4.13-stable updates for XFS

4.13-stable updates for XFS

[Christoph Hellwig]

Hi all,

below are backports up to 4.14-rc1 for XFS.  I've selected all
clear bugfixes for reported bugs, and a few harmless prep patches
that allow us to carry late bug fixes as-is instead of providing
special versions for -stable, which in the past have caused more
harm than good.

[PATCH 01/25] xfs: write unmount record for ro mounts

[Christoph Hellwig]

From: Eric Sandeen <sandeen@sandeen.net>

commit 757a69ef6cf2bf839bd4088e5609ddddd663b0c4 upstream.

There are dueling comments in the xfs code about intent
for log writes when unmounting a readonly filesystem.

In xfs_mountfs, we see the intent:

/*
 * Now the log is fully replayed, we can transition to full read-only
 * mode for read-only mounts. This will sync all the metadata and clean
 * the log so that the recovery we just performed does not have to be
 * replayed again on the next mount.
 */

and it calls xfs_quiesce_attr(), but by the time we get to
xfs_log_unmount_write(), it returns early for a RDONLY mount:

 * Don't write out unmount record on read-only mounts.

Because of this, sequential ro mounts of a filesystem with
a dirty log will replay the log each time, which seems odd.

Fix this by writing an unmount record even for RO mounts, as long
as norecovery wasn't specified (don't write a clean log record
if a dirty log may still be there!) and the log device is
writable.

Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-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/xfs_log.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c
index 4ebd0bafc914..972eda87db2b 100644
--- a/fs/xfs/xfs_log.c
+++ b/fs/xfs/xfs_log.c
@@ -812,11 +812,14 @@ xfs_log_unmount_write(xfs_mount_t *mp)
 	int		 error;
 
 	/*
-	 * Don't write out unmount record on read-only mounts.
+	 * Don't write out unmount record on norecovery mounts or ro devices.
 	 * Or, if we are doing a forced umount (typically because of IO errors).
 	 */
-	if (mp->m_flags & XFS_MOUNT_RDONLY)
+	if (mp->m_flags & XFS_MOUNT_NORECOVERY ||
+	    xfs_readonly_buftarg(log->l_mp->m_logdev_targp)) {
+		ASSERT(mp->m_flags & XFS_MOUNT_RDONLY);
 		return 0;
+	}
 
 	error = _xfs_log_force(mp, XFS_LOG_SYNC, NULL);
 	ASSERT(error || !(XLOG_FORCED_SHUTDOWN(log)));
-- 
2.14.1

Patch "xfs: write unmount record for ro mounts" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: write unmount record for ro mounts

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-write-unmount-record-for-ro-mounts.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:07 -0700
Subject: xfs: write unmount record for ro mounts
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, Eric Sandeen <sandeen@sandeen.net>, Eric Sandeen <sandeen@redhat.com>, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-2-hch@lst.de>

From: Eric Sandeen <sandeen@sandeen.net>

commit 757a69ef6cf2bf839bd4088e5609ddddd663b0c4 upstream.

There are dueling comments in the xfs code about intent
for log writes when unmounting a readonly filesystem.

In xfs_mountfs, we see the intent:

/*
 * Now the log is fully replayed, we can transition to full read-only
 * mode for read-only mounts. This will sync all the metadata and clean
 * the log so that the recovery we just performed does not have to be
 * replayed again on the next mount.
 */

and it calls xfs_quiesce_attr(), but by the time we get to
xfs_log_unmount_write(), it returns early for a RDONLY mount:

 * Don't write out unmount record on read-only mounts.

Because of this, sequential ro mounts of a filesystem with
a dirty log will replay the log each time, which seems odd.

Fix this by writing an unmount record even for RO mounts, as long
as norecovery wasn't specified (don't write a clean log record
if a dirty log may still be there!) and the log device is
writable.

Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-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>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/xfs_log.c |    7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

--- a/fs/xfs/xfs_log.c
+++ b/fs/xfs/xfs_log.c
@@ -812,11 +812,14 @@ xfs_log_unmount_write(xfs_mount_t *mp)
 	int		 error;
 
 	/*
-	 * Don't write out unmount record on read-only mounts.
+	 * Don't write out unmount record on norecovery mounts or ro devices.
 	 * Or, if we are doing a forced umount (typically because of IO errors).
 	 */
-	if (mp->m_flags & XFS_MOUNT_RDONLY)
+	if (mp->m_flags & XFS_MOUNT_NORECOVERY ||
+	    xfs_readonly_buftarg(log->l_mp->m_logdev_targp)) {
+		ASSERT(mp->m_flags & XFS_MOUNT_RDONLY);
 		return 0;
+	}
 
 	error = _xfs_log_force(mp, XFS_LOG_SYNC, NULL);
 	ASSERT(error || !(XLOG_FORCED_SHUTDOWN(log)));


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 02/25] xfs: toggle readonly state around xfs_log_mount_finish

[Christoph Hellwig]

From: Eric Sandeen <sandeen@sandeen.net>

commit 6f4a1eefdd0ad4561543270a7fceadabcca075dd upstream.

When we do log recovery on a readonly mount, unlinked inode
processing does not happen due to the readonly checks in
xfs_inactive(), which are trying to prevent any I/O on a
readonly mount.

This is misguided - we do I/O on readonly mounts all the time,
for consistency; for example, log recovery.  So do the same
RDONLY flag twiddling around xfs_log_mount_finish() as we
do around xfs_log_mount(), for the same reason.

This all cries out for a big rework but for now this is a
simple fix to an obvious problem.

Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-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/xfs_log.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c
index 972eda87db2b..f467f1230444 100644
--- a/fs/xfs/xfs_log.c
+++ b/fs/xfs/xfs_log.c
@@ -743,10 +743,14 @@ xfs_log_mount_finish(
 	struct xfs_mount	*mp)
 {
 	int	error = 0;
+	bool	readonly = (mp->m_flags & XFS_MOUNT_RDONLY);
 
 	if (mp->m_flags & XFS_MOUNT_NORECOVERY) {
 		ASSERT(mp->m_flags & XFS_MOUNT_RDONLY);
 		return 0;
+	} else if (readonly) {
+		/* Allow unlinked processing to proceed */
+		mp->m_flags &= ~XFS_MOUNT_RDONLY;
 	}
 
 	/*
@@ -764,6 +768,9 @@ xfs_log_mount_finish(
 		xfs_log_work_queue(mp);
 	mp->m_super->s_flags &= ~MS_ACTIVE;
 
+	if (readonly)
+		mp->m_flags |= XFS_MOUNT_RDONLY;
+
 	return error;
 }
 
-- 
2.14.1

Patch "xfs: toggle readonly state around xfs_log_mount_finish" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: toggle readonly state around xfs_log_mount_finish

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:08 -0700
Subject: xfs: toggle readonly state around xfs_log_mount_finish
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, Eric Sandeen <sandeen@sandeen.net>, Eric Sandeen <sandeen@redhat.com>, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-3-hch@lst.de>

From: Eric Sandeen <sandeen@sandeen.net>

commit 6f4a1eefdd0ad4561543270a7fceadabcca075dd upstream.

When we do log recovery on a readonly mount, unlinked inode
processing does not happen due to the readonly checks in
xfs_inactive(), which are trying to prevent any I/O on a
readonly mount.

This is misguided - we do I/O on readonly mounts all the time,
for consistency; for example, log recovery.  So do the same
RDONLY flag twiddling around xfs_log_mount_finish() as we
do around xfs_log_mount(), for the same reason.

This all cries out for a big rework but for now this is a
simple fix to an obvious problem.

Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-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>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/xfs_log.c |    7 +++++++
 1 file changed, 7 insertions(+)

--- a/fs/xfs/xfs_log.c
+++ b/fs/xfs/xfs_log.c
@@ -743,10 +743,14 @@ xfs_log_mount_finish(
 	struct xfs_mount	*mp)
 {
 	int	error = 0;
+	bool	readonly = (mp->m_flags & XFS_MOUNT_RDONLY);
 
 	if (mp->m_flags & XFS_MOUNT_NORECOVERY) {
 		ASSERT(mp->m_flags & XFS_MOUNT_RDONLY);
 		return 0;
+	} else if (readonly) {
+		/* Allow unlinked processing to proceed */
+		mp->m_flags &= ~XFS_MOUNT_RDONLY;
 	}
 
 	/*
@@ -764,6 +768,9 @@ xfs_log_mount_finish(
 		xfs_log_work_queue(mp);
 	mp->m_super->s_flags &= ~MS_ACTIVE;
 
+	if (readonly)
+		mp->m_flags |= XFS_MOUNT_RDONLY;
+
 	return error;
 }
 


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 03/25] xfs: Add infrastructure needed for error propagation during buffer IO failure

[Christoph Hellwig]

From: Carlos Maiolino <cmaiolino@redhat.com>

commit 0b80ae6ed13169bd3a244e71169f2cc020b0c57a upstream.

With the current code, XFS never re-submit a failed buffer for IO,
because the failed item in the buffer is kept in the flush locked state
forever.

To be able to resubmit an log item for IO, we need a way to mark an item
as failed, if, for any reason the buffer which the item belonged to
failed during writeback.

Add a new log item callback to be used after an IO completion failure
and make the needed clean ups.

Reviewed-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-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/xfs_buf_item.c | 32 +++++++++++++++++++++++++++++++-
 fs/xfs/xfs_trans.h    |  7 +++++--
 2 files changed, 36 insertions(+), 3 deletions(-)

diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c
index f6a8422e9562..7573a1f0bc9a 100644
--- a/fs/xfs/xfs_buf_item.c
+++ b/fs/xfs/xfs_buf_item.c
@@ -29,6 +29,7 @@
 #include "xfs_error.h"
 #include "xfs_trace.h"
 #include "xfs_log.h"
+#include "xfs_inode.h"
 
 
 kmem_zone_t	*xfs_buf_item_zone;
@@ -1054,6 +1055,31 @@ xfs_buf_do_callbacks(
 	}
 }
 
+/*
+ * Invoke the error state callback for each log item affected by the failed I/O.
+ *
+ * If a metadata buffer write fails with a non-permanent error, the buffer is
+ * eventually resubmitted and so the completion callbacks are not run. The error
+ * state may need to be propagated to the log items attached to the buffer,
+ * however, so the next AIL push of the item knows hot to handle it correctly.
+ */
+STATIC void
+xfs_buf_do_callbacks_fail(
+	struct xfs_buf		*bp)
+{
+	struct xfs_log_item	*next;
+	struct xfs_log_item	*lip = bp->b_fspriv;
+	struct xfs_ail		*ailp = lip->li_ailp;
+
+	spin_lock(&ailp->xa_lock);
+	for (; lip; lip = next) {
+		next = lip->li_bio_list;
+		if (lip->li_ops->iop_error)
+			lip->li_ops->iop_error(lip, bp);
+	}
+	spin_unlock(&ailp->xa_lock);
+}
+
 static bool
 xfs_buf_iodone_callback_error(
 	struct xfs_buf		*bp)
@@ -1123,7 +1149,11 @@ xfs_buf_iodone_callback_error(
 	if ((mp->m_flags & XFS_MOUNT_UNMOUNTING) && mp->m_fail_unmount)
 		goto permanent_error;
 
-	/* still a transient error, higher layers will retry */
+	/*
+	 * Still a transient error, run IO completion failure callbacks and let
+	 * the higher layers retry the buffer.
+	 */
+	xfs_buf_do_callbacks_fail(bp);
 	xfs_buf_ioerror(bp, 0);
 	xfs_buf_relse(bp);
 	return true;
diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h
index 6bdad6f58934..442d679210a1 100644
--- a/fs/xfs/xfs_trans.h
+++ b/fs/xfs/xfs_trans.h
@@ -64,11 +64,13 @@ typedef struct xfs_log_item {
 } xfs_log_item_t;
 
 #define	XFS_LI_IN_AIL	0x1
-#define XFS_LI_ABORTED	0x2
+#define	XFS_LI_ABORTED	0x2
+#define	XFS_LI_FAILED	0x4
 
 #define XFS_LI_FLAGS \
 	{ XFS_LI_IN_AIL,	"IN_AIL" }, \
-	{ XFS_LI_ABORTED,	"ABORTED" }
+	{ XFS_LI_ABORTED,	"ABORTED" }, \
+	{ XFS_LI_FAILED,	"FAILED" }
 
 struct xfs_item_ops {
 	void (*iop_size)(xfs_log_item_t *, int *, int *);
@@ -79,6 +81,7 @@ struct xfs_item_ops {
 	void (*iop_unlock)(xfs_log_item_t *);
 	xfs_lsn_t (*iop_committed)(xfs_log_item_t *, xfs_lsn_t);
 	void (*iop_committing)(xfs_log_item_t *, xfs_lsn_t);
+	void (*iop_error)(xfs_log_item_t *, xfs_buf_t *);
 };
 
 void	xfs_log_item_init(struct xfs_mount *mp, struct xfs_log_item *item,
-- 
2.14.1

Patch "xfs: Add infrastructure needed for error propagation during buffer IO failure" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: Add infrastructure needed for error propagation during buffer IO failure

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:09 -0700
Subject: xfs: Add infrastructure needed for error propagation during buffer IO failure
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, Carlos Maiolino <cmaiolino@redhat.com>, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-4-hch@lst.de>

From: Carlos Maiolino <cmaiolino@redhat.com>

commit 0b80ae6ed13169bd3a244e71169f2cc020b0c57a upstream.

With the current code, XFS never re-submit a failed buffer for IO,
because the failed item in the buffer is kept in the flush locked state
forever.

To be able to resubmit an log item for IO, we need a way to mark an item
as failed, if, for any reason the buffer which the item belonged to
failed during writeback.

Add a new log item callback to be used after an IO completion failure
and make the needed clean ups.

Reviewed-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-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>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/xfs_buf_item.c |   32 +++++++++++++++++++++++++++++++-
 fs/xfs/xfs_trans.h    |    7 +++++--
 2 files changed, 36 insertions(+), 3 deletions(-)

--- a/fs/xfs/xfs_buf_item.c
+++ b/fs/xfs/xfs_buf_item.c
@@ -29,6 +29,7 @@
 #include "xfs_error.h"
 #include "xfs_trace.h"
 #include "xfs_log.h"
+#include "xfs_inode.h"
 
 
 kmem_zone_t	*xfs_buf_item_zone;
@@ -1054,6 +1055,31 @@ xfs_buf_do_callbacks(
 	}
 }
 
+/*
+ * Invoke the error state callback for each log item affected by the failed I/O.
+ *
+ * If a metadata buffer write fails with a non-permanent error, the buffer is
+ * eventually resubmitted and so the completion callbacks are not run. The error
+ * state may need to be propagated to the log items attached to the buffer,
+ * however, so the next AIL push of the item knows hot to handle it correctly.
+ */
+STATIC void
+xfs_buf_do_callbacks_fail(
+	struct xfs_buf		*bp)
+{
+	struct xfs_log_item	*next;
+	struct xfs_log_item	*lip = bp->b_fspriv;
+	struct xfs_ail		*ailp = lip->li_ailp;
+
+	spin_lock(&ailp->xa_lock);
+	for (; lip; lip = next) {
+		next = lip->li_bio_list;
+		if (lip->li_ops->iop_error)
+			lip->li_ops->iop_error(lip, bp);
+	}
+	spin_unlock(&ailp->xa_lock);
+}
+
 static bool
 xfs_buf_iodone_callback_error(
 	struct xfs_buf		*bp)
@@ -1123,7 +1149,11 @@ xfs_buf_iodone_callback_error(
 	if ((mp->m_flags & XFS_MOUNT_UNMOUNTING) && mp->m_fail_unmount)
 		goto permanent_error;
 
-	/* still a transient error, higher layers will retry */
+	/*
+	 * Still a transient error, run IO completion failure callbacks and let
+	 * the higher layers retry the buffer.
+	 */
+	xfs_buf_do_callbacks_fail(bp);
 	xfs_buf_ioerror(bp, 0);
 	xfs_buf_relse(bp);
 	return true;
--- a/fs/xfs/xfs_trans.h
+++ b/fs/xfs/xfs_trans.h
@@ -64,11 +64,13 @@ typedef struct xfs_log_item {
 } xfs_log_item_t;
 
 #define	XFS_LI_IN_AIL	0x1
-#define XFS_LI_ABORTED	0x2
+#define	XFS_LI_ABORTED	0x2
+#define	XFS_LI_FAILED	0x4
 
 #define XFS_LI_FLAGS \
 	{ XFS_LI_IN_AIL,	"IN_AIL" }, \
-	{ XFS_LI_ABORTED,	"ABORTED" }
+	{ XFS_LI_ABORTED,	"ABORTED" }, \
+	{ XFS_LI_FAILED,	"FAILED" }
 
 struct xfs_item_ops {
 	void (*iop_size)(xfs_log_item_t *, int *, int *);
@@ -79,6 +81,7 @@ struct xfs_item_ops {
 	void (*iop_unlock)(xfs_log_item_t *);
 	xfs_lsn_t (*iop_committed)(xfs_log_item_t *, xfs_lsn_t);
 	void (*iop_committing)(xfs_log_item_t *, xfs_lsn_t);
+	void (*iop_error)(xfs_log_item_t *, xfs_buf_t *);
 };
 
 void	xfs_log_item_init(struct xfs_mount *mp, struct xfs_log_item *item,


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 04/25] xfs: Properly retry failed inode items in case of error during buffer writeback

[Christoph Hellwig]

From: Carlos Maiolino <cmaiolino@redhat.com>

commit d3a304b6292168b83b45d624784f973fdc1ca674 upstream.

When a buffer has been failed during writeback, the inode items into it
are kept flush locked, and are never resubmitted due the flush lock, so,
if any buffer fails to be written, the items in AIL are never written to
disk and never unlocked.

This causes unmount operation to hang due these items flush locked in AIL,
but this also causes the items in AIL to never be written back, even when
the IO device comes back to normal.

I've been testing this patch with a DM-thin device, creating a
filesystem larger than the real device.

When writing enough data to fill the DM-thin device, XFS receives ENOSPC
errors from the device, and keep spinning on xfsaild (when 'retry
forever' configuration is set).

At this point, the filesystem can not be unmounted because of the flush locked
items in AIL, but worse, the items in AIL are never retried at all
(once xfs_inode_item_push() will skip the items that are flush locked),
even if the underlying DM-thin device is expanded to the proper size.

This patch fixes both cases, retrying any item that has been failed
previously, using the infra-structure provided by the previous patch.

Reviewed-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/xfs_buf_item.c   | 28 ++++++++++++++++++++++++++++
 fs/xfs/xfs_buf_item.h   |  3 +++
 fs/xfs/xfs_inode_item.c | 47 +++++++++++++++++++++++++++++++++++++++++++----
 fs/xfs/xfs_trans.h      |  1 +
 fs/xfs/xfs_trans_ail.c  |  3 ++-
 fs/xfs/xfs_trans_priv.h | 31 +++++++++++++++++++++++++++++++
 6 files changed, 108 insertions(+), 5 deletions(-)

diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c
index 7573a1f0bc9a..573fc72c3f23 100644
--- a/fs/xfs/xfs_buf_item.c
+++ b/fs/xfs/xfs_buf_item.c
@@ -1234,3 +1234,31 @@ xfs_buf_iodone(
 	xfs_trans_ail_delete(ailp, lip, SHUTDOWN_CORRUPT_INCORE);
 	xfs_buf_item_free(BUF_ITEM(lip));
 }
+
+/*
+ * Requeue a failed buffer for writeback
+ *
+ * Return true if the buffer has been re-queued properly, false otherwise
+ */
+bool
+xfs_buf_resubmit_failed_buffers(
+	struct xfs_buf		*bp,
+	struct xfs_log_item	*lip,
+	struct list_head	*buffer_list)
+{
+	struct xfs_log_item	*next;
+
+	/*
+	 * Clear XFS_LI_FAILED flag from all items before resubmit
+	 *
+	 * XFS_LI_FAILED set/clear is protected by xa_lock, caller  this
+	 * function already have it acquired
+	 */
+	for (; lip; lip = next) {
+		next = lip->li_bio_list;
+		xfs_clear_li_failed(lip);
+	}
+
+	/* Add this buffer back to the delayed write list */
+	return xfs_buf_delwri_queue(bp, buffer_list);
+}
diff --git a/fs/xfs/xfs_buf_item.h b/fs/xfs/xfs_buf_item.h
index f7eba99d19dd..530686e1afb9 100644
--- a/fs/xfs/xfs_buf_item.h
+++ b/fs/xfs/xfs_buf_item.h
@@ -70,6 +70,9 @@ void	xfs_buf_attach_iodone(struct xfs_buf *,
 			      xfs_log_item_t *);
 void	xfs_buf_iodone_callbacks(struct xfs_buf *);
 void	xfs_buf_iodone(struct xfs_buf *, struct xfs_log_item *);
+bool	xfs_buf_resubmit_failed_buffers(struct xfs_buf *,
+					struct xfs_log_item *,
+					struct list_head *);
 
 extern kmem_zone_t	*xfs_buf_item_zone;
 
diff --git a/fs/xfs/xfs_inode_item.c b/fs/xfs/xfs_inode_item.c
index 013cc78d7daf..6d0f74ec31e8 100644
--- a/fs/xfs/xfs_inode_item.c
+++ b/fs/xfs/xfs_inode_item.c
@@ -27,6 +27,7 @@
 #include "xfs_error.h"
 #include "xfs_trace.h"
 #include "xfs_trans_priv.h"
+#include "xfs_buf_item.h"
 #include "xfs_log.h"
 
 
@@ -475,6 +476,23 @@ xfs_inode_item_unpin(
 		wake_up_bit(&ip->i_flags, __XFS_IPINNED_BIT);
 }
 
+/*
+ * Callback used to mark a buffer with XFS_LI_FAILED when items in the buffer
+ * have been failed during writeback
+ *
+ * This informs the AIL that the inode is already flush locked on the next push,
+ * and acquires a hold on the buffer to ensure that it isn't reclaimed before
+ * dirty data makes it to disk.
+ */
+STATIC void
+xfs_inode_item_error(
+	struct xfs_log_item	*lip,
+	struct xfs_buf		*bp)
+{
+	ASSERT(xfs_isiflocked(INODE_ITEM(lip)->ili_inode));
+	xfs_set_li_failed(lip, bp);
+}
+
 STATIC uint
 xfs_inode_item_push(
 	struct xfs_log_item	*lip,
@@ -484,13 +502,28 @@ xfs_inode_item_push(
 {
 	struct xfs_inode_log_item *iip = INODE_ITEM(lip);
 	struct xfs_inode	*ip = iip->ili_inode;
-	struct xfs_buf		*bp = NULL;
+	struct xfs_buf		*bp = lip->li_buf;
 	uint			rval = XFS_ITEM_SUCCESS;
 	int			error;
 
 	if (xfs_ipincount(ip) > 0)
 		return XFS_ITEM_PINNED;
 
+	/*
+	 * The buffer containing this item failed to be written back
+	 * previously. Resubmit the buffer for IO.
+	 */
+	if (lip->li_flags & XFS_LI_FAILED) {
+		if (!xfs_buf_trylock(bp))
+			return XFS_ITEM_LOCKED;
+
+		if (!xfs_buf_resubmit_failed_buffers(bp, lip, buffer_list))
+			rval = XFS_ITEM_FLUSHING;
+
+		xfs_buf_unlock(bp);
+		return rval;
+	}
+
 	if (!xfs_ilock_nowait(ip, XFS_ILOCK_SHARED))
 		return XFS_ITEM_LOCKED;
 
@@ -622,7 +655,8 @@ static const struct xfs_item_ops xfs_inode_item_ops = {
 	.iop_unlock	= xfs_inode_item_unlock,
 	.iop_committed	= xfs_inode_item_committed,
 	.iop_push	= xfs_inode_item_push,
-	.iop_committing = xfs_inode_item_committing
+	.iop_committing = xfs_inode_item_committing,
+	.iop_error	= xfs_inode_item_error
 };
 
 
@@ -710,7 +744,8 @@ xfs_iflush_done(
 		 * the AIL lock.
 		 */
 		iip = INODE_ITEM(blip);
-		if (iip->ili_logged && blip->li_lsn == iip->ili_flush_lsn)
+		if ((iip->ili_logged && blip->li_lsn == iip->ili_flush_lsn) ||
+		    lip->li_flags & XFS_LI_FAILED)
 			need_ail++;
 
 		blip = next;
@@ -718,7 +753,8 @@ xfs_iflush_done(
 
 	/* make sure we capture the state of the initial inode. */
 	iip = INODE_ITEM(lip);
-	if (iip->ili_logged && lip->li_lsn == iip->ili_flush_lsn)
+	if ((iip->ili_logged && lip->li_lsn == iip->ili_flush_lsn) ||
+	    lip->li_flags & XFS_LI_FAILED)
 		need_ail++;
 
 	/*
@@ -739,6 +775,9 @@ xfs_iflush_done(
 			if (INODE_ITEM(blip)->ili_logged &&
 			    blip->li_lsn == INODE_ITEM(blip)->ili_flush_lsn)
 				mlip_changed |= xfs_ail_delete_one(ailp, blip);
+			else {
+				xfs_clear_li_failed(blip);
+			}
 		}
 
 		if (mlip_changed) {
diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h
index 442d679210a1..7d627721e4b3 100644
--- a/fs/xfs/xfs_trans.h
+++ b/fs/xfs/xfs_trans.h
@@ -49,6 +49,7 @@ typedef struct xfs_log_item {
 	struct xfs_ail			*li_ailp;	/* ptr to AIL */
 	uint				li_type;	/* item type */
 	uint				li_flags;	/* misc flags */
+	struct xfs_buf			*li_buf;	/* real buffer pointer */
 	struct xfs_log_item		*li_bio_list;	/* buffer item list */
 	void				(*li_cb)(struct xfs_buf *,
 						 struct xfs_log_item *);
diff --git a/fs/xfs/xfs_trans_ail.c b/fs/xfs/xfs_trans_ail.c
index 9056c0f34a3c..70f5ab017323 100644
--- a/fs/xfs/xfs_trans_ail.c
+++ b/fs/xfs/xfs_trans_ail.c
@@ -687,12 +687,13 @@ xfs_trans_ail_update_bulk(
 bool
 xfs_ail_delete_one(
 	struct xfs_ail		*ailp,
-	struct xfs_log_item 	*lip)
+	struct xfs_log_item	*lip)
 {
 	struct xfs_log_item	*mlip = xfs_ail_min(ailp);
 
 	trace_xfs_ail_delete(lip, mlip->li_lsn, lip->li_lsn);
 	xfs_ail_delete(ailp, lip);
+	xfs_clear_li_failed(lip);
 	lip->li_flags &= ~XFS_LI_IN_AIL;
 	lip->li_lsn = 0;
 
diff --git a/fs/xfs/xfs_trans_priv.h b/fs/xfs/xfs_trans_priv.h
index d91706c56c63..b317a3644c00 100644
--- a/fs/xfs/xfs_trans_priv.h
+++ b/fs/xfs/xfs_trans_priv.h
@@ -164,4 +164,35 @@ xfs_trans_ail_copy_lsn(
 	*dst = *src;
 }
 #endif
+
+static inline void
+xfs_clear_li_failed(
+	struct xfs_log_item	*lip)
+{
+	struct xfs_buf	*bp = lip->li_buf;
+
+	ASSERT(lip->li_flags & XFS_LI_IN_AIL);
+	lockdep_assert_held(&lip->li_ailp->xa_lock);
+
+	if (lip->li_flags & XFS_LI_FAILED) {
+		lip->li_flags &= ~XFS_LI_FAILED;
+		lip->li_buf = NULL;
+		xfs_buf_rele(bp);
+	}
+}
+
+static inline void
+xfs_set_li_failed(
+	struct xfs_log_item	*lip,
+	struct xfs_buf		*bp)
+{
+	lockdep_assert_held(&lip->li_ailp->xa_lock);
+
+	if (!(lip->li_flags & XFS_LI_FAILED)) {
+		xfs_buf_hold(bp);
+		lip->li_flags |= XFS_LI_FAILED;
+		lip->li_buf = bp;
+	}
+}
+
 #endif	/* __XFS_TRANS_PRIV_H__ */
-- 
2.14.1

Patch "xfs: Properly retry failed inode items in case of error during buffer writeback" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: Properly retry failed inode items in case of error during buffer writeback

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:10 -0700
Subject: xfs: Properly retry failed inode items in case of error during buffer writeback
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, Carlos Maiolino <cmaiolino@redhat.com>, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-5-hch@lst.de>

From: Carlos Maiolino <cmaiolino@redhat.com>

commit d3a304b6292168b83b45d624784f973fdc1ca674 upstream.

