4.9-stable updates for XFS

4.9-stable updates for XFS

[Christoph Hellwig]

Hi all,

below are backports up to (almost) 4.12-rc4 for XFS.  I've selected all
clear bugfixes for reported bugs.  In addition to the patches in
4.11-stable there also are 3 patches that are in 4.11 but not in
4.9-stable yet at the beginning of the series.

[PATCH 01/25] xfs: verify inline directory data forks

[Christoph Hellwig]

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

When we're reading or writing the data fork of an inline directory,
check the contents to make sure we're not overflowing buffers or eating
garbage data.  xfs/348 corrupts an inline symlink into an inline
directory, triggering a buffer overflow bug.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
---
v2: add more checks consistent with _dir2_sf_check and make the verifier
usable from anywhere.
---
 fs/xfs/libxfs/xfs_dir2_priv.h  |  2 +
 fs/xfs/libxfs/xfs_dir2_sf.c    | 87 ++++++++++++++++++++++++++++++++++++++++++
 fs/xfs/libxfs/xfs_inode_fork.c | 26 +++++++++++--
 fs/xfs/libxfs/xfs_inode_fork.h |  2 +-
 fs/xfs/xfs_dir2_readdir.c      | 11 ------
 fs/xfs/xfs_inode.c             | 12 ++++--
 6 files changed, 122 insertions(+), 18 deletions(-)

diff --git a/fs/xfs/libxfs/xfs_dir2_priv.h b/fs/xfs/libxfs/xfs_dir2_priv.h
index ef9f6ead96a4..860be99407bf 100644
--- a/fs/xfs/libxfs/xfs_dir2_priv.h
+++ b/fs/xfs/libxfs/xfs_dir2_priv.h
@@ -126,6 +126,8 @@ extern int xfs_dir2_sf_create(struct xfs_da_args *args, xfs_ino_t pino);
 extern int xfs_dir2_sf_lookup(struct xfs_da_args *args);
 extern int xfs_dir2_sf_removename(struct xfs_da_args *args);
 extern int xfs_dir2_sf_replace(struct xfs_da_args *args);
+extern int xfs_dir2_sf_verify(struct xfs_mount *mp, struct xfs_dir2_sf_hdr *sfp,
+		int size);
 
 /* xfs_dir2_readdir.c */
 extern int xfs_readdir(struct xfs_inode *dp, struct dir_context *ctx,
diff --git a/fs/xfs/libxfs/xfs_dir2_sf.c b/fs/xfs/libxfs/xfs_dir2_sf.c
index c6809ff41197..96b45cd6c63f 100644
--- a/fs/xfs/libxfs/xfs_dir2_sf.c
+++ b/fs/xfs/libxfs/xfs_dir2_sf.c
@@ -629,6 +629,93 @@ xfs_dir2_sf_check(
 }
 #endif	/* DEBUG */
 
+/* Verify the consistency of an inline directory. */
+int
+xfs_dir2_sf_verify(
+	struct xfs_mount		*mp,
+	struct xfs_dir2_sf_hdr		*sfp,
+	int				size)
+{
+	struct xfs_dir2_sf_entry	*sfep;
+	struct xfs_dir2_sf_entry	*next_sfep;
+	char				*endp;
+	const struct xfs_dir_ops	*dops;
+	xfs_ino_t			ino;
+	int				i;
+	int				i8count;
+	int				offset;
+	__uint8_t			filetype;
+
+	dops = xfs_dir_get_ops(mp, NULL);
+
+	/*
+	 * Give up if the directory is way too short.
+	 */
+	XFS_WANT_CORRUPTED_RETURN(mp, size >
+			offsetof(struct xfs_dir2_sf_hdr, parent));
+	XFS_WANT_CORRUPTED_RETURN(mp, size >=
+			xfs_dir2_sf_hdr_size(sfp->i8count));
+
+	endp = (char *)sfp + size;
+
+	/* Check .. entry */
+	ino = dops->sf_get_parent_ino(sfp);
+	i8count = ino > XFS_DIR2_MAX_SHORT_INUM;
+	XFS_WANT_CORRUPTED_RETURN(mp, !xfs_dir_ino_validate(mp, ino));
+	offset = dops->data_first_offset;
+
+	/* Check all reported entries */
+	sfep = xfs_dir2_sf_firstentry(sfp);
+	for (i = 0; i < sfp->count; i++) {
+		/*
+		 * struct xfs_dir2_sf_entry has a variable length.
+		 * Check the fixed-offset parts of the structure are
+		 * within the data buffer.
+		 */
+		XFS_WANT_CORRUPTED_RETURN(mp,
+				((char *)sfep + sizeof(*sfep)) < endp);
+
+		/* Don't allow names with known bad length. */
+		XFS_WANT_CORRUPTED_RETURN(mp, sfep->namelen > 0);
+		XFS_WANT_CORRUPTED_RETURN(mp, sfep->namelen < MAXNAMELEN);
+
+		/*
+		 * Check that the variable-length part of the structure is
+		 * within the data buffer.  The next entry starts after the
+		 * name component, so nextentry is an acceptable test.
+		 */
+		next_sfep = dops->sf_nextentry(sfp, sfep);
+		XFS_WANT_CORRUPTED_RETURN(mp, endp >= (char *)next_sfep);
+
+		/* Check that the offsets always increase. */
+		XFS_WANT_CORRUPTED_RETURN(mp,
+				xfs_dir2_sf_get_offset(sfep) >= offset);
+
+		/* Check the inode number. */
+		ino = dops->sf_get_ino(sfp, sfep);
+		i8count += ino > XFS_DIR2_MAX_SHORT_INUM;
+		XFS_WANT_CORRUPTED_RETURN(mp, !xfs_dir_ino_validate(mp, ino));
+
+		/* Check the file type. */
+		filetype = dops->sf_get_ftype(sfep);
+		XFS_WANT_CORRUPTED_RETURN(mp, filetype < XFS_DIR3_FT_MAX);
+
+		offset = xfs_dir2_sf_get_offset(sfep) +
+				dops->data_entsize(sfep->namelen);
+
+		sfep = next_sfep;
+	}
+	XFS_WANT_CORRUPTED_RETURN(mp, i8count == sfp->i8count);
+	XFS_WANT_CORRUPTED_RETURN(mp, (void *)sfep == (void *)endp);
+
+	/* Make sure this whole thing ought to be in local format. */
+	XFS_WANT_CORRUPTED_RETURN(mp, offset +
+	       (sfp->count + 2) * (uint)sizeof(xfs_dir2_leaf_entry_t) +
+	       (uint)sizeof(xfs_dir2_block_tail_t) <= mp->m_dir_geo->blksize);
+
+	return 0;
+}
+
 /*
  * Create a new (shortform) directory.
  */
diff --git a/fs/xfs/libxfs/xfs_inode_fork.c b/fs/xfs/libxfs/xfs_inode_fork.c
index 25c1e078aef6..9653e964eda4 100644
--- a/fs/xfs/libxfs/xfs_inode_fork.c
+++ b/fs/xfs/libxfs/xfs_inode_fork.c
@@ -33,6 +33,8 @@
 #include "xfs_trace.h"
 #include "xfs_attr_sf.h"
 #include "xfs_da_format.h"
+#include "xfs_da_btree.h"
+#include "xfs_dir2_priv.h"
 
 kmem_zone_t *xfs_ifork_zone;
 
@@ -320,6 +322,7 @@ xfs_iformat_local(
 	int		whichfork,
 	int		size)
 {
+	int		error;
 
 	/*
 	 * If the size is unreasonable, then something
@@ -336,6 +339,14 @@ xfs_iformat_local(
 		return -EFSCORRUPTED;
 	}
 
+	if (S_ISDIR(VFS_I(ip)->i_mode) && whichfork == XFS_DATA_FORK) {
+		error = xfs_dir2_sf_verify(ip->i_mount,
+				(struct xfs_dir2_sf_hdr *)XFS_DFORK_DPTR(dip),
+				size);
+		if (error)
+			return error;
+	}
+
 	xfs_init_local_fork(ip, whichfork, XFS_DFORK_PTR(dip, whichfork), size);
 	return 0;
 }
@@ -856,7 +867,7 @@ xfs_iextents_copy(
  * In these cases, the format always takes precedence, because the
  * format indicates the current state of the fork.
  */
-void
+int
 xfs_iflush_fork(
 	xfs_inode_t		*ip,
 	xfs_dinode_t		*dip,
@@ -866,6 +877,7 @@ xfs_iflush_fork(
 	char			*cp;
 	xfs_ifork_t		*ifp;
 	xfs_mount_t		*mp;
+	int			error;
 	static const short	brootflag[2] =
 		{ XFS_ILOG_DBROOT, XFS_ILOG_ABROOT };
 	static const short	dataflag[2] =
@@ -874,7 +886,7 @@ xfs_iflush_fork(
 		{ XFS_ILOG_DEXT, XFS_ILOG_AEXT };
 
 	if (!iip)
-		return;
+		return 0;
 	ifp = XFS_IFORK_PTR(ip, whichfork);
 	/*
 	 * This can happen if we gave up in iformat in an error path,
@@ -882,12 +894,19 @@ xfs_iflush_fork(
 	 */
 	if (!ifp) {
 		ASSERT(whichfork == XFS_ATTR_FORK);
-		return;
+		return 0;
 	}
 	cp = XFS_DFORK_PTR(dip, whichfork);
 	mp = ip->i_mount;
 	switch (XFS_IFORK_FORMAT(ip, whichfork)) {
 	case XFS_DINODE_FMT_LOCAL:
+		if (S_ISDIR(VFS_I(ip)->i_mode) && whichfork == XFS_DATA_FORK) {
+			error = xfs_dir2_sf_verify(mp,
+					(struct xfs_dir2_sf_hdr *)ifp->if_u1.if_data,
+					ifp->if_bytes);
+			if (error)
+				return error;
+		}
 		if ((iip->ili_fields & dataflag[whichfork]) &&
 		    (ifp->if_bytes > 0)) {
 			ASSERT(ifp->if_u1.if_data != NULL);
@@ -940,6 +959,7 @@ xfs_iflush_fork(
 		ASSERT(0);
 		break;
 	}
+	return 0;
 }
 
 /*
diff --git a/fs/xfs/libxfs/xfs_inode_fork.h b/fs/xfs/libxfs/xfs_inode_fork.h
index 7fb8365326d1..132dc59fdde6 100644
--- a/fs/xfs/libxfs/xfs_inode_fork.h
+++ b/fs/xfs/libxfs/xfs_inode_fork.h
@@ -140,7 +140,7 @@ typedef struct xfs_ifork {
 struct xfs_ifork *xfs_iext_state_to_fork(struct xfs_inode *ip, int state);
 
 int		xfs_iformat_fork(struct xfs_inode *, struct xfs_dinode *);
-void		xfs_iflush_fork(struct xfs_inode *, struct xfs_dinode *,
+int		xfs_iflush_fork(struct xfs_inode *, struct xfs_dinode *,
 				struct xfs_inode_log_item *, int);
 void		xfs_idestroy_fork(struct xfs_inode *, int);
 void		xfs_idata_realloc(struct xfs_inode *, int, int);
diff --git a/fs/xfs/xfs_dir2_readdir.c b/fs/xfs/xfs_dir2_readdir.c
index 29816981b50a..6645a221bf8e 100644
--- a/fs/xfs/xfs_dir2_readdir.c
+++ b/fs/xfs/xfs_dir2_readdir.c
@@ -71,22 +71,11 @@ xfs_dir2_sf_getdents(
 	struct xfs_da_geometry	*geo = args->geo;
 
 	ASSERT(dp->i_df.if_flags & XFS_IFINLINE);
-	/*
-	 * Give up if the directory is way too short.
-	 */
-	if (dp->i_d.di_size < offsetof(xfs_dir2_sf_hdr_t, parent)) {
-		ASSERT(XFS_FORCED_SHUTDOWN(dp->i_mount));
-		return -EIO;
-	}
-
 	ASSERT(dp->i_df.if_bytes == dp->i_d.di_size);
 	ASSERT(dp->i_df.if_u1.if_data != NULL);
 
 	sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
 
-	if (dp->i_d.di_size < xfs_dir2_sf_hdr_size(sfp->i8count))
-		return -EFSCORRUPTED;
-
 	/*
 	 * If the block number in the offset is out of range, we're done.
 	 */
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index e50636c9a89c..5af5e76db7a3 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -3491,6 +3491,7 @@ xfs_iflush_int(
 	struct xfs_inode_log_item *iip = ip->i_itemp;
 	struct xfs_dinode	*dip;
 	struct xfs_mount	*mp = ip->i_mount;
+	int			error;
 
 	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_ILOCK_SHARED));
 	ASSERT(xfs_isiflocked(ip));
@@ -3573,9 +3574,14 @@ xfs_iflush_int(
 	if (ip->i_d.di_flushiter == DI_MAX_FLUSH)
 		ip->i_d.di_flushiter = 0;
 
-	xfs_iflush_fork(ip, dip, iip, XFS_DATA_FORK);
-	if (XFS_IFORK_Q(ip))
-		xfs_iflush_fork(ip, dip, iip, XFS_ATTR_FORK);
+	error = xfs_iflush_fork(ip, dip, iip, XFS_DATA_FORK);
+	if (error)
+		return error;
+	if (XFS_IFORK_Q(ip)) {
+		error = xfs_iflush_fork(ip, dip, iip, XFS_ATTR_FORK);
+		if (error)
+			return error;
+	}
 	xfs_inobp_check(mp, bp);
 
 	/*
-- 
2.11.0

Re: [PATCH 01/25] xfs: verify inline directory data forks

[Greg KH]

On Sat, Jun 03, 2017 at 03:18:12PM +0200, Christoph Hellwig wrote:
> From: "Darrick J. Wong" <darrick.wong@oracle.com>
> 
> When we're reading or writing the data fork of an inline directory,
> check the contents to make sure we're not overflowing buffers or eating
> garbage data.  xfs/348 corrupts an inline symlink into an inline
> directory, triggering a buffer overflow bug.
> 
> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
> Reviewed-by: Brian Foster <bfoster@redhat.com>
> ---
> v2: add more checks consistent with _dir2_sf_check and make the verifier
> usable from anywhere.

