GNOME Bugzilla – Bug 332214
Rhythmbox Freezes
Last modified: 2006-03-08 20:06:59 UTC
Steps to reproduce: 1. Open a huge library 2. Start playing 3. Keep jumping songs without leting them finish 4. It should crash sooner or later Stack trace: Parent process: Process 9044 attached - interrupt to quit futex(0x86bbfd0, FUTEX_WAIT, 2, NULL Child process: Process 9096 attached - interrupt to quit poll([{fd=26, events=POLLIN|POLLERR|POLLHUP}], 1, 500) = 0 semop(557056, 0x7ffe43a8, 2) = 0 shmctl(6324240, IPC_64|IPC_STAT, 0x7ffe4354) = 0 semop(557056, 0x7ffe43b4, 1) = 0 poll([{fd=26, events=POLLIN|POLLERR|POLLHUP}], 1, 500) = 0 semop(557056, 0x7ffe43a8, 2) = 0 shmctl(6324240, IPC_64|IPC_STAT, 0x7ffe4354) = 0 semop(557056, 0x7ffe43b4, 1) = 0 poll([{fd=26, events=POLLIN|POLLERR|POLLHUP}], 1, 500) = 0 semop(557056, 0x7ffe43a8, 2) = 0 shmctl(6324240, IPC_64|IPC_STAT, 0x7ffe4354) = 0 semop(557056, 0x7ffe43b4, 1) = 0 [repeat ad nauseum] Other information: Running on Ubuntu Dapper 6.04, not sharing songs, AudioScrobbler configured correctly.
It just happened again and I attached gdb on the child. All I got was this (not sure if it is useful, but...): 0xffffe410 in __kernel_vsyscall () (gdb) n Single stepping until exit from function __kernel_vsyscall, which has no line number information. [Switching to Thread 1994180288 (LWP 9794)] 0x77036904 in poll () from /lib/tls/i686/cmov/libc.so.6 (gdb) Single stepping until exit from function poll, which has no line number information. 0x76f3532c in snd_pcm_direct_check_interleave () from /usr/lib/libasound.so.2 (gdb) Single stepping until exit from function snd_pcm_direct_check_interleave, which has no line number information.
Can you get a full stack trace for all threads ('thread apply all bt' in gdb)? This may be the same problem as reported in bug #330748.
unfortunately, Ubuntu packages are stripped and I can't get a full trace, but... Parent process: 0xffffe410 in __kernel_vsyscall () (gdb) where
+ Trace 66415
And rhythmbox is spawning only a parent and a child process, nothing more. I would probably mark this bug as duplicate, anyway. :)
I'm not sure where you're getting this idea of parent and child processes from. Those stack traces are identical, and they're not from the main thread. Please use 'thread apply all bt', as without debug info, I can still determine whether this is the same problem as bug 330748 or not.
Unfortunatelly (or maybe fortunatelly), since last night update, I can't make Rhythmbox freeze again, no matter how many jumps I do. I think you can close the bug. (Oh, btw: I called parent the process that appeared on 'ps -ef' with '1' as parent pid and child the process that had the first process pid as parent. And maybe I did a bad copy'n'paste of the gdb on the previous post -- my bad)
Please reopen the bug (and include stack traces for all threads) if you see this again.
Ok, it happened again. Again, I had two processes pointed on 'ps -ef', and I did a "thread apply all bt" on both process: The first process dropped this: (gdb) thread apply all bt
+ Trace 66777
This is the same problem as reported in bug 330748. *** This bug has been marked as a duplicate of 330748 ***