Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.kernel > #1572650

[PATCH 4.9 46/51] xfs: fix COW writeback race

From Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Newsgroups linux.kernel
Subject [PATCH 4.9 46/51] xfs: fix COW writeback race
Date 2017-02-02 19:50 +0100
Message-ID <t6uCf-7FA-53@gated-at.bofh.it> (permalink)
References <t6usx-7Bs-3@gated-at.bofh.it>
Organization linux.* mail to news gateway

Show all headers | View raw


4.9-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Christoph Hellwig <hch@lst.de>

commit d2b3964a0780d2d2994eba57f950d6c9fe489ed8 upstream.

Due to the way how xfs_iomap_write_allocate tries to convert the whole
found extents from delalloc to real space we can run into a race
condition with multiple threads doing writes to this same extent.
For the non-COW case that is harmless as the only thing that can happen
is that we call xfs_bmapi_write on an extent that has already been
converted to a real allocation.  For COW writes where we move the extent
from the COW to the data fork after I/O completion the race is, however,
not quite as harmless.  In the worst case we are now calling
xfs_bmapi_write on a region that contains hole in the COW work, which
will trip up an assert in debug builds or lead to file system corruption
in non-debug builds.  This seems to be reproducible with workloads of
small O_DSYNC write, although so far I've not managed to come up with
a with an isolated reproducer.

The fix for the issue is relatively simple:  tell xfs_bmapi_write
that we are only asked to convert delayed allocations and skip holes
in that case.

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/libxfs/xfs_bmap.c |   44 ++++++++++++++++++++++++++++++++------------
 fs/xfs/libxfs/xfs_bmap.h |    6 +++++-
 fs/xfs/xfs_iomap.c       |    2 +-
 3 files changed, 38 insertions(+), 14 deletions(-)

--- a/fs/xfs/libxfs/xfs_bmap.c
+++ b/fs/xfs/libxfs/xfs_bmap.c
@@ -4607,8 +4607,6 @@ xfs_bmapi_write(
 	int			n;		/* current extent index */
 	xfs_fileoff_t		obno;		/* old block number (offset) */
 	int			whichfork;	/* data or attr fork */
-	char			inhole;		/* current location is hole in file */
-	char			wasdelay;	/* old extent was delayed */
 
 #ifdef DEBUG
 	xfs_fileoff_t		orig_bno;	/* original block number value */
@@ -4694,22 +4692,44 @@ xfs_bmapi_write(
 	bma.firstblock = firstblock;
 
 	while (bno < end && n < *nmap) {
-		inhole = eof || bma.got.br_startoff > bno;
-		wasdelay = !inhole && isnullstartblock(bma.got.br_startblock);
+		bool			need_alloc = false, wasdelay = false;
 
-		/*
-		 * Make sure we only reflink into a hole.
-		 */
-		if (flags & XFS_BMAPI_REMAP)
-			ASSERT(inhole);
-		if (flags & XFS_BMAPI_COWFORK)
-			ASSERT(!inhole);
+		/* in hole or beyoned EOF? */
+		if (eof || bma.got.br_startoff > bno) {
+			if (flags & XFS_BMAPI_DELALLOC) {
+				/*
+				 * For the COW fork we can reasonably get a
+				 * request for converting an extent that races
+				 * with other threads already having converted
+				 * part of it, as there converting COW to
+				 * regular blocks is not protected using the
+				 * IOLOCK.
+				 */
+				ASSERT(flags & XFS_BMAPI_COWFORK);
+				if (!(flags & XFS_BMAPI_COWFORK)) {
+					error = -EIO;
+					goto error0;
+				}
+
+				if (eof || bno >= end)
+					break;
+			} else {
+				need_alloc = true;
+			}
+		} else {
+			/*
+			 * Make sure we only reflink into a hole.
+			 */
+			ASSERT(!(flags & XFS_BMAPI_REMAP));
+			if (isnullstartblock(bma.got.br_startblock))
+				wasdelay = true;
+		}
 
 		/*
 		 * First, deal with the hole before the allocated space
 		 * that we found, if any.
 		 */
-		if (inhole || wasdelay) {
+		if (need_alloc || wasdelay) {
 			bma.eof = eof;
 			bma.conv = !!(flags & XFS_BMAPI_CONVERT);
 			bma.wasdel = wasdelay;
--- a/fs/xfs/libxfs/xfs_bmap.h
+++ b/fs/xfs/libxfs/xfs_bmap.h
@@ -110,6 +110,9 @@ struct xfs_extent_free_item
 /* Map something in the CoW fork. */
 #define XFS_BMAPI_COWFORK	0x200
 