When a buffer has been failed during writeback, the inode items into it
are kept flush locked, and are never resubmitted due the flush lock, so,
if any buffer fails to be written, the items in AIL are never written to
disk and never unlocked.

This causes unmount operation to hang due these items flush locked in AIL,
but this also causes the items in AIL to never be written back, even when
the IO device comes back to normal.

I've been testing this patch with a DM-thin device, creating a
filesystem larger than the real device.

When writing enough data to fill the DM-thin device, XFS receives ENOSPC
errors from the device, and keep spinning on xfsaild (when 'retry
forever' configuration is set).

At this point, the filesystem can not be unmounted because of the flush locked
items in AIL, but worse, the items in AIL are never retried at all
(once xfs_inode_item_push() will skip the items that are flush locked),
even if the underlying DM-thin device is expanded to the proper size.

This patch fixes both cases, retrying any item that has been failed
previously, using the infra-structure provided by the previous patch.

Reviewed-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/xfs_buf_item.c   |   28 ++++++++++++++++++++++++++++
 fs/xfs/xfs_buf_item.h   |    3 +++
 fs/xfs/xfs_inode_item.c |   47 +++++++++++++++++++++++++++++++++++++++++++----
 fs/xfs/xfs_trans.h      |    1 +
 fs/xfs/xfs_trans_ail.c  |    3 ++-
 fs/xfs/xfs_trans_priv.h |   31 +++++++++++++++++++++++++++++++
 6 files changed, 108 insertions(+), 5 deletions(-)

--- a/fs/xfs/xfs_buf_item.c
+++ b/fs/xfs/xfs_buf_item.c
@@ -1234,3 +1234,31 @@ xfs_buf_iodone(
 	xfs_trans_ail_delete(ailp, lip, SHUTDOWN_CORRUPT_INCORE);
 	xfs_buf_item_free(BUF_ITEM(lip));
 }
+
+/*
+ * Requeue a failed buffer for writeback
+ *
+ * Return true if the buffer has been re-queued properly, false otherwise
+ */
+bool
+xfs_buf_resubmit_failed_buffers(
+	struct xfs_buf		*bp,
+	struct xfs_log_item	*lip,
+	struct list_head	*buffer_list)
+{
+	struct xfs_log_item	*next;
+
+	/*
+	 * Clear XFS_LI_FAILED flag from all items before resubmit
+	 *
+	 * XFS_LI_FAILED set/clear is protected by xa_lock, caller  this
+	 * function already have it acquired
+	 */
+	for (; lip; lip = next) {
+		next = lip->li_bio_list;
+		xfs_clear_li_failed(lip);
+	}
+
+	/* Add this buffer back to the delayed write list */
+	return xfs_buf_delwri_queue(bp, buffer_list);
+}
--- a/fs/xfs/xfs_buf_item.h
+++ b/fs/xfs/xfs_buf_item.h
@@ -70,6 +70,9 @@ void	xfs_buf_attach_iodone(struct xfs_bu
 			      xfs_log_item_t *);
 void	xfs_buf_iodone_callbacks(struct xfs_buf *);
 void	xfs_buf_iodone(struct xfs_buf *, struct xfs_log_item *);
+bool	xfs_buf_resubmit_failed_buffers(struct xfs_buf *,
+					struct xfs_log_item *,
+					struct list_head *);
 
 extern kmem_zone_t	*xfs_buf_item_zone;
 
--- a/fs/xfs/xfs_inode_item.c
+++ b/fs/xfs/xfs_inode_item.c
@@ -27,6 +27,7 @@
 #include "xfs_error.h"
 #include "xfs_trace.h"
 #include "xfs_trans_priv.h"
+#include "xfs_buf_item.h"
 #include "xfs_log.h"
 
 
@@ -475,6 +476,23 @@ xfs_inode_item_unpin(
 		wake_up_bit(&ip->i_flags, __XFS_IPINNED_BIT);
 }
 
+/*
+ * Callback used to mark a buffer with XFS_LI_FAILED when items in the buffer
+ * have been failed during writeback
+ *
+ * This informs the AIL that the inode is already flush locked on the next push,
+ * and acquires a hold on the buffer to ensure that it isn't reclaimed before
+ * dirty data makes it to disk.
+ */
+STATIC void
+xfs_inode_item_error(
+	struct xfs_log_item	*lip,
+	struct xfs_buf		*bp)
+{
+	ASSERT(xfs_isiflocked(INODE_ITEM(lip)->ili_inode));
+	xfs_set_li_failed(lip, bp);
+}
+
 STATIC uint
 xfs_inode_item_push(
 	struct xfs_log_item	*lip,
@@ -484,13 +502,28 @@ xfs_inode_item_push(
 {
 	struct xfs_inode_log_item *iip = INODE_ITEM(lip);
 	struct xfs_inode	*ip = iip->ili_inode;
-	struct xfs_buf		*bp = NULL;
+	struct xfs_buf		*bp = lip->li_buf;
 	uint			rval = XFS_ITEM_SUCCESS;
 	int			error;
 
 	if (xfs_ipincount(ip) > 0)
 		return XFS_ITEM_PINNED;
 
+	/*
+	 * The buffer containing this item failed to be written back
+	 * previously. Resubmit the buffer for IO.
+	 */
+	if (lip->li_flags & XFS_LI_FAILED) {
+		if (!xfs_buf_trylock(bp))
+			return XFS_ITEM_LOCKED;
+
+		if (!xfs_buf_resubmit_failed_buffers(bp, lip, buffer_list))
+			rval = XFS_ITEM_FLUSHING;
+
+		xfs_buf_unlock(bp);
+		return rval;
+	}
+
 	if (!xfs_ilock_nowait(ip, XFS_ILOCK_SHARED))
 		return XFS_ITEM_LOCKED;
 
@@ -622,7 +655,8 @@ static const struct xfs_item_ops xfs_ino
 	.iop_unlock	= xfs_inode_item_unlock,
 	.iop_committed	= xfs_inode_item_committed,
 	.iop_push	= xfs_inode_item_push,
-	.iop_committing = xfs_inode_item_committing
+	.iop_committing = xfs_inode_item_committing,
+	.iop_error	= xfs_inode_item_error
 };
 
 
@@ -710,7 +744,8 @@ xfs_iflush_done(
 		 * the AIL lock.
 		 */
 		iip = INODE_ITEM(blip);
-		if (iip->ili_logged && blip->li_lsn == iip->ili_flush_lsn)
+		if ((iip->ili_logged && blip->li_lsn == iip->ili_flush_lsn) ||
+		    lip->li_flags & XFS_LI_FAILED)
 			need_ail++;
 
 		blip = next;
@@ -718,7 +753,8 @@ xfs_iflush_done(
 
 	/* make sure we capture the state of the initial inode. */
 	iip = INODE_ITEM(lip);
-	if (iip->ili_logged && lip->li_lsn == iip->ili_flush_lsn)
+	if ((iip->ili_logged && lip->li_lsn == iip->ili_flush_lsn) ||
+	    lip->li_flags & XFS_LI_FAILED)
 		need_ail++;
 
 	/*
@@ -739,6 +775,9 @@ xfs_iflush_done(
 			if (INODE_ITEM(blip)->ili_logged &&
 			    blip->li_lsn == INODE_ITEM(blip)->ili_flush_lsn)
 				mlip_changed |= xfs_ail_delete_one(ailp, blip);
+			else {
+				xfs_clear_li_failed(blip);
+			}
 		}
 
 		if (mlip_changed) {
--- a/fs/xfs/xfs_trans.h
+++ b/fs/xfs/xfs_trans.h
@@ -49,6 +49,7 @@ typedef struct xfs_log_item {
 	struct xfs_ail			*li_ailp;	/* ptr to AIL */
 	uint				li_type;	/* item type */
 	uint				li_flags;	/* misc flags */
+	struct xfs_buf			*li_buf;	/* real buffer pointer */
 	struct xfs_log_item		*li_bio_list;	/* buffer item list */
 	void				(*li_cb)(struct xfs_buf *,
 						 struct xfs_log_item *);
--- a/fs/xfs/xfs_trans_ail.c
+++ b/fs/xfs/xfs_trans_ail.c
@@ -687,12 +687,13 @@ xfs_trans_ail_update_bulk(
 bool
 xfs_ail_delete_one(
 	struct xfs_ail		*ailp,
-	struct xfs_log_item 	*lip)
+	struct xfs_log_item	*lip)
 {
 	struct xfs_log_item	*mlip = xfs_ail_min(ailp);
 
 	trace_xfs_ail_delete(lip, mlip->li_lsn, lip->li_lsn);
 	xfs_ail_delete(ailp, lip);
+	xfs_clear_li_failed(lip);
 	lip->li_flags &= ~XFS_LI_IN_AIL;
 	lip->li_lsn = 0;
 
--- a/fs/xfs/xfs_trans_priv.h
+++ b/fs/xfs/xfs_trans_priv.h
@@ -164,4 +164,35 @@ xfs_trans_ail_copy_lsn(
 	*dst = *src;
 }
 #endif
+
+static inline void
+xfs_clear_li_failed(
+	struct xfs_log_item	*lip)
+{
+	struct xfs_buf	*bp = lip->li_buf;
+
+	ASSERT(lip->li_flags & XFS_LI_IN_AIL);
+	lockdep_assert_held(&lip->li_ailp->xa_lock);
+
+	if (lip->li_flags & XFS_LI_FAILED) {
+		lip->li_flags &= ~XFS_LI_FAILED;
+		lip->li_buf = NULL;
+		xfs_buf_rele(bp);
+	}
+}
+
+static inline void
+xfs_set_li_failed(
+	struct xfs_log_item	*lip,
+	struct xfs_buf		*bp)
+{
+	lockdep_assert_held(&lip->li_ailp->xa_lock);
+
+	if (!(lip->li_flags & XFS_LI_FAILED)) {
+		xfs_buf_hold(bp);
+		lip->li_flags |= XFS_LI_FAILED;
+		lip->li_buf = bp;
+	}
+}
+
 #endif	/* __XFS_TRANS_PRIV_H__ */


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 05/25] xfs: fix recovery failure when log record header wraps log end

[Christoph Hellwig]

From: Brian Foster <bfoster@redhat.com>

commit 284f1c2c9bebf871861184b0e2c40fa921dd380b upstream.

The high-level log recovery algorithm consists of two loops that
walk the physical log and process log records from the tail to the
head. The first loop handles the case where the tail is beyond the
head and processes records up to the end of the physical log. The
subsequent loop processes records from the beginning of the physical
log to the head.

Because log records can wrap around the end of the physical log, the
first loop mentioned above must handle this case appropriately.
Records are processed from in-core buffers, which means that this
algorithm must split the reads of such records into two partial
I/Os: 1.) from the beginning of the record to the end of the log and
2.) from the beginning of the log to the end of the record. This is
further complicated by the fact that the log record header and log
record data are read into independent buffers.

The current handling of each buffer correctly splits the reads when
either the header or data starts before the end of the log and wraps
around the end. The data read does not correctly handle the case
where the prior header read wrapped or ends on the physical log end
boundary. blk_no is incremented to or beyond the log end after the
header read to point to the record data, but the split data read
logic triggers, attempts to read from an invalid log block and
ultimately causes log recovery to fail. This can be reproduced
fairly reliably via xfstests tests generic/047 and generic/388 with
large iclog sizes (256k) and small (10M) logs.

If the record header read has pushed beyond the end of the physical
log, the subsequent data read is actually contiguous. Update the
data read logic to detect the case where blk_no has wrapped, mod it
against the log size to read from the correct address and issue one
contiguous read for the log data buffer. The log record is processed
as normal from the buffer(s), the loop exits after the current
iteration and the subsequent loop picks up with the first new record
after the start of the log.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/xfs_log_recover.c | 18 ++++++++++++++----
 1 file changed, 14 insertions(+), 4 deletions(-)

diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index 9549188f5a36..36a179f2c931 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -5218,7 +5218,7 @@ xlog_do_recovery_pass(
 	xfs_daddr_t		*first_bad)	/* out: first bad log rec */
 {
 	xlog_rec_header_t	*rhead;
-	xfs_daddr_t		blk_no;
+	xfs_daddr_t		blk_no, rblk_no;
 	xfs_daddr_t		rhead_blk;
 	char			*offset;
 	xfs_buf_t		*hbp, *dbp;
@@ -5371,9 +5371,19 @@ xlog_do_recovery_pass(
 			bblks = (int)BTOBB(be32_to_cpu(rhead->h_len));
 			blk_no += hblks;
 
-			/* Read in data for log record */
-			if (blk_no + bblks <= log->l_logBBsize) {
-				error = xlog_bread(log, blk_no, bblks, dbp,
+			/*
+			 * Read the log record data in multiple reads if it
+			 * wraps around the end of the log. Note that if the
+			 * header already wrapped, blk_no could point past the
+			 * end of the log. The record data is contiguous in
+			 * that case.
+			 */
+			if (blk_no + bblks <= log->l_logBBsize ||
+			    blk_no >= log->l_logBBsize) {
+				/* mod blk_no in case the header wrapped and
+				 * pushed it beyond the end of the log */
+				rblk_no = do_mod(blk_no, log->l_logBBsize);
+				error = xlog_bread(log, rblk_no, bblks, dbp,
 						   &offset);
 				if (error)
 					goto bread_err2;
-- 
2.14.1

Patch "xfs: fix recovery failure when log record header wraps log end" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: fix recovery failure when log record header wraps log end

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:11 -0700
Subject: xfs: fix recovery failure when log record header wraps log end
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, Brian Foster <bfoster@redhat.com>, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-6-hch@lst.de>

From: Brian Foster <bfoster@redhat.com>

commit 284f1c2c9bebf871861184b0e2c40fa921dd380b upstream.

The high-level log recovery algorithm consists of two loops that
walk the physical log and process log records from the tail to the
head. The first loop handles the case where the tail is beyond the
head and processes records up to the end of the physical log. The
subsequent loop processes records from the beginning of the physical
log to the head.

Because log records can wrap around the end of the physical log, the
first loop mentioned above must handle this case appropriately.
Records are processed from in-core buffers, which means that this
algorithm must split the reads of such records into two partial
I/Os: 1.) from the beginning of the record to the end of the log and
2.) from the beginning of the log to the end of the record. This is
further complicated by the fact that the log record header and log
record data are read into independent buffers.

The current handling of each buffer correctly splits the reads when
either the header or data starts before the end of the log and wraps
around the end. The data read does not correctly handle the case
where the prior header read wrapped or ends on the physical log end
boundary. blk_no is incremented to or beyond the log end after the
header read to point to the record data, but the split data read
logic triggers, attempts to read from an invalid log block and
ultimately causes log recovery to fail. This can be reproduced
fairly reliably via xfstests tests generic/047 and generic/388 with
large iclog sizes (256k) and small (10M) logs.

If the record header read has pushed beyond the end of the physical
log, the subsequent data read is actually contiguous. Update the
data read logic to detect the case where blk_no has wrapped, mod it
against the log size to read from the correct address and issue one
contiguous read for the log data buffer. The log record is processed
as normal from the buffer(s), the loop exits after the current
iteration and the subsequent loop picks up with the first new record
after the start of the log.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/xfs_log_recover.c |   18 ++++++++++++++----
 1 file changed, 14 insertions(+), 4 deletions(-)

--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -5218,7 +5218,7 @@ xlog_do_recovery_pass(
 	xfs_daddr_t		*first_bad)	/* out: first bad log rec */
 {
 	xlog_rec_header_t	*rhead;
-	xfs_daddr_t		blk_no;
+	xfs_daddr_t		blk_no, rblk_no;
 	xfs_daddr_t		rhead_blk;
 	char			*offset;
 	xfs_buf_t		*hbp, *dbp;
@@ -5371,9 +5371,19 @@ xlog_do_recovery_pass(
 			bblks = (int)BTOBB(be32_to_cpu(rhead->h_len));
 			blk_no += hblks;
 
-			/* Read in data for log record */
-			if (blk_no + bblks <= log->l_logBBsize) {
-				error = xlog_bread(log, blk_no, bblks, dbp,
+			/*
+			 * Read the log record data in multiple reads if it
+			 * wraps around the end of the log. Note that if the
+			 * header already wrapped, blk_no could point past the
+			 * end of the log. The record data is contiguous in
+			 * that case.
+			 */
+			if (blk_no + bblks <= log->l_logBBsize ||
+			    blk_no >= log->l_logBBsize) {
+				/* mod blk_no in case the header wrapped and
+				 * pushed it beyond the end of the log */
+				rblk_no = do_mod(blk_no, log->l_logBBsize);
+				error = xlog_bread(log, rblk_no, bblks, dbp,
 						   &offset);
 				if (error)
 					goto bread_err2;


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 06/25] xfs: always verify the log tail during recovery

[Christoph Hellwig]

From: Brian Foster <bfoster@redhat.com>

commit 5297ac1f6d7cbf45464a49b9558831f271dfc559 upstream.

Log tail verification currently only occurs when torn writes are
detected at the head of the log. This was introduced because a
change in the head block due to torn writes can lead to a change in
the tail block (each log record header references the current tail)
and the tail block should be verified before log recovery proceeds.

Tail corruption is possible outside of torn write scenarios,
however. For example, partial log writes can be detected and cleared
during the initial head/tail block discovery process. If the partial
write coincides with a tail overwrite, the log tail is corrupted and
recovery fails.

To facilitate correct handling of log tail overwites, update log
recovery to always perform tail verification. This is necessary to
detect potential tail overwrite conditions when torn writes may not
have occurred. This changes normal (i.e., no torn writes) recovery
behavior slightly to detect and return CRC related errors near the
tail before actual recovery starts.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/xfs_log_recover.c | 26 +++-----------------------
 1 file changed, 3 insertions(+), 23 deletions(-)

diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index 36a179f2c931..c1337127d8fb 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -1183,31 +1183,11 @@ xlog_verify_head(
 			ASSERT(0);
 			return 0;
 		}
-
-		/*
-		 * Now verify the tail based on the updated head. This is
-		 * required because the torn writes trimmed from the head could
-		 * have been written over the tail of a previous record. Return
-		 * any errors since recovery cannot proceed if the tail is
-		 * corrupt.
-		 *
-		 * XXX: This leaves a gap in truly robust protection from torn
-		 * writes in the log. If the head is behind the tail, the tail
-		 * pushes forward to create some space and then a crash occurs
-		 * causing the writes into the previous record's tail region to
-		 * tear, log recovery isn't able to recover.
-		 *
-		 * How likely is this to occur? If possible, can we do something
-		 * more intelligent here? Is it safe to push the tail forward if
-		 * we can determine that the tail is within the range of the
-		 * torn write (e.g., the kernel can only overwrite the tail if
-		 * it has actually been pushed forward)? Alternatively, could we
-		 * somehow prevent this condition at runtime?
-		 */
-		error = xlog_verify_tail(log, *head_blk, *tail_blk);
 	}
+	if (error)
+		return error;
 
-	return error;
+	return xlog_verify_tail(log, *head_blk, *tail_blk);
 }
 
 /*
-- 
2.14.1

Patch "xfs: always verify the log tail during recovery" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: always verify the log tail during recovery

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-always-verify-the-log-tail-during-recovery.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:12 -0700
Subject: xfs: always verify the log tail during recovery
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, Brian Foster <bfoster@redhat.com>, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-7-hch@lst.de>

From: Brian Foster <bfoster@redhat.com>

commit 5297ac1f6d7cbf45464a49b9558831f271dfc559 upstream.

Log tail verification currently only occurs when torn writes are
detected at the head of the log. This was introduced because a
change in the head block due to torn writes can lead to a change in
the tail block (each log record header references the current tail)
and the tail block should be verified before log recovery proceeds.

Tail corruption is possible outside of torn write scenarios,
however. For example, partial log writes can be detected and cleared
during the initial head/tail block discovery process. If the partial
write coincides with a tail overwrite, the log tail is corrupted and
recovery fails.

To facilitate correct handling of log tail overwites, update log
recovery to always perform tail verification. This is necessary to
detect potential tail overwrite conditions when torn writes may not
have occurred. This changes normal (i.e., no torn writes) recovery
behavior slightly to detect and return CRC related errors near the
tail before actual recovery starts.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/xfs_log_recover.c |   26 +++-----------------------
 1 file changed, 3 insertions(+), 23 deletions(-)

--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -1183,31 +1183,11 @@ xlog_verify_head(
 			ASSERT(0);
 			return 0;
 		}
-
-		/*
-		 * Now verify the tail based on the updated head. This is
-		 * required because the torn writes trimmed from the head could
-		 * have been written over the tail of a previous record. Return
-		 * any errors since recovery cannot proceed if the tail is
-		 * corrupt.
-		 *
-		 * XXX: This leaves a gap in truly robust protection from torn
-		 * writes in the log. If the head is behind the tail, the tail
-		 * pushes forward to create some space and then a crash occurs
-		 * causing the writes into the previous record's tail region to
-		 * tear, log recovery isn't able to recover.
-		 *
-		 * How likely is this to occur? If possible, can we do something
-		 * more intelligent here? Is it safe to push the tail forward if
-		 * we can determine that the tail is within the range of the
-		 * torn write (e.g., the kernel can only overwrite the tail if
-		 * it has actually been pushed forward)? Alternatively, could we
-		 * somehow prevent this condition at runtime?
-		 */
-		error = xlog_verify_tail(log, *head_blk, *tail_blk);
 	}
+	if (error)
+		return error;
 
-	return error;
+	return xlog_verify_tail(log, *head_blk, *tail_blk);
 }
 
 /*


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 07/25] xfs: fix log recovery corruption error due to tail overwrite

[Christoph Hellwig]

From: Brian Foster <bfoster@redhat.com>

commit 4a4f66eac4681378996a1837ad1ffec3a2e2981f upstream.

If we consider the case where the tail (T) of the log is pinned long
enough for the head (H) to push and block behind the tail, we can
end up blocked in the following state without enough free space (f)
in the log to satisfy a transaction reservation:

	0	phys. log	N
	[-------HffT---H'--T'---]

The last good record in the log (before H) refers to T. The tail
eventually pushes forward (T') leaving more free space in the log
for writes to H. At this point, suppose space frees up in the log
for the maximum of 8 in-core log buffers to start flushing out to
the log. If this pushes the head from H to H', these next writes
overwrite the previous tail T. This is safe because the items logged
from T to T' have been written back and removed from the AIL.

If the next log writes (H -> H') happen to fail and result in
partial records in the log, the filesystem shuts down having
overwritten T with invalid data. Log recovery correctly locates H on
the subsequent mount, but H still refers to the now corrupted tail
T. This results in log corruption errors and recovery failure.

