iio:imu:mpu6050: icm20608 initial support

Introduce support for Invense ICM20608 IMU, a 6-axis motion tracking device
that combines a 3-axis gyroscope and a 3-axis accelerometer:
http://www.invensense.com/products/motion-tracking/6-axis/icm-20608-2

Signed-off-by: Gregor Boirie <gregor.boirie@parrot.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: adc: max1363: Add device tree binding

This patch adds the necessary device tree binding to allow DT probing of
currently supported parts.

Signed-off-by: Florian Vaussard <florian.vaussard@heig-vd.ch>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: adc: max1363: Add device tree binding documentation

Add the device tree documentation for all the supported parts. Mandatory
binding is the compatible string and the slave I2C address.

Optional properties can be used to specify the Vcc / Vref regulators, as
well as the IRQ line if available.

Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Florian Vaussard <florian.vaussard@heig-vd.ch>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: bmg160: add callbacks for the filter frequency

The filter frequency and sample rate have a fixed relationship.
Only the filter frequency is unique, however.
Currently the driver ignores the filter settings for 32 Hz and
64 Hz.

This patch adds the necessary callbacks to be able to configure
and read the filter setting from sysfs.

Signed-off-by: Steffen Trumtrar <s.trumtrar@pengutronix.de>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: Add driver for Broadcom iproc-static-adc

This patch adds basic driver implementation for Broadcom's
static adc controller used in iProc SoC's family.

Signed-off-by: Raveendra Padasalagi <raveendra.padasalagi@broadcom.com>
Reviewed-by: Ray Jui <ray.jui@broadcom.com>
Reviewed-by: Scott Branden <scott.branden@broadcom.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

Documentation: DT: Add iproc-static-adc binding

The patch adds devicetree binding document for broadcom's
iproc-static-adc controller driver.

Signed-off-by: Raveendra Padasalagi <raveendra.padasalagi@broadcom.com>
Reviewed-by: Ray Jui <ray.jui@broadcom.com>
Reviewed-by: Scott Branden <scott.branden@broadcom.com>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: pressure: bmp280: read calibration data once

The calibration data is described as coming from an E2PROM and that
means it does not change. Just read it once at probe time and store
it in the device state container. Also toss the calibration data
into the entropy pool since it is device unique.

Reviewed-by: Vlad Dogaru <vlad.dogaru@intel.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: pressure: bmp280: add power management

The PM280 has an internal standby-mode, but to really save power
we should shut the sensor down and disconnect the power. With
the proper .pm hooks we can enable both runtime and system power
management of the sensor. We use the *force callbacks from the
system PM hooks. When the sensor comes back we always reconfigure
it to make sure it is ready to roll as expected.

Cc: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: pressure: bmp280: add support for BMP085 EOC interrupt

The first version of this sensor, BMP085, supports sending an
End-of-Conversion (EOC) interrupt. Add code to support this using
a completion, in a similar vein as drivers/misc/bmp085.c does.

Make sure to check that we are given a rising edge, because the
EOC line goes from low-to-high when the conversion is ready.

Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: pressure: bmp280: add SPI interface driver

This patch mimics the SPI functionality found in the misc driver in
drivers/misc/bh085-spi.c to make it possible to reuse the existing
BMP280/BMP180/BMP085 driver with all clients of the other driver.
The adoption is straight-forward since like the other driver, it is
a simple matter of using regmap.

This driver is also so obviously inspired/copied from the old misc
driver in drivers/misc/bmp085.c that I just took the liberty to
add in the authors of the other drivers + self in the core driver
file.

The MISC driver also supports a variant named "BMP181" so include
that here to be complete in comparison to the old driver.

The bus mapping code for SPI was written by Akinobu Mita.

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Tested-by: Akinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: pressure: bmp280: split off an I2C Kconfig entry

This creates a separate BMP280_I2C Kconfig entry that gets selected
by BMP280 for I2C transport. As we currently only support I2C
transport there is not much practical change other than getting
a separate object file (or module) for the I2C driver part. The
old Kconfig symbol BMP280 will still select the stuff we need so
that oldconfig and old defconfigs works fine.