+/* Only convert delalloc space, don't allocate entirely new extents */
+#define XFS_BMAPI_DELALLOC	0x400
+
 #define XFS_BMAPI_FLAGS \
 	{ XFS_BMAPI_ENTIRE,	"ENTIRE" }, \
 	{ XFS_BMAPI_METADATA,	"METADATA" }, \
@@ -120,7 +123,8 @@ struct xfs_extent_free_item
 	{ XFS_BMAPI_CONVERT,	"CONVERT" }, \
 	{ XFS_BMAPI_ZERO,	"ZERO" }, \
 	{ XFS_BMAPI_REMAP,	"REMAP" }, \
-	{ XFS_BMAPI_COWFORK,	"COWFORK" }
+	{ XFS_BMAPI_COWFORK,	"COWFORK" }, \
+	{ XFS_BMAPI_DELALLOC,	"DELALLOC" }
 
 
 static inline int xfs_bmapi_aflag(int w)
--- a/fs/xfs/xfs_iomap.c
+++ b/fs/xfs/xfs_iomap.c
@@ -681,7 +681,7 @@ xfs_iomap_write_allocate(
 	xfs_trans_t	*tp;
 	int		nimaps;
 	int		error = 0;
-	int		flags = 0;
+	int		flags = XFS_BMAPI_DELALLOC;
 	int		nres;
 
 	if (whichfork == XFS_COW_FORK)

Back to linux.kernel | Previous | NextPrevious in thread | Next in thread | Find similar | Unroll thread


Thread

[PATCH 4.9 00/51] 4.9.8-stable review Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:40 +0100
  [PATCH 4.9 35/51] xfs: adjust allocation length in xfs_alloc_space_available Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:40 +0100
  [PATCH 4.9 36/51] xfs: dont rely on ->total in xfs_alloc_space_available Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:40 +0100
  [PATCH 4.9 09/51] tcp: fix tcp_fastopen unaligned access complaints on sparc Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:40 +0100
  [PATCH 4.9 05/51] mlxsw: pci: Fix EQE structure definition Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:40 +0100
  [PATCH 4.9 04/51] mlxsw: switchx2: Fix memory leak at skb reallocation Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:40 +0100
  [PATCH 4.9 03/51] mlxsw: spectrum: Fix memory leak at skb reallocation Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:40 +0100
  [PATCH 4.9 26/51] net: mpls: Fix multipath selection for LSR use case Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:40 +0100
  [PATCH 4.9 11/51] mlx4: do not call napi_schedule() without care Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 14/51] ax25: Fix segfault after sock connection timeout Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 47/51] xfs: verify dirblocklog correctly Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 43/51] xfs: sanity check inode di_mode Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 39/51] xfs: sanity check directory inode di_size Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 15/51] net sched actions: fix refcnt when GETing of action after bind Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 50/51] xfs: clear _XBF_PAGES from buffers when readahead page Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 22/51] ipv6: addrconf: Avoid addrconf_disable_change() using RCU read-side lock Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 31/51] lwtunnel: Fix oops on state free after encap module unload Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 23/51] tcp: initialize max window for a new fastopen socket Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 49/51] xfs: extsize hints are not unlikely in xfs_bmap_btalloc Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 24/51] net/mlx5e: Do not recycle pages from emergency reserve Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 08/51] net: systemport: Decouple flow control from __bcm_sysport_tx_reclaim Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 30/51] net: Specify the owning module for lwtunnel ops Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 20/51] net: phy: bcm63xx: Utilize correct config_intr function Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 27/51] r8152: dont execute runtime suspend if the tx is not empty Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 46/51] xfs: fix COW writeback race Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 44/51] xfs: dont wrap ID in xfs_dq_get_next_id Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 06/51] net: lwtunnel: Handle lwtunnel_fill_encap failure Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 42/51] xfs: sanity check inode mode when creating new dentry Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 32/51] net: dsa: Bring back device detaching in dsa_slave_suspend() Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 41/51] xfs: replace xfs_mode_to_ftype table with switch statement Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 10/51] openvswitch: maintain correct checksum state in conntrack actions Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 13/51] ip6_tunnel: Account for tunnel header in tunnel MTU Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 45/51] xfs: fix xfs_mode_to_ftype() prototype Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 28/51] af_unix: move unix_mknod() out of bindlock Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 25/51] bridge: netlink: call br_changelink() during br_dev_newlink() Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 34/51] xfs: fix bogus minleft manipulations Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 37/51] xfs: dont print warnings when xfs_log_force fails Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 48/51] xfs: remove racy hasattr check from attr ops Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 12/51] ravb: do not use zero-length alignment DMA descriptor Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 51/51] xfs: fix bmv_count confusion w/ shared extents Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 33/51] xfs: bump up reserved blocks in xfs_alloc_set_aside Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 19:50 +0100
  [PATCH 4.9 02/51] netvsc: add rcu_read locking to netvsc callback Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 20:00 +0100
  [PATCH 4.9 17/51] virtio-net: restore VIRTIO_HDR_F_DATA_VALID on receiving Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 20:00 +0100
  [PATCH 4.9 07/51] net: ipv4: fix table id in getroute response Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 20:00 +0100
  [PATCH 4.9 01/51] r8152: fix the sw rx checksum is unavailable Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 20:00 +0100
  Re: [PATCH 4.9 00/51] 4.9.8-stable review Shuah Khan <shuahkh@osg.samsung.com> - 2017-02-02 21:40 +0100
    Re: [PATCH 4.9 00/51] 4.9.8-stable review Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-02 22:00 +0100
  Re: [PATCH 4.9 00/51] 4.9.8-stable review Guenter Roeck <linux@roeck-us.net> - 2017-02-03 06:20 +0100
    Re: [PATCH 4.9 00/51] 4.9.8-stable review Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-02-03 08:20 +0100

csiph-web