Since the tail overwrite results from otherwise correct runtime
behavior, it is up to log recovery to try and deal with this
situation. Update log recovery tail verification to run a CRC pass
from the first record past the tail to the head. This facilitates
error detection at T and moves the recovery tail to the first good
record past H' (similar to truncating the head on torn write
detection). If corruption is detected beyond the range possibly
affected by the max number of iclogs, the log is legitimately
corrupted and log recovery failure is expected.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/xfs_log_recover.c | 108 +++++++++++++++++++++++++++++++++--------------
 1 file changed, 77 insertions(+), 31 deletions(-)

diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index c1337127d8fb..a5e2ca8f5cd6 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -1029,61 +1029,106 @@ xlog_seek_logrec_hdr(
 }
 
 /*
- * Check the log tail for torn writes. This is required when torn writes are
- * detected at the head and the head had to be walked back to a previous record.
- * The tail of the previous record must now be verified to ensure the torn
- * writes didn't corrupt the previous tail.
+ * Calculate distance from head to tail (i.e., unused space in the log).
+ */
+static inline int
+xlog_tail_distance(
+	struct xlog	*log,
+	xfs_daddr_t	head_blk,
+	xfs_daddr_t	tail_blk)
+{
+	if (head_blk < tail_blk)
+		return tail_blk - head_blk;
+
+	return tail_blk + (log->l_logBBsize - head_blk);
+}
+
+/*
+ * Verify the log tail. This is particularly important when torn or incomplete
+ * writes have been detected near the front of the log and the head has been
+ * walked back accordingly.
+ *
+ * We also have to handle the case where the tail was pinned and the head
+ * blocked behind the tail right before a crash. If the tail had been pushed
+ * immediately prior to the crash and the subsequent checkpoint was only
+ * partially written, it's possible it overwrote the last referenced tail in the
+ * log with garbage. This is not a coherency problem because the tail must have
+ * been pushed before it can be overwritten, but appears as log corruption to
+ * recovery because we have no way to know the tail was updated if the
+ * subsequent checkpoint didn't write successfully.
  *
- * Return an error if CRC verification fails as recovery cannot proceed.
+ * Therefore, CRC check the log from tail to head. If a failure occurs and the
+ * offending record is within max iclog bufs from the head, walk the tail
+ * forward and retry until a valid tail is found or corruption is detected out
+ * of the range of a possible overwrite.
  */
 STATIC int
 xlog_verify_tail(
 	struct xlog		*log,
 	xfs_daddr_t		head_blk,
-	xfs_daddr_t		tail_blk)
+	xfs_daddr_t		*tail_blk,
+	int			hsize)
 {
 	struct xlog_rec_header	*thead;
 	struct xfs_buf		*bp;
 	xfs_daddr_t		first_bad;
-	int			count;
 	int			error = 0;
 	bool			wrapped;
-	xfs_daddr_t		tmp_head;
+	xfs_daddr_t		tmp_tail;
+	xfs_daddr_t		orig_tail = *tail_blk;
 
 	bp = xlog_get_bp(log, 1);
 	if (!bp)
 		return -ENOMEM;
 
 	/*
-	 * Seek XLOG_MAX_ICLOGS + 1 records past the current tail record to get
-	 * a temporary head block that points after the last possible
-	 * concurrently written record of the tail.
+	 * Make sure the tail points to a record (returns positive count on
+	 * success).
 	 */
-	count = xlog_seek_logrec_hdr(log, head_blk, tail_blk,
-				     XLOG_MAX_ICLOGS + 1, bp, &tmp_head, &thead,
-				     &wrapped);
-	if (count < 0) {
-		error = count;
+	error = xlog_seek_logrec_hdr(log, head_blk, *tail_blk, 1, bp,
+			&tmp_tail, &thead, &wrapped);
+	if (error < 0)
 		goto out;
-	}
+	if (*tail_blk != tmp_tail)
+		*tail_blk = tmp_tail;
 
 	/*
-	 * If the call above didn't find XLOG_MAX_ICLOGS + 1 records, we ran
-	 * into the actual log head. tmp_head points to the start of the record
-	 * so update it to the actual head block.
+	 * Run a CRC check from the tail to the head. We can't just check
+	 * MAX_ICLOGS records past the tail because the tail may point to stale
+	 * blocks cleared during the search for the head/tail. These blocks are
+	 * overwritten with zero-length records and thus record count is not a
+	 * reliable indicator of the iclog state before a crash.
 	 */
-	if (count < XLOG_MAX_ICLOGS + 1)
-		tmp_head = head_blk;
-
-	/*
-	 * We now have a tail and temporary head block that covers at least
-	 * XLOG_MAX_ICLOGS records from the tail. We need to verify that these
-	 * records were completely written. Run a CRC verification pass from
-	 * tail to head and return the result.
-	 */
-	error = xlog_do_recovery_pass(log, tmp_head, tail_blk,
+	first_bad = 0;
+	error = xlog_do_recovery_pass(log, head_blk, *tail_blk,
 				      XLOG_RECOVER_CRCPASS, &first_bad);
+	while (error == -EFSBADCRC && first_bad) {
+		int	tail_distance;
+
+		/*
+		 * Is corruption within range of the head? If so, retry from
+		 * the next record. Otherwise return an error.
+		 */
+		tail_distance = xlog_tail_distance(log, head_blk, first_bad);
+		if (tail_distance > BTOBB(XLOG_MAX_ICLOGS * hsize))
+			break;
 
+		/* skip to the next record; returns positive count on success */
+		error = xlog_seek_logrec_hdr(log, head_blk, first_bad, 2, bp,
+				&tmp_tail, &thead, &wrapped);
+		if (error < 0)
+			goto out;
+
+		*tail_blk = tmp_tail;
+		first_bad = 0;
+		error = xlog_do_recovery_pass(log, head_blk, *tail_blk,
+					      XLOG_RECOVER_CRCPASS, &first_bad);
+	}
+
+	if (!error && *tail_blk != orig_tail)
+		xfs_warn(log->l_mp,
+		"Tail block (0x%llx) overwrite detected. Updated to 0x%llx",
+			 orig_tail, *tail_blk);
 out:
 	xlog_put_bp(bp);
 	return error;
@@ -1187,7 +1232,8 @@ xlog_verify_head(
 	if (error)
 		return error;
 
-	return xlog_verify_tail(log, *head_blk, *tail_blk);
+	return xlog_verify_tail(log, *head_blk, tail_blk,
+				be32_to_cpu((*rhead)->h_size));
 }
 
 /*
-- 
2.14.1

Patch "xfs: fix log recovery corruption error due to tail overwrite" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: fix log recovery corruption error due to tail overwrite

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:13 -0700
Subject: xfs: fix log recovery corruption error due to tail overwrite
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, Brian Foster <bfoster@redhat.com>, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-8-hch@lst.de>

From: Brian Foster <bfoster@redhat.com>

commit 4a4f66eac4681378996a1837ad1ffec3a2e2981f upstream.

If we consider the case where the tail (T) of the log is pinned long
enough for the head (H) to push and block behind the tail, we can
end up blocked in the following state without enough free space (f)
in the log to satisfy a transaction reservation:

	0	phys. log	N
	[-------HffT---H'--T'---]

The last good record in the log (before H) refers to T. The tail
eventually pushes forward (T') leaving more free space in the log
for writes to H. At this point, suppose space frees up in the log
for the maximum of 8 in-core log buffers to start flushing out to
the log. If this pushes the head from H to H', these next writes
overwrite the previous tail T. This is safe because the items logged
from T to T' have been written back and removed from the AIL.

If the next log writes (H -> H') happen to fail and result in
partial records in the log, the filesystem shuts down having
overwritten T with invalid data. Log recovery correctly locates H on
the subsequent mount, but H still refers to the now corrupted tail
T. This results in log corruption errors and recovery failure.

Since the tail overwrite results from otherwise correct runtime
behavior, it is up to log recovery to try and deal with this
situation. Update log recovery tail verification to run a CRC pass
from the first record past the tail to the head. This facilitates
error detection at T and moves the recovery tail to the first good
record past H' (similar to truncating the head on torn write
detection). If corruption is detected beyond the range possibly
affected by the max number of iclogs, the log is legitimately
corrupted and log recovery failure is expected.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/xfs_log_recover.c |  108 +++++++++++++++++++++++++++++++++--------------
 1 file changed, 77 insertions(+), 31 deletions(-)

--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -1029,61 +1029,106 @@ out_error:
 }
 
 /*
- * Check the log tail for torn writes. This is required when torn writes are
- * detected at the head and the head had to be walked back to a previous record.
- * The tail of the previous record must now be verified to ensure the torn
- * writes didn't corrupt the previous tail.
+ * Calculate distance from head to tail (i.e., unused space in the log).
+ */
+static inline int
+xlog_tail_distance(
+	struct xlog	*log,
+	xfs_daddr_t	head_blk,
+	xfs_daddr_t	tail_blk)
+{
+	if (head_blk < tail_blk)
+		return tail_blk - head_blk;
+
+	return tail_blk + (log->l_logBBsize - head_blk);
+}
+
+/*
+ * Verify the log tail. This is particularly important when torn or incomplete
+ * writes have been detected near the front of the log and the head has been
+ * walked back accordingly.
+ *
+ * We also have to handle the case where the tail was pinned and the head
+ * blocked behind the tail right before a crash. If the tail had been pushed
+ * immediately prior to the crash and the subsequent checkpoint was only
+ * partially written, it's possible it overwrote the last referenced tail in the
+ * log with garbage. This is not a coherency problem because the tail must have
+ * been pushed before it can be overwritten, but appears as log corruption to
+ * recovery because we have no way to know the tail was updated if the
+ * subsequent checkpoint didn't write successfully.
  *
- * Return an error if CRC verification fails as recovery cannot proceed.
+ * Therefore, CRC check the log from tail to head. If a failure occurs and the
+ * offending record is within max iclog bufs from the head, walk the tail
+ * forward and retry until a valid tail is found or corruption is detected out
+ * of the range of a possible overwrite.
  */
 STATIC int
 xlog_verify_tail(
 	struct xlog		*log,
 	xfs_daddr_t		head_blk,
-	xfs_daddr_t		tail_blk)
+	xfs_daddr_t		*tail_blk,
+	int			hsize)
 {
 	struct xlog_rec_header	*thead;
 	struct xfs_buf		*bp;
 	xfs_daddr_t		first_bad;
-	int			count;
 	int			error = 0;
 	bool			wrapped;
-	xfs_daddr_t		tmp_head;
+	xfs_daddr_t		tmp_tail;
+	xfs_daddr_t		orig_tail = *tail_blk;
 
 	bp = xlog_get_bp(log, 1);
 	if (!bp)
 		return -ENOMEM;
 
 	/*
-	 * Seek XLOG_MAX_ICLOGS + 1 records past the current tail record to get
-	 * a temporary head block that points after the last possible
-	 * concurrently written record of the tail.
+	 * Make sure the tail points to a record (returns positive count on
+	 * success).
 	 */
-	count = xlog_seek_logrec_hdr(log, head_blk, tail_blk,
-				     XLOG_MAX_ICLOGS + 1, bp, &tmp_head, &thead,
-				     &wrapped);
-	if (count < 0) {
-		error = count;
+	error = xlog_seek_logrec_hdr(log, head_blk, *tail_blk, 1, bp,
+			&tmp_tail, &thead, &wrapped);
+	if (error < 0)
 		goto out;
-	}
+	if (*tail_blk != tmp_tail)
+		*tail_blk = tmp_tail;
 
 	/*
-	 * If the call above didn't find XLOG_MAX_ICLOGS + 1 records, we ran
-	 * into the actual log head. tmp_head points to the start of the record
-	 * so update it to the actual head block.
+	 * Run a CRC check from the tail to the head. We can't just check
+	 * MAX_ICLOGS records past the tail because the tail may point to stale
+	 * blocks cleared during the search for the head/tail. These blocks are
+	 * overwritten with zero-length records and thus record count is not a
+	 * reliable indicator of the iclog state before a crash.
 	 */
-	if (count < XLOG_MAX_ICLOGS + 1)
-		tmp_head = head_blk;
-
-	/*
-	 * We now have a tail and temporary head block that covers at least
-	 * XLOG_MAX_ICLOGS records from the tail. We need to verify that these
-	 * records were completely written. Run a CRC verification pass from
-	 * tail to head and return the result.
-	 */
-	error = xlog_do_recovery_pass(log, tmp_head, tail_blk,
+	first_bad = 0;
+	error = xlog_do_recovery_pass(log, head_blk, *tail_blk,
 				      XLOG_RECOVER_CRCPASS, &first_bad);
+	while (error == -EFSBADCRC && first_bad) {
+		int	tail_distance;
+
+		/*
+		 * Is corruption within range of the head? If so, retry from
+		 * the next record. Otherwise return an error.
+		 */
+		tail_distance = xlog_tail_distance(log, head_blk, first_bad);
+		if (tail_distance > BTOBB(XLOG_MAX_ICLOGS * hsize))
+			break;
+
+		/* skip to the next record; returns positive count on success */
+		error = xlog_seek_logrec_hdr(log, head_blk, first_bad, 2, bp,
+				&tmp_tail, &thead, &wrapped);
+		if (error < 0)
+			goto out;
+
+		*tail_blk = tmp_tail;
+		first_bad = 0;
+		error = xlog_do_recovery_pass(log, head_blk, *tail_blk,
+					      XLOG_RECOVER_CRCPASS, &first_bad);
+	}
 
+	if (!error && *tail_blk != orig_tail)
+		xfs_warn(log->l_mp,
+		"Tail block (0x%llx) overwrite detected. Updated to 0x%llx",
+			 orig_tail, *tail_blk);
 out:
 	xlog_put_bp(bp);
 	return error;
@@ -1187,7 +1232,8 @@ xlog_verify_head(
 	if (error)
 		return error;
 
-	return xlog_verify_tail(log, *head_blk, *tail_blk);
+	return xlog_verify_tail(log, *head_blk, tail_blk,
+				be32_to_cpu((*rhead)->h_size));
 }
 
 /*


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 08/25] xfs: handle -EFSCORRUPTED during head/tail verification

[Christoph Hellwig]

From: Brian Foster <bfoster@redhat.com>

commit a4c9b34d6a17081005ec459b57b8effc08f4c731 upstream.

Torn write and tail overwrite detection both trigger only on
-EFSBADCRC errors. While this is the most likely failure scenario
for each condition, -EFSCORRUPTED is still possible in certain cases
depending on what ends up on disk when a torn write or partial tail
overwrite occurs. For example, an invalid log record h_len can lead
to an -EFSCORRUPTED error when running the log recovery CRC pass.

Therefore, update log head and tail verification to trigger the
associated head/tail fixups in the event of -EFSCORRUPTED errors
along with -EFSBADCRC. Also, -EFSCORRUPTED can currently be returned
from xlog_do_recovery_pass() before rhead_blk is initialized if the
first record encountered happens to be corrupted. This leads to an
incorrect 'first_bad' return value. Initialize rhead_blk earlier in
the function to address that problem as well.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/xfs_log_recover.c | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)

diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index a5e2ca8f5cd6..43d5df3a563f 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -1102,7 +1102,7 @@ xlog_verify_tail(
 	first_bad = 0;
 	error = xlog_do_recovery_pass(log, head_blk, *tail_blk,
 				      XLOG_RECOVER_CRCPASS, &first_bad);
-	while (error == -EFSBADCRC && first_bad) {
+	while ((error == -EFSBADCRC || error == -EFSCORRUPTED) && first_bad) {
 		int	tail_distance;
 
 		/*
@@ -1188,7 +1188,7 @@ xlog_verify_head(
 	 */
 	error = xlog_do_recovery_pass(log, *head_blk, tmp_rhead_blk,
 				      XLOG_RECOVER_CRCPASS, &first_bad);
-	if (error == -EFSBADCRC) {
+	if ((error == -EFSBADCRC || error == -EFSCORRUPTED) && first_bad) {
 		/*
 		 * We've hit a potential torn write. Reset the error and warn
 		 * about it.
@@ -5257,7 +5257,7 @@ xlog_do_recovery_pass(
 	LIST_HEAD		(buffer_list);
 
 	ASSERT(head_blk != tail_blk);
-	rhead_blk = 0;
+	blk_no = rhead_blk = tail_blk;
 
 	for (i = 0; i < XLOG_RHASH_SIZE; i++)
 		INIT_HLIST_HEAD(&rhash[i]);
@@ -5335,7 +5335,6 @@ xlog_do_recovery_pass(
 	}
 
 	memset(rhash, 0, sizeof(rhash));
-	blk_no = rhead_blk = tail_blk;
 	if (tail_blk > head_blk) {
 		/*
 		 * Perform recovery around the end of the physical log.
-- 
2.14.1

Patch "xfs: handle -EFSCORRUPTED during head/tail verification" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: handle -EFSCORRUPTED during head/tail verification

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-handle-efscorrupted-during-head-tail-verification.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:14 -0700
Subject: xfs: handle -EFSCORRUPTED during head/tail verification
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, Brian Foster <bfoster@redhat.com>, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-9-hch@lst.de>

From: Brian Foster <bfoster@redhat.com>

commit a4c9b34d6a17081005ec459b57b8effc08f4c731 upstream.

Torn write and tail overwrite detection both trigger only on
-EFSBADCRC errors. While this is the most likely failure scenario
for each condition, -EFSCORRUPTED is still possible in certain cases
depending on what ends up on disk when a torn write or partial tail
overwrite occurs. For example, an invalid log record h_len can lead
to an -EFSCORRUPTED error when running the log recovery CRC pass.

Therefore, update log head and tail verification to trigger the
associated head/tail fixups in the event of -EFSCORRUPTED errors
along with -EFSBADCRC. Also, -EFSCORRUPTED can currently be returned
from xlog_do_recovery_pass() before rhead_blk is initialized if the
first record encountered happens to be corrupted. This leads to an
incorrect 'first_bad' return value. Initialize rhead_blk earlier in
the function to address that problem as well.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/xfs_log_recover.c |    7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)

--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -1102,7 +1102,7 @@ xlog_verify_tail(
 	first_bad = 0;
 	error = xlog_do_recovery_pass(log, head_blk, *tail_blk,
 				      XLOG_RECOVER_CRCPASS, &first_bad);
-	while (error == -EFSBADCRC && first_bad) {
+	while ((error == -EFSBADCRC || error == -EFSCORRUPTED) && first_bad) {
 		int	tail_distance;
 
 		/*
@@ -1188,7 +1188,7 @@ xlog_verify_head(
 	 */
 	error = xlog_do_recovery_pass(log, *head_blk, tmp_rhead_blk,
 				      XLOG_RECOVER_CRCPASS, &first_bad);
-	if (error == -EFSBADCRC) {
+	if ((error == -EFSBADCRC || error == -EFSCORRUPTED) && first_bad) {
 		/*
 		 * We've hit a potential torn write. Reset the error and warn
 		 * about it.
@@ -5257,7 +5257,7 @@ xlog_do_recovery_pass(
 	LIST_HEAD		(buffer_list);
 
 	ASSERT(head_blk != tail_blk);
-	rhead_blk = 0;
+	blk_no = rhead_blk = tail_blk;
 
 	for (i = 0; i < XLOG_RHASH_SIZE; i++)
 		INIT_HLIST_HEAD(&rhash[i]);
@@ -5335,7 +5335,6 @@ xlog_do_recovery_pass(
 	}
 
 	memset(rhash, 0, sizeof(rhash));
-	blk_no = rhead_blk = tail_blk;
 	if (tail_blk > head_blk) {
 		/*
 		 * Perform recovery around the end of the physical log.


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 09/25] xfs: stop searching for free slots in an inode chunk when there are none

[Christoph Hellwig]

From: Carlos Maiolino <cmaiolino@redhat.com>

commit 2d32311cf19bfb8c1d2b4601974ddd951f9cfd0b upstream.

In a filesystem without finobt, the Space manager selects an AG to alloc a new
inode, where xfs_dialloc_ag_inobt() will search the AG for the free slot chunk.

When the new inode is in the same AG as its parent, the btree will be searched
starting on the parent's record, and then retried from the top if no slot is
available beyond the parent's record.

To exit this loop though, xfs_dialloc_ag_inobt() relies on the fact that the
btree must have a free slot available, once its callers relied on the
agi->freecount when deciding how/where to allocate this new inode.

In the case when the agi->freecount is corrupted, showing available inodes in an
AG, when in fact there is none, this becomes an infinite loop.

Add a way to stop the loop when a free slot is not found in the btree, making
the function to fall into the whole AG scan which will then, be able to detect
the corruption and shut the filesystem down.

As pointed by Brian, this might impact performance, giving the fact we
don't reset the search distance anymore when we reach the end of the
tree, giving it fewer tries before falling back to the whole AG search, but
it will only affect searches that start within 10 records to the end of the tree.

Signed-off-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/libxfs/xfs_ialloc.c | 55 +++++++++++++++++++++++-----------------------
 1 file changed, 27 insertions(+), 28 deletions(-)

diff --git a/fs/xfs/libxfs/xfs_ialloc.c b/fs/xfs/libxfs/xfs_ialloc.c
index abf5beaae907..1e0658a3f155 100644
--- a/fs/xfs/libxfs/xfs_ialloc.c
+++ b/fs/xfs/libxfs/xfs_ialloc.c
@@ -1133,6 +1133,7 @@ xfs_dialloc_ag_inobt(
 	int			error;
 	int			offset;
 	int			i, j;
+	int			searchdistance = 10;
 
 	pag = xfs_perag_get(mp, agno);
 
@@ -1159,7 +1160,6 @@ xfs_dialloc_ag_inobt(
 	if (pagno == agno) {
 		int		doneleft;	/* done, to the left */
 		int		doneright;	/* done, to the right */
-		int		searchdistance = 10;
 
 		error = xfs_inobt_lookup(cur, pagino, XFS_LOOKUP_LE, &i);
 		if (error)
@@ -1220,21 +1220,9 @@ xfs_dialloc_ag_inobt(
 		/*
 		 * Loop until we find an inode chunk with a free inode.
 		 */
-		while (!doneleft || !doneright) {
+		while (--searchdistance > 0 && (!doneleft || !doneright)) {
 			int	useleft;  /* using left inode chunk this time */
 
-			if (!--searchdistance) {
-				/*
-				 * Not in range - save last search
-				 * location and allocate a new inode
-				 */
-				xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR);
-				pag->pagl_leftrec = trec.ir_startino;
-				pag->pagl_rightrec = rec.ir_startino;
-				pag->pagl_pagino = pagino;
-				goto newino;
-			}
-
 			/* figure out the closer block if both are valid. */
 			if (!doneleft && !doneright) {
 				useleft = pagino -
@@ -1278,26 +1266,37 @@ xfs_dialloc_ag_inobt(
 				goto error1;
 		}
 
-		/*
-		 * We've reached the end of the btree. because
-		 * we are only searching a small chunk of the
-		 * btree each search, there is obviously free
-		 * inodes closer to the parent inode than we
-		 * are now. restart the search again.
-		 */
-		pag->pagl_pagino = NULLAGINO;
-		pag->pagl_leftrec = NULLAGINO;
-		pag->pagl_rightrec = NULLAGINO;
-		xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR);
-		xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR);
-		goto restart_pagno;
+		if (searchdistance <= 0) {
+			/*
+			 * Not in range - save last search
+			 * location and allocate a new inode
+			 */
+			xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR);
+			pag->pagl_leftrec = trec.ir_startino;
+			pag->pagl_rightrec = rec.ir_startino;
+			pag->pagl_pagino = pagino;
+
+		} else {
+			/*
+			 * We've reached the end of the btree. because
+			 * we are only searching a small chunk of the
+			 * btree each search, there is obviously free
+			 * inodes closer to the parent inode than we
+			 * are now. restart the search again.
+			 */
+			pag->pagl_pagino = NULLAGINO;
+			pag->pagl_leftrec = NULLAGINO;
+			pag->pagl_rightrec = NULLAGINO;
+			xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR);
+			xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR);
+			goto restart_pagno;
+		}
 	}
 
 	/*
 	 * In a different AG from the parent.
 	 * See if the most recently allocated block has any free.
 	 */
-newino:
 	if (agi->agi_newino != cpu_to_be32(NULLAGINO)) {
 		error = xfs_inobt_lookup(cur, be32_to_cpu(agi->agi_newino),
 					 XFS_LOOKUP_EQ, &i);
-- 
2.14.1

Patch "xfs: stop searching for free slots in an inode chunk when there are none" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: stop searching for free slots in an inode chunk when there are none

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:15 -0700
Subject: xfs: stop searching for free slots in an inode chunk when there are none
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, Carlos Maiolino <cmaiolino@redhat.com>, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-10-hch@lst.de>

From: Carlos Maiolino <cmaiolino@redhat.com>

commit 2d32311cf19bfb8c1d2b4601974ddd951f9cfd0b upstream.

In a filesystem without finobt, the Space manager selects an AG to alloc a new
inode, where xfs_dialloc_ag_inobt() will search the AG for the free slot chunk.

When the new inode is in the same AG as its parent, the btree will be searched
starting on the parent's record, and then retried from the top if no slot is
available beyond the parent's record.

To exit this loop though, xfs_dialloc_ag_inobt() relies on the fact that the
btree must have a free slot available, once its callers relied on the
agi->freecount when deciding how/where to allocate this new inode.

In the case when the agi->freecount is corrupted, showing available inodes in an
AG, when in fact there is none, this becomes an infinite loop.

Add a way to stop the loop when a free slot is not found in the btree, making
the function to fall into the whole AG scan which will then, be able to detect
the corruption and shut the filesystem down.

As pointed by Brian, this might impact performance, giving the fact we
don't reset the search distance anymore when we reach the end of the
tree, giving it fewer tries before falling back to the whole AG search, but
it will only affect searches that start within 10 records to the end of the tree.

Signed-off-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/libxfs/xfs_ialloc.c |   55 ++++++++++++++++++++++-----------------------
 1 file changed, 27 insertions(+), 28 deletions(-)

--- a/fs/xfs/libxfs/xfs_ialloc.c
+++ b/fs/xfs/libxfs/xfs_ialloc.c
@@ -1133,6 +1133,7 @@ xfs_dialloc_ag_inobt(
 	int			error;
 	int			offset;
 	int			i, j;
+	int			searchdistance = 10;
 
 	pag = xfs_perag_get(mp, agno);
 
@@ -1159,7 +1160,6 @@ xfs_dialloc_ag_inobt(
 	if (pagno == agno) {
 		int		doneleft;	/* done, to the left */
 		int		doneright;	/* done, to the right */
-		int		searchdistance = 10;
 
 		error = xfs_inobt_lookup(cur, pagino, XFS_LOOKUP_LE, &i);
 		if (error)
@@ -1220,21 +1220,9 @@ xfs_dialloc_ag_inobt(
 		/*
 		 * Loop until we find an inode chunk with a free inode.
 		 */
-		while (!doneleft || !doneright) {
+		while (--searchdistance > 0 && (!doneleft || !doneright)) {
 			int	useleft;  /* using left inode chunk this time */
 
-			if (!--searchdistance) {
-				/*
-				 * Not in range - save last search
-				 * location and allocate a new inode
-				 */
-				xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR);
-				pag->pagl_leftrec = trec.ir_startino;
-				pag->pagl_rightrec = rec.ir_startino;
-				pag->pagl_pagino = pagino;
-				goto newino;
-			}
-
 			/* figure out the closer block if both are valid. */
 			if (!doneleft && !doneright) {
 				useleft = pagino -
@@ -1278,26 +1266,37 @@ xfs_dialloc_ag_inobt(
 				goto error1;
 		}
 
-		/*
-		 * We've reached the end of the btree. because
-		 * we are only searching a small chunk of the
-		 * btree each search, there is obviously free
-		 * inodes closer to the parent inode than we
-		 * are now. restart the search again.
-		 */
-		pag->pagl_pagino = NULLAGINO;
-		pag->pagl_leftrec = NULLAGINO;
-		pag->pagl_rightrec = NULLAGINO;
-		xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR);
-		xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR);
-		goto restart_pagno;
+		if (searchdistance <= 0) {
+			/*
+			 * Not in range - save last search
+			 * location and allocate a new inode
+			 */
+			xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR);
+			pag->pagl_leftrec = trec.ir_startino;
+			pag->pagl_rightrec = rec.ir_startino;
+			pag->pagl_pagino = pagino;
+
+		} else {
+			/*
+			 * We've reached the end of the btree. because
+			 * we are only searching a small chunk of the
+			 * btree each search, there is obviously free
+			 * inodes closer to the parent inode than we
+			 * are now. restart the search again.
+			 */
+			pag->pagl_pagino = NULLAGINO;
+			pag->pagl_leftrec = NULLAGINO;
+			pag->pagl_rightrec = NULLAGINO;
+			xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR);
+			xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR);
+			goto restart_pagno;
+		}
 	}
 
 	/*
 	 * In a different AG from the parent.
 	 * See if the most recently allocated block has any free.
 	 */
-newino:
 	if (agi->agi_newino != cpu_to_be32(NULLAGINO)) {
 		error = xfs_inobt_lookup(cur, be32_to_cpu(agi->agi_newino),
 					 XFS_LOOKUP_EQ, &i);


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 10/25] xfs: evict all inodes involved with log redo item

[Christoph Hellwig]

From: "Darrick J. Wong" <darrick.wong@oracle.com>

commit 799ea9e9c59949008770aab4e1da87f10e99dbe4 upstream.

