Traditionally ext2/3/4 has returned a 32-bit hash value from llseek()
to appease NFSv2, which can only handle a 32-bit cookie for seekdir()
and telldir().  However, this causes problems if there are 32-bit hash
collisions, since the NFSv2 server can get stuck resending the same
entries from the directory repeatedly.

Allow ext4 to return a full 64-bit hash (both major and minor) for
telldir to decrease the chance of hash collisions.  This still needs
integration on the NFS side.
Patch-updated-by: Bernd Schubert <bernd.schubert@itwm.fraunhofer.de>
(blame me if something is not correct)
Signed-off-by: Fan Yong <yong.fan@whamcloud.com>
Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Signed-off-by: Bernd Schubert <bernd.schubert@itwm.fraunhofer.de>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>

Those flags are supposed to be set by NFS readdir() to tell ext3/ext4
to 32bit (NFSv2) or 64bit hash values (offsets) in seekdir().
Signed-off-by: Bernd Schubert <bernd.schubert@itwm.fraunhofer.de>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>

There are few important bug fixes for LogFS

* tag 'for-linus' of git://github.com/prasad-joshi/logfs_upstream:
  Logfs: Allow NULL block_isbad() methods
  logfs: Grow inode in delete path
  logfs: Free areas before calling generic_shutdown_super()
  logfs: remove useless BUG_ON
  MAINTAINERS: Add Prasad Joshi in LogFS maintiners
  logfs: Propagate page parameter to __logfs_write_inode
  logfs: set superblock shutdown flag after generic sb shutdown
  logfs: take write mutex lock during fsync and sync
  logfs: Prevent memory corruption
  logfs: update page reference count for pined pages

Fix up conflict in fs/logfs/dev_mtd.c due to semantic change in what
"mtd->block_isbad" means in commit f2933e86: "Logfs: Allow NULL
block_isbad() methods" clashing with the abstraction changes in the
commits 7086c19d: "mtd: introduce mtd_block_isbad interface" and
d58b27ed: "logfs: do not use 'mtd->block_isbad' directly".

This resolution takes the semantics from commit f2933e86, and just
makes mtd_block_isbad() return zero (false) if the 'block_isbad'
function is NULL.  But that also means that now "mtd_can_have_bb()"
always returns 0.

Now, "mtd_block_markbad()" will obviously return an error if the
low-level driver doesn't support bad blocks, so this is somewhat
non-symmetric, but it actually makes sense if a NULL "block_isbad"
function is considered to mean "I assume that all my blocks are always
good".

Merge branch 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging

* 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging:
  hwmon: (w83627ehf) Disable setting DC mode for pwm2, pwm3 on NCT6776F
  hwmon: (sht15) fix bad error code
  MAINTAINERS: Drop maintainer for MAX1668 hwmon driver
  MAINTAINERS: Add hwmon entries for Wolfson
  hwmon: (f71805f) Fix clamping of temperature limits

Here are some fixes to the pin control system that has accumulated since
-rc1.  Mainly Tony Lindgren fixed the module load/unload logic and the
rest are minor fixes and documentation.

* 'for-torvalds' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl:
  pinctrl: add checks for empty function names
  pinctrl: fix pinmux_hog_maps when ctrl_dev_name is not set
  pinctrl: fix some pinmux typos
  pinctrl: free debugfs entries when unloading a pinmux driver
  pinctrl: unbreak error messages
  Documentation/pinctrl: fix a few syntax errors in code examples
  pinctrl: fix pinconf_pins_show iteration

Here are some tty/serial patches for 3.3-rc1

Big thing here is the movement of the 8250 serial drivers to their own
directory, now that the patch churn has calmed down.

Other than that, only minor stuff (omap patches were reverted as they
were found to be wrong), and another broken driver removed from the
system.
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

* tag 'tty-3.3-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty:
  serial: Kill off Moorestown code
  Revert "tty: serial: OMAP: ensure FIFO levels are set correctly in non-DMA mode"
  Revert "tty: serial: OMAP: transmit FIFO threshold interrupts don't wake the chip"
  serial: Fix wakeup init logic to speed up startup
  docbook: don't use serial_core.h in device-drivers book
  serial: amba-pl011: lock console writes against interrupts
  amba-pl011: do not disable RTS during shutdown
  tty: serial: OMAP: transmit FIFO threshold interrupts don't wake the chip
  tty: serial: OMAP: ensure FIFO levels are set correctly in non-DMA mode
  omap-serial: make serial_omap_restore_context depend on CONFIG_PM_RUNTIME
  omap-serial :Make the suspend/resume functions depend on CONFIG_PM_SLEEP.
  TTY: fix UV serial console regression
  jsm: Fixed EEH recovery error
  Updated TTY MAINTAINERS info
  serial: group all the 8250 related code together

Here are a bunch of USB patches for 3.3-rc1.

