Re: [Ekiga-devel-list] Win32 master crash on hang up

[Damien Sandras <dsandras seconix com>]

Le jeudi 29 octobre 2009 à 19:04 +0100, Michael Rickmann a écrit :
> Damien Sandras schrieb:
> > Le dimanche 25 octobre 2009 à 22:56 +0100, Michael Rickmann a écrit :
> >> I got this under XP-SP2 when calling 500 ekiga net after hanging up. For 
> >> the attached gdb backtrace I did a "bt" then "thread apply all bt", not 
> >> very meaningful I think. The d4 output I do not dare to attach, first 
> >> normal then
> >>
> >> 2009/10/25 22:18:35.750	  2:10.605	       Media Patch:1892 
> >> AudioOutputCore	Falling back to SILENT (Ekiga/Ekiga)
> >> 2009/10/25 22:18:35.750	  2:10.605	       Media Patch:1892 
> >> AudioOutputCore	Tried to set unexisting primary device SILENT (Ekiga/Ekiga)
> >> 2009/10/25 22:18:35.750	  2:10.605	       Media Patch:1892 
> >> AudioOutputCore	Falling back to SILENT (Ekiga/Ekiga)
> >> 2009/10/25 22:18:35.750	  2:10.605	       Media Patch:1892 
> >> AudioOutputCore	Tried to set unexisting primary device SILENT (Ekiga/Ekiga)
> >>
> >> interleaved and in the end above messages for about 6 M pure.
> >> This seems to be related to 031069d8eb 2009-10-15 18:07:15 and the 
> >> trouble we had in stable 
> >> https://bugzilla.gnome.org/show_bug.cgi?id=586531#c6 , comment #16.
> >> Except of that I am sure to have compiled a debug version.
> > 
> > I possibly fixed that one.
> > 
> > Can you check ?
> > 
> > Thanks !
> > 
> It is fixed !! I observed a heap instability at the same point of Win32 
> Ekiga, but that should be separated. Thank you.
> Nevertheless, I would like to build the Win32 debug version with the 
> same principle code base as the release one. Can we get rid of the 
> silent devices in debug and release?
Well, release would require a patch in that case. I would recommend against doing it.

---

	Damien Sandras
	http://www.ekiga.org http://www.beip.be

---