When we introduced the bmap redo log items, we set MS_ACTIVE on the
mountpoint and XFS_IRECOVERY on the inode to prevent unlinked inodes
from being truncated prematurely during log recovery.  This also had the
effect of putting linked inodes on the lru instead of evicting them.

Unfortunately, we neglected to find all those unreferenced lru inodes
and evict them after finishing log recovery, which means that we leak
them if anything goes wrong in the rest of xfs_mountfs, because the lru
is only cleaned out on unmount.

Therefore, evict unreferenced inodes in the lru list immediately
after clearing MS_ACTIVE.

Fixes: 17c12bcd30 ("xfs: when replaying bmap operations, don't let unlinked inodes get reaped")
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Cc: viro@ZenIV.linux.org.uk
Reviewed-by: Brian Foster <bfoster@redhat.com>
---
 fs/inode.c         |  1 +
 fs/internal.h      |  1 -
 fs/xfs/xfs_log.c   | 12 ++++++++++++
 include/linux/fs.h |  1 +
 4 files changed, 14 insertions(+), 1 deletion(-)

diff --git a/fs/inode.c b/fs/inode.c
index 50370599e371..6a1626e0edaf 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -637,6 +637,7 @@ void evict_inodes(struct super_block *sb)
 
 	dispose_list(&dispose);
 }
+EXPORT_SYMBOL_GPL(evict_inodes);
 
 /**
  * invalidate_inodes	- attempt to free all inodes on a superblock
diff --git a/fs/internal.h b/fs/internal.h
index 9676fe11c093..fedfe94d84ba 100644
--- a/fs/internal.h
+++ b/fs/internal.h
@@ -132,7 +132,6 @@ static inline bool atime_needs_update_rcu(const struct path *path,
 extern void inode_io_list_del(struct inode *inode);
 
 extern long get_nr_dirty_inodes(void);
-extern void evict_inodes(struct super_block *);
 extern int invalidate_inodes(struct super_block *, bool);
 
 /*
diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c
index f467f1230444..bcb2f860e508 100644
--- a/fs/xfs/xfs_log.c
+++ b/fs/xfs/xfs_log.c
@@ -761,12 +761,24 @@ xfs_log_mount_finish(
 	 * inodes.  Turn it off immediately after recovery finishes
 	 * so that we don't leak the quota inodes if subsequent mount
 	 * activities fail.
+	 *
+	 * We let all inodes involved in redo item processing end up on
+	 * the LRU instead of being evicted immediately so that if we do
+	 * something to an unlinked inode, the irele won't cause
+	 * premature truncation and freeing of the inode, which results
+	 * in log recovery failure.  We have to evict the unreferenced
+	 * lru inodes after clearing MS_ACTIVE because we don't
+	 * otherwise clean up the lru if there's a subsequent failure in
+	 * xfs_mountfs, which leads to us leaking the inodes if nothing
+	 * else (e.g. quotacheck) references the inodes before the
+	 * mount failure occurs.
 	 */
 	mp->m_super->s_flags |= MS_ACTIVE;
 	error = xlog_recover_finish(mp->m_log);
 	if (!error)
 		xfs_log_work_queue(mp);
 	mp->m_super->s_flags &= ~MS_ACTIVE;
+	evict_inodes(mp->m_super);
 
 	if (readonly)
 		mp->m_flags |= XFS_MOUNT_RDONLY;
diff --git a/include/linux/fs.h b/include/linux/fs.h
index cbfe127bccf8..d0c0ca8ea8c1 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -2831,6 +2831,7 @@ static inline void lockdep_annotate_inode_mutex_key(struct inode *inode) { };
 #endif
 extern void unlock_new_inode(struct inode *);
 extern unsigned int get_next_ino(void);
+extern void evict_inodes(struct super_block *sb);
 
 extern void __iget(struct inode * inode);
 extern void iget_failed(struct inode *);
-- 
2.14.1

Patch "xfs: evict all inodes involved with log redo item" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: evict all inodes involved with log redo item

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-evict-all-inodes-involved-with-log-redo-item.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:16 -0700
Subject: xfs: evict all inodes involved with log redo item
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, "Darrick J. Wong" <darrick.wong@oracle.com>, viro@ZenIV.linux.org.uk
Message-ID: <20170917210631.10725-11-hch@lst.de>

From: "Darrick J. Wong" <darrick.wong@oracle.com>

commit 799ea9e9c59949008770aab4e1da87f10e99dbe4 upstream.

When we introduced the bmap redo log items, we set MS_ACTIVE on the
mountpoint and XFS_IRECOVERY on the inode to prevent unlinked inodes
from being truncated prematurely during log recovery.  This also had the
effect of putting linked inodes on the lru instead of evicting them.

Unfortunately, we neglected to find all those unreferenced lru inodes
and evict them after finishing log recovery, which means that we leak
them if anything goes wrong in the rest of xfs_mountfs, because the lru
is only cleaned out on unmount.

Therefore, evict unreferenced inodes in the lru list immediately
after clearing MS_ACTIVE.

Fixes: 17c12bcd30 ("xfs: when replaying bmap operations, don't let unlinked inodes get reaped")
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Cc: viro@ZenIV.linux.org.uk
Reviewed-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/inode.c         |    1 +
 fs/internal.h      |    1 -
 fs/xfs/xfs_log.c   |   12 ++++++++++++
 include/linux/fs.h |    1 +
 4 files changed, 14 insertions(+), 1 deletion(-)

--- a/fs/inode.c
+++ b/fs/inode.c
@@ -637,6 +637,7 @@ again:
 
 	dispose_list(&dispose);
 }
+EXPORT_SYMBOL_GPL(evict_inodes);
 
 /**
  * invalidate_inodes	- attempt to free all inodes on a superblock
--- a/fs/internal.h
+++ b/fs/internal.h
@@ -132,7 +132,6 @@ static inline bool atime_needs_update_rc
 extern void inode_io_list_del(struct inode *inode);
 
 extern long get_nr_dirty_inodes(void);
-extern void evict_inodes(struct super_block *);
 extern int invalidate_inodes(struct super_block *, bool);
 
 /*
--- a/fs/xfs/xfs_log.c
+++ b/fs/xfs/xfs_log.c
@@ -761,12 +761,24 @@ xfs_log_mount_finish(
 	 * inodes.  Turn it off immediately after recovery finishes
 	 * so that we don't leak the quota inodes if subsequent mount
 	 * activities fail.
+	 *
+	 * We let all inodes involved in redo item processing end up on
+	 * the LRU instead of being evicted immediately so that if we do
+	 * something to an unlinked inode, the irele won't cause
+	 * premature truncation and freeing of the inode, which results
+	 * in log recovery failure.  We have to evict the unreferenced
+	 * lru inodes after clearing MS_ACTIVE because we don't
+	 * otherwise clean up the lru if there's a subsequent failure in
+	 * xfs_mountfs, which leads to us leaking the inodes if nothing
+	 * else (e.g. quotacheck) references the inodes before the
+	 * mount failure occurs.
 	 */
 	mp->m_super->s_flags |= MS_ACTIVE;
 	error = xlog_recover_finish(mp->m_log);
 	if (!error)
 		xfs_log_work_queue(mp);
 	mp->m_super->s_flags &= ~MS_ACTIVE;
+	evict_inodes(mp->m_super);
 
 	if (readonly)
 		mp->m_flags |= XFS_MOUNT_RDONLY;
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -2831,6 +2831,7 @@ static inline void lockdep_annotate_inod
 #endif
 extern void unlock_new_inode(struct inode *);
 extern unsigned int get_next_ino(void);
+extern void evict_inodes(struct super_block *sb);
 
 extern void __iget(struct inode * inode);
 extern void iget_failed(struct inode *);


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 11/25] xfs: check for race with xfs_reclaim_inode() in xfs_ifree_cluster()

[Christoph Hellwig]

From: Omar Sandoval <osandov@fb.com>

commit f2e9ad212def50bcf4c098c6288779dd97fff0f0 upstream.

After xfs_ifree_cluster() finds an inode in the radix tree and verifies
that the inode number is what it expected, xfs_reclaim_inode() can swoop
in and free it. xfs_ifree_cluster() will then happily continue working
on the freed inode. Most importantly, it will mark the inode stale,
which will probably be overwritten when the inode slab object is
reallocated, but if it has already been reallocated then we can end up
with an inode spuriously marked stale.

In 8a17d7ddedb4 ("xfs: mark reclaimed inodes invalid earlier") we added
a second check to xfs_iflush_cluster() to detect this race, but the
similar RCU lookup in xfs_ifree_cluster() needs the same treatment.

Signed-off-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/xfs_icache.c | 10 +++++-----
 fs/xfs/xfs_inode.c  | 23 ++++++++++++++++++-----
 2 files changed, 23 insertions(+), 10 deletions(-)

diff --git a/fs/xfs/xfs_icache.c b/fs/xfs/xfs_icache.c
index 0a9e6985a0d0..34227115a5d6 100644
--- a/fs/xfs/xfs_icache.c
+++ b/fs/xfs/xfs_icache.c
@@ -1124,11 +1124,11 @@ xfs_reclaim_inode(
 	 * Because we use RCU freeing we need to ensure the inode always appears
 	 * to be reclaimed with an invalid inode number when in the free state.
 	 * We do this as early as possible under the ILOCK so that
-	 * xfs_iflush_cluster() can be guaranteed to detect races with us here.
-	 * By doing this, we guarantee that once xfs_iflush_cluster has locked
-	 * XFS_ILOCK that it will see either a valid, flushable inode that will
-	 * serialise correctly, or it will see a clean (and invalid) inode that
-	 * it can skip.
+	 * xfs_iflush_cluster() and xfs_ifree_cluster() can be guaranteed to
+	 * detect races with us here. By doing this, we guarantee that once
+	 * xfs_iflush_cluster() or xfs_ifree_cluster() has locked XFS_ILOCK that
+	 * it will see either a valid inode that will serialise correctly, or it
+	 * will see an invalid inode that it can skip.
 	 */
 	spin_lock(&ip->i_flags_lock);
 	ip->i_flags = XFS_IRECLAIM;
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index ff48f0096810..97045e8dfed5 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -2359,11 +2359,24 @@ xfs_ifree_cluster(
 			 * already marked stale. If we can't lock it, back off
 			 * and retry.
 			 */
-			if (ip != free_ip &&
-			    !xfs_ilock_nowait(ip, XFS_ILOCK_EXCL)) {
-				rcu_read_unlock();
-				delay(1);
-				goto retry;
+			if (ip != free_ip) {
+				if (!xfs_ilock_nowait(ip, XFS_ILOCK_EXCL)) {
+					rcu_read_unlock();
+					delay(1);
+					goto retry;
+				}
+
+				/*
+				 * Check the inode number again in case we're
+				 * racing with freeing in xfs_reclaim_inode().
+				 * See the comments in that function for more
+				 * information as to why the initial check is
+				 * not sufficient.
+				 */
+				if (ip->i_ino != inum + i) {
+					xfs_iunlock(ip, XFS_ILOCK_EXCL);
+					continue;
+				}
 			}
 			rcu_read_unlock();
 
-- 
2.14.1

Patch "xfs: check for race with xfs_reclaim_inode() in xfs_ifree_cluster()" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: check for race with xfs_reclaim_inode() in xfs_ifree_cluster()

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:17 -0700
Subject: xfs: check for race with xfs_reclaim_inode() in xfs_ifree_cluster()
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, Omar Sandoval <osandov@fb.com>, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-12-hch@lst.de>

From: Omar Sandoval <osandov@fb.com>

commit f2e9ad212def50bcf4c098c6288779dd97fff0f0 upstream.

After xfs_ifree_cluster() finds an inode in the radix tree and verifies
that the inode number is what it expected, xfs_reclaim_inode() can swoop
in and free it. xfs_ifree_cluster() will then happily continue working
on the freed inode. Most importantly, it will mark the inode stale,
which will probably be overwritten when the inode slab object is
reallocated, but if it has already been reallocated then we can end up
with an inode spuriously marked stale.

In 8a17d7ddedb4 ("xfs: mark reclaimed inodes invalid earlier") we added
a second check to xfs_iflush_cluster() to detect this race, but the
similar RCU lookup in xfs_ifree_cluster() needs the same treatment.

Signed-off-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/xfs_icache.c |   10 +++++-----
 fs/xfs/xfs_inode.c  |   23 ++++++++++++++++++-----
 2 files changed, 23 insertions(+), 10 deletions(-)

--- a/fs/xfs/xfs_icache.c
+++ b/fs/xfs/xfs_icache.c
@@ -1124,11 +1124,11 @@ reclaim:
 	 * Because we use RCU freeing we need to ensure the inode always appears
 	 * to be reclaimed with an invalid inode number when in the free state.
 	 * We do this as early as possible under the ILOCK so that
-	 * xfs_iflush_cluster() can be guaranteed to detect races with us here.
-	 * By doing this, we guarantee that once xfs_iflush_cluster has locked
-	 * XFS_ILOCK that it will see either a valid, flushable inode that will
-	 * serialise correctly, or it will see a clean (and invalid) inode that
-	 * it can skip.
+	 * xfs_iflush_cluster() and xfs_ifree_cluster() can be guaranteed to
+	 * detect races with us here. By doing this, we guarantee that once
+	 * xfs_iflush_cluster() or xfs_ifree_cluster() has locked XFS_ILOCK that
+	 * it will see either a valid inode that will serialise correctly, or it
+	 * will see an invalid inode that it can skip.
 	 */
 	spin_lock(&ip->i_flags_lock);
 	ip->i_flags = XFS_IRECLAIM;
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -2359,11 +2359,24 @@ retry:
 			 * already marked stale. If we can't lock it, back off
 			 * and retry.
 			 */
-			if (ip != free_ip &&
-			    !xfs_ilock_nowait(ip, XFS_ILOCK_EXCL)) {
-				rcu_read_unlock();
-				delay(1);
-				goto retry;
+			if (ip != free_ip) {
+				if (!xfs_ilock_nowait(ip, XFS_ILOCK_EXCL)) {
+					rcu_read_unlock();
+					delay(1);
+					goto retry;
+				}
+
+				/*
+				 * Check the inode number again in case we're
+				 * racing with freeing in xfs_reclaim_inode().
+				 * See the comments in that function for more
+				 * information as to why the initial check is
+				 * not sufficient.
+				 */
+				if (ip->i_ino != inum + i) {
+					xfs_iunlock(ip, XFS_ILOCK_EXCL);
+					continue;
+				}
 			}
 			rcu_read_unlock();
 


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 12/25] xfs: open-code xfs_buf_item_dirty()

[Christoph Hellwig]

From: Brian Foster <bfoster@redhat.com>

commit a4f6cf6b2b6b60ec2a05a33a32e65caa4149aa2b upstream.

It checks a single flag and has one caller. It probably isn't worth
its own function.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/xfs_buf_item.c  | 11 -----------
 fs/xfs/xfs_buf_item.h  |  1 -
 fs/xfs/xfs_trans_buf.c |  2 +-
 3 files changed, 1 insertion(+), 13 deletions(-)

diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c
index 573fc72c3f23..cdae0ad5e0a5 100644
--- a/fs/xfs/xfs_buf_item.c
+++ b/fs/xfs/xfs_buf_item.c
@@ -945,17 +945,6 @@ xfs_buf_item_log(
 }
 
 
-/*
- * Return 1 if the buffer has been logged or ordered in a transaction (at any
- * point, not just the current transaction) and 0 if not.
- */
-uint
-xfs_buf_item_dirty(
-	xfs_buf_log_item_t	*bip)
-{
-	return (bip->bli_flags & XFS_BLI_DIRTY);
-}
-
 STATIC void
 xfs_buf_item_free(
 	xfs_buf_log_item_t	*bip)
diff --git a/fs/xfs/xfs_buf_item.h b/fs/xfs/xfs_buf_item.h
index 530686e1afb9..e0e744aefaa8 100644
--- a/fs/xfs/xfs_buf_item.h
+++ b/fs/xfs/xfs_buf_item.h
@@ -64,7 +64,6 @@ typedef struct xfs_buf_log_item {
 int	xfs_buf_item_init(struct xfs_buf *, struct xfs_mount *);
 void	xfs_buf_item_relse(struct xfs_buf *);
 void	xfs_buf_item_log(xfs_buf_log_item_t *, uint, uint);
-uint	xfs_buf_item_dirty(xfs_buf_log_item_t *);
 void	xfs_buf_attach_iodone(struct xfs_buf *,
 			      void(*)(struct xfs_buf *, xfs_log_item_t *),
 			      xfs_log_item_t *);
diff --git a/fs/xfs/xfs_trans_buf.c b/fs/xfs/xfs_trans_buf.c
index 86987d823d76..cac8abbeca3f 100644
--- a/fs/xfs/xfs_trans_buf.c
+++ b/fs/xfs/xfs_trans_buf.c
@@ -435,7 +435,7 @@ xfs_trans_brelse(xfs_trans_t	*tp,
 	if (XFS_FORCED_SHUTDOWN(tp->t_mountp) && freed) {
 		xfs_trans_ail_remove(&bip->bli_item, SHUTDOWN_LOG_IO_ERROR);
 		xfs_buf_item_relse(bp);
-	} else if (!xfs_buf_item_dirty(bip)) {
+	} else if (!(bip->bli_flags & XFS_BLI_DIRTY)) {
 /***
 		ASSERT(bp->b_pincount == 0);
 ***/
-- 
2.14.1

Patch "xfs: open-code xfs_buf_item_dirty()" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: open-code xfs_buf_item_dirty()

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-open-code-xfs_buf_item_dirty.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:18 -0700
Subject: xfs: open-code xfs_buf_item_dirty()
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, Brian Foster <bfoster@redhat.com>, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-13-hch@lst.de>

From: Brian Foster <bfoster@redhat.com>

commit a4f6cf6b2b6b60ec2a05a33a32e65caa4149aa2b upstream.

It checks a single flag and has one caller. It probably isn't worth
its own function.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/xfs_buf_item.c  |   11 -----------
 fs/xfs/xfs_buf_item.h  |    1 -
 fs/xfs/xfs_trans_buf.c |    2 +-
 3 files changed, 1 insertion(+), 13 deletions(-)

--- a/fs/xfs/xfs_buf_item.c
+++ b/fs/xfs/xfs_buf_item.c
@@ -945,17 +945,6 @@ xfs_buf_item_log(
 }
 
 
-/*
- * Return 1 if the buffer has been logged or ordered in a transaction (at any
- * point, not just the current transaction) and 0 if not.
- */
-uint
-xfs_buf_item_dirty(
-	xfs_buf_log_item_t	*bip)
-{
-	return (bip->bli_flags & XFS_BLI_DIRTY);
-}
-
 STATIC void
 xfs_buf_item_free(
 	xfs_buf_log_item_t	*bip)
--- a/fs/xfs/xfs_buf_item.h
+++ b/fs/xfs/xfs_buf_item.h
@@ -64,7 +64,6 @@ typedef struct xfs_buf_log_item {
 int	xfs_buf_item_init(struct xfs_buf *, struct xfs_mount *);
 void	xfs_buf_item_relse(struct xfs_buf *);
 void	xfs_buf_item_log(xfs_buf_log_item_t *, uint, uint);
-uint	xfs_buf_item_dirty(xfs_buf_log_item_t *);
 void	xfs_buf_attach_iodone(struct xfs_buf *,
 			      void(*)(struct xfs_buf *, xfs_log_item_t *),
 			      xfs_log_item_t *);
--- a/fs/xfs/xfs_trans_buf.c
+++ b/fs/xfs/xfs_trans_buf.c
@@ -435,7 +435,7 @@ xfs_trans_brelse(xfs_trans_t	*tp,
 	if (XFS_FORCED_SHUTDOWN(tp->t_mountp) && freed) {
 		xfs_trans_ail_remove(&bip->bli_item, SHUTDOWN_LOG_IO_ERROR);
 		xfs_buf_item_relse(bp);
-	} else if (!xfs_buf_item_dirty(bip)) {
+	} else if (!(bip->bli_flags & XFS_BLI_DIRTY)) {
 /***
 		ASSERT(bp->b_pincount == 0);
 ***/


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 13/25] xfs: remove unnecessary dirty bli format check for ordered bufs

[Christoph Hellwig]

From: Brian Foster <bfoster@redhat.com>

commit 6453c65d3576bc3e602abb5add15f112755c08ca upstream.

xfs_buf_item_unlock() historically checked the dirty state of the
buffer by manually checking the buffer log formats for dirty
segments. The introduction of ordered buffers invalidated this check
because ordered buffers have dirty bli's but no dirty (logged)
segments. The check was updated to accommodate ordered buffers by
looking at the bli state first and considering the blf only if the
bli is clean.

This logic is safe but unnecessary. There is no valid case where the
bli is clean yet the blf has dirty segments. The bli is set dirty
whenever the blf is logged (via xfs_trans_log_buf()) and the blf is
cleared in the only place BLI_DIRTY is cleared (xfs_trans_binval()).

Remove the conditional blf dirty checks and replace with an assert
that should catch any discrepencies between bli and blf dirty
states. Refactor the old blf dirty check into a helper function to
be used by the assert.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/xfs_buf_item.c | 62 ++++++++++++++++++++++++++-------------------------
 fs/xfs/xfs_buf_item.h |  1 +
 2 files changed, 33 insertions(+), 30 deletions(-)

diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c
index cdae0ad5e0a5..ff076d11804a 100644
--- a/fs/xfs/xfs_buf_item.c
+++ b/fs/xfs/xfs_buf_item.c
@@ -575,26 +575,18 @@ xfs_buf_item_unlock(
 {
 	struct xfs_buf_log_item	*bip = BUF_ITEM(lip);
 	struct xfs_buf		*bp = bip->bli_buf;
-	bool			clean;
-	bool			aborted;
-	int			flags;
+	bool			aborted = !!(lip->li_flags & XFS_LI_ABORTED);
+	bool			hold = !!(bip->bli_flags & XFS_BLI_HOLD);
+	bool			dirty = !!(bip->bli_flags & XFS_BLI_DIRTY);
+	bool			ordered = !!(bip->bli_flags & XFS_BLI_ORDERED);
 
 	/* Clear the buffer's association with this transaction. */
 	bp->b_transp = NULL;
 
 	/*
-	 * If this is a transaction abort, don't return early.  Instead, allow
-	 * the brelse to happen.  Normally it would be done for stale
-	 * (cancelled) buffers at unpin time, but we'll never go through the
-	 * pin/unpin cycle if we abort inside commit.
+	 * The per-transaction state has been copied above so clear it from the
+	 * bli.
 	 */
-	aborted = (lip->li_flags & XFS_LI_ABORTED) ? true : false;
-	/*
-	 * Before possibly freeing the buf item, copy the per-transaction state
-	 * so we can reference it safely later after clearing it from the
-	 * buffer log item.
-	 */
-	flags = bip->bli_flags;
 	bip->bli_flags &= ~(XFS_BLI_LOGGED | XFS_BLI_HOLD | XFS_BLI_ORDERED);
 
 	/*
@@ -602,7 +594,7 @@ xfs_buf_item_unlock(
 	 * unlock the buffer and free the buf item when the buffer is unpinned
 	 * for the last time.
 	 */
-	if (flags & XFS_BLI_STALE) {
+	if (bip->bli_flags & XFS_BLI_STALE) {
 		trace_xfs_buf_item_unlock_stale(bip);
 		ASSERT(bip->__bli_format.blf_flags & XFS_BLF_CANCEL);
 		if (!aborted) {
@@ -620,20 +612,11 @@ xfs_buf_item_unlock(
 	 * regardless of whether it is dirty or not. A dirty abort implies a
 	 * shutdown, anyway.
 	 *
-	 * Ordered buffers are dirty but may have no recorded changes, so ensure
-	 * we only release clean items here.
+	 * The bli dirty state should match whether the blf has logged segments
+	 * except for ordered buffers, where only the bli should be dirty.
 	 */
-	clean = (flags & XFS_BLI_DIRTY) ? false : true;
-	if (clean) {
-		int i;
-		for (i = 0; i < bip->bli_format_count; i++) {
-			if (!xfs_bitmap_empty(bip->bli_formats[i].blf_data_map,
-				     bip->bli_formats[i].blf_map_size)) {
-				clean = false;
-				break;
-			}
-		}
-	}
+	ASSERT((!ordered && dirty == xfs_buf_item_dirty_format(bip)) ||
+	       (ordered && dirty && !xfs_buf_item_dirty_format(bip)));
 
 	/*
 	 * Clean buffers, by definition, cannot be in the AIL. However, aborted
@@ -652,11 +635,11 @@ xfs_buf_item_unlock(
 			ASSERT(XFS_FORCED_SHUTDOWN(lip->li_mountp));
 			xfs_trans_ail_remove(lip, SHUTDOWN_LOG_IO_ERROR);
 			xfs_buf_item_relse(bp);
-		} else if (clean)
+		} else if (!dirty)
 			xfs_buf_item_relse(bp);
 	}
 
-	if (!(flags & XFS_BLI_HOLD))
+	if (!hold)
 		xfs_buf_relse(bp);
 }
 
@@ -945,6 +928,25 @@ xfs_buf_item_log(
 }
 
 
+/*
+ * Return true if the buffer has any ranges logged/dirtied by a transaction,
+ * false otherwise.
+ */
+bool
+xfs_buf_item_dirty_format(
+	struct xfs_buf_log_item	*bip)
+{
+	int			i;
+
+	for (i = 0; i < bip->bli_format_count; i++) {
+		if (!xfs_bitmap_empty(bip->bli_formats[i].blf_data_map,
+			     bip->bli_formats[i].blf_map_size))
+			return true;
+	}
+
+	return false;
+}
+
 STATIC void
 xfs_buf_item_free(
 	xfs_buf_log_item_t	*bip)
diff --git a/fs/xfs/xfs_buf_item.h b/fs/xfs/xfs_buf_item.h
index e0e744aefaa8..9690ce62c9a7 100644
--- a/fs/xfs/xfs_buf_item.h
+++ b/fs/xfs/xfs_buf_item.h
@@ -64,6 +64,7 @@ typedef struct xfs_buf_log_item {
 int	xfs_buf_item_init(struct xfs_buf *, struct xfs_mount *);
 void	xfs_buf_item_relse(struct xfs_buf *);
 void	xfs_buf_item_log(xfs_buf_log_item_t *, uint, uint);
+bool	xfs_buf_item_dirty_format(struct xfs_buf_log_item *);
 void	xfs_buf_attach_iodone(struct xfs_buf *,
 			      void(*)(struct xfs_buf *, xfs_log_item_t *),
 			      xfs_log_item_t *);
-- 
2.14.1

Patch "xfs: remove unnecessary dirty bli format check for ordered bufs" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: remove unnecessary dirty bli format check for ordered bufs

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:19 -0700
Subject: xfs: remove unnecessary dirty bli format check for ordered bufs
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, Brian Foster <bfoster@redhat.com>, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-14-hch@lst.de>

From: Brian Foster <bfoster@redhat.com>

commit 6453c65d3576bc3e602abb5add15f112755c08ca upstream.

xfs_buf_item_unlock() historically checked the dirty state of the
buffer by manually checking the buffer log formats for dirty
segments. The introduction of ordered buffers invalidated this check
because ordered buffers have dirty bli's but no dirty (logged)
segments. The check was updated to accommodate ordered buffers by
looking at the bli state first and considering the blf only if the
bli is clean.

