From mboxrd@z Thu Jan 1 00:00:00 1970 From: Kumar S Subject: Memory leak in ip_dst_cache Date: Thu, 8 Sep 2011 22:04:34 -0700 (PDT) Message-ID: <1315544674.20226.YahooMailNeo@web113902.mail.gq1.yahoo.com> Reply-To: Kumar S Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE To: "netdev@vger.kernel.org" Return-path: Received: from nm3.bullet.mail.sp2.yahoo.com ([98.139.91.73]:27141 "HELO nm3.bullet.mail.sp2.yahoo.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with SMTP id S1753148Ab1IIFEf convert rfc822-to-8bit (ORCPT ); Fri, 9 Sep 2011 01:04:35 -0400 References: Sender: netdev-owner@vger.kernel.org List-ID: Hi, We are=A0running Linux-2.6.24 kernel on MPC8360. Though forwarding is d= isabled, when connected to public network we see the system running out= of memory and rebooting frequently. After doing some study we found ou= t slowly "ip_dst-cache" is growing, and doesn't release entries. Intere= stingly route entries displayed with command "ip route ls cache" show f= ewer than the active objects listed under "cat /proc/slabinfo | grep ip= _dst_cache". After doing some study, we could reproduce it in the lab b= y injecting packets withdifferent source IP addresses. Ideally ageing s= hould happen, and old entries are supposed to be cleared out, but that = doesn't happen. We do see one or two entries getting agedout but not al= l.=20 After some time we did see that "rt_run_flush" kicks in and flushes out= the ip_dst_cache. That's why the "ip route ls cache" fewer entries. Bu= t looks like __chache_free() doesn't get called, that's why these entri= es are not really released. This results in the leak. =A0 Any idea what is going wrong here. Is it a known bug? =A0 Regards psk