Opened 11 years ago

Closed 11 years ago

Last modified 11 years ago

#6142 closed defect (invalid)

Mythfrontend segfaults when exiting PBP

Reported by: eric.bosch@… Owned by: Isaac Richards
Priority: trivial Milestone: 0.22
Component: mythtv Version: head
Severity: low Keywords:
Cc: Ticket locked: no

Description

Mythfrontend is segfaulting when exiting PBP mode. PIP works fine. 2009-01-18 13:45:10.077 Realtime priority would require SUID as root. 2009-01-18 13:45:10.080 Video timing method: USleep with busy wait 2009-01-18 13:45:29.433 PxPTeardownView() 2009-01-18 13:45:29.454 TV: Attempting to change from WatchingLiveTV to None [Thread 0x7fa0c10fc950 (LWP 21452) exited] [Thread 0x7fa0c18fd950 (LWP 21450) exited] 2009-01-18 13:45:29.903 TV: Changing from WatchingLiveTV to None 2009-01-18 13:45:29.909 PIPRemovePlayer ok: 1 [Thread 0x7fa0c20fe950 (LWP 21449) exited] [Thread 0x7fa0c28ff950 (LWP 21448) exited] [Thread 0x7fa0c3100950 (LWP 21447) exited]

Program received signal SIGSEGV, Segmentation fault. [Switching to Thread 0x7fa0c71f1950 (LWP 21446)] 0x00007fa0c64b1170 in ?? () (gdb) bt #0 0x00007fa0c64b1170 in ?? () #1 0x00007fa0e7516da6 in glXMakeCurrentReadSGI ()

from usr/lib64/opengl/ati/lib/libGL.so.1

#2 0x00007fa0e75170b1 in glXMakeCurrent ()

from usr/lib64/opengl/ati/lib/libGL.so.1

#3 0x00007fa0e6d4beaf in OpenGLContextGLX::MakeContextCurrent? (

this=0x474e1c0, current=true) at openglcontext.cpp:1241

#4 0x00007fa0e6d4938e in OpenGLContext::MakeCurrent? (this=0x474e1c0,

current=true) at openglcontext.cpp:200

#5 0x00007fa0e6d11648 in VideoOutputXv::DeleteBuffers? (this=0x3da95b0,

subtype=XVideo, delete_pause_frame=true) at videoout_xv.cpp:2793

#6 0x00007fa0e6d2e946 in ~VideoOutputXv? (this=0x3da95b0)

at videoout_xv.cpp:220

#7 0x00007fa0e6bafe2f in NuppelVideoPlayer::OutputVideoLoop? (this=0x47717c0)

at NuppelVideoPlayer?.cpp:3017

#8 0x00007fa0e6bafeea in NuppelVideoPlayer::kickoffOutputVideoLoop (

player=0x47717c0) at NuppelVideoPlayer?.cpp:3087

#9 0x00007fa0dfb51097 in start_thread (arg=<value optimized out>)

at pthread_create.c:297

#10 0x00007fa0def2a6ad in clone ()

at ../sysdeps/unix/sysv/linux/x86_64/clone.S:112

#11 0x0000000000000000 in ?? ()

Attachments (2)

mythfrontend.log (49.2 KB) - added by eric.bosch@… 11 years ago.
Logfile of mythfrontend when segfault on exit PBP
gdb.txt (48.4 KB) - added by eric.bosch@… 11 years ago.
gdb.txt from debugging session

Download all attachments as: .zip

Change History (11)

comment:1 Changed 11 years ago by danielk

Priority: majortrivial
Severity: mediumlow
Status: newinfoneeded_new

need the rest of the backtrace

comment:2 Changed 11 years ago by eric.bosch@…

Sorry, that was all the backtrace I had available for this specific machine. I found, however that this was working fine on a similar system, just slightly different video hardware. The failure apparently was due to the fact it was using VDPAU, and was sort of working until exiting. It appears now that PBP refuses to start if rendering VDPAU.

comment:3 Changed 11 years ago by eric.bosch@…

This is still happening on an ATI HD3400. It seems to work fine on nvidia cards (8600 GTX and 9800 GTX). I will do some more work to try to get additional backtrace or debugging information and will upload it later.

