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] xfs: fail if xattr inactivation hits a hole
Date: Thu, 12 Oct 2017 07:27:32 -0400	[thread overview]
Message-ID: <20171012112732.39137-1-bfoster@redhat.com> (raw)
In-Reply-To: <20171010164756.28390-2-bfoster@redhat.com>

The child buffer read in xfs_attr3_node_inactive() should never
reach a hole in the attr fork. If this occurs, it is likely due to a
bug. Prior to commit cd87d867 ("xfs: don't crash on unexpected holes
in dir/attr btrees"), this would result in a crash. Now that the
crash has been fixed, this is a silent failure.

Pass -1 to xfs_da3_node_read() from xfs_da3_node_inactive() to
indicate that reading from a hole is an error. This logs an error to
syslog and fails the inode inactivation, leaving the inode on the AG
unlinked list until removed by xfs_repair (or log recovery). Also
update the subsequent code to reflect that the read now returns a
non-NULL buffer or an error.

Signed-off-by: Brian Foster <bfoster@redhat.com>
---
 fs/xfs/xfs_attr_inactive.c | 69 ++++++++++++++++++++++------------------------
 1 file changed, 33 insertions(+), 36 deletions(-)

diff --git a/fs/xfs/xfs_attr_inactive.c b/fs/xfs/xfs_attr_inactive.c
index e3a950e..52818ea 100644
--- a/fs/xfs/xfs_attr_inactive.c
+++ b/fs/xfs/xfs_attr_inactive.c
@@ -251,47 +251,44 @@ xfs_attr3_node_inactive(
 		 * traversal of the tree so we may deal with many blocks
 		 * before we come back to this one.
 		 */
-		error = xfs_da3_node_read(*trans, dp, child_fsb, -2, &child_bp,
-						XFS_ATTR_FORK);
+		error = xfs_da3_node_read(*trans, dp, child_fsb, -1, &child_bp,
+					  XFS_ATTR_FORK);
 		if (error)
 			return error;
-		if (child_bp) {
-						/* save for re-read later */
-			child_blkno = XFS_BUF_ADDR(child_bp);
 
-			/*
-			 * Invalidate the subtree, however we have to.
-			 */
-			info = child_bp->b_addr;
-			switch (info->magic) {
-			case cpu_to_be16(XFS_DA_NODE_MAGIC):
-			case cpu_to_be16(XFS_DA3_NODE_MAGIC):
-				error = xfs_attr3_node_inactive(trans, dp,
-							child_bp, level + 1);
-				break;
-			case cpu_to_be16(XFS_ATTR_LEAF_MAGIC):
-			case cpu_to_be16(XFS_ATTR3_LEAF_MAGIC):
-				error = xfs_attr3_leaf_inactive(trans, dp,
-							child_bp);
-				break;
-			default:
-				error = -EIO;
-				xfs_trans_brelse(*trans, child_bp);
-				break;
-			}
-			if (error)
-				return error;
+		/* save for re-read later */
+		child_blkno = XFS_BUF_ADDR(child_bp);
 
-			/*
-			 * Remove the subsidiary block from the cache
-			 * and from the log.
-			 */
-			error = xfs_da_get_buf(*trans, dp, 0, child_blkno,
-				&child_bp, XFS_ATTR_FORK);
-			if (error)
-				return error;
-			xfs_trans_binval(*trans, child_bp);
+		/*
+		 * Invalidate the subtree, however we have to.
+		 */
+		info = child_bp->b_addr;
+		switch (info->magic) {
+		case cpu_to_be16(XFS_DA_NODE_MAGIC):
+		case cpu_to_be16(XFS_DA3_NODE_MAGIC):
+			error = xfs_attr3_node_inactive(trans, dp, child_bp,
+							level + 1);
+			break;
+		case cpu_to_be16(XFS_ATTR_LEAF_MAGIC):
+		case cpu_to_be16(XFS_ATTR3_LEAF_MAGIC):
+			error = xfs_attr3_leaf_inactive(trans, dp, child_bp);
+			break;
+		default:
+			error = -EIO;
+			xfs_trans_brelse(*trans, child_bp);
+			break;
 		}
+		if (error)
+			return error;
+
+		/*
+		 * Remove the subsidiary block from the cache and from the log.
+		 */
+		error = xfs_da_get_buf(*trans, dp, 0, child_blkno, &child_bp,
+				       XFS_ATTR_FORK);
+		if (error)
+			return error;
+		xfs_trans_binval(*trans, child_bp);
 
 		/*
 		 * If we're not done, re-read the parent to get the next
-- 
2.9.5


  parent reply	other threads:[~2017-10-12 11:27 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-10-10 16:47 [PATCH 0/2] xfs: test support for xattr inactivate Brian Foster
2017-10-10 16:47 ` [PATCH 1/2] xfs: assert that xattr inactivation never reaches a hole Brian Foster
2017-10-10 16:55   ` Darrick J. Wong
2017-10-10 17:09     ` Brian Foster
2017-10-12 11:27   ` Brian Foster [this message]
2017-10-12 19:55     ` [PATCH v2] xfs: fail if xattr inactivation hits " Darrick J. Wong
2017-10-10 16:47 ` [PATCH 2/2] xfs: buffer lru reference count error injection tag Brian Foster
2017-10-10 16:59   ` Darrick J. Wong

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=20171012112732.39137-1-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).