X-Authentication-Warning: delorie.com: mailnull set sender to djgpp-workers-bounces using -f
From: sandmann AT clio DOT rice DOT edu (Charles Sandmann)
Message-Id: <10203022059.AA24801@clio.rice.edu>
Subject: Re: Malloc/free DJGPP code
To: djgpp-workers AT delorie DOT com
Date: Sat, 2 Mar 2002 14:59:56 -0600 (CST)
In-Reply-To: <1438-Sat02Mar2002205944+0200-eliz@is.elta.co.il> from "Eli Zaretskii" at Mar 02, 2002 08:59:44 PM
X-Mailer: ELM [version 2.5 PL2]
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
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

> > Eli asked to see the one Martin likes, but hasn't commented (didn't like
> > the faster patch so much) - maybe he didn't like the one liner either :-)
> 
> I like the one liner, and am actually using it for the past few weeks.

Okay, two votes for the one liner, and some testing too ...

> I think the speed difference is insignificant, but perhaps someone
> could time that and see.

It won't be measurable probably.  I did an instruction count and that
line will add about 50 instructions (mostly simple, so probably around
50 clocks) to each malloc which must call sbrk.  Each brk/sbrk that 
doesn't call DPMI is about 20 or so instructions.

It really doesn't matter which way we do it, which is why I asked ...