mtd: nand: omap: fix ecclayout->oobfree->offset

commit aa6092f9835893290e77c3e24649def49dac1583 upstream.

1) In current implementation, ecclayout->oobfree->offset is calculated with
 respect to ecclayout->eccpos[0] which is incorrect because ECC bytes may not
 be stored contiguously in OOB.
 So, this patch calculates ecclayout->oobfree->offset with respect to last
 ECC byte-position 'eccpos[ecclayout->eccbytes-1]'.

2) ECC layout of some ecc-schemes expects reserved-markers at specific eccpos[]
 which should not be over-written by any file-system metadata.
 So this patch aligns oobfree->offset taking into account of such markers.

Tested-by: Enric Balletbo i Serra <eballetbo@gmail.com>
Tested-by: Stefan Roese <sr@denx.de>
Signed-off-by: Pekon Gupta <pekon@ti.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mtd: nand: omap: fix ecclayout to be in sync with u-boot NAND driver

commit eae39cb4934d3daab3ec828c5201c955b2e56af9 upstream.

Fixes: commit a919e51161b58ed7e6e663daba99ab7d558808f3
       mtd: nand: omap2: clean-up BCHx_HW and BCHx_SW ECC configurations in device_probe

Fixes ecclayout mismatch introduced in above commit for following ecc-schemes:
 - OMAP_ECC_BCH4_CODE_HW_DETECTION_SW
 - OMAP_ECC_BCH8_CODE_HW_DETECTION_SW
 However, this patch also touches other ecc-schemes as the fix required
 refactoring common code, into ecc-scheme specific code.

This patch aligns ecc-layout for below ecc-schemes as per reference [1],[2],[3]

 +---+------------+-------------++-------------+-------------+
 |OOB|BCH8_CODE_HW|BCH8_CODE_HW_||HAM1_CODE_HW |HAM1_CODE_HW |
 |pos|            | DETECTION_SW||(x8 device)  |(x16 device) |
 +---+------------+-------------++-------------+-------------+
 | 0 |BADBLK_MARK | BADBLK_MARK || BADBLK_MARK | BADBLK_MARK |
 | 1 |BADBLK_MARK | BADBLK_MARK || eccpos[0]   | BADBLK_MARK |
 | 2 | eccpos[0]  | eccpos[0]   || eccpos[1]   | eccpos[0]   |
 | 3 | eccpos[1]  | eccpos[1]   || eccpos[2]   | eccpos[1]   |
 | 4 | eccpos[2]  | eccpos[2]   || eccpos[3]   | eccpos[2]   |
 | 5 | eccpos[3]  | eccpos[3]   || eccpos[4]   | eccpos[3]   |
 | 6 | eccpos[4]  | eccpos[4]   || eccpos[5]   | eccpos[4]   |
 | 7 | eccpos[5]  | eccpos[5]   || eccpos[6]   | eccpos[5]   |
 | 8 | eccpos[6]  | eccpos[6]   || eccpos[7]   | eccpos[6]   |
 | 9 | eccpos[7]  | eccpos[7]   || eccpos[8]   | eccpos[7]   |
 |10 | eccpos[8]  | eccpos[8]   || eccpos[9]   | eccpos[8]   |
 |11 | eccpos[9]  | eccpos[9]   || eccpos[10]  | eccpos[9]   |
 |12 | eccpos[10] | eccpos[10]  || eccpos[11]  | eccpos[10]  |
 |13 | eccpos[11] | eccpos[11]  || oobfree[0]  | eccpos[11]  |
 |14 | eccpos[12] | eccpos[12]  || oobfree[1]  | oobfree[0]  |
 |15 | eccpos[13] | <reserved>  || oobfree[2]  | oobfree[1]  |
 +---+------------+-------------++-------------+-------------+
 |16 | eccpos[14] | eccpos[13]  || oobfree[3]  | oobfree[2]  |
 |...| [...]      | [...]       || [...]       | [...]       |
 |56 | eccpos[54] | eccpos[51]  || oobfree[43] | oobfree[42] |
 |57 | eccpos[55] | <reserved>  || oobfree[44] | oobfree[43] |
 +===+============+=============+==============+=============+
 |58 | oobfree[0] | oobfree[0]  || oobfree[45] | oobfree[44] |
 |59 | oobfree[1] | oobfree[1]  || oobfree[46] | oobfree[45] |
 |60 | oobfree[2] | oobfree[2]  || oobfree[47] | oobfree[46] |
 |61 | oobfree[3] | oobfree[3]  || oobfree[48] | oobfree[47] |
 |62 | oobfree[4] | oobfree[4]  || oobfree[49] | oobfree[48] |
 |63 | oobfree[5] | oobfree[5]  || oobfree[50] | oobfree[49] |
 +---+------------+-------------+--------------+-------------+

[1] ecc-layout expected by ROM code, as specified in SoC TRM under:
      Chapter="Initialization"
        Section="Device Initialization by ROM code"
            Sub-Section="Memory Booting"
                Heading="NAND"
                Figure="ECC Locations in NAND Spare Areas"

[2] ecc-layout updates in u-boot
    http://lists.denx.de/pipermail/u-boot/2013-November/167551.html

[3] u-boot configurations to match above ecc-layout are documented at
    https://processors.wiki.ti.com/index.php/Linux_Core_NAND_User%27s_Guide

Reported-by: Enric Balletbo Serra <eballetbo@iseebcn.com>
Tested-by: Enric Balletbo i Serra <eballetbo@gmail.com>
Tested-by: Stefan Roese <sr@denx.de>
Signed-off-by: Pekon Gupta <pekon@ti.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

regulator: da9063: Bug fix when setting max voltage on LDOs 5-11

commit ebf6dad0de89677aa58a4d8b009014ff88a23452 upstream.

Bug fix to allow the setting of maximum voltage for certain LDOs.

What the bug is:

There is a problem caused by an invalid calculation of n_voltages
in the driver. This n_voltages value has the potential to be
different for each regulator.

The value for linear_min_sel is set as DA9063_V##regl_name#
which can be different depending upon the regulator. This is
chosen according to the following definitions in the DA9063
registers.h file:

DA9063_VLDO1_BIAS 0
DA9063_VLDO2_BIAS 0
DA9063_VLDO3_BIAS 0
DA9063_VLDO4_BIAS 0
DA9063_VLDO5_BIAS 2
DA9063_VLDO6_BIAS 2
DA9063_VLDO7_BIAS 2
DA9063_VLDO8_BIAS 2
DA9063_VLDO9_BIAS 3
DA9063_VLDO10_BIAS 2
DA9063_VLDO11_BIAS 2

The calculation for n_voltages is valid for LDOs whose BIAS value
is zero but this is not correct for those LDOs which have a
non-zero value.

What the fix is:

In order to take into account the non-zero linear_min_sel value which
is set for the regulators LDO5, LDO6, LDO7, LDO8, LDO9, LDO10 and
LDO11, the calculation for n_voltages should take into account the
missing term defined by DA9063_V##regl_name#.

This will in turn allow the core constraints calculation to set the
maximum voltage limits correctly and therefore allow users to apply
the maximum expected voltage to all of the LDOs.

Signed-off-by: Steve Twiss <stwiss.opensource@diasemi.com>
Signed-off-by: Mark Brown <broonie@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

workqueue: ensure @task is valid across kthread_stop()

commit 5bdfff96c69a4d5ab9c49e60abf9e070ecd2acbb upstream.

When a kworker should die, the kworkre is notified through WORKER_DIE
flag instead of kthread_should_stop().  This, IIRC, is primarily to
keep the test synchronized inside worker_pool lock.  WORKER_DIE is
first set while holding pool->lock, the lock is dropped and
kthread_stop() is called.

Unfortunately, this means that there's a slight chance that the target
kworker may see WORKER_DIE before kthread_stop() finishes and exits
and frees the target task before or during kthread_stop().

Fix it by pinning the target task before setting WORKER_DIE and
putting it after kthread_stop() is done.

tj: Improved patch description and comment.  Moved pinning above
    WORKER_DIE for better signify what it's protecting.

Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

hwmon: (max1668) Fix writing the minimum temperature

commit 500a91571f0a5d0d3242d83802ea2fd1faccc66e upstream.

When trying to set the minimum temperature, the driver was erroneously
writing the maximum temperature into the chip.

Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Reviewed-by: Jean Delvare <jdelvare@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mei: set client's read_cb to NULL when flow control fails

commit accb884b32e82f943340688c9cd30290531e73e0 upstream.

In mei_cl_read_start(), if it fails to send flow control request, it
will release "cl->read_cb" but forget to set pointer to NULL, leaving
"cl->read_cb" still pointing to random memory, next time this client is
operated like mei_release(), it has chance to refer to this wrong pointer.

Fixes: PANIC at kfree in mei_release()
[228781.826904] Call Trace:
[228781.829737]  [<c16249b8>] ? mei_cl_unlink+0x48/0xa0
[228781.835283]  [<c1624487>] mei_io_cb_free+0x17/0x30
[228781.840733]  [<c16265d8>] mei_release+0xa8/0x180
[228781.845989]  [<c135c610>] ? __fsnotify_parent+0xa0/0xf0
[228781.851925]  [<c1325a69>] __fput+0xd9/0x200
[228781.856696]  [<c1325b9d>] ____fput+0xd/0x10
[228781.861467]  [<c125cae1>] task_work_run+0x81/0xb0
[228781.866821]  [<c1242e53>] do_exit+0x283/0xa00
[228781.871786]  [<c1a82b36>] ? kprobe_flush_task+0x66/0xc0
[228781.877722]  [<c124eeb8>] ? __dequeue_signal+0x18/0x1a0
[228781.883657]  [<c124f072>] ? dequeue_signal+0x32/0x190
[228781.889397]  [<c1243744>] do_group_exit+0x34/0xa0
[228781.894750]  [<c12517b6>] get_signal_to_deliver+0x206/0x610
[228781.901075]  [<c12018d8>] do_signal+0x38/0x100
[228781.906136]  [<c1626d1c>] ? mei_read+0x42c/0x4e0
[228781.911393]  [<c12600a0>] ? wake_up_bit+0x30/0x30
[228781.916745]  [<c16268f0>] ? mei_poll+0x120/0x120
[228781.922001]  [<c1324be9>] ? vfs_read+0x89/0x160
[228781.927158]  [<c16268f0>] ? mei_poll+0x120/0x120
[228781.932414]  [<c133ca34>] ? fget_light+0x44/0xe0
[228781.937670]  [<c1324e58>] ? SyS_read+0x68/0x80
[228781.942730]  [<c12019f5>] do_notify_resume+0x55/0x70
[228781.948376]  [<c1a7de5d>] work_notifysig+0x29/0x30
[228781.953827]  [<c1a70000>] ? bad_area+0x5/0x3e

Signed-off-by: Chao Bi <chao.bi@intel.com>
Signed-off-by: Tomas Winkler <tomas.winkler@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

USB: ftdi_sio: add Cressi Leonardo PID

commit 6dbd46c849e071e6afc1e0cad489b0175bca9318 upstream.

Hello,

the following patch adds an entry for the PID of a Cressi Leonardo
diving computer interface to kernel 3.13.0.
It is detected as FT232RL.
Works with subsurface.

Signed-off-by: Joerg Dorchain <joerg@dorchain.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usb: ehci: fix deadlock when threadirqs option is used

commit a1227f3c1030e96ebc51d677d2f636268845c5fb upstream.

ehci_irq() and ehci_hrtimer_func() can deadlock on ehci->lock when
threadirqs option is used. To prevent the deadlock use
spin_lock_irqsave() in ehci_irq().

This change can be reverted when hrtimer callbacks become threaded.

Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

USB: EHCI: add delay during suspend to prevent erroneous wakeups

commit 3e8d6d85adedc59115a564c0a54b36e42087c4d9 upstream.

High-speed USB connections revert back to full-speed signalling when
the device goes into suspend.  This takes several milliseconds, and
during that time it's not possible to tell reliably whether the device
has been disconnected.

On some platforms, the Wake-On-Disconnect circuitry gets confused
during this intermediate state.  It generates a false wakeup signal,
which can prevent the controller from going to sleep.

To avoid this problem, this patch adds a 5-ms delay to the
ehci_bus_suspend() routine if any ports have to switch over to
full-speed signalling.  (Actually, the delay was already present for
devices using a particular kind of PHY power management; the patch
merely causes the delay to be used more widely.)

