Advanced search

Message boards : Number crunching : SIGSEGV, but still Validated with Credit?

Author Message
Jacob Klein
Send message
Joined: 11 Oct 08
Posts: 1127
Credit: 1,901,927,545
RAC: 0
Level
His
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 33303 - Posted: 1 Oct 2013 | 13:56:51 UTC
Last modified: 1 Oct 2013 | 13:57:46 UTC

Is it normal for a task to say that it encountered "SIGSEGV: segmentation violation", but still be status "Valid" with credit granted?
Surely this is a problem, isn't it?

Note: This isn't my computer - I was just looking at other users' results for my recently-errored tasks. (I am computer 153764)

See:
http://www.gpugrid.net/result.php?resultid=7304027

Name I66R8-NATHAN_KIDKIXc22_6-9-50-RND7714_0
Workunit 4795185
Created 23 Sep 2013 | 11:32:41 UTC
Sent 24 Sep 2013 | 9:39:39 UTC
Received 30 Sep 2013 | 20:00:24 UTC
Server state Over
Outcome Success
Client state Done
Exit status 0 (0x0)
Computer ID 158149
Report deadline 29 Sep 2013 | 9:39:39 UTC
Run time 148,550.97
CPU time 99,086.94
Validate state Valid
Credit 98,300.00
Application version Long runs (8-12 hours on fastest card) v8.03 (cuda55)
Stderr output

<core_client_version>7.0.65</core_client_version>
<![CDATA[
<stderr_txt>
SIGSEGV: segmentation violation
Stack trace (23 frames):
../../projects/www.gpugrid.net/acemd.800-55.bin(boinc_catch_signal+0x4d)[0x55bf0d]
/lib64/libpthread.so.0(+0xf1f0)[0x7f93aabb61f0]
/lib64/libc.so.6(+0x37f11)[0x7f93a6e9ef11]
/lib64/libc.so.6(+0x37fe5)[0x7f93a6e9efe5]
../../projects/www.gpugrid.net/acemd.800-55.bin[0x5580e9]
../../projects/www.gpugrid.net/acemd.800-55.bin[0x55822b]
/lib64/libpthread.so.0(+0xf1f0)[0x7f93aabb61f0]
/usr/lib64/libcuda.so.1(+0x1d1570)[0x7f93a9e82570]
/usr/lib64/libcuda.so.1(+0x1d2873)[0x7f93a9e83873]
/usr/lib64/libcuda.so.1(+0x1d6e68)[0x7f93a9e87e68]
/usr/lib64/libcuda.so.1(+0x1d294e)[0x7f93a9e8394e]
/usr/lib64/libcuda.so.1(+0x1d4c8f)[0x7f93a9e85c8f]
/usr/lib64/libcuda.so.1(+0x1b4178)[0x7f93a9e65178]
/usr/lib64/libcuda.so.1(+0x1c9104)[0x7f93a9e7a104]
/usr/lib64/libcuda.so.1(+0xe8675)[0x7f93a9d99675]
/usr/lib64/libcuda.so.1(cuLaunchKernel+0xcc)[0x7f93a9d7e36c]
../../projects/www.gpugrid.net/acemd.800-55.bin[0x56f8d1]
../../projects/www.gpugrid.net/acemd.800-55.bin[0x47627d]
../../projects/www.gpugrid.net/acemd.800-55.bin[0x423d56]
../../projects/www.gpugrid.net/acemd.800-55.bin[0x408d98]
../../projects/www.gpugrid.net/acemd.800-55.bin[0x407de3]
/lib64/libc.so.6(__libc_start_main+0xf5)[0x7f93a6e88a15]
../../projects/www.gpugrid.net/acemd.800-55.bin[0x407c79]

Exiting...
18:27:46 (23534): No heartbeat from core client for 30 sec - exiting
# Time per step (avg over 175000 steps): 21.393 ms
# Approximate elapsed time for entire WU: 106963.315 s
01:20:26 (29700): called boinc_finish

</stderr_txt>
]]>

Post to thread

Message boards : Number crunching : SIGSEGV, but still Validated with Credit?

//