Attachment 609686 Details for Bug 821580 – block and scsi error throttling patch

[patch] block and scsi error throttling patch

v3-scsi_lib-rate-limit-the-error-message-from-failing-commands.patch (text/plain), 4.00 KB, created by Mike Snitzer on 2012-09-04 13:33:42 UTC

(hide)

Description:

Filename:

MIME Type:

Creator: Mike Snitzer

Created: 2012-09-04 13:33:42 UTC

Size: 4.00 KB

patch

obsolete

>From patchwork Thu Aug 30 21:06:07 2012
>Content-Type: text/plain; charset="utf-8"
>MIME-Version: 1.0
>Content-Transfer-Encoding: 7bit
>Subject: [v3] scsi_lib: rate-limit the error message from failing commands
>Date: Fri, 31 Aug 2012 05:06:07 -0000
>From: Yi Zou <yi.zou@intel.com>
>X-Patchwork-Id: 2655
>Message-Id: <20120830210607.28119.96438.stgit@localhost6.localdomain6>
>To: axboe@kernel.dk, James.Bottomley@HansenPartnership.com
>Cc: Tomas Henzl <thenzl@redhat.com>,
> "www.Open-FCoE.org" <devel@open-fcoe.org>, linux-scsi@vger.kernel.org
>
>[
>Jens/James,
>
>This is a rather old rate limt patch but never gets picked up in upstream, so I
>am resending it here as v3, with some minor changes so it is directly applicable
>to the current linux-2.6.git tree. I don't have that many LUNs that the original
>issue was reported but the patch was already reviewed and acked in the past,
>I did retest w/ stress I/O on 4 LUNs I have right now.
>
>Original thread can be found at:
>http://comments.gmane.org/gmane.linux.scsi/73497
>http://www.open-fcoe.org/patchwork/patch/2436/
>
>Can you please pull this patch in? it's been hanging there for a long time.
>Original patch description is below.
>
>thanks,
>yi
>]
>
>When performing a cable pull test w/ active stress I/O using fio over
>a dual port Intel 82599 FCoE CNA, w/ 256LUNs on one port and about 32LUNs
>on the other, it is observed that the system becomes not usable due to
>scsi-ml being busy printing the error messages for all the failing commands.
>I don't believe this problem is specific to FCoE and these commands are
>anyway failing due to link being down (DID_NO_CONNECT), just rate-limit
>the messages here to solve this issue.
>
>v2->v1: use __ratelimit() as Tomas Henzl mentioned as the proper way for
>rate-limit per function. However, in this case, the failed i/o gets to
>blk_end_request_err() and then blk_update_request(), which also has to
>be rate-limited, as added in the v2 of this patch.
>
>v3-v2: resolved conflict to apply on current 3.6-rc3 upstream tip.
>
>Signed-off-by: Yi Zou <yi.zou@intel.com>
>Cc: www.Open-FCoE.org <devel@open-fcoe.org>
>Cc: Tomas Henzl <thenzl@redhat.com>
>Cc: <linux-scsi@vger.kernel.org>
>
>---
>block/blk-core.c        |    8 +++++---
> drivers/scsi/scsi_lib.c |    5 ++++-
> 2 files changed, 9 insertions(+), 4 deletions(-)
>
>diff --git a/block/blk-core.c b/block/blk-core.c
>index 4b4dbdf..ee3cb3a 100644
>--- a/block/blk-core.c
>+++ b/block/blk-core.c
>@@ -2254,9 +2254,11 @@ bool blk_update_request(struct request *req, int error, unsigned int nr_bytes)
> 			error_type = "I/O";
> 			break;
> 		}
>-		printk(KERN_ERR "end_request: %s error, dev %s, sector %llu\n",
>-		       error_type, req->rq_disk ? req->rq_disk->disk_name : "?",
>-		       (unsigned long long)blk_rq_pos(req));
>+		printk_ratelimited(KERN_ERR "end_request: %s error, dev %s, sector %llu\n",
>+				   error_type, req->rq_disk ?
>+				   req->rq_disk->disk_name : "?",
>+				   (unsigned long long)blk_rq_pos(req));
>+
> 	}
> 
> 	blk_account_io_completion(req, nr_bytes);
>diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
>index ffd7773..d784e68 100644
>--- a/drivers/scsi/scsi_lib.c
>+++ b/drivers/scsi/scsi_lib.c
>@@ -20,6 +20,7 @@
> #include <linux/delay.h>
> #include <linux/hardirq.h>
> #include <linux/scatterlist.h>
>+#include <linux/ratelimit.h>
> 
> #include <scsi/scsi.h>
> #include <scsi/scsi_cmnd.h>
>@@ -768,6 +769,8 @@ void scsi_io_completion(struct scsi_cmnd *cmd, unsigned int good_bytes)
> 	enum {ACTION_FAIL, ACTION_REPREP, ACTION_RETRY,
> 	      ACTION_DELAYED_RETRY} action;
> 	char *description = NULL;
>+	static DEFINE_RATELIMIT_STATE(rs,  DEFAULT_RATELIMIT_INTERVAL,
>+				      DEFAULT_RATELIMIT_BURST);
> 
> 	if (result) {
> 		sense_valid = scsi_command_normalize_sense(cmd, &sshdr);
>@@ -958,7 +961,7 @@ void scsi_io_completion(struct scsi_cmnd *cmd, unsigned int good_bytes)
> 	case ACTION_FAIL:
> 		/* Give up and fail the remainder of the request */
> 		scsi_release_buffers(cmd);
>-		if (!(req->cmd_flags & REQ_QUIET)) {
>+		if (!(req->cmd_flags & REQ_QUIET) && __ratelimit(&rs)) {
> 			if (description)
> 				scmd_printk(KERN_INFO, cmd, "%s\n",
> 					    description);

Actions: View | Diff

Attachments on bug 821580: 602962 | 609686