Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753518AbXLFMLX (ORCPT ); Thu, 6 Dec 2007 07:11:23 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1752435AbXLFMLO (ORCPT ); Thu, 6 Dec 2007 07:11:14 -0500 Received: from brick.kernel.dk ([87.55.233.238]:7718 "EHLO kernel.dk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752047AbXLFMLM (ORCPT ); Thu, 6 Dec 2007 07:11:12 -0500 Date: Thu, 6 Dec 2007 13:08:43 +0100 From: Jens Axboe To: Hannes Reinecke Cc: Alexey Dobriyan , Andrew Morton , linux-kernel@vger.kernel.org, James.Bottomley@steeleye.com, linux-scsi@vger.kernel.org Subject: Re: 2.6.24-rc4-mm1: hostbyte=0x01 driverbyte=0x00 (now bisected) Message-ID: <20071206120843.GC5231@kernel.dk> References: <20071204211701.994dfce6.akpm@linux-foundation.org> <20071205234154.GA4041@martell.zuzino.mipt.ru> <4757AA3D.4060003@suse.de> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: <4757AA3D.4060003@suse.de> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 4547 Lines: 145 On Thu, Dec 06 2007, Hannes Reinecke wrote: > Alexey Dobriyan wrote: > >> git-scsi-misc.patch > > > > Apologies for not looking into the problem earlier. See > > http://marc.info/?t=119628022300005&r=1&w=2 > > "2.6.24-rc3-mm2: Result: hostbyte=0x01 driverbyte=0x00\nend_request: I/O error" > > for previous installment. > > > > I've bisected it to the following patch in git-scsi-misc branch. > > Revert on top of 2.6.24-rc4-mm1 also helps. > > > > commit 8655a546c83fc43f0a73416bbd126d02de7ad6c0 > > Author: Hannes Reinecke > > Date: Tue Nov 6 09:23:40 2007 +0100 > > > > [SCSI] Do not requeue requests if REQ_FAILFAST is set > > > > Any requests with the REQ_FAILFAST flag set should not be requeued > > to the requeust queue, but rather terminated directly. > > Otherwise the multipath failover will stall until the command > > timeout triggers. > > > > Signed-off-by: Hannes Reinecke > > Signed-off-by: James Bottomley > > > > diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c > > index 0f44bdb..0da0dd0 100644 > > --- a/drivers/scsi/scsi_lib.c > > +++ b/drivers/scsi/scsi_lib.c > > @@ -1286,6 +1286,11 @@ int scsi_prep_state_check(struct scsi_device *sdev, struct request *req) > > */ > > if (!(req->cmd_flags & REQ_PREEMPT)) > > ret = BLKPREP_DEFER; > > + /* > > + * Return failfast requests immediately > > + */ > > + if (req->cmd_flags & REQ_FAILFAST) > > + ret = BLKPREP_KILL; > > break; > > default: > > /* > > @@ -1414,6 +1419,17 @@ static inline int scsi_host_queue_ready(struct request_queue *q, > > return 1; > > } > > > > +static void __scsi_kill_request(struct request *req) > > +{ > > + struct scsi_cmnd *cmd = req->special; > > + struct scsi_device *sdev = cmd->device; > > + > > + cmd->result = DID_NO_CONNECT << 16; > > + atomic_inc(&cmd->device->iorequest_cnt); > > + sdev->device_busy--; > > + __scsi_done(cmd); > > +} > > + > > /* > > * Kill a request for a dead device > > */ > > @@ -1527,8 +1543,16 @@ static void scsi_request_fn(struct request_queue *q) > > * accept it. > > */ > > req = elv_next_request(q); > > - if (!req || !scsi_dev_queue_ready(q, sdev)) > > + if (!req) > > + break; > > + > > + if (!scsi_dev_queue_ready(q, sdev)) { > > + if (req->cmd_flags & REQ_FAILFAST) { > > + scsi_kill_request(req, q); > > + continue; > > + } > > break; > > + } > > > > if (unlikely(!scsi_device_online(sdev))) { > > sdev_printk(KERN_ERR, sdev, > > @@ -1609,8 +1633,12 @@ static void scsi_request_fn(struct request_queue *q) > > * later time. > > */ > > spin_lock_irq(q->queue_lock); > > - blk_requeue_request(q, req); > > - sdev->device_busy--; > > + if (unlikely(req->cmd_flags & REQ_FAILFAST)) > > + __scsi_kill_request(req); > > + else { > > + blk_requeue_request(q, req); > > + sdev->device_busy--; > > + } > > if(sdev->device_busy == 0) > > blk_plug_device(q); > > out: > Yeah, sorry. That patch was bad. Please use the attached one instead. > Andrew, can you replace them? > > Cheers, > > Hannes > -- > Dr. Hannes Reinecke zSeries & Storage > hare@suse.de +49 911 74053 688 > SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 N?rnberg > GF: Markus Rex, HRB 16746 (AG N?rnberg) > diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c > index 13e7e09..9ec1566 100644 > --- a/drivers/scsi/scsi_lib.c > +++ b/drivers/scsi/scsi_lib.c > @@ -1284,13 +1284,15 @@ int scsi_prep_state_check(struct scsi_device *sdev, struct request *req) > /* > * If the devices is blocked we defer normal commands. > */ > - if (!(req->cmd_flags & REQ_PREEMPT)) > - ret = BLKPREP_DEFER; > - /* > - * Return failfast requests immediately > - */ > - if (req->cmd_flags & REQ_FAILFAST) > - ret = BLKPREP_KILL; > + if (!(req->cmd_flags & REQ_PREEMPT)) { > + /* > + * Return failfast requests immediately > + */ > + if (req->cmd_flags & REQ_FAILFAST) > + ret = BLKPREP_KILL; > + else > + ret = BLKPREP_DEFER; > + } > break; > default: > /* can we please stick to using blk_noretry_request() consistently, instead of thwrowing REQ_FAILFAST tests in there? -- Jens Axboe -- 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/