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Reviewed-by: Peter Chen <Peter.Chen@freescale.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

USB: serial: option: blacklist interface 4 for Cinterion PHS8 and PXS8

commit 12df84d4a80278a5b1abfec3206795291da52fc9 upstream.

This interface is to be handled by the qmi_wwan driver.

CC: Hans-Christoph Schemmel <hans-christoph.schemmel@gemalto.com>
CC: Christian Schmiedl <christian.schmiedl@gemalto.com>
CC: Nicolaus Colberg <nicolaus.colberg@gemalto.com>
CC: David McCullough <david.mccullough@accelecon.com>
Signed-off-by: Aleksander Morgado <aleksander@aleksander.es>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usb: gadget: bcm63xx_udc: fix build failure on DMA channel code

commit 2d1f7af3d60dd09794e0738a915d272c6c27abc5 upstream.

Commit 3dc6475 ("bcm63xx_enet: add support Broadcom BCM6345 Ethernet")
changed the ENETDMA[CS] macros such that they are no longer macros, but
actual register offset definitions. The bcm63xx_udc driver was not
updated, and as a result, causes the following build error to pop up:

 CC      drivers/usb/gadget/u_ether.o
drivers/usb/gadget/bcm63xx_udc.c: In function 'iudma_write':
drivers/usb/gadget/bcm63xx_udc.c:642:24: error: called object '0' is not
a function
drivers/usb/gadget/bcm63xx_udc.c: In function 'iudma_reset_channel':
drivers/usb/gadget/bcm63xx_udc.c:698:46: error: called object '0' is not
a function
drivers/usb/gadget/bcm63xx_udc.c:700:49: error: called object '0' is not
a function

Fix this by updating usb_dmac_{read,write}l and usb_dmas_{read,write}l to
take an extra channel argument, and use the channel width
(ENETDMA_CHAN_WIDTH) to offset the register we want to access, hence
doing again what the macro implicitely did for us.

Cc: Kevin Cernekee <cernekee@gmail.com>
Cc: Jonas Gorski <jogo@openwrt.org>
Signed-off-by: Florian Fainelli <florian@openwrt.org>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usb: chipidea: need to mask when writting endptflush and endptprime

commit 5bf5dbeda2454296f1984adfbfc8e6f5965ac389 upstream.

ENDPTFLUSH and ENDPTPRIME registers are set by software and clear
by hardware. There is a bit for each endpoint. When we are setting
a bit for an endpoint we should make sure we do not touch other
endpoint bit. There is a race condition if the hardware clear the
bit between the read and the write in hw_write.

Signed-off-by: Peter Chen <peter.chen@freescale.com>
Signed-off-by: Matthieu CASTET <matthieu.castet@parrot.com>
Tested-by: Michael Grzeschik <mgrzeschik@pengutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

can: kvaser_usb: check number of channels returned by HW

commit 862474f8b46f6c1e600d4934e40ba40646c696ec upstream.

It is needed to check the number of channels returned by the HW because it
cannot be greater than MAX_NET_DEVICES otherwise it will crash.

Signed-off-by: Olivier Sobrie <olivier@sobrie.be>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

intel_pstate: Use LFM bus ratio as min ratio/P state

commit 4042e7570cff740460b75c6fc604c629621d3dd2 upstream.

LFM (max efficiency ratio) is the max frequency at minimum voltage
supported by the processor.  Using LFM as the minimum P state
increases performmance without affecting power. By not using P states
below LFM we avoid using P states that are less power efficient.

Signed-off-by: Dirk Brandewie <dirk.j.brandewie@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ACPI / processor: Rework processor throttling with work_on_cpu()

commit f3ca4164529b875374c410193bbbac0ee960895f upstream.

acpi_processor_set_throttling() uses set_cpus_allowed_ptr() to make
sure that the (struct acpi_processor)->acpi_processor_set_throttling()
callback will run on the right CPU.  However, the function may be
called from a worker thread already bound to a different CPU in which
case that won't work.

Make acpi_processor_set_throttling() use work_on_cpu() as appropriate
instead of abusing set_cpus_allowed_ptr().

Reported-and-tested-by: Jiri Olsa <jolsa@redhat.com>
Signed-off-by: Lan Tianyu <tianyu.lan@intel.com>
[rjw: Changelog]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ACPI / video: Filter the _BCL table for duplicate brightness values

commit bd8ba20597f0cfef3ef65c3fd2aa92ab23d4c8e1 upstream.

Some devices have duplicate entries in there brightness levels table, ie
on my Dell Latitude E6430 the table looks like this:

[    3.686060] acpi backlight index   0, val 80
[    3.686095] acpi backlight index   1, val 50
[    3.686122] acpi backlight index   2, val 5
[    3.686147] acpi backlight index   3, val 5
[    3.686172] acpi backlight index   4, val 5
[    3.686197] acpi backlight index   5, val 5
[    3.686223] acpi backlight index   6, val 5
[    3.686248] acpi backlight index   7, val 5
[    3.686273] acpi backlight index   8, val 6
[    3.686332] acpi backlight index   9, val 7
[    3.686356] acpi backlight index  10, val 8
[    3.686380] acpi backlight index  11, val 9
etc.

Notice that brightness values 0-5 are all mapped to 5. This means that
if userspace writes any value between 0 and 5 to the brightness sysfs attribute
and then reads it, it will always return 0, which is somewhat unexpected.

This is a problem for ie gnome-settings-daemon, which uses read-modify-write
logic when the users presses the brightness up or down keys. This is done
this way to take brightness changes from other sources into account.

On this specific laptop what happens once the brightness has been set to 0,
is that gsd reads 0, adds 5, writes 5, and on the next brightness up key press
again reads 0, so things get stuck at the lowest brightness setting.

Filtering out the duplicate table entries, makes any write to brightness
read back as the written value as one would expect, fixing this.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Aaron Lu <aaron.lu@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

i7core_edac: Fix PCI device reference count

commit c0f5eeed0f4cef4f05b74883a7160e7edde58b6a upstream.

The reference count changes done by pci_get_device can be a little
misleading when the usage diverges from the most common scheme. The
reference count of the device passed as the last parameter is always
decreased, even if the function returns no new device. So if we are
going to try alternative device IDs, we must manually increment the
device reference count before each retry. If we don't, we end up
decreasing the reference count, and after a few modprobe/rmmod cycles
the PCI devices will vanish.

In other words and as Alan put it: without this fix the EDAC code
corrupts the PCI device list.

This fixes kernel bug #50491:
https://bugzilla.kernel.org/show_bug.cgi?id=50491

Signed-off-by: Jean Delvare <jdelvare@suse.de>
Link: http://lkml.kernel.org/r/20140224093927.7659dd9d@endymion.delvare
Reviewed-by: Alan Cox <alan@linux.intel.com>
Cc: Mauro Carvalho Chehab <m.chehab@samsung.com>
Cc: Doug Thompson <dougthompson@xmission.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ACPI / PCI: Fix memory leak in acpi_pci_irq_enable()

commit b685f3b1744061aa9ad822548ba9c674de5be7c6 upstream.

acpi_pci_link_allocate_irq() can return negative gsi even if
entry != NULL.  For that case we have a memory leak, so free
entry before returning from acpi_pci_irq_enable() for gsi < 0.

Signed-off-by: Tomasz Nowicki <tomasz.nowicki@linaro.org>
[rjw: Subject and changelog]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

PCI: Enable INTx if BIOS left them disabled

commit 1f42db786b14a31bf807fc41ee5583a00c08fcb1 upstream.

Some firmware leaves the Interrupt Disable bit set even if the device uses
INTx interrupts.  Clear Interrupt Disable so we get those interrupts.

Based on the report mentioned below, if the user selects the "EHCI only"
option in the Intel Baytrail BIOS, the EHCI device is handed off to the OS
with the PCI_COMMAND_INTX_DISABLE bit set.

Link: http://lkml.kernel.org/r/20140114181721.GC12126@xanatos
Link: https://bugzilla.kernel.org/show_bug.cgi?id=70601
Reported-by: Chris Cheng <chris.cheng@atrustcorp.com>
Reported-and-tested-by: Jamie Chen <jamie.chen@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
CC: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

PCI: mvebu: Use Device ID and revision from underlying endpoint

commit 322a8e91844f4ae2093e0d3d8a318d0ef2596756 upstream.

Marvell SoCs place the SoC number into the PCIe endpoint device ID.  The
SoC stepping is placed into the PCIe revision. The old plat-orion PCIe
driver allowed this information to be seen in user space with a simple
lspci command.

The new driver places a virtual PCI-PCI bridge on top of these endpoints.
It has its own hard coded PCI device ID. Thus it is no longer possible to
see what the SoC is using lspci.

When initializing the PCI-PCI bridge, set its device ID and revision from
the underlying endpoint, thus restoring this functionality.  Debian would
like to use this in order to aid installing the correct DTB file.

Fixes: 45361a4fe4464 ("pci: PCIe driver for Marvell Armada 370/XP systems")
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: Jason Cooper <jason@lakedaemon.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Revert "writeback: do not sync data dirtied after sync start"

commit 0dc83bd30b0bf5410c0933cfbbf8853248eff0a9 upstream.

This reverts commit c4a391b53a72d2df4ee97f96f78c1d5971b47489. Dave
Chinner <david@fromorbit.com> has reported the commit may cause some
inodes to be left out from sync(2). This is because we can call
redirty_tail() for some inode (which sets i_dirtied_when to current time)
after sync(2) has started or similarly requeue_inode() can set
i_dirtied_when to current time if writeback had to skip some pages. The
real problem is in the functions clobbering i_dirtied_when but fixing
that isn't trivial so revert is a safer choice for now.

Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

cpufreq: powernow-k8: Initialize per-cpu data-structures properly

commit c3274763bfc3bf1ececa269ed6e6c4d7ec1c3e5e upstream.

The powernow-k8 driver maintains a per-cpu data-structure called
powernow_data that is used to perform the frequency transitions.
It initializes this data structure only for the policy->cpu. So,
accesses to this data structure by other CPUs results in various
problems because they would have been uninitialized.

Specifically, if a cpu (!= policy->cpu) invokes the drivers' ->get()
function, it returns 0 as the KHz value, since its per-cpu memory
doesn't point to anything valid. This causes problems during
suspend/resume since cpufreq_update_policy() tries to enforce this
(0 KHz) as the current frequency of the CPU, and this madness gets
propagated to adjust_jiffies() as well. Eventually, lots of things
start breaking down, including the r8169 ethernet card, in one
particularly interesting case reported by Pierre Ossman.

Fix this by initializing the per-cpu data-structures of all the CPUs
in the policy appropriately.

References: https://bugzilla.kernel.org/show_bug.cgi?id=70311
Reported-by: Pierre Ossman <pierre@ossman.eu>
Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

sata_sil: apply MOD15WRITE quirk to TOSHIBA MK2561GSYN

commit 9f9c47f00ce99329b1a82e2ac4f70f0fe3db549c upstream.

It's a bit odd to see a newer device showing mod15write; however, the
reported behavior is highly consistent and other factors which could
contribute seem to have been verified well enough.  Also, both
sata_sil itself and the drive are fairly outdated at this point making
the risk of this change fairly low.  It is possible, probably likely,
that other drive models in the same family have the same problem;
however, for now, let's just add the specific model which was tested.

Signed-off-by: Tejun Heo <tj@kernel.org>
Reported-by: matson <lists-matsonpa@luxsci.me>
References: http://lkml.kernel.org/g/201401211912.s0LJCk7F015058@rs103.luxsci.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ata: enable quirk from jmicron JMB350 for JMB394

commit efb9e0f4f43780f0ae0c6428d66bd03e805c7539 upstream.

Without the patch the kernel generates the following error.

 ata11.15: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
 ata11.15: Port Multiplier vendor mismatch '0x197b' != '0x123'
 ata11.15: PMP revalidation failed (errno=-19)
 ata11.15: failed to recover PMP after 5 tries, giving up

This patch helps to bypass this error and the device becomes
functional.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: <linux-ide@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

perf/x86: Fix event scheduling

commit 26e61e8939b1fe8729572dabe9a9e97d930dd4f6 upstream.

Vince "Super Tester" Weaver reported a new round of syscall fuzzing (Trinity) failures,
with perf WARN_ON()s triggering. He also provided traces of the failures.

