All of lore.kernel.org
 help / color / mirror / Atom feed
From: Christoph Hellwig <hch@lst.de>
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, "Darrick J . Wong" <darrick.wong@oracle.com>
Subject: [PATCH 26/26] xfs: try any AG when allocating the first btree block when reflinking
Date: Sat,  1 Apr 2017 08:35:12 +0200	[thread overview]
Message-ID: <20170401063512.25313-27-hch@lst.de> (raw)
In-Reply-To: <20170401063512.25313-1-hch@lst.de>

commit 2fcc319d2467a5f5b78f35f79fd6e22741a31b1e upstream.

When a reflink operation causes the bmap code to allocate a btree block
we're currently doing single-AG allocations due to having ->firstblock
set and then try any higher AG due a little reflink quirk we've put in
when adding the reflink code.  But given that we do not have a minleft
reservation of any kind in this AG we can still not have any space in
the same or higher AG even if the file system has enough free space.
To fix this use a XFS_ALLOCTYPE_FIRST_AG allocation in this fall back
path instead.

[And yes, we need to redo this properly instead of piling hacks over
 hacks.  I'm working on that, but it's not going to be a small series.
 In the meantime this fixes the customer reported issue]

Also add a warning for failing allocations to make it easier to debug.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/libxfs/xfs_bmap.c       | 10 +++++++---
 fs/xfs/libxfs/xfs_bmap_btree.c |  6 +++---
 2 files changed, 10 insertions(+), 6 deletions(-)

diff --git a/fs/xfs/libxfs/xfs_bmap.c b/fs/xfs/libxfs/xfs_bmap.c
index eb34879d1a0c..3365ecb9074d 100644
--- a/fs/xfs/libxfs/xfs_bmap.c
+++ b/fs/xfs/libxfs/xfs_bmap.c
@@ -769,8 +769,8 @@ xfs_bmap_extents_to_btree(
 		args.type = XFS_ALLOCTYPE_START_BNO;
 		args.fsbno = XFS_INO_TO_FSB(mp, ip->i_ino);
 	} else if (dfops->dop_low) {
-try_another_ag:
 		args.type = XFS_ALLOCTYPE_START_BNO;
+try_another_ag:
 		args.fsbno = *firstblock;
 	} else {
 		args.type = XFS_ALLOCTYPE_NEAR_BNO;
@@ -796,13 +796,17 @@ xfs_bmap_extents_to_btree(
 	if (xfs_sb_version_hasreflink(&cur->bc_mp->m_sb) &&
 	    args.fsbno == NULLFSBLOCK &&
 	    args.type == XFS_ALLOCTYPE_NEAR_BNO) {
-		dfops->dop_low = true;
+		args.type = XFS_ALLOCTYPE_FIRST_AG;
 		goto try_another_ag;
 	}
+	if (WARN_ON_ONCE(args.fsbno == NULLFSBLOCK)) {
+		xfs_iroot_realloc(ip, -1, whichfork);
+		xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
+		return -ENOSPC;
+	}
 	/*
 	 * Allocation can't fail, the space was reserved.
 	 */
-	ASSERT(args.fsbno != NULLFSBLOCK);
 	ASSERT(*firstblock == NULLFSBLOCK ||
 	       args.agno >= XFS_FSB_TO_AGNO(mp, *firstblock));
 	*firstblock = cur->bc_private.b.firstblock = args.fsbno;
diff --git a/fs/xfs/libxfs/xfs_bmap_btree.c b/fs/xfs/libxfs/xfs_bmap_btree.c
index d9be241fc86f..999cc5878890 100644
--- a/fs/xfs/libxfs/xfs_bmap_btree.c
+++ b/fs/xfs/libxfs/xfs_bmap_btree.c
@@ -453,8 +453,8 @@ xfs_bmbt_alloc_block(
 
 	if (args.fsbno == NULLFSBLOCK) {
 		args.fsbno = be64_to_cpu(start->l);
-try_another_ag:
 		args.type = XFS_ALLOCTYPE_START_BNO;
+try_another_ag:
 		/*
 		 * Make sure there is sufficient room left in the AG to
 		 * complete a full tree split for an extent insert.  If
@@ -494,8 +494,8 @@ xfs_bmbt_alloc_block(
 	if (xfs_sb_version_hasreflink(&cur->bc_mp->m_sb) &&
 	    args.fsbno == NULLFSBLOCK &&
 	    args.type == XFS_ALLOCTYPE_NEAR_BNO) {
-		cur->bc_private.b.dfops->dop_low = true;
 		args.fsbno = cur->bc_private.b.firstblock;
+		args.type = XFS_ALLOCTYPE_FIRST_AG;
 		goto try_another_ag;
 	}
 
