right - in full screen - "kill -9" is the new pink.
but in "large viewer" - everything works fine (for me).

> .. I'll try to see if can gather some more info and get a
bugzilla report submitted to the Mozilla team.

I'm skeptical due to adobe sending all dev work to google folks.(Pepper).



On Wed, Nov 14, 2012 at 2:55 PM, David Sommerseth <da...@sommerseths.net>wrote:

> On 14/11/12 20:32, Andrew Z wrote:
> > David,
> >  we are running the same mozilla/adobe setup. Unlike you i don't have
> > anything captured in xsession-errors file.
> > If you run the same HD (720p) in the "large player" mode in youtube - it
> > (probably) won't crash.
>
> That's correct!  But I don't have a crash, I have a hard freeze, and I
> can not even escape the fullscreen mode normally and have to kill the
> xulrunner/plugin-container.  And this freeze only happens in full screen.
>
> > .. //frustration
>
> Yep! Agreed.  I'll try to see if can gather some more info and get a
> bugzilla report submitted to the Mozilla team.
>
>
> kind regards,
>
> David Sommerseth
>
>
> > On Wed, Nov 14, 2012 at 10:55 AM, David Sommerseth
> > <da...@sommerseths.net <mailto:da...@sommerseths.net>> wrote:
> >
> >     On 14/11/12 14:44, Andrew Z wrote:
> >
> >         David,
> >           i'm using the one from adobe. See here:
> >         yum repolist
> >         <skip>
> >         repo id            repo name
> >                   status
> >         Dropbox            Dropbox Repository
> >                        4
> >         adobe-linux-x86_64 Adobe Systems Incorporated
> >                        2
> >
> >         installing:
> >           flash-plugin    x86_64    11.2.202.251-release
> >          adobe-linux-x86_64    6.9 M
> >
> >         and it's the one i'm bi%% heartlessly about.
> >         can you try this one
> >         (
> http://www.youtube.com/watch?__v=NBO9z4Hy9rI&feature=share&__list=PLB7667BC178F2658F
> >         <
> http://www.youtube.com/watch?v=NBO9z4Hy9rI&feature=share&list=PLB7667BC178F2658F
> >
> >         <
> http://www.youtube.com/watch?__v=NBO9z4Hy9rI&feature=share&__list=PLB7667BC178F2658F
> >         <
> http://www.youtube.com/watch?v=NBO9z4Hy9rI&feature=share&list=PLB7667BC178F2658F
> >>)
> >
> >         in HD full screen for say 15 minutes? Works fine when in
> >         anything but full
> >         screen for me.
> >
> >
> >     I'm using these packages:
> >
> >     flash-plugin-11.2.202.251-__release.x86_64
> >     firefox-10.0.10-1.el6_3.x86_64
> >
> >     This is an up-to-date SL6.3 x86_64 system with sl and sl6x repos
> >     enabled.
> >
> >     I tested that youtube video in fullscreen and it ran without any
> >     issues for about 30 minutes in 480p mode with 1920x1200 resolution.
> >      But with 720p mode, it halts completely (audio and video) on my
> >     laptop within 5 minutes - but only with certain circumstances.  I
> >     then found this error in
> >     ~/.xsession-errors:
> >
> >     WARNING: pipe error (67): Connection reset by peer: file
> >
> /builddir/build/BUILD/__xulrunner-10.0.10/mozilla-__esr10/ipc/chromium/src/chrome/__common/ipc_channel_posix.cc,
> >     line 419
> >
> >     Checking it with gdb, attaching to the xulrunner-2/plugin-container
> >     process, I see something which looks like a locking issue - probably