This logic is safe but unnecessary. There is no valid case where the
bli is clean yet the blf has dirty segments. The bli is set dirty
whenever the blf is logged (via xfs_trans_log_buf()) and the blf is
cleared in the only place BLI_DIRTY is cleared (xfs_trans_binval()).

Remove the conditional blf dirty checks and replace with an assert
that should catch any discrepencies between bli and blf dirty
states. Refactor the old blf dirty check into a helper function to
be used by the assert.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/xfs_buf_item.c |   62 +++++++++++++++++++++++++-------------------------
 fs/xfs/xfs_buf_item.h |    1 
 2 files changed, 33 insertions(+), 30 deletions(-)

--- a/fs/xfs/xfs_buf_item.c
+++ b/fs/xfs/xfs_buf_item.c
@@ -575,26 +575,18 @@ xfs_buf_item_unlock(
 {
 	struct xfs_buf_log_item	*bip = BUF_ITEM(lip);
 	struct xfs_buf		*bp = bip->bli_buf;
-	bool			clean;
-	bool			aborted;
-	int			flags;
+	bool			aborted = !!(lip->li_flags & XFS_LI_ABORTED);
+	bool			hold = !!(bip->bli_flags & XFS_BLI_HOLD);
+	bool			dirty = !!(bip->bli_flags & XFS_BLI_DIRTY);
+	bool			ordered = !!(bip->bli_flags & XFS_BLI_ORDERED);
 
 	/* Clear the buffer's association with this transaction. */
 	bp->b_transp = NULL;
 
 	/*
-	 * If this is a transaction abort, don't return early.  Instead, allow
-	 * the brelse to happen.  Normally it would be done for stale
-	 * (cancelled) buffers at unpin time, but we'll never go through the
-	 * pin/unpin cycle if we abort inside commit.
-	 */
-	aborted = (lip->li_flags & XFS_LI_ABORTED) ? true : false;
-	/*
-	 * Before possibly freeing the buf item, copy the per-transaction state
-	 * so we can reference it safely later after clearing it from the
-	 * buffer log item.
+	 * The per-transaction state has been copied above so clear it from the
+	 * bli.
 	 */
-	flags = bip->bli_flags;
 	bip->bli_flags &= ~(XFS_BLI_LOGGED | XFS_BLI_HOLD | XFS_BLI_ORDERED);
 
 	/*
@@ -602,7 +594,7 @@ xfs_buf_item_unlock(
 	 * unlock the buffer and free the buf item when the buffer is unpinned
 	 * for the last time.
 	 */
-	if (flags & XFS_BLI_STALE) {
+	if (bip->bli_flags & XFS_BLI_STALE) {
 		trace_xfs_buf_item_unlock_stale(bip);
 		ASSERT(bip->__bli_format.blf_flags & XFS_BLF_CANCEL);
 		if (!aborted) {
@@ -620,20 +612,11 @@ xfs_buf_item_unlock(
 	 * regardless of whether it is dirty or not. A dirty abort implies a
 	 * shutdown, anyway.
 	 *
-	 * Ordered buffers are dirty but may have no recorded changes, so ensure
-	 * we only release clean items here.
+	 * The bli dirty state should match whether the blf has logged segments
+	 * except for ordered buffers, where only the bli should be dirty.
 	 */
-	clean = (flags & XFS_BLI_DIRTY) ? false : true;
-	if (clean) {
-		int i;
-		for (i = 0; i < bip->bli_format_count; i++) {
-			if (!xfs_bitmap_empty(bip->bli_formats[i].blf_data_map,
-				     bip->bli_formats[i].blf_map_size)) {
-				clean = false;
-				break;
-			}
-		}
-	}
+	ASSERT((!ordered && dirty == xfs_buf_item_dirty_format(bip)) ||
+	       (ordered && dirty && !xfs_buf_item_dirty_format(bip)));
 
 	/*
 	 * Clean buffers, by definition, cannot be in the AIL. However, aborted
@@ -652,11 +635,11 @@ xfs_buf_item_unlock(
 			ASSERT(XFS_FORCED_SHUTDOWN(lip->li_mountp));
 			xfs_trans_ail_remove(lip, SHUTDOWN_LOG_IO_ERROR);
 			xfs_buf_item_relse(bp);
-		} else if (clean)
+		} else if (!dirty)
 			xfs_buf_item_relse(bp);
 	}
 
-	if (!(flags & XFS_BLI_HOLD))
+	if (!hold)
 		xfs_buf_relse(bp);
 }
 
@@ -945,6 +928,25 @@ xfs_buf_item_log(
 }
 
 
+/*
+ * Return true if the buffer has any ranges logged/dirtied by a transaction,
+ * false otherwise.
+ */
+bool
+xfs_buf_item_dirty_format(
+	struct xfs_buf_log_item	*bip)
+{
+	int			i;
+
+	for (i = 0; i < bip->bli_format_count; i++) {
+		if (!xfs_bitmap_empty(bip->bli_formats[i].blf_data_map,
+			     bip->bli_formats[i].blf_map_size))
+			return true;
+	}
+
+	return false;
+}
+
 STATIC void
 xfs_buf_item_free(
 	xfs_buf_log_item_t	*bip)
--- a/fs/xfs/xfs_buf_item.h
+++ b/fs/xfs/xfs_buf_item.h
@@ -64,6 +64,7 @@ typedef struct xfs_buf_log_item {
 int	xfs_buf_item_init(struct xfs_buf *, struct xfs_mount *);
 void	xfs_buf_item_relse(struct xfs_buf *);
 void	xfs_buf_item_log(xfs_buf_log_item_t *, uint, uint);
+bool	xfs_buf_item_dirty_format(struct xfs_buf_log_item *);
 void	xfs_buf_attach_iodone(struct xfs_buf *,
 			      void(*)(struct xfs_buf *, xfs_log_item_t *),
 			      xfs_log_item_t *);


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 14/25] xfs: ordered buffer log items are never formatted

[Christoph Hellwig]

From: Brian Foster <bfoster@redhat.com>

commit e9385cc6fb7edf23702de33a2dc82965d92d9392 upstream.

Ordered buffers pass through the logging infrastructure without ever
being written to the log. The way this works is that the ordered
buffer status is transferred to the log vector at commit time via
the ->iop_size() callback. In xlog_cil_insert_format_items(),
ordered log vectors bypass ->iop_format() processing altogether.

Therefore it is unnecessary for xfs_buf_item_format() to handle
ordered buffers. Remove the unnecessary logic and assert that an
ordered buffer never reaches this point.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-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/xfs_buf_item.c | 12 ++----------
 fs/xfs/xfs_trace.h    |  1 -
 2 files changed, 2 insertions(+), 11 deletions(-)

diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c
index ff076d11804a..ef2c1375f092 100644
--- a/fs/xfs/xfs_buf_item.c
+++ b/fs/xfs/xfs_buf_item.c
@@ -323,6 +323,8 @@ xfs_buf_item_format(
 	ASSERT((bip->bli_flags & XFS_BLI_STALE) ||
 	       (xfs_blft_from_flags(&bip->__bli_format) > XFS_BLFT_UNKNOWN_BUF
 	        && xfs_blft_from_flags(&bip->__bli_format) < XFS_BLFT_MAX_BUF));
+	ASSERT(!(bip->bli_flags & XFS_BLI_ORDERED) ||
+	       (bip->bli_flags & XFS_BLI_STALE));
 
 
 	/*
@@ -347,16 +349,6 @@ xfs_buf_item_format(
 		bip->bli_flags &= ~XFS_BLI_INODE_BUF;
 	}
 
-	if ((bip->bli_flags & (XFS_BLI_ORDERED|XFS_BLI_STALE)) ==
-							XFS_BLI_ORDERED) {
-		/*
-		 * The buffer has been logged just to order it.  It is not being
-		 * included in the transaction commit, so don't format it.
-		 */
-		trace_xfs_buf_item_format_ordered(bip);
-		return;
-	}
-
 	for (i = 0; i < bip->bli_format_count; i++) {
 		xfs_buf_item_format_segment(bip, lv, &vecp, offset,
 					    &bip->bli_formats[i]);
diff --git a/fs/xfs/xfs_trace.h b/fs/xfs/xfs_trace.h
index bcc3cdf8e1c5..bb0099708827 100644
--- a/fs/xfs/xfs_trace.h
+++ b/fs/xfs/xfs_trace.h
@@ -517,7 +517,6 @@ DEFINE_BUF_ITEM_EVENT(xfs_buf_item_size);
 DEFINE_BUF_ITEM_EVENT(xfs_buf_item_size_ordered);
 DEFINE_BUF_ITEM_EVENT(xfs_buf_item_size_stale);
 DEFINE_BUF_ITEM_EVENT(xfs_buf_item_format);
-DEFINE_BUF_ITEM_EVENT(xfs_buf_item_format_ordered);
 DEFINE_BUF_ITEM_EVENT(xfs_buf_item_format_stale);
 DEFINE_BUF_ITEM_EVENT(xfs_buf_item_ordered);
 DEFINE_BUF_ITEM_EVENT(xfs_buf_item_pin);
-- 
2.14.1

Patch "xfs: ordered buffer log items are never formatted" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: ordered buffer log items are never formatted

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-ordered-buffer-log-items-are-never-formatted.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:20 -0700
Subject: xfs: ordered buffer log items are never formatted
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, Brian Foster <bfoster@redhat.com>, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-15-hch@lst.de>

From: Brian Foster <bfoster@redhat.com>

commit e9385cc6fb7edf23702de33a2dc82965d92d9392 upstream.

Ordered buffers pass through the logging infrastructure without ever
being written to the log. The way this works is that the ordered
buffer status is transferred to the log vector at commit time via
the ->iop_size() callback. In xlog_cil_insert_format_items(),
ordered log vectors bypass ->iop_format() processing altogether.

Therefore it is unnecessary for xfs_buf_item_format() to handle
ordered buffers. Remove the unnecessary logic and assert that an
ordered buffer never reaches this point.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-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>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/xfs_buf_item.c |   12 ++----------
 fs/xfs/xfs_trace.h    |    1 -
 2 files changed, 2 insertions(+), 11 deletions(-)

--- a/fs/xfs/xfs_buf_item.c
+++ b/fs/xfs/xfs_buf_item.c
@@ -323,6 +323,8 @@ xfs_buf_item_format(
 	ASSERT((bip->bli_flags & XFS_BLI_STALE) ||
 	       (xfs_blft_from_flags(&bip->__bli_format) > XFS_BLFT_UNKNOWN_BUF
 	        && xfs_blft_from_flags(&bip->__bli_format) < XFS_BLFT_MAX_BUF));
+	ASSERT(!(bip->bli_flags & XFS_BLI_ORDERED) ||
+	       (bip->bli_flags & XFS_BLI_STALE));
 
 
 	/*
@@ -347,16 +349,6 @@ xfs_buf_item_format(
 		bip->bli_flags &= ~XFS_BLI_INODE_BUF;
 	}
 
-	if ((bip->bli_flags & (XFS_BLI_ORDERED|XFS_BLI_STALE)) ==
-							XFS_BLI_ORDERED) {
-		/*
-		 * The buffer has been logged just to order it.  It is not being
-		 * included in the transaction commit, so don't format it.
-		 */
-		trace_xfs_buf_item_format_ordered(bip);
-		return;
-	}
-
 	for (i = 0; i < bip->bli_format_count; i++) {
 		xfs_buf_item_format_segment(bip, lv, &vecp, offset,
 					    &bip->bli_formats[i]);
--- a/fs/xfs/xfs_trace.h
+++ b/fs/xfs/xfs_trace.h
@@ -517,7 +517,6 @@ DEFINE_BUF_ITEM_EVENT(xfs_buf_item_size)
 DEFINE_BUF_ITEM_EVENT(xfs_buf_item_size_ordered);
 DEFINE_BUF_ITEM_EVENT(xfs_buf_item_size_stale);
 DEFINE_BUF_ITEM_EVENT(xfs_buf_item_format);
-DEFINE_BUF_ITEM_EVENT(xfs_buf_item_format_ordered);
 DEFINE_BUF_ITEM_EVENT(xfs_buf_item_format_stale);
 DEFINE_BUF_ITEM_EVENT(xfs_buf_item_ordered);
 DEFINE_BUF_ITEM_EVENT(xfs_buf_item_pin);


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 15/25] xfs: refactor buffer logging into buffer dirtying helper

[Christoph Hellwig]

From: Brian Foster <bfoster@redhat.com>

commit 9684010d38eccda733b61106765e9357cf436f65 upstream.

xfs_trans_log_buf() is responsible for logging the dirty segments of
a buffer along with setting all of the necessary state on the
transaction, buffer, bli, etc., to ensure that the associated items
are marked as dirty and prepared for I/O. We have a couple use cases
that need to to dirty a buffer in a transaction without actually
logging dirty ranges of the buffer.  One existing use case is
ordered buffers, which are currently logged with arbitrary ranges to
accomplish this even though the content of ordered buffers is never
written to the log. Another pending use case is to relog an already
dirty buffer across rolled transactions within the deferred
operations infrastructure. This is required to prevent a held
(XFS_BLI_HOLD) buffer from pinning the tail of the log.

Refactor xfs_trans_log_buf() into a new function that contains all
of the logic responsible to dirty the transaction, lidp, buffer and
bli. This new function can be used in the future for the use cases
outlined above. This patch does not introduce functional changes.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/xfs_trans.h     |  4 +++-
 fs/xfs/xfs_trans_buf.c | 46 ++++++++++++++++++++++++++++++----------------
 2 files changed, 33 insertions(+), 17 deletions(-)

diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h
index 7d627721e4b3..b3632ebe9d9d 100644
--- a/fs/xfs/xfs_trans.h
+++ b/fs/xfs/xfs_trans.h
@@ -217,7 +217,9 @@ void		xfs_trans_dquot_buf(xfs_trans_t *, struct xfs_buf *, uint);
 void		xfs_trans_inode_alloc_buf(xfs_trans_t *, struct xfs_buf *);
 void		xfs_trans_ichgtime(struct xfs_trans *, struct xfs_inode *, int);
 void		xfs_trans_ijoin(struct xfs_trans *, struct xfs_inode *, uint);
-void		xfs_trans_log_buf(xfs_trans_t *, struct xfs_buf *, uint, uint);
+void		xfs_trans_log_buf(struct xfs_trans *, struct xfs_buf *, uint,
+				  uint);
+void		xfs_trans_dirty_buf(struct xfs_trans *, struct xfs_buf *);
 void		xfs_trans_log_inode(xfs_trans_t *, struct xfs_inode *, uint);
 
 void		xfs_extent_free_init_defer_op(void);
diff --git a/fs/xfs/xfs_trans_buf.c b/fs/xfs/xfs_trans_buf.c
index cac8abbeca3f..8c99813e5377 100644
--- a/fs/xfs/xfs_trans_buf.c
+++ b/fs/xfs/xfs_trans_buf.c
@@ -493,25 +493,17 @@ xfs_trans_bhold_release(xfs_trans_t	*tp,
 }
 
 /*
- * This is called to mark bytes first through last inclusive of the given
- * buffer as needing to be logged when the transaction is committed.
- * The buffer must already be associated with the given transaction.
- *
- * First and last are numbers relative to the beginning of this buffer,
- * so the first byte in the buffer is numbered 0 regardless of the
- * value of b_blkno.
+ * Mark a buffer dirty in the transaction.
  */
 void
-xfs_trans_log_buf(xfs_trans_t	*tp,
-		  xfs_buf_t	*bp,
-		  uint		first,
-		  uint		last)
+xfs_trans_dirty_buf(
+	struct xfs_trans	*tp,
+	struct xfs_buf		*bp)
 {
-	xfs_buf_log_item_t	*bip = bp->b_fspriv;
+	struct xfs_buf_log_item	*bip = bp->b_fspriv;
 
 	ASSERT(bp->b_transp == tp);
 	ASSERT(bip != NULL);
-	ASSERT(first <= last && last < BBTOB(bp->b_length));
 	ASSERT(bp->b_iodone == NULL ||
 	       bp->b_iodone == xfs_buf_iodone_callbacks);
 
@@ -531,8 +523,6 @@ xfs_trans_log_buf(xfs_trans_t	*tp,
 	bp->b_iodone = xfs_buf_iodone_callbacks;
 	bip->bli_item.li_cb = xfs_buf_iodone;
 
-	trace_xfs_trans_log_buf(bip);
-
 	/*
 	 * If we invalidated the buffer within this transaction, then
 	 * cancel the invalidation now that we're dirtying the buffer
@@ -545,15 +535,39 @@ xfs_trans_log_buf(xfs_trans_t	*tp,
 		bp->b_flags &= ~XBF_STALE;
 		bip->__bli_format.blf_flags &= ~XFS_BLF_CANCEL;
 	}
+	bip->bli_flags |= XFS_BLI_DIRTY | XFS_BLI_LOGGED;
 
 	tp->t_flags |= XFS_TRANS_DIRTY;
 	bip->bli_item.li_desc->lid_flags |= XFS_LID_DIRTY;
+}
+
+/*
+ * This is called to mark bytes first through last inclusive of the given
+ * buffer as needing to be logged when the transaction is committed.
+ * The buffer must already be associated with the given transaction.
+ *
+ * First and last are numbers relative to the beginning of this buffer,
+ * so the first byte in the buffer is numbered 0 regardless of the
+ * value of b_blkno.
+ */
+void
+xfs_trans_log_buf(
+	struct xfs_trans	*tp,
+	struct xfs_buf		*bp,
+	uint			first,
+	uint			last)
+{
+	struct xfs_buf_log_item	*bip = bp->b_fspriv;
+
+	ASSERT(first <= last && last < BBTOB(bp->b_length));
+
+	xfs_trans_dirty_buf(tp, bp);
 
 	/*
 	 * If we have an ordered buffer we are not logging any dirty range but
 	 * it still needs to be marked dirty and that it has been logged.
 	 */
-	bip->bli_flags |= XFS_BLI_DIRTY | XFS_BLI_LOGGED;
+	trace_xfs_trans_log_buf(bip);
 	if (!(bip->bli_flags & XFS_BLI_ORDERED))
 		xfs_buf_item_log(bip, first, last);
 }
-- 
2.14.1

Patch "xfs: refactor buffer logging into buffer dirtying helper" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: refactor buffer logging into buffer dirtying helper

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:21 -0700
Subject: xfs: refactor buffer logging into buffer dirtying helper
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, Brian Foster <bfoster@redhat.com>, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-16-hch@lst.de>

From: Brian Foster <bfoster@redhat.com>

commit 9684010d38eccda733b61106765e9357cf436f65 upstream.

xfs_trans_log_buf() is responsible for logging the dirty segments of
a buffer along with setting all of the necessary state on the
transaction, buffer, bli, etc., to ensure that the associated items
are marked as dirty and prepared for I/O. We have a couple use cases
that need to to dirty a buffer in a transaction without actually
logging dirty ranges of the buffer.  One existing use case is
ordered buffers, which are currently logged with arbitrary ranges to
accomplish this even though the content of ordered buffers is never
written to the log. Another pending use case is to relog an already
dirty buffer across rolled transactions within the deferred
operations infrastructure. This is required to prevent a held
(XFS_BLI_HOLD) buffer from pinning the tail of the log.

Refactor xfs_trans_log_buf() into a new function that contains all
of the logic responsible to dirty the transaction, lidp, buffer and
bli. This new function can be used in the future for the use cases
outlined above. This patch does not introduce functional changes.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/xfs_trans.h     |    4 +++-
 fs/xfs/xfs_trans_buf.c |   46 ++++++++++++++++++++++++++++++----------------
 2 files changed, 33 insertions(+), 17 deletions(-)

--- a/fs/xfs/xfs_trans.h
+++ b/fs/xfs/xfs_trans.h
@@ -217,7 +217,9 @@ void		xfs_trans_dquot_buf(xfs_trans_t *,
 void		xfs_trans_inode_alloc_buf(xfs_trans_t *, struct xfs_buf *);
 void		xfs_trans_ichgtime(struct xfs_trans *, struct xfs_inode *, int);
 void		xfs_trans_ijoin(struct xfs_trans *, struct xfs_inode *, uint);
-void		xfs_trans_log_buf(xfs_trans_t *, struct xfs_buf *, uint, uint);
+void		xfs_trans_log_buf(struct xfs_trans *, struct xfs_buf *, uint,
+				  uint);
+void		xfs_trans_dirty_buf(struct xfs_trans *, struct xfs_buf *);
 void		xfs_trans_log_inode(xfs_trans_t *, struct xfs_inode *, uint);
 
 void		xfs_extent_free_init_defer_op(void);
--- a/fs/xfs/xfs_trans_buf.c
+++ b/fs/xfs/xfs_trans_buf.c
@@ -493,25 +493,17 @@ xfs_trans_bhold_release(xfs_trans_t	*tp,
 }
 
 /*
- * This is called to mark bytes first through last inclusive of the given
- * buffer as needing to be logged when the transaction is committed.
- * The buffer must already be associated with the given transaction.
- *
- * First and last are numbers relative to the beginning of this buffer,
- * so the first byte in the buffer is numbered 0 regardless of the
- * value of b_blkno.
+ * Mark a buffer dirty in the transaction.
  */
 void
-xfs_trans_log_buf(xfs_trans_t	*tp,
-		  xfs_buf_t	*bp,
-		  uint		first,
-		  uint		last)
+xfs_trans_dirty_buf(
+	struct xfs_trans	*tp,
+	struct xfs_buf		*bp)
 {
-	xfs_buf_log_item_t	*bip = bp->b_fspriv;
+	struct xfs_buf_log_item	*bip = bp->b_fspriv;
 
 	ASSERT(bp->b_transp == tp);
 	ASSERT(bip != NULL);
-	ASSERT(first <= last && last < BBTOB(bp->b_length));
 	ASSERT(bp->b_iodone == NULL ||
 	       bp->b_iodone == xfs_buf_iodone_callbacks);
 
@@ -531,8 +523,6 @@ xfs_trans_log_buf(xfs_trans_t	*tp,
 	bp->b_iodone = xfs_buf_iodone_callbacks;
 	bip->bli_item.li_cb = xfs_buf_iodone;
 
-	trace_xfs_trans_log_buf(bip);
-
 	/*
 	 * If we invalidated the buffer within this transaction, then
 	 * cancel the invalidation now that we're dirtying the buffer
@@ -545,15 +535,39 @@ xfs_trans_log_buf(xfs_trans_t	*tp,
 		bp->b_flags &= ~XBF_STALE;
 		bip->__bli_format.blf_flags &= ~XFS_BLF_CANCEL;
 	}
+	bip->bli_flags |= XFS_BLI_DIRTY | XFS_BLI_LOGGED;
 
 	tp->t_flags |= XFS_TRANS_DIRTY;
 	bip->bli_item.li_desc->lid_flags |= XFS_LID_DIRTY;
+}
+
+/*
+ * This is called to mark bytes first through last inclusive of the given
+ * buffer as needing to be logged when the transaction is committed.
+ * The buffer must already be associated with the given transaction.
+ *
+ * First and last are numbers relative to the beginning of this buffer,
+ * so the first byte in the buffer is numbered 0 regardless of the
+ * value of b_blkno.
+ */
+void
+xfs_trans_log_buf(
+	struct xfs_trans	*tp,
+	struct xfs_buf		*bp,
+	uint			first,
+	uint			last)
+{
+	struct xfs_buf_log_item	*bip = bp->b_fspriv;
+
+	ASSERT(first <= last && last < BBTOB(bp->b_length));
+
+	xfs_trans_dirty_buf(tp, bp);
 
 	/*
 	 * If we have an ordered buffer we are not logging any dirty range but
 	 * it still needs to be marked dirty and that it has been logged.
 	 */
-	bip->bli_flags |= XFS_BLI_DIRTY | XFS_BLI_LOGGED;
+	trace_xfs_trans_log_buf(bip);
 	if (!(bip->bli_flags & XFS_BLI_ORDERED))
 		xfs_buf_item_log(bip, first, last);
 }


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 16/25] xfs: don't log dirty ranges for ordered buffers

[Christoph Hellwig]

From: Brian Foster <bfoster@redhat.com>

commit 8dc518dfa7dbd079581269e51074b3c55a65a880 upstream.

Ordered buffers are attached to transactions and pushed through the
logging infrastructure just like normal buffers with the exception
that they are not actually written to the log. Therefore, we don't
need to log dirty ranges of ordered buffers. xfs_trans_log_buf() is
called on ordered buffers to set up all of the dirty state on the
transaction, buffer and log item and prepare the buffer for I/O.

