bug#17753: Cygwin emacs-X11 core dump

[Markus Hoenicka]

Am 2014-06-11 04:51, schrieb Eli Zaretskii:
> > Date: Mon, 9 Jun 2014 23:55:32 +0200
> > From: markus.hoenicka@mhoenicka.de
> >
> > As I started to collect crash data only about two weeks ago, it
> > is too early to search for patterns or for ways to reproduce the
> > crashes. This is also why I can't offer backtraces after starting
> > Emacs with the -Q switch. I hope that someone can make sense of the
> > backtrace data anyway.
>
> Thanks.  Your backtrace looks strange: was that binary stripped or
> something?  Why does the backtrace end at CHAR_TABLE_REF, which is a
> very internal function?  Why the "??" in the backtrace?
>
> Are you sure you were in the right thread when you collected the
> backtrace?

No, I'm not sure at all. How do I find out? When I start gdb, I see the 
following:

$ gdb /usr/bin/emacs-X11.exe emacs-X11.exe.core
GNU gdb (GDB) 7.6.50.20130728-cvs (cygwin-special)
[... copyright blurb ...]
Reading symbols from /usr/bin/emacs-X11.exe...Reading symbols from 
/usr/lib/debug/usr/bin/emacs-X11.exe.dbg...done.
done.

warning: core file may not match specified executable file.
[New Thread 0x6d8]
[New Thread 0x64c]
[New Thread 0xda0]
[New Thread 0xf34]
[New Thread 0x8dc]
[New Thread 0xc20]
[New Thread 0x8fc]
[New Thread 0xb48]
[New Thread 0xe14]
#0  0x0000003300001fa4 in ?? ()

This is where I type "bt" or "bt full".

> > I'll be happy to provide further information and run tests if needed.
>
> A this point, we need as many informative backtraces as you can
> provide.
>
> (Btw, the other Cygwin-64 related crashes were in the w32 build, not a
> GTK build of the Cygwin Emacs.)

Yes, I'm aware of that. This is why I felt compelled to report crash 
data from emacs-X11.exe as well.

regards,
Markus

--
Markus Hoenicka
http://www.mhoenicka.de
AQ score 38