GNOME Bugzilla – Bug 617941
Process /usr/libexec/tracker-miner-fs was killed by signal 11
Last modified: 2011-01-03 01:47:18 UTC
Created attachment 160450 [details] Crash backtace Another crash in tracker detected by Fedora's Abrt tool (https://bugzilla.redhat.com/show_bug.cgi?id=589602). Relevant info in the original bug include; ... cmdline: /usr/libexec/tracker-miner-fs component: tracker crash_function: magazine_chain_pop_head executable: /usr/libexec/tracker-miner-fs global_uuid: 9d664742fb55284ccc0dbee8173a4ed0a5aa0b5b kernel: 2.6.33.3-79.fc13.x86_64 package: tracker-0.8.4-1.fc13 rating: 3 reason: Process /usr/libexec/tracker-miner-fs was killed by signal 11 (SIGSEGV) release: Fedora release 13 (Goddard) How to reproduce: Nothing special. But i know it was crawling a shared samba dir
At first glance it either looks to me like a memory corruption, or a GIO bug. On the crashing thread there is nothing but GIO functions, and I find it quite suspicious that it crashes on allocating memory (Usually a symptom of memory corruption) Can you ask the reporter about running tracker-miner-fs under valgrind? $ G_SLICE=always_malloc valgrind --leak-check=full --num-callers=30 --log-file=valgrind.log ./tracker-miner-fs -v 3 and come back with the valgrind.log file?
Created attachment 160978 [details] Valgrind log The requested valgrind log file.
Reopening as the valgring log has been provided.
(In reply to comment #2) > Created an attachment (id=160978) [details] > Valgrind log > > The requested valgrind log file. This valgrind log doesn't show any invalid read/write error. Was the log obtained while actually reproducing the crash?
Closing this bug report as no further information has been provided. Please feel free to reopen this bug if you can provide the information asked for. Thanks!