error 38 (0x26)
7 Mar 2008 14:55:07 UTC
Topic 193555
(moderation:
Dear searchers,
I have installed the boinc and einstein@home on a new dell laptop. I get two work units that run for 10-20% and then die with error code 0x26. How could I solve this very unfortunate behavior? I browsed the net for a possible explanation of the problem, however out of luck. The log of the last few hours of computation is attached. Thanks!
Fri Mar 7 12:55:27 2008||Starting BOINC client version 5.4.11 for i686-pc-linux-gnu
Fri Mar 7 12:55:27 2008||libcurl/7.15.5 OpenSSL/0.9.8c zlib/1.2.3 libidn/0.6.5
Fri Mar 7 12:55:27 2008||Data directory: /var/lib/boinc-client
Fri Mar 7 12:55:27 2008||Processor: 2 GenuineIntel Intel(R) Core(TM)2 Duo CPU T7500 @ 2.20GHz
Fri Mar 7 12:55:27 2008||Memory: 1.98 GB physical, 1.99 GB virtual
Fri Mar 7 12:55:27 2008||Disk: 19.69 GB total, 7.84 GB free
Fri Mar 7 12:55:27 2008|Einstein@Home|URL: http://einstein.phys.uwm.edu/; Computer ID: 1126481; location: home; project prefs: default
Fri Mar 7 12:55:27 2008||General prefs: from Einstein@Home (last modified 2005-03-15 11:57:20)
Fri Mar 7 12:55:27 2008||General prefs: no separate prefs for home; using your defaults
Fri Mar 7 12:55:27 2008||Reading preferences override file
Fri Mar 7 12:55:27 2008||Local control only allowed
Fri Mar 7 12:55:27 2008||Listening on port 31416
Fri Mar 7 12:55:27 2008|Einstein@Home|Resuming task h1_0887.95_S5R3__255_S5R3b_1 using einstein_S5R3 version 431
Fri Mar 7 12:55:27 2008|Einstein@Home|Resuming task h1_0887.95_S5R3__253_S5R3b_0 using einstein_S5R3 version 431
Fri Mar 7 15:02:25 2008|Einstein@Home|Unrecoverable error for result h1_0887.95_S5R3__253_S5R3b_0 (process exited with code 38 (0x26))
Fri Mar 7 15:02:25 2008|Einstein@Home|Deferring scheduler requests for 1 minutes and 0 seconds
Fri Mar 7 15:02:25 2008||Rescheduling CPU: application exited
Fri Mar 7 15:02:25 2008|Einstein@Home|Computation for task h1_0887.95_S5R3__253_S5R3b_0 finished
Fri Mar 7 15:03:26 2008|Einstein@Home|Sending scheduler request to http://einstein.phys.uwm.edu/EinsteinAtHome_cgi/cgi
Fri Mar 7 15:03:26 2008|Einstein@Home|Reason: To fetch work
Fri Mar 7 15:03:26 2008|Einstein@Home|Requesting 17280 seconds of new work, and reporting 1 completed tasks
Fri Mar 7 15:03:31 2008|Einstein@Home|Scheduler request succeeded
Fri Mar 7 15:03:33 2008||Rescheduling CPU: files downloaded
Fri Mar 7 15:03:33 2008|Einstein@Home|Starting task h1_0887.95_S5R3__221_S5R3b_1 using einstein_S5R3 version 431
Fri Mar 7 15:27:32 2008|Einstein@Home|Unrecoverable error for result h1_0887.95_S5R3__255_S5R3b_1 (process exited with code 38 (0x26))
Fri Mar 7 15:27:32 2008|Einstein@Home|Deferring scheduler requests for 1 minutes and 0 seconds
Fri Mar 7 15:27:32 2008||Rescheduling CPU: application exited
Fri Mar 7 15:27:32 2008|Einstein@Home|Computation for task h1_0887.95_S5R3__255_S5R3b_1 finished
Fri Mar 7 15:28:36 2008|Einstein@Home|Sending scheduler request to http://einstein.phys.uwm.edu/EinsteinAtHome_cgi/cgi
Fri Mar 7 15:28:36 2008|Einstein@Home|Reason: To fetch work
Fri Mar 7 15:28:36 2008|Einstein@Home|Requesting 17280 seconds of new work, and reporting 1 completed tasks
Fri Mar 7 15:28:41 2008|Einstein@Home|Scheduler request succeeded
Fri Mar 7 15:28:44 2008||Rescheduling CPU: files downloaded
Fri Mar 7 15:28:44 2008|Einstein@Home|Starting task h1_0887.95_S5R3__219_S5R3b_0 using einstein_S5R3 version 431
Fri Mar 7 15:48:33 2008|Einstein@Home|Unrecoverable error for result h1_0887.95_S5R3__219_S5R3b_0 (process exited with code 38 (0x26))
Fri Mar 7 15:48:33 2008|Einstein@Home|Deferring scheduler requests for 1 minutes and 0 seconds
Fri Mar 7 15:48:33 2008||Rescheduling CPU: application exited
Fri Mar 7 15:48:33 2008|Einstein@Home|Computation for task h1_0887.95_S5R3__219_S5R3b_0 finished
error 38 (0x26)
)
First impression, looking at the actual stderr.txt, you are getting signal 8 all the time. Signal 8 is a floating point exception. This might be a problem with your CPU's FPU.
But I also see that you're using BOINC 5.4.11, which in BOINC terms is quite old.
First things first, try to update to a newer version of BOINC. We're in the 5.10 versions at this moment. You can download the latest from the download site. Whether or not you want to try the development version is up to you, else go for the recommended version for your OS.
Check here on how to install the new version.
If that doesn't fix things though, consider checking with other programs if your FPU is still working as it should. I'm sure someone else here will know of a way to check it under Linux.
Hi! It's not all clear to
)
Hi!
It's not all clear to me what causes this, but there seems to be a correlation of theis error happenening and certain flavours of Linux kernels/libraries. For example, Gentoo distribution Linux seems to be far overrepresented compared to other distributions when it comes to this particular error.
So, is there anything special about the kernel or libraries? Did E@H work before? Was there a lib update between then and the first occurance of this error??
CU
Bikeman
I've updated boinc to 5.8.15,
)
I've updated boinc to 5.8.15, and run einstein@home over the night without any errors so far.. It seems that the update solved the issue. Thanks!