Re: MS-Windows build using Posix configury

[Dmitry Gutov]

On 17.05.2013 22:43, Eli Zaretskii wrote:
> > Date: Fri, 17 May 2013 20:40:45 +0300
> > Sun-Java-System-SMTP-Warning: Lines longer than SMTP allows found and
> >         truncated.
> > From: Eli Zaretskii <address@hidden>
> > Cc: address@hidden
> >
> > > Looks like another good candidate for mentioning in INSTALL.MSYS.
> >
> > I'd rather we unload all the DLLs before we exit, to avoid such
> > problems in the first place.
>
> Like in the patch below.  Can you try that?  (I hope you still have
> that libintl-8.dll somewhere which required libgcc_s_dw2-1.dll, and
> can reinstate it at least for the purposes of testing this patch.)

No such luck, but the libintl build here:

http://sourceforge.net/projects/mingw/files/MinGW/Base/gettext/gettext-0.18.1.1-2/

(third item from the top) has the same dependency problem.

The patch doesn't seem to have the desired effect (and I've tried make 
clean && make CFLAGS="-g3").

Moreover, despite the above CFLAGS value, I'm unable to obtain detailed 
backtrace here:

Breakpoint 3, 0x74c98e76 in msvcrt!abort ()
   from C:\windows\syswow64\msvcrt.dll
(gdb) bt full
#0  0x74c98e76 in msvcrt!abort () from C:\windows\syswow64\msvcrt.dll
No symbol table info available.
#1  0x6e956f62 in libgcc_s_dw2-1!__deregister_frame_info_bases ()
   from J:\Apps\System\MinGW\bin\libgcc_s_dw2-1.dll
No symbol table info available.
#2  0x00000000 in ?? ()
No symbol table info available.

Lisp Backtrace:
"kill-emacs" (0x88f028)
"save-buffers-kill-emacs" (0x88f318)
"save-buffers-kill-terminal" (0x88f644)
"call-interactively" (0x88f880)
"command-execute" (0x88fb7c)
(gdb)

If I revert the patch, recompile and reproduce it again, 'bt full' 
outputs it full, no problem.