linux-xfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Brian Foster <bfoster@redhat.com>
To: linux-xfs@vger.kernel.org
Subject: [PATCH v2 05/13] xfs: ratelimit unmount time per-buffer I/O error message
Date: Wed, 22 Apr 2020 13:54:21 -0400	[thread overview]
Message-ID: <20200422175429.38957-6-bfoster@redhat.com> (raw)
In-Reply-To: <20200422175429.38957-1-bfoster@redhat.com>

At unmount time, XFS emits a warning for every in-core buffer that
might have undergone a write error. In practice this behavior is
probably reasonable given that the filesystem is likely short lived
once I/O errors begin to occur consistently. Under certain test or
otherwise expected error conditions, this can spam the logs and slow
down the unmount.

We already have a ratelimit state defined for buffers failing
writeback. Fold this state into the buftarg and reuse it for the
unmount time errors.

Signed-off-by: Brian Foster <bfoster@redhat.com>
---
 fs/xfs/xfs_buf.c      | 13 +++++++++++--
 fs/xfs/xfs_buf.h      |  1 +
 fs/xfs/xfs_buf_item.c | 10 +---------
 3 files changed, 13 insertions(+), 11 deletions(-)

diff --git a/fs/xfs/xfs_buf.c b/fs/xfs/xfs_buf.c
index 7a6bc617f0a9..c28a93d2fd8c 100644
--- a/fs/xfs/xfs_buf.c
+++ b/fs/xfs/xfs_buf.c
@@ -1684,10 +1684,12 @@ xfs_wait_buftarg(
 			struct xfs_buf *bp;
 			bp = list_first_entry(&dispose, struct xfs_buf, b_lru);
 			list_del_init(&bp->b_lru);
-			if (bp->b_flags & XBF_WRITE_FAIL) {
+			if (bp->b_flags & XBF_WRITE_FAIL &&
+			    ___ratelimit(&bp->b_target->bt_ioerror_rl,
+					 "XFS: Corruption Alert")) {
 				xfs_alert(btp->bt_mount,
 "Corruption Alert: Buffer at daddr 0x%llx had permanent write failures!",
-					(long long)bp->b_bn);
+					  (long long)bp->b_bn);
 				xfs_alert(btp->bt_mount,
 "Please run xfs_repair to determine the extent of the problem.");
 			}
@@ -1828,6 +1830,13 @@ xfs_alloc_buftarg(
 	btp->bt_bdev = bdev;
 	btp->bt_daxdev = dax_dev;
 
+	/*
+	 * Buffer IO error rate limiting. Limit it to no more than 10 messages
+	 * per 30 seconds so as to not spam logs too much on repeated errors.
+	 */
+	ratelimit_state_init(&btp->bt_ioerror_rl, 30 * HZ,
+			     DEFAULT_RATELIMIT_BURST);
+
 	if (xfs_setsize_buftarg_early(btp, bdev))
 		goto error_free;
 
diff --git a/fs/xfs/xfs_buf.h b/fs/xfs/xfs_buf.h
index 598b93b17d95..10492f68fd4b 100644
--- a/fs/xfs/xfs_buf.h
+++ b/fs/xfs/xfs_buf.h
@@ -91,6 +91,7 @@ typedef struct xfs_buftarg {
 	struct list_lru		bt_lru;
 
 	struct percpu_counter	bt_io_count;
+	struct ratelimit_state	bt_ioerror_rl;
 } xfs_buftarg_t;
 
 struct xfs_buf;
diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c
index e34298227f87..23cbfeb82183 100644
--- a/fs/xfs/xfs_buf_item.c
+++ b/fs/xfs/xfs_buf_item.c
@@ -480,14 +480,6 @@ xfs_buf_item_unpin(
 	}
 }
 
-/*
- * Buffer IO error rate limiting. Limit it to no more than 10 messages per 30
- * seconds so as to not spam logs too much on repeated detection of the same
- * buffer being bad..
- */
-
-static DEFINE_RATELIMIT_STATE(xfs_buf_write_fail_rl_state, 30 * HZ, 10);
-
 STATIC uint
 xfs_buf_item_push(
 	struct xfs_log_item	*lip,
@@ -518,7 +510,7 @@ xfs_buf_item_push(
 