Tested-by: Akinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: pressure: bmp280: split driver in logical parts

This splits the BMP280 driver in three logical parts: the core driver
bmp280-core that only operated on a struct device * and a struct regmap *,
the regmap driver bmp280-regmap that can be shared between I2C and other
transports and the I2C module driver bmp280-i2c.

Cleverly bake all functionality into a single object bmp280.o so that
we still get the same module binary built for the device in the end,
without any fuzz exporting symbols to the left and right.

Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: pressure: bmp280: support supply regulators

The BMP085/BMP180/BMP280 is supplied with two power sources:
VDDA (analog power) and VDDD (digital power). As these may come
from regulators (as on the APQ8060 Dragonboard) we need the driver
to attempt to fetch and enable these regulators.

We FAIL if we cannot: boards should either define:
- Proper regulators if present
- Define fixed regulators if power is hardwired to the component
- Rely on dummy regulators (will be present on all DT systems and
  any boardfile system that calls regulator_has_full_constraints().

Cc: Mark Brown <broonie@kernel.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

drivers:iio:light:isl29125: added macros for sensing range

Added macros for sensing range as the corresponding magic numbers
were used at multiple places.
   - ISL29125_SENSING_RANGE_0 for 375 lux full range
   - ISL29125_SENSING_RANGE_1 for 10k lux full range

Signed-off-by: Bijosh Thykkoottathil <bijosh.t@hotmail.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: st_sensors: harden interrupt handling

Leonard Crestez observed the following phenomenon: when using
hard interrupt triggers (the DRDY line coming out of an ST
sensor) sometimes a new value would arrive while reading the
previous value, due to latencies in the system.

We discovered that the ST hardware as far as can be observed
is designed for level interrupts: the DRDY line will be held
asserted as long as there are new values coming. The interrupt
handler should be re-entered until we're out of values to
handle from the sensor.

If interrupts were handled as occurring on the edges (usually
low-to-high) new values could appear and the line be held
asserted after that, and these values would be missed, the
interrupt handler would also lock up as new data was
available, but as no new edges occurs on the DRDY signal,
nothing happens: the edge detector only detects edges.

To counter this, do the following:

- Accept interrupt lines to be flagged as level interrupts
  using IRQF_TRIGGER_HIGH and IRQF_TRIGGER_LOW. If the line
  is marked like this (in the device tree node or ACPI
  table or similar) it will be utilized as a level IRQ.
  We mark the line with IRQF_ONESHOT and mask the IRQ
  while processing a sample, then the top half will be
  entered again if new values are available.

- If we are flagged as using edge interrupts with
  IRQF_TRIGGER_RISING or IRQF_TRIGGER_FALLING: remove
  IRQF_ONESHOT so that the interrupt line is not
  masked while running the thread part of the interrupt.
  This way we will never miss an interrupt, then introduce
  a loop that polls the data ready registers repeatedly
  until no new samples are available, then exit the
  interrupt handler. This way we know no new values are
  available when the interrupt handler exits and
  new (edge) interrupts will be triggered when data arrives.
  Take some extra care to update the timestamp in the poll
  loop if this happens. The timestamp will not be 100%
  perfect, but it will at least be closer to the actual
  events. Usually the extra poll loop will handle the new
  samples, but once in a blue moon, we get a new IRQ
  while exiting the loop, before returning from the
  thread IRQ bottom half with IRQ_HANDLED. On these rare
  occasions, the removal of IRQF_ONESHOT means the
  interrupt will immediately fire again.

- If no interrupt type is indicated from the DT/ACPI,
  choose IRQF_TRIGGER_RISING as default, as this is necessary
  for legacy boards.

Tested successfully on the LIS331DL and L3G4200D by setting
sampling frequency to 400Hz/800Hz and stressing the system:
extra reads in the threaded interrupt handler occurs.

Cc: Giuseppe Barba <giuseppe.barba@st.com>
Cc: Denis Ciocca <denis.ciocca@st.com>
Tested-by: Crestez Dan Leonard <cdleonard@gmail.com>
Reported-by: Crestez Dan Leonard <cdleonard@gmail.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: magn: ak8975: deploy runtime and system PM

This adds runtime PM support to the AK8975 driver. It solves two
problems:

- After reading the first value the chip was left in MODE_ONCE,
  meaning (presumably) it may be consuming more power. Now the
  runtime PM hooks kick in and set it to POWER_DOWN.

