X-Authentication-Warning: ieva01.lanet.lv: pavenis owned process doing -bs
Date: Wed, 12 May 1999 10:28:16 +0300 (WET)
From: Andris Pavenis <pavenis AT lanet DOT lv>
To: Robert Hoehne <robert DOT hoehne AT gmx DOT net>
cc: djgpp-workers AT delorie DOT com
Subject: Re: gdb 4.18 for DJGPP (alpha)
In-Reply-To: <199905112055.QAA17308@delorie.com>
Message-ID: <Pine.A41.4.05.9905121019170.55766-100000@ieva01.lanet.lv>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Reply-To: djgpp-workers AT delorie DOT com
X-Mailing-List: djgpp-workers AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk



On Tue, 11 May 1999, Robert Hoehne wrote:

> > > 	- moved definitions and initialization of gdb_stdout and gdb_stderr and 
> > >           some other global variables from main.c to top.c. Some time ago it was
> > >           discussed in one of gdb mailing lists
> > 
> > These seem like they wouldn't be djgpp patches.  If they aren't, they
> > need to be sent to the usual gdb list.
> 
> They were discussed on the gdb-list and not applied. Since the main
> problem with it was the integration in RHIDE I solved the problem now
> for the coming RHIDE version in an other way.

Yes it was so. But then my patch used __attribute__((constructor)) which
was not accepted for understandable reasons. Current patch does not use
any similar gcc specific hacks or multiple initialization levels. Of
course one can duplicate the code I suggest to move to top.c to any
program that uses gdb in form of library (like rhide). But if additional
similar initialization will be added to gdb/main.c then we'll have to
again patch all programs that uses it.

Andris