From owner-freebsd-current  Mon May 13 10:53:21 1996
Return-Path: owner-current
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id KAA08520
          for current-outgoing; Mon, 13 May 1996 10:53:21 -0700 (PDT)
Received: from rocky.sri.MT.net (rocky.sri.MT.net [204.182.243.10])
          by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id KAA08513
          for <FreeBSD-current@freebsd.org>; Mon, 13 May 1996 10:53:18 -0700 (PDT)
Received: (from nate@localhost) by rocky.sri.MT.net (8.6.12/8.6.12) id LAA11182; Mon, 13 May 1996 11:52:59 -0600
Date: Mon, 13 May 1996 11:52:59 -0600
From: Nate Williams <nate@sri.MT.net>
Message-Id: <199605131752.LAA11182@rocky.sri.MT.net>
To: Chuck Robey <chuckr@Glue.umd.edu>
Cc: FreeBSD-current@freebsd.org
Subject: Re: PHK malloc
In-Reply-To: <Pine.OSF.3.91.960513122423.3531A-100000@ginger.eng.umd.edu>
References: <Pine.OSF.3.91.960513122423.3531A-100000@ginger.eng.umd.edu>
Sender: owner-current@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

> I was looking at the code for ld.so, and I noticed that it carries it's 
> own version of malloc with it.  How come Poul's malloc isn't used for the 
> run time loader?

For one, Poul's malloc doesn't exist in -stable, and second of all the
malloc has been tweaked a bit from the stock malloc.  I suspect that it
might be fairly easy to bring in the phk-malloc, but the savings would
be pretty minor since the code doesn't malloc much memory at all.


Nate