- Regulators were simply enabled and left on, making it
  impossible to turn the power consuming regulators off because
  of the increased refcount. We now disable the regulators at
  autosuspend.

- We also handle system suspend: by using pm_runtime_force_suspend()
  and pm_runtime_force_resume() from the system PM sleep hooks,
  the runtime PM code is managing the power also for this case.
  It is currently not completely optimal: when the system resumes
  the AK8975 goes into active mode even if noone is going to use
  it: currently the force calls need to be paired, but the runtime
  PM people are working on making it possible to leave devices
  runtime suspended when coming back from sleep.

Inspired by my work on the BH1780 light sensor driver.

Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: magn: ak8975: make sure to power down at remove()

The code was not powering the magnetometer down properly at
remove(): just cutting the regulators without first setting the
device in power off mode. Fix this.

Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: magn: ak8975: allow a delay after enabling regulators

The datasheet actually specifies that we need to wait atleast
500us after powering on the device before trying to set mode.

Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: magn: ak8975: refactor regulator handlers

Move the regulator_get() calls directly into the probe() function,
keep only the power_on()/power_off() functions to flick the
regulators on/off.

Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: magn: ak8975: add Vid regulator

The AK8975 has two power sources: Vdd (analog voltage supply)
and Vid (digital voltage supply). Optionally also obtain the Vid
supply regulator and enable it.

If an error occurs when enabling one of the regulators: bail out.

Cc: Gregor Boirie <gregor.boirie@parrot.com>
Cc: Richard Leitner <dev@g0hl1n.net>
Cc: Krzysztof Kozlowski <k.kozlowski@samsung.com>
Cc: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: magn: ak8975: fix regulator usage

IS_ERR_OR_NULL() should never be used with regulators because
a NULL pointer may be a perfectly valid dummy regulator

We should always succeed to fetch and enable a regulator, but
it may be a dummy. That is fine, so bail out for any real
errors or probe deferrals

Include the error code in the warning print so we know what
kind of problem we're dealing with (for example it is nice to
see if it is a probe deferral).

As we will bail out of probe if the regulator is erroneous,
just issue regulator_disable() on the poweroff path: it will
succeed.

Cc: Mark Brown <broonie@kernel.org>
Cc: Lars-Peter Clausen Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: pressure: bmp280: add reset GPIO line handling

On the APQ8060 Dragonboard the reset line to the BMP085 pressure
sensor is not deasserted on boot, so the driver needs to handle
this. For a simple GPIO line supplied as a descriptor (from a board
file, device tree or ACPI) this does the trick.

Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: pressure: bmp280: support device tree initialization

This adds device tree support to the BMP085, BMP180 and BMP280
pressure sensors. Tested on the Qualcomm APQ8060 Dragonboard:

iio:device1$ cat in_temp_input
26700
iio:device1$ cat in_pressure_input
99.185000000

Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: pressure: bmp280: augment DT bindings

This adds standard device tree bindings for a reset GPIO line, and
the VDDD and VDDA power regulators.

Cc: devicetree@vger.kernel.org
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio:core: timestamping clock selection support

Adds a new per-device sysfs attribute "current_timestamp_clock" to allow
userspace to select a particular POSIX clock for buffered samples and
events timestamping.

Following clocks, as listed in clock_gettime(2), are supported:
CLOCK_REALTIME, CLOCK_MONOTONIC, CLOCK_MONOTONIC_RAW,
CLOCK_REALTIME_COARSE, CLOCK_MONOTONIC_COARSE, CLOCK_BOOTTIME and
CLOCK_TAI.

Signed-off-by: Gregor Boirie <gregor.boirie@parrot.com>
Acked-by: Sanchayan Maity <maitysanchayan@gmail.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

timekeeping: export get_monotonic_coarse64 symbol

EXPORT_SYMBOL() get_monotonic_coarse64 for new IIO timestamping clock
selection usage. This provides user apps the ability to request a
particular IIO device to timestamp samples using a monotonic coarse clock
granularity.