What is the git commit id of this patch upstream?

thanks,

greg k-h

Re: [PATCH 01/25] xfs: verify inline directory data forks

[Greg KH]

On Mon, Jun 05, 2017 at 04:04:07PM +0200, Greg KH wrote:
> On Sat, Jun 03, 2017 at 03:18:12PM +0200, Christoph Hellwig wrote:
> > From: "Darrick J. Wong" <darrick.wong@oracle.com>
> > 
> > When we're reading or writing the data fork of an inline directory,
> > check the contents to make sure we're not overflowing buffers or eating
> > garbage data.  xfs/348 corrupts an inline symlink into an inline
> > directory, triggering a buffer overflow bug.
> > 
> > Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
> > Reviewed-by: Brian Foster <bfoster@redhat.com>
> > ---
> > v2: add more checks consistent with _dir2_sf_check and make the verifier
> > usable from anywhere.
> 
> What is the git commit id of this patch upstream?

Nevermind, I found it, 630a04e79dd41ff746b545d4fc052e0abb836120

[PATCH 02/25] xfs: rework the inline directory verifiers

[Christoph Hellwig]

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

The inline directory verifiers should be called on the inode fork data,
which means after iformat_local on the read side, and prior to
ifork_flush on the write side.  This makes the fork verifier more
consistent with the way buffer verifiers work -- i.e. they will operate
on the memory buffer that the code will be reading and writing directly.

Furthermore, revise the verifier function to return -EFSCORRUPTED so
that we don't flood the logs with corruption messages and assert
notices.  This has been a particular problem with xfs/348, which
triggers the XFS_WANT_CORRUPTED_RETURN assertions, which halts the
kernel when CONFIG_XFS_DEBUG=y.  Disk corruption isn't supposed to do
that, at least not in a verifier.

Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/libxfs/xfs_dir2_priv.h  |  3 +-
 fs/xfs/libxfs/xfs_dir2_sf.c    | 63 +++++++++++++++++++++++++++---------------
 fs/xfs/libxfs/xfs_inode_fork.c | 35 +++++++++--------------
 fs/xfs/libxfs/xfs_inode_fork.h |  2 +-
 fs/xfs/xfs_inode.c             | 19 +++++++------
 5 files changed, 66 insertions(+), 56 deletions(-)

diff --git a/fs/xfs/libxfs/xfs_dir2_priv.h b/fs/xfs/libxfs/xfs_dir2_priv.h
index 860be99407bf..699a51bb2cbe 100644
--- a/fs/xfs/libxfs/xfs_dir2_priv.h
+++ b/fs/xfs/libxfs/xfs_dir2_priv.h
@@ -126,8 +126,7 @@ extern int xfs_dir2_sf_create(struct xfs_da_args *args, xfs_ino_t pino);
 extern int xfs_dir2_sf_lookup(struct xfs_da_args *args);
 extern int xfs_dir2_sf_removename(struct xfs_da_args *args);
 extern int xfs_dir2_sf_replace(struct xfs_da_args *args);
-extern int xfs_dir2_sf_verify(struct xfs_mount *mp, struct xfs_dir2_sf_hdr *sfp,
-		int size);
+extern int xfs_dir2_sf_verify(struct xfs_inode *ip);
 
 /* xfs_dir2_readdir.c */
 extern int xfs_readdir(struct xfs_inode *dp, struct dir_context *ctx,
diff --git a/fs/xfs/libxfs/xfs_dir2_sf.c b/fs/xfs/libxfs/xfs_dir2_sf.c
index 96b45cd6c63f..e84af093b2ab 100644
--- a/fs/xfs/libxfs/xfs_dir2_sf.c
+++ b/fs/xfs/libxfs/xfs_dir2_sf.c
@@ -632,36 +632,49 @@ xfs_dir2_sf_check(
 /* Verify the consistency of an inline directory. */
 int
 xfs_dir2_sf_verify(
-	struct xfs_mount		*mp,
-	struct xfs_dir2_sf_hdr		*sfp,
-	int				size)
+	struct xfs_inode		*ip)
 {
+	struct xfs_mount		*mp = ip->i_mount;
+	struct xfs_dir2_sf_hdr		*sfp;
 	struct xfs_dir2_sf_entry	*sfep;
 	struct xfs_dir2_sf_entry	*next_sfep;
 	char				*endp;
 	const struct xfs_dir_ops	*dops;
+	struct xfs_ifork		*ifp;
 	xfs_ino_t			ino;
 	int				i;
 	int				i8count;
 	int				offset;
+	int				size;
+	int				error;
 	__uint8_t			filetype;
 
+	ASSERT(ip->i_d.di_format == XFS_DINODE_FMT_LOCAL);
+	/*
+	 * xfs_iread calls us before xfs_setup_inode sets up ip->d_ops,
+	 * so we can only trust the mountpoint to have the right pointer.
+	 */
 	dops = xfs_dir_get_ops(mp, NULL);
 
+	ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
+	sfp = (struct xfs_dir2_sf_hdr *)ifp->if_u1.if_data;
+	size = ifp->if_bytes;
+
 	/*
 	 * Give up if the directory is way too short.
 	 */
-	XFS_WANT_CORRUPTED_RETURN(mp, size >
-			offsetof(struct xfs_dir2_sf_hdr, parent));
-	XFS_WANT_CORRUPTED_RETURN(mp, size >=
-			xfs_dir2_sf_hdr_size(sfp->i8count));
+	if (size <= offsetof(struct xfs_dir2_sf_hdr, parent) ||
+	    size < xfs_dir2_sf_hdr_size(sfp->i8count))
+		return -EFSCORRUPTED;
 
 	endp = (char *)sfp + size;
 
 	/* Check .. entry */
 	ino = dops->sf_get_parent_ino(sfp);
 	i8count = ino > XFS_DIR2_MAX_SHORT_INUM;
-	XFS_WANT_CORRUPTED_RETURN(mp, !xfs_dir_ino_validate(mp, ino));
+	error = xfs_dir_ino_validate(mp, ino);
+	if (error)
+		return error;
 	offset = dops->data_first_offset;
 
 	/* Check all reported entries */
@@ -672,12 +685,12 @@ xfs_dir2_sf_verify(
 		 * Check the fixed-offset parts of the structure are
 		 * within the data buffer.
 		 */
-		XFS_WANT_CORRUPTED_RETURN(mp,
-				((char *)sfep + sizeof(*sfep)) < endp);
+		if (((char *)sfep + sizeof(*sfep)) >= endp)
+			return -EFSCORRUPTED;
 
 		/* Don't allow names with known bad length. */
-		XFS_WANT_CORRUPTED_RETURN(mp, sfep->namelen > 0);
-		XFS_WANT_CORRUPTED_RETURN(mp, sfep->namelen < MAXNAMELEN);
+		if (sfep->namelen == 0)
+			return -EFSCORRUPTED;
 
 		/*
 		 * Check that the variable-length part of the structure is
@@ -685,33 +698,39 @@ xfs_dir2_sf_verify(
 		 * name component, so nextentry is an acceptable test.
 		 */
 		next_sfep = dops->sf_nextentry(sfp, sfep);
-		XFS_WANT_CORRUPTED_RETURN(mp, endp >= (char *)next_sfep);
+		if (endp < (char *)next_sfep)
+			return -EFSCORRUPTED;
 
 		/* Check that the offsets always increase. */
-		XFS_WANT_CORRUPTED_RETURN(mp,
-				xfs_dir2_sf_get_offset(sfep) >= offset);
+		if (xfs_dir2_sf_get_offset(sfep) < offset)
+			return -EFSCORRUPTED;
 
 		/* Check the inode number. */
 		ino = dops->sf_get_ino(sfp, sfep);
 		i8count += ino > XFS_DIR2_MAX_SHORT_INUM;
-		XFS_WANT_CORRUPTED_RETURN(mp, !xfs_dir_ino_validate(mp, ino));
+		error = xfs_dir_ino_validate(mp, ino);
+		if (error)
+			return error;
 
 		/* Check the file type. */
 		filetype = dops->sf_get_ftype(sfep);
-		XFS_WANT_CORRUPTED_RETURN(mp, filetype < XFS_DIR3_FT_MAX);
+		if (filetype >= XFS_DIR3_FT_MAX)
+			return -EFSCORRUPTED;
 
 		offset = xfs_dir2_sf_get_offset(sfep) +
 				dops->data_entsize(sfep->namelen);
 
 		sfep = next_sfep;
 	}
-	XFS_WANT_CORRUPTED_RETURN(mp, i8count == sfp->i8count);
-	XFS_WANT_CORRUPTED_RETURN(mp, (void *)sfep == (void *)endp);
+	if (i8count != sfp->i8count)
+		return -EFSCORRUPTED;
+	if ((void *)sfep != (void *)endp)
+		return -EFSCORRUPTED;
 
 	/* Make sure this whole thing ought to be in local format. */
-	XFS_WANT_CORRUPTED_RETURN(mp, offset +
-	       (sfp->count + 2) * (uint)sizeof(xfs_dir2_leaf_entry_t) +
-	       (uint)sizeof(xfs_dir2_block_tail_t) <= mp->m_dir_geo->blksize);
+	if (offset + (sfp->count + 2) * (uint)sizeof(xfs_dir2_leaf_entry_t) +
+	    (uint)sizeof(xfs_dir2_block_tail_t) > mp->m_dir_geo->blksize)
+		return -EFSCORRUPTED;
 
 	return 0;
 }
diff --git a/fs/xfs/libxfs/xfs_inode_fork.c b/fs/xfs/libxfs/xfs_inode_fork.c
index 9653e964eda4..8a37efe04de3 100644
--- a/fs/xfs/libxfs/xfs_inode_fork.c
+++ b/fs/xfs/libxfs/xfs_inode_fork.c
@@ -212,6 +212,16 @@ xfs_iformat_fork(
 	if (error)
 		return error;
 
+	/* Check inline dir contents. */
+	if (S_ISDIR(VFS_I(ip)->i_mode) &&
+	    dip->di_format == XFS_DINODE_FMT_LOCAL) {
+		error = xfs_dir2_sf_verify(ip);
+		if (error) {
+			xfs_idestroy_fork(ip, XFS_DATA_FORK);
+			return error;
+		}
+	}
+
 	if (xfs_is_reflink_inode(ip)) {
 		ASSERT(ip->i_cowfp == NULL);
 		xfs_ifork_init_cow(ip);
@@ -322,8 +332,6 @@ xfs_iformat_local(
 	int		whichfork,
 	int		size)
 {
-	int		error;
-
 	/*
 	 * If the size is unreasonable, then something
 	 * is wrong and we just bail out rather than crash in
@@ -339,14 +347,6 @@ xfs_iformat_local(
 		return -EFSCORRUPTED;
 	}
 
-	if (S_ISDIR(VFS_I(ip)->i_mode) && whichfork == XFS_DATA_FORK) {
-		error = xfs_dir2_sf_verify(ip->i_mount,
-				(struct xfs_dir2_sf_hdr *)XFS_DFORK_DPTR(dip),
-				size);
-		if (error)
-			return error;
-	}
-
 	xfs_init_local_fork(ip, whichfork, XFS_DFORK_PTR(dip, whichfork), size);
 	return 0;
 }
@@ -867,7 +867,7 @@ xfs_iextents_copy(
  * In these cases, the format always takes precedence, because the
  * format indicates the current state of the fork.
  */
-int
+void
 xfs_iflush_fork(
 	xfs_inode_t		*ip,
 	xfs_dinode_t		*dip,
@@ -877,7 +877,6 @@ xfs_iflush_fork(
 	char			*cp;
 	xfs_ifork_t		*ifp;
 	xfs_mount_t		*mp;
-	int			error;
 	static const short	brootflag[2] =
 		{ XFS_ILOG_DBROOT, XFS_ILOG_ABROOT };
 	static const short	dataflag[2] =
@@ -886,7 +885,7 @@ xfs_iflush_fork(
 		{ XFS_ILOG_DEXT, XFS_ILOG_AEXT };
 
 	if (!iip)
-		return 0;
+		return;
 	ifp = XFS_IFORK_PTR(ip, whichfork);
 	/*
 	 * This can happen if we gave up in iformat in an error path,
@@ -894,19 +893,12 @@ xfs_iflush_fork(
 	 */
 	if (!ifp) {
 		ASSERT(whichfork == XFS_ATTR_FORK);
-		return 0;
+		return;
 	}
 	cp = XFS_DFORK_PTR(dip, whichfork);
 	mp = ip->i_mount;
 	switch (XFS_IFORK_FORMAT(ip, whichfork)) {
 	case XFS_DINODE_FMT_LOCAL:
-		if (S_ISDIR(VFS_I(ip)->i_mode) && whichfork == XFS_DATA_FORK) {
-			error = xfs_dir2_sf_verify(mp,
-					(struct xfs_dir2_sf_hdr *)ifp->if_u1.if_data,
-					ifp->if_bytes);
-			if (error)
-				return error;
-		}
 		if ((iip->ili_fields & dataflag[whichfork]) &&
 		    (ifp->if_bytes > 0)) {
 			ASSERT(ifp->if_u1.if_data != NULL);
@@ -959,7 +951,6 @@ xfs_iflush_fork(
 		ASSERT(0);
 		break;
 	}
-	return 0;
 }
 
