| Summary: | ktorrent still crashes when used for a few hours at a time | ||
|---|---|---|---|
| Product: | [Applications] ktorrent | Reporter: | Richard Hartmann <richih-kde> |
| Component: | general | Assignee: | Joris Guisson <joris.guisson> |
| Status: | RESOLVED FIXED | ||
| Severity: | crash | CC: | axela74, chriswhy, cwgundlach4, gordey, lashloe, mail, marcos.caceres, mastercactapus, muziofg, niky.el45, vvasaitis, zorael |
| Priority: | NOR | ||
| Version: | unspecified | ||
| Target Milestone: | --- | ||
| Platform: | Debian unstable | ||
| OS: | Linux | ||
| Latest Commit: | Version Fixed In: | ||
| Sentry Crash Report: | |||
| Attachments: | New crash information added by DrKonqi | ||
|
Description
Richard Hartmann
2010-06-18 01:48:28 UTC
Post the log file from the crash. It is probably a std::bad_alloc exception, caused by running out of memory. "Log file" as in "start on the shell and look at whatever comes out of STDOUT"? If not, where can I find the log? In ~/.kde/share/apps/ktorrent/ there are files named log-1.gz, log-2.gz ... These are the log files of the last 10 sessions, there is also a file named log, which is the log file of the current session (or last if KT isn't running) The last second before the crash, log privacy scraped, but consistant. (D keeps on being D, etc) Sat Jun 19 10:55:39 2010: Authentication to A : failure Sat Jun 19 10:55:39 2010: Error thrown: Illegal token: 65 Sat Jun 19 10:55:39 2010: Error happened during parsing : Illegal token: 65 Sat Jun 19 10:55:39 2010: Timeout occurred Sat Jun 19 10:55:39 2010: Authentication to B : failure Sat Jun 19 10:55:39 2010: Doing scrape request to url : C Sat Jun 19 10:55:39 2010: Accepted connection from [::ffff:D]:60982 Sat Jun 19 10:55:39 2010: DHT: Sending ping response Sat Jun 19 10:55:39 2010: Error thrown: Illegal token: 17 Sat Jun 19 10:55:39 2010: Error happened during parsing : Illegal token: 17 Sat Jun 19 10:55:39 2010: Authentication(S) to D : failure Sat Jun 19 10:55:39 2010: Initiating connection to E via (UTP) Sat Jun 19 10:55:39 2010: Request timed out 3089 49152 16384 Transmission 1.5.1.0 Sat Jun 19 10:55:39 2010: Initiating connection to F via (UTP) Sat Jun 19 10:55:39 2010: Initiating connection to G via (UTP) Sat Jun 19 10:55:39 2010: Timeout occurred Sat Jun 19 10:55:39 2010: Authentication to H : failure Sat Jun 19 10:55:39 2010: Initiating connection to H via (TCP) Sat Jun 19 10:55:39 2010: Timeout occurred Sat Jun 19 10:55:39 2010: Authentication to I : failure Sat Jun 19 10:55:39 2010: Uncaught exception: std::bad_alloc Sat Jun 19 12:35:57 2010: Uncaught exception: std::bad_alloc Seem to stay the same. How much memory was KT using before the crash ? I honestly don't think total RAM is a problem (though this is i386, not x64):
richih@adamantium ~ % free -m
total used free shared buffers cached
Mem: 3295 3078 216 0 0 2136
-/+ buffers/cache: 942 2353
Swap: 4769 0 4768
richih@adamantium ~ %
I can run ps aux in a loop and redirect the output to a log file if you want.
*** Bug 242362 has been marked as a duplicate of this bug. *** *** Bug 242741 has been marked as a duplicate of this bug. *** *** Bug 242613 has been marked as a duplicate of this bug. *** Created attachment 48328 [details]
New crash information added by DrKonqi
using ktorrent for a few hours results in crash
*** Bug 241570 has been marked as a duplicate of this bug. *** Is there any ETA on a fix? This issue makes ktorrent almost useless. Is there any ETA on a fix? This issue makes ktorrent almost useless. Yesterday, I'm add ktorrent in cron for run every 20 minutes. I see crash-message only one. May be, kernel of ktorrent collect date in background and they call crash, but if you restore ktorrent's window this date reset. > Is there any ETA on a fix? This issue makes ktorrent almost useless.
I'm working on it, it's an annoying hard to track down memory leak of some sort, turn of µTP in the mean time.
If I can help tracking it down, just tell me :) Thanks for the hint with uTP; I didn't know that uTP caused this. I will report back if it crashes again. I'm fairly certain now what the problem is, will have to come up with a more memory efficient delay window. SVN commit 1145357 by guisson: Fix bug with sequence number wrapping around in µTP causing bad_alloc's BUG: 242042 M +1 -0 ChangeLog M +13 -4 src/utp/connection.cpp M +12 -7 src/utp/localwindow.cpp M +29 -0 src/utp/utpprotocol.h WebSVN link: http://websvn.kde.org/?view=rev&revision=1145357 *** Bug 243784 has been marked as a duplicate of this bug. *** Are there any plans to release 4.0.2 now that the bug is (hopefully) squashed? That would reduce your workload due to duplicate bug reports, too :) I was planning to tag the release today, and release it today or tomorrow (depends on how much time I have) *** Bug 243937 has been marked as a duplicate of this bug. *** *** Bug 245392 has been marked as a duplicate of this bug. *** *** Bug 245853 has been marked as a duplicate of this bug. *** *** Bug 246494 has been marked as a duplicate of this bug. *** *** Bug 248868 has been marked as a duplicate of this bug. *** |