Signed-off-by: Gregor Boirie <gregor.boirie@parrot.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

Merge tag 'iio-for-4.8b' of git://git./linux/kernel/git/jic23/iio into staging-next

Jonathan writes:

Second round of new iio device support, features and cleanups in the 4.8 cycle

Firstly some contact detail updates:
* NXP took over freescale. Update the mma8452 header to reflect this.
* Martin Kepplinger email address change in mma8452 header.
* Adriana Reus has changed email address. Update .mailmap.
* Matt Ranostay has changed email address. Update .mailmap.

New Device Support
* max1363
  - add the missing i2c_device_ids for a couple of parts so they can actually
    be used.
* ms5867
  - add device ids for ms5805 and ms5837 parts.

New Features
* ad5755
  - DT support.  This one was a bit controversial and under review for a long
    time.  Still no one could come up with a better solution.
* stx104
  - add gpio support
* ti-adc081c
  - Add ACPI device ID matching.

Core changes
* Refuse to register triggers with duplicate names.  There is no way to
  distinguish between them so this makes no sense.  A few drivers do not
  generate unique names for each instance of the device present.  We can't
  fix this without changing ABI so leave them and wait for someone to
  actually take the rare step of two identical accelerometers on the same
  board.
* buffer-dma
  - use ARRAY_SIZE in a few appropriate locations.

Tools
* Fix the fact that the --trigger-num option in generic_buffer didn't allow
  0 which is perfectly valid in the ABI.

Cleanups
* as3935
  - improve error reporting.
  - remove redundant zeroing of a field in iio_priv.
* gp2ap020a00f
  - use the iio_device_claim_*_mode helpers rather than open coding locking
  around mode changes.
* isl29125
  - use the iio_device_claim_*_mode helpers rather than open coding locking.
* lidar
  - use the iio_device_claim_*_mode helpers rather than open coding locking.
* mma8452
  - more detail in devices supported description in comments (addresses and
  similar)
* sca3000
  - add a missing error check.
* tcs3414
  - use the iio_device_claim_*_mode helpers rather than open coding locking.
* tcs3472
  - use the iio_device_claim_*_mode helpers rather than open coding locking.

iio: ad5755: Add DT binding documentation

Signed-off-by: Sean Nyekjaer <sean.nyekjaer@prevas.dk>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: ad5755: add support for dt bindings

Devicetree can provide platform data

Signed-off-by: Sean Nyekjaer <sean.nyekjaer@prevas.dk>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: buffer-dma: Use ARRAY_SIZE in for loop range

Use the ARRAY_SIZE macro in the for loops that access queue->fileio.blocks.
Macro is already used in a couple of places where this access occurs,
but range was hardcoded in these locations.

Signed-off-by: Phil Reid <preid@electromag.com.au>
Acked-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: adc: max1363: Fix missing i2c_device_id for MAX1164x parts

The driver supports MAX11644, MAX11645, MAX11646 and MAX11647 parts. But
the corresponding i2c_device_id are missing. Add them!

Signed-off-by: Florian Vaussard <florian.vaussard@heig-vd.ch>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: proximity: as3935: remove redundant zeroing of tune_cap

This is redundant as the containing stucture is allocated as part of
iio_device_alloc using kzalloc and hence is already 0.

Signed-off-by: Matt Ranostay <mranostay@gmail.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: stx104: Add GPIO support for the Apex Embedded Systems STX104

The Apex Embedded Systems STX104 device features eight lines of digital
I/O (four digital inputs and four digital outputs). This patch adds GPIO
support for these eight lines of digital I/O via GPIOLIB.

Cc: Alexandre Courbot <gnurou@gmail.com>
Signed-off-by: William Breathitt Gray <vilhelm.gray@gmail.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: as3935: improve error reporting in as3935_event_work

gcc warns about a potentially uninitialized variable use
in as3935_event_work:

drivers/iio/proximity/as3935.c: In function ‘as3935_event_work’:
drivers/iio/proximity/as3935.c:231:6: error: ‘val’ may be used uninitialized in this function [-Werror=maybe-uninitialized]

This case specifically happens when spi_w8r8() fails with a
negative return code. We check all other users of this function
except this one.