Now that xfs_trans_dirty_buf() is available, call it from
xfs_trans_ordered_buf() so the latter is now mutually exclusive with
xfs_trans_log_buf(). This reflects the implementation of ordered
buffers and helps eliminate confusion over the need to log ranges of
ordered buffers just to set up internal log state.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/libxfs/xfs_btree.c  |  6 ++----
 fs/xfs/libxfs/xfs_ialloc.c |  2 --
 fs/xfs/xfs_trans_buf.c     | 26 ++++++++++++++------------
 3 files changed, 16 insertions(+), 18 deletions(-)

diff --git a/fs/xfs/libxfs/xfs_btree.c b/fs/xfs/libxfs/xfs_btree.c
index e0bcc4a59efd..0b7905a4e27a 100644
--- a/fs/xfs/libxfs/xfs_btree.c
+++ b/fs/xfs/libxfs/xfs_btree.c
@@ -4464,12 +4464,10 @@ xfs_btree_block_change_owner(
 	 * though, so everything is consistent in memory.
 	 */
 	if (bp) {
-		if (cur->bc_tp) {
+		if (cur->bc_tp)
 			xfs_trans_ordered_buf(cur->bc_tp, bp);
-			xfs_btree_log_block(cur, bp, XFS_BB_OWNER);
-		} else {
+		else
 			xfs_buf_delwri_queue(bp, bbcoi->buffer_list);
-		}
 	} else {
 		ASSERT(cur->bc_flags & XFS_BTREE_ROOT_IN_INODE);
 		ASSERT(level == cur->bc_nlevels - 1);
diff --git a/fs/xfs/libxfs/xfs_ialloc.c b/fs/xfs/libxfs/xfs_ialloc.c
index 1e0658a3f155..988bb3f31446 100644
--- a/fs/xfs/libxfs/xfs_ialloc.c
+++ b/fs/xfs/libxfs/xfs_ialloc.c
@@ -378,8 +378,6 @@ xfs_ialloc_inode_init(
 				 * transaction and pin the log appropriately.
 				 */
 				xfs_trans_ordered_buf(tp, fbuf);
-				xfs_trans_log_buf(tp, fbuf, 0,
-						  BBTOB(fbuf->b_length) - 1);
 			}
 		} else {
 			fbuf->b_flags |= XBF_DONE;
diff --git a/fs/xfs/xfs_trans_buf.c b/fs/xfs/xfs_trans_buf.c
index 8c99813e5377..3089e8015369 100644
--- a/fs/xfs/xfs_trans_buf.c
+++ b/fs/xfs/xfs_trans_buf.c
@@ -560,16 +560,12 @@ xfs_trans_log_buf(
 	struct xfs_buf_log_item	*bip = bp->b_fspriv;
 
 	ASSERT(first <= last && last < BBTOB(bp->b_length));
+	ASSERT(!(bip->bli_flags & XFS_BLI_ORDERED));
 
 	xfs_trans_dirty_buf(tp, bp);
 
-	/*
-	 * If we have an ordered buffer we are not logging any dirty range but
-	 * it still needs to be marked dirty and that it has been logged.
-	 */
 	trace_xfs_trans_log_buf(bip);
-	if (!(bip->bli_flags & XFS_BLI_ORDERED))
-		xfs_buf_item_log(bip, first, last);
+	xfs_buf_item_log(bip, first, last);
 }
 
 
@@ -722,12 +718,11 @@ xfs_trans_inode_alloc_buf(
 }
 
 /*
- * Mark the buffer as ordered for this transaction. This means
- * that the contents of the buffer are not recorded in the transaction
- * but it is tracked in the AIL as though it was. This allows us
- * to record logical changes in transactions rather than the physical
- * changes we make to the buffer without changing writeback ordering
- * constraints of metadata buffers.
+ * Mark the buffer as ordered for this transaction. This means that the contents
+ * of the buffer are not recorded in the transaction but it is tracked in the
+ * AIL as though it was. This allows us to record logical changes in
+ * transactions rather than the physical changes we make to the buffer without
+ * changing writeback ordering constraints of metadata buffers.
  */
 void
 xfs_trans_ordered_buf(
@@ -739,9 +734,16 @@ xfs_trans_ordered_buf(
 	ASSERT(bp->b_transp == tp);
 	ASSERT(bip != NULL);
 	ASSERT(atomic_read(&bip->bli_refcount) > 0);
+	ASSERT(!xfs_buf_item_dirty_format(bip));
 
 	bip->bli_flags |= XFS_BLI_ORDERED;
 	trace_xfs_buf_item_ordered(bip);
+
+	/*
+	 * We don't log a dirty range of an ordered buffer but it still needs
+	 * to be marked dirty and that it has been logged.
+	 */
+	xfs_trans_dirty_buf(tp, bp);
 }
 
 /*
-- 
2.14.1

Patch "xfs: don't log dirty ranges for ordered buffers" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: don't log dirty ranges for ordered buffers

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:22 -0700
Subject: xfs: don't log dirty ranges for ordered buffers
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, Brian Foster <bfoster@redhat.com>, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-17-hch@lst.de>

From: Brian Foster <bfoster@redhat.com>

commit 8dc518dfa7dbd079581269e51074b3c55a65a880 upstream.

Ordered buffers are attached to transactions and pushed through the
logging infrastructure just like normal buffers with the exception
that they are not actually written to the log. Therefore, we don't
need to log dirty ranges of ordered buffers. xfs_trans_log_buf() is
called on ordered buffers to set up all of the dirty state on the
transaction, buffer and log item and prepare the buffer for I/O.

Now that xfs_trans_dirty_buf() is available, call it from
xfs_trans_ordered_buf() so the latter is now mutually exclusive with
xfs_trans_log_buf(). This reflects the implementation of ordered
buffers and helps eliminate confusion over the need to log ranges of
ordered buffers just to set up internal log state.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/libxfs/xfs_btree.c  |    6 ++----
 fs/xfs/libxfs/xfs_ialloc.c |    2 --
 fs/xfs/xfs_trans_buf.c     |   26 ++++++++++++++------------
 3 files changed, 16 insertions(+), 18 deletions(-)

--- a/fs/xfs/libxfs/xfs_btree.c
+++ b/fs/xfs/libxfs/xfs_btree.c
@@ -4464,12 +4464,10 @@ xfs_btree_block_change_owner(
 	 * though, so everything is consistent in memory.
 	 */
 	if (bp) {
-		if (cur->bc_tp) {
+		if (cur->bc_tp)
 			xfs_trans_ordered_buf(cur->bc_tp, bp);
-			xfs_btree_log_block(cur, bp, XFS_BB_OWNER);
-		} else {
+		else
 			xfs_buf_delwri_queue(bp, bbcoi->buffer_list);
-		}
 	} else {
 		ASSERT(cur->bc_flags & XFS_BTREE_ROOT_IN_INODE);
 		ASSERT(level == cur->bc_nlevels - 1);
--- a/fs/xfs/libxfs/xfs_ialloc.c
+++ b/fs/xfs/libxfs/xfs_ialloc.c
@@ -378,8 +378,6 @@ xfs_ialloc_inode_init(
 				 * transaction and pin the log appropriately.
 				 */
 				xfs_trans_ordered_buf(tp, fbuf);
-				xfs_trans_log_buf(tp, fbuf, 0,
-						  BBTOB(fbuf->b_length) - 1);
 			}
 		} else {
 			fbuf->b_flags |= XBF_DONE;
--- a/fs/xfs/xfs_trans_buf.c
+++ b/fs/xfs/xfs_trans_buf.c
@@ -560,16 +560,12 @@ xfs_trans_log_buf(
 	struct xfs_buf_log_item	*bip = bp->b_fspriv;
 
 	ASSERT(first <= last && last < BBTOB(bp->b_length));
+	ASSERT(!(bip->bli_flags & XFS_BLI_ORDERED));
 
 	xfs_trans_dirty_buf(tp, bp);
 
-	/*
-	 * If we have an ordered buffer we are not logging any dirty range but
-	 * it still needs to be marked dirty and that it has been logged.
-	 */
 	trace_xfs_trans_log_buf(bip);
-	if (!(bip->bli_flags & XFS_BLI_ORDERED))
-		xfs_buf_item_log(bip, first, last);
+	xfs_buf_item_log(bip, first, last);
 }
 
 
@@ -722,12 +718,11 @@ xfs_trans_inode_alloc_buf(
 }
 
 /*
- * Mark the buffer as ordered for this transaction. This means
- * that the contents of the buffer are not recorded in the transaction
- * but it is tracked in the AIL as though it was. This allows us
- * to record logical changes in transactions rather than the physical
- * changes we make to the buffer without changing writeback ordering
- * constraints of metadata buffers.
+ * Mark the buffer as ordered for this transaction. This means that the contents
+ * of the buffer are not recorded in the transaction but it is tracked in the
+ * AIL as though it was. This allows us to record logical changes in
+ * transactions rather than the physical changes we make to the buffer without
+ * changing writeback ordering constraints of metadata buffers.
  */
 void
 xfs_trans_ordered_buf(
@@ -739,9 +734,16 @@ xfs_trans_ordered_buf(
 	ASSERT(bp->b_transp == tp);
 	ASSERT(bip != NULL);
 	ASSERT(atomic_read(&bip->bli_refcount) > 0);
+	ASSERT(!xfs_buf_item_dirty_format(bip));
 
 	bip->bli_flags |= XFS_BLI_ORDERED;
 	trace_xfs_buf_item_ordered(bip);
+
+	/*
+	 * We don't log a dirty range of an ordered buffer but it still needs
+	 * to be marked dirty and that it has been logged.
+	 */
+	xfs_trans_dirty_buf(tp, bp);
 }
 
 /*


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 17/25] xfs: skip bmbt block ino validation during owner change

[Christoph Hellwig]

From: Brian Foster <bfoster@redhat.com>

commit 99c794c639a65cc7b74f30a674048fd100fe9ac8 upstream.

Extent swap uses xfs_btree_visit_blocks() to fix up bmbt block
owners on v5 (!rmapbt) filesystems. The bmbt scan uses
xfs_btree_lookup_get_block() to read bmbt blocks which verifies the
current owner of the block against the parent inode of the bmbt.
This works during extent swap because the bmbt owners are updated to
the opposite inode number before the inode extent forks are swapped.

The modified bmbt blocks are marked as ordered buffers which allows
everything to commit in a single transaction. If the transaction
commits to the log and the system crashes such that recovery of the
extent swap is required, log recovery restarts the bmbt scan to fix
up any bmbt blocks that may have not been written back before the
crash. The log recovery bmbt scan occurs after the inode forks have
been swapped, however. This causes the bmbt block owner verification
to fail, leads to log recovery failure and requires xfs_repair to
zap the log to recover.

Define a new invalid inode owner flag to inform the btree block
lookup mechanism that the current inode may be invalid with respect
to the current owner of the bmbt block. Set this flag on the cursor
used for change owner scans to allow this operation to work at
runtime and during log recovery.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Fixes: bb3be7e7c ("xfs: check for bogus values in btree block headers")
Cc: stable@vger.kernel.org
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/libxfs/xfs_bmap_btree.c | 1 +
 fs/xfs/libxfs/xfs_btree.c      | 1 +
 fs/xfs/libxfs/xfs_btree.h      | 3 ++-
 3 files changed, 4 insertions(+), 1 deletion(-)

diff --git a/fs/xfs/libxfs/xfs_bmap_btree.c b/fs/xfs/libxfs/xfs_bmap_btree.c
index 85de22513014..a6331ffa51e3 100644
--- a/fs/xfs/libxfs/xfs_bmap_btree.c
+++ b/fs/xfs/libxfs/xfs_bmap_btree.c
@@ -858,6 +858,7 @@ xfs_bmbt_change_owner(
 	cur = xfs_bmbt_init_cursor(ip->i_mount, tp, ip, whichfork);
 	if (!cur)
 		return -ENOMEM;
+	cur->bc_private.b.flags |= XFS_BTCUR_BPRV_INVALID_OWNER;
 
 	error = xfs_btree_change_owner(cur, new_owner, buffer_list);
 	xfs_btree_del_cursor(cur, error ? XFS_BTREE_ERROR : XFS_BTREE_NOERROR);
diff --git a/fs/xfs/libxfs/xfs_btree.c b/fs/xfs/libxfs/xfs_btree.c
index 0b7905a4e27a..1d15d04e29a5 100644
--- a/fs/xfs/libxfs/xfs_btree.c
+++ b/fs/xfs/libxfs/xfs_btree.c
@@ -1791,6 +1791,7 @@ xfs_btree_lookup_get_block(
 
 	/* Check the inode owner since the verifiers don't. */
 	if (xfs_sb_version_hascrc(&cur->bc_mp->m_sb) &&
+	    !(cur->bc_private.b.flags & XFS_BTCUR_BPRV_INVALID_OWNER) &&
 	    (cur->bc_flags & XFS_BTREE_LONG_PTRS) &&
 	    be64_to_cpu((*blkp)->bb_u.l.bb_owner) !=
 			cur->bc_private.b.ip->i_ino)
diff --git a/fs/xfs/libxfs/xfs_btree.h b/fs/xfs/libxfs/xfs_btree.h
index 9c95e965cfe5..f2a88c3b1159 100644
--- a/fs/xfs/libxfs/xfs_btree.h
+++ b/fs/xfs/libxfs/xfs_btree.h
@@ -233,7 +233,8 @@ typedef struct xfs_btree_cur
 			short		forksize;	/* fork's inode space */
 			char		whichfork;	/* data or attr fork */
 			char		flags;		/* flags */
-#define	XFS_BTCUR_BPRV_WASDEL	1			/* was delayed */
+#define	XFS_BTCUR_BPRV_WASDEL		(1<<0)		/* was delayed */
+#define	XFS_BTCUR_BPRV_INVALID_OWNER	(1<<1)		/* for ext swap */
 		} b;
 	}		bc_private;	/* per-btree type data */
 } xfs_btree_cur_t;
-- 
2.14.1

Patch "xfs: skip bmbt block ino validation during owner change" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: skip bmbt block ino validation during owner change

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:23 -0700
Subject: xfs: skip bmbt block ino validation during owner change
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, Brian Foster <bfoster@redhat.com>, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-18-hch@lst.de>

From: Brian Foster <bfoster@redhat.com>

commit 99c794c639a65cc7b74f30a674048fd100fe9ac8 upstream.

Extent swap uses xfs_btree_visit_blocks() to fix up bmbt block
owners on v5 (!rmapbt) filesystems. The bmbt scan uses
xfs_btree_lookup_get_block() to read bmbt blocks which verifies the
current owner of the block against the parent inode of the bmbt.
This works during extent swap because the bmbt owners are updated to
the opposite inode number before the inode extent forks are swapped.

The modified bmbt blocks are marked as ordered buffers which allows
everything to commit in a single transaction. If the transaction
commits to the log and the system crashes such that recovery of the
extent swap is required, log recovery restarts the bmbt scan to fix
up any bmbt blocks that may have not been written back before the
crash. The log recovery bmbt scan occurs after the inode forks have
been swapped, however. This causes the bmbt block owner verification
to fail, leads to log recovery failure and requires xfs_repair to
zap the log to recover.

Define a new invalid inode owner flag to inform the btree block
lookup mechanism that the current inode may be invalid with respect
to the current owner of the bmbt block. Set this flag on the cursor
used for change owner scans to allow this operation to work at
runtime and during log recovery.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Fixes: bb3be7e7c ("xfs: check for bogus values in btree block headers")
Cc: stable@vger.kernel.org
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/libxfs/xfs_bmap_btree.c |    1 +
 fs/xfs/libxfs/xfs_btree.c      |    1 +
 fs/xfs/libxfs/xfs_btree.h      |    3 ++-
 3 files changed, 4 insertions(+), 1 deletion(-)

--- a/fs/xfs/libxfs/xfs_bmap_btree.c
+++ b/fs/xfs/libxfs/xfs_bmap_btree.c
@@ -858,6 +858,7 @@ xfs_bmbt_change_owner(
 	cur = xfs_bmbt_init_cursor(ip->i_mount, tp, ip, whichfork);
 	if (!cur)
 		return -ENOMEM;
+	cur->bc_private.b.flags |= XFS_BTCUR_BPRV_INVALID_OWNER;
 
 	error = xfs_btree_change_owner(cur, new_owner, buffer_list);
 	xfs_btree_del_cursor(cur, error ? XFS_BTREE_ERROR : XFS_BTREE_NOERROR);
--- a/fs/xfs/libxfs/xfs_btree.c
+++ b/fs/xfs/libxfs/xfs_btree.c
@@ -1791,6 +1791,7 @@ xfs_btree_lookup_get_block(
 
 	/* Check the inode owner since the verifiers don't. */
 	if (xfs_sb_version_hascrc(&cur->bc_mp->m_sb) &&
+	    !(cur->bc_private.b.flags & XFS_BTCUR_BPRV_INVALID_OWNER) &&
 	    (cur->bc_flags & XFS_BTREE_LONG_PTRS) &&
 	    be64_to_cpu((*blkp)->bb_u.l.bb_owner) !=
 			cur->bc_private.b.ip->i_ino)
--- a/fs/xfs/libxfs/xfs_btree.h
+++ b/fs/xfs/libxfs/xfs_btree.h
@@ -233,7 +233,8 @@ typedef struct xfs_btree_cur
 			short		forksize;	/* fork's inode space */
 			char		whichfork;	/* data or attr fork */
 			char		flags;		/* flags */
-#define	XFS_BTCUR_BPRV_WASDEL	1			/* was delayed */
+#define	XFS_BTCUR_BPRV_WASDEL		(1<<0)		/* was delayed */
+#define	XFS_BTCUR_BPRV_INVALID_OWNER	(1<<1)		/* for ext swap */
 		} b;
 	}		bc_private;	/* per-btree type data */
 } xfs_btree_cur_t;


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 18/25] xfs: move bmbt owner change to last step of extent swap

[Christoph Hellwig]

From: Brian Foster <bfoster@redhat.com>

commit 6fb10d6d22094bc4062f92b9ccbcee2f54033d04 upstream.

The extent swap operation currently resets bmbt block owners before
the inode forks are swapped. The bmbt buffers are marked as ordered
so they do not have to be physically logged in the transaction.

This use of ordered buffers is not safe as bmbt buffers may have
been previously physically logged. The bmbt owner change algorithm
needs to be updated to physically log buffers that are already dirty
when/if they are encountered. This means that an extent swap will
eventually require multiple rolling transactions to handle large
btrees. In addition, all inode related changes must be logged before
the bmbt owner change scan begins and can roll the transaction for
the first time to preserve fs consistency via log recovery.

In preparation for such fixes to the bmbt owner change algorithm,
refactor the bmbt scan out of the extent fork swap code to the last
operation before the transaction is committed. Update
xfs_swap_extent_forks() to only set the inode log flags when an
owner change scan is necessary. Update xfs_swap_extents() to trigger
the owner change based on the inode log flags. Note that since the
owner change now occurs after the extent fork swap, the inode btrees
must be fixed up with the inode number of the current inode (similar
to log recovery).

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/xfs_bmap_util.c | 44 ++++++++++++++++++++++++++------------------
 1 file changed, 26 insertions(+), 18 deletions(-)

diff --git a/fs/xfs/xfs_bmap_util.c b/fs/xfs/xfs_bmap_util.c
index 93e955262d07..ee8fb9a9ed73 100644
--- a/fs/xfs/xfs_bmap_util.c
+++ b/fs/xfs/xfs_bmap_util.c
@@ -1840,29 +1840,18 @@ xfs_swap_extent_forks(
 	}
 
 	/*
-	 * Before we've swapped the forks, lets set the owners of the forks
-	 * appropriately. We have to do this as we are demand paging the btree
-	 * buffers, and so the validation done on read will expect the owner
-	 * field to be correctly set. Once we change the owners, we can swap the
-	 * inode forks.
+	 * Btree format (v3) inodes have the inode number stamped in the bmbt
+	 * block headers. We can't start changing the bmbt blocks until the
+	 * inode owner change is logged so recovery does the right thing in the
+	 * event of a crash. Set the owner change log flags now and leave the
+	 * bmbt scan as the last step.
 	 */
 	if (ip->i_d.di_version == 3 &&
-	    ip->i_d.di_format == XFS_DINODE_FMT_BTREE) {
+	    ip->i_d.di_format == XFS_DINODE_FMT_BTREE)
 		(*target_log_flags) |= XFS_ILOG_DOWNER;
-		error = xfs_bmbt_change_owner(tp, ip, XFS_DATA_FORK,
-					      tip->i_ino, NULL);
-		if (error)
-			return error;
-	}
-
 	if (tip->i_d.di_version == 3 &&
-	    tip->i_d.di_format == XFS_DINODE_FMT_BTREE) {
+	    tip->i_d.di_format == XFS_DINODE_FMT_BTREE)
 		(*src_log_flags) |= XFS_ILOG_DOWNER;
-		error = xfs_bmbt_change_owner(tp, tip, XFS_DATA_FORK,
-					      ip->i_ino, NULL);
-		if (error)
-			return error;
-	}
 
 	/*
 	 * Swap the data forks of the inodes
@@ -2091,6 +2080,25 @@ xfs_swap_extents(
 	xfs_trans_log_inode(tp, ip,  src_log_flags);
 	xfs_trans_log_inode(tp, tip, target_log_flags);
 
+	/*
+	 * The extent forks have been swapped, but crc=1,rmapbt=0 filesystems
+	 * have inode number owner values in the bmbt blocks that still refer to
+	 * the old inode. Scan each bmbt to fix up the owner values with the
+	 * inode number of the current inode.
+	 */
+	if (src_log_flags & XFS_ILOG_DOWNER) {
+		error = xfs_bmbt_change_owner(tp, ip, XFS_DATA_FORK,
+					      ip->i_ino, NULL);
+		if (error)
+			goto out_trans_cancel;
+	}
+	if (target_log_flags & XFS_ILOG_DOWNER) {
+		error = xfs_bmbt_change_owner(tp, tip, XFS_DATA_FORK,
+					      tip->i_ino, NULL);
+		if (error)
+			goto out_trans_cancel;
+	}
+
 	/*
 	 * If this is a synchronous mount, make sure that the
 	 * transaction goes to disk before returning to the user.
-- 
2.14.1

Patch "xfs: move bmbt owner change to last step of extent swap" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: move bmbt owner change to last step of extent swap

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:24 -0700
Subject: xfs: move bmbt owner change to last step of extent swap
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, Brian Foster <bfoster@redhat.com>, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-19-hch@lst.de>

From: Brian Foster <bfoster@redhat.com>

commit 6fb10d6d22094bc4062f92b9ccbcee2f54033d04 upstream.

The extent swap operation currently resets bmbt block owners before
the inode forks are swapped. The bmbt buffers are marked as ordered
so they do not have to be physically logged in the transaction.

This use of ordered buffers is not safe as bmbt buffers may have
been previously physically logged. The bmbt owner change algorithm
needs to be updated to physically log buffers that are already dirty
when/if they are encountered. This means that an extent swap will
eventually require multiple rolling transactions to handle large
btrees. In addition, all inode related changes must be logged before
the bmbt owner change scan begins and can roll the transaction for
the first time to preserve fs consistency via log recovery.

In preparation for such fixes to the bmbt owner change algorithm,
refactor the bmbt scan out of the extent fork swap code to the last
operation before the transaction is committed. Update
xfs_swap_extent_forks() to only set the inode log flags when an
owner change scan is necessary. Update xfs_swap_extents() to trigger
the owner change based on the inode log flags. Note that since the
owner change now occurs after the extent fork swap, the inode btrees
must be fixed up with the inode number of the current inode (similar
to log recovery).

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/xfs_bmap_util.c |   44 ++++++++++++++++++++++++++------------------
 1 file changed, 26 insertions(+), 18 deletions(-)

--- a/fs/xfs/xfs_bmap_util.c
+++ b/fs/xfs/xfs_bmap_util.c
@@ -1840,29 +1840,18 @@ xfs_swap_extent_forks(
 	}
 
 	/*
-	 * Before we've swapped the forks, lets set the owners of the forks
-	 * appropriately. We have to do this as we are demand paging the btree
-	 * buffers, and so the validation done on read will expect the owner
-	 * field to be correctly set. Once we change the owners, we can swap the
-	 * inode forks.
+	 * Btree format (v3) inodes have the inode number stamped in the bmbt
+	 * block headers. We can't start changing the bmbt blocks until the
+	 * inode owner change is logged so recovery does the right thing in the
+	 * event of a crash. Set the owner change log flags now and leave the
+	 * bmbt scan as the last step.
 	 */
 	if (ip->i_d.di_version == 3 &&
-	    ip->i_d.di_format == XFS_DINODE_FMT_BTREE) {
+	    ip->i_d.di_format == XFS_DINODE_FMT_BTREE)
 		(*target_log_flags) |= XFS_ILOG_DOWNER;
-		error = xfs_bmbt_change_owner(tp, ip, XFS_DATA_FORK,
-					      tip->i_ino, NULL);
-		if (error)
-			return error;
-	}
-
 	if (tip->i_d.di_version == 3 &&
-	    tip->i_d.di_format == XFS_DINODE_FMT_BTREE) {
+	    tip->i_d.di_format == XFS_DINODE_FMT_BTREE)
 		(*src_log_flags) |= XFS_ILOG_DOWNER;
-		error = xfs_bmbt_change_owner(tp, tip, XFS_DATA_FORK,
-					      ip->i_ino, NULL);
-		if (error)
-			return error;
-	}
 
 	/*
 	 * Swap the data forks of the inodes
@@ -2092,6 +2081,25 @@ xfs_swap_extents(
 	xfs_trans_log_inode(tp, tip, target_log_flags);
 
 	/*
+	 * The extent forks have been swapped, but crc=1,rmapbt=0 filesystems
+	 * have inode number owner values in the bmbt blocks that still refer to
+	 * the old inode. Scan each bmbt to fix up the owner values with the
+	 * inode number of the current inode.
+	 */
+	if (src_log_flags & XFS_ILOG_DOWNER) {
+		error = xfs_bmbt_change_owner(tp, ip, XFS_DATA_FORK,
+					      ip->i_ino, NULL);
+		if (error)
+			goto out_trans_cancel;
+	}
+	if (target_log_flags & XFS_ILOG_DOWNER) {
+		error = xfs_bmbt_change_owner(tp, tip, XFS_DATA_FORK,
+					      tip->i_ino, NULL);
+		if (error)
+			goto out_trans_cancel;
+	}
+
+	/*
 	 * If this is a synchronous mount, make sure that the
 	 * transaction goes to disk before returning to the user.
 	 */


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 19/25] xfs: disallow marking previously dirty buffers as ordered

[Christoph Hellwig]

From: Brian Foster <bfoster@redhat.com>

commit a5814bceea48ee1c57c4db2bd54b0c0246daf54a upstream.

Ordered buffers are used in situations where the buffer is not
physically logged but must pass through the transaction/logging
pipeline for a particular transaction. As a result, ordered buffers
are not unpinned and written back until the transaction commits to
the log. Ordered buffers have a strict requirement that the target
buffer must not be currently dirty and resident in the log pipeline
at the time it is marked ordered. If a dirty+ordered buffer is
committed, the buffer is reinserted to the AIL but not physically
relogged at the LSN of the associated checkpoint. The buffer log
item is assigned the LSN of the latest checkpoint and the AIL
effectively releases the previously logged buffer content from the
active log before the buffer has been written back. If the tail
pushes forward and a filesystem crash occurs while in this state, an
inconsistent filesystem could result.

