From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751115AbWDLNU4 (ORCPT ); Wed, 12 Apr 2006 09:20:56 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751118AbWDLNU4 (ORCPT ); Wed, 12 Apr 2006 09:20:56 -0400 Received: from 220-130-178-143.HINET-IP.hinet.net ([220.130.178.143]:64499 "EHLO areca.com.tw") by vger.kernel.org with ESMTP id S1751115AbWDLNUz (ORCPT ); Wed, 12 Apr 2006 09:20:55 -0400 Message-ID: <002901c65e33$ceac9e00$b100a8c0@erich2003> From: "erich" To: "Jens Axboe" Cc: , =?UTF-8?B?Iijlu6Plronnp5HmioAp5a6J5Y+vTyI=?= , "\"Al Viro\"" , "\"Andrew Morton\"" , "\"Randy.Dunlap\"" , "\"Matti Aarnio\"" , , "\"James Bottomley\"" , "Chris Caputo" References: <007701c653d7$8b8ee670$b100a8c0@erich2003> <004a01c65470$412daaa0$b100a8c0@erich2003> <20060330192057.4bd8c568.akpm@osdl.org> <20060331074237.GH14022@suse.de> Subject: Re: new Areca driver in 2.6.16-rc6-mm2 appears to be broken Date: Wed, 12 Apr 2006 21:20:00 +0800 MIME-Version: 1.0 Content-Type: text/plain; format=flowed; charset="UTF-8"; reply-type=original Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.3790.1830 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.1830 X-OriginalArrivalTime: 12 Apr 2006 13:15:43.0921 (UTC) FILETIME=[343A3210:01C65E33] Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Dear Jens Axboe, I had found a big difference of generic_make_request(struct bio *bio) got message : sdb1: rw=0, want=...., limit=..... ***************** ** TEST 1 ***************** I used "MAX_XFER_SECTORS 4096" driver to do mkfs.ext2 with ARECA RAID volume sdb1. and copy a big file (900MB) into sdb1. If I copy this file from sdb1, the message rw=.... ,want=......, limit=...... will appear immediately. When I reboot the system and used "MAX_XFER_SECTORS 512" driver. I copy this big file from sdb1, the message rw=.... ,want=......, limit=...... still appear immediately. When I redo mkfs.ext2 with ARECA RAID volume sdb1 (used "MAX_XFER_SECTORS 512") and copy the same file (900MB) into sdb1. I copy this file from sdb1 again,the message rw=.... ,want=......, limit=...... disappear. **************** ** TEST 2 **************** When I do another test with "MAX_XFER_SECTORS 4096" driver #echo 0 > /sys/block/sdb/queue/read_ahead_kb #mkfs.ext2 /dev/sdb1 then copy the same file (900MB) into sdb1. I copy this file from sdb1 again,the message rw=.... ,want=......, limit=...... disappear. Even I redo "echo 4096 > /sys/block/sdb/queue/read_ahead_kb" with ARECA RAID volume. The message rw=.... ,want=......, limit=...... never appear at my "copy compare" test script. Now I can say the bug come from "block read ahead". And this bug will certainly appear when do "mkfs.ext2" if read_ahead_kb value large enough. Best Regards Erich Chen ----- Original Message ----- From: "Jens Axboe" To: "Andrew Morton" Cc: "James Bottomley" ; Sent: Friday, March 31, 2006 3:42 PM Subject: Re: new Areca driver in 2.6.16-rc6-mm2 appears to be broken > > (irk, Erich wasn't in the cc, sorry to Andrew and James for getting this > mail twice) > > On Thu, Mar 30 2006, Andrew Morton wrote: >> "erich" wrote: >> > >> > Dear Chris Caputo, >> > >> > Thanks you to conform this issue again, my colleague assisted me and >> > to >> > double check my older version driver yesterday. >> > and the old driver is working fine as your mention before. >> > >> > The ARCMSR_MAX_XFER_SECTORS is the reason why cause "attempt to access >> > beyond end of device". >> > >> > #define ARCMSR_MAX_XFER_SECTORS >> > 256 -----old >> > #define ARCMSR_MAX_XFER_SECTORS >> > 4096 -----new >> >> That seems odd. ARCMSR_MAX_XFER_SECTORS just gets put into >> scsi_host_template.max_sectors. Could it be a scsi core buglet? > > Perhaps the larger max sectors setting is causing read-ahead to be > overly optimistic and going beyond the end? Should not happen. > > Erich, can you try and shrink read-ahead on that device and retest? > Basically just do > > # echo 0 > /sys/block/sdX/queue/read_ahead_kb > > and see if it still complains. > > -- > Jens Axboe >