As the error is rather unlikely to happen after the device
has already been initialized, this just adds a dev_warn().
Another warning already exists in the same function, but is
missing a trailing '\n' character, so I'm fixing that too.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Matt Ranostay <mranostay@gmail.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

staging: iio: accel: add error check

Go to error_ret if sca3000_read_ctrl_reg() failed.

Signed-off-by: Luis de Bethencourt <luisbg@osg.samsung.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: adc: ti-adc081c: add ACPI device ID matching

Add ACPI device ID matching for TI ADC081C/ADC101C/ADC121C ADCs.

Signed-off-by: Dan O'Donovan <dan@emutex.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

iio: Refuse to register triggers with duplicate names

The trigger name is documented as unique but drivers are currently
allowed to register triggers with duplicate names. This should be
considered a bug since it makes the 'current_trigger' interface
unusable.

Signed-off-by: Crestez Dan Leonard <leonard.crestez@intel.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>

staging: fsl-mc: convert mc command build/parse to use C structs

The layer abstracting the building of commands and extracting
responses is currently based on macros that shift and mask the command
fields and requires exposing offset/size values as macro parameters
and makes the code harder to read.

For clarity and maintainability, instead use an implementation based on
mapping the MC command definitions to C structures. These structures
contain the hardware command fields (which are naturally-aligned)
and individual fields are little-endian ordering (the byte ordering
of the hardware).

As such, there is no need to perform the conversion between core and
hardware (LE) endianness in mc_send_command(), but instead each
individual field in a command will be converted separately if needed
by the function building the command or extracting the response.

This patch does not introduce functional changes, both the hardware
ABIs and the APIs exposed for the DPAA2 objects remain the same.

Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Stuart Yoder <stuart.yoder@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: fsl-mc: properly set hwirq in msi set_desc

For an MSI domain the hwirq is an arbitrary but unique
id to identify an interrupt.  Previously the hwirq was set to
the MSI index of the interrupt, but that only works if there is
one DPRC.  Additional DPRCs require an expanded namespace.  Use
both the ICID (which is unique per DPRC) and the MSI index to
compose a hwirq value.

Signed-off-by: Stuart Yoder <stuart.yoder@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: fsl-mc: dprc: fix ordering problem freeing resources in remove of dprc

When unbinding a dprc from the dprc driver the cleanup of
the resource pools must happen after irq pool cleanup
is done.

Signed-off-by: Stuart Yoder <stuart.yoder@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: fsl-mc: dprc: add missing irq free

add missing free of the Linux irq when tearing down interrupts

Signed-off-by: Stuart Yoder <stuart.yoder@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: fsl-mc: fix asymmetry in destroy of mc_io

An mc_io represents a mapped MC portal.  Previously, an mc_io was
created for the root dprc in fsl_mc_bus_probe() and for child dprcs
in dprc_probe().  But the free of that data structure happened in the
general bus remove callback.  This asymmetry resulted in some bugs due
to unwanted destroys of mc_io object in some scenarios (e.g. vfio).

Fix this bug by making things symmetric-- mc_io created in
fsl_mc_bus_probe() is freed in fsl_mc_bus_remove().  The mc_io created
in dprc_probe() is freed in dprc_remove().

Signed-off-by: Bharat Bhushan <Bharat.Bhushan@nxp.com>
[Stuart: added check for root dprc and reworded commit message]
Signed-off-by: Stuart Yoder <stuart.yoder@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: fsl-mc: make fsl_mc_is_root_dprc() global

make fsl_mc_is_root_dprc() global so that the dprc driver
can use it

Signed-off-by: Stuart Yoder <stuart.yoder@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: fsl-mc: export mc_get_version

some drivers (built as modules) rely on mc_get_version()

Signed-off-by: Stuart Yoder <stuart.yoder@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: fsl-mc: add support for device table matching

Move the definition of fsl_mc_device_id to its proper location in
mod_devicetable.h, and add fsl-mc bus support to devicetable-offsets.c
and file2alias.c to enable device table matching.  With this patch udev
based module loading of fsl-mc drivers is supported.