 /*
diff --git a/fs/xfs/libxfs/xfs_inode_fork.h b/fs/xfs/libxfs/xfs_inode_fork.h
index 132dc59fdde6..7fb8365326d1 100644
--- a/fs/xfs/libxfs/xfs_inode_fork.h
+++ b/fs/xfs/libxfs/xfs_inode_fork.h
@@ -140,7 +140,7 @@ typedef struct xfs_ifork {
 struct xfs_ifork *xfs_iext_state_to_fork(struct xfs_inode *ip, int state);
 
 int		xfs_iformat_fork(struct xfs_inode *, struct xfs_dinode *);
-int		xfs_iflush_fork(struct xfs_inode *, struct xfs_dinode *,
+void		xfs_iflush_fork(struct xfs_inode *, struct xfs_dinode *,
 				struct xfs_inode_log_item *, int);
 void		xfs_idestroy_fork(struct xfs_inode *, int);
 void		xfs_idata_realloc(struct xfs_inode *, int, int);
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index 5af5e76db7a3..eded851cbd85 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -50,6 +50,7 @@
 #include "xfs_log.h"
 #include "xfs_bmap_btree.h"
 #include "xfs_reflink.h"
+#include "xfs_dir2_priv.h"
 
 kmem_zone_t *xfs_inode_zone;
 
@@ -3491,7 +3492,6 @@ xfs_iflush_int(
 	struct xfs_inode_log_item *iip = ip->i_itemp;
 	struct xfs_dinode	*dip;
 	struct xfs_mount	*mp = ip->i_mount;
-	int			error;
 
 	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_ILOCK_SHARED));
 	ASSERT(xfs_isiflocked(ip));
@@ -3563,6 +3563,12 @@ xfs_iflush_int(
 	if (ip->i_d.di_version < 3)
 		ip->i_d.di_flushiter++;
 
+	/* Check the inline directory data. */
+	if (S_ISDIR(VFS_I(ip)->i_mode) &&
+	    ip->i_d.di_format == XFS_DINODE_FMT_LOCAL &&
+	    xfs_dir2_sf_verify(ip))
+		goto corrupt_out;
+
 	/*
 	 * Copy the dirty parts of the inode into the on-disk inode.  We always
 	 * copy out the core of the inode, because if the inode is dirty at all
@@ -3574,14 +3580,9 @@ xfs_iflush_int(
 	if (ip->i_d.di_flushiter == DI_MAX_FLUSH)
 		ip->i_d.di_flushiter = 0;
 
-	error = xfs_iflush_fork(ip, dip, iip, XFS_DATA_FORK);
-	if (error)
-		return error;
-	if (XFS_IFORK_Q(ip)) {
-		error = xfs_iflush_fork(ip, dip, iip, XFS_ATTR_FORK);
-		if (error)
-			return error;
-	}
+	xfs_iflush_fork(ip, dip, iip, XFS_DATA_FORK);
+	if (XFS_IFORK_Q(ip))
+		xfs_iflush_fork(ip, dip, iip, XFS_ATTR_FORK);
 	xfs_inobp_check(mp, bp);
 
 	/*
-- 
2.11.0

Re: [PATCH 02/25] xfs: rework the inline directory verifiers

[Greg KH]

On Sat, Jun 03, 2017 at 03:18:13PM +0200, Christoph Hellwig wrote:
> From: "Darrick J. Wong" <darrick.wong@oracle.com>
> 
> The inline directory verifiers should be called on the inode fork data,
> which means after iformat_local on the read side, and prior to
> ifork_flush on the write side.  This makes the fork verifier more
> consistent with the way buffer verifiers work -- i.e. they will operate
> on the memory buffer that the code will be reading and writing directly.
> 
> Furthermore, revise the verifier function to return -EFSCORRUPTED so
> that we don't flood the logs with corruption messages and assert
> notices.  This has been a particular problem with xfs/348, which
> triggers the XFS_WANT_CORRUPTED_RETURN assertions, which halts the
> kernel when CONFIG_XFS_DEBUG=y.  Disk corruption isn't supposed to do
> that, at least not in a verifier.
> 
> Reviewed-by: Brian Foster <bfoster@redhat.com>
> Reviewed-by: Christoph Hellwig <hch@lst.de>
> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>

Commit 78420281a9d74014af7616958806c3aba056319e, right?

Re: [PATCH 02/25] xfs: rework the inline directory verifiers

[Christoph Hellwig]

On Mon, Jun 05, 2017 at 04:06:00PM +0200, Greg KH wrote:
> On Sat, Jun 03, 2017 at 03:18:13PM +0200, Christoph Hellwig wrote:
> > From: "Darrick J. Wong" <darrick.wong@oracle.com>
> > 
> > The inline directory verifiers should be called on the inode fork data,
> > which means after iformat_local on the read side, and prior to
> > ifork_flush on the write side.  This makes the fork verifier more
> > consistent with the way buffer verifiers work -- i.e. they will operate
> > on the memory buffer that the code will be reading and writing directly.
> > 
> > Furthermore, revise the verifier function to return -EFSCORRUPTED so
> > that we don't flood the logs with corruption messages and assert
> > notices.  This has been a particular problem with xfs/348, which
> > triggers the XFS_WANT_CORRUPTED_RETURN assertions, which halts the
> > kernel when CONFIG_XFS_DEBUG=y.  Disk corruption isn't supposed to do
> > that, at least not in a verifier.
> > 
> > Reviewed-by: Brian Foster <bfoster@redhat.com>
> > Reviewed-by: Christoph Hellwig <hch@lst.de>
> > Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
> 
> Commit 78420281a9d74014af7616958806c3aba056319e, right?

Yes.  Sorry for not including it.

[PATCH 03/25] xfs: fix kernel memory exposure problems

[Christoph Hellwig]

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

commit bf9216f922612d2db7666aae01e65064da2ffb3a upstream.

Fix a memory exposure problems in inumbers where we allocate an array of
structures with holes, fail to zero the holes, then blindly copy the
kernel memory contents (junk and all) into userspace.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/xfs/xfs_itable.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/xfs/xfs_itable.c b/fs/xfs/xfs_itable.c
index 66e881790c17..d8a77dbf4e3a 100644
--- a/fs/xfs/xfs_itable.c
+++ b/fs/xfs/xfs_itable.c
@@ -585,7 +585,7 @@ xfs_inumbers(
 		return error;
 
 	bcount = MIN(left, (int)(PAGE_SIZE / sizeof(*buffer)));
-	buffer = kmem_alloc(bcount * sizeof(*buffer), KM_SLEEP);
+	buffer = kmem_zalloc(bcount * sizeof(*buffer), KM_SLEEP);
 	do {
 		struct xfs_inobt_rec_incore	r;
 		int				stat;
-- 
2.11.0

[PATCH 04/25] xfs: use dedicated log worker wq to avoid deadlock with cil wq

[Christoph Hellwig]

From: Brian Foster <bfoster@redhat.com>

commit 696a562072e3c14bcd13ae5acc19cdf27679e865 upstream.

The log covering background task used to be part of the xfssyncd
workqueue. That workqueue was removed as of commit 5889608df ("xfs:
syncd workqueue is no more") and the associated work item scheduled
to the xfs-log wq. The latter is used for log buffer I/O completion.

Since xfs_log_worker() can invoke a log flush, a deadlock is
possible between the xfs-log and xfs-cil workqueues. Consider the
following codepath from xfs_log_worker():

xfs_log_worker()
  xfs_log_force()
    _xfs_log_force()
      xlog_cil_force()
        xlog_cil_force_lsn()
          xlog_cil_push_now()
            flush_work()

The above is in xfs-log wq context and blocked waiting on the
completion of an xfs-cil work item. Concurrently, the cil push in
progress can end up blocked here:

xlog_cil_push_work()
  xlog_cil_push()
    xlog_write()
      xlog_state_get_iclog_space()
        xlog_wait(&log->l_flush_wait, ...)

The above is in xfs-cil context waiting on log buffer I/O
completion, which executes in xfs-log wq context. In this scenario
both workqueues are deadlocked waiting on eachother.

Add a new workqueue specifically for the high level log covering and
ail pushing worker, as was the case prior to commit 5889608df.

Diagnosed-by: David Jeffery <djeffery@redhat.com>
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.c   | 2 +-
 fs/xfs/xfs_mount.h | 1 +
 fs/xfs/xfs_super.c | 8 ++++++++
 3 files changed, 10 insertions(+), 1 deletion(-)

diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c
index 4017aa967331..b57ab34fbf3c 100644
--- a/fs/xfs/xfs_log.c
+++ b/fs/xfs/xfs_log.c
@@ -1293,7 +1293,7 @@ void
 xfs_log_work_queue(
 	struct xfs_mount        *mp)
 {
-	queue_delayed_work(mp->m_log_workqueue, &mp->m_log->l_work,
+	queue_delayed_work(mp->m_sync_workqueue, &mp->m_log->l_work,
 				msecs_to_jiffies(xfs_syncd_centisecs * 10));
 }
 
diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h
index 1bf878b0492c..5415f9031ef8 100644
--- a/fs/xfs/xfs_mount.h
+++ b/fs/xfs/xfs_mount.h
@@ -183,6 +183,7 @@ typedef struct xfs_mount {
 	struct workqueue_struct	*m_reclaim_workqueue;
 	struct workqueue_struct	*m_log_workqueue;
 	struct workqueue_struct *m_eofblocks_workqueue;
+	struct workqueue_struct	*m_sync_workqueue;
 
 	/*
 	 * Generation of the filesysyem layout.  This is incremented by each
diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
index dbbd3f1fd2b7..882fb8524fcb 100644
--- a/fs/xfs/xfs_super.c
+++ b/fs/xfs/xfs_super.c
@@ -872,8 +872,15 @@ xfs_init_mount_workqueues(
 	if (!mp->m_eofblocks_workqueue)
 		goto out_destroy_log;
 
+	mp->m_sync_workqueue = alloc_workqueue("xfs-sync/%s", WQ_FREEZABLE, 0,
+					       mp->m_fsname);
+	if (!mp->m_sync_workqueue)
+		goto out_destroy_eofb;
+
 	return 0;
 
+out_destroy_eofb:
+	destroy_workqueue(mp->m_eofblocks_workqueue);
 out_destroy_log:
 	destroy_workqueue(mp->m_log_workqueue);
 out_destroy_reclaim:
@@ -894,6 +901,7 @@ STATIC void
 xfs_destroy_mount_workqueues(
 	struct xfs_mount	*mp)
 {
+	destroy_workqueue(mp->m_sync_workqueue);
 	destroy_workqueue(mp->m_eofblocks_workqueue);
 	destroy_workqueue(mp->m_log_workqueue);
 	destroy_workqueue(mp->m_reclaim_workqueue);
-- 
2.11.0

[PATCH 05/25] xfs: fix over-copying of getbmap parameters from userspace

[Christoph Hellwig]

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

commit be6324c00c4d1e0e665f03ed1fc18863a88da119 upstream.

In xfs_ioc_getbmap, we should only copy the fields of struct getbmap
from userspace, or else we end up copying random stack contents into the
kernel.  struct getbmap is a strict subset of getbmapx, so a partial
structure copy should work fine.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/xfs/xfs_ioctl.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/fs/xfs/xfs_ioctl.c b/fs/xfs/xfs_ioctl.c
index a39197501a7c..73cfc7179124 100644
--- a/fs/xfs/xfs_ioctl.c
+++ b/fs/xfs/xfs_ioctl.c
@@ -1542,10 +1542,11 @@ xfs_ioc_getbmap(
 	unsigned int		cmd,
 	void			__user *arg)
 {
-	struct getbmapx		bmx;
+	struct getbmapx		bmx = { 0 };
 	int			error;
 
-	if (copy_from_user(&bmx, arg, sizeof(struct getbmapx)))
+	/* struct getbmap is a strict subset of struct getbmapx. */
+	if (copy_from_user(&bmx, arg, offsetof(struct getbmapx, bmv_iflags)))
 		return -EFAULT;
 
 	if (bmx.bmv_count < 2)
-- 
2.11.0

[PATCH 06/25] xfs: actually report xattr extents via iomap

[Christoph Hellwig]

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

commit 84358536dc355a9c8978ee425f87e116186bed16 upstream.

Apparently FIEMAP for xattrs has been broken since we switched to
the iomap backend because of an incorrect check for xattr presence.
Also fix the broken locking.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/xfs/xfs_iomap.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c
index 360562484e7b..65740d1cbd92 100644
--- a/fs/xfs/xfs_iomap.c
+++ b/fs/xfs/xfs_iomap.c
@@ -1151,10 +1151,10 @@ xfs_xattr_iomap_begin(
 	if (XFS_FORCED_SHUTDOWN(mp))
 		return -EIO;
 
-	lockmode = xfs_ilock_data_map_shared(ip);
+	lockmode = xfs_ilock_attr_map_shared(ip);
 
 	/* if there are no attribute fork or extents, return ENOENT */
-	if (XFS_IFORK_Q(ip) || !ip->i_d.di_anextents) {
+	if (!XFS_IFORK_Q(ip) || !ip->i_d.di_anextents) {
 		error = -ENOENT;
 		goto out_unlock;
 	}
-- 
2.11.0

[PATCH 07/25] xfs: drop iolock from reclaim context to appease lockdep

[Christoph Hellwig]

From: Brian Foster <bfoster@redhat.com>

commit 3b4683c294095b5f777c03307ef8c60f47320e12 upstream.