> >     caused buffer underflows due to network issues.
> >
> >     (gdb) bt
> >     #0  __lll_lock_wait () at
> >     ../nptl/sysdeps/unix/sysv/__linux/x86_64/lowlevellock.S:__136
> >     #1  0x00007fcebdc463be in _L_lock_995 () from /lib64/libpthread.so.0
> >     #2  0x00007fcebdc46326 in __pthread_mutex_lock
> >     (mutex=0x7fcebcc0dd70) at pthread_mutex_lock.c:101
> >     #3  0x0000003761c23f79 in PR_Lock (lock=0x7fcebcc0dd70) at
> >     ../../../mozilla/nsprpub/pr/__src/pthreads/ptsynch.c:174
> >     #4  0x0000003e12982f56 in Lock (this=0x7fcebcc1b840) at
> >     ../../dist/include/mozilla/__Mutex.h:106
> >     #5  Lock (this=0x7fcebcc1b840) at
> >     ../../dist/include/mozilla/__Monitor.h:70
> >     #6  MonitorAutoLock (this=0x7fcebcc1b840) at
> >     ../../dist/include/mozilla/__Monitor.h:125
> >     #7  mozilla::ipc::RPCChannel::__OnMaybeDequeueOne
> >     (this=0x7fcebcc1b840) at
> >
> /usr/src/debug/xulrunner-10.0.__10/mozilla-esr10/ipc/glue/__RPCChannel.cpp:402
> >     #8  0x0000003e12a308b2 in MessageLoop::RunTask (this=0x7fffb048c200,
> >     task=0x7fceaefae6d0) at
> >
> /usr/src/debug/xulrunner-10.0.__10/mozilla-esr10/ipc/chromium/__src/base/message_loop.cc:318
> >     #9  0x0000003e12a32bd6 in MessageLoop::__DeferOrRunPendingTask
> >     (this=<value optimized out>, pending_task=<value optimized out>)
> >         at
> >
> /usr/src/debug/xulrunner-10.0.__10/mozilla-esr10/ipc/chromium/__src/base/message_loop.cc:326
> >     #10 0x0000003e12a32fa7 in MessageLoop::DoWork (this=0x7fffb048c200)
> >     at
> >
> /usr/src/debug/xulrunner-10.0.__10/mozilla-esr10/ipc/chromium/__src/base/message_loop.cc:426
> >     #11 0x0000003e12a4352a in
> >     base::MessagePumpForUI::__RunWithDispatcher (this=0x7fcebcc300a0,
> >     delegate=<value optimized out>, dispatcher=<value optimized out>)
> >         at
> >
> /usr/src/debug/xulrunner-10.0.__10/mozilla-esr10/ipc/chromium/__src/base/message_pump_glib.cc:__199
> >     #12 0x0000003e12a30852 in RunHandler (this=<value optimized out>) at
> >
> /usr/src/debug/xulrunner-10.0.__10/mozilla-esr10/ipc/chromium/__src/base/message_loop.cc:201
> >     #13 MessageLoop::Run (this=<value optimized out>) at
> >
> /usr/src/debug/xulrunner-10.0.__10/mozilla-esr10/ipc/chromium/__src/base/message_loop.cc:175
> >     #14 0x0000003e11ff90fd in XRE_InitChildProcess (aArgc=6,
> >     aArgv=<value optimized out>, aProcess=<value optimized out>)
> >         at
> >
> /usr/src/debug/xulrunner-10.0.__10/mozilla-esr10/toolkit/xre/__nsEmbedFunctions.cpp:524
> >     #15 0x0000000000401314 in main (argc=7, argv=<value optimized out>)
> >     at
> >
> /usr/src/debug/xulrunner-10.0.__10/mozilla-esr10/ipc/app/__MozillaRuntimeMain.cpp:81
> >     (gdb)
> >
> >     I also noticed that if youtube could cache up more data (you know
> >     the grey shading in front of the red progress marker) ... This issue
> >     did not happen. So I'd expect there to be an issue in xulrunner, not
> >     the flash plug-in itself and only when there is a lag in the data
> >     stream.
> >
> >     In 480p mode I saw buffer-underflows as well - but those times it
> >     would just lag for less than 0.5 second and then continue.
> >
> >
> >     kind regards,
> >
> >     David Sommerseth
> >
>

Reply via email to