[PATCH 3.13.y-ckt 42/60] xfs: remote attributes need to be considered data

From	Kamal Mostafa <kamal@canonical.com>
Newsgroups	linux.kernel
Subject	[PATCH 3.13.y-ckt 42/60] xfs: remote attributes need to be considered data
Date	2015-09-02 03:10 +0200
Message-ID	<q44IP-6Dz-45@gated-at.bofh.it> (permalink)
References	<q44z8-6cH-7@gated-at.bofh.it>
Organization	linux.* mail to news gateway

Show all headers | View raw

3.13.11-ckt26 -stable review patch.  If anyone has any objections, please let me know.

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

From: Dave Chinner <dchinner@redhat.com>

commit df150ed102baa0e78c06e08e975dfb47147dd677 upstream.

We don't log remote attribute contents, and instead write them
synchronously before we commit the block allocation and attribute
tree update transaction. As a result we are writing to the allocated
space before the allcoation has been made permanent.

As a result, we cannot consider this allocation to be a metadata
allocation. Metadata allocation can take blocks from the free list
and so reuse them before the transaction that freed the block is
committed to disk. This behaviour is perfectly fine for journalled
metadata changes as log recovery will ensure the free operation is
replayed before the overwrite, but for remote attribute writes this
is not the case.

Hence we have to consider the remote attribute blocks to contain
data and allocate accordingly. We do this by dropping the
XFS_BMAPI_METADATA flag from the block allocation. This means the
allocation will not use blocks that are on the busy list without
first ensuring that the freeing transaction has been committed to
disk and the blocks removed from the busy list. This ensures we will
never overwrite a freed block without first ensuring that it is
really free.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
[ luis: backported to 3.16:
  - file rename: fs/xfs/libxfs/xfs_attr_remote.c -> fs/xfs/xfs_attr_remote.c ]
Signed-off-by: Luis Henriques <luis.henriques@canonical.com>
Signed-off-by: Kamal Mostafa <kamal@canonical.com>
---
 fs/xfs/xfs_attr_remote.c | 15 +++++++++++----
 1 file changed, 11 insertions(+), 4 deletions(-)

diff --git a/fs/xfs/xfs_attr_remote.c b/fs/xfs/xfs_attr_remote.c
index 5549d69..1aec8ad 100644
--- a/fs/xfs/xfs_attr_remote.c
+++ b/fs/xfs/xfs_attr_remote.c
@@ -435,14 +435,21 @@ xfs_attr_rmtval_set(
 
 		/*
 		 * Allocate a single extent, up to the size of the value.
+		 *
+		 * Note that we have to consider this a data allocation as we
+		 * write the remote attribute without logging the contents.
+		 * Hence we must ensure that we aren't using blocks that are on
+		 * the busy list so that we don't overwrite blocks which have
+		 * recently been freed but their transactions are not yet
+		 * committed to disk. If we overwrite the contents of a busy
+		 * extent and then crash then the block may not contain the
+		 * correct metadata after log recovery occurs.
 		 */
 		xfs_bmap_init(args->flist, args->firstblock);
 		nmap = 1;
 		error = xfs_bmapi_write(args->trans, dp, (xfs_fileoff_t)lblkno,
-				  blkcnt,
-				  XFS_BMAPI_ATTRFORK | XFS_BMAPI_METADATA,
-				  args->firstblock, args->total, &map, &nmap,
-				  args->flist);
+				  blkcnt, XFS_BMAPI_ATTRFORK, args->firstblock,
+				  args->total, &map, &nmap, args->flist);
 		if (!error) {
 			error = xfs_bmap_finish(&args->trans, args->flist,
 						&committed);
-- 
1.9.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Thread