@@ -512,7 +512,7 @@ xfs_bmbt_alloc_block(
 			goto error0;
 		cur->bc_private.b.dfops->dop_low = true;
 	}
-	if (args.fsbno == NULLFSBLOCK) {
+	if (WARN_ON_ONCE(args.fsbno == NULLFSBLOCK)) {
 		XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT);
 		*stat = 0;
 		return 0;
-- 
2.11.0


  parent reply	other threads:[~2017-04-01  6:37 UTC|newest]

Thread overview: 35+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-04-01  6:34 4.10-stable updates for XFS Christoph Hellwig
2017-04-01  6:34 ` [PATCH 01/26] xfs: pull up iolock from xfs_free_eofblocks() Christoph Hellwig
2017-04-01  6:34 ` [PATCH 02/26] xfs: sync eofblocks scans under iolock are livelock prone Christoph Hellwig
2017-04-01  6:34 ` [PATCH 03/26] xfs: fix eofblocks race with file extending async dio writes Christoph Hellwig
2017-04-01  6:34 ` [PATCH 04/26] xfs: fix toctou race when locking an inode to access the data map Christoph Hellwig
2017-04-01  6:34 ` [PATCH 05/26] xfs: fail _dir_open when readahead fails Christoph Hellwig
2017-04-01  6:34 ` [PATCH 06/26] xfs: filter out obviously bad btree pointers Christoph Hellwig
2017-04-01  6:34 ` [PATCH 07/26] xfs: check for obviously bad level values in the bmbt root Christoph Hellwig
2017-04-01  6:34 ` [PATCH 08/26] xfs: verify free block header fields Christoph Hellwig
2017-04-01  6:34 ` [PATCH 09/26] xfs: allow unwritten extents in the CoW fork Christoph Hellwig
2017-04-01  6:34 ` [PATCH 10/26] xfs: mark speculative prealloc CoW fork extents unwritten Christoph Hellwig
2017-04-01  6:34 ` [PATCH 11/26] xfs: reset b_first_retry_time when clear the retry status of xfs_buf_t Christoph Hellwig
2017-04-01  6:34 ` [PATCH 12/26] xfs: reject all unaligned direct writes to reflinked files Christoph Hellwig
2017-04-01 17:21   ` Greg KH
2017-04-01 17:22     ` Christoph Hellwig
2017-04-01 17:26     ` Greg KH
2017-04-01 17:42       ` Christoph Hellwig
2017-04-01  6:34 ` [PATCH 13/26] xfs: update ctime and mtime on clone destinatation inodes Christoph Hellwig
2017-04-01  6:35 ` [PATCH 14/26] xfs: correct null checks and error processing in xfs_initialize_perag Christoph Hellwig
2017-04-01  6:35 ` [PATCH 15/26] xfs: don't fail xfs_extent_busy allocation Christoph Hellwig
2017-04-01  6:35 ` [PATCH 16/26] xfs: handle indlen shortage on delalloc extent merge Christoph Hellwig
2017-04-01  6:35 ` [PATCH 17/26] xfs: split indlen reservations fairly when under reserved Christoph Hellwig
2017-04-01  6:35 ` [PATCH 18/26] xfs: fix uninitialized variable in _reflink_convert_cow Christoph Hellwig
2017-04-01  6:35 ` [PATCH 19/26] xfs: don't reserve blocks for right shift transactions Christoph Hellwig
2017-04-01  6:35 ` [PATCH 20/26] xfs: Use xfs_icluster_size_fsb() to calculate inode chunk alignment Christoph Hellwig
2017-04-01  6:35 ` [PATCH 21/26] xfs: tune down agno asserts in the bmap code Christoph Hellwig
2017-04-01  6:35 ` [PATCH 22/26] xfs: only reclaim unwritten COW extents periodically Christoph Hellwig
2017-04-01  6:35 ` [PATCH 23/26] xfs: fix and streamline error handling in xfs_end_io Christoph Hellwig
2017-04-01  6:35 ` [PATCH 24/26] xfs: Use xfs_icluster_size_fsb() to calculate inode alignment mask Christoph Hellwig
2017-04-01  6:35 ` [PATCH 25/26] xfs: use iomap new flag for newly allocated delalloc blocks Christoph Hellwig
2017-04-01  6:35 ` Christoph Hellwig [this message]
2017-04-01  7:43 ` 4.10-stable updates for XFS Eryu Guan
2017-04-01 13:37   ` Christoph Hellwig
2017-04-01 17:32 ` Greg KH
  -- strict thread matches above, loose matches on Subject: below --
2017-03-27  8:38 Christoph Hellwig
2017-03-27  8:38 ` [PATCH 26/26] xfs: try any AG when allocating the first btree block when reflinking 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=20170401063512.25313-27-hch@lst.de \
    --to=hch@lst.de \
    --cc=darrick.wong@oracle.com \
    --cc=linux-xfs@vger.kernel.org \
    --cc=stable@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.