From owner-freebsd-stable Thu Jul 11 19:17:03 1996 Return-Path: owner-stable Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id TAA21124 for stable-outgoing; Thu, 11 Jul 1996 19:17:03 -0700 (PDT) Received: from Cypress.Com (janus.cypress.com [157.95.1.1]) by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id TAA21114 for ; Thu, 11 Jul 1996 19:16:55 -0700 (PDT) Received: from cypress.com (diamond.cadc.cypress.com) by Cypress.Com; Thu, 11 Jul 1996 19:22:05 -0700 Received: from onyx.cadc by cypress.com (4.1/SMI-4.1) id AA28357; Thu, 11 Jul 96 21:16:08 CDT Received: by onyx.cadc (SMI-8.6/SMI-SVR4) id VAA25222; Thu, 11 Jul 1996 21:16:04 -0500 Date: Thu, 11 Jul 1996 21:16:04 -0500 From: bab@cypress.com (Barry Boes/CADC Datacomm CAD) Message-Id: <199607120216.VAA25222@onyx.cadc> To: andrew@pubnix.net Cc: imp@village.org, stable@FreeBSD.ORG In-Reply-To: (message from Andrew Webster on Thu, 11 Jul 1996 18:44:33 -0400 (EDT)) Subject: Re: Uptime report Sender: owner-stable@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > Very easy to reproduce, run netscape for a couple of days. It seems that > when it allocates memory for all those .gif and .jpg icons and pictures, > the memory becomes fragmented and is never recovered fully. > > After a while you run out of swap space. > > Happens on an SGI running IRIX 5.3 as well as under FreeBSD 2.1.0-stable > > Solution, quit X every so often, all your memory comes back. > > One could then deduce that the leak is somewhere in X11R6. As another data point, I have also seen Suns running Solaris 2.5 and the openwindows version of X do the same thing. I have had the X server alone consume 300 Mb of swap over a months time. Kill the X server and it all comes back. However, ours might be a different problem since killing the X server frees the swap space. Warner specifies that the used swap remained used even after killing the X server. - Warner : you did kill the X server, right? Your original message says some xdm processes were still running?