This is I think the relevant bit:

>    pec_1076_warn-2804  [000] d...   147.926153: x86_pmu_disable: x86_pmu_disable
>    pec_1076_warn-2804  [000] d...   147.926153: x86_pmu_state: Events: {
>    pec_1076_warn-2804  [000] d...   147.926156: x86_pmu_state:   0: state: .R config: ffffffffffffffff (          (null))
>    pec_1076_warn-2804  [000] d...   147.926158: x86_pmu_state:   33: state: AR config: 0 (ffff88011ac99800)
>    pec_1076_warn-2804  [000] d...   147.926159: x86_pmu_state: }
>    pec_1076_warn-2804  [000] d...   147.926160: x86_pmu_state: n_events: 1, n_added: 0, n_txn: 1
>    pec_1076_warn-2804  [000] d...   147.926161: x86_pmu_state: Assignment: {
>    pec_1076_warn-2804  [000] d...   147.926162: x86_pmu_state:   0->33 tag: 1 config: 0 (ffff88011ac99800)
>    pec_1076_warn-2804  [000] d...   147.926163: x86_pmu_state: }
>    pec_1076_warn-2804  [000] d...   147.926166: collect_events: Adding event: 1 (ffff880119ec8800)

So we add the insn:p event (fd[23]).

At this point we should have:

  n_events = 2, n_added = 1, n_txn = 1

>    pec_1076_warn-2804  [000] d...   147.926170: collect_events: Adding event: 0 (ffff8800c9e01800)
>    pec_1076_warn-2804  [000] d...   147.926172: collect_events: Adding event: 4 (ffff8800cbab2c00)

We try and add the {BP,cycles,br_insn} group (fd[3], fd[4], fd[15]).
These events are 0:cycles and 4:br_insn, the BP event isn't x86_pmu so
that's not visible.

group_sched_in()
  pmu->start_txn() /* nop - BP pmu */
  event_sched_in()
     event->pmu->add()

So here we should end up with:

  0: n_events = 3, n_added = 2, n_txn = 2
  4: n_events = 4, n_added = 3, n_txn = 3

But seeing the below state on x86_pmu_enable(), the must have failed,
because the 0 and 4 events aren't there anymore.

Looking at group_sched_in(), since the BP is the leader, its
event_sched_in() must have succeeded, for otherwise we would not have
seen the sibling adds.

But since neither 0 or 4 are in the below state; their event_sched_in()
must have failed; but I don't see why, the complete state: 0,0,1:p,4
fits perfectly fine on a core2.

However, since we try and schedule 4 it means the 0 event must have
succeeded!  Therefore the 4 event must have failed, its failure will
have put group_sched_in() into the fail path, which will call:

event_sched_out()
  event->pmu->del()

on 0 and the BP event.

Now x86_pmu_del() will reduce n_events; but it will not reduce n_added;
giving what we see below:

 n_event = 2, n_added = 2, n_txn = 2

>    pec_1076_warn-2804  [000] d...   147.926177: x86_pmu_enable: x86_pmu_enable
>    pec_1076_warn-2804  [000] d...   147.926177: x86_pmu_state: Events: {
>    pec_1076_warn-2804  [000] d...   147.926179: x86_pmu_state:   0: state: .R config: ffffffffffffffff (          (null))
>    pec_1076_warn-2804  [000] d...   147.926181: x86_pmu_state:   33: state: AR config: 0 (ffff88011ac99800)
>    pec_1076_warn-2804  [000] d...   147.926182: x86_pmu_state: }
>    pec_1076_warn-2804  [000] d...   147.926184: x86_pmu_state: n_events: 2, n_added: 2, n_txn: 2
>    pec_1076_warn-2804  [000] d...   147.926184: x86_pmu_state: Assignment: {
>    pec_1076_warn-2804  [000] d...   147.926186: x86_pmu_state:   0->33 tag: 1 config: 0 (ffff88011ac99800)
>    pec_1076_warn-2804  [000] d...   147.926188: x86_pmu_state:   1->0 tag: 1 config: 1 (ffff880119ec8800)
>    pec_1076_warn-2804  [000] d...   147.926188: x86_pmu_state: }
>    pec_1076_warn-2804  [000] d...   147.926190: x86_pmu_enable: S0: hwc->idx: 33, hwc->last_cpu: 0, hwc->last_tag: 1 hwc->state: 0

So the problem is that x86_pmu_del(), when called from a
group_sched_in() that fails (for whatever reason), and without x86_pmu
TXN support (because the leader is !x86_pmu), will corrupt the n_added
state.

Reported-and-Tested-by: Vince Weaver <vincent.weaver@maine.edu>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Dave Jones <davej@redhat.com>
Link: http://lkml.kernel.org/r/20140221150312.GF3104@twins.programming.kicks-ass.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

perf trace: Fix ioctl 'request' beautifier build problems on !(i386 || x86_64) arches

commit 844ae5b46c08dbc7ba695b543c023f9cf3bbf9ff upstream.

Supporting decoding the ioctl 'request' parameter needs more work to
properly support more architectures, the current approach doesn't work
on at least powerpc and sparc, as reported by Ben Hutchings in
http://lkml.kernel.org/r/1391593985.3003.48.camel@deadeye.wl.decadent.org.uk .

Work around that by making it to be ifdefed for the architectures known
to work with the current, limited approach, i386 and x86_64 till better
code is written.

Reported-by: Ben Hutchings <ben@decadent.org.uk>
Acked-by: Ben Hutchings <ben@decadent.org.uk>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/n/tip-ss04k11insqlu329xh5g02q0@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

x86: dma-mapping: fix GFP_ATOMIC macro usage

commit c091c71ad2218fc50a07b3d1dab85783f3b77efd upstream.

GFP_ATOMIC is not a single gfp flag, but a macro which expands to the other
flags, where meaningful is the LACK of __GFP_WAIT flag. To check if caller
wants to perform an atomic allocation, the code must test for a lack of the
__GFP_WAIT flag. This patch fixes the issue introduced in v3.5-rc1.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ahci: disable NCQ on Samsung pci-e SSDs on macbooks

commit 67809f85d31eac600f6b28defa5386c9d2a13b1d upstream.

Samsung's pci-e SSDs with device ID 0x1600 which are found on some
macbooks time out on NCQ commands.  Blacklist NCQ on the device so
that the affected machines can at least boot.

Original-patch-by: Levente Kurusa <levex@linux.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=60731
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

powerpc/powernv: Fix indirect XSCOM unmangling

commit e0cf957614976896111e676e5134ac98ee227d3d upstream.

We need to unmangle the full address, not just the register
number, and we also need to support the real indirect bit
being set for in-kernel uses.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

powerpc/powernv: Fix opal_xscom_{read,write} prototype

commit 2f3f38e4d3d03dd4125cc9a1f49ab3cc91d8d670 upstream.

The OPAL firmware functions opal_xscom_read and opal_xscom_write
take a 64-bit argument for the XSCOM (PCB) address in order to
support the indirect mode on P8.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

powerpc/crashdump : Fix page frame number check in copy_oldmem_page

commit f5295bd8ea8a65dc5eac608b151386314cb978f1 upstream.

In copy_oldmem_page, the current check using max_pfn and min_low_pfn to
decide if the page is backed or not, is not valid when the memory layout is
not continuous.

This happens when running as a QEMU/KVM guest, where RTAS is mapped higher
in the memory. In that case max_pfn points to the end of RTAS, and a hole
between the end of the kdump kernel and RTAS is not backed by PTEs. As a
consequence, the kdump kernel is crashing in copy_oldmem_page when accessing
in a direct way the pages in that hole.

This fix relies on the memblock's service memblock_is_region_memory to
check if the read page is part or not of the directly accessible memory.

Signed-off-by: Laurent Dufour <ldufour@linux.vnet.ibm.com>
Tested-by: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

powerpc/le: Ensure that the 'stop-self' RTAS token is handled correctly

commit 41dd03a94c7d408d2ef32530545097f7d1befe5c upstream.

Currently we're storing a host endian RTAS token in
rtas_stop_self_args.token.  We then pass that directly to rtas.  This is
fine on big endian however on little endian the token is not what we
expect.

This will typically result in hitting:
panic("Alas, I survived.\n");

To fix this we always use the stop-self token in host order and always
convert it to be32 before passing this to rtas.

Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

powerpc: Increase stack redzone for 64-bit userspace to 512 bytes

commit 573ebfa6601fa58b439e7f15828762839ccd306a upstream.

The new ELFv2 little-endian ABI increases the stack redzone -- the
area below the stack pointer that can be used for storing data --
from 288 bytes to 512 bytes.  This means that we need to allow more
space on the user stack when delivering a signal to a 64-bit process.

To make the code a bit clearer, we define new USER_REDZONE_SIZE and
KERNEL_REDZONE_SIZE symbols in ptrace.h.  For now, we leave the
kernel redzone size at 288 bytes, since increasing it to 512 bytes
would increase the size of interrupt stack frames correspondingly.

Gcc currently only makes use of 288 bytes of redzone even when
compiling for the new little-endian ABI, and the kernel cannot
currently be compiled with the new ABI anyway.

In the future, hopefully gcc will provide an option to control the
amount of redzone used, and then we could reduce it even more.

This also changes the code in arch_compat_alloc_user_space() to
preserve the expanded redzone.  It is not clear why this function would
ever be used on a 64-bit process, though.

Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

SUNRPC: Ensure that gss_auth isn't freed before its upcall messages

commit 9eb2ddb48ce3a7bd745c14a933112994647fa3cd upstream.

Fix a race in which the RPC client is shutting down while the
gss daemon is processing a downcall. If the RPC client manages to
shut down before the gss daemon is done, then the struct gss_auth
used in gss_release_msg() may have already been freed.

Link: http://lkml.kernel.org/r/1392494917.71728.YahooMailNeo@web140002.mail.bf1.yahoo.com
Reported-by: John <da_audiophile@yahoo.com>
Reported-by: Borislav Petkov <bp@alien8.de>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

SUNRPC: Fix races in xs_nospace()

commit 06ea0bfe6e6043cb56a78935a19f6f8ebc636226 upstream.

When a send failure occurs due to the socket being out of buffer space,
we call xs_nospace() in order to have the RPC task wait until the
socket has drained enough to make it worth while trying again.
The current patch fixes a race in which the socket is drained before
we get round to setting up the machinery in xs_nospace(), and which
is reported to cause hangs.

Link: http://lkml.kernel.org/r/20140210170315.33dfc621@notabene.brown
Fixes: a9a6b52ee1ba (SUNRPC: Don't start the retransmission timer...)
Reported-by: Neil Brown <neilb@suse.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ASoC: wm8958-dsp: Fix firmware block loading

commit 548da08fc1e245faf9b0d7c41ecd8e07984fc332 upstream.

The codec->control_data contains a pointer to the device's regmap struct. But
wm8994_bulk_write() expects a pointer to the parent wm8998 device.

The issue was introduced in commit d9a7666f ("ASoC: Remove ASoC-specific
WM8994 I/O code").

Fixes: d9a7666f ("ASoC: Remove ASoC-specific WM8994 I/O code")
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ASoC: sta32x: Fix array access overflow

commit 025c3fa9256d4c54506b7a29dc3befac54f5c68d upstream.

Preset EQ enum of sta32x codec driver declares too many number of
items and it may lead to the access over the actual array size.

Use SOC_ENUM_SINGLE_DECL() helper and it's automatically fixed.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Acked-by: Liam Girdwood <liam.r.girdwood@linux.intel.com>
Acked-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ASoC: sta32x: Fix wrong enum for limiter2 release rate

commit b3619b288b621e63f66908045f48495869a996a6 upstream.

There is a typo in the Limiter2 Release Rate control, a wrong enum for
Limiter1 is assigned.  It must point to Limiter2.
Spotted by a compile warning:

In file included from sound/soc/codecs/sta32x.c:34:0:
sound/soc/codecs/sta32x.c:223:29: warning: ‘sta32x_limiter2_release_rate_enum’ defined but not used [-Wunused-variable]
 static SOC_ENUM_SINGLE_DECL(sta32x_limiter2_release_rate_enum,
                             ^
include/sound/soc.h:275:18: note: in definition of macro ‘SOC_ENUM_DOUBLE_DECL’
  struct soc_enum name = SOC_ENUM_DOUBLE(xreg, xshift_l, xshift_r, \
                  ^
sound/soc/codecs/sta32x.c:223:8: note: in expansion of macro ‘SOC_ENUM_SINGLE_DECL’
 static SOC_ENUM_SINGLE_DECL(sta32x_limiter2_release_rate_enum,
        ^

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Mark Brown <broonie@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ASoC: sta32x: Fix cache sync

commit 70ff00f82a6af0ff68f8f7b411738634ce2f20d0 upstream.

codec->control_data contains a pointer to the regmap struct of the device, not
to the device private data. Use snd_soc_codec_get_drvdata() instead.