Changed 11 years ago by eric.bosch@…

Attachment: mythfrontend.log added

Logfile of mythfrontend when segfault on exit PBP

comment:4 Changed 11 years ago by anonymous

The full backtrace that I am able to get is following. I also attached a mythtv -v playback log.

[Thread 0x7fdfcb4a2950 (LWP 8407) exited] [Thread 0x7fdfcaca1950 (LWP 8408) exited] [Thread 0x7fdfc949e950 (LWP 8394) exited]

Program received signal SIGSEGV, Segmentation fault. [Switching to Thread 0x7fdfcf71d950 (LWP 8406)] 0x00007fdfce9dd170 in ?? () (gdb) bt #0 0x00007fdfce9dd170 in ?? () #1 0x00007fdfed617da6 in glXMakeCurrentReadSGI ()

from usr/lib64/opengl/ati/lib/libGL.so.1

#2 0x00007fdfed6180b1 in glXMakeCurrent () from usr/lib64/opengl/ati/lib/libGL.so.1 #3 0x00007fdfece525ef in OpenGLContextGLX::MakeContextCurrent? (this=0x8ecb50,

current=true) at openglcontext.cpp:1241

#4 0x00007fdfece4face in OpenGLContext::MakeCurrent? (this=0x8ecb50, current=true)

at openglcontext.cpp:200

#5 0x00007fdfece22f50 in VideoOutputXv::DeleteBuffers? (this=0x4610dc0, subtype=XVideo,

delete_pause_frame=true) at videoout_xv.cpp:2793

#6 0x00007fdfece3e71c in ~VideoOutputXv? (this=0x4610dc0) at videoout_xv.cpp:220 #7 0x00007fdfeccd14af in NuppelVideoPlayer::OutputVideoLoop? (this=0x4799350)

at NuppelVideoPlayer?.cpp:3017

#8 0x00007fdfeccd156a in NuppelVideoPlayer::kickoffOutputVideoLoop (player=0x4799350)

at NuppelVideoPlayer?.cpp:3087

#9 0x00007fdfe5e6e097 in start_thread (arg=<value optimized out>)

at pthread_create.c:297

#10 0x00007fdfe52476ad in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:112 #11 0x0000000000000000 in ?? ()

comment:5 Changed 11 years ago by Stuart Auchterlonie

Status: infoneeded_newnew

I'm also getting this in a number of situations. Primarily exiting playback (recording or livetv)

There's no VDPAU in my system, I run ATI with the open source drivers.

Stuart

comment:6 Changed 11 years ago by danielk

Status: newinfoneeded_new

Guys, unless you have a full backtrace, see TicketHowTo this is an invalid ticket. Either this is a threading issue, in which case we need to see where the rest of the threads are, or this is a bug in the ATI video drivers which we're not going to do anything about.

Changed 11 years ago by eric.bosch@…

Attachment: gdb.txt added

gdb.txt from debugging session

comment:7 Changed 11 years ago by eric.bosch@…

I have attached gdb.txt as directed in the TicketHowTo. The backtrace after the segfault remains the same as that posted above. If there is a way to put in a breakpoint before the segfault, please let me know the process and I will try to get a more complete backtrace.

comment:8 Changed 11 years ago by danielk

Resolution: invalid
Status: infoneeded_newclosed

Sorry, this looks like a driver bug from the info provided :[

comment:9 Changed 11 years ago by eric.bosch@…

Ok, well thanks a lot. I had a feeling that's where it was. Hopefully ATI will come out with an update soon!

Note: See TracTickets for help on using tickets.