Nothing major, largest thing here is the removal of some drivers that
did not work at all.  Other than that, the normal collection of bugfixes
and new device ids.
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

* tag 'usb-3.3-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb: (52 commits)
  uwb & wusb: fix kconfig error
  USB: Realtek cr: fix autopm scheduling while atomic
  USB: ftdi_sio: Add more identifiers
  xHCI: Cleanup isoc transfer ring when TD length mismatch found
  usb: musb: omap2430: minor cleanups.
  qcaux: add more Pantech UML190 and UML290 ports
  Revert "drivers: usb: Fix dependency for USB_HWA_HCD"
  usb: mv-otg - Fix build if CONFIG_USB is not set
  USB: cdc-wdm: Avoid hanging on interface with no USB_CDC_DMM_TYPE
  usb: add support for STA2X11 host driver
  drivers: usb: Fix dependency for USB_HWA_HCD
  kernel-doc: fix new warning in usb.h
  USB: OHCI: fix new compiler warnings
  usb: serial: kobil_sct: fix compile warning:
  drivers/usb/host/ehci-fsl.c: add missing iounmap
  USB: cdc-wdm: better allocate a buffer that is at least as big as we tell the USB core
  USB: cdc-wdm: call wake_up_all to allow driver to shutdown on device removal
  USB: cdc-wdm: use two mutexes to allow simultaneous read and write
  USB: cdc-wdm: updating desc->length must be protected by spin_lock
  USB: usbsevseg: fix max length
  ...

1) Setting link attributes can modify the size of the attributes that
   would be reported on a subsequent getlink netlink operation,
   therefore min_ifinfo_dump_size needs to be adjusted.  From Stefan
   Gula.

2) Resegmentation of TSO frames while trimming can violate invariants
   expected by callers, namely that the number of segments can only stay
   the same or decrease, never increase.  If MSS changes, however, we
   can trim data but then end up with more segments.  Fix this by only
   segmenting to the MSS already recorded in the SKB.  That's the
   simplest fix for now and if we want to get more fancy in the future
   that's a more involved change.

   This probably explains some retransmit counter inaccuracies.

   From Neal Cardwell.

3) Fix too-many-wakeups in POLL with AF_UNIX sockets, from Eric Dumazet.

4) Fix CAIF crashes wrt.  namespace handling.  From Eric Dumazet and
   Eric W. Biederman.

5) TCP port selection fixes from Flavio Leitner.

6) More socket memory cgroup build fixes in certain randonfig
   situations.  From Glauber Costa.

7) Fix TCP memory sysctl regression reported by Ingo Molnar, also from
   Glauber Costa.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net:
  af_unix: fix EPOLLET regression for stream sockets
  tcp: fix tcp_trim_head() to adjust segment count with skb MSS
  net/tcp: Fix tcp memory limits initialization when !CONFIG_SYSCTL
  net caif: Register properly as a pernet subsystem.
  netns: Fail conspicously if someone uses net_generic at an inappropriate time.
  net: explicitly add jump_label.h header to sock.h
  net: RTNETLINK adjusting values of min_ifinfo_dump_size
  ipv6: Fix ip_gre lockless xmits.
  xen-netfront: correct MAX_TX_TARGET calculation.
  netns: fix net_alloc_generic()
  tcp: bind() optimize port allocation
  tcp: bind() fix autoselection to share ports
  l2tp: l2tp_ip - fix possible oops on packet receive
  iwlwifi: fix PCI-E transport "inta" race
  mac80211: set bss_conf.idle when vif is connected
  mac80211: update oper_channel on ibss join

This fixes an integration issue with the regulator device tree bindings
which shook out in -rc.  The bindings were overly enthusiatic when
deciding to set a voltage on a regulator and would try to set zero volts
on an unconfigured regulator which isn't supported.

* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regulator:
  regulator: Set apply_uV only when min and max voltages are defined

Commit 0884d7aa (AF_UNIX: Fix poll blocking problem when reading from
a stream socket) added a regression for epoll() in Edge Triggered mode
(EPOLLET)

Appropriate fix is to use skb_peek()/skb_unlink() instead of
skb_dequeue(), and only call skb_unlink() when skb is fully consumed.

This remove the need to requeue a partial skb into sk_receive_queue head
and the extra sk->sk_data_ready() calls that added the regression.