Lockdep complains about use of the iolock in inode reclaim context
because it doesn't understand that reclaim has the last reference to
the inode, and thus an iolock->reclaim->iolock deadlock is not
possible.

The iolock is technically not necessary in xfs_inactive() and was
only added to appease an assert in xfs_free_eofblocks(), which can
be called from other non-reclaim contexts. Therefore, just kill the
assert and drop the use of the iolock from reclaim context to quiet
lockdep.

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_bmap_util.c | 8 +++-----
 fs/xfs/xfs_inode.c     | 9 +++++----
 2 files changed, 8 insertions(+), 9 deletions(-)

diff --git a/fs/xfs/xfs_bmap_util.c b/fs/xfs/xfs_bmap_util.c
index 5328ecdd03d4..8a1a62ee84ad 100644
--- a/fs/xfs/xfs_bmap_util.c
+++ b/fs/xfs/xfs_bmap_util.c
@@ -911,9 +911,9 @@ xfs_can_free_eofblocks(struct xfs_inode *ip, bool force)
 }
 
 /*
- * This is called by xfs_inactive to free any blocks beyond eof
- * when the link count isn't zero and by xfs_dm_punch_hole() when
- * punching a hole to EOF.
+ * This is called to free any blocks beyond eof. The caller must hold
+ * IOLOCK_EXCL unless we are in the inode reclaim path and have the only
+ * reference to the inode.
  */
 int
 xfs_free_eofblocks(
@@ -928,8 +928,6 @@ xfs_free_eofblocks(
 	struct xfs_bmbt_irec	imap;
 	struct xfs_mount	*mp = ip->i_mount;
 
-	ASSERT(xfs_isilocked(ip, XFS_IOLOCK_EXCL));
-
 	/*
 	 * Figure out if there are any blocks beyond the end
 	 * of the file.  If not, then there is nothing to do.
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index eded851cbd85..7a0b4eeb99e4 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -1915,12 +1915,13 @@ xfs_inactive(
 		 * force is true because we are evicting an inode from the
 		 * cache. Post-eof blocks must be freed, lest we end up with
 		 * broken free space accounting.
+		 *
+		 * Note: don't bother with iolock here since lockdep complains
+		 * about acquiring it in reclaim context. We have the only
+		 * reference to the inode at this point anyways.
 		 */
-		if (xfs_can_free_eofblocks(ip, true)) {
-			xfs_ilock(ip, XFS_IOLOCK_EXCL);
+		if (xfs_can_free_eofblocks(ip, true))
 			xfs_free_eofblocks(ip);
-			xfs_iunlock(ip, XFS_IOLOCK_EXCL);
-		}
 
 		return;
 	}
-- 
2.11.0

[PATCH 08/25] xfs: fix integer truncation in xfs_bmap_remap_alloc

[Christoph Hellwig]

commit 52813fb13ff90bd9c39a93446cbf1103c290b6e9 upstream.

bno should be a xfs_fsblock_t, which is 64-bit wides instead of a
xfs_aglock_t, which truncates the value to 32 bits.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/libxfs/xfs_bmap.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/xfs/libxfs/xfs_bmap.c b/fs/xfs/libxfs/xfs_bmap.c
index 5a508b011e27..23f2b4b36f38 100644
--- a/fs/xfs/libxfs/xfs_bmap.c
+++ b/fs/xfs/libxfs/xfs_bmap.c
@@ -3964,7 +3964,7 @@ xfs_bmap_remap_alloc(
 {
 	struct xfs_trans	*tp = ap->tp;
 	struct xfs_mount	*mp = tp->t_mountp;
-	xfs_agblock_t		bno;
+	xfs_fsblock_t		bno;
 	struct xfs_alloc_arg	args;
 	int			error;
 
-- 
2.11.0

[PATCH 09/25] xfs: handle array index overrun in xfs_dir2_leaf_readbuf()

[Christoph Hellwig]

From: Eric Sandeen <sandeen@redhat.com>

commit 023cc840b40fad95c6fe26fff1d380a8c9d45939 upstream.

Carlos had a case where "find" seemed to start spinning
forever and never return.

This was on a filesystem with non-default multi-fsb (8k)
directory blocks, and a fragmented directory with extents
like this:

0:[0,133646,2,0]
1:[2,195888,1,0]
2:[3,195890,1,0]
3:[4,195892,1,0]
4:[5,195894,1,0]
5:[6,195896,1,0]
6:[7,195898,1,0]
7:[8,195900,1,0]
8:[9,195902,1,0]
9:[10,195908,1,0]
10:[11,195910,1,0]
11:[12,195912,1,0]
12:[13,195914,1,0]
...

i.e. the first extent is a contiguous 2-fsb dir block, but
after that it is fragmented into 1 block extents.

At the top of the readdir path, we allocate a mapping array
which (for this filesystem geometry) can hold 10 extents; see
the assignment to map_info->map_size.  During readdir, we are
therefore able to map extents 0 through 9 above into the array
for readahead purposes.  If we count by 2, we see that the last
mapped index (9) is the first block of a 2-fsb directory block.

At the end of xfs_dir2_leaf_readbuf() we have 2 loops to fill
more readahead; the outer loop assumes one full dir block is
processed each loop iteration, and an inner loop that ensures
that this is so by advancing to the next extent until a full
directory block is mapped.

The problem is that this inner loop may step past the last
extent in the mapping array as it tries to reach the end of
the directory block.  This will read garbage for the extent
length, and as a result the loop control variable 'j' may
become corrupted and never fail the loop conditional.

The number of valid mappings we have in our array is stored
in map->map_valid, so stop this inner loop based on that limit.

There is an ASSERT at the top of the outer loop for this
same condition, but we never made it out of the inner loop,
so the ASSERT never fired.

Huge appreciation for Carlos for debugging and isolating
the problem.

Debugged-and-analyzed-by: Carlos Maiolino <cmaiolino@redhat.com>
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Tested-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: Bill O'Donnell <billodo@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/xfs_dir2_readdir.c | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)

diff --git a/fs/xfs/xfs_dir2_readdir.c b/fs/xfs/xfs_dir2_readdir.c
index 6645a221bf8e..ba95846f8f8a 100644
--- a/fs/xfs/xfs_dir2_readdir.c
+++ b/fs/xfs/xfs_dir2_readdir.c
@@ -394,6 +394,7 @@ xfs_dir2_leaf_readbuf(
 
 	/*
 	 * Do we need more readahead?
+	 * Each loop tries to process 1 full dir blk; last may be partial.
 	 */
 	blk_start_plug(&plug);
 	for (mip->ra_index = mip->ra_offset = i = 0;
@@ -425,9 +426,14 @@ xfs_dir2_leaf_readbuf(
 		}
 
 		/*
-		 * Advance offset through the mapping table.
+		 * Advance offset through the mapping table, processing a full
+		 * dir block even if it is fragmented into several extents.
+		 * But stop if we have consumed all valid mappings, even if
+		 * it's not yet a full directory block.
 		 */
-		for (j = 0; j < geo->fsbcount; j += length ) {
+		for (j = 0;
+		     j < geo->fsbcount && mip->ra_index < mip->map_valid;
+		     j += length ) {
 			/*
 			 * The rest of this extent but not more than a dir
 			 * block.
-- 
2.11.0

[PATCH 10/25] xfs: prevent multi-fsb dir readahead from reading random blocks

[Christoph Hellwig]

From: Brian Foster <bfoster@redhat.com>

commit cb52ee334a45ae6c78a3999e4b473c43ddc528f4 upstream.

Directory block readahead uses a complex iteration mechanism to map
between high-level directory blocks and underlying physical extents.
This mechanism attempts to traverse the higher-level dir blocks in a
manner that handles multi-fsb directory blocks and simultaneously
maintains a reference to the corresponding physical blocks.

This logic doesn't handle certain (discontiguous) physical extent
layouts correctly with multi-fsb directory blocks. For example,
consider the case of a 4k FSB filesystem with a 2 FSB (8k) directory
block size and a directory with the following extent layout:

 EXT: FILE-OFFSET      BLOCK-RANGE      AG AG-OFFSET        TOTAL
   0: [0..7]:          88..95            0 (88..95)             8
   1: [8..15]:         80..87            0 (80..87)             8
   2: [16..39]:        168..191          0 (168..191)          24
   3: [40..63]:        5242952..5242975  1 (72..95)            24

Directory block 0 spans physical extents 0 and 1, dirblk 1 lies
entirely within extent 2 and dirblk 2 spans extents 2 and 3. Because
extent 2 is larger than the directory block size, the readahead code
erroneously assumes the block is contiguous and issues a readahead
based on the physical mapping of the first fsb of the dirblk. This
results in read verifier failure and a spurious corruption or crc
failure, depending on the filesystem format.

Further, the subsequent readahead code responsible for walking
through the physical table doesn't correctly advance the physical
block reference for dirblk 2. Instead of advancing two physical
filesystem blocks, the first iteration of the loop advances 1 block
(correctly), but the subsequent iteration advances 2 more physical
blocks because the next physical extent (extent 3, above) happens to
cover more than dirblk 2. At this point, the higher-level directory
block walking is completely off the rails of the actual physical
layout of the directory for the respective mapping table.

Update the contiguous dirblock logic to consider the current offset
in the physical extent to avoid issuing directory readahead to
unrelated blocks. Also, update the mapping table advancing code to
consider the current offset within the current dirblock to avoid
advancing the mapping reference too far beyond the dirblock.

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_dir2_readdir.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/fs/xfs/xfs_dir2_readdir.c b/fs/xfs/xfs_dir2_readdir.c
index ba95846f8f8a..eba63160be6c 100644
--- a/fs/xfs/xfs_dir2_readdir.c
+++ b/fs/xfs/xfs_dir2_readdir.c
@@ -405,7 +405,8 @@ xfs_dir2_leaf_readbuf(
 		 * Read-ahead a contiguous directory block.
 		 */
 		if (i > mip->ra_current &&
-		    map[mip->ra_index].br_blockcount >= geo->fsbcount) {
+		    (map[mip->ra_index].br_blockcount - mip->ra_offset) >=
+		    geo->fsbcount) {
 			xfs_dir3_data_readahead(dp,
 				map[mip->ra_index].br_startoff + mip->ra_offset,
 				XFS_FSB_TO_DADDR(dp->i_mount,
@@ -438,7 +439,7 @@ xfs_dir2_leaf_readbuf(
 			 * The rest of this extent but not more than a dir
 			 * block.
 			 */
-			length = min_t(int, geo->fsbcount,
+			length = min_t(int, geo->fsbcount - j,
 					map[mip->ra_index].br_blockcount -
 							mip->ra_offset);
 			mip->ra_offset += length;
-- 
2.11.0

[PATCH 11/25] xfs: fix up quotacheck buffer list error handling

[Christoph Hellwig]

From: Brian Foster <bfoster@redhat.com>

commit 20e8a063786050083fe05b4f45be338c60b49126 upstream.

The quotacheck error handling of the delwri buffer list assumes the
resident buffers are locked and doesn't clear the _XBF_DELWRI_Q flag
on the buffers that are dequeued. This can lead to assert failures
on buffer release and possibly other locking problems.

Move this code to a delwri queue cancel helper function to
encapsulate the logic required to properly release buffers from a
delwri queue. Update the helper to clear the delwri queue flag and
call it from quotacheck.

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.c | 24 ++++++++++++++++++++++++
 fs/xfs/xfs_buf.h |  1 +
 fs/xfs/xfs_qm.c  |  7 +------
 3 files changed, 26 insertions(+), 6 deletions(-)

diff --git a/fs/xfs/xfs_buf.c b/fs/xfs/xfs_buf.c
index d7a67d7fbc7f..88f4086308c0 100644
--- a/fs/xfs/xfs_buf.c
+++ b/fs/xfs/xfs_buf.c
@@ -1052,6 +1052,8 @@ void
 xfs_buf_unlock(
 	struct xfs_buf		*bp)
 {
+	ASSERT(xfs_buf_islocked(bp));
+
 	XB_CLEAR_OWNER(bp);
 	up(&bp->b_sema);
 
@@ -1790,6 +1792,28 @@ xfs_alloc_buftarg(
 }
 
 /*
+ * Cancel a delayed write list.
+ *
+ * Remove each buffer from the list, clear the delwri queue flag and drop the
+ * associated buffer reference.
+ */
+void
+xfs_buf_delwri_cancel(
+	struct list_head	*list)
+{
+	struct xfs_buf		*bp;
+
+	while (!list_empty(list)) {
+		bp = list_first_entry(list, struct xfs_buf, b_list);
+
+		xfs_buf_lock(bp);
+		bp->b_flags &= ~_XBF_DELWRI_Q;
+		list_del_init(&bp->b_list);
+		xfs_buf_relse(bp);
+	}
+}
+
+/*
  * Add a buffer to the delayed write list.
  *
  * This queues a buffer for writeout if it hasn't already been.  Note that
diff --git a/fs/xfs/xfs_buf.h b/fs/xfs/xfs_buf.h
index 1c2e52b2d926..24da7ed1c787 100644
--- a/fs/xfs/xfs_buf.h
+++ b/fs/xfs/xfs_buf.h
@@ -330,6 +330,7 @@ extern void *xfs_buf_offset(struct xfs_buf *, size_t);
 extern void xfs_buf_stale(struct xfs_buf *bp);
 
 /* Delayed Write Buffer Routines */
+extern void xfs_buf_delwri_cancel(struct list_head *);
 extern bool xfs_buf_delwri_queue(struct xfs_buf *, struct list_head *);
 extern int xfs_buf_delwri_submit(struct list_head *);
 extern int xfs_buf_delwri_submit_nowait(struct list_head *);
diff --git a/fs/xfs/xfs_qm.c b/fs/xfs/xfs_qm.c
index b669b123287b..8b9a9f15f022 100644
--- a/fs/xfs/xfs_qm.c
+++ b/fs/xfs/xfs_qm.c
@@ -1384,12 +1384,7 @@ xfs_qm_quotacheck(
 	mp->m_qflags |= flags;
 
  error_return:
-	while (!list_empty(&buffer_list)) {
-		struct xfs_buf *bp =
-			list_first_entry(&buffer_list, struct xfs_buf, b_list);
-		list_del_init(&bp->b_list);
-		xfs_buf_relse(bp);
-	}
+	xfs_buf_delwri_cancel(&buffer_list);
 
 	if (error) {
 		xfs_warn(mp,
-- 
2.11.0

[PATCH 12/25] xfs: support ability to wait on new inodes

[Christoph Hellwig]

From: Brian Foster <bfoster@redhat.com>

commit 756baca27fff3ecaeab9dbc7a5ee35a1d7bc0c7f upstream.

