Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 17 Jun 2004 18:54:30 -0700
From:      Sean McNeil <sean@mcneil.com>
To:        Joe Marcus Clarke <marcus@marcuscom.com>
Cc:        freebsd-gnome@freebsd.org
Subject:   Re: firefox core dump
Message-ID:  <1087523669.56010.2.camel@server.mcneil.com>
In-Reply-To: <1087522521.27689.6.camel@shumai.marcuscom.com>
References:  <1087349838.99586.2.camel@server.mcneil.com> <1087364265.24564.8.camel@shumai.marcuscom.com> <1087499331.73513.3.camel@server.mcneil.com> <1087522521.27689.6.camel@shumai.marcuscom.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, 2004-06-17 at 18:35, Joe Marcus Clarke wrote:
> On Thu, 2004-06-17 at 15:08, Sean McNeil wrote:
> > On Tue, 2004-06-15 at 22:37, Joe Marcus Clarke wrote:
> > > On Tue, 2004-06-15 at 21:37, Sean McNeil wrote:
> > > > I just got done rebuilding firefox on amd64 and I get a sigbus on
> > > > startup.  I had
> > > > 
> > > > CFLAGS= -O -pipe -g
> > > > STRIP=
> > > > 
> > > > in my make.conf, but it didn't make a difference for the traceback:
> > > > 
> > > > (gdb) bt
> > > > #0  0x0000000200835a64 in non-virtual thunk to
> > > > nsDirectoryService::AddRef() ()
> > > >    from /usr/X11R6/lib/firefox/lib/firefox-0.9/libxpcom.so
> > > > #1  0x00000002008350a3 in nsDirectoryService::~nsDirectoryService ()
> > > >    from /usr/X11R6/lib/firefox/lib/firefox-0.9/libxpcom.so
> > > > #2  0x0000000200834ddc in nsDirectoryService::Create ()
> > > >    from /usr/X11R6/lib/firefox/lib/firefox-0.9/libxpcom.so
> > > > #3  0x000000020081c4a2 in NS_InitXPCOM2 ()
> > > >    from /usr/X11R6/lib/firefox/lib/firefox-0.9/libxpcom.so
> > > > #4  0x0000000000407747 in ScopedXPCOMStartup::Initialize ()
> > > > #5  0x00000000004089b2 in ScopedXPCOMStartup::SetWindowCreator ()
> > > > #6  0x0000000000408e74 in ScopedXPCOMStartup::SetWindowCreator ()
> > > > #7  0x00000000004098fa in xre_main ()
> > > > #8  0x00000000004058a0 in main ()
> > > > 
> > > > Is that DirectoryService as in ldap?
> > > 
> > > No, this is the component registration stuff.  However, this stack trace
> > > doesn't make much sense.  I have a feeling there is some more amd64 work
> > > that needs to be done for Mozilla 1.7-based browsers.  I was kind of
> > > hoping the amd64 team could have a look since I'm shy on hardware, and
> > > they have the requisite architecture knowledge.
> > > 
> > > Mezz's recommendation is the best course of action right now.  Try to
> > > get a more coherent stack trace, and maybe this will turn out to be an
> > > easy solution.
> > > 
> > > Joe
> > 
> > Sorry, I cannot get any further information on this.  I tried removing
> > everything in /usr/X11R6/lib/firefox but that didn't help.  Setting
> > WITH_DEBUG caused the problem compiling that I posted previously.
> 
> Debugging with Mozilla and Firefox has been problematic due to the lack
> of KSE debugging support.  In general, libmap'ing libpthread to libc_r
> should get Firefox to compile.  However, running under libc_r might not
> produce any crashes.  At least this is what I'm seeing with a
> thunderbird-0.7 crash.
> 
> Joe

In this case, I think it is a compiler problem.  I just CC'd -gnome with
the results of looking into a warning message that turned out to be
invalid code.  Hopefully, this will solve my problem with firefox.  It's
in the right area of code at least.  Still building, but I should know
soon enough.

Sean




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1087523669.56010.2.camel>