This is safe because once skb is given to sk_receive_queue, it is not
modified by a writer, and readers are serialized by u->readlock mutex.

This also reduce number of spinlock acquisition for small reads or
MSG_PEEK users so should improve overall performance.
Reported-by: Nick Mathewson <nickm@freehaven.net>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Alexey Moiseytsev <himeraster@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This commit fixes tcp_trim_head() to recalculate the number of
segments in the skb with the skb's existing MSS, so trimming the head
causes the skb segment count to be monotonically non-increasing - it
should stay the same or go down, but not increase.

Previously tcp_trim_head() used the current MSS of the connection. But
if there was a decrease in MSS between original transmission and ACK
(e.g. due to PMTUD), this could cause tcp_trim_head() to
counter-intuitively increase the segment count when trimming bytes off
the head of an skb. This violated assumptions in tcp_tso_acked() that
tcp_trim_head() only decreases the packet count, so that packets_acked
in tcp_tso_acked() could underflow, leading tcp_clean_rtx_queue() to
pass u32 pkts_acked values as large as 0xffffffff to
ca_ops->pkts_acked().

As an aside, if tcp_trim_head() had really wanted the skb to reflect
the current MSS, it should have called tcp_set_skb_tso_segs()
unconditionally, since a decrease in MSS would mean that a
single-packet skb should now be sliced into multiple segments.
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>

sysctl_tcp_mem() initialization was moved to sysctl_tcp_ipv4.c
in commit 3dc43e3e, since it
became a per-ns value.

That code, however, will never run when CONFIG_SYSCTL is
disabled, leading to bogus values on those fields - causing hung
TCP sockets.

This patch fixes it by keeping an initialization code in
tcp_init(). It will be overwritten by the first net namespace
init if CONFIG_SYSCTL is compiled in, and do the right thing if
it is compiled out.

It is also named properly as tcp_init_mem(), to properly signal
its non-sysctl side effect on TCP limits.
Reported-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Glauber Costa <glommer@parallels.com>
Cc: David S. Miller <davem@davemloft.net>
Link: http://lkml.kernel.org/r/4F22D05A.8030604@parallels.com
[ renamed the function, tidied up the changelog a bit ]
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: David S. Miller <davem@davemloft.net>

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux:
  [S390] dasd: revalidate server for new pathgroup
  [S390] dasd: revert LCU optimization
  [S390] cleanup entry point definition

* 'next' of git://git.monstr.eu/linux-2.6-microblaze:
  microblaze: generic atomic64 support

* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux:
  vmwgfx: Fix assignment in vmw_framebuffer_create_handle
  drm/radeon/kms: Fix device tree linkage of i2c buses
  drm: Pass the real error code back during GEM bo initialisation
  Revert "drm/i810: cleanup reclaim_buffers"

NFS client bugfixes for Linux 3.3 (pull 3)

* tag 'nfs-for-3.3-3' of git://git.linux-nfs.org/projects/trondmy/linux-nfs:
  SUNRPC: Fix machine creds in generic_create_cred and generic_match

Power management fix for 3.3-rc2

Fix for a hibernate (s2disk) regression introduced during the 3.2
merge window that causes s2disk to trigger BUG_ON() in
freeze_workqueues_begin() if there is not enough swap space to save
the image.

* tag 'pm-fix-for-3.3-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  PM / Hibernate: Fix s2disk regression related to freezing workqueues

The assignment of handle in vmw_framebuffer_create_handle doesn't actually do anything useful and is incorrectly assigning an integer value to a pointer argument. It appears that this is a typo and should be dereferencing handle rather than assigning to it directly. This fixes a bug where an undefined handle value is potentially returned to user-space.
Signed-off-by: Ryan Mallon <rmallon@gmail.com>
Reviewed-by: Jakob Bornecrantz<jakob@vmware.com>
Cc: stable@vger.kernel.org
Signed-off-by: Dave Airlie <airlied@redhat.com>

Properly set the parent device of i2c buses before registering them so
that they will show at the right place in the device tree (rather than
in /sys/devices directly.)
Signed-off-by: Jean Delvare <jdelvare@suse.de>
Cc: Dave Airlie <airlied@gmail.com>
Reviewed-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

In particular, I found I was hitting the max-file limit in the VFS,
and the EFILE was being magically transformed into ENOMEM. Confusion
reigns.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Dave Airlie <airlied@redhat.com>

This reverts commit 87499ffd.

Where is that paper bag ... ah here.