It is currently the caller responsibility to ensure an ordered
buffer is not already dirty from a previous modification. This is
unclear and error prone when not used in situations where it is
guaranteed a buffer has not been previously modified (such as new
metadata allocations).

To facilitate general purpose use of ordered buffers, update
xfs_trans_ordered_buf() to conditionally order the buffer based on
state of the log item and return the status of the result. If the
bli is dirty, do not order the buffer and return false. The caller
must either physically log the buffer (having acquired the
appropriate log reservation) or push it from the AIL to clean it
before it can be marked ordered in the current transaction.

Note that ordered buffers are currently only used in two situations:
1.) inode chunk allocation where previously logged buffers are not
possible and 2.) extent swap which will be updated to handle ordered
buffer failures in a separate patch.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/xfs_trans.h     | 2 +-
 fs/xfs/xfs_trans_buf.c | 7 +++++--
 2 files changed, 6 insertions(+), 3 deletions(-)

diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h
index b3632ebe9d9d..4709823e04b9 100644
--- a/fs/xfs/xfs_trans.h
+++ b/fs/xfs/xfs_trans.h
@@ -212,7 +212,7 @@ void		xfs_trans_bhold_release(xfs_trans_t *, struct xfs_buf *);
 void		xfs_trans_binval(xfs_trans_t *, struct xfs_buf *);
 void		xfs_trans_inode_buf(xfs_trans_t *, struct xfs_buf *);
 void		xfs_trans_stale_inode_buf(xfs_trans_t *, struct xfs_buf *);
-void		xfs_trans_ordered_buf(xfs_trans_t *, struct xfs_buf *);
+bool		xfs_trans_ordered_buf(xfs_trans_t *, struct xfs_buf *);
 void		xfs_trans_dquot_buf(xfs_trans_t *, struct xfs_buf *, uint);
 void		xfs_trans_inode_alloc_buf(xfs_trans_t *, struct xfs_buf *);
 void		xfs_trans_ichgtime(struct xfs_trans *, struct xfs_inode *, int);
diff --git a/fs/xfs/xfs_trans_buf.c b/fs/xfs/xfs_trans_buf.c
index 3089e8015369..3ba7a96a8abd 100644
--- a/fs/xfs/xfs_trans_buf.c
+++ b/fs/xfs/xfs_trans_buf.c
@@ -724,7 +724,7 @@ xfs_trans_inode_alloc_buf(
  * transactions rather than the physical changes we make to the buffer without
  * changing writeback ordering constraints of metadata buffers.
  */
-void
+bool
 xfs_trans_ordered_buf(
 	struct xfs_trans	*tp,
 	struct xfs_buf		*bp)
@@ -734,7 +734,9 @@ xfs_trans_ordered_buf(
 	ASSERT(bp->b_transp == tp);
 	ASSERT(bip != NULL);
 	ASSERT(atomic_read(&bip->bli_refcount) > 0);
-	ASSERT(!xfs_buf_item_dirty_format(bip));
+
+	if (xfs_buf_item_dirty_format(bip))
+		return false;
 
 	bip->bli_flags |= XFS_BLI_ORDERED;
 	trace_xfs_buf_item_ordered(bip);
@@ -744,6 +746,7 @@ xfs_trans_ordered_buf(
 	 * to be marked dirty and that it has been logged.
 	 */
 	xfs_trans_dirty_buf(tp, bp);
+	return true;
 }
 
 /*
-- 
2.14.1

Patch "xfs: disallow marking previously dirty buffers as ordered" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: disallow marking previously dirty buffers as ordered

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:25 -0700
Subject: xfs: disallow marking previously dirty buffers as ordered
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, Brian Foster <bfoster@redhat.com>, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-20-hch@lst.de>

From: Brian Foster <bfoster@redhat.com>

commit a5814bceea48ee1c57c4db2bd54b0c0246daf54a upstream.

Ordered buffers are used in situations where the buffer is not
physically logged but must pass through the transaction/logging
pipeline for a particular transaction. As a result, ordered buffers
are not unpinned and written back until the transaction commits to
the log. Ordered buffers have a strict requirement that the target
buffer must not be currently dirty and resident in the log pipeline
at the time it is marked ordered. If a dirty+ordered buffer is
committed, the buffer is reinserted to the AIL but not physically
relogged at the LSN of the associated checkpoint. The buffer log
item is assigned the LSN of the latest checkpoint and the AIL
effectively releases the previously logged buffer content from the
active log before the buffer has been written back. If the tail
pushes forward and a filesystem crash occurs while in this state, an
inconsistent filesystem could result.

It is currently the caller responsibility to ensure an ordered
buffer is not already dirty from a previous modification. This is
unclear and error prone when not used in situations where it is
guaranteed a buffer has not been previously modified (such as new
metadata allocations).

To facilitate general purpose use of ordered buffers, update
xfs_trans_ordered_buf() to conditionally order the buffer based on
state of the log item and return the status of the result. If the
bli is dirty, do not order the buffer and return false. The caller
must either physically log the buffer (having acquired the
appropriate log reservation) or push it from the AIL to clean it
before it can be marked ordered in the current transaction.

Note that ordered buffers are currently only used in two situations:
1.) inode chunk allocation where previously logged buffers are not
possible and 2.) extent swap which will be updated to handle ordered
buffer failures in a separate patch.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/xfs_trans.h     |    2 +-
 fs/xfs/xfs_trans_buf.c |    7 +++++--
 2 files changed, 6 insertions(+), 3 deletions(-)

--- a/fs/xfs/xfs_trans.h
+++ b/fs/xfs/xfs_trans.h
@@ -212,7 +212,7 @@ void		xfs_trans_bhold_release(xfs_trans_
 void		xfs_trans_binval(xfs_trans_t *, struct xfs_buf *);
 void		xfs_trans_inode_buf(xfs_trans_t *, struct xfs_buf *);
 void		xfs_trans_stale_inode_buf(xfs_trans_t *, struct xfs_buf *);
-void		xfs_trans_ordered_buf(xfs_trans_t *, struct xfs_buf *);
+bool		xfs_trans_ordered_buf(xfs_trans_t *, struct xfs_buf *);
 void		xfs_trans_dquot_buf(xfs_trans_t *, struct xfs_buf *, uint);
 void		xfs_trans_inode_alloc_buf(xfs_trans_t *, struct xfs_buf *);
 void		xfs_trans_ichgtime(struct xfs_trans *, struct xfs_inode *, int);
--- a/fs/xfs/xfs_trans_buf.c
+++ b/fs/xfs/xfs_trans_buf.c
@@ -724,7 +724,7 @@ xfs_trans_inode_alloc_buf(
  * transactions rather than the physical changes we make to the buffer without
  * changing writeback ordering constraints of metadata buffers.
  */
-void
+bool
 xfs_trans_ordered_buf(
 	struct xfs_trans	*tp,
 	struct xfs_buf		*bp)
@@ -734,7 +734,9 @@ xfs_trans_ordered_buf(
 	ASSERT(bp->b_transp == tp);
 	ASSERT(bip != NULL);
 	ASSERT(atomic_read(&bip->bli_refcount) > 0);
-	ASSERT(!xfs_buf_item_dirty_format(bip));
+
+	if (xfs_buf_item_dirty_format(bip))
+		return false;
 
 	bip->bli_flags |= XFS_BLI_ORDERED;
 	trace_xfs_buf_item_ordered(bip);
@@ -744,6 +746,7 @@ xfs_trans_ordered_buf(
 	 * to be marked dirty and that it has been logged.
 	 */
 	xfs_trans_dirty_buf(tp, bp);
+	return true;
 }
 
 /*


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 20/25] xfs: relog dirty buffers during swapext bmbt owner change

[Christoph Hellwig]

From: Brian Foster <bfoster@redhat.com>

commit 2dd3d709fc4338681a3aa61658122fa8faa5a437 upstream.

The owner change bmbt scan that occurs during extent swap operations
does not handle ordered buffer failures. Buffers that cannot be
marked ordered must be physically logged so previously dirty ranges
of the buffer can be relogged in the transaction.

Since the bmbt scan may need to process and potentially log a large
number of blocks, we can't expect to complete this operation in a
single transaction. Update extent swap to use a permanent
transaction with enough log reservation to physically log a buffer.
Update the bmbt scan to physically log any buffers that cannot be
ordered and to terminate the scan with -EAGAIN. On -EAGAIN, the
caller rolls the transaction and restarts the scan. Finally, update
the bmbt scan helper function to skip bmbt blocks that already match
the expected owner so they are not reprocessed after scan restarts.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
[darrick: fix the xfs_trans_roll call]
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/libxfs/xfs_btree.c | 26 ++++++++++++++-------
 fs/xfs/xfs_bmap_util.c    | 57 ++++++++++++++++++++++++++++++++++++++---------
 2 files changed, 65 insertions(+), 18 deletions(-)

diff --git a/fs/xfs/libxfs/xfs_btree.c b/fs/xfs/libxfs/xfs_btree.c
index 1d15d04e29a5..5bfb88261c7e 100644
--- a/fs/xfs/libxfs/xfs_btree.c
+++ b/fs/xfs/libxfs/xfs_btree.c
@@ -4452,10 +4452,15 @@ xfs_btree_block_change_owner(
 
 	/* modify the owner */
 	block = xfs_btree_get_block(cur, level, &bp);
-	if (cur->bc_flags & XFS_BTREE_LONG_PTRS)
+	if (cur->bc_flags & XFS_BTREE_LONG_PTRS) {
+		if (block->bb_u.l.bb_owner == cpu_to_be64(bbcoi->new_owner))
+			return 0;
 		block->bb_u.l.bb_owner = cpu_to_be64(bbcoi->new_owner);
-	else
+	} else {
+		if (block->bb_u.s.bb_owner == cpu_to_be32(bbcoi->new_owner))
+			return 0;
 		block->bb_u.s.bb_owner = cpu_to_be32(bbcoi->new_owner);
+	}
 
 	/*
 	 * If the block is a root block hosted in an inode, we might not have a
@@ -4464,14 +4469,19 @@ xfs_btree_block_change_owner(
 	 * block is formatted into the on-disk inode fork. We still change it,
 	 * though, so everything is consistent in memory.
 	 */
-	if (bp) {
-		if (cur->bc_tp)
-			xfs_trans_ordered_buf(cur->bc_tp, bp);
-		else
-			xfs_buf_delwri_queue(bp, bbcoi->buffer_list);
-	} else {
+	if (!bp) {
 		ASSERT(cur->bc_flags & XFS_BTREE_ROOT_IN_INODE);
 		ASSERT(level == cur->bc_nlevels - 1);
+		return 0;
+	}
+
+	if (cur->bc_tp) {
+		if (!xfs_trans_ordered_buf(cur->bc_tp, bp)) {
+			xfs_btree_log_block(cur, bp, XFS_BB_OWNER);
+			return -EAGAIN;
+		}
+	} else {
+		xfs_buf_delwri_queue(bp, bbcoi->buffer_list);
 	}
 
 	return 0;
diff --git a/fs/xfs/xfs_bmap_util.c b/fs/xfs/xfs_bmap_util.c
index ee8fb9a9ed73..3e9b7a4fb8fd 100644
--- a/fs/xfs/xfs_bmap_util.c
+++ b/fs/xfs/xfs_bmap_util.c
@@ -1929,6 +1929,48 @@ xfs_swap_extent_forks(
 	return 0;
 }
 
+/*
+ * Fix up the owners of the bmbt blocks to refer to the current inode. The
+ * change owner scan attempts to order all modified buffers in the current
+ * transaction. In the event of ordered buffer failure, the offending buffer is
+ * physically logged as a fallback and the scan returns -EAGAIN. We must roll
+ * the transaction in this case to replenish the fallback log reservation and
+ * restart the scan. This process repeats until the scan completes.
+ */
+static int
+xfs_swap_change_owner(
+	struct xfs_trans	**tpp,
+	struct xfs_inode	*ip,
+	struct xfs_inode	*tmpip)
+{
+	int			error;
+	struct xfs_trans	*tp = *tpp;
+
+	do {
+		error = xfs_bmbt_change_owner(tp, ip, XFS_DATA_FORK, ip->i_ino,
+					      NULL);
+		/* success or fatal error */
+		if (error != -EAGAIN)
+			break;
+
+		error = xfs_trans_roll(tpp, NULL);
+		if (error)
+			break;
+		tp = *tpp;
+
+		/*
+		 * Redirty both inodes so they can relog and keep the log tail
+		 * moving forward.
+		 */
+		xfs_trans_ijoin(tp, ip, 0);
+		xfs_trans_ijoin(tp, tmpip, 0);
+		xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
+		xfs_trans_log_inode(tp, tmpip, XFS_ILOG_CORE);
+	} while (true);
+
+	return error;
+}
+
 int
 xfs_swap_extents(
 	struct xfs_inode	*ip,	/* target inode */
@@ -1943,7 +1985,7 @@ xfs_swap_extents(
 	int			lock_flags;
 	struct xfs_ifork	*cowfp;
 	uint64_t		f;
-	int			resblks;
+	int			resblks = 0;
 
 	/*
 	 * Lock the inodes against other IO, page faults and truncate to
@@ -1991,11 +2033,8 @@ xfs_swap_extents(
 			  XFS_SWAP_RMAP_SPACE_RES(mp,
 				XFS_IFORK_NEXTENTS(tip, XFS_DATA_FORK),
 				XFS_DATA_FORK);
-		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_write, resblks,
-				0, 0, &tp);
-	} else
-		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_ichange, 0,
-				0, 0, &tp);
+	}
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_write, resblks, 0, 0, &tp);
 	if (error)
 		goto out_unlock;
 
@@ -2087,14 +2126,12 @@ xfs_swap_extents(
 	 * inode number of the current inode.
 	 */
 	if (src_log_flags & XFS_ILOG_DOWNER) {
-		error = xfs_bmbt_change_owner(tp, ip, XFS_DATA_FORK,
-					      ip->i_ino, NULL);
+		error = xfs_swap_change_owner(&tp, ip, tip);
 		if (error)
 			goto out_trans_cancel;
 	}
 	if (target_log_flags & XFS_ILOG_DOWNER) {
-		error = xfs_bmbt_change_owner(tp, tip, XFS_DATA_FORK,
-					      tip->i_ino, NULL);
+		error = xfs_swap_change_owner(&tp, tip, ip);
 		if (error)
 			goto out_trans_cancel;
 	}
-- 
2.14.1

Patch "xfs: relog dirty buffers during swapext bmbt owner change" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: relog dirty buffers during swapext bmbt owner change

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:26 -0700
Subject: xfs: relog dirty buffers during swapext bmbt owner change
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, Brian Foster <bfoster@redhat.com>, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-21-hch@lst.de>

From: Brian Foster <bfoster@redhat.com>

commit 2dd3d709fc4338681a3aa61658122fa8faa5a437 upstream.

The owner change bmbt scan that occurs during extent swap operations
does not handle ordered buffer failures. Buffers that cannot be
marked ordered must be physically logged so previously dirty ranges
of the buffer can be relogged in the transaction.

Since the bmbt scan may need to process and potentially log a large
number of blocks, we can't expect to complete this operation in a
single transaction. Update extent swap to use a permanent
transaction with enough log reservation to physically log a buffer.
Update the bmbt scan to physically log any buffers that cannot be
ordered and to terminate the scan with -EAGAIN. On -EAGAIN, the
caller rolls the transaction and restarts the scan. Finally, update
the bmbt scan helper function to skip bmbt blocks that already match
the expected owner so they are not reprocessed after scan restarts.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
[darrick: fix the xfs_trans_roll call]
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/libxfs/xfs_btree.c |   26 ++++++++++++++------
 fs/xfs/xfs_bmap_util.c    |   57 +++++++++++++++++++++++++++++++++++++---------
 2 files changed, 65 insertions(+), 18 deletions(-)

--- a/fs/xfs/libxfs/xfs_btree.c
+++ b/fs/xfs/libxfs/xfs_btree.c
@@ -4452,10 +4452,15 @@ xfs_btree_block_change_owner(
 
 	/* modify the owner */
 	block = xfs_btree_get_block(cur, level, &bp);
-	if (cur->bc_flags & XFS_BTREE_LONG_PTRS)
+	if (cur->bc_flags & XFS_BTREE_LONG_PTRS) {
+		if (block->bb_u.l.bb_owner == cpu_to_be64(bbcoi->new_owner))
+			return 0;
 		block->bb_u.l.bb_owner = cpu_to_be64(bbcoi->new_owner);
-	else
+	} else {
+		if (block->bb_u.s.bb_owner == cpu_to_be32(bbcoi->new_owner))
+			return 0;
 		block->bb_u.s.bb_owner = cpu_to_be32(bbcoi->new_owner);
+	}
 
 	/*
 	 * If the block is a root block hosted in an inode, we might not have a
@@ -4464,14 +4469,19 @@ xfs_btree_block_change_owner(
 	 * block is formatted into the on-disk inode fork. We still change it,
 	 * though, so everything is consistent in memory.
 	 */
-	if (bp) {
-		if (cur->bc_tp)
-			xfs_trans_ordered_buf(cur->bc_tp, bp);
-		else
-			xfs_buf_delwri_queue(bp, bbcoi->buffer_list);
-	} else {
+	if (!bp) {
 		ASSERT(cur->bc_flags & XFS_BTREE_ROOT_IN_INODE);
 		ASSERT(level == cur->bc_nlevels - 1);
+		return 0;
+	}
+
+	if (cur->bc_tp) {
+		if (!xfs_trans_ordered_buf(cur->bc_tp, bp)) {
+			xfs_btree_log_block(cur, bp, XFS_BB_OWNER);
+			return -EAGAIN;
+		}
+	} else {
+		xfs_buf_delwri_queue(bp, bbcoi->buffer_list);
 	}
 
 	return 0;
--- a/fs/xfs/xfs_bmap_util.c
+++ b/fs/xfs/xfs_bmap_util.c
@@ -1929,6 +1929,48 @@ xfs_swap_extent_forks(
 	return 0;
 }
 
+/*
+ * Fix up the owners of the bmbt blocks to refer to the current inode. The
+ * change owner scan attempts to order all modified buffers in the current
+ * transaction. In the event of ordered buffer failure, the offending buffer is
+ * physically logged as a fallback and the scan returns -EAGAIN. We must roll
+ * the transaction in this case to replenish the fallback log reservation and
+ * restart the scan. This process repeats until the scan completes.
+ */
+static int
+xfs_swap_change_owner(
+	struct xfs_trans	**tpp,
+	struct xfs_inode	*ip,
+	struct xfs_inode	*tmpip)
+{
+	int			error;
+	struct xfs_trans	*tp = *tpp;
+
+	do {
+		error = xfs_bmbt_change_owner(tp, ip, XFS_DATA_FORK, ip->i_ino,
+					      NULL);
+		/* success or fatal error */
+		if (error != -EAGAIN)
+			break;
+
+		error = xfs_trans_roll(tpp, NULL);
+		if (error)
+			break;
+		tp = *tpp;
+
+		/*
+		 * Redirty both inodes so they can relog and keep the log tail
+		 * moving forward.
+		 */
+		xfs_trans_ijoin(tp, ip, 0);
+		xfs_trans_ijoin(tp, tmpip, 0);
+		xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
+		xfs_trans_log_inode(tp, tmpip, XFS_ILOG_CORE);
+	} while (true);
+
+	return error;
+}
+
 int
 xfs_swap_extents(
 	struct xfs_inode	*ip,	/* target inode */
@@ -1943,7 +1985,7 @@ xfs_swap_extents(
 	int			lock_flags;
 	struct xfs_ifork	*cowfp;
 	uint64_t		f;
-	int			resblks;
+	int			resblks = 0;
 
 	/*
 	 * Lock the inodes against other IO, page faults and truncate to
@@ -1991,11 +2033,8 @@ xfs_swap_extents(
 			  XFS_SWAP_RMAP_SPACE_RES(mp,
 				XFS_IFORK_NEXTENTS(tip, XFS_DATA_FORK),
 				XFS_DATA_FORK);
-		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_write, resblks,
-				0, 0, &tp);
-	} else
-		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_ichange, 0,
-				0, 0, &tp);
+	}
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_write, resblks, 0, 0, &tp);
 	if (error)
 		goto out_unlock;
 
@@ -2087,14 +2126,12 @@ xfs_swap_extents(
 	 * inode number of the current inode.
 	 */
 	if (src_log_flags & XFS_ILOG_DOWNER) {
-		error = xfs_bmbt_change_owner(tp, ip, XFS_DATA_FORK,
-					      ip->i_ino, NULL);
+		error = xfs_swap_change_owner(&tp, ip, tip);
 		if (error)
 			goto out_trans_cancel;
 	}
 	if (target_log_flags & XFS_ILOG_DOWNER) {
-		error = xfs_bmbt_change_owner(tp, tip, XFS_DATA_FORK,
-					      tip->i_ino, NULL);
+		error = xfs_swap_change_owner(&tp, tip, ip);
 		if (error)
 			goto out_trans_cancel;
 	}


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 21/25] xfs: disable per-inode DAX flag

[Christoph Hellwig]

commit 742d84290739ae908f1b61b7d17ea382c8c0073a upstream.

Currently flag switching can be used to easily crash the kernel.  Disable
the per-inode DAX flag until that is sorted out.

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/xfs_ioctl.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/fs/xfs/xfs_ioctl.c b/fs/xfs/xfs_ioctl.c
index 9c0c7a920304..06ca244c323a 100644
--- a/fs/xfs/xfs_ioctl.c
+++ b/fs/xfs/xfs_ioctl.c
@@ -1008,11 +1008,12 @@ xfs_diflags_to_linux(
 		inode->i_flags |= S_NOATIME;
 	else
 		inode->i_flags &= ~S_NOATIME;
+#if 0	/* disabled until the flag switching races are sorted out */
 	if (xflags & FS_XFLAG_DAX)
 		inode->i_flags |= S_DAX;
 	else
 		inode->i_flags &= ~S_DAX;
-
+#endif
 }
 
 static int
-- 
2.14.1

Patch "xfs: disable per-inode DAX flag" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: disable per-inode DAX flag

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-disable-per-inode-dax-flag.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:27 -0700
Subject: xfs: disable per-inode DAX flag
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-22-hch@lst.de>

From: Christoph Hellwig <hch@lst.de>

commit 742d84290739ae908f1b61b7d17ea382c8c0073a upstream.

Currently flag switching can be used to easily crash the kernel.  Disable
the per-inode DAX flag until that is sorted out.

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>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/xfs_ioctl.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

--- a/fs/xfs/xfs_ioctl.c
+++ b/fs/xfs/xfs_ioctl.c
@@ -1008,11 +1008,12 @@ xfs_diflags_to_linux(
 		inode->i_flags |= S_NOATIME;
 	else
 		inode->i_flags &= ~S_NOATIME;
+#if 0	/* disabled until the flag switching races are sorted out */
 	if (xflags & FS_XFLAG_DAX)
 		inode->i_flags |= S_DAX;
 	else
 		inode->i_flags &= ~S_DAX;
-
+#endif
 }
 
 static int


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 22/25] xfs: fix incorrect log_flushed on fsync

[Christoph Hellwig]

From: Amir Goldstein <amir73il@gmail.com>

commit 47c7d0b19502583120c3f396c7559e7a77288a68 upstream.

When calling into _xfs_log_force{,_lsn}() with a pointer
to log_flushed variable, log_flushed will be set to 1 if:
1. xlog_sync() is called to flush the active log buffer
AND/OR
2. xlog_wait() is called to wait on a syncing log buffers

xfs_file_fsync() checks the value of log_flushed after
_xfs_log_force_lsn() call to optimize away an explicit
PREFLUSH request to the data block device after writing
out all the file's pages to disk.

This optimization is incorrect in the following sequence of events:

 Task A                    Task B
 -------------------------------------------------------
 xfs_file_fsync()
   _xfs_log_force_lsn()
     xlog_sync()
        [submit PREFLUSH]
                           xfs_file_fsync()
                             file_write_and_wait_range()
                               [submit WRITE X]
                               [endio  WRITE X]
                             _xfs_log_force_lsn()
                               xlog_wait()
        [endio  PREFLUSH]

The write X is not guarantied to be on persistent storage
when PREFLUSH request in completed, because write A was submitted
after the PREFLUSH request, but xfs_file_fsync() of task A will
be notified of log_flushed=1 and will skip explicit flush.

If the system crashes after fsync of task A, write X may not be
present on disk after reboot.

This bug was discovered and demonstrated using Josef Bacik's
dm-log-writes target, which can be used to record block io operations
and then replay a subset of these operations onto the target device.
The test goes something like this:
- Use fsx to execute ops of a file and record ops on log device
- Every now and then fsync the file, store md5 of file and mark
  the location in the log
- Then replay log onto device for each mark, mount fs and compare
  md5 of file to stored value

Cc: Christoph Hellwig <hch@lst.de>
Cc: Josef Bacik <jbacik@fb.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/xfs_log.c | 7 -------
 1 file changed, 7 deletions(-)

diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c
index bcb2f860e508..c5107c7bc4bf 100644
--- a/fs/xfs/xfs_log.c
+++ b/fs/xfs/xfs_log.c
@@ -3375,8 +3375,6 @@ _xfs_log_force(
 		 */
 		if (iclog->ic_state & XLOG_STATE_IOERROR)
 			return -EIO;
-		if (log_flushed)
-			*log_flushed = 1;
 	} else {
 
 no_sleep:
@@ -3480,8 +3478,6 @@ _xfs_log_force_lsn(
 
 				xlog_wait(&iclog->ic_prev->ic_write_wait,
 							&log->l_icloglock);
-				if (log_flushed)
-					*log_flushed = 1;
 				already_slept = 1;
 				goto try_again;
 			}
@@ -3515,9 +3511,6 @@ _xfs_log_force_lsn(
 			 */
 			if (iclog->ic_state & XLOG_STATE_IOERROR)
 				return -EIO;
-
-			if (log_flushed)
-				*log_flushed = 1;
 		} else {		/* just return */
 			spin_unlock(&log->l_icloglock);
 		}
-- 
2.14.1

Patch "xfs: fix incorrect log_flushed on fsync" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: fix incorrect log_flushed on fsync

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-fix-incorrect-log_flushed-on-fsync.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:28 -0700
Subject: xfs: fix incorrect log_flushed on fsync
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, Amir Goldstein <amir73il@gmail.com>, Josef Bacik <jbacik@fb.com>, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-23-hch@lst.de>

From: Amir Goldstein <amir73il@gmail.com>

commit 47c7d0b19502583120c3f396c7559e7a77288a68 upstream.

