From owner-freebsd-hackers  Fri Jun 30 20:52:59 1995
Return-Path: hackers-owner
Received: (from majordom@localhost)
          by freefall.cdrom.com (8.6.10/8.6.6) id UAA01270
          for hackers-outgoing; Fri, 30 Jun 1995 20:52:59 -0700
Received: from bunyip.cc.uq.oz.au (bunyip.cc.uq.oz.au [130.102.2.1])
          by freefall.cdrom.com (8.6.10/8.6.6) with SMTP id UAA01264
          for <hackers@freebsd.org>; Fri, 30 Jun 1995 20:52:58 -0700
Received: from cc.uq.oz.au by bunyip.cc.uq.oz.au 
          id <24374-0@bunyip.cc.uq.oz.au>; Sat, 1 Jul 1995 13:52:49 +1000
Received: from netfl15a.devetir.qld.gov.au 
          by pandora.devetir.qld.gov.au (8.6.10/DEVETIR-E0.3a) with ESMTP 
          id NAA05501 for <hackers@freebsd.org>; Sat, 1 Jul 1995 13:57:11 +1000
Received: by netfl15a.devetir.qld.gov.au (8.6.8.1/DEVETIR-0.1) id DAA25726;
          Sat, 1 Jul 1995 03:54:12 GMT
Date: Sat, 1 Jul 1995 03:54:12 GMT
From: Stephen Hocking <sysseh@devetir.qld.gov.au>
Message-Id: <199507010354.DAA25726@netfl15a.devetir.qld.gov.au>
To: hackers@freebsd.org
Subject: Re: Memory leak somewhere?
Sender: hackers-owner@freebsd.org
Precedence: bulk


>|From: Amancio Hasty <hasty@rah.star-gate.com>
>|I switched to gnumalloc on XF86_S3 and I have not seen any problems.
>|It is kind of early to report bugs. So far xman is not hugging the
>|X server's memory, at one point xman managed to make the X server
>|grow to 20MB over here with libc's malloc. My guess is that there
>|is a memory leak on the X server. Given that xman exasperates the
>|problem it may be worth a try to use re-link the X server with
>|mprof and running xman againt the X server.
>|
>|From: J Wunsch <j@uriah.heep.sax.de>
>|Hmm, except for XFree86 3.1.1, i've always modified my xf86site.def to
>|use -lgmalloc (i simply forgot it when re-vamping the last official
>|version from scratch).  I've never noticed any problems.  (And due to
>|the modification of the site.def, this has been inherited by all
>|clients, too.)
>
>I agree with you both. -- It gnumalloc works for the server
>'in my hands' as well.
>
>Yet the beta testers did report problems.  At the moment
>it's far easier for me to create binaries for beta testing.
>So unless there are objections I think I'll use gnumalloc in
>the next couple rounds of beta tests and see if it can pass
>the beta tests.  Rich

	I was one of those beta testers reporting problems, back in the
early days of R6. It's my view that all gnumalloc did was expose another
bug, which was happily scribbling away in the huge holes that normal malloc
leaves. That bug seems to have been fixed, as I've been happily using gnu
malloc for some time now.

	Stephen