I've failed to take an odd interaction between my other cleanups and
this reclaim_buffers patch into account and also failed to properly
test it. Looks like there are more dragons and hidden trapdoors in the
drm release path than actual lines of code.

Until I get a clue, let's just revert this.
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Dave Airlie <airlied@redhat.com>

NCT6776F only supports pwm mode for pwm2 and pwm3. Return error if an attempt
is made to set those pwm channels to DC mode.
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Acked-by: Jean Delvare <khali@linux-fr.org>
Cc: stable@vger.kernel.org # 3.0+
Signed-off-by: Guenter Roeck <guenter.roeck@ericsson.com>

Commit 2aede851

  PM / Hibernate: Freeze kernel threads after preallocating memory

introduced a mechanism by which kernel threads were frozen after
the preallocation of hibernate image memory to avoid problems with
frozen kernel threads not responding to memory freeing requests.
However, it overlooked the s2disk code path in which the
SNAPSHOT_CREATE_IMAGE ioctl was run directly after SNAPSHOT_FREE,
which caused freeze_workqueues_begin() to BUG(), because it saw
that worqueues had been already frozen.

Although in principle this issue might be addressed by removing
the relevant BUG_ON() from freeze_workqueues_begin(), that would
reintroduce the very problem that commit 2aede851
attempted to avoid into that particular code path.  For this reason,
to fix the issue at hand, introduce thaw_kernel_threads() and make
the SNAPSHOT_FREE ioctl execute it.

Special thanks to Srivatsa S. Bhat for detailed analysis of the
problem.
Reported-and-tested-by: Jiri Slaby <jslaby@suse.cz>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Acked-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
Cc: stable@kernel.org

When no platform data was supplied, returned error code was 0.
Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Cc: stable@vger.kernel.org # 2.6.32+
Signed-off-by: Guenter Roeck <guenter.roeck@ericsson.com>

Merge tag 'driver-core-3.3-rc1-bugfixes' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core

Here are some patches for the 3.3-rc1 tree.

It contains the removal of the sysdev code, now that all users of it are
gone, as well as some sysfs bugfixes that have been reported by users.
There are also some documentation updates here as well.

* tag 'driver-core-3.3-rc1-bugfixes' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core:
  sysfs: Complain bitterly about attempts to remove files from nonexistent directories.
  stable: update documentation to ask for kernel version
  base/core.c:fix typo in comment in function device_add
  Documentation: devres: add allocation functions to list of supported calls
  Documentation update for the driver model core
  kernel-doc: fix new warnings in driver-core
  kernel-doc: fix new warnings in debugfs
  kernel-doc: fix new warnings in device.h
  driver core: remove drivers/base/sys.c and include/linux/sysdev.h

* tag 'for-linus' of git://github.com/rustyrussell/linux:
  lguest: remove reference from Documentation/virtual/00-INDEX
  virtio: correct the memory barrier in virtqueue_kick_prepare()
  virtio: fix typos of memory barriers

Merge branch 'stable/for-linus-fixes-3.3' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen

* 'stable/for-linus-fixes-3.3' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen:
  xen/granttable: Disable grant v2 for HVM domains.
  x86: xen: size struct xen_spinlock to always fit in arch_spinlock_t

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs:
  Btrfs: fix reservations in btrfs_page_mkwrite
  Btrfs: advance window_start if we're using a bitmap
  btrfs: mask out gfp flags in releasepage
  Btrfs: fix enospc error caused by wrong checks of the chunk
  Btrfs: do not defrag a file partially
  Btrfs: fix warning for 32-bit build of fs/btrfs/check-integrity.c
  Btrfs: use cluster->window_start when allocating from a cluster bitmap
  Btrfs: Check for NULL page in extent_range_uptodate
  btrfs: Fix busyloops in transaction waiting code
  Btrfs: make sure a bitmap has enough bytes
  Btrfs: fix uninit warning in backref.c

* git://www.linux-watchdog.org/linux-watchdog:
  watchdog: iTCO_wdt: add Intel Lynx Point DeviceIDs
  watchdog: via_wdt: Set min_timeout and max_timeout for wdt_dev
  watchdog: Fix typo "unexpectdly"
  watchdog: wafer5823wdt: Fix handling WDIOS_DISABLECARD/WDIOS_ENABLECARD options
  watchdog: wm8350_wdt: Fix handling WDIOS_DISABLECARD/WDIOS_ENABLECARD options
  watchdog: Return proper error in nuc900wdt_probe if misc_register fails
  watchdog: Staticise nuc900_wdt
  watchdog: via_wdt: Staticise wdt_pci_table
  watchdog: omap_wdt.c: Fix the mismatch of pm_runtime enable and disable
  watchdog: dw_wdt.c: use devm_request_and_ioremap
  watchdog: imx2_wdt.c: use devm_request_and_ioremap