The issue was introduced in commit 29fdf4fbbe ("ASoC: sta32x: Convert to
regmap").

Fixes: 29fdf4fbbe (ASoC: sta32x: Convert to regmap)
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ASoC: da732x: Mark DC offset control registers volatile

commit 75306820248e26d15d84acf4e297b9fb27dd3bb2 upstream.

The driver reads from the DC offset control registers during callibration
but since the registers are marked as volatile and there is a register
cache the values will not be read from the hardware after the first reading
rendering the callibration ineffective.

It appears that the driver was originally written for the ASoC level
register I/O code but converted to regmap prior to merge and this issue
was missed during the conversion as the framework level volatile register
functionality was not being used.

Signed-off-by: Mark Brown <broonie@linaro.org>
Acked-by: Adam Thomson <Adam.Thomson.Opensource@diasemi.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ASoC: wm8770: Fix wrong number of enum items

commit 7a6c0a58dc824523966f212c76322d47c5b0e6fe upstream.

wm8770 codec driver defines ain_enum with a wrong number of items.

Use SOC_ENUM_DOUBLE_DECL() macro and it's automatically fixed.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Acked-by: Liam Girdwood <liam.r.girdwood@linux.intel.com>
Acked-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Acked-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ASoC: max98090: sync regcache on entering STANDBY

commit c42c8922c46d33ed769e99618bdfba06866a0c72 upstream.

Sync regcache when entering STANDBY from OFF.  ON isn't entered with
OFF as the current state, so the registers were not being re-synced
after suspend/resume.

The 98088 and 98095 already call regcache_sync from STANDBY.

Signed-off-by: Dylan Reid <dgreid@chromium.org>
Signed-off-by: Mark Brown <broonie@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ASoC: fsl: fix pm support of machine drivers

commit 47cf84e17ebb79a20e6244b954c4ea4e18a82d43 upstream.

The commit 1abe729 (ASoC: fsl: Add missing pm to current machine
drivers) enables pm support for a few IMX machine drivers.  But it does
not update dev drvdata to be the pointer to 'card'.  This causes the
kernel dump below in system suspend, because snd_soc_suspend() expects
that the dev drvdata points to 'card', while it still points to the
private data of machine driver.

This patch fixes imx-sgtl5000 and imx-wm8962 by attaching 'card' to dev
drvdata and private data to card drvdata.  For imx-mc13783, I simply
revert the pm change because it must be broken for the same reason and
I don't have hardware to test pm enabling code.

$ echo mem > /sys/power/state
PM: Syncing filesystems ... done.
PM: Preparing system for mem sleep
mmc1: card e624 removed
Freezing user space processes ... (elapsed 0.002 seconds) done.
Freezing remaining freezable tasks ... (elapsed 0.002 seconds) done.
PM: Entering mem sleep
INFO: trying to register non-static key.
the code is fine but needs lockdep annotation.
turning off the locking correctness validator.
CPU: 0 PID: 1861 Comm: bash Not tainted 3.14.0-rc1+ #1648
Backtrace:
[<80012144>] (dump_backtrace) from [<800122e4>] (show_stack+0x18/0x1c)
 r6:8079c77c r5:00000c5a r4:00000000 r3:00000000
[<800122cc>] (show_stack) from [<80637ac0>] (dump_stack+0x78/0x94)
[<80637a48>] (dump_stack) from [<80028918>] (warn_slowpath_common+0x6c/0x8c)
 r4:bdb21c38 r3:be62df00
[<800288ac>] (warn_slowpath_common) from [<800289dc>] (warn_slowpath_fmt+0x38/0x40)
 r8:be62e3a8 r7:bf122960 r6:00000005 r5:00000000 r4:00000000
[<800289a8>] (warn_slowpath_fmt) from [<8006518c>] (__lock_acquire+0x1ae0/0x1ce0)
 r3:8079d598 r2:80799e70
[<800636ac>] (__lock_acquire) from [<80065894>] (lock_acquire+0x68/0x7c)
 r10:bdb20000 r9:be62df00 r8:00000000 r7:00000000 r6:60000013 r5:bdb20000
 r4:00000000
[<8006582c>] (lock_acquire) from [<8063c938>] (mutex_lock_nested+0x5c/0x3b8)
 r7:00000000 r6:80dfc78c r5:804be444 r4:bf122928
[<8063c8dc>] (mutex_lock_nested) from [<804be444>] (snd_soc_suspend+0x34/0x42c)
 r10:00000000 r9:00000000 r8:00000000 r7:bf1c4444 r6:bf1c4410 r5:be978150
 r4:be978010
[<804be410>] (snd_soc_suspend) from [<8034392c>] (platform_pm_suspend+0x34/0x64)
 r10:00000000 r8:00000000 r7:bf1c4444 r6:bf1c4410 r5:803438f8 r4:bf1c4410
[<803438f8>] (platform_pm_suspend) from [<80348e18>] (dpm_run_callback.isra.7+0x34/0x6c)
[<80348de4>] (dpm_run_callback.isra.7) from [<80349354>] (__device_suspend+0x10c/0x220)
 r9:808dd974 r8:808c4a5c r6:00000002 r5:80e5001c r4:bf1c4410
[<80349248>] (__device_suspend) from [<8034a338>] (dpm_suspend+0x60/0x220)
 r7:bf1c4410 r6:808dd90c r5:80e5001c r4:bf1c44c0
[<8034a2d8>] (dpm_suspend) from [<8034a790>] (dpm_suspend_start+0x60/0x68)
 r10:8079a818 r9:00000000 r8:00000004 r7:80dfbe90 r6:80641eec r5:00000000
 r4:00000002
[<8034a730>] (dpm_suspend_start) from [<8006a788>] (suspend_devices_and_enter+0x74/0x318)
 r4:00000003 r3:80dfbe98
[<8006a714>] (suspend_devices_and_enter) from [<8006abd8>] (pm_suspend+0x1ac/0x244)
 r10:8079a818 r8:00000004 r7:00000003 r6:80641eec r5:00000000 r4:00000003
[<8006aa2c>] (pm_suspend) from [<80069a4c>] (state_store+0x70/0xc0)
 r5:00000003 r4:bd85ea40
[<800699dc>] (state_store) from [<80294034>] (kobj_attr_store+0x1c/0x28)
 r10:beb9fe08 r8:00000000 r7:bdb21f78 r6:bd85ea40 r5:00000004 r4:beb9fe00
[<80294018>] (kobj_attr_store) from [<80140f90>] (sysfs_kf_write+0x54/0x58)
[<80140f3c>] (sysfs_kf_write) from [<8014474c>] (kernfs_fop_write+0xc4/0x160)
 r6:bd85ea40 r5:beb9fe00 r4:00000004 r3:80140f3c
[<80144688>] (kernfs_fop_write) from [<800dfa14>] (vfs_write+0xbc/0x184)
 r10:00000000 r9:00000000 r8:00000000 r7:bdb21f78 r6:00500c08 r5:00000004
 r4:be782600
[<800df958>] (vfs_write) from [<800dfe00>] (SyS_write+0x48/0x70)
 r10:00000000 r8:00000000 r7:00000004 r6:00500c08 r5:00000000 r4:be782600
[<800dfdb8>] (SyS_write) from [<8000e800>] (ret_fast_syscall+0x0/0x48)
 r9:bdb20000 r8:8000e9c4 r7:00000004 r6:00500c08 r5:00000004 r4:76eb65e0

Fixes: 1abe729 (ASoC: fsl: Add missing pm to current machine drivers)
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Signed-off-by: Mark Brown <broonie@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ASoC: txx9aclc_ac97: Fix kernel crash on probe

commit 9febd494d15c4a351e9c9cae7184643144eea892 upstream.

This patch fixes a crash caused by commit 3bed3344c826
(ASoC: txx9aclc_ac97: Convert to devm_ioremap_resource()).
This is an attempt to assign "drvdata->base" while memory
for "drvdata" is not already allocated.

Fixes: 3bed3344c826 (ASoC: txx9aclc_ac97: Convert to devm_ioremap_resource())
Signed-off-by: Alexander Shiyan <shc_work@mail.ru>
Signed-off-by: Mark Brown <broonie@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ASoC: rt5640: Add ACPI ID for Intel Baytrail

commit b31b2b6d5de71c569413d8dc4f7b050cbe25a09e upstream.

Realtek RT5640 uses ACPI ID "10EC5640" for Intel Baytrail platforms.

Signed-off-by: Jarkko Nikula <jarkko.nikula@linux.intel.com>
Signed-off-by: Mark Brown <broonie@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ASoC: da9055: Fix device registration of PMIC and CODEC devices

commit 07b0e5b10258b48e5edfb6c8ac156f05510eb775 upstream.

Currently the I2C device Ids conflict for the MFD and CODEC so
cannot be both instantiated on one platform. This patch updates
the Ids and names to make them unique from each other.

It should be noted that the I2C addresses for both PMIC and CODEC
are modifiable so instantiation of the two are kept as separate
devices, rather than instantiating the CODEC from the MFD code.

Signed-off-by: Adam Thomson <Adam.Thomson.Opensource@diasemi.com>
Acked-by: Mark Brown <broonie@linaro.org>
Signed-off-by: Mark Brown <broonie@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

kvm, vmx: Really fix lazy FPU on nested guest

commit 1b385cbdd74aa803e966e01e5fe49490d6044e30 upstream.

Commit e504c9098ed6 (kvm, vmx: Fix lazy FPU on nested guest, 2013-11-13)
highlighted a real problem, but the fix was subtly wrong.

nested_read_cr0 is the CR0 as read by L2, but here we want to look at
the CR0 value reflecting L1's setup.  In other words, L2 might think
that TS=0 (so nested_read_cr0 has the bit clear); but if L1 is actually
running it with TS=1, we should inject the fault into L1.

The effective value of CR0 in L2 is contained in vmcs12->guest_cr0, use
it.

Fixes: e504c9098ed6acd9e1079c5e10e4910724ad429f
Reported-by: Kashyap Chamarty <kchamart@redhat.com>
Reported-by: Stefan Bader <stefan.bader@canonical.com>
Tested-by: Kashyap Chamarty <kchamart@redhat.com>
Tested-by: Anthoine Bourgeois <bourgeois@bertin.fr>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

kvm: x86: fix emulator buffer overflow (CVE-2014-0049)

commit a08d3b3b99efd509133946056531cdf8f3a0c09b upstream.

The problem occurs when the guest performs a pusha with the stack
address pointing to an mmio address (or an invalid guest physical
address) to start with, but then extending into an ordinary guest
physical address.  When doing repeated emulated pushes
emulator_read_write sets mmio_needed to 1 on the first one.  On a
later push when the stack points to regular memory,
mmio_nr_fragments is set to 0, but mmio_is_needed is not set to 0.

As a result, KVM exits to userspace, and then returns to
complete_emulated_mmio.  In complete_emulated_mmio
vcpu->mmio_cur_fragment is incremented.  The termination condition of
vcpu->mmio_cur_fragment == vcpu->mmio_nr_fragments is never achieved.
The code bounces back and fourth to userspace incrementing
mmio_cur_fragment past it's buffer.  If the guest does nothing else it
eventually leads to a a crash on a memcpy from invalid memory address.

However if a guest code can cause the vm to be destroyed in another
vcpu with excellent timing, then kvm_clear_async_pf_completion_queue
can be used by the guest to control the data that's pointed to by the
call to cancel_work_item, which can be used to gain execution.

Fixes: f78146b0f9230765c6315b2e14f56112513389ad
Signed-off-by: Andrew Honig <ahonig@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

export: declare ksymtab symbols

commit 7b4ec8dd7d4ac467e9eee4d49f2c9574d773efbb upstream.

sparse complains about any __ksymtab symbols with the following:

 warning: symbol '__ksymtab_...' was not declared. Should it be static?

due to Andi's patch making it non-static.

Mollify sparse by declaring the symbol extern, otherwise we get
drowned in sparse warnings for anything that uses EXPORT_SYMBOL
in the sources, making it easy to miss real warnings.

Fixes: e0f244c63fc9 ("asmlinkage, module: Make ksymtab [...] __visible")
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Acked-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

fs: fix iversion handling

commit dff6efc326a4d5f305797d4a6bba14f374fdd633 upstream.