Signed-off-by: Stuart Yoder <stuart.yoder@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: fsl-mc: clean up the device id struct

-rename the struct used for fsl-mc device ids to be more
 consistent with other busses
-remove the now obsolete and unused version fields

Signed-off-by: Stuart Yoder <stuart.yoder@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: fsl-mc: implement uevent callback and set the modalias

Replace placeholder code in the uevent callback to properly
set the MODALIAS env variable.

Signed-off-by: Stuart Yoder <stuart.yoder@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: fsl-mc: add support for the modalias sysfs attribute

In order to support uevent based module loading implement modalias support
for the fsl-mc bus driver. Aliases are based on vendor and object/device
id and are of the form "fsl-mc:vNdN".

Signed-off-by: Stuart Yoder <stuart.yoder@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: fix return value check in wlan_initialize_threads()

In case of error, the function kthread_run() returns ERR_PTR()
and never returns NULL. The NULL test in the return value check
should be replaced with IS_ERR().

Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn>
Reviewed-by: Julian Calaby <julian.calaby@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: arrays can't be NULL

hif_drv->usr_scan_req.net.net_info[i] contains found_net_info structs
which have the following element:
u8 bssid[6];

pstrNetworkInfo, of type network_info, also contains an u8 array named
bssid.

request->ssids is an array of cfg80211_ssid structs. Making ssid:
u8 ssid[IEEE80211_MAX_SSID_LEN];

In these 3 cases the arrays are being checked against NULL, which can't
happen. Removing the checks since they will always be true.

Found with smatch:
drivers/staging/wilc1000/host_interface.c:1234 Handle_RcvdNtwrkInfo() warn: this array is probably non-NULL. 'hif_drv->usr_scan_req.net_info[i].bssid'
drivers/staging/wilc1000/host_interface.c:1235 Handle_RcvdNtwrkInfo() warn: this array is probably non-NULL. 'pstrNetworkInfo->bssid'
drivers/staging/wilc1000/host_interface.c:1253 Handle_RcvdNtwrkInfo() warn: this array is probably non-NULL. 'hif_drv->usr_scan_req.net_info[hif_drv->usr_scan_req.rcvd_ch_cnt].bssid'
drivers/staging/wilc1000/host_interface.c:1254 Handle_RcvdNtwrkInfo() warn: this array is probably non-NULL. 'pstrNetworkInfo->bssid'

Signed-off-by: Luis de Bethencourt <luisbg@osg.samsung.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: Change interface wilc_mq_send to wilc_enqueue_cmd

Replace the interface 'wilc_mq_send' with 'wilc_enqueue_cmd'
and remove the now unused structures 'message' and 'message_queue'.
Restructure switch statement in the work queue helper function
host_if_work and remove unwanted indentation.

Signed-off-by: Binoy Jayan <binoy.jayan@linaro.org>
Reviewed-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: Replace kthread with workqueue for host interface

Deconstruct the kthread / message_queue logic, replacing it with
create_singlethread_workqueue() / queue_work() setup, by adding a
'struct work_struct' to 'struct host_if_msg'. The current kthread
hostIFthread() is converted to a work queue helper with the name
'host_if_work'.

Signed-off-by: Binoy Jayan <binoy.jayan@linaro.org>
Reviewed-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: message_queue: Move code to host interface

Move the contents of wilc_msgqueue.c and wilc_msgqueue.h into
host_interface.c, remove 'wilc_msgqueue.c' and 'wilc_msgqueue.h'.
This is done so as to restructure the implementation of the kthread
'hostIFthread' using a work queue.

Signed-off-by: Binoy Jayan <binoy.jayan@linaro.org>
Reviewed-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: Remove semaphore close_exit_sync

The semaphore 'close_exit_sync' does not serve any purpose other
than delaying the deregistration of the device which it is trying
to protect from shared access. 'up' is called only when a subdevice
is closed and not when it is opened. So, the semaphore count only
goes up when the device is used.

Signed-off-by: Binoy Jayan <binoy.jayan@linaro.org>
Reviewed-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: Replace semaphore sync_event with completion

The semaphore 'sync_event' is used as completion, so convert
it to a struct completion type. Also, return -ETIME if the return
value of wait_for_completion_timeout is 0.

