From owner-freebsd-current  Mon Aug  5 13:06:53 1996
Return-Path: owner-current
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id NAA15731
          for current-outgoing; Mon, 5 Aug 1996 13:06:53 -0700 (PDT)
Received: from peedub.gj.org (newpc.muc.ditec.de [194.120.126.33])
          by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id NAA15710
          for <freebsd-current@freebsd.org>; Mon, 5 Aug 1996 13:06:45 -0700 (PDT)
Received: from peedub.gj.org (localhost [127.0.0.1]) by peedub.gj.org (8.7.5/8.6.9) with ESMTP id WAA00720; Mon, 5 Aug 1996 22:05:14 GMT
Message-Id: <199608052205.WAA00720@peedub.gj.org>
X-Mailer: exmh version 1.6.7 5/3/96
To: Michael Smith <msmith@atrad.adelaide.edu.au>
Cc: freebsd-current@freebsd.org
Subject: Re: question about grinding system (Linux guru attn. reqd) 
Reply-To: Gary Jennejohn <Gary.Jennejohn@munich.netsurf.de>
In-reply-to: Your message of "Mon, 05 Aug 1996 14:14:02 +0930."
             <199608050444.OAA10882@genesis.atrad.adelaide.edu.au> 
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Mon, 05 Aug 1996 22:05:14 +0000
From: Gary Jennejohn <garyj@munich.netsurf.de>
Sender: owner-current@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

Michael Smith writes:
>
>I've just been watching this slowly-dying -current system trying to find 
>something that looks wrong, and I think I have something.
>
>Watching the output of vmstat -m, I notice that the numbers for the 
>'temp' category keep growing, and that there's about 8M allocated to it
>at this point in time.
>
>Following this and the suspicion that the problem is linux-emulation related
>I went looking for M_TEMP allocations in i386/linux, and I'd like someone
>who's better qualified than I to look over the exit conditions in
>linux_emul_find() in linux_util.c because I'm fairly sure that it's
>possible for this function to return without freeing its buffer,
>and this would lead to the sort of leak that I'm seeing.
>
>Naturally, I'll try to test this as soon as I get a chance, but I'd
>appreciate any comments.
>

sure looks like the places where a "goto done" happens fail to free the
buffer.

---
Gary Jennejohn
Home - Gary.Jennejohn@munich.netsurf.de
Work - gjennejohn@frt.dec.com