From owner-freebsd-geom@FreeBSD.ORG Mon Aug 20 13:20:07 2007 Return-Path: Delivered-To: freebsd-geom@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 5830016A41A for ; Mon, 20 Aug 2007 13:20:07 +0000 (UTC) (envelope-from hg@queue.to) Received: from pickle.queue.to (pickle.queue.to [71.180.69.18]) by mx1.freebsd.org (Postfix) with ESMTP id F319713C45A for ; Mon, 20 Aug 2007 13:20:06 +0000 (UTC) (envelope-from hg@queue.to) Received: (qmail 83138 invoked from network); 20 Aug 2007 09:20:06 -0400 Received: from cally.queue.to (172.16.0.6) by pickle.queue.to with ESMTP; 20 Aug 2007 09:20:06 -0400 Message-ID: <46C99506.2040106@queue.to> Date: Mon, 20 Aug 2007 09:20:06 -0400 From: Howard Goldstein User-Agent: Thunderbird 2.0.0.6 (X11/20070802) MIME-Version: 1.0 To: arne_woerner@yahoo.com References: <541513.41122.qm@web30315.mail.mud.yahoo.com> In-Reply-To: <541513.41122.qm@web30315.mail.mud.yahoo.com> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Cc: freebsd-geom@freebsd.org Subject: Re: graid5, 3 consumers, unaligned access X-BeenThere: freebsd-geom@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: GEOM-specific discussions and implementations List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 20 Aug 2007 13:20:07 -0000 Arne Wörner wrote: > --- Howard Goldstein wrote: >> The twe driver has a design flaw that depends on malloc()ing bounce >> buffers when it's handed data not aligned on 512 byte boundaries. When >> malloc fails, the driver syslogs a unique error that only can come from >> > I had a look at that file (twe...c) and found that it is not 512 bytes but 64 > bytes (in 6.2R) and that it is about the virtual memory address and not about > the on-disk-offset... Ahh you're believing the code comment :( (Rule #1: Never believe the code comments). The code comment is wrong, the place where it claims 64 bytes it actually wants 512 bytes, you have to look at the #define and not the comment. The problem for me is much worse since it happens all the time with that requirement. Can the device driver ever expect to see 512 byte aligned buffers from geom? > > So it is not a GEOM problem... > > Maybe u could try to reduce the graid5 write cache by setting .maxwql and > .maxmem to something smaller. OK