* 'fixes' of git://git.linaro.org/people/rmk/linux-arm: (31 commits)
  ARM: 7304/1: ioremap: fix boundary check when reusing static mapping
  ARM: 7301/1: Rename the T() macro to TUSER() to avoid namespace conflicts
  ARM: 7299/1: ftrace: clear zero bit in reported IPs for Thumb-2
  ARM: 7298/1: realview: fix mapping of MPCore private memory region
  PCMCIA: fix sa1111 oops on remove
  ARM: 7288/1: mach-sa1100: add missing module_init() call
  ARM: 7297/1: smp_twd: make sure timer is stopped before registering it
  ARM: 7296/1: proc-v7.S: remove HARVARD_CACHE preprocessor guards
  ARM: 7295/1: cortex-a7: move proc_info out of !CONFIG_ARM_LPAE block
  ARM: 7293/1: logical_cpu_map: decouple CPU mapping from SMP
  ARM: 7291/1: cache: assume 64-byte L1 cachelines for ARMv7 CPUs
  ARM: 7290/1: vmlinux.lds.S: align the exception fixup table to a 4-byte boundary
  ARM: 7289/1: vmlinux.lds.S: do not hardcode cacheline size as 32 bytes
  MFD: ucb1x00-ts: fix resume failure
  MFD: ucb1x00-core: fix gpiolib direction_output handling
  MFD: ucb1x00-core: fix missing restore of io output data on resume
  MFD: mcp-core: fix mcp_priv() to be more type safe
  MFD: mcp-core: fix complaints from the genirq layer
  Revert "ARM: sa11x0: Implement autoloading of codec and codec pdata for mcp bus."
  Revert "ARM: sa1100: Refactor mcp-sa11x0 to use platform resources."
  ...

Fix up conflict due to arch/arm/mach-mx5/Kconfig having been merged into
mach-imx5 (commit 784a90c0: "ARM i.MX: Merge i.MX5 support into
mach-imx"), but the ARM_L1_CACHE_SHIFT_6 entry was moved to be driven by
the CPU_V7 logic from it in the old location in rmk's branch (commit
a092f2b1: "ARM: 7291/1: cache: assume 64-byte L1 cachelines for
ARMv7 CPUs").

arm-soc fixes for 3.3-rc:

AT91 needed reset fixes which resulted in some minor code refactoring,
it also adds a feature-removal for one of their platforms for 3.4.
The USB patches have been acked by Greg K-H.

i.MX and ux500 both have some minor fixes, nothing controversial.

* tag 'fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc:
  arch/arm/mach-imx/mach-mx53_ard.c: add missing iounmap
  ARM: imx: iomux-v1.h: Fix build error due to __init annotation
  ARM: at91: Fix at91sam9g45 and at91cap9 reset
  ARM: at91: make rstc soc independent
  ARM: at91: introduce AT91_SAM9_ALT_RESET to select the at91sam9 alternative reset
  ARM: at91: merge at91cap9_ddrsdr.h in at91sam9_ddrsdr.h
  ARM: at91: fix cap9 ddrsdr register
  ARM/USB: at91/ohci-at91: rename vbus_pin_inverted to vbus_pin_active_low
  USB: at91: fix clk_get error handling
  ARM: at91: removal of CAP9 SoC family
  ARM: at91: fix at91rm9200 soc subtype handling
  mach-ux500: no MMC_CAP_SD_HIGHSPEED on Snowball
  mach-ux500: enable ARM errata 764369
  mach-ux500: do not override outer.inv_all
  mach-ux500: musb: now musb is always in OTG mode
  ARM: imx6: add missing twd_clk for imx6q clock

Not all mtd drivers define block_isbad().  Let's assume no bad blocks
instead of refusing to mount.
Signed-off-by: Joern Engel <joern@logfs.org>

Can be necessary if an inode gets deleted (through -ENOSPC) before being
written.  Might be better to move this into logfs_write_rec(), but for
now go with the stupid&safe patch.
Signed-off-by: Joern Engel <joern@logfs.org>

Or hit an assertion in map_invalidatepage() instead.
Signed-off-by: Joern Engel <joern@logfs.org>

It prevents write sizes >4k.
Signed-off-by: Joern Engel <joern@logfs.org>

Acked-by: Joern Engel <joern@logfs.org>
Signed-off-by: Prasad Joshi <prasadjoshi.linux@gmail.com>

During GC LogFS has to rewrite each valid block to a separate segment.
Rewrite operation reads data from an old segment and writes it to a
newly allocated segment. Since every write operation changes data
block pointers maintained in inode, inode should also be rewritten.

