On Sun, Jun 02, 2013 at 05:35:27PM +0200, Dimitry Andric wrote: > On May 29, 2013, at 05:53, Ted Faber <fa...@lunabase.org> wrote: > > On Tue, May 28, 2013 at 08:52:35AM -0700, Ted Faber wrote: > >> On Tue, May 28, 2013 at 02:00:00PM +0200, Dimitry Andric wrote: > >>> On 2013-05-26 01:07, Ted Faber wrote: > >>>> I'm seeing a repeatable, consistent segmentation fault before the first > >>>> window appears (though firefox -ProfileManager brings up the > >>>> profile manager, but crashes when I try to actually start the browser). > ... > > OK, I improved my svn fu, pulled the tree, extracted the patch, applied > > it, made and installed world. > > > > Now I see different behavior, but no better. Still gets a SEGV, but a > > different trace. (Attached) > > Ok, I think I have figured out what the problem is: Firefox's JavaScript > OSFile implementation loads libc.so.6 if it is available! That is, if > you have the compat6x package installed, it loads > /usr/local/lib/libc.so.6, even while /lib/libc.so.7 is already loaded. > > This wreaks havoc when it later tries to call libc functions via libffi, > and the result is that it calls into a completely bogus PLT entry. Most > of the time, this leads to a crash. > > So the easiest way to avoid the crash for now is to uninstall the > compat6x package, and making sure there is no libc.so.6 in > /usr/local/lib/compat. Then Firefox should start with no problems.
Removing the compat6x package does fix my problem. Thanks enormously. I'll reinstall and give your patch a try next, but this is clearly the right bug. Thanks enormously! -- http://www.lunabase.org/~faber Unexpected attachment? http://www.lunabase.org/~faber/FAQ.html#SIG
pgpEzlXowhLRK.pgp
Description: PGP signature