Inodes that are inserted into the perag tree but still under
construction are flagged with the XFS_INEW bit. Most contexts either
skip such inodes when they are encountered or have the ability to
handle them.

The runtime quotaoff sequence introduces a context that must wait
for construction of such inodes to correctly ensure that all dquots
in the fs are released. In anticipation of this, support the ability
to wait on new inodes. Wake the appropriate bit when XFS_INEW is
cleared.

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_icache.c | 5 ++++-
 fs/xfs/xfs_inode.h  | 4 +++-
 2 files changed, 7 insertions(+), 2 deletions(-)

diff --git a/fs/xfs/xfs_icache.c b/fs/xfs/xfs_icache.c
index 3fb1f3fb8efe..027cf8ba2235 100644
--- a/fs/xfs/xfs_icache.c
+++ b/fs/xfs/xfs_icache.c
@@ -368,14 +368,17 @@ xfs_iget_cache_hit(
 
 		error = xfs_reinit_inode(mp, inode);
 		if (error) {
+			bool wake;
 			/*
 			 * Re-initializing the inode failed, and we are in deep
 			 * trouble.  Try to re-add it to the reclaim list.
 			 */
 			rcu_read_lock();
 			spin_lock(&ip->i_flags_lock);
-
+			wake = !!__xfs_iflags_test(ip, XFS_INEW);
 			ip->i_flags &= ~(XFS_INEW | XFS_IRECLAIM);
+			if (wake)
+				wake_up_bit(&ip->i_flags, __XFS_INEW_BIT);
 			ASSERT(ip->i_flags & XFS_IRECLAIMABLE);
 			trace_xfs_iget_reclaim_fail(ip);
 			goto out_error;
diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h
index 71e8a81c91a3..c038f6eecc28 100644
--- a/fs/xfs/xfs_inode.h
+++ b/fs/xfs/xfs_inode.h
@@ -217,7 +217,8 @@ static inline bool xfs_is_reflink_inode(struct xfs_inode *ip)
 #define XFS_IRECLAIM		(1 << 0) /* started reclaiming this inode */
 #define XFS_ISTALE		(1 << 1) /* inode has been staled */
 #define XFS_IRECLAIMABLE	(1 << 2) /* inode can be reclaimed */
-#define XFS_INEW		(1 << 3) /* inode has just been allocated */
+#define __XFS_INEW_BIT		3	 /* inode has just been allocated */
+#define XFS_INEW		(1 << __XFS_INEW_BIT)
 #define XFS_ITRUNCATED		(1 << 5) /* truncated down so flush-on-close */
 #define XFS_IDIRTY_RELEASE	(1 << 6) /* dirty release already seen */
 #define __XFS_IFLOCK_BIT	7	 /* inode is being flushed right now */
@@ -467,6 +468,7 @@ static inline void xfs_finish_inode_setup(struct xfs_inode *ip)
 	xfs_iflags_clear(ip, XFS_INEW);
 	barrier();
 	unlock_new_inode(VFS_I(ip));
+	wake_up_bit(&ip->i_flags, __XFS_INEW_BIT);
 }
 
 static inline void xfs_setup_existing_inode(struct xfs_inode *ip)
-- 
2.11.0

[PATCH 13/25] xfs: update ag iterator to support wait on new inodes

[Christoph Hellwig]

From: Brian Foster <bfoster@redhat.com>

commit ae2c4ac2dd39b23a87ddb14ceddc3f2872c6aef5 upstream.

The AG inode iterator currently skips new inodes as such inodes are
inserted into the inode radix tree before they are fully
constructed. Certain contexts require the ability to wait on the
construction of new inodes, however. The fs-wide dquot release from
the quotaoff sequence is an example of this.

Update the AG inode iterator to support the ability to wait on
inodes flagged with XFS_INEW upon request. Create a new
xfs_inode_ag_iterator_flags() interface and support a set of
iteration flags to modify the iteration behavior. When the
XFS_AGITER_INEW_WAIT flag is set, include XFS_INEW flags in the
radix tree inode lookup and wait on them before the callback is
executed.

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_icache.c | 53 +++++++++++++++++++++++++++++++++++++++++++++--------
 fs/xfs/xfs_icache.h |  8 ++++++++
 2 files changed, 53 insertions(+), 8 deletions(-)

diff --git a/fs/xfs/xfs_icache.c b/fs/xfs/xfs_icache.c
index 027cf8ba2235..74304b6ce84b 100644
--- a/fs/xfs/xfs_icache.c
+++ b/fs/xfs/xfs_icache.c
@@ -264,6 +264,22 @@ xfs_inode_clear_reclaim_tag(
 	xfs_perag_clear_reclaim_tag(pag);
 }
 
+static void
+xfs_inew_wait(
+	struct xfs_inode	*ip)
+{
+	wait_queue_head_t *wq = bit_waitqueue(&ip->i_flags, __XFS_INEW_BIT);
+	DEFINE_WAIT_BIT(wait, &ip->i_flags, __XFS_INEW_BIT);
+
+	do {
+		prepare_to_wait(wq, &wait.wait, TASK_UNINTERRUPTIBLE);
+		if (!xfs_iflags_test(ip, XFS_INEW))
+			break;
+		schedule();
+	} while (true);
+	finish_wait(wq, &wait.wait);
+}
+
 /*
  * When we recycle a reclaimable inode, we need to re-initialise the VFS inode
  * part of the structure. This is made more complex by the fact we store
@@ -628,9 +644,11 @@ xfs_iget(
 
 STATIC int
 xfs_inode_ag_walk_grab(
-	struct xfs_inode	*ip)
+	struct xfs_inode	*ip,
+	int			flags)
 {
 	struct inode		*inode = VFS_I(ip);
+	bool			newinos = !!(flags & XFS_AGITER_INEW_WAIT);
 
 	ASSERT(rcu_read_lock_held());
 
@@ -648,7 +666,8 @@ xfs_inode_ag_walk_grab(
 		goto out_unlock_noent;
 
 	/* avoid new or reclaimable inodes. Leave for reclaim code to flush */
-	if (__xfs_iflags_test(ip, XFS_INEW | XFS_IRECLAIMABLE | XFS_IRECLAIM))
+	if ((!newinos && __xfs_iflags_test(ip, XFS_INEW)) ||
+	    __xfs_iflags_test(ip, XFS_IRECLAIMABLE | XFS_IRECLAIM))
 		goto out_unlock_noent;
 	spin_unlock(&ip->i_flags_lock);
 
@@ -676,7 +695,8 @@ xfs_inode_ag_walk(
 					   void *args),
 	int			flags,
 	void			*args,
-	int			tag)
+	int			tag,
+	int			iter_flags)
 {
 	uint32_t		first_index;
 	int			last_error = 0;
@@ -718,7 +738,7 @@ xfs_inode_ag_walk(
 		for (i = 0; i < nr_found; i++) {
 			struct xfs_inode *ip = batch[i];
 
-			if (done || xfs_inode_ag_walk_grab(ip))
+			if (done || xfs_inode_ag_walk_grab(ip, iter_flags))
 				batch[i] = NULL;
 
 			/*
@@ -746,6 +766,9 @@ xfs_inode_ag_walk(
 		for (i = 0; i < nr_found; i++) {
 			if (!batch[i])
 				continue;
+			if ((iter_flags & XFS_AGITER_INEW_WAIT) &&
+			    xfs_iflags_test(batch[i], XFS_INEW))
+				xfs_inew_wait(batch[i]);
 			error = execute(batch[i], flags, args);
 			IRELE(batch[i]);
 			if (error == -EAGAIN) {
@@ -825,12 +848,13 @@ xfs_cowblocks_worker(
 }
 
 int
-xfs_inode_ag_iterator(
+xfs_inode_ag_iterator_flags(
 	struct xfs_mount	*mp,
 	int			(*execute)(struct xfs_inode *ip, int flags,
 					   void *args),
 	int			flags,
-	void			*args)
+	void			*args,
+	int			iter_flags)
 {
 	struct xfs_perag	*pag;
 	int			error = 0;
@@ -840,7 +864,8 @@ xfs_inode_ag_iterator(
 	ag = 0;
 	while ((pag = xfs_perag_get(mp, ag))) {
 		ag = pag->pag_agno + 1;
-		error = xfs_inode_ag_walk(mp, pag, execute, flags, args, -1);
+		error = xfs_inode_ag_walk(mp, pag, execute, flags, args, -1,
+					  iter_flags);
 		xfs_perag_put(pag);
 		if (error) {
 			last_error = error;
@@ -852,6 +877,17 @@ xfs_inode_ag_iterator(
 }
 
 int
+xfs_inode_ag_iterator(
+	struct xfs_mount	*mp,
+	int			(*execute)(struct xfs_inode *ip, int flags,
+					   void *args),
+	int			flags,
+	void			*args)
+{
+	return xfs_inode_ag_iterator_flags(mp, execute, flags, args, 0);
+}
+
+int
 xfs_inode_ag_iterator_tag(
 	struct xfs_mount	*mp,
 	int			(*execute)(struct xfs_inode *ip, int flags,
@@ -868,7 +904,8 @@ xfs_inode_ag_iterator_tag(
 	ag = 0;
 	while ((pag = xfs_perag_get_tag(mp, ag, tag))) {
 		ag = pag->pag_agno + 1;
-		error = xfs_inode_ag_walk(mp, pag, execute, flags, args, tag);
+		error = xfs_inode_ag_walk(mp, pag, execute, flags, args, tag,
+					  0);
 		xfs_perag_put(pag);
 		if (error) {
 			last_error = error;
diff --git a/fs/xfs/xfs_icache.h b/fs/xfs/xfs_icache.h
index 8a7c849b4dea..9183f77958ef 100644
--- a/fs/xfs/xfs_icache.h
+++ b/fs/xfs/xfs_icache.h
@@ -48,6 +48,11 @@ struct xfs_eofblocks {
 #define XFS_IGET_UNTRUSTED	0x2
 #define XFS_IGET_DONTCACHE	0x4
 
+/*
+ * flags for AG inode iterator
+ */
+#define XFS_AGITER_INEW_WAIT	0x1	/* wait on new inodes */
+
 int xfs_iget(struct xfs_mount *mp, struct xfs_trans *tp, xfs_ino_t ino,
 	     uint flags, uint lock_flags, xfs_inode_t **ipp);
 
@@ -79,6 +84,9 @@ void xfs_cowblocks_worker(struct work_struct *);
 int xfs_inode_ag_iterator(struct xfs_mount *mp,
 	int (*execute)(struct xfs_inode *ip, int flags, void *args),
 	int flags, void *args);
+int xfs_inode_ag_iterator_flags(struct xfs_mount *mp,
+	int (*execute)(struct xfs_inode *ip, int flags, void *args),
+	int flags, void *args, int iter_flags);
 int xfs_inode_ag_iterator_tag(struct xfs_mount *mp,
 	int (*execute)(struct xfs_inode *ip, int flags, void *args),
 	int flags, void *args, int tag);
-- 
2.11.0

[PATCH 14/25] xfs: wait on new inodes during quotaoff dquot release

[Christoph Hellwig]

From: Brian Foster <bfoster@redhat.com>

commit e20c8a517f259cb4d258e10b0cd5d4b30d4167a0 upstream.

The quotaoff operation has a race with inode allocation that results
in a livelock. An inode allocation that occurs before the quota
status flags are updated acquires the appropriate dquots for the
inode via xfs_qm_vop_dqalloc(). It then inserts the XFS_INEW inode
into the perag radix tree, sometime later attaches the dquots to the
inode and finally clears the XFS_INEW flag. Quotaoff expects to
release the dquots from all inodes in the filesystem via
xfs_qm_dqrele_all_inodes(). This invokes the AG inode iterator,
which skips inodes in the XFS_INEW state because they are not fully
constructed. If the scan occurs after dquots have been attached to
an inode, but before XFS_INEW is cleared, the newly allocated inode
will continue to hold a reference to the applicable dquots. When
quotaoff invokes xfs_qm_dqpurge_all(), the reference count of those
dquot(s) remain elevated and the dqpurge scan spins indefinitely.

To address this problem, update the xfs_qm_dqrele_all_inodes() scan
to wait on inodes marked on the XFS_INEW state. We wait on the
inodes explicitly rather than skip and retry to avoid continuous
retry loops due to a parallel inode allocation workload. Since
quotaoff updates the quota state flags and uses a synchronous
transaction before the dqrele scan, and dquots are attached to
inodes after radix tree insertion iff quota is enabled, one INEW
waiting pass through the AG guarantees that the scan has processed
all inodes that could possibly hold dquot references.

