X-Authentication-Warning:	ieva01.lanet.lv: pavenis owned process doing -bs
Date:	Mon, 15 Feb 1999 12:24:07 +0200 (WET)
From:	Andris Pavenis <pavenis AT lanet DOT lv>
To:	Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
cc:	Kbwms AT aol DOT com, djgpp-workers AT delorie DOT com,
	Robert Hoehne <robert DOT hoehne AT gmx DOT net>
Subject:	Re: Speaking of Groff
In-Reply-To:	<Pine.SUN.3.91.990215100345.10128C-100000@is>
Message-ID:	<Pine.A41.4.05.9902151215100.34566-100000@ieva01.lanet.lv>
MIME-Version:	1.0
Reply-To:	djgpp-workers AT delorie DOT com

On Mon, 15 Feb 1999, Eli Zaretskii wrote:

> 
> On Sun, 14 Feb 1999 Kbwms AT aol DOT com wrote:
> 
> > Appended is a diff that represents a 'fix' to $djdir/lang/cxx/_G_config.h
> > that enables the code in archive gro110s.zip to be compiled completely.
> 
> Thanks.  I think this is a bug in gpp281b.zip which should be fixed.

The problem were that I more likely had libsocket installed (perhaps
lsck073b.zip) when built gcc-2.8.1 in October 1998. 

About rebuilding it. I think better would be to change to egcs-1.1.1.
There should be no big problems with C++ exceptions if I apply one small
patch to gcc/dwarf2out.c. C++ exceptions was main reason why I asked DJ to
put egcs-1.1.1 in v2gnu/beta instead of v2gnu.

More serious problem is that building egcs-1.1.1 from source archives I
have uploaded requires LFN support. Unfortunatelly fixing that
perhaps should require too much time...

Maybe we should still make a release of DJGPP port of egcs-1.1.1 (or wait
egcs-1.1.2 that should be out in some months, I hope)

> 
> > The word `fix' is in quotes because I do not know the ramifications of
> > commenting out the line
> > 
> >     #undef NULL 
> 
> Does anybody know why did _G_config.h need to #undef NULL?
> 

Andris

- Raw text -