[3.13.y-ckt stable] Linux 3.13.11-ckt26 stable review Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:00 +0200
  [PATCH 3.13.y-ckt 46/60] ipr: Fix locking for unit attention handling Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:00 +0200
  [PATCH 3.13.y-ckt 32/60] usb: gadget: mv_udc_core: fix phy_regs I/O memory leak Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:00 +0200
  [PATCH 3.13.y-ckt 57/60] ALSA: hda - fix cs4210_spdif_automute() Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 22/60] Input: usbtouchscreen - avoid unresponsive TSC-30 touch screen Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 30/60] net: Clone skb before setting peeked flag Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 43/60] ALSA: usb-audio: add dB range mapping for some devices Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 50/60] x86/xen: Probe target addresses in set_aliased_prot() before the hypercall Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 20/60] blkcg: fix gendisk reference leak in blkg_conf_prep() Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 34/60] bonding: fix destruction of bond with devices different from arphrd_ether Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 58/60] niu: don't count tx error twice in case of headroom realloc fails Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 23/60] md/raid1: fix test for 'was read error from last working device'. Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 49/60] ALSA: hda - Fix MacBook Pro 5,2 quirk Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 55/60] netfilter: nf_conntrack: Support expectations in different zones Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 38/60] iscsi-target: Fix iscsit_start_kthreads failure OOPs Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 53/60] ipvs: fix crash with sync protocol v0 and FTP Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 26/60] mmc: sdhci-pxav3: fix platform_data is not initialized Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 27/60] mmc: block: Add missing mmc_blk_put() in power_ro_lock_show() Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 44/60] drm/radeon/combios: add some validation of lvds values Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 24/60] mmc: omap_hsmmc: Fix DTO and DCRC handling Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 39/60] iscsi-target: Fix iser explicit logout TX kthread leak Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 37/60] iscsi-target: Fix use-after-free during TPG session shutdown Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 40/60] ALSA: hda - Apply fixup for another Toshiba Satellite S50D Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 33/60] inet: frags: fix defragmented packet's IP header for af_packet Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 42/60] xfs: remote attributes need to be considered data Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 52/60] bridge: mdb: fix delmdb state in the notification Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 29/60] bonding: correctly handle bonding type change on enslave failure Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 47/60] ipr: Fix incorrect trace indexing Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 51/60] netfilter: ctnetlink: put back references to master ct and expect objects Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 45/60] x86/efi: Use all 64 bit of efi_memmap in setup_e820() Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 31/60] bridge: mdb: fix double add notification Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 35/60] ARM: OMAP2+: hwmod: Fix _wait_target_ready() for hwmods without sysc Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 48/60] ipr: Fix invalid array indexing for HRRQ Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 41/60] vhost: actually track log eventfd file Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 36/60] ASoC: pcm1681: Fix setting de-emphasis sampling rate selection Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:10 +0200
  [PATCH 3.13.y-ckt 03/60] libata: add ATA_HORKAGE_BROKEN_FPDMA_AA quirk for HP 250GB SATA disk VB0250EAVER Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:20 +0200
  [PATCH 3.13.y-ckt 17/60] xhci: do not report PLC when link is in internal resume state Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:20 +0200
  [PATCH 3.13.y-ckt 18/60] USB: OHCI: Fix race between ED unlink and URB submission Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:20 +0200
  [PATCH 3.13.y-ckt 08/60] mac80211: clear subdir_stations when removing debugfs Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:20 +0200
  [PATCH 3.13.y-ckt 10/60] md/raid10: always set reshape_safe when initializing reshape_position. Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:20 +0200
    Re: [PATCH 3.13.y-ckt 10/60] md/raid10: always set reshape_safe  when initializing reshape_position. Kamal Mostafa <kamal@canonical.com> - 2015-09-04 21:40 +0200
  [PATCH 3.13.y-ckt 05/60] libata: force disable trim for SuperSSpeed S238 Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:20 +0200
  [PATCH 3.13.y-ckt 13/60] usb: xhci: Bugfix for NULL pointer deference in xhci_endpoint_init() function Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:20 +0200
  [PATCH 3.13.y-ckt 06/60] libata: increase the timeout when setting transfer mode Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:20 +0200
  [PATCH 3.13.y-ckt 11/60] drm: Stop resetting connector state to unknown Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:20 +0200
  [PATCH 3.13.y-ckt 09/60] ALSA: hda - Add new GPU codec ID 0x10de007d to snd-hda Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:20 +0200
  [PATCH 3.13.y-ckt 16/60] xhci: prevent bus_suspend if SS port resuming in phase 1 Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:20 +0200
  [PATCH 3.13.y-ckt 25/60] isdn/gigaset: reset tty->receive_room when attaching ser_gigaset Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:20 +0200
  [PATCH 3.13.y-ckt 19/60] usb-storage: ignore ZTE MF 823 card reader in mode 0x1225 Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:20 +0200
  [PATCH 3.13.y-ckt 02/60] ata: pmp: add quirk for Marvell 4140 SATA PMP Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:20 +0200
  [PATCH 3.13.y-ckt 04/60] libata: add ATA_HORKAGE_NOTRIM Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:20 +0200
  [PATCH 3.13.y-ckt 15/60] xhci: report U3 when link is in resume state Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:20 +0200
  [PATCH 3.13.y-ckt 21/60] tile: use free_bootmem_late() for initrd Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:20 +0200
  [PATCH 3.13.y-ckt 12/60] usb: dwc3: Reset the transfer resource index on SET_INTERFACE Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:20 +0200
  [PATCH 3.13.y-ckt 07/60] libata: Do not blacklist M510DC Kamal Mostafa <kamal@canonical.com> - 2015-09-02 03:20 +0200

csiph-web