Reported-by: Eryu Guan <eguan@redhat.com>
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_qm_syscalls.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/fs/xfs/xfs_qm_syscalls.c b/fs/xfs/xfs_qm_syscalls.c
index 475a3882a81f..9cb5c381b01c 100644
--- a/fs/xfs/xfs_qm_syscalls.c
+++ b/fs/xfs/xfs_qm_syscalls.c
@@ -759,5 +759,6 @@ xfs_qm_dqrele_all_inodes(
 	uint		 flags)
 {
 	ASSERT(mp->m_quotainfo);
-	xfs_inode_ag_iterator(mp, xfs_dqrele_inode, flags, NULL);
+	xfs_inode_ag_iterator_flags(mp, xfs_dqrele_inode, flags, NULL,
+				    XFS_AGITER_INEW_WAIT);
 }
-- 
2.11.0

[PATCH 15/25] xfs: reserve enough blocks to handle btree splits when remapping

[Christoph Hellwig]

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

commit fe0be23e68200573de027de9b8cc2b27e7fce35e upstream.

In xfs_reflink_end_cow, we erroneously reserve only enough blocks to
handle adding 1 extent.  This is problematic if we fragment free space,
have to do CoW, and then have to perform multiple bmap btree expansions.
Furthermore, the BUI recovery routine doesn't reserve /any/ blocks to
handle btree splits, so log recovery fails after our first error causes
the filesystem to go down.

Therefore, refactor the transaction block reservation macros until we
have a macro that works for our deferred (re)mapping activities, and fix
both problems by using that macro.

With 1k blocks we can hit this fairly often in g/187 if the scratch fs
is big enough.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/xfs/libxfs/xfs_trans_space.h | 23 +++++++++++++++++------
 fs/xfs/xfs_bmap_item.c          |  5 ++++-
 fs/xfs/xfs_reflink.c            | 18 ++++++++++++++++--
 3 files changed, 37 insertions(+), 9 deletions(-)

diff --git a/fs/xfs/libxfs/xfs_trans_space.h b/fs/xfs/libxfs/xfs_trans_space.h
index 7917f6e44286..d787c677d2a3 100644
--- a/fs/xfs/libxfs/xfs_trans_space.h
+++ b/fs/xfs/libxfs/xfs_trans_space.h
@@ -21,8 +21,20 @@
 /*
  * Components of space reservations.
  */
+
+/* Worst case number of rmaps that can be held in a block. */
 #define XFS_MAX_CONTIG_RMAPS_PER_BLOCK(mp)    \
 		(((mp)->m_rmap_mxr[0]) - ((mp)->m_rmap_mnr[0]))
+
+/* Adding one rmap could split every level up to the top of the tree. */
+#define XFS_RMAPADD_SPACE_RES(mp) ((mp)->m_rmap_maxlevels)
+
+/* Blocks we might need to add "b" rmaps to a tree. */
+#define XFS_NRMAPADD_SPACE_RES(mp, b)\
+	(((b + XFS_MAX_CONTIG_RMAPS_PER_BLOCK(mp) - 1) / \
+	  XFS_MAX_CONTIG_RMAPS_PER_BLOCK(mp)) * \
+	  XFS_RMAPADD_SPACE_RES(mp))
+
 #define XFS_MAX_CONTIG_EXTENTS_PER_BLOCK(mp)    \
 		(((mp)->m_alloc_mxr[0]) - ((mp)->m_alloc_mnr[0]))
 #define	XFS_EXTENTADD_SPACE_RES(mp,w)	(XFS_BM_MAXLEVELS(mp,w) - 1)
@@ -30,13 +42,12 @@
 	(((b + XFS_MAX_CONTIG_EXTENTS_PER_BLOCK(mp) - 1) / \
 	  XFS_MAX_CONTIG_EXTENTS_PER_BLOCK(mp)) * \
 	  XFS_EXTENTADD_SPACE_RES(mp,w))
+
+/* Blocks we might need to add "b" mappings & rmappings to a file. */
 #define XFS_SWAP_RMAP_SPACE_RES(mp,b,w)\
-	(((b + XFS_MAX_CONTIG_EXTENTS_PER_BLOCK(mp) - 1) / \
-	  XFS_MAX_CONTIG_EXTENTS_PER_BLOCK(mp)) * \
-	  XFS_EXTENTADD_SPACE_RES(mp,w) + \
-	 ((b + XFS_MAX_CONTIG_RMAPS_PER_BLOCK(mp) - 1) / \
-	  XFS_MAX_CONTIG_RMAPS_PER_BLOCK(mp)) * \
-	  (mp)->m_rmap_maxlevels)
+	(XFS_NEXTENTADD_SPACE_RES((mp), (b), (w)) + \
+	 XFS_NRMAPADD_SPACE_RES((mp), (b)))
+
 #define	XFS_DAENTER_1B(mp,w)	\
 	((w) == XFS_DATA_FORK ? (mp)->m_dir_geo->fsbcount : 1)
 #define	XFS_DAENTER_DBS(mp,w)	\
diff --git a/fs/xfs/xfs_bmap_item.c b/fs/xfs/xfs_bmap_item.c
index 9bf57c76623b..c4b90e794e41 100644
--- a/fs/xfs/xfs_bmap_item.c
+++ b/fs/xfs/xfs_bmap_item.c
@@ -34,6 +34,8 @@
 #include "xfs_bmap.h"
 #include "xfs_icache.h"
 #include "xfs_trace.h"
+#include "xfs_bmap_btree.h"
+#include "xfs_trans_space.h"
 
 
 kmem_zone_t	*xfs_bui_zone;
@@ -446,7 +448,8 @@ xfs_bui_recover(
 		return -EIO;
 	}
 
-	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_itruncate, 0, 0, 0, &tp);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_itruncate,
+			XFS_EXTENTADD_SPACE_RES(mp, XFS_DATA_FORK), 0, 0, &tp);
 	if (error)
 		return error;
 	budp = xfs_trans_get_bud(tp, buip);
diff --git a/fs/xfs/xfs_reflink.c b/fs/xfs/xfs_reflink.c
index 2252f163c38f..29a75ecb2425 100644
--- a/fs/xfs/xfs_reflink.c
+++ b/fs/xfs/xfs_reflink.c
@@ -736,8 +736,22 @@ xfs_reflink_end_cow(
 	offset_fsb = XFS_B_TO_FSBT(ip->i_mount, offset);
 	end_fsb = XFS_B_TO_FSB(ip->i_mount, offset + count);
 
-	/* Start a rolling transaction to switch the mappings */
-	resblks = XFS_EXTENTADD_SPACE_RES(ip->i_mount, XFS_DATA_FORK);
+	/*
+	 * Start a rolling transaction to switch the mappings.  We're
+	 * unlikely ever to have to remap 16T worth of single-block
+	 * extents, so just cap the worst case extent count to 2^32-1.
+	 * Stick a warning in just in case, and avoid 64-bit division.
+	 */
+	BUILD_BUG_ON(MAX_RW_COUNT > UINT_MAX);
+	if (end_fsb - offset_fsb > UINT_MAX) {
+		error = -EFSCORRUPTED;
+		xfs_force_shutdown(ip->i_mount, SHUTDOWN_CORRUPT_INCORE);
+		ASSERT(0);
+		goto out;
+	}
+	resblks = XFS_NEXTENTADD_SPACE_RES(ip->i_mount,
+			(unsigned int)(end_fsb - offset_fsb),
+			XFS_DATA_FORK);
 	error = xfs_trans_alloc(ip->i_mount, &M_RES(ip->i_mount)->tr_write,
 			resblks, 0, 0, &tp);
 	if (error)
-- 
2.11.0

[PATCH 16/25] xfs: fix use-after-free in xfs_finish_page_writeback

[Christoph Hellwig]

From: Eryu Guan <eguan@redhat.com>

commit 161f55efba5ddccc690139fae9373cafc3447a97 upstream.

Commit 28b783e47ad7 ("xfs: bufferhead chains are invalid after
end_page_writeback") fixed one use-after-free issue by
pre-calculating the loop conditionals before calling bh->b_end_io()
in the end_io processing loop, but it assigned 'next' pointer before
checking end offset boundary & breaking the loop, at which point the
bh might be freed already, and caused use-after-free.

This is caught by KASAN when running fstests generic/127 on sub-page
block size XFS.

[ 2517.244502] run fstests generic/127 at 2017-04-27 07:30:50
[ 2747.868840] ==================================================================
[ 2747.876949] BUG: KASAN: use-after-free in xfs_destroy_ioend+0x3d3/0x4e0 [xfs] at addr ffff8801395ae698
...
[ 2747.918245] Call Trace:
[ 2747.920975]  dump_stack+0x63/0x84
[ 2747.924673]  kasan_object_err+0x21/0x70
[ 2747.928950]  kasan_report+0x271/0x530
[ 2747.933064]  ? xfs_destroy_ioend+0x3d3/0x4e0 [xfs]
[ 2747.938409]  ? end_page_writeback+0xce/0x110
[ 2747.943171]  __asan_report_load8_noabort+0x19/0x20
[ 2747.948545]  xfs_destroy_ioend+0x3d3/0x4e0 [xfs]
[ 2747.953724]  xfs_end_io+0x1af/0x2b0 [xfs]
[ 2747.958197]  process_one_work+0x5ff/0x1000
[ 2747.962766]  worker_thread+0xe4/0x10e0
[ 2747.966946]  kthread+0x2d3/0x3d0
[ 2747.970546]  ? process_one_work+0x1000/0x1000
[ 2747.975405]  ? kthread_create_on_node+0xc0/0xc0
[ 2747.980457]  ? syscall_return_slowpath+0xe6/0x140
[ 2747.985706]  ? do_page_fault+0x30/0x80
[ 2747.989887]  ret_from_fork+0x2c/0x40
[ 2747.993874] Object at ffff8801395ae690, in cache buffer_head size: 104
[ 2748.001155] Allocated:
[ 2748.003782] PID = 8327
[ 2748.006411]  save_stack_trace+0x1b/0x20
[ 2748.010688]  save_stack+0x46/0xd0
[ 2748.014383]  kasan_kmalloc+0xad/0xe0
[ 2748.018370]  kasan_slab_alloc+0x12/0x20
[ 2748.022648]  kmem_cache_alloc+0xb8/0x1b0
[ 2748.027024]  alloc_buffer_head+0x22/0xc0
[ 2748.031399]  alloc_page_buffers+0xd1/0x250
[ 2748.035968]  create_empty_buffers+0x30/0x410
[ 2748.040730]  create_page_buffers+0x120/0x1b0
[ 2748.045493]  __block_write_begin_int+0x17a/0x1800
[ 2748.050740]  iomap_write_begin+0x100/0x2f0
[ 2748.055308]  iomap_zero_range_actor+0x253/0x5c0
[ 2748.060362]  iomap_apply+0x157/0x270
[ 2748.064347]  iomap_zero_range+0x5a/0x80
[ 2748.068624]  iomap_truncate_page+0x6b/0xa0
[ 2748.073227]  xfs_setattr_size+0x1f7/0xa10 [xfs]
[ 2748.078312]  xfs_vn_setattr_size+0x68/0x140 [xfs]
[ 2748.083589]  xfs_file_fallocate+0x4ac/0x820 [xfs]
[ 2748.088838]  vfs_fallocate+0x2cf/0x780
[ 2748.093021]  SyS_fallocate+0x48/0x80
[ 2748.097006]  do_syscall_64+0x18a/0x430
[ 2748.101186]  return_from_SYSCALL_64+0x0/0x6a
[ 2748.105948] Freed:
[ 2748.108189] PID = 8327
[ 2748.110816]  save_stack_trace+0x1b/0x20
[ 2748.115093]  save_stack+0x46/0xd0
[ 2748.118788]  kasan_slab_free+0x73/0xc0
[ 2748.122969]  kmem_cache_free+0x7a/0x200
[ 2748.127247]  free_buffer_head+0x41/0x80
[ 2748.131524]  try_to_free_buffers+0x178/0x250
[ 2748.136316]  xfs_vm_releasepage+0x2e9/0x3d0 [xfs]
[ 2748.141563]  try_to_release_page+0x100/0x180
[ 2748.146325]  invalidate_inode_pages2_range+0x7da/0xcf0
[ 2748.152087]  xfs_shift_file_space+0x37d/0x6e0 [xfs]
[ 2748.157557]  xfs_collapse_file_space+0x49/0x120 [xfs]
[ 2748.163223]  xfs_file_fallocate+0x2a7/0x820 [xfs]
[ 2748.168462]  vfs_fallocate+0x2cf/0x780
[ 2748.172642]  SyS_fallocate+0x48/0x80
[ 2748.176629]  do_syscall_64+0x18a/0x430
[ 2748.180810]  return_from_SYSCALL_64+0x0/0x6a

Fixed it by checking on offset against end & breaking out first,
dereference bh only if there're still bufferheads to process.

Signed-off-by: Eryu Guan <eguan@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_aops.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/fs/xfs/xfs_aops.c b/fs/xfs/xfs_aops.c
index 0457abe4118a..6df0a7ce3e8a 100644
--- a/fs/xfs/xfs_aops.c
+++ b/fs/xfs/xfs_aops.c
@@ -116,11 +116,11 @@ xfs_finish_page_writeback(
 
 	bsize = bh->b_size;
 	do {
+		if (off > end)
+			break;
 		next = bh->b_this_page;
 		if (off < bvec->bv_offset)
 			goto next_bh;
-		if (off > end)
-			break;
 		bh->b_end_io(bh, !error);
 next_bh:
 		off += bsize;
-- 
2.11.0

[PATCH 17/25] xfs: fix indlen accounting error on partial delalloc conversion

[Christoph Hellwig]

From: Brian Foster <bfoster@redhat.com>

commit 0daaecacb83bc6b656a56393ab77a31c28139bc7 upstream.

