thanks.

JG
Send message
Joined: 28 Jun 09
Posts: 7
Credit: 0
RAC: 0
Level

Scientific publications
wat
Message 11247 - Posted: 22 Jul 2009 | 12:45:45 UTC - in response to Message 11246.

Never mind... I have seen that is a problem with the 185.* drivers in Linux with GPUGRID. Hopefully this is something that can be sorted soon.

Post to thread

Message boards : Graphics cards (GPUs) : Computation Error - Linux - 185.18.08

//