WARNING: This website is obsolete! Please follow this link to get to the new Albert@Home website!

Posts by Gaurav Khanna
1) Message boards : Problems and Bug Reports : Issue with padding to power of 2 length? (Message 111669)
Posted 9 Jan 2012 by Profile Gaurav Khanna
Post:
Just noticed an odd error popping up on OpenCL ATI GPU clients:

[13:32:30][9263][ERROR] The padded time series length 12582912 isn't a power of 2 (required by OpenCL FFT)!
[13:32:30][9263][ERROR] Demodulation failed (error: 4)!

For example:
http://albert.phys.uwm.edu/result.php?resultid=85848

Regards,
Gaurav
2) Message boards : Problems and Bug Reports : Running OpenCl on Llano can it be done? (Message 111611)
Posted 21 Dec 2011 by Profile Gaurav Khanna
Post:
An update: With the the just-released 11.12 catalyst, I actually have things working on the APU! I seem to have a minor problem left to address. After the first data-set processing in a workunit is complete, it is unable to start the next one. For example see below:

http://albert.phys.uwm.edu/result.php?resultid=64181
http://albert.phys.uwm.edu/result.php?resultid=64159
http://albert.phys.uwm.edu/result.php?resultid=64137

They all crash with: Error during OpenCL kernel setup: PS (error: -5) during the processing of the second data-file AFTER completing the processing of the first data file successfully.

Comments?
3) Message boards : News : New App S6LV1 (Message 111593)
Posted 16 Dec 2011 by Profile Gaurav Khanna
Post:
Hmm. All the S6 LineVeto work units are crashing immediately for me:

http://albert.phys.uwm.edu/result.php?resultid=65949
http://albert.phys.uwm.edu/result.php?resultid=65791

4) Message boards : Problems and Bug Reports : Running OpenCl on Llano can it be done? (Message 111564)
Posted 14 Dec 2011 by Profile Gaurav Khanna
Post:
Thanks, HBE. Unfortunately, this does not appear to be the problem, it seems. I am planning to experiment with slightly older drivers (11.5 -- 11.10) to see if anything by chance just works out ..

Will report back soon.
5) Message boards : Problems and Bug Reports : Running OpenCl on Llano can it be done? (Message 111556)
Posted 13 Dec 2011 by Profile Gaurav Khanna
Post:
I've been trying on an AMD Fusion E350 box too. I have the latest drivers; updated to the AMD SDK 2.5 and have 1GB of VRAM. BOINC is working fine for me and sending the GPU work.

However the OpenCL ATI BRP4 application crashes for me each time with this output:

http://albert.phys.uwm.edu/result.php?resultid=61097

[15:42:24][2052][ERROR] Application caught signal 11.

------> Obtained 3 stack frames for this thread.
------> Backtrace:
Frame 3:
Binary file: ../../projects/albert.phys.uwm.edu/einsteinbinary_BRP4_1.19_i686-pc-linux-gnu__atiOpenCL (0x80b6ae2)
Offset info: pthread_mutex_lock+0x64a
Source file: erp_boinc_wrapper.cpp (Function: sighandler / Line: 166)
Frame 2:
Binary file: /opt/AMDAPP/lib/x86/libamdocl32.so (0xf6054199)
Offset info: +0xae199
Frame 1:
Binary file: /opt/AMDAPP/lib/x86/libamdocl32.so (0xf6054199)
Offset info: +0xae199
------> End of backtrace

15:42:24 (2052): called boinc_finish
6) Message boards : News : Sending work (Message 111531)
Posted 9 Dec 2011 by Profile Gaurav Khanna
Post:
Thanks Oliver. That was helpful. Figured it out .. it was a 64-bit Vs 32-bit issue. The ATI app is 32-bit, but the BOINC client I built is 64-bit. So, I had to make both 32 and 64 bit OpenCL libs available in my library path.

Its working now (finished a result successfully):
http://albert.phys.uwm.edu/result.php?resultid=54521
7) Message boards : News : Sending work (Message 111517)
Posted 8 Dec 2011 by Profile Gaurav Khanna
Post:
What exactly does BOINC do to check for OpenCL?
8) Message boards : News : Sending work (Message 111516)
Posted 8 Dec 2011 by Profile Gaurav Khanna
Post:
Weird .. with BOINC 6.13 the same ATI GPU completed scores of work units:

For example:
http://albert.phys.uwm.edu/result.php?resultid=53182

And I use the ATI GPU for OpenCL development all the time ..
9) Message boards : News : Sending work (Message 111511)
Posted 8 Dec 2011 by Profile Gaurav Khanna
Post:
Since the upgrade to 7.0.2 I'm not getting any work for the GPUs ..

08-Dec-2011 07:20:02 [Albert@Home] Sending scheduler request: To fetch work.
08-Dec-2011 07:20:02 [Albert@Home] Requesting new tasks for ATI
08-Dec-2011 07:20:05 [Albert@Home] Scheduler request completed: got 0 new tasks
08-Dec-2011 07:20:05 [Albert@Home] No tasks sent
08-Dec-2011 07:58:09 [Albert@Home] Sending scheduler request: To fetch work.
08-Dec-2011 07:58:09 [Albert@Home] Requesting new tasks for NVIDIA
08-Dec-2011 07:58:11 [Albert@Home] Scheduler request completed: got 0 new tasks
08-Dec-2011 07:58:11 [Albert@Home] No tasks sent

host details are here:
http://albert.phys.uwm.edu/show_host_detail.php?hostid=1396

Any thoughts?





This material is based upon work supported by the National Science Foundation (NSF) under Grant PHY-0555655 and by the Max Planck Gesellschaft (MPG). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the investigators and do not necessarily reflect the views of the NSF or the MPG.

Copyright © 2024 Bruce Allen for the LIGO Scientific Collaboration