The delalloc -> real block conversion path uses an incorrect
calculation in the case where the middle part of a delalloc extent
is being converted. This is documented as a rare situation because
XFS generally attempts to maximize contiguity by converting as much
of a delalloc extent as possible.

If this situation does occur, the indlen reservation for the two new
delalloc extents left behind by the conversion of the middle range
is calculated and compared with the original reservation. If more
blocks are required, the delta is allocated from the global block
pool. This delta value can be characterized as the difference
between the new total requirement (temp + temp2) and the currently
available reservation minus those blocks that have already been
allocated (startblockval(PREV.br_startblock) - allocated).

The problem is that the current code does not account for previously
allocated blocks correctly. It subtracts the current allocation
count from the (new - old) delta rather than the old indlen
reservation. This means that more indlen blocks than have been
allocated end up stashed in the remaining extents and free space
accounting is broken as a result.

Fix up the calculation to subtract the allocated block count from
the original extent indlen and thus correctly allocate the
reservation delta based on the difference between the new total
requirement and the unused blocks from the original reservation.
Also remove a bogus assert that contradicts the fact that the new
indlen reservation can be larger than the original indlen
reservation.

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/libxfs/xfs_bmap.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/fs/xfs/libxfs/xfs_bmap.c b/fs/xfs/libxfs/xfs_bmap.c
index 23f2b4b36f38..2a8cbd15d5d1 100644
--- a/fs/xfs/libxfs/xfs_bmap.c
+++ b/fs/xfs/libxfs/xfs_bmap.c
@@ -2208,8 +2208,10 @@ xfs_bmap_add_extent_delay_real(
 		}
 		temp = xfs_bmap_worst_indlen(bma->ip, temp);
 		temp2 = xfs_bmap_worst_indlen(bma->ip, temp2);
-		diff = (int)(temp + temp2 - startblockval(PREV.br_startblock) -
-			(bma->cur ? bma->cur->bc_private.b.allocated : 0));
+		diff = (int)(temp + temp2 -
+			     (startblockval(PREV.br_startblock) -
+			      (bma->cur ?
+			       bma->cur->bc_private.b.allocated : 0)));
 		if (diff > 0) {
 			error = xfs_mod_fdblocks(bma->ip->i_mount,
 						 -((int64_t)diff), false);
@@ -2266,7 +2268,6 @@ xfs_bmap_add_extent_delay_real(
 		temp = da_new;
 		if (bma->cur)
 			temp += bma->cur->bc_private.b.allocated;
-		ASSERT(temp <= da_old);
 		if (temp < da_old)
 			xfs_mod_fdblocks(bma->ip->i_mount,
 					(int64_t)(da_old - temp), false);
-- 
2.11.0

[PATCH 18/25] xfs: BMAPX shouldn't barf on inline-format directories

[Christoph Hellwig]

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

commit 6eadbf4c8ba816c10d1c97bed9aa861d9fd17809 upstream.

When we're fulfilling a BMAPX request, jump out early if the data fork
is in local format.  This prevents us from hitting a debugging check in
bmapi_read and barfing errors back to userspace.  The on-disk extent
count check later isn't sufficient for IF_DELALLOC mode because da
extents are in memory and not on disk.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/xfs/xfs_bmap_util.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/fs/xfs/xfs_bmap_util.c b/fs/xfs/xfs_bmap_util.c
index 8a1a62ee84ad..15be2c973c40 100644
--- a/fs/xfs/xfs_bmap_util.c
+++ b/fs/xfs/xfs_bmap_util.c
@@ -588,9 +588,13 @@ xfs_getbmap(
 		}
 		break;
 	default:
+		/* Local format data forks report no extents. */
+		if (ip->i_d.di_format == XFS_DINODE_FMT_LOCAL) {
+			bmv->bmv_entries = 0;
+			return 0;
+		}
 		if (ip->i_d.di_format != XFS_DINODE_FMT_EXTENTS &&
-		    ip->i_d.di_format != XFS_DINODE_FMT_BTREE &&
-		    ip->i_d.di_format != XFS_DINODE_FMT_LOCAL)
+		    ip->i_d.di_format != XFS_DINODE_FMT_BTREE)
 			return -EINVAL;
 
 		if (xfs_get_extsz_hint(ip) ||
-- 
2.11.0

[PATCH 19/25] xfs: bad assertion for delalloc an extent that start at i_size

[Christoph Hellwig]

From: Zorro Lang <zlang@redhat.com>

commit 892d2a5f705723b2cb488bfb38bcbdcf83273184 upstream.

By run fsstress long enough time enough in RHEL-7, I find an
assertion failure (harder to reproduce on linux-4.11, but problem
is still there):

  XFS: Assertion failed: (iflags & BMV_IF_DELALLOC) != 0, file: fs/xfs/xfs_bmap_util.c

The assertion is in xfs_getbmap() funciton:

  if (map[i].br_startblock == DELAYSTARTBLOCK &&
-->   map[i].br_startoff <= XFS_B_TO_FSB(mp, XFS_ISIZE(ip)))
          ASSERT((iflags & BMV_IF_DELALLOC) != 0);

When map[i].br_startoff == XFS_B_TO_FSB(mp, XFS_ISIZE(ip)), the
startoff is just at EOF. But we only need to make sure delalloc
extents that are within EOF, not include EOF.

Signed-off-by: Zorro Lang <zlang@redhat.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_bmap_util.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/xfs/xfs_bmap_util.c b/fs/xfs/xfs_bmap_util.c
index 15be2c973c40..87b495e2f15a 100644
--- a/fs/xfs/xfs_bmap_util.c
+++ b/fs/xfs/xfs_bmap_util.c
@@ -722,7 +722,7 @@ xfs_getbmap(
 			 * extents.
 			 */
 			if (map[i].br_startblock == DELAYSTARTBLOCK &&
-			    map[i].br_startoff <= XFS_B_TO_FSB(mp, XFS_ISIZE(ip)))
+			    map[i].br_startoff < XFS_B_TO_FSB(mp, XFS_ISIZE(ip)))
 				ASSERT((iflags & BMV_IF_DELALLOC) != 0);
 
                         if (map[i].br_startblock == HOLESTARTBLOCK &&
-- 
2.11.0

[PATCH 20/25] xfs: xfs_trans_alloc_empty

[Christoph Hellwig]

This is a partial cherry-pick of commit e89c041338
("xfs: implement the GETFSMAP ioctl"), which also adds this helper, and
a great example of why feature patches should be properly split into
their parts.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
[hch: split from the larger patch for -stable]
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 fs/xfs/xfs_trans.c | 22 ++++++++++++++++++++++
 fs/xfs/xfs_trans.h |  2 ++
 2 files changed, 24 insertions(+)

diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c
index 70f42ea86dfb..a280e126491f 100644
--- a/fs/xfs/xfs_trans.c
+++ b/fs/xfs/xfs_trans.c
@@ -263,6 +263,28 @@ xfs_trans_alloc(
 }
 
 /*
+ * Create an empty transaction with no reservation.  This is a defensive
+ * mechanism for routines that query metadata without actually modifying
+ * them -- if the metadata being queried is somehow cross-linked (think a
+ * btree block pointer that points higher in the tree), we risk deadlock.
+ * However, blocks grabbed as part of a transaction can be re-grabbed.
+ * The verifiers will notice the corrupt block and the operation will fail
+ * back to userspace without deadlocking.
+ *
+ * Note the zero-length reservation; this transaction MUST be cancelled
+ * without any dirty data.
+ */
+int
+xfs_trans_alloc_empty(
+	struct xfs_mount		*mp,
+	struct xfs_trans		**tpp)
+{
+	struct xfs_trans_res		resv = {0};
+
+	return xfs_trans_alloc(mp, &resv, 0, 0, XFS_TRANS_NO_WRITECOUNT, tpp);
+}
+
+/*
  * Record the indicated change to the given field for application
  * to the file system's superblock when the transaction commits.
  * For now, just store the change in the transaction structure.
diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h
index 61b7fbdd3ebd..98024cb933ef 100644
--- a/fs/xfs/xfs_trans.h
+++ b/fs/xfs/xfs_trans.h
@@ -159,6 +159,8 @@ typedef struct xfs_trans {
 int		xfs_trans_alloc(struct xfs_mount *mp, struct xfs_trans_res *resp,
 			uint blocks, uint rtextents, uint flags,
 			struct xfs_trans **tpp);
+int		xfs_trans_alloc_empty(struct xfs_mount *mp,
+			struct xfs_trans **tpp);
 void		xfs_trans_mod_sb(xfs_trans_t *, uint, int64_t);
 
 struct xfs_buf	*xfs_trans_get_buf_map(struct xfs_trans *tp,
-- 
2.11.0

[PATCH 21/25] xfs: avoid mount-time deadlock in CoW extent recovery

[Christoph Hellwig]

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

commit 3ecb3ac7b950ff8f6c6a61e8b7b0d6e3546429a0 upstream.

If a malicious user corrupts the refcount btree to cause a cycle between
different levels of the tree, the next mount attempt will deadlock in
the CoW recovery routine while grabbing buffer locks.  We can use the
ability to re-grab a buffer that was previous locked to a transaction to
avoid deadlocks, so do that here.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
---
 fs/xfs/libxfs/xfs_refcount.c | 43 +++++++++++++++++++++++++++++++------------
 1 file changed, 31 insertions(+), 12 deletions(-)

diff --git a/fs/xfs/libxfs/xfs_refcount.c b/fs/xfs/libxfs/xfs_refcount.c
index b177ef33cd4c..82a38d86ebad 100644
--- a/fs/xfs/libxfs/xfs_refcount.c
+++ b/fs/xfs/libxfs/xfs_refcount.c
@@ -1629,13 +1629,28 @@ xfs_refcount_recover_cow_leftovers(
 	if (mp->m_sb.sb_agblocks >= XFS_REFC_COW_START)
 		return -EOPNOTSUPP;
 
-	error = xfs_alloc_read_agf(mp, NULL, agno, 0, &agbp);
+	INIT_LIST_HEAD(&debris);
+
+	/*
+	 * In this first part, we use an empty transaction to gather up
+	 * all the leftover CoW extents so that we can subsequently
+	 * delete them.  The empty transaction is used to avoid
+	 * a buffer lock deadlock if there happens to be a loop in the
+	 * refcountbt because we're allowed to re-grab a buffer that is
+	 * already attached to our transaction.  When we're done
+	 * recording the CoW debris we cancel the (empty) transaction
+	 * and everything goes away cleanly.
+	 */
+	error = xfs_trans_alloc_empty(mp, &tp);
 	if (error)
 		return error;
-	cur = xfs_refcountbt_init_cursor(mp, NULL, agbp, agno, NULL);
+
+	error = xfs_alloc_read_agf(mp, tp, agno, 0, &agbp);
+	if (error)
+		goto out_trans;
+	cur = xfs_refcountbt_init_cursor(mp, tp, agbp, agno, NULL);
 
 	/* Find all the leftover CoW staging extents. */
-	INIT_LIST_HEAD(&debris);
 	memset(&low, 0, sizeof(low));
 	memset(&high, 0, sizeof(high));
 	low.rc.rc_startblock = XFS_REFC_COW_START;
@@ -1645,10 +1660,11 @@ xfs_refcount_recover_cow_leftovers(
 	if (error)
 		goto out_cursor;
 	xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR);
-	xfs_buf_relse(agbp);
+	xfs_trans_brelse(tp, agbp);
+	xfs_trans_cancel(tp);
 
 	/* Now iterate the list to free the leftovers */
-	list_for_each_entry(rr, &debris, rr_list) {
+	list_for_each_entry_safe(rr, n, &debris, rr_list) {
 		/* Set up transaction. */
 		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_write, 0, 0, 0, &tp);
 		if (error)
@@ -1676,8 +1692,16 @@ xfs_refcount_recover_cow_leftovers(
 		error = xfs_trans_commit(tp);
 		if (error)
 			goto out_free;
+
+		list_del(&rr->rr_list);
+		kmem_free(rr);
 	}
 
+	return error;
+out_defer:
+	xfs_defer_cancel(&dfops);
+out_trans:
+	xfs_trans_cancel(tp);
 out_free:
 	/* Free the leftover list */
 	list_for_each_entry_safe(rr, n, &debris, rr_list) {
@@ -1688,11 +1712,6 @@ xfs_refcount_recover_cow_leftovers(
 
 out_cursor:
 	xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
-	xfs_buf_relse(agbp);
-	goto out_free;
-
-out_defer:
-	xfs_defer_cancel(&dfops);
-	xfs_trans_cancel(tp);
-	goto out_free;
+	xfs_trans_brelse(tp, agbp);
+	goto out_trans;
 }
-- 
2.11.0

[PATCH 22/25] xfs: fix unaligned access in xfs_btree_visit_blocks

[Christoph Hellwig]

From: Eric Sandeen <sandeen@sandeen.net>

commit a4d768e702de224cc85e0c8eac9311763403b368 upstream.

This structure copy was throwing unaligned access warnings on sparc64:

Kernel unaligned access at TPC[1043c088] xfs_btree_visit_blocks+0x88/0xe0 [xfs]

xfs_btree_copy_ptrs does a memcpy, which avoids it.