When calling into _xfs_log_force{,_lsn}() with a pointer
to log_flushed variable, log_flushed will be set to 1 if:
1. xlog_sync() is called to flush the active log buffer
AND/OR
2. xlog_wait() is called to wait on a syncing log buffers

xfs_file_fsync() checks the value of log_flushed after
_xfs_log_force_lsn() call to optimize away an explicit
PREFLUSH request to the data block device after writing
out all the file's pages to disk.

This optimization is incorrect in the following sequence of events:

 Task A                    Task B
 -------------------------------------------------------
 xfs_file_fsync()
   _xfs_log_force_lsn()
     xlog_sync()
        [submit PREFLUSH]
                           xfs_file_fsync()
                             file_write_and_wait_range()
                               [submit WRITE X]
                               [endio  WRITE X]
                             _xfs_log_force_lsn()
                               xlog_wait()
        [endio  PREFLUSH]

The write X is not guarantied to be on persistent storage
when PREFLUSH request in completed, because write A was submitted
after the PREFLUSH request, but xfs_file_fsync() of task A will
be notified of log_flushed=1 and will skip explicit flush.

If the system crashes after fsync of task A, write X may not be
present on disk after reboot.

This bug was discovered and demonstrated using Josef Bacik's
dm-log-writes target, which can be used to record block io operations
and then replay a subset of these operations onto the target device.
The test goes something like this:
- Use fsx to execute ops of a file and record ops on log device
- Every now and then fsync the file, store md5 of file and mark
  the location in the log
- Then replay log onto device for each mark, mount fs and compare
  md5 of file to stored value

Cc: Christoph Hellwig <hch@lst.de>
Cc: Josef Bacik <jbacik@fb.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/xfs_log.c |    7 -------
 1 file changed, 7 deletions(-)

--- a/fs/xfs/xfs_log.c
+++ b/fs/xfs/xfs_log.c
@@ -3375,8 +3375,6 @@ maybe_sleep:
 		 */
 		if (iclog->ic_state & XLOG_STATE_IOERROR)
 			return -EIO;
-		if (log_flushed)
-			*log_flushed = 1;
 	} else {
 
 no_sleep:
@@ -3480,8 +3478,6 @@ try_again:
 
 				xlog_wait(&iclog->ic_prev->ic_write_wait,
 							&log->l_icloglock);
-				if (log_flushed)
-					*log_flushed = 1;
 				already_slept = 1;
 				goto try_again;
 			}
@@ -3515,9 +3511,6 @@ try_again:
 			 */
 			if (iclog->ic_state & XLOG_STATE_IOERROR)
 				return -EIO;
-
-			if (log_flushed)
-				*log_flushed = 1;
 		} else {		/* just return */
 			spin_unlock(&log->l_icloglock);
 		}


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 23/25] xfs: don't set v3 xflags for v2 inodes

[Christoph Hellwig]

commit dd60687ee541ca3f6df8758f38e6f22f57c42a37 upstream.

Reject attempts to set XFLAGS that correspond to di_flags2 inode flags
if the inode isn't a v3 inode, because di_flags2 only exists on v3.

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/xfs_ioctl.c | 38 +++++++++++++++++++++++++-------------
 1 file changed, 25 insertions(+), 13 deletions(-)

diff --git a/fs/xfs/xfs_ioctl.c b/fs/xfs/xfs_ioctl.c
index 06ca244c323a..5049e8ab6e30 100644
--- a/fs/xfs/xfs_ioctl.c
+++ b/fs/xfs/xfs_ioctl.c
@@ -931,16 +931,15 @@ xfs_ioc_fsgetxattr(
 	return 0;
 }
 
-STATIC void
-xfs_set_diflags(
+STATIC uint16_t
+xfs_flags2diflags(
 	struct xfs_inode	*ip,
 	unsigned int		xflags)
 {
-	unsigned int		di_flags;
-	uint64_t		di_flags2;
-
 	/* can't set PREALLOC this way, just preserve it */
-	di_flags = (ip->i_d.di_flags & XFS_DIFLAG_PREALLOC);
+	uint16_t		di_flags =
+		(ip->i_d.di_flags & XFS_DIFLAG_PREALLOC);
+
 	if (xflags & FS_XFLAG_IMMUTABLE)
 		di_flags |= XFS_DIFLAG_IMMUTABLE;
 	if (xflags & FS_XFLAG_APPEND)
@@ -970,19 +969,24 @@ xfs_set_diflags(
 		if (xflags & FS_XFLAG_EXTSIZE)
 			di_flags |= XFS_DIFLAG_EXTSIZE;
 	}
-	ip->i_d.di_flags = di_flags;
 
-	/* diflags2 only valid for v3 inodes. */
-	if (ip->i_d.di_version < 3)
-		return;
+	return di_flags;
+}
+
+STATIC uint64_t
+xfs_flags2diflags2(
+	struct xfs_inode	*ip,
+	unsigned int		xflags)
+{
+	uint64_t		di_flags2 =
+		(ip->i_d.di_flags2 & XFS_DIFLAG2_REFLINK);
 
-	di_flags2 = (ip->i_d.di_flags2 & XFS_DIFLAG2_REFLINK);
 	if (xflags & FS_XFLAG_DAX)
 		di_flags2 |= XFS_DIFLAG2_DAX;
 	if (xflags & FS_XFLAG_COWEXTSIZE)
 		di_flags2 |= XFS_DIFLAG2_COWEXTSIZE;
 
-	ip->i_d.di_flags2 = di_flags2;
+	return di_flags2;
 }
 
 STATIC void
@@ -1023,6 +1027,7 @@ xfs_ioctl_setattr_xflags(
 	struct fsxattr		*fa)
 {
 	struct xfs_mount	*mp = ip->i_mount;
+	uint64_t		di_flags2;
 
 	/* Can't change realtime flag if any extents are allocated. */
 	if ((ip->i_d.di_nextents || ip->i_delayed_blks) &&
@@ -1053,7 +1058,14 @@ xfs_ioctl_setattr_xflags(
 	    !capable(CAP_LINUX_IMMUTABLE))
 		return -EPERM;
 
-	xfs_set_diflags(ip, fa->fsx_xflags);
+	/* diflags2 only valid for v3 inodes. */
+	di_flags2 = xfs_flags2diflags2(ip, fa->fsx_xflags);
+	if (di_flags2 && ip->i_d.di_version < 3)
+		return -EINVAL;
+
+	ip->i_d.di_flags = xfs_flags2diflags(ip, fa->fsx_xflags);
+	ip->i_d.di_flags2 = di_flags2;
+
 	xfs_diflags_to_linux(ip);
 	xfs_trans_ichgtime(tp, ip, XFS_ICHGTIME_CHG);
 	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
-- 
2.14.1

Patch "xfs: don't set v3 xflags for v2 inodes" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: don't set v3 xflags for v2 inodes

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-don-t-set-v3-xflags-for-v2-inodes.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:29 -0700
Subject: xfs: don't set v3 xflags for v2 inodes
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-24-hch@lst.de>

From: Christoph Hellwig <hch@lst.de>

commit dd60687ee541ca3f6df8758f38e6f22f57c42a37 upstream.

Reject attempts to set XFLAGS that correspond to di_flags2 inode flags
if the inode isn't a v3 inode, because di_flags2 only exists on v3.

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>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/xfs_ioctl.c |   38 +++++++++++++++++++++++++-------------
 1 file changed, 25 insertions(+), 13 deletions(-)

--- a/fs/xfs/xfs_ioctl.c
+++ b/fs/xfs/xfs_ioctl.c
@@ -931,16 +931,15 @@ xfs_ioc_fsgetxattr(
 	return 0;
 }
 
-STATIC void
-xfs_set_diflags(
+STATIC uint16_t
+xfs_flags2diflags(
 	struct xfs_inode	*ip,
 	unsigned int		xflags)
 {
-	unsigned int		di_flags;
-	uint64_t		di_flags2;
-
 	/* can't set PREALLOC this way, just preserve it */
-	di_flags = (ip->i_d.di_flags & XFS_DIFLAG_PREALLOC);
+	uint16_t		di_flags =
+		(ip->i_d.di_flags & XFS_DIFLAG_PREALLOC);
+
 	if (xflags & FS_XFLAG_IMMUTABLE)
 		di_flags |= XFS_DIFLAG_IMMUTABLE;
 	if (xflags & FS_XFLAG_APPEND)
@@ -970,19 +969,24 @@ xfs_set_diflags(
 		if (xflags & FS_XFLAG_EXTSIZE)
 			di_flags |= XFS_DIFLAG_EXTSIZE;
 	}
-	ip->i_d.di_flags = di_flags;
 
-	/* diflags2 only valid for v3 inodes. */
-	if (ip->i_d.di_version < 3)
-		return;
+	return di_flags;
+}
+
+STATIC uint64_t
+xfs_flags2diflags2(
+	struct xfs_inode	*ip,
+	unsigned int		xflags)
+{
+	uint64_t		di_flags2 =
+		(ip->i_d.di_flags2 & XFS_DIFLAG2_REFLINK);
 
-	di_flags2 = (ip->i_d.di_flags2 & XFS_DIFLAG2_REFLINK);
 	if (xflags & FS_XFLAG_DAX)
 		di_flags2 |= XFS_DIFLAG2_DAX;
 	if (xflags & FS_XFLAG_COWEXTSIZE)
 		di_flags2 |= XFS_DIFLAG2_COWEXTSIZE;
 
-	ip->i_d.di_flags2 = di_flags2;
+	return di_flags2;
 }
 
 STATIC void
@@ -1023,6 +1027,7 @@ xfs_ioctl_setattr_xflags(
 	struct fsxattr		*fa)
 {
 	struct xfs_mount	*mp = ip->i_mount;
+	uint64_t		di_flags2;
 
 	/* Can't change realtime flag if any extents are allocated. */
 	if ((ip->i_d.di_nextents || ip->i_delayed_blks) &&
@@ -1053,7 +1058,14 @@ xfs_ioctl_setattr_xflags(
 	    !capable(CAP_LINUX_IMMUTABLE))
 		return -EPERM;
 
-	xfs_set_diflags(ip, fa->fsx_xflags);
+	/* diflags2 only valid for v3 inodes. */
+	di_flags2 = xfs_flags2diflags2(ip, fa->fsx_xflags);
+	if (di_flags2 && ip->i_d.di_version < 3)
+		return -EINVAL;
+
+	ip->i_d.di_flags = xfs_flags2diflags(ip, fa->fsx_xflags);
+	ip->i_d.di_flags2 = di_flags2;
+
 	xfs_diflags_to_linux(ip);
 	xfs_trans_ichgtime(tp, ip, XFS_ICHGTIME_CHG);
 	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 24/25] xfs: open code end_buffer_async_write in xfs_finish_page_writeback

[Christoph Hellwig]

commit 8353a814f2518dcfa79a5bb77afd0e7dfa391bb1 upstream.

Our loop in xfs_finish_page_writeback, which iterates over all buffer
heads in a page and then calls end_buffer_async_write, which also
iterates over all buffers in the page to check if any I/O is in flight
is not only inefficient, but also potentially dangerous as
end_buffer_async_write can cause the page and all buffers to be freed.

Replace it with a single loop that does the work of end_buffer_async_write
on a per-page basis.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/xfs_aops.c | 71 ++++++++++++++++++++++++++++++++++++-------------------
 1 file changed, 47 insertions(+), 24 deletions(-)

diff --git a/fs/xfs/xfs_aops.c b/fs/xfs/xfs_aops.c
index 6bf120bb1a17..f9efd67f6fa1 100644
--- a/fs/xfs/xfs_aops.c
+++ b/fs/xfs/xfs_aops.c
@@ -85,11 +85,11 @@ xfs_find_bdev_for_inode(
  * associated buffer_heads, paying attention to the start and end offsets that
  * we need to process on the page.
  *
- * Landmine Warning: bh->b_end_io() will call end_page_writeback() on the last
- * buffer in the IO. Once it does this, it is unsafe to access the bufferhead or
- * the page at all, as we may be racing with memory reclaim and it can free both
- * the bufferhead chain and the page as it will see the page as clean and
- * unused.
+ * Note that we open code the action in end_buffer_async_write here so that we
+ * only have to iterate over the buffers attached to the page once.  This is not
+ * only more efficient, but also ensures that we only calls end_page_writeback
+ * at the end of the iteration, and thus avoids the pitfall of having the page
+ * and buffers potentially freed after every call to end_buffer_async_write.
  */
 static void
 xfs_finish_page_writeback(
@@ -97,29 +97,44 @@ xfs_finish_page_writeback(
 	struct bio_vec		*bvec,
 	int			error)
 {
-	unsigned int		end = bvec->bv_offset + bvec->bv_len - 1;
-	struct buffer_head	*head, *bh, *next;
+	struct buffer_head	*head = page_buffers(bvec->bv_page), *bh = head;
+	bool			busy = false;
 	unsigned int		off = 0;
-	unsigned int		bsize;
+	unsigned long		flags;
 
 	ASSERT(bvec->bv_offset < PAGE_SIZE);
 	ASSERT((bvec->bv_offset & (i_blocksize(inode) - 1)) == 0);
-	ASSERT(end < PAGE_SIZE);
+	ASSERT(bvec->bv_offset + bvec->bv_len <= PAGE_SIZE);
 	ASSERT((bvec->bv_len & (i_blocksize(inode) - 1)) == 0);
 
-	bh = head = page_buffers(bvec->bv_page);
-
-	bsize = bh->b_size;
+	local_irq_save(flags);
+	bit_spin_lock(BH_Uptodate_Lock, &head->b_state);
 	do {
-		if (off > end)
-			break;
-		next = bh->b_this_page;
-		if (off < bvec->bv_offset)
-			goto next_bh;
-		bh->b_end_io(bh, !error);
-next_bh:
-		off += bsize;
-	} while ((bh = next) != head);
+		if (off >= bvec->bv_offset &&
+		    off < bvec->bv_offset + bvec->bv_len) {
+			ASSERT(buffer_async_write(bh));
+			ASSERT(bh->b_end_io == NULL);
+
+			if (error) {
+				mark_buffer_write_io_error(bh);
+				clear_buffer_uptodate(bh);
+				SetPageError(bvec->bv_page);
+			} else {
+				set_buffer_uptodate(bh);
+			}
+			clear_buffer_async_write(bh);
+			unlock_buffer(bh);
+		} else if (buffer_async_write(bh)) {
+			ASSERT(buffer_locked(bh));
+			busy = true;
+		}
+		off += bh->b_size;
+	} while ((bh = bh->b_this_page) != head);
+	bit_spin_unlock(BH_Uptodate_Lock, &head->b_state);
+	local_irq_restore(flags);
+
+	if (!busy)
+		end_page_writeback(bvec->bv_page);
 }
 
 /*
@@ -133,8 +148,10 @@ xfs_destroy_ioend(
 	int			error)
 {
 	struct inode		*inode = ioend->io_inode;
-	struct bio		*last = ioend->io_bio;
-	struct bio		*bio, *next;
+	struct bio		*bio = &ioend->io_inline_bio;
+	struct bio		*last = ioend->io_bio, *next;
+	u64			start = bio->bi_iter.bi_sector;
+	bool			quiet = bio_flagged(bio, BIO_QUIET);
 
 	for (bio = &ioend->io_inline_bio; bio; bio = next) {
 		struct bio_vec	*bvec;
@@ -155,6 +172,11 @@ xfs_destroy_ioend(
 
 		bio_put(bio);
 	}
+
+	if (unlikely(error && !quiet)) {
+		xfs_err_ratelimited(XFS_I(inode)->i_mount,
+			"writeback error on sector %llu", start);
+	}
 }
 
 /*
@@ -423,7 +445,8 @@ xfs_start_buffer_writeback(
 	ASSERT(!buffer_delay(bh));
 	ASSERT(!buffer_unwritten(bh));
 
-	mark_buffer_async_write(bh);
+	bh->b_end_io = NULL;
+	set_buffer_async_write(bh);
 	set_buffer_uptodate(bh);
 	clear_buffer_dirty(bh);
 }
-- 
2.14.1

Patch "xfs: open code end_buffer_async_write in xfs_finish_page_writeback" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: open code end_buffer_async_write in xfs_finish_page_writeback

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:30 -0700
Subject: xfs: open code end_buffer_async_write in xfs_finish_page_writeback
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-25-hch@lst.de>

From: Christoph Hellwig <hch@lst.de>

commit 8353a814f2518dcfa79a5bb77afd0e7dfa391bb1 upstream.

Our loop in xfs_finish_page_writeback, which iterates over all buffer
heads in a page and then calls end_buffer_async_write, which also
iterates over all buffers in the page to check if any I/O is in flight
is not only inefficient, but also potentially dangerous as
end_buffer_async_write can cause the page and all buffers to be freed.

Replace it with a single loop that does the work of end_buffer_async_write
on a per-page basis.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/xfs_aops.c |   71 +++++++++++++++++++++++++++++++++++-------------------
 1 file changed, 47 insertions(+), 24 deletions(-)

--- a/fs/xfs/xfs_aops.c
+++ b/fs/xfs/xfs_aops.c
@@ -85,11 +85,11 @@ xfs_find_bdev_for_inode(
  * associated buffer_heads, paying attention to the start and end offsets that
  * we need to process on the page.
  *
- * Landmine Warning: bh->b_end_io() will call end_page_writeback() on the last
- * buffer in the IO. Once it does this, it is unsafe to access the bufferhead or
- * the page at all, as we may be racing with memory reclaim and it can free both
- * the bufferhead chain and the page as it will see the page as clean and
- * unused.
+ * Note that we open code the action in end_buffer_async_write here so that we
+ * only have to iterate over the buffers attached to the page once.  This is not
+ * only more efficient, but also ensures that we only calls end_page_writeback
+ * at the end of the iteration, and thus avoids the pitfall of having the page
+ * and buffers potentially freed after every call to end_buffer_async_write.
  */
 static void
 xfs_finish_page_writeback(
@@ -97,29 +97,44 @@ xfs_finish_page_writeback(
 	struct bio_vec		*bvec,
 	int			error)
 {
-	unsigned int		end = bvec->bv_offset + bvec->bv_len - 1;
-	struct buffer_head	*head, *bh, *next;
+	struct buffer_head	*head = page_buffers(bvec->bv_page), *bh = head;
+	bool			busy = false;
 	unsigned int		off = 0;
-	unsigned int		bsize;
+	unsigned long		flags;
 
 	ASSERT(bvec->bv_offset < PAGE_SIZE);
 	ASSERT((bvec->bv_offset & (i_blocksize(inode) - 1)) == 0);
-	ASSERT(end < PAGE_SIZE);
+	ASSERT(bvec->bv_offset + bvec->bv_len <= PAGE_SIZE);
 	ASSERT((bvec->bv_len & (i_blocksize(inode) - 1)) == 0);
 
-	bh = head = page_buffers(bvec->bv_page);
-
-	bsize = bh->b_size;
+	local_irq_save(flags);
+	bit_spin_lock(BH_Uptodate_Lock, &head->b_state);
 	do {
-		if (off > end)
-			break;
-		next = bh->b_this_page;
-		if (off < bvec->bv_offset)
-			goto next_bh;
-		bh->b_end_io(bh, !error);
-next_bh:
-		off += bsize;
-	} while ((bh = next) != head);
+		if (off >= bvec->bv_offset &&
+		    off < bvec->bv_offset + bvec->bv_len) {
+			ASSERT(buffer_async_write(bh));
+			ASSERT(bh->b_end_io == NULL);
+
+			if (error) {
+				mark_buffer_write_io_error(bh);
+				clear_buffer_uptodate(bh);
+				SetPageError(bvec->bv_page);
+			} else {
+				set_buffer_uptodate(bh);
+			}
+			clear_buffer_async_write(bh);
+			unlock_buffer(bh);
+		} else if (buffer_async_write(bh)) {
+			ASSERT(buffer_locked(bh));
+			busy = true;
+		}
+		off += bh->b_size;
+	} while ((bh = bh->b_this_page) != head);
+	bit_spin_unlock(BH_Uptodate_Lock, &head->b_state);
+	local_irq_restore(flags);
+
+	if (!busy)
+		end_page_writeback(bvec->bv_page);
 }
 
 /*
@@ -133,8 +148,10 @@ xfs_destroy_ioend(
 	int			error)
 {
 	struct inode		*inode = ioend->io_inode;
-	struct bio		*last = ioend->io_bio;
-	struct bio		*bio, *next;
+	struct bio		*bio = &ioend->io_inline_bio;
+	struct bio		*last = ioend->io_bio, *next;
+	u64			start = bio->bi_iter.bi_sector;
+	bool			quiet = bio_flagged(bio, BIO_QUIET);
 
 	for (bio = &ioend->io_inline_bio; bio; bio = next) {
 		struct bio_vec	*bvec;
@@ -155,6 +172,11 @@ xfs_destroy_ioend(
 
 		bio_put(bio);
 	}
+
+	if (unlikely(error && !quiet)) {
+		xfs_err_ratelimited(XFS_I(inode)->i_mount,
+			"writeback error on sector %llu", start);
+	}
 }
 
 /*
@@ -423,7 +445,8 @@ xfs_start_buffer_writeback(
 	ASSERT(!buffer_delay(bh));
 	ASSERT(!buffer_unwritten(bh));
 
-	mark_buffer_async_write(bh);
+	bh->b_end_io = NULL;
+	set_buffer_async_write(bh);
 	set_buffer_uptodate(bh);
 	clear_buffer_dirty(bh);
 }


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

[PATCH 25/25] xfs: use kmem_free to free return value of kmem_zalloc

[Christoph Hellwig]

From: Pan Bian <bianpan2016@163.com>

commit 6c370590cfe0c36bcd62d548148aa65c984540b7 upstream.

In function xfs_test_remount_options(), kfree() is used to free memory
allocated by kmem_zalloc(). But it is better to use kmem_free().

Signed-off-by: Pan Bian <bianpan2016@163.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/xfs_super.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
index 38aaacdbb8b3..c1c4c2ea1014 100644
--- a/fs/xfs/xfs_super.c
+++ b/fs/xfs/xfs_super.c
@@ -1220,7 +1220,7 @@ xfs_test_remount_options(
 	tmp_mp->m_super = sb;
 	error = xfs_parseargs(tmp_mp, options);
 	xfs_free_fsname(tmp_mp);
-	kfree(tmp_mp);
+	kmem_free(tmp_mp);
 
 	return error;
 }
-- 
2.14.1

Patch "xfs: use kmem_free to free return value of kmem_zalloc" has been added to the 4.13-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: use kmem_free to free return value of kmem_zalloc

to the 4.13-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
and it can be found in the queue-4.13 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From foo@baz Mon Sep 18 10:25:08 CEST 2017
From: Christoph Hellwig <hch@lst.de>
Date: Sun, 17 Sep 2017 14:06:31 -0700
Subject: xfs: use kmem_free to free return value of kmem_zalloc
To: stable@vger.kernel.org
Cc: linux-xfs@vger.kernel.org, Pan Bian <bianpan2016@163.com>, "Darrick J . Wong" <darrick.wong@oracle.com>
Message-ID: <20170917210631.10725-26-hch@lst.de>

From: Pan Bian <bianpan2016@163.com>

commit 6c370590cfe0c36bcd62d548148aa65c984540b7 upstream.

In function xfs_test_remount_options(), kfree() is used to free memory
allocated by kmem_zalloc(). But it is better to use kmem_free().

Signed-off-by: Pan Bian <bianpan2016@163.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/xfs/xfs_super.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

--- a/fs/xfs/xfs_super.c
+++ b/fs/xfs/xfs_super.c
@@ -1220,7 +1220,7 @@ xfs_test_remount_options(
 	tmp_mp->m_super = sb;
 	error = xfs_parseargs(tmp_mp, options);
 	xfs_free_fsname(tmp_mp);
-	kfree(tmp_mp);
+	kmem_free(tmp_mp);
 
 	return error;
 }


Patches currently in stable-queue which might be from hch@lst.de are

queue-4.13/xfs-open-code-xfs_buf_item_dirty.patch
queue-4.13/xfs-properly-retry-failed-inode-items-in-case-of-error-during-buffer-writeback.patch
queue-4.13/xfs-use-kmem_free-to-free-return-value-of-kmem_zalloc.patch
queue-4.13/xfs-add-infrastructure-needed-for-error-propagation-during-buffer-io-failure.patch
queue-4.13/xfs-don-t-set-v3-xflags-for-v2-inodes.patch
queue-4.13/xfs-toggle-readonly-state-around-xfs_log_mount_finish.patch
queue-4.13/xfs-fix-log-recovery-corruption-error-due-to-tail-overwrite.patch
queue-4.13/xfs-move-bmbt-owner-change-to-last-step-of-extent-swap.patch
queue-4.13/xfs-check-for-race-with-xfs_reclaim_inode-in-xfs_ifree_cluster.patch
queue-4.13/xfs-always-verify-the-log-tail-during-recovery.patch
queue-4.13/xfs-open-code-end_buffer_async_write-in-xfs_finish_page_writeback.patch
queue-4.13/xfs-relog-dirty-buffers-during-swapext-bmbt-owner-change.patch
queue-4.13/xfs-disable-per-inode-dax-flag.patch
queue-4.13/xfs-refactor-buffer-logging-into-buffer-dirtying-helper.patch
queue-4.13/xfs-fix-recovery-failure-when-log-record-header-wraps-log-end.patch
queue-4.13/xfs-skip-bmbt-block-ino-validation-during-owner-change.patch
queue-4.13/xfs-don-t-log-dirty-ranges-for-ordered-buffers.patch
queue-4.13/xfs-stop-searching-for-free-slots-in-an-inode-chunk-when-there-are-none.patch
queue-4.13/xfs-fix-incorrect-log_flushed-on-fsync.patch
queue-4.13/xfs-evict-all-inodes-involved-with-log-redo-item.patch
queue-4.13/xfs-write-unmount-record-for-ro-mounts.patch
queue-4.13/xfs-remove-unnecessary-dirty-bli-format-check-for-ordered-bufs.patch
queue-4.13/xfs-disallow-marking-previously-dirty-buffers-as-ordered.patch
queue-4.13/xfs-handle-efscorrupted-during-head-tail-verification.patch
queue-4.13/xfs-ordered-buffer-log-items-are-never-formatted.patch

Re: 4.13-stable updates for XFS

[Greg KH]

On Sun, Sep 17, 2017 at 02:06:06PM -0700, Christoph Hellwig wrote:
> Hi all,
> 
> below are backports up to 4.14-rc1 for XFS.  I've selected all
> clear bugfixes for reported bugs, and a few harmless prep patches
> that allow us to carry late bug fixes as-is instead of providing
> special versions for -stable, which in the past have caused more
> harm than good.

Thanks for these, all now queued up.

And I agree with the "as-is" stuff, with the prep patches, that's the
best way to go.

thanks,

greg k-h