Currently notify_change directly updates i_version for size updates,
which not only is counter to how all other fields are updated through
struct iattr, but also breaks XFS, which need inode updates to happen
under its own lock, and synchronized to the structure that gets written
to the log.

Remove the update in the common code, and it to btrfs and ext4,
XFS already does a proper updaste internally and currently gets a
double update with the existing code.

IMHO this is 3.13 and -stable material and should go in through the XFS
tree.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Andreas Dilger <adilger@dilger.ca>
Acked-by: Jan Kara <jack@suse.cz>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Chris Mason <clm@fb.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

cgroup: update cgroup_enable_task_cg_lists() to grab siglock

commit 532de3fc72adc2a6525c4d53c07bf81e1732083d upstream.

Currently, there's nothing preventing cgroup_enable_task_cg_lists()
from missing set PF_EXITING and race against cgroup_exit().  Depending
on the timing, cgroup_exit() may finish with the task still linked on
css_set leading to list corruption.  Fix it by grabbing siglock in
cgroup_enable_task_cg_lists() so that PF_EXITING is guaranteed to be
visible.

This whole on-demand cg_list optimization is extremely fragile and has
ample possibility to lead to bugs which can cause things like
once-a-year oops during boot.  I'm wondering whether the better
approach would be just adding "cgroup_disable=all" handling which
disables the whole cgroup rather than tempting fate with this
on-demand craziness.

Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: Li Zefan <lizefan@huawei.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

cgroup: fix locking in cgroup_cfts_commit()

commit 48573a893303986e3b0b2974d6fb11f3d1bb7064 upstream.

cgroup_cfts_commit() walks the cgroup hierarchy that the target
subsystem is attached to and tries to apply the file changes.  Due to
the convolution with inode locking, it can't keep cgroup_mutex locked
while iterating.  It currently holds only RCU read lock around the
actual iteration and then pins the found cgroup using dget().

Unfortunately, this is incorrect.  Although the iteration does check
cgroup_is_dead() before invoking dget(), there's nothing which
prevents the dentry from going away inbetween.  Note that this is
different from the usual css iterations where css_tryget() is used to
pin the css - css_tryget() tests whether the css can be pinned and
fails if not.

The problem can be solved by simply holding cgroup_mutex instead of
RCU read lock around the iteration, which actually reduces LOC.

Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: Li Zefan <lizefan@huawei.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

cgroup: fix error return from cgroup_create()

commit b58c89986a77a23658682a100eb15d8edb571ebb upstream.

cgroup_create() was returning 0 after allocation failures.  Fix it.

Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: Li Zefan <lizefan@huawei.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

cgroup: fix error return value in cgroup_mount()

commit eb46bf89696972b856a9adb6aebd5c7b65c266e4 upstream.

When cgroup_mount() fails to allocate an id for the root, it didn't
set ret before jumping to unlock_drop ending up returning 0 after a
failure.  Fix it.

Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: Li Zefan <lizefan@huawei.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ALSA: hda - Enable front audio jacks on one HP desktop model

commit 1de7ca5e844866f56bebb2fc47fa18e090677e88 upstream.

The front headphone and mic jackes on a HP desktop model (Vendor Id:
0x111d76c7 Subsystem Id: 0x103c2b17) can not work, the codec on this
machine has 8 physical ports, 6 of them are routed to rear jackes
and all of them work very well, while the remaining 2 ports are
routed to front headphone and mic jackes, but the corresponding
pin complex node are not defined correctly.

After apply this fix, the front audio jackes can work very well.

[trivial fix of enum definition by tiwai]

BugLink: https://bugs.launchpad.net/bugs/1282369
Cc: David Henningsson <david.henningsson@canonical.com>
Tested-by: Gerald Yang <gerald.yang@canonical.com>
Signed-off-by: Hui Wang <hui.wang@canonical.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ALSA: hda/ca0132 - Fix recording from mode id 0x8

commit 13c12dbe3a2ce17227f7ddef652b6a53c78fa51f upstream.

Incorrect ADC is picked in ca0132_capture_pcm_prepare(),
where it assumes multiple streams while there is one stream
per ADC. Note that ca0132_capture_pcm_cleanup() already does
the right thing.

The Chromebook Pixel has a microphone under the keyboard that
is attached to node id 0x8. Before this fix, recording would
always go to the main internal mic (node id 0x7).

Signed-off-by: Hsin-Yu Chao <hychao@chromium.org>
Reviewed-by: Dylan Reid <dgreid@chromium.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ALSA: hda/ca0132 - setup/cleanup streams

commit 28fba95087a7f3d107a3a6728aef7dbfaf3fd782 upstream.

When a HDMI stream is opened with the same stream tag
as a following opened stream to ca0132, audio will be
heard from two ports simultaneously.
Fix this issue by change to use snd_hda_codec_setup_stream
and snd_hda_codec_cleanup_stream instead, so that an
inactive stream can be marked as 'dirty' when found
with a conflict stream tag, and then get purified.

Signed-off-by: Hsin-Yu Chao <hychao@chromium.org>
Reviewed-by: Chih-Chung Chang <chihchung@chromium.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ALSA: hda - add headset mic detect quirks for two Dell laptops

commit 4913e0bf239dafee356bc7fab61806cc2518930c upstream.

When we plug a 3-ring headset on the Dell machines (Vendor ID:
0x10ec0255, Subsystem ID: 0x10280657; Vendor ID: 0x10ec0255,
Subsystem ID: 0x1028065f), the headset mic can't be
detected, after apply this patch, the headset mic can work well.

BugLink: https://bugs.launchpad.net/bugs/1260303
Cc: David Henningsson <david.henningsson@canonical.com>
Tested-by: Cyrus Lien <cyrus.lien@canonical.com>
Signed-off-by: Hui Wang <hui.wang@canonical.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ALSA: usb-audio: work around KEF X300A firmware bug

commit 624aef494f86ed0c58056361c06347ad62b26806 upstream.

When the driver tries to access Function Unit 10, the KEF X300A
speakers' firmware apparently locks up, making even PCM streaming
impossible.  Work around this by ignoring this FU.

Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

batman-adv: fix potential kernel paging error for unicast transmissions

[ Upstream commit 70b271a78beba787155d6696aacd7c4d4a251c50 ]

batadv_send_skb_prepare_unicast(_4addr) might reallocate the
skb's data. If it does then our ethhdr pointer is not valid
anymore in batadv_send_skb_unicast(), resulting in a kernel
paging error.

Fixing this by refetching the ethhdr pointer after the
potential reallocation.

Signed-off-by: Linus Lüssing <linus.luessing@web.de>
Signed-off-by: Antonio Quartulli <antonio@meshcoding.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

batman-adv: avoid double free when orig_node initialization fails

[ Upstream commit a5a5cb8cab526af2f6cbe9715f8ca843192f0d81 ]

In the failure path of the orig_node initialization routine
the orig_node->bat_iv.bcast_own field is free'd twice: first
in batadv_iv_ogm_orig_get() and then later in
batadv_orig_node_free_rcu().

Fix it by removing the kfree in batadv_iv_ogm_orig_get().

Signed-off-by: Antonio Quartulli <antonio@meshcoding.com>
Signed-off-by: Marek Lindner <mareklindner@neomailbox.ch>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

batman-adv: free skb on TVLV parsing success

[ Upstream commit 05c3c8a636aa9ee35ce13f65afc5b665615cc786 ]

When the TVLV parsing routine succeed the skb is left
untouched thus leading to a memory leak.

Fix this by consuming the skb in case of success.

Introduced by ef26157747d42254453f6b3ac2bd8bd3c53339c3
("batman-adv: tvlv - basic infrastructure")

Reported-by: Russel Senior <russell@personaltelco.net>
Signed-off-by: Antonio Quartulli <antonio@open-mesh.com>
Tested-by: Russell Senior <russell@personaltelco.net>
Signed-off-by: Marek Lindner <mareklindner@neomailbox.ch>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

batman-adv: fix TT CRC computation by ensuring byte order

[ Upstream commit a30e22ca8464c2dc573e0144a972221c2f06c2cd ]

When computing the CRC on a 2byte variable the order of
the bytes obviously alters the final result. This means
that computing the CRC over the same value on two archs
having different endianess leads to different numbers.

The global and local translation table CRC computation
routine makes this mistake while processing the clients
VIDs. The result is a continuous CRC mismatching between
nodes having different endianess.

Fix this by converting the VID to Network Order before
processing it. This guarantees that every node uses the same
byte order.

Introduced by 7ea7b4a142758deaf46c1af0ca9ceca6dd55138b
("batman-adv: make the TT CRC logic VLAN specific")

Reported-by: Russel Senior <russell@personaltelco.net>
Signed-off-by: Antonio Quartulli <antonio@open-mesh.com>
Tested-by: Russell Senior <russell@personaltelco.net>
Signed-off-by: Marek Lindner <mareklindner@neomailbox.ch>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

batman-adv: fix potential orig_node reference leak

[ Upstream commit b2262df7fcf2c395eca564df83238e931d88d7bf ]

Since batadv_orig_node_new() sets the refcount to two, assuming that
the calling function will use a reference for putting the orig_node into
a hash or similar, both references must be freed if initialization of
the orig_node fails. Otherwise that object may be leaked in that error
case.

Reported-by: Antonio Quartulli <antonio@meshcoding.com>
Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de>
Signed-off-by: Marek Lindner <mareklindner@neomailbox.ch>
Signed-off-by: Antonio Quartulli <antonio@meshcoding.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

batman-adv: avoid potential race condition when adding a new neighbour

[ Upstream commit 08bf0ed29c7ded45c477d08618220dd200c3524a ]

When adding a new neighbour it is important to atomically
perform the following:
- check if the neighbour already exists
- append the neighbour to the proper list

If the two operations are not performed in an atomic context
it is possible that two concurrent insertions add the same
neighbour twice.

Signed-off-by: Antonio Quartulli <antonio@open-mesh.com>
Signed-off-by: Marek Lindner <mareklindner@neomailbox.ch>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

batman-adv: properly check pskb_may_pull return value

[ Upstream commit f1791425cf0bcda43ab9a9a37df1ad3ccb1f6654 ]

pskb_may_pull() returns 1 on success and 0 in case of failure,
therefore checking for the return value being negative does
not make sense at all.

This way if the function fails we will probably read beyond the current
skb data buffer. Fix this by doing the proper check.

Signed-off-by: Antonio Quartulli <antonio@meshcoding.com>
Signed-off-by: Marek Lindner <mareklindner@neomailbox.ch>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

batman-adv: release vlan object after checking the CRC

[ Upstream commit 91c2b1a9f680ff105369d49abc7e19ca7efb33e1 ]

There is a refcounter unbalance in the CRC checking routine
invoked on OGM reception. A vlan object is retrieved (thus
its refcounter is increased by one) but it is never properly
released. This leads to a memleak because the vlan object
will never be free'd.

Fix this by releasing the vlan object after having read the
CRC.

Reported-by: Russell Senior <russell@personaltelco.net>
Reported-by: Daniel <daniel@makrotopia.org>
Reported-by: cmsv <cmsv@wirelesspt.net>
Signed-off-by: Antonio Quartulli <antonio@meshcoding.com>
Signed-off-by: Marek Lindner <mareklindner@neomailbox.ch>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

batman-adv: fix TT-TVLV parsing on OGM reception

[ Upstream commit e889241f45f9cecbc84a6ffed577083ab52e62ee ]

When accessing a TT-TVLV container in the OGM RX path
the variable pointing to the list of changes to apply is
altered by mistake.

This makes the TT component read data at the wrong position
in the OGM packet buffer.

Fix it by removing the bogus pointer alteration.

Signed-off-by: Antonio Quartulli <antonio@meshcoding.com>
Signed-off-by: Marek Lindner <mareklindner@neomailbox.ch>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

batman-adv: fix soft-interface MTU computation

[ Upstream commit 930cd6e46eadce8b8ed2a232ee536e5fd286c152 ]

The current MTU computation always returns a value
smaller than 1500bytes even if the real interfaces
have an MTU large enough to compensate the batman-adv
overhead.

Fix the computation by properly returning the highest
admitted value.

Introduced by a19d3d85e1b854e4a483a55d740a42458085560d
("batman-adv: limit local translation table max size")

Reported-by: Russell Senior <russell@personaltelco.net>
Signed-off-by: Antonio Quartulli <antonio@meshcoding.com>
Signed-off-by: Marek Lindner <mareklindner@neomailbox.ch>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

net: use __GFP_NORETRY for high order allocations

[ Upstream commit ed98df3361f059db42786c830ea96e2d18b8d4db ]

sock_alloc_send_pskb() & sk_page_frag_refill()
have a loop trying high order allocations to prepare
skb with low number of fragments as this increases performance.

