From: "Ian Miller" <Ian AT shelob DOT force9 DOT co DOT uk>
Newsgroups: comp.os.msdos.djgpp
Subject: Re: ...and now problem when trying to recompile libm.a !
Date: Sat, 27 Jun 1998 09:22:22 +0100
Organization: Force 9 Internet
Lines: 44
Message-ID: <6n28ri$km3@news3.force9.net>
References: <Pine DOT SUN DOT 3 DOT 91 DOT 980625202115 DOT 29169A-100000 AT is>
NNTP-Posting-Host: 02.usr04.shef.dialup.force9.net
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Precedence: bulk


Eli Zaretskii wrote in message ...
>
>On 25 Jun 1998, Ian Miller wrote:
>> I'd be investigating this [emacs 20.2 not compiling under gcc 2.81]
>> more actively if there were not lots of *really important* World Cup
>> football matches on at the moment :-)
>
>Emacs 20.3 is about to be released, so it'd be swell if it would compile
>with 2.8.1 as well.  As you understand, I cannot debug a problem that I
>don't see.

As it happens, I overslept yesterday and took the day off, so I had some
time to take a look at this problem. After a little effort I realised that
the dump process was expecting temacs to be an *unstubbed* coff
file. Under the port of gcc 2.8x, it isn't. exe2coff sorts things out. The
patch to emacs-20.2/msdos/sed1v2.inp below sorts things out for me.

Note: I'm posting from home using MS Outlook Express so I
shouldn't be at all surprised if tabs are replaced by spaces and so
on in the patch. I'll post this again from work on Monday (where I
use emacs under Linux which I know will do exactly what I tell it to
and no more).

Regards,
Ian
--
Ian Miller, Dorset, UK

*** sed1v2.inp.orig     Sun Sep  7 17:22:46 1997
--- sed1v2.inp  Fri Jun 26 22:42:18 1998
***************
*** 6,11 ****
--- 6,12 ----
  s/^ / /
  s/\.h\.in/.h-in/
  /^    \.\/temacs/i\
+       exe2coff temacs.exe\
        stubedit temacs.exe minstack=100k
  /^MAKE *=/s/^/# /
  /^SHELL *=/s/^/# /