Signed-off-by: Eric Sandeen <sandeen@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_btree.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/xfs/libxfs/xfs_btree.c b/fs/xfs/libxfs/xfs_btree.c
index 2849d3fa3d0b..91c68913d495 100644
--- a/fs/xfs/libxfs/xfs_btree.c
+++ b/fs/xfs/libxfs/xfs_btree.c
@@ -4376,7 +4376,7 @@ xfs_btree_visit_blocks(
 			xfs_btree_readahead_ptr(cur, ptr, 1);
 
 			/* save for the next iteration of the loop */
-			lptr = *ptr;
+			xfs_btree_copy_ptrs(cur, &lptr, ptr, 1);
 		}
 
 		/* for each buffer in the level */
-- 
2.11.0

[PATCH 23/25] xfs: fix off-by-one on max nr_pages in xfs_find_get_desired_pgoff()

[Christoph Hellwig]

From: Eryu Guan <eguan@redhat.com>

commit 8affebe16d79ebefb1d9d6d56a46dc89716f9453 upstream.

xfs_find_get_desired_pgoff() is used to search for offset of hole or
data in page range [index, end] (both inclusive), and the max number
of pages to search should be at least one, if end == index.
Otherwise the only page is missed and no hole or data is found,
which is not correct.

When block size is smaller than page size, this can be demonstrated
by preallocating a file with size smaller than page size and writing
data to the last block. E.g. run this xfs_io command on a 1k block
size XFS on x86_64 host.

  # xfs_io -fc "falloc 0 3k" -c "pwrite 2k 1k" \
  	    -c "seek -d 0" /mnt/xfs/testfile
  wrote 1024/1024 bytes at offset 2048
  1 KiB, 1 ops; 0.0000 sec (33.675 MiB/sec and 34482.7586 ops/sec)
  Whence  Result
  DATA    EOF

Data at offset 2k was missed, and lseek(2) returned ENXIO.

This is uncovered by generic/285 subtest 07 and 08 on ppc64 host,
where pagesize is 64k. Because a recent change to generic/285
reduced the preallocated file size to smaller than 64k.

Signed-off-by: Eryu Guan <eguan@redhat.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/xfs_file.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c
index 1209ad29e902..a4efc61fa775 100644
--- a/fs/xfs/xfs_file.c
+++ b/fs/xfs/xfs_file.c
@@ -1136,7 +1136,7 @@ xfs_find_get_desired_pgoff(
 		unsigned	nr_pages;
 		unsigned int	i;
 
-		want = min_t(pgoff_t, end - index, PAGEVEC_SIZE);
+		want = min_t(pgoff_t, end - index, PAGEVEC_SIZE - 1) + 1;
 		nr_pages = pagevec_lookup(&pvec, inode->i_mapping, index,
 					  want);
 		/*
-- 
2.11.0

[PATCH 24/25] xfs: Fix missed holes in SEEK_HOLE implementation

[Christoph Hellwig]

From: Jan Kara <jack@suse.cz>

commit 5375023ae1266553a7baa0845e82917d8803f48c upstream.

XFS SEEK_HOLE implementation could miss a hole in an unwritten extent as
can be seen by the following command:

xfs_io -c "falloc 0 256k" -c "pwrite 0 56k" -c "pwrite 128k 8k"
       -c "seek -h 0" file
wrote 57344/57344 bytes at offset 0
56 KiB, 14 ops; 0.0000 sec (49.312 MiB/sec and 12623.9856 ops/sec)
wrote 8192/8192 bytes at offset 131072
8 KiB, 2 ops; 0.0000 sec (70.383 MiB/sec and 18018.0180 ops/sec)
Whence	Result
HOLE	139264

Where we can see that hole at offset 56k was just ignored by SEEK_HOLE
implementation. The bug is in xfs_find_get_desired_pgoff() which does
not properly detect the case when pages are not contiguous.

Fix the problem by properly detecting when found page has larger offset
than expected.

Fixes: d126d43f631f996daeee5006714fed914be32368
Signed-off-by: Jan Kara <jack@suse.cz>
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_file.c | 29 +++++++++--------------------
 1 file changed, 9 insertions(+), 20 deletions(-)

diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c
index a4efc61fa775..9292a59efbfa 100644
--- a/fs/xfs/xfs_file.c
+++ b/fs/xfs/xfs_file.c
@@ -1163,17 +1163,6 @@ xfs_find_get_desired_pgoff(
 			break;
 		}
 
-		/*
-		 * At lease we found one page.  If this is the first time we
-		 * step into the loop, and if the first page index offset is
-		 * greater than the given search offset, a hole was found.
-		 */
-		if (type == HOLE_OFF && lastoff == startoff &&
-		    lastoff < page_offset(pvec.pages[0])) {
-			found = true;
-			break;
-		}
-
 		for (i = 0; i < nr_pages; i++) {
 			struct page	*page = pvec.pages[i];
 			loff_t		b_offset;
@@ -1185,18 +1174,18 @@ xfs_find_get_desired_pgoff(
 			 * file mapping. However, page->index will not change
 			 * because we have a reference on the page.
 			 *
-			 * Searching done if the page index is out of range.
-			 * If the current offset is not reaches the end of
-			 * the specified search range, there should be a hole
-			 * between them.
+			 * If current page offset is beyond where we've ended,
+			 * we've found a hole.
 			 */
-			if (page->index > end) {
-				if (type == HOLE_OFF && lastoff < endoff) {
-					*offset = lastoff;
-					found = true;
-				}
+			if (type == HOLE_OFF && lastoff < endoff &&
+			    lastoff < page_offset(pvec.pages[i])) {
+				found = true;
+				*offset = lastoff;
 				goto out;
 			}
+			/* Searching done if the page index is out of range. */
+			if (page->index > end)
+				goto out;
 
 			lock_page(page);
 			/*
-- 
2.11.0

[PATCH 25/25] xfs: Fix off-by-in in loop termination in xfs_find_get_desired_pgoff()

[Christoph Hellwig]

From: Jan Kara <jack@suse.cz>

commit d7fd24257aa60316bf81093f7f909dc9475ae974 upstream.

There is an off-by-one error in loop termination conditions in
xfs_find_get_desired_pgoff() since 'end' may index a page beyond end of
desired range if 'endoff' is page aligned. It doesn't have any visible
effects but still it is good to fix it.

Signed-off-by: Jan Kara <jack@suse.cz>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/xfs_file.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c
index 9292a59efbfa..a90ec3fad69f 100644
--- a/fs/xfs/xfs_file.c
+++ b/fs/xfs/xfs_file.c
@@ -1130,7 +1130,7 @@ xfs_find_get_desired_pgoff(
 
 	index = startoff >> PAGE_SHIFT;
 	endoff = XFS_FSB_TO_B(mp, map->br_startoff + map->br_blockcount);
-	end = endoff >> PAGE_SHIFT;
+	end = (endoff - 1) >> PAGE_SHIFT;
 	do {
 		int		want;
 		unsigned	nr_pages;
-- 
2.11.0

4.9-stable updates for XFS

[Christoph Hellwig]

Hi all,

below are backports up to 4.14-rc5 for XFS.

Re: 4.9-stable updates for XFS

[Greg KH]

On Thu, Oct 19, 2017 at 04:22:43PM +0200, Christoph Hellwig wrote:
> Hi all,
> 
> below are backports up to 4.14-rc5 for XFS.

All now applied, thanks.

greg k-h

4.9-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.

In addition to the patches for 4.13-stable there also are various
patches that are in 4.13 but not in 4.9-stable yet.

Re: 4.9-stable updates for XFS

[Greg KH]

On Sun, Sep 17, 2017 at 02:06:25PM -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.
> 
> In addition to the patches for 4.13-stable there also are various
> patches that are in 4.13 but not in 4.9-stable yet.

Thanks for all of these, now queued up.

greg k-h

4.9-stable updates for XFS

[Christoph Hellwig]

Hi all,

below are backports up to 4.11-rc4 for XFS.  I've selected all clear
bugfixes for reported bugs.  In addition to the patches in 4.10-stable
there also are 2 patches that are in 4.10 but not in 4.9-stable yet
at the beginning of the series.

Re: 4.9-stable updates for XFS

[Greg KH]

On Sat, Apr 01, 2017 at 08:39:59AM +0200, Christoph Hellwig wrote:
> Hi all,
> 
> below are backports up to 4.11-rc4 for XFS.  I've selected all clear
> bugfixes for reported bugs.  In addition to the patches in 4.10-stable
> there also are 2 patches that are in 4.10 but not in 4.9-stable yet
> at the beginning of the series.

All now applied, thanks for the series.

greg k-h

4.9-stable updates for XFS

[Christoph Hellwig]

Hi all,

below are backports up to 4.10-rc6 for XFS.  I've selected all clear
bugfixes for reported bugs.  I've skipped the finobt space reservation
patch for now - I think it is stable material, but I want it to settle
a bit more first.

Re: 4.9-stable updates for XFS

[Greg KH]

On Thu, Feb 02, 2017 at 08:55:52AM +0100, Christoph Hellwig wrote:
> Hi all,
> 
> below are backports up to 4.10-rc6 for XFS.  I've selected all clear
> bugfixes for reported bugs.  I've skipped the finobt space reservation
> patch for now - I think it is stable material, but I want it to settle
> a bit more first.

Thanks for all of these, now all applied.

greg k-h

Re: 4.9-stable updates for XFS

[Eric Sandeen]

On 2/2/17 1:55 AM, Christoph Hellwig wrote:
> Hi all,
> 
> below are backports up to 4.10-rc6 for XFS.  I've selected all clear
> bugfixes for reported bugs.  I've skipped the finobt space reservation
> patch for now - I think it is stable material, but I want it to settle
> a bit more first.

fwiw, the finobt space reservation thing also seems to be doing something
odd with allocation on striped storage, unless I'm missing something.

If I stick an xfs_bmap -v into generic/108 after we write the initial
file, we find that it's not stripe aligned with that patch in place,
which seems unexpected.  I haven't yet worked out why, was first trying
to sort out that lvm error propagation problem...

-Eric

Re: 4.9-stable updates for XFS

[Eric Sandeen]

On 2/2/17 10:00 AM, Eric Sandeen wrote:
> On 2/2/17 1:55 AM, Christoph Hellwig wrote:
>> Hi all,
>>
>> below are backports up to 4.10-rc6 for XFS.  I've selected all clear
>> bugfixes for reported bugs.  I've skipped the finobt space reservation
>> patch for now - I think it is stable material, but I want it to settle
>> a bit more first.
> 
> fwiw, the finobt space reservation thing also seems to be doing something
> odd with allocation on striped storage, unless I'm missing something.
> 
> If I stick an xfs_bmap -v into generic/108 after we write the initial
> file, we find that it's not stripe aligned with that patch in place,
> which seems unexpected.  I haven't yet worked out why, was first trying
> to sort out that lvm error propagation problem...

Ok, I guess this may not be a big deal.  Because this is a small filesystem
in the test, and striped, we end up with small AGs:

meta-data=/dev/mapper/vg_108-lv_108 isize=512    agcount=9, agsize=7168 blks
         =                       sectsz=512   attr=2, projid32bit=1
         =                       crc=1        finobt=1, sparse=0, rmapbt=0, reflink=0
data     =                       bsize=4096   blocks=63488, imaxpct=25
         =                       sunit=1024   swidth=2048 blks
naming   =version 2              bsize=4096   ascii-ci=0 ftype=1
log      =/dev/sdb3              bsize=4096   blocks=32768, version=2
         =                       sectsz=512   sunit=0 blks, lazy-count=1
realtime =none                   extsz=4096   blocks=0, rtextents=0

each only 7168 blocks, or about 28M.

We try to do a 16M write, and xfs would like that to be stripe aligned.
I'm not quite sure why it's failing, but for a larger filesystem things
do come out aligned again.  I'll keep digging but I don't think there's
any fundamental problem here, this is a weird, small filesystem.

-Eric

4.9-stable updates for XFS

[Christoph Hellwig]

Hi all,

below are backports from 4.10-rc3 for XFS.  I've selected all clear
bugfixes for reported bugs, and added about a handful of small
infrastructure backports so that we can take the fixes as is instead of
having to redo them in a different way than upstream.

These have passed xfstests on top of Linux 4.9.2.  Even if some of the 
patches might apply to older stable trees please do not backport them 
unless someone volunteers to QA them first.

Re: 4.9-stable updates for XFS

[Darrick J. Wong]

On Mon, Jan 09, 2017 at 04:38:31PM +0100, Christoph Hellwig wrote:
> Hi all,
> 
> below are backports from 4.10-rc3 for XFS.  I've selected all clear
> bugfixes for reported bugs, and added about a handful of small
> infrastructure backports so that we can take the fixes as is instead of
> having to redo them in a different way than upstream.
> 
> These have passed xfstests on top of Linux 4.9.2.  Even if some of the 
> patches might apply to older stable trees please do not backport them 
> unless someone volunteers to QA them first.

FWIW the list looks reasonable to me.

--D

> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: 4.9-stable updates for XFS

[Greg KH]

On Mon, Jan 09, 2017 at 04:38:31PM +0100, Christoph Hellwig wrote:
> Hi all,
> 
> below are backports from 4.10-rc3 for XFS.  I've selected all clear
> bugfixes for reported bugs, and added about a handful of small
> infrastructure backports so that we can take the fixes as is instead of
> having to redo them in a different way than upstream.
> 
> These have passed xfstests on top of Linux 4.9.2.  Even if some of the 
> patches might apply to older stable trees please do not backport them 
> unless someone volunteers to QA them first.

Many thanks for these, all now queued up.

greg k-h