bug#16132: Crash on Windows

[Andy Moreton <andrewjmoreton@gmail.com>]

On Sat 14 Dec 2013, Eli Zaretskii wrote:

>> From: Andy Moreton <andrewjmoreton@gmail.com>
>> Date: Fri, 13 Dec 2013 22:04:53 +0000
>> 
>> On Fri 13 Dec 2013, Eli Zaretskii wrote:
>> 
>> >> Mingw build of emacs from r115506 running on win7 64bit. Emacs displayed
>> >> the abort dialog during startup, while desktop mode was restoring
>> >> buffers. Attaching a debugger:
>> >
>> > Please start Emacs from the debugger, and tell what kind of backtrace
>> > you see then.  The backtrace you collected is not very informative.
>> 
>> I tried that before sending the report - it was an even less informative
>> SEGV. I've not had much success with gdb recently, and have not yet
>> determined what chaegs to emacs or the toolchain are responsible.
>
> Sorry for asking the obvious: you did check all the threads before
> deciding that the backtraces are not useful?

Yes - gdb aborted when I tried that, so my gdb setup is clearly not
working well.

>> >> Lisp Backtrace:
>> >> "file-system-info" (0x88d2dc)
>> >
>> > It would be good to step into file-system-info and see what goes wrong
>> > there.
>> 
>> I'll report back if I can persuade gdb to report anything more useful.
>
> At the very least, you should be able to set a breakpoint in
> Ffile_system_info, and step through that function when it breaks.
> Assuming there aren't too many buffers in Dired mode in your desktop
> file, you should be able to do that without too much hassle.

After a bootstrap from r115533 and then rebuilding at r115539 I now have
a working emacs. I assume this bug was caused by one or more of the
issues that have been fixed over the last couple of days.

    AndyM