 	/* has a previous flush failed due to IO errors? */
 	if ((bp->b_flags & XBF_WRITE_FAIL) &&
-	    ___ratelimit(&xfs_buf_write_fail_rl_state, "XFS: Failing async write")) {
+	    ___ratelimit(&bp->b_target->bt_ioerror_rl, "XFS: Failing async write")) {
 		xfs_warn(bp->b_mount,
 "Failing async write on buffer block 0x%llx. Retrying async write.",
 			 (long long)bp->b_bn);
-- 
2.21.1


  parent reply	other threads:[~2020-04-22 17:54 UTC|newest]

Thread overview: 46+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-04-22 17:54 [PATCH v2 00/13] xfs: flush related error handling cleanups Brian Foster
2020-04-22 17:54 ` [PATCH v2 01/13] xfs: refactor failed buffer resubmission into xfsaild Brian Foster
2020-04-23  4:09   ` Dave Chinner
2020-04-25 17:21   ` Christoph Hellwig
2020-04-22 17:54 ` [PATCH v2 02/13] xfs: factor out buffer I/O failure simulation code Brian Foster
2020-04-23  4:10   ` Dave Chinner
2020-04-25 17:23   ` Christoph Hellwig
2020-04-27 11:11     ` Brian Foster
2020-04-22 17:54 ` [PATCH v2 03/13] xfs: fallthru to buffer attach on error and simplify error handling Brian Foster
2020-04-23  4:18   ` Dave Chinner
2020-04-23 14:28     ` Brian Foster
2020-04-25 17:26   ` Christoph Hellwig
2020-04-22 17:54 ` [PATCH v2 04/13] xfs: remove unnecessary shutdown check from xfs_iflush() Brian Foster
2020-04-25 17:27   ` Christoph Hellwig
2020-04-22 17:54 ` Brian Foster [this message]
2020-04-23  4:46   ` [PATCH v2 05/13] xfs: ratelimit unmount time per-buffer I/O error message Dave Chinner
2020-04-23 14:29     ` Brian Foster
2020-04-23 21:14       ` Dave Chinner
2020-04-24 11:12         ` Brian Foster
2020-04-24 22:08           ` Dave Chinner
2020-04-27 11:11             ` Brian Foster
2020-04-22 17:54 ` [PATCH v2 06/13] xfs: fix duplicate verification from xfs_qm_dqflush() Brian Foster
2020-04-23  4:47   ` Dave Chinner
2020-04-25 17:28   ` Christoph Hellwig
2020-04-22 17:54 ` [PATCH v2 07/13] xfs: abort consistently on dquot flush failure Brian Foster
2020-04-25 17:30   ` Christoph Hellwig
2020-04-22 17:54 ` [PATCH v2 08/13] xfs: elide the AIL lock on log item failure tracking Brian Foster
2020-04-23  5:59   ` Dave Chinner
2020-04-23 14:36     ` Brian Foster
2020-04-23 21:38       ` Dave Chinner
2020-04-24 11:14         ` Brian Foster
2020-04-22 17:54 ` [PATCH v2 09/13] xfs: clean up AIL log item removal functions Brian Foster
2020-04-23  4:54   ` Dave Chinner
2020-04-25 17:37   ` Christoph Hellwig
2020-04-27 11:12     ` Brian Foster
2020-04-22 17:54 ` [PATCH v2 10/13] xfs: combine xfs_trans_ail_[remove|delete]() Brian Foster
2020-04-23  4:55   ` Dave Chinner
2020-04-22 17:54 ` [PATCH v2 11/13] xfs: remove unused iflush stale parameter Brian Foster
2020-04-25 17:37   ` Christoph Hellwig
2020-04-22 17:54 ` [PATCH v2 12/13] xfs: random buffer write failure errortag Brian Foster
2020-04-23  5:11   ` Dave Chinner
2020-04-25 17:38   ` Christoph Hellwig
2020-04-27 11:12     ` Brian Foster
2020-04-22 17:54 ` [PATCH v2 13/13] xfs: remove unused shutdown types Brian Foster
2020-04-23  5:13   ` Dave Chinner
2020-04-25 17:39   ` Christoph Hellwig

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20200422175429.38957-6-bfoster@redhat.com \
    --to=bfoster@redhat.com \
    --cc=linux-xfs@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).