In GC path to avoid AB-BA deadlock LogFS marks a page with
PG_pre_locked in addition to locking the page (PG_locked). The page
lock is ignored iff the page is pre-locked.

LogFS uses a special file called segment file. The segment file
maintains an 8 bytes entry for every segment. It keeps track of erase
count, level etc. for every segment.

Bad things happen with a segment belonging to the segment file is GCed

 ------------[ cut here ]------------
kernel BUG at /home/prasad/logfs/readwrite.c:297!
invalid opcode: 0000 [#1] SMP
Modules linked in: logfs joydev usbhid hid psmouse e1000 i2c_piix4
		serio_raw [last unloaded: logfs]
Pid: 20161, comm: mount Not tainted 3.1.0-rc3+ #3 innotek GmbH
		VirtualBox
EIP: 0060:[<f809132a>] EFLAGS: 00010292 CPU: 0
EIP is at logfs_lock_write_page+0x6a/0x70 [logfs]
EAX: 00000027 EBX: f73f5b20 ECX: c16007c8 EDX: 00000094
ESI: 00000000 EDI: e59be6e4 EBP: c7337b28 ESP: c7337b18
DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068
Process mount (pid: 20161, ti=c7336000 task=eb323f70 task.ti=c7336000)
Stack:
f8099a3d c7337b24 f73f5b20 00001002 c7337b50 f8091f6d f8099a4d f80994e4
00000003 00000000 c7337b68 00000000 c67e4400 00001000 c7337b80 f80935e5
00000000 00000000 00000000 00000000 e1fcf000 0000000f e59be618 c70bf900
Call Trace:
[<f8091f6d>] logfs_get_write_page.clone.16+0xdd/0x100 [logfs]
[<f80935e5>] logfs_mod_segment_entry+0x55/0x110 [logfs]
[<f809460d>] logfs_get_segment_entry+0x1d/0x20 [logfs]
[<f8091060>] ? logfs_cleanup_journal+0x50/0x50 [logfs]
[<f809521b>] ostore_get_erase_count+0x1b/0x40 [logfs]
[<f80965b8>] logfs_open_area+0xc8/0x150 [logfs]
[<c141a7ec>] ? kmemleak_alloc+0x2c/0x60
[<f809668e>] __logfs_segment_write.clone.16+0x4e/0x1b0 [logfs]
[<c10dd563>] ? mempool_kmalloc+0x13/0x20
[<c10dd563>] ? mempool_kmalloc+0x13/0x20
[<f809696f>] logfs_segment_write+0x17f/0x1d0 [logfs]
[<f8092e8c>] logfs_write_i0+0x11c/0x180 [logfs]
[<f8092f35>] logfs_write_direct+0x45/0x90 [logfs]
[<f80934cd>] __logfs_write_buf+0xbd/0xf0 [logfs]
[<c102900e>] ? kmap_atomic_prot+0x4e/0xe0
[<f809424b>] logfs_write_buf+0x3b/0x60 [logfs]
[<f80947a9>] __logfs_write_inode+0xa9/0x110 [logfs]
[<f8094cb0>] logfs_rewrite_block+0xc0/0x110 [logfs]
[<f8095300>] ? get_mapping_page+0x10/0x60 [logfs]
[<f8095aa0>] ? logfs_load_object_aliases+0x2e0/0x2f0 [logfs]
[<f808e57d>] logfs_gc_segment+0x2ad/0x310 [logfs]
[<f808e62a>] __logfs_gc_once+0x4a/0x80 [logfs]
[<f808ed43>] logfs_gc_pass+0x683/0x6a0 [logfs]
[<f8097a89>] logfs_mount+0x5a9/0x680 [logfs]
[<c1126b21>] mount_fs+0x21/0xd0
[<c10f6f6f>] ? __alloc_percpu+0xf/0x20
[<c113da41>] ? alloc_vfsmnt+0xb1/0x130
[<c113db4b>] vfs_kern_mount+0x4b/0xa0
[<c113e06e>] do_kern_mount+0x3e/0xe0
[<c113f60d>] do_mount+0x34d/0x670
[<c10f2749>] ? strndup_user+0x49/0x70
[<c113fcab>] sys_mount+0x6b/0xa0
[<c142d87c>] syscall_call+0x7/0xb
Code: f8 e8 8b 93 39 c9 8b 45 f8 3e 0f ba 28 00 19 d2 85 d2 74 ca eb d0 0f 0b 8d 45 fc 89 44 24 04 c7 04 24 3d 9a 09 f8 e8 09 92 39 c9 <0f> 0b 8d 74 26 00 55 89 e5 3e 8d 74 26 00 8b 10 80 e6 01 74 09
EIP: [<f809132a>] logfs_lock_write_page+0x6a/0x70 [logfs] SS:ESP 0068:c7337b18
---[ end trace 96e67d5b3aa3d6ca ]---