Problem is that under memory pressure/fragmentation, this can
trigger OOM while the intent was only to try the high order
allocations, then fallback to order-0 allocations.

We had various reports from unexpected regressions.

According to David, setting __GFP_NORETRY should be fine,
as the asynchronous compaction is still enabled, and this
will prevent OOM from kicking as in :

CFSClientEventm invoked oom-killer: gfp_mask=0x42d0, order=3, oom_adj=0,
oom_score_adj=0, oom_score_badness=2 (enabled),memcg_scoring=disabled
CFSClientEventm

Call Trace:
 [<ffffffff8043766c>] dump_header+0xe1/0x23e
 [<ffffffff80437a02>] oom_kill_process+0x6a/0x323
 [<ffffffff80438443>] out_of_memory+0x4b3/0x50d
 [<ffffffff8043a4a6>] __alloc_pages_may_oom+0xa2/0xc7
 [<ffffffff80236f42>] __alloc_pages_nodemask+0x1002/0x17f0
 [<ffffffff8024bd23>] alloc_pages_current+0x103/0x2b0
 [<ffffffff8028567f>] sk_page_frag_refill+0x8f/0x160
 [<ffffffff80295fa0>] tcp_sendmsg+0x560/0xee0
 [<ffffffff802a5037>] inet_sendmsg+0x67/0x100
 [<ffffffff80283c9c>] __sock_sendmsg_nosec+0x6c/0x90
 [<ffffffff80283e85>] sock_sendmsg+0xc5/0xf0
 [<ffffffff802847b6>] __sys_sendmsg+0x136/0x430
 [<ffffffff80284ec8>] sys_sendmsg+0x88/0x110
 [<ffffffff80711472>] system_call_fastpath+0x16/0x1b
Out of Memory: Kill process 2856 (bash) score 9999 or sacrifice child

Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: David Rientjes <rientjes@google.com>
Acked-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

net: mvneta: replace Tx timer with a real interrupt

[ Upstream commit 71f6d1b31fb1f278a345a30a2180515adc7d80ae ]

Right now the mvneta driver doesn't handle Tx IRQ, and relies on two
mechanisms to flush Tx descriptors : a flush at the end of mvneta_tx()
and a timer. If a burst of packets is emitted faster than the device
can send them, then the queue is stopped until next wake-up of the
timer 10ms later. This causes jerky output traffic with bursts and
pauses, making it difficult to reach line rate with very few streams.

A test on UDP traffic shows that it's not possible to go beyond 134
Mbps / 12 kpps of outgoing traffic with 1500-bytes IP packets. Routed
traffic tends to observe pauses as well if the traffic is bursty,
making it even burstier after the wake-up.

It seems that this feature was inherited from the original driver but
nothing there mentions any reason for not using the interrupt instead,
which the chip supports.

Thus, this patch enables Tx interrupts and removes the timer. It does
the two at once because it's not really possible to make the two
mechanisms coexist, so a split patch doesn't make sense.

First tests performed on a Mirabox (Armada 370) show that less CPU
seems to be used when sending traffic. One reason might be that we now
call the mvneta_tx_done_gbe() with a mask indicating which queues have
been done instead of looping over all of them.

The same UDP test above now happily reaches 987 Mbps / 87.7 kpps.
Single-stream TCP traffic can now more easily reach line rate. HTTP
transfers of 1 MB objects over a single connection went from 730 to
840 Mbps. It is even possible to go significantly higher (>900 Mbps)
by tweaking tcp_tso_win_divisor.

Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Gregory CLEMENT <gregory.clement@free-electrons.com>
Cc: Arnaud Ebalard <arno@natisbad.org>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Tested-by: Arnaud Ebalard <arno@natisbad.org>
Signed-off-by: Willy Tarreau <w@1wt.eu>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

net: mvneta: add missing bit descriptions for interrupt masks and causes

[ Upstream commit 40ba35e74fa56866918d2f3bc0528b5b92725d5e ]

Marvell has not published the chip's datasheet yet, so it's very hard
to find the relevant bits to manipulate to change the IRQ behaviour.
Fortunately, these bits are described in the proprietary LSP patch set
which is publicly available here :

    http://www.plugcomputer.org/downloads/mirabox/

So let's put them back in the driver in order to reduce the burden of
current and future maintenance.

Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Gregory CLEMENT <gregory.clement@free-electrons.com>
Tested-by: Arnaud Ebalard <arno@natisbad.org>
Signed-off-by: Willy Tarreau <w@1wt.eu>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

net: mvneta: do not schedule in mvneta_tx_timeout

[ Upstream commit 290213667ab53a95456397763205e4b1e30f46b5 ]

If a queue timeout is reported, we can oops because of some
schedules while the caller is atomic, as shown below :

  mvneta d0070000.ethernet eth0: tx timeout
  BUG: scheduling while atomic: bash/1528/0x00000100
  Modules linked in: slhttp_ethdiv(C) [last unloaded: slhttp_ethdiv]
  CPU: 2 PID: 1528 Comm: bash Tainted: G        WC   3.13.0-rc4-mvebu-nf #180
  [<c0011bd9>] (unwind_backtrace+0x1/0x98) from [<c000f1ab>] (show_stack+0xb/0xc)
  [<c000f1ab>] (show_stack+0xb/0xc) from [<c02ad323>] (dump_stack+0x4f/0x64)
  [<c02ad323>] (dump_stack+0x4f/0x64) from [<c02abe67>] (__schedule_bug+0x37/0x4c)
  [<c02abe67>] (__schedule_bug+0x37/0x4c) from [<c02ae261>] (__schedule+0x325/0x3ec)
  [<c02ae261>] (__schedule+0x325/0x3ec) from [<c02adb97>] (schedule_timeout+0xb7/0x118)
  [<c02adb97>] (schedule_timeout+0xb7/0x118) from [<c0020a67>] (msleep+0xf/0x14)
  [<c0020a67>] (msleep+0xf/0x14) from [<c01dcbe5>] (mvneta_stop_dev+0x21/0x194)
  [<c01dcbe5>] (mvneta_stop_dev+0x21/0x194) from [<c01dcfe9>] (mvneta_tx_timeout+0x19/0x24)
  [<c01dcfe9>] (mvneta_tx_timeout+0x19/0x24) from [<c024afc7>] (dev_watchdog+0x18b/0x1c4)
  [<c024afc7>] (dev_watchdog+0x18b/0x1c4) from [<c0020b53>] (call_timer_fn.isra.27+0x17/0x5c)
  [<c0020b53>] (call_timer_fn.isra.27+0x17/0x5c) from [<c0020cad>] (run_timer_softirq+0x115/0x170)
  [<c0020cad>] (run_timer_softirq+0x115/0x170) from [<c001ccb9>] (__do_softirq+0xbd/0x1a8)
  [<c001ccb9>] (__do_softirq+0xbd/0x1a8) from [<c001cfad>] (irq_exit+0x61/0x98)
  [<c001cfad>] (irq_exit+0x61/0x98) from [<c000d4bf>] (handle_IRQ+0x27/0x60)
  [<c000d4bf>] (handle_IRQ+0x27/0x60) from [<c000843b>] (armada_370_xp_handle_irq+0x33/0xc8)
  [<c000843b>] (armada_370_xp_handle_irq+0x33/0xc8) from [<c000fba9>] (__irq_usr+0x49/0x60)

Ben Hutchings attempted to propose a better fix consisting in using a
scheduled work for this, but while it fixed this panic, it caused other
random freezes and panics proving that the reset sequence in the driver
is unreliable and that additional fixes should be investigated.

When sending multiple streams over a link limited to 100 Mbps, Tx timeouts
happen from time to time, and the driver correctly recovers only when the
function is disabled.

Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Gregory CLEMENT <gregory.clement@free-electrons.com>
Cc: Ben Hutchings <ben@decadent.org.uk>
Tested-by: Arnaud Ebalard <arno@natisbad.org>
Signed-off-by: Willy Tarreau <w@1wt.eu>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

net: mvneta: use per_cpu stats to fix an SMP lock up

[ Upstream commit 74c41b048db1073a04827d7f39e95ac1935524cc ]

Stats writers are mvneta_rx() and mvneta_tx(). They don't lock anything
when they update the stats, and as a result, it randomly happens that
the stats freeze on SMP if two updates happen during stats retrieval.
This is very easily reproducible by starting two HTTP servers and binding
each of them to a different CPU, then consulting /proc/net/dev in loops
during transfers, the interface should immediately lock up. This issue
also randomly happens upon link state changes during transfers, because
the stats are collected in this situation, but it takes more attempts to
reproduce it.

The comments in netdevice.h suggest using per_cpu stats instead to get
rid of this issue.

This patch implements this. It merges both rx_stats and tx_stats into
a single "stats" member with a single syncp. Both mvneta_rx() and
mvneta_rx() now only update the a single CPU's counters.

In turn, mvneta_get_stats64() does the summing by iterating over all CPUs
to get their respective stats.

With this change, stats are still correct and no more lockup is encountered.

Note that this bug was present since the first import of the mvneta
driver.  It might make sense to backport it to some stable trees. If
so, it depends on "d33dc73 net: mvneta: increase the 64-bit rx/tx stats
out of the hot path".

Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Gregory CLEMENT <gregory.clement@free-electrons.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Tested-by: Arnaud Ebalard <arno@natisbad.org>
Signed-off-by: Willy Tarreau <w@1wt.eu>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

net: mvneta: increase the 64-bit rx/tx stats out of the hot path

[ Upstream commit dc4277dd41a80fd5f29a90412ea04bc3ba54fbf1 ]

Better count packets and bytes in the stack and on 32 bit then
accumulate them at the end for once. This saves two memory writes
and two memory barriers per packet. The incoming packet rate was
increased by 4.7% on the Openblocks AX3 thanks to this.

Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Gregory CLEMENT <gregory.clement@free-electrons.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Tested-by: Arnaud Ebalard <arno@natisbad.org>
Signed-off-by: Willy Tarreau <w@1wt.eu>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

net: ip, ipv6: handle gso skbs in forwarding path

commit fe6cc55f3a9a053482a76f5a6b2257cee51b4663 upstream.

Marcelo Ricardo Leitner reported problems when the forwarding link path
has a lower mtu than the incoming one if the inbound interface supports GRO.

Given:
Host <mtu1500> R1 <mtu1200> R2

Host sends tcp stream which is routed via R1 and R2.  R1 performs GRO.

In this case, the kernel will fail to send ICMP fragmentation needed
messages (or pkt too big for ipv6), as GSO packets currently bypass dstmtu
checks in forward path. Instead, Linux tries to send out packets exceeding
the mtu.

When locking route MTU on Host (i.e., no ipv4 DF bit set), R1 does
not fragment the packets when forwarding, and again tries to send out
packets exceeding R1-R2 link mtu.

This alters the forwarding dstmtu checks to take the individual gso
segment lengths into account.

For ipv6, we send out pkt too big error for gso if the individual
segments are too big.

For ipv4, we either send icmp fragmentation needed, or, if the DF bit
is not set, perform software segmentation and let the output path
create fragments when the packet is leaving the machine.
It is not 100% correct as the error message will contain the headers of
the GRO skb instead of the original/segmented one, but it seems to
work fine in my (limited) tests.

Eric Dumazet suggested to simply shrink mss via ->gso_size to avoid
sofware segmentation.

However it turns out that skb_segment() assumes skb nr_frags is related
to mss size so we would BUG there.  I don't want to mess with it considering
Herbert and Eric disagree on what the correct behavior should be.

Hannes Frederic Sowa notes that when we would shrink gso_size
skb_segment would then also need to deal with the case where
SKB_MAX_FRAGS would be exceeded.

This uses sofware segmentation in the forward path when we hit ipv4
non-DF packets and the outgoing link mtu is too small.  Its not perfect,
but given the lack of bug reports wrt. GRO fwd being broken this is a
rare case anyway.  Also its not like this could not be improved later
once the dust settles.

Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Reported-by: Marcelo Ricardo Leitner <mleitner@redhat.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

net: core: introduce netif_skb_dev_features

commit d206940319c41df4299db75ed56142177bb2e5f6 upstream.

Will be used by upcoming ipv4 forward path change that needs to
determine feature mask using skb->dst->dev instead of skb->dev.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

net: add and use skb_gso_transport_seglen()

commit de960aa9ab4decc3304959f69533eef64d05d8e8 upstream.

