Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932389AbWAZSMy (ORCPT ); Thu, 26 Jan 2006 13:12:54 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S932385AbWAZSMy (ORCPT ); Thu, 26 Jan 2006 13:12:54 -0500 Received: from sccrmhc12.comcast.net ([204.127.202.56]:50111 "EHLO sccrmhc12.comcast.net") by vger.kernel.org with ESMTP id S932359AbWAZSMx (ORCPT ); Thu, 26 Jan 2006 13:12:53 -0500 Date: Thu, 26 Jan 2006 13:12:42 -0500 (EST) From: Ariel To: Jamie Heilman cc: Chase Venters , Arjan van de Ven , linux-ide@vger.kernel.org, linux-kernel@vger.kernel.org, linux-scsi@vger.kernel.org Subject: Re: memory leak in scsi_cmd_cache 2.6.15 In-Reply-To: <20060123072556.GC15490@fifty-fifty.audible.transient.net> Message-ID: References: <1137997104.2977.7.camel@laptopd505.fenrus.org> <200601230029.12674.chase.venters@clientec.com> <20060123072556.GC15490@fifty-fifty.audible.transient.net> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1329 Lines: 39 On Sun, 22 Jan 2006, Jamie Heilman wrote: > Ariel wrote: >> ata_piix seems like it's in common for all, but this is not a lot of > Hmm. I just moved my sata_sil stuff out of the way and rebooted: > $ uptime; grep scsi_cmd_cache /proc/slabinfo > 23:22:16 up 4 min, 1 user, load average: 0.00, 0.03, 0.00 > scsi_cmd_cache 1200 1200 384 10 1 : tunables 54 27 8 : slabdata 120 120 $ Is this good or bad? I'm guessing it means it's still leaking. So it's really starting to look like ata_piix is the problem. But we need someone who has the leak to remove that and see if it helps. I can't, since my drives are connected to it. And developers: Can anyone PLEASE take a look at this? We've got 4 confirmed reports of it, and it's nothing to do with a tainted module. Take a look at this slabtop output: OBJS ACTIVE USE OBJ SIZE SLABS OBJ/SLAB CACHE SIZE NAME 1230120 1230120 100% 0.38K 123012 10 492048K scsi_cmd_cache I have a 492MB leak! And notice how objects never become unused - that looks like the problem. -Ariel - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/