The patch passes locked page to __logfs_write_inode. It calls function
logfs_get_wblocks() to pre-lock the page. This ensures any further
attempts to lock the page are ignored (esp from get_erase_count).
Acked-by: Joern Engel <joern@logfs.org>
Signed-off-by: Prasad Joshi <prasadjoshi.linux@gmail.com>

While unmounting the file system LogFS calls generic_shutdown_super.
The function does file system independent superblock shutdown.
However, it might result in call file system specific inode eviction.

LogFS marks FS shutting down by setting bit LOGFS_SB_FLAG_SHUTDOWN in
super->s_flags. Since, inode eviction might call truncate on inode,
following BUG is observed when file system is unmounted:

------------[ cut here ]------------
kernel BUG at /home/prasad/logfs/segment.c:362!
invalid opcode: 0000 [#1] PREEMPT SMP
CPU 3
Modules linked in: logfs binfmt_misc ppdev virtio_blk parport_pc lp
	parport psmouse floppy virtio_pci serio_raw virtio_ring virtio

Pid: 1933, comm: umount Not tainted 3.0.0+ #4 Bochs Bochs
RIP: 0010:[<ffffffffa008c841>]  [<ffffffffa008c841>]
		logfs_segment_write+0x211/0x230 [logfs]
RSP: 0018:ffff880062d7b9e8  EFLAGS: 00010202
RAX: 000000000000000e RBX: ffff88006eca9000 RCX: 0000000000000000
RDX: ffff88006fd87c40 RSI: ffffea00014ff468 RDI: ffff88007b68e000
RBP: ffff880062d7ba48 R08: 8000000020451430 R09: 0000000000000000
R10: dead000000100100 R11: 0000000000000000 R12: ffff88006fd87c40
R13: ffffea00014ff468 R14: ffff88005ad0a460 R15: 0000000000000000
FS:  00007f25d50ea760(0000) GS:ffff88007fd80000(0000)
	knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 0000000000d05e48 CR3: 0000000062c72000 CR4: 00000000000006e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process umount (pid: 1933, threadinfo ffff880062d7a000,
	task ffff880070b44500)
Stack:
ffff880062d7ba38 ffff88005ad0a508 0000000000001000 0000000000000000
8000000020451430 ffffea00014ff468 ffff880062d7ba48 ffff88005ad0a460
ffff880062d7bad8 ffffea00014ff468 ffff88006fd87c40 0000000000000000
Call Trace:
[<ffffffffa0088fee>] logfs_write_i0+0x12e/0x190 [logfs]
[<ffffffffa0089360>] __logfs_write_rec+0x140/0x220 [logfs]
[<ffffffffa0089312>] __logfs_write_rec+0xf2/0x220 [logfs]
[<ffffffffa00894a4>] logfs_write_rec+0x64/0xd0 [logfs]
[<ffffffffa0089616>] __logfs_write_buf+0x106/0x110 [logfs]
[<ffffffffa008a19e>] logfs_write_buf+0x4e/0x80 [logfs]
[<ffffffffa008a6b8>] __logfs_write_inode+0x98/0x110 [logfs]
[<ffffffffa008a7c4>] logfs_truncate+0x54/0x290 [logfs]
[<ffffffffa008abfc>] logfs_evict_inode+0xdc/0x190 [logfs]
[<ffffffff8115eef5>] evict+0x85/0x170
[<ffffffff8115f126>] iput+0xe6/0x1b0
[<ffffffff8115b4a8>] shrink_dcache_for_umount_subtree+0x218/0x280
[<ffffffff8115ce91>] shrink_dcache_for_umount+0x51/0x90
[<ffffffff8114796c>] generic_shutdown_super+0x2c/0x100
[<ffffffffa008cc47>] logfs_kill_sb+0x57/0xf0 [logfs]
[<ffffffff81147de5>] deactivate_locked_super+0x45/0x70
[<ffffffff811487ea>] deactivate_super+0x4a/0x70
[<ffffffff81163934>] mntput_no_expire+0xa4/0xf0
[<ffffffff8116469f>] sys_umount+0x6f/0x380
[<ffffffff814dd46b>] system_call_fastpath+0x16/0x1b
Code: 55 c8 49 8d b6 a8 00 00 00 45 89 f9 45 89 e8 4c 89 e1 4c 89 55
b8 c7 04 24 00 00 00 00 e8 68 fc ff ff 4c 8b 55 b8 e9 3c ff ff ff <0f>
0b 0f 0b c7 45 c0 00 00 00 00 e9 44 fe ff ff 66 66 66 66 66
RIP  [<ffffffffa008c841>] logfs_segment_write+0x211/0x230 [logfs]
RSP <ffff880062d7b9e8>
---[ end trace fe6b040cea952290 ]---

