Date: Sat, 10 May 2003 20:32:54 +0300
From: "Eli Zaretskii" <eliz AT elta DOT co DOT il>
Sender: halo1 AT zahav DOT net DOT il
To: djgpp-workers AT delorie DOT com
Message-Id: <3791-Sat10May2003203254+0300-eliz@elta.co.il>
X-Mailer: emacs 21.3.50 (via feedmail 8 I) and Blat ver 1.8.9
In-reply-to: <3EBCE61D.6861C887@yahoo.com> (message from CBFalconer on Sat, 10
	May 2003 07:44:29 -0400)
Subject: Re: Reducing sbrk test on nmalloc]
References: <3EBBF6E4 DOT CF2893B9 AT yahoo DOT com> <2950-Sat10May2003125017+0300-eliz AT elta DOT co DOT il> <3EBCE61D DOT 6861C887 AT yahoo DOT com>
Reply-To: djgpp-workers AT delorie DOT com
Errors-To: nobody AT delorie DOT com
X-Mailing-List: djgpp-workers AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk

> Date: Sat, 10 May 2003 07:44:29 -0400
> From: CBFalconer <cbfalconer AT yahoo DOT com>
> > 
> > The library build doesn't tolerate warnings, as we use -Werror.  So
> > please make changes to eliminate _all_ warnings.  It is advisable to
> > build the code using the exact set of compiler switches used by the
> > library build (see the file src/gcc.opt in the library sources).
> 
> To do that I need a safe name to guard the development switches,
> with something like:
> 
> #ifndef <safename>
> ...
> #endif
> 
> at present I depend on NDEBUG being defined for production use.  I
> don't want to have to physically revise the source to enable debug
> runs.  This has been in there from day one and nobody said
> anything.

It's very easy: add the necessary -D<WHATEVER> flag to the switches
passed to the compiler in the makefile you use to build nmalloc.  The
library build will not use that makefile; instead, it will at most
change the makefile already present in the appropriate subdirectory of
the library source distribution (ideally, no changes should be
necessary at all).  Therefore, these two makefiles can use different
sets of compiler switches.

> I am leaving tomorrow for about two weeks so it will have to wait.

I don't see any problems with that: the first alpha has just hit the
FTP sites a couple of weeks ago, and Richard said he will be busy for
a few weeks anyway.