Re: Building Emacs from a new MinGW environment

[Andy Moreton]

On Wed 18 Sep 2013, Eli Zaretskii wrote:

[snipped]
>> #12 0x01160b7e in sprintf (__stream=0x0, __format=0x1419ae0 "Copying raw
>> data for %.8s...", __format=0x1419ae0 "Copying raw data for %.8s...") at
>> c:/mingw/bin/../lib/gcc/mingw32/4.7.2/../../../../include/stdio.h:269
>
> __stream being NULL in the call to sprintf is the problem, I think.
> But the only instances of "Copying raw data for %.8s..." I could find
> are in src/unexw32.c and nt/addsection.c, and there the first argument
> is a local array that cannot possibly be NULL, so I'm unsure how this
> could happen.  Moreover, both unexw32.c and addsection.c are unrelated
> to byte-compiling, which just adds to the mystery...

I suspect that gdb is slightly confused. The mingw stdio.h does include
a redirect to mingw supplied versions, but the argument names don't
quite match.

> Do you see the same crash if you byte-compile db-ebrowse.el by hand?
> If so, try running the same command under GDB, you might get a better
> backtrace.  Using the latest port of GDB (7.6.1 should be available)
> might also help.

Byte compiling that file from an uninstalled emacs works properly - it
only seems to fall over when byte compliling during bootstrap. I'll try
to update gdb and report back if that gives a more meaningful backtrace.

    AndyM