Therefore, move super->s_flags setting after the fs-indenpendent work
has been finished.
Reviewed-by: Joern Engel <joern@logfs.org>
Signed-off-by: Prasad Joshi <prasadjoshi.linux@gmail.com>

LogFS uses super->s_write_mutex while writing data to disk. Taking the
same mutex lock in sync and fsync code path solves the following BUG:

------------[ cut here ]------------
kernel BUG at /home/prasad/logfs/dev_bdev.c:134!

Pid: 2387, comm: flush-253:16 Not tainted 3.0.0+ #4 Bochs Bochs
RIP: 0010:[<ffffffffa007deed>]  [<ffffffffa007deed>]
                bdev_writeseg+0x25d/0x270 [logfs]
Call Trace:
[<ffffffffa007c381>] logfs_open_area+0x91/0x150 [logfs]
[<ffffffff8128dcb2>] ? find_level.clone.9+0x62/0x100
[<ffffffffa007c49c>] __logfs_segment_write.clone.20+0x5c/0x190 [logfs]
[<ffffffff810ef005>] ? mempool_kmalloc+0x15/0x20
[<ffffffff810ef383>] ? mempool_alloc+0x53/0x130
[<ffffffffa007c7a4>] logfs_segment_write+0x1d4/0x230 [logfs]
[<ffffffffa0078f8e>] logfs_write_i0+0x12e/0x190 [logfs]
[<ffffffffa0079300>] __logfs_write_rec+0x140/0x220 [logfs]
[<ffffffffa0079444>] logfs_write_rec+0x64/0xd0 [logfs]
[<ffffffffa00795b6>] __logfs_write_buf+0x106/0x110 [logfs]
[<ffffffffa007a13e>] logfs_write_buf+0x4e/0x80 [logfs]
[<ffffffffa0073e33>] __logfs_writepage+0x23/0x80 [logfs]
[<ffffffffa007410c>] logfs_writepage+0xdc/0x110 [logfs]
[<ffffffff810f5ba7>] __writepage+0x17/0x40
[<ffffffff810f6208>] write_cache_pages+0x208/0x4f0
[<ffffffff810f5b90>] ? set_page_dirty+0x70/0x70
[<ffffffff810f653a>] generic_writepages+0x4a/0x70
[<ffffffff810f75d1>] do_writepages+0x21/0x40
[<ffffffff8116b9d1>] writeback_single_inode+0x101/0x250
[<ffffffff8116bdbd>] writeback_sb_inodes+0xed/0x1c0
[<ffffffff8116c5fb>] writeback_inodes_wb+0x7b/0x1e0
[<ffffffff8116cc23>] wb_writeback+0x4c3/0x530
[<ffffffff814d984d>] ? sub_preempt_count+0x9d/0xd0
[<ffffffff8116cd6b>] wb_do_writeback+0xdb/0x290
[<ffffffff814d984d>] ? sub_preempt_count+0x9d/0xd0
[<ffffffff814d6208>] ? _raw_spin_unlock_irqrestore+0x18/0x40
[<ffffffff8105aa5a>] ? del_timer+0x8a/0x120
[<ffffffff8116cfac>] bdi_writeback_thread+0x8c/0x2e0
[<ffffffff8116cf20>] ? wb_do_writeback+0x290/0x290
[<ffffffff8106d2e6>] kthread+0x96/0xa0
[<ffffffff814de514>] kernel_thread_helper+0x4/0x10
[<ffffffff8106d250>] ? kthread_worker_fn+0x190/0x190
[<ffffffff814de510>] ? gs_change+0xb/0xb
RIP  [<ffffffffa007deed>] bdev_writeseg+0x25d/0x270 [logfs]
---[ end trace 0211ad60a57657c4 ]---
Reviewed-by: Joern Engel <joern@logfs.org>
Signed-off-by: Prasad Joshi <prasadjoshi.linux@gmail.com>