This moves part of Eric Dumazets skb_gso_seglen helper from tbf sched to
skbuff core so it may be reused by upcoming ip forwarding path patch.

Signed-off-by: Florian Westphal <fw@strlen.de>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

net: sctp: fix sctp_connectx abi for ia32 emulation/compat mode

[ Upstream commit ffd5939381c609056b33b7585fb05a77b4c695f3 ]

SCTP's sctp_connectx() abi breaks for 64bit kernels compiled with 32bit
emulation (e.g. ia32 emulation or x86_x32). Due to internal usage of
'struct sctp_getaddrs_old' which includes a struct sockaddr pointer,
sizeof(param) check will always fail in kernel as the structure in
64bit kernel space is 4bytes larger than for user binaries compiled
in 32bit mode. Thus, applications making use of sctp_connectx() won't
be able to run under such circumstances.

Introduce a compat interface in the kernel to deal with such
situations by using a 'struct compat_sctp_getaddrs_old' structure
where user data is copied into it, and then sucessively transformed
into a 'struct sctp_getaddrs_old' structure with the help of
compat_ptr(). That fixes sctp_connectx() abi without any changes
needed in user space, and lets the SCTP test suite pass when compiled
in 32bit and run on 64bit kernels.

Fixes: f9c67811ebc0 ("sctp: Fix regression introduced by new sctp_connectx api")
Signed-off-by: Daniel Borkmann <dborkman@redhat.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Acked-by: Vlad Yasevich <vyasevich@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ipv4: fix counter in_slow_tot

[ Upstream commit a6254864c08109c66a194612585afc0439005286 ]

since commit 89aef8921bf("ipv4: Delete routing cache."), the counter
in_slow_tot can't work correctly.

The counter in_slow_tot increase by one when fib_lookup() return successfully
in ip_route_input_slow(), but actually the dst struct maybe not be created and
cached, so we can increase in_slow_tot after the dst struct is created.

Signed-off-by: Duan Jiong <duanj.fnst@cn.fujitsu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

bonding: 802.3ad: make aggregator_identifier bond-private

[ Upstream commit 163c8ff30dbe473abfbb24a7eac5536c87f3baa9 ]

aggregator_identifier is used to assign unique aggregator identifiers
to aggregators of a bond during device enslaving.

aggregator_identifier is currently a global variable that is zeroed in
bond_3ad_initialize().

This sequence will lead to duplicate aggregator identifiers for eth1 and eth3:

create bond0
change bond0 mode to 802.3ad
enslave eth0 to bond0  //eth0 gets agg id 1
enslave eth1 to bond0  //eth1 gets agg id 2
create bond1
change bond1 mode to 802.3ad
enslave eth2 to bond1 //aggregator_identifier is reset to 0
//eth2 gets agg id 1
enslave eth3 to bond0  //eth3 gets agg id 2

Fix this by making aggregator_identifier private to the bond.

Signed-off-by: Jiri Bohac <jbohac@suse.cz>
Acked-by: Veaceslav Falico <vfalico@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usbnet: remove generic hard_header_len check

[ Upstream commit eb85569fe2d06c2fbf4de7b66c263ca095b397aa ]

This patch removes a generic hard_header_len check from the usbnet
module that is causing dropped packages under certain circumstances
for devices that send rx packets that cross urb boundaries.

One example is the AX88772B which occasionally send rx packets that
cross urb boundaries where the remaining partial packet is sent with
no hardware header. When the buffer with a partial packet is of less
number of octets than the value of hard_header_len the buffer is
discarded by the usbnet module.

With AX88772B this can be reproduced by using ping with a packet
size between 1965-1976.

The bug has been reported here:

https://bugzilla.kernel.org/show_bug.cgi?id=29082

This patch introduces the following changes:
- Removes the generic hard_header_len check in the rx_complete
  function in the usbnet module.
- Introduces a ETH_HLEN check for skbs that are not cloned from
  within a rx_fixup callback.
- For safety a hard_header_len check is added to each rx_fixup
  callback function that could be affected by this change.
  These extra checks could possibly be removed by someone
  who has the hardware to test.
- Removes a call to dev_kfree_skb_any() and instead utilizes the
  dev->done list to queue skbs for cleanup.

The changes place full responsibility on the rx_fixup callback
functions that clone skbs to only pass valid skbs to the
usbnet_skb_return function.

Signed-off-by: Emil Goode <emilgoode@gmail.com>
Reported-by: Igor Gnatenko <i.gnatenko.brain@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

gre: add link local route when local addr is any

[ Upstream commit 08b44656c08c8c2f73cdac2a058be2880e3361f2 ]

This bug was reported by Steinar H. Gunderson and was introduced by commit
f7cb8886335d ("sit/gre6: don't try to add the same route two times").

root@morgental:~# ip tunnel add foo mode gre remote 1.2.3.4 ttl 64
root@morgental:~# ip link set foo up mtu 1468
root@morgental:~# ip -6 route show dev foo
fe80::/64  proto kernel  metric 256

but after the above commit, no such route shows up.

There is no link local route because dev->dev_addr is 0 (because local ipv4
address is 0), hence no link local address is configured.

In this scenario, the link local address is added manually: 'ip -6 addr add
fe80::1 dev foo' and because prefix is /128, no link local route is added by the
kernel.

Even if the right things to do is to add the link local address with a /64
prefix, we need to restore the previous behavior to avoid breaking userpace.

Reported-by: Steinar H. Gunderson <sesse@samfundet.no>
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

net: asix: add missing flag to struct driver_info

[ Upstream commit d43ff4cd798911736fb39025ec8004284b1b0bc2 ]

The struct driver_info ax88178_info is assigned the function
asix_rx_fixup_common as it's rx_fixup callback. This means that
FLAG_MULTI_PACKET must be set as this function is cloning the
data and calling usbnet_skb_return. Not setting this flag leads
to usbnet_skb_return beeing called a second time from within
the rx_process function in the usbnet module.

Signed-off-by: Emil Goode <emilgoode@gmail.com>
Reported-by: Bjørn Mork <bjorn@mork.no>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

hyperv: Fix the carrier status setting

[ Upstream commit 891de74d693bb4fefe2efcc6432a4a9a9bee561e ]

Without this patch, the "cat /sys/class/net/ethN/operstate" shows
"unknown", and "ethtool ethN" shows "Link detected: yes", when VM
boots up with or without vNIC connected.

This patch fixed the problem.

Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com>
Reviewed-by: K. Y. Srinivasan <kys@microsoft.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

vhost: fix ref cnt checking deadlock

[ Upstream commit 0ad8b480d6ee916aa84324f69acf690142aecd0e ]

vhost checked the counter within the refcnt before decrementing.  It
really wanted to know that it is the one that has the last reference, as
a way to batch freeing resources a bit more efficiently.

Note: we only let refcount go to 0 on device release.

This works well but we now access the ref counter twice so there's a
race: all users might see a high count and decide to defer freeing
resources.
In the end no one initiates freeing resources until the last reference
is gone (which is on VM shotdown so might happen after a looooong time).

Let's do what we probably should have done straight away:
switch from kref to plain atomic, documenting the
semantics, return the refcount value atomically after decrement,
then use that to avoid the deadlock.

Reported-by: Qin Chuanyu <qinchuanyu@huawei.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

tg3: Fix deadlock in tg3_change_mtu()

[ Upstream commit c6993dfd7db9b0c6b7ca7503a56fda9236a4710f ]

Quoting David Vrabel -
"5780 cards cannot have jumbo frames and TSO enabled together.  When
jumbo frames are enabled by setting the MTU, the TSO feature must be
cleared.  This is done indirectly by calling netdev_update_features()
which will call tg3_fix_features() to actually clear the flags.

netdev_update_features() will also trigger a new netlink message for the
feature change event which will result in a call to tg3_get_stats64()
which deadlocks on the tg3 lock."

tg3_set_mtu() does not need to be under the tg3 lock since converting
the flags to use set_bit(). Move it out to after tg3_netif_stop().

Reported-by: David Vrabel <david.vrabel@citrix.com>
Tested-by: David Vrabel <david.vrabel@citrix.com>
Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: Nithin Nayak Sujir <nsujir@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

tcp: tsq: fix nonagle handling

[ Upstream commit bf06200e732de613a1277984bf34d1a21c2de03d ]

Commit 46d3ceabd8d9 ("tcp: TCP Small Queues") introduced a possible
regression for applications using TCP_NODELAY.

If TCP session is throttled because of tsq, we should consult
tp->nonagle when TX completion is done and allow us to send additional
segment, especially if this segment is not a full MSS.
Otherwise this segment is sent after an RTO.

[edumazet] : Cooked the changelog, added another fix about testing
sk_wmem_alloc twice because TX completion can happen right before
setting TSQ_THROTTLED bit.

This problem is particularly visible with recent auto corking,
but might also be triggered with low tcp_limit_output_bytes
values or NIC drivers delaying TX completion by hundred of usec,
and very low rtt.

Thomas Glanzmann for example reported an iscsi regression, caused
by tcp auto corking making this bug quite visible.

Fixes: 46d3ceabd8d9 ("tcp: TCP Small Queues")
Signed-off-by: John Ogness <john.ogness@linutronix.de>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Thomas Glanzmann <thomas@glanzmann.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

net: qmi_wwan: add Netgear Aircard 340U

[ Upstream commit fbd3a77d813f211060f86cc7a2f8416caf0e03b1 ]

This device was mentioned in an OpenWRT forum.  Seems to have a "standard"
Sierra Wireless ifnumber to function layout:
 0: qcdm
 2: nmea
 3: modem
 8: qmi
 9: storage

Signed-off-by: Bjørn Mork <bjorn@mork.no>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

netpoll: fix netconsole IPv6 setup

[ Upstream commit 00fe11b3c67dc670fe6391d22f1fe64e7c99a8ec ]

Currently, to make netconsole start over IPv6, the source address
needs to be specified. Without a source address, netpoll_parse_options
assumes we're setting up over IPv4 and the destination IPv6 address is
rejected.

Check if the IP version has been forced by a source address before
checking for a version mismatch when parsing the destination address.

Signed-off-by: Sabrina Dubroca <sd@queasysnail.net>
Acked-by: Cong Wang <cwang@twopensource.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

net: fix 'ip rule' iif/oif device rename

[ Upstream commit 946c032e5a53992ea45e062ecb08670ba39b99e3 ]

ip rules with iif/oif references do not update:
(detach/attach) across interface renames.

Signed-off-by: Maciej Żenczykowski <maze@google.com>
CC: Willem de Bruijn <willemb@google.com>
CC: Eric Dumazet <edumazet@google.com>
CC: Chris Davis <chrismd@google.com>
CC: Carlo Contavalli <ccontavalli@google.com>
Google-Bug-Id: 12936021
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ipv4: Fix runtime WARNING in rtmsg_ifa()

[ Upstream commit 63b5f152eb4a5bb79b9caf7ec37b4201d12f6e66 ]

On m68k/ARAnyM:

WARNING: CPU: 0 PID: 407 at net/ipv4/devinet.c:1599 0x316a99()
Modules linked in:
CPU: 0 PID: 407 Comm: ifconfig Not tainted
3.13.0-atari-09263-g0c71d68014d1 #1378
Stack from 10c4fdf0:
        10c4fdf0 002ffabb 000243e8 00000000 008ced6c 00024416 00316a99 0000063f
        00316a99 00000009 00000000 002501b4 00316a99 0000063f c0a86117 00000080
        c0a86117 00ad0c90 00250a5a 00000014 00ad0c90 00000000 00000000 00000001
        00b02dd0 00356594 00000000 00356594 c0a86117 eff6c9e4 008ced6c 00000002
        008ced60 0024f9b4 00250b52 00ad0c90 00000000 00000000 00252390 00ad0c90
        eff6c9e4 0000004f 00000000 00000000 eff6c9e4 8000e25c eff6c9e4 80001020
Call Trace: [<000243e8>] warn_slowpath_common+0x52/0x6c
 [<00024416>] warn_slowpath_null+0x14/0x1a
 [<002501b4>] rtmsg_ifa+0xdc/0xf0
 [<00250a5a>] __inet_insert_ifa+0xd6/0x1c2
 [<0024f9b4>] inet_abc_len+0x0/0x42
 [<00250b52>] inet_insert_ifa+0xc/0x12
 [<00252390>] devinet_ioctl+0x2ae/0x5d6

Adding some debugging code reveals that net_fill_ifaddr() fails in

    put_cacheinfo(skb, ifa->ifa_cstamp, ifa->ifa_tstamp,
                              preferred, valid))

nla_put complains:

    lib/nlattr.c:454: skb_tailroom(skb) = 12, nla_total_size(attrlen) = 20

Apparently commit 5c766d642bcaffd0c2a5b354db2068515b3846cf ("ipv4:
introduce address lifetime") forgot to take into account the addition of
struct ifa_cacheinfo in inet_nlmsg_size(). Hence add it, like is already
done for ipv6.

