[desktop-discuss] [osol-bugs] rhythmbox snv_b64 x86 crash

[Laszlo (Laca) Peter]

Right.  LSARC/2007/378 - "libexpat upgrade and move to /usr" was 
approved last week.  I'm working on getting these changes into the
next SFW build.  Then I will need to patch python to use the
shared libexpat instead of statically linking in a different version.
It will probably happen at least one build later than the libexpat
upgrade.

Laca

On Mon, 2007-07-02 at 09:42 +0800, Jerry Tan wrote:
> It is a known bug. already filed buy QA
> because rhythmbox has one plugin depend on python expat library.
> but the python expat library used is 1.95.8 ,  is not the same as 
> libexpat under /usr/sfw/(which is 1.95.7)
> so this bug happened.
> 
> CC: laca, he know this more clearly.
> 
> Shawn Walker wrote:
> > On 30/06/07, Liane Praza <lianep at eng.sun.com> wrote:
> >> =?UTF-8?Q?Thommy_M=2E_Malmstr=C3=B6m?= writes:
> >> > Running snv_b64 on a Ferrari 4005 importing music from CD with 
> >> Sound Recorder
> >> >  in Ogg format.
> >> > Trying to listen to it with Rhytmbox just makes it crash. I 
> >> actually plays th
> >> > e first song but then crashes. Same  if I choose mp3 files...
> >>
> >> I've been hitting something similar on builds 65 and 67.  My crash is
> >> definitely:
> >>   http://bugs.opensolaris.org/view_bug.do?bug_id=6531456
> >>
> >> Don't be fooled by the synopsis.  I've not enabled accessability nor
> >> am I running Alacarte when this happens.
> >>
> >> Unfortunately, rhythmbox avoids saving its core.  To confirm you're
> >> hitting the same bug, run rhythmbox under mdb.
> >>
> >>    $ mdb /usr/bin/rhythmbox
> >>
> >> You'll get a prompt, then type "::run".
> >>
> >> Then you'll get another prompt asking (p)arent or (c)child?  Type "p".
> >>
> >> Rhythmbox will startup, then play a song to get it to crash.  mdb will
> >> tell you it SIGSEGVed, and you'll get a prompt again.  Type "$c".  Then
> >> type "$q" to quit.
> >>
> >> If the top of the output from $c looks like this, you're seeing the bug
> >> I referenced above.
> >>   libc.so.1`strlen+0x30(0, c99e8054, 8044b2c, c99ad8a1, 8401ddc, 
> >> c99eb158)
> >>   libpython2.4.so.1.0`PyModule_AddStringConstant+0xe(8401ddc, 
> >> c99eb158, 0)
> >>   pyexpat.so`initpyexpat+0x841(8403604, 804503c, 8731b4c, 1, 7, 
> >> d2b89d08)
> >>   libpython2.4.so.1.0`_PyImport_LoadDynamicModule+0x83(804503c, 
> >> 8044bac, d2b89d08
> >> )
> >>   libpython2.4.so.1.0`load_module+0x80(804503c, d2b89d08, 8044bac, 3, 0)
> >>
> >> If the top of your stack doesn't look like that, send the output to
> >> desktop-discussed (cc'ed), and they might be able to help.
> >
> > I have the same output on a fresh install of b66, so its definitely 
> > that bug.
> >
>