Arch Linux

Please read this before reporting a bug:
https://wiki.archlinux.org/title/Bug_reporting_guidelines

Do NOT report bugs when a package is just outdated, or it is in the AUR. Use the 'flag out of date' link on the package page, or the Mailing List.

REPEAT: Do NOT report bugs for outdated packages!
Tasklist

FS#23913 - orc 0.4.13 update breaks gstreamer

Attached to Project: Arch Linux
Opened by Hussam Al-Tayeb (hussam) - Sunday, 24 April 2011, 13:49 GMT
Last edited by Ionut Biru (wonder) - Saturday, 30 April 2011, 15:12 GMT
Task Type Bug Report
Category Packages: Extra
Status Closed
Assigned To Ionut Biru (wonder)
Architecture i686
Severity Medium
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

after upgrading from orc 0.4.11 to 0.4.13, some gstreamer applications (clementine in particular) segfault when playing music. At first I didn't know what caused this
A quick search on google showed other gstreamer applications are affected by this and that it's orc's fault. I downgraded to orc 0.4.11 and the segfaults stopped.
This task depends upon

Closed by  Ionut Biru (wonder)
Saturday, 30 April 2011, 15:12 GMT
Reason for closing:  Fixed
Additional comments about closing:  orc 0.4.14
Comment by Ionut Biru (wonder) - Sunday, 24 April 2011, 14:41 GMT
can you try to see if a rebuild fix this problem?

depending on what packages you have, this packages needs a rebuild:
gstreamer0.10-bad
gstreamer0.10-bad-plugins
gstreamer0.10-base
gstreamer0.10-ffmpeg
gstreamer0.10-good
gstreamer0.10-ugly-plugins
Comment by Hussam Al-Tayeb (hussam) - Sunday, 24 April 2011, 17:21 GMT
Ok, I'm rebuilding now starting by gstreamer0.10-base. I will let you know as soon as they are done.
Comment by Hussam Al-Tayeb (hussam) - Sunday, 24 April 2011, 18:41 GMT
Ok, I rebuilt all the gstreamer0.10 packages and double checked from the dates of the files in /usr/lib/gstreamer-0.10/ that everything was rebuilt.
The same problem happens.
Comment by Ionut Biru (wonder) - Sunday, 24 April 2011, 18:50 GMT
it will be nice to build orc with debug symbols and run clementine or other application in gdb and see if it really crash with an orc trace.

https://wiki.archlinux.org/index.php/Debug_-_Getting_Traces#One_package_settings_only
Comment by Hussam Al-Tayeb (hussam) - Sunday, 24 April 2011, 18:53 GMT
Even after the gstreamer rebuild, downgrading again to orc 0.4.11 stops the segfaults
Comment by Hussam Al-Tayeb (hussam) - Sunday, 24 April 2011, 18:53 GMT
ok, I'll try that now.
Comment by Hussam Al-Tayeb (hussam) - Sunday, 24 April 2011, 21:19 GMT
Sorry it's taking time. I rebuilt orc with debug symbols. I'm doing the same with gstreamer0.10 and gstreamer0.10-base now too.

btw, I found this
http://blog.doliver.co.uk/2011/03/fixing-clementine-and-other-music-players-crashing-in-ubuntu-10-10-maverick-meerkat-gstreamer/
Comment by Hussam Al-Tayeb (hussam) - Sunday, 24 April 2011, 21:28 GMT
(gdb) bt full
#0 0xa95fc9f4 in ?? ()
No symbol table info available.
#1 0x00000900 in ?? ()
qt_meta_data_QSQLiteDriver = {5, 0 <repeats 14 times>}
qt_meta_stringdata_QSQLiteDriver = "QSQLiteDriver\000"
QSQLiteDriver::staticMetaObject = {d = {superdata = 0x8b165e0,
stringdata = 0x89e425c "QSQLiteDriver", data = 0x89e4220, extradata = 0x0}}
#2 0x00000000 in ?? ()
No symbol table info available.
(gdb) info threads
19 Thread 0xa8af5b70 (LWP 7818) 0xb7fdf424 in __kernel_vsyscall ()
* 18 Thread 0xa934bb70 (LWP 7817) 0xa95fc9f4 in ?? ()
17 Thread 0xa9e39b70 (LWP 7816) 0xb7fdf424 in __kernel_vsyscall ()
16 Thread 0xaa6a3b70 (LWP 7815) 0xb7fdf424 in __kernel_vsyscall ()
15 Thread 0xab0a8b70 (LWP 7814) 0xb7fdf424 in __kernel_vsyscall ()
14 Thread 0xab8a9b70 (LWP 7813) 0xb7fdf424 in __kernel_vsyscall ()
13 Thread 0xac0aab70 (LWP 7812) 0xb7fdf424 in __kernel_vsyscall ()
12 Thread 0xac8fab70 (LWP 7809) 0xb7fdf424 in __kernel_vsyscall ()
11 Thread 0xad0fbb70 (LWP 7808) 0xb7fdf424 in __kernel_vsyscall ()
10 Thread 0xad8fcb70 (LWP 7807) 0xb7fdf424 in __kernel_vsyscall ()
9 Thread 0xae0fdb70 (LWP 7806) 0xb7fdf424 in __kernel_vsyscall ()
8 Thread 0xae8feb70 (LWP 7805) 0xb7fdf424 in __kernel_vsyscall ()
7 Thread 0xaf0ffb70 (LWP 7804) 0xb7fdf424 in __kernel_vsyscall ()
6 Thread 0xafa39b70 (LWP 7803) 0xb7fdf424 in __kernel_vsyscall ()
5 Thread 0xb023ab70 (LWP 7802) 0xb7fdf424 in __kernel_vsyscall ()
4 Thread 0xb2c5bb70 (LWP 7801) 0xb7fdf424 in __kernel_vsyscall ()
2 Thread 0xb345cb70 (LWP 7799) 0xb7fdf424 in __kernel_vsyscall ()
1 Thread 0xb4544720 (LWP 7796) 0xb7fdf424 in __kernel_vsyscall ()
(gdb)


btw, 0.4.12 also causes the crash. the last working release is 0.4.11

I must be a change between 0.4.11 and 0.4.12 that causes this. I'll try to find out which but it may take time. Any help is greatly welcomed.
Comment by Hussam Al-Tayeb (hussam) - Monday, 25 April 2011, 13:16 GMT
Ok, I did git bisect on orc sources and I found the bad commit:


f769c8ab92398f110b6620ad98a1271b87e99a49 is the first bad commit
commit f769c8ab92398f110b6620ad98a1271b87e99a49
Author: David Schleef <ds@schleef.org>
Date: Sun Oct 31 13:51:47 2010 +0100

sse: Fix param order of orc_x86_emit_modrm_memoffset

:040000 040000 4ea7fcce35db1702d52931ecdc2388cc721facf7 670715eb9657699c13397e796727f0501035060c M orc


I will report this to orc developers.
Comment by Hussam Al-Tayeb (hussam) - Monday, 25 April 2011, 21:44 GMT
Ok, orc developer fixed the orc problem and the crash no longer happens. today's checkins fix the problem http://code.entropywave.com/git?p=orc.git;a=summary
Comment by Ionut Biru (wonder) - Monday, 25 April 2011, 21:48 GMT
thanks a lot for doing this. i see that he bumped the version to 0.4.13.1 and i guess he will do a release soon

Loading...