Suggested-by: Cong Wang <cwang@twopensource.com>
Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Cong Wang <cwang@twopensource.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

can: add destructor for self generated skbs

[ Upstream commit 0ae89beb283a0db5980d1d4781c7d7be2f2810d6 ]

Self generated skbuffs in net/can/bcm.c are setting a skb->sk reference but
no explicit destructor which is enforced since Linux 3.11 with commit
376c7311bdb6 (net: add a temporary sanity check in skb_orphan()).

This patch adds some helper functions to make sure that a destructor is
properly defined when a sock reference is assigned to a CAN related skb.
To create an unshared skb owned by the original sock a common helper function
has been introduced to replace open coded functions to create CAN echo skbs.

Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Tested-by: Andre Naujoks <nautsch2@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

bridge: fix netconsole setup over bridge

[ Upstream commit dbe173079ab58a444e12dbebe96f5aec1e0bed1a ]

Commit 93d8bf9fb8f3 ("bridge: cleanup netpoll code") introduced
a check in br_netpoll_enable(), but this check is incorrect for
br_netpoll_setup(). This patch moves the code after the check
into __br_netpoll_enable() and calls it in br_netpoll_setup().
For br_add_if(), the check is still needed.

Fixes: 93d8bf9fb8f3 ("bridge: cleanup netpoll code")
Cc: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp>
Cc: Stephen Hemminger <stephen@networkplumber.org>
Cc: David S. Miller <davem@davemloft.net>
Signed-off-by: Cong Wang <cwang@twopensource.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Acked-by: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp>
Tested-by: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

9p/trans_virtio.c: Fix broken zero-copy on vmalloc() buffers

[ Upstream commit b6f52ae2f0d32387bde2b89883e3b64d88b9bfe8 ]

The 9p-virtio transport does zero copy on things larger than 1024 bytes
in size. It accomplishes this by returning the physical addresses of
pages to the virtio-pci device. At present, the translation is usually a
bit shift.

That approach produces an invalid page address when we read/write to
vmalloc buffers, such as those used for Linux kernel modules. Any
attempt to load a Linux kernel module from 9p-virtio produces the
following stack.

[<ffffffff814878ce>] p9_virtio_zc_request+0x45e/0x510
[<ffffffff814814ed>] p9_client_zc_rpc.constprop.16+0xfd/0x4f0
[<ffffffff814839dd>] p9_client_read+0x15d/0x240
[<ffffffff811c8440>] v9fs_fid_readn+0x50/0xa0
[<ffffffff811c84a0>] v9fs_file_readn+0x10/0x20
[<ffffffff811c84e7>] v9fs_file_read+0x37/0x70
[<ffffffff8114e3fb>] vfs_read+0x9b/0x160
[<ffffffff81153571>] kernel_read+0x41/0x60
[<ffffffff810c83ab>] copy_module_from_fd.isra.34+0xfb/0x180

Subsequently, QEMU will die printing:

qemu-system-x86_64: virtio: trying to map MMIO memory

This patch enables 9p-virtio to correctly handle this case. This not
only enables us to load Linux kernel modules off virtfs, but also
enables ZFS file-based vdevs on virtfs to be used without killing QEMU.

Special thanks to both Avi Kivity and Alexander Graf for their
interpretation of QEMU backtraces. Without their guidence, tracking down
this bug would have taken much longer. Also, special thanks to Linus
Torvalds for his insightful explanation of why this should use
is_vmalloc_addr() instead of is_vmalloc_or_module_addr():

https://lkml.org/lkml/2014/2/8/272

Signed-off-by: Richard Yao <ryao@gentoo.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

6lowpan: fix lockdep splats

[ Upstream commit 20e7c4e80dcd01dad5e6c8b32455228b8fe9c619 ]

When a device ndo_start_xmit() calls again dev_queue_xmit(),
lockdep can complain because dev_queue_xmit() is re-entered and the
spinlocks protecting tx queues share a common lockdep class.

Same issue was fixed for bonding/l2tp/ppp in commits

0daa2303028a6 ("[PATCH] bonding: lockdep annotation")
49ee49202b4ac ("bonding: set qdisc_tx_busylock to avoid LOCKDEP splat")
23d3b8bfb8eb2 ("net: qdisc busylock needs lockdep annotations ")
303c07db487be ("ppp: set qdisc_tx_busylock to avoid LOCKDEP splat ")

Reported-by: Alexander Aring <alex.aring@gmail.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Tested-by: Alexander Aring <alex.aring@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

NFS fix error return in nfs4_select_rw_stateid

commit 146d70caaa1b87f64597743429d7da4b8073d0c9 upstream.

Do not return an error when nfs4_copy_delegation_stateid succeeds.

Signed-off-by: Andy Adamson <andros@netapp.com>
Link: http://lkml.kernel.org/r/1392737765-41942-1-git-send-email-andros@netapp.com
Fixes: ef1820f9be27b (NFSv4: Don't try to recover NFSv4 locks when...)
Cc: NeilBrown <neilb@suse.de>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

NFS: Do not set NFS_INO_INVALID_LABEL unless server supports labeled NFS

commit fd1defc257e2b12ab69bc0b379105c00eca4e112 upstream.

Commit aa9c2669626c (NFS: Client implementation of Labeled-NFS) introduces
a performance regression. When nfs_zap_caches_locked is called, it sets
the NFS_INO_INVALID_LABEL flag irrespectively of whether or not the
NFS server supports security labels. Since that flag is never cleared,
it means that all calls to nfs_revalidate_inode() will now trigger
an on-the-wire GETATTR call.

This patch ensures that we never set the NFS_INO_INVALID_LABEL unless the
server advertises support for labeled NFS.
It also causes nfs_setsecurity() to clear NFS_INO_INVALID_LABEL when it
has successfully set the security label for the inode.
Finally it gets rid of the NFS_INO_INVALID_LABEL cruft from nfs_update_inode,
which has nothing to do with labeled NFS.

Reported-by: Neil Brown <neilb@suse.de>
Tested-by: Neil Brown <neilb@suse.de>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

rtlwifi: rtl8192ce: Fix too long disable of IRQs

commit f78bccd79ba3cd9d9664981b501d57bdb81ab8a4 upstream.

rtl8192ce is disabling for too long the local interrupts during hw initiatialisation when performing scans

The observable symptoms in dmesg can be:

- underruns from ALSA playback
- clock freezes (tstamps do not change for several dmesg entries until irqs are finaly reenabled):

[  250.817669] rtlwifi:rtl_op_config():<0-0-0> 0x100
[  250.817685] rtl8192ce:_rtl92ce_phy_set_rf_power_state():<0-1-0> IPS Set eRf nic enable
[  250.817732] rtl8192ce:_rtl92ce_init_mac():<0-1-0> reg0xec:18051d59:11
[  250.817796] rtl8192ce:_rtl92ce_init_mac():<0-1-0> reg0xec:18051d59:11
[  250.817910] rtl8192ce:_rtl92ce_init_mac():<0-1-0> reg0xec:18051d59:11
[  250.818024] rtl8192ce:_rtl92ce_init_mac():<0-1-0> reg0xec:18051d59:11
[  250.818139] rtl8192ce:_rtl92ce_init_mac():<0-1-0> reg0xec:18051d59:11
[  250.818253] rtl8192ce:_rtl92ce_init_mac():<0-1-0> reg0xec:18051d59:11
[  250.818367] rtl8192ce:_rtl92ce_init_mac():<0-1-0> reg0xec:18051d59:11
[  250.818472] rtl8192ce:_rtl92ce_init_mac():<0-1-0> reg0xec:18051d59:11
[  250.818472] rtl8192ce:_rtl92ce_init_mac():<0-1-0> reg0xec:18051d59:11
[  250.818472] rtl8192ce:_rtl92ce_init_mac():<0-1-0> reg0xec:18051d59:11
[  250.818472] rtl8192ce:_rtl92ce_init_mac():<0-1-0> reg0xec:18051d59:11
[  250.818472] rtl8192ce:_rtl92ce_init_mac():<0-1-0> reg0xec:98053f15:10
[  250.818472] rtl8192ce:rtl92ce_sw_led_on():<0-1-0> LedAddr:4E ledpin=1
[  250.818472] rtl8192c_common:rtl92c_download_fw():<0-1-0> Firmware Version(49), Signature(0x88c1),Size(32)
[  250.818472] rtl8192ce:rtl92ce_enable_hw_security_config():<0-1-0> PairwiseEncAlgorithm = 0 GroupEncAlgorithm = 0
[  250.818472] rtl8192ce:rtl92ce_enable_hw_security_config():<0-1-0> The SECR-value cc
[  250.818472] rtl8192c_common:rtl92c_dm_check_txpower_tracking_thermal_meter():<0-1-0> Schedule TxPowerTracking direct call!!
[  250.818472] rtl8192c_common:rtl92c_dm_txpower_tracking_callback_thermalmeter():<0-1-0> rtl92c_dm_txpower_tracking_callback_thermalmeter
[  250.818472] rtl8192c_common:rtl92c_dm_txpower_tracking_callback_thermalmeter():<0-1-0> Readback Thermal Meter = 0xe pre thermal meter 0xf eeprom_thermalmeter 0xf
[  250.818472] rtl8192c_common:rtl92c_dm_txpower_tracking_callback_thermalmeter():<0-1-0> Initial pathA ele_d reg0xc80 = 0x40000000, ofdm_index=0xc
[  250.818472] rtl8192c_common:rtl92c_dm_txpower_tracking_callback_thermalmeter():<0-1-0> Initial reg0xa24 = 0x90e1317, cck_index=0xc, ch14 0
[  250.818472] rtl8192c_common:rtl92c_dm_txpower_tracking_callback_thermalmeter():<0-1-0> Readback Thermal Meter = 0xe pre thermal meter 0xf eeprom_thermalmeter 0xf delta 0x1 delta_lck 0x0 delta_iqk 0x0
[  250.818472] rtl8192c_common:rtl92c_dm_txpower_tracking_callback_thermalmeter():<0-1-0> <===
[  250.818472] rtl8192c_common:rtl92c_dm_initialize_txpower_tracking_thermalmeter():<0-1-0> pMgntInfo->txpower_tracking = 1
[  250.818472] rtl8192ce:rtl92ce_led_control():<0-1-0> ledaction 3
[  250.818472] rtl8192ce:rtl92ce_sw_led_on():<0-1-0> LedAddr:4E ledpin=1
[  250.818472] rtlwifi:rtl_ips_nic_on():<0-1-0> before spin_unlock_irqrestore
[  251.154656] PCM: Lost interrupts? [Q]-0 (stream=0, delta=15903, new_hw_ptr=293408, old_hw_ptr=277505)

The exact code flow that causes that is:

1. wpa_supplicant send a start_scan request to the nl80211 driver
2. mac80211 module call rtl_op_config with IEEE80211_CONF_CHANGE_IDLE
3.   rtl_ips_nic_on is called which disable local irqs
4.     rtl92c_phy_set_rf_power_state() is called
5.       rtl_ps_enable_nic() is called and hw_init()is executed and then the interrupts on the device are enabled

A good solution could be to refactor the code to avoid calling rtl92ce_hw_init() with the irqs disabled
but a quick and dirty solution that has proven to work is
to reenable the irqs during the function rtl92ce_hw_init().

I think that it is safe doing so since the device interrupt will only be enabled after the init function succeed.

Signed-off-by: Olivier Langlois <olivier@trillion01.com>
Acked-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

rtlwifi: Fix incorrect return from rtl_ps_enable_nic()

commit 2e8c5e56b307271c2dab6f8bfd1d8a3822ca2390 upstream.

rtl_ps_enable_nic() is called from loops that will loop until this function returns true or a
maximum number of retries is performed.

hw_init() returns non-zero on error. In that situation return false to
restore the original design intent to retry hw init when it fails.

Signed-off-by: Olivier Langlois <olivier@trillion01.com>
Acked-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>