Signed-off-by: Binoy Jayan <binoy.jayan@linaro.org>
Reviewed-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: Replace semaphore cfg_event with completion

The semaphore 'cfg_event' is used as completion, so convert
it to a struct completion type.

Signed-off-by: Binoy Jayan <binoy.jayan@linaro.org>
Reviewed-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: Replace semaphore txq_add_to_head_cs with mutex

The semaphore 'txq_add_to_head_cs' is a simple mutex, so it should be
written as one. Semaphores are going away in the future. Also, removing
the timeout scenario as the error handling code does not propagate the
timeout properly.

Signed-off-by: Binoy Jayan <binoy.jayan@linaro.org>
Reviewed-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: Replace semaphore txq_event with completion

The semaphore 'txq_event' is used as completion, so convert it
to a struct completion type.

Signed-off-by: Binoy Jayan <binoy.jayan@linaro.org>
Reviewed-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: fix typo

firmare -> firmware

Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: fix spelling mistake: "interupts" -> "interrupts"

trivial fix to spelling mistake in dev_err messages

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Reviewed-by: Julian Calaby <julian.calaby@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: remove unnecesary type cast of bss_type

There is no need to use type cast of bss_type because
hif_drv->cfg_values.bss_type is u8.

Signed-off-by: Chaehyun Lim <chaehyun.lim@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: add bss_type to remove line over 80 characters

A local bss_type variable is added to remove checkpatch warning of
line over 80 characters.

Signed-off-by: Chaehyun Lim <chaehyun.lim@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: remove unused struct set_mac_addr

struct set_mac_addr is not used anymore, so just remove it.

Signed-off-by: Chaehyun Lim <chaehyun.lim@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: rename result in handle_cfg_param

This patch renames result to ret that is used to get return value from
wilc_send_config_pkt. Some handle_*() functions are used as result,
others are used as ret. It will be changed as ret in all handle_*()
functions to match variable name.

Signed-off-by: Chaehyun Lim <chaehyun.lim@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: change data type of result in handle_cfg_param

This patch changes data type of result variable from s32 to int. result
is used to get return value from wilc_send_config_pkt that has return
type of int.

Signed-off-by: Chaehyun Lim <chaehyun.lim@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: change handle_cfg_param's return type to void

When handle_cfg_param is called in hostIFthread that is a kernel thread,
it is not checked return type of this function. This patch changes
return type to void.

Signed-off-by: Chaehyun Lim <chaehyun.lim@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: lustre: quiet lockdep recursive lock warning

Lockdep complains about potential recursive locking during mount
because the client configuration log is holding a lock on the MGC
obd_device to prevent it from being torn down, while also getting
mutexes on the MDC and OSC devices as they are instantiated:

 Lustre: Mounted myth-client
=============================================
[ INFO: possible recursive locking detected ]
4.7.0-rc2-vm-nfs+ #127 Tainted: G         C
---------------------------------------------

 May be due to missing lock nesting notation
2 locks held by ll_cfg_requeue/5928:
 #0:  (&cli->cl_sem){.+.+.+}, at: mgc_requeue_thread+0x15d/0x730 [mgc]
 #1:  (&cld->cld_lock){+.+.+.}, at: mgc_process_log+0x5e/0xf80 [mgc]
CPU: 0 PID: 5928 Comm: ll_cfg_requeue
Call Trace:
 [<ffffffff814a0855>] dump_stack+0x86/0xc1
 [<ffffffff810e7766>] __lock_acquire+0x726/0x1210
 [<ffffffff810e86be>] lock_acquire+0xfe/0x1f0
 [<ffffffff81888171>] down_read+0x51/0xa0
 [<ffffffffa04a8477>] sptlrpc_conf_client_adapt+0x47/0x150 [ptlrpc]
 [<ffffffffa0186b16>] mdc_set_info_async+0x2b6/0x470 [mdc]
 [<ffffffffa0294090>] class_notify_sptlrpc_conf+0x190/0x360 [obdclass]
 [<ffffffffa01a9e85>] mgc_process_log+0x925/0xf80 [mgc]
 [<ffffffffa01abafa>] mgc_requeue_thread+0x1fa/0x730 [mgc]
 [<ffffffff810af331>] kthread+0x101/0x120
 [<ffffffff8188ad6f>] ret_from_fork+0x1f/0x40

Add a separate lock class for the MGC callpath, since it will always
be held first, and none of the other obd_device locks should ever
be held concurrently.

Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/libcfs: Do not call kthread_run in wrong state

kthread_run might sleep during an allocation, and so
it's considered unsafe to call with a state that's not
RUNNABLE.
Move the state setting to after kthread_run call.

Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/osc: glimpse lock should match only with granted locks

A deadlock is possible during ccc_prep_size()->ldlm_lock_match() vs
cl_io_lock() which is waiting for a matched lock and conflicts with
already taken lock before ccc_prep_size().

It is better to send an additional lock request to avoid deadlock.

Seagate-bug-id: MRP-3312
Signed-off-by: Andriy Skulysh <andriy.skulysh@seagate.com>
Reviewed-on: http://review.whamcloud.com/18738
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7829
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Bobi Jam <bobijam@hotmail.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre: Add documentation for unstable_stats in sysfs

commit ac5b14810952 ("staging: lustre: osc: Track and limit
"unstable" pages") added a new sysfs variable, but corresponding bit of
documentation was not forgotten.

Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/osc: fix signed one bit field

Bit field 'oi_lockless' and 'oi_is_active' has one bit and is signed
which is confusing.

Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-on: http://review.whamcloud.com/19196
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7258
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Frank Zago <fzago@cray.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/llite: IOC_MDC_GETFILEINFO returns the wrong ino

req_capsule_server_get() through  __req_capsule_get in ll_dir_ioctl()
returns a pointer to a PTLRPC request or reply buffer, which is assigned
to struct mdt_body.

If the command is IOC_MDS_GETFILEINFO then the inode "st.st_ino" should
be assigned from one extracted from mdt_body through cl_fid_build_ino().

Signed-off-by: John Hammond <john.hammond@intel.com>
Signed-off-by: akam kumar bharathi <azurelustre@gmail.com>
Reviewed-on: http://review.whamcloud.com/17618
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-5954
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/llite: ll_revalidate_dentry update

There are a couple of cases in ll_revalidate_dentry() where
we are pretty sure the dentry is valid, so check for them early
and save more expensive checks for later.

Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/llite: Restore proper opencache operations

Mark dentries that came to us via NFS in a special way so that
we can tell them apart during open and activate open cache
(we really don't want to do open/close RPC for every NFS IO).

This became needed since dentry revlidate no longer reimplements
any RPCs for lookup, and as such if a dentry is valid,
ll_revalidate_dentry returns 1 and ll_lookup_it() is never visited
during opens, we get straght into ll_file_open() without a valid
intent/RPC. This used to be only true for NFS, so opencache was
engaged needlessly, and it carries a cost of it's own if there is
in fact no repetitive file opening-closing going on

Signed-off-by: Oleg Drokin <oleg.drokin@intel.com>
Reviewed-on: http://review.whamcloud.com/20354
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-8019
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Li Xi <lixi@ddn.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/llite: don't panic when fid is insane

LASSERT should never be done on data that is
received to over the network. Return EINVAL
when server returns invalid fid despite of
it_status == 0.

Signed-off-by: Sergey Cheremencev <sergey.cheremencev@seagate.com>
Seagate-bug-id: MRP-3073
Reviewed-on: http://review.whamcloud.com/17985
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7422
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/mdc: Zero atime in close RPC

While atime on close is supposed to only increase, there's
a bug in some older server versions where atime from a client
is taken no matter the value that allows a stale client atime
to overwrite a correct value.

Update atime in close rpc to 0 to help such servers out.

Signed-off-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-on: http://review.whamcloud.com/19932
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-8041
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/llite: ensure obd is effective in onu_upcall

The watched obd device may still not setup while onu_upcall
invoked. So we need verify it in cl_ocd_update.

Signed-off-by: Yang Sheng <yang.sheng@intel.com>
Reviewed-on: http://review.whamcloud.com/19597
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-8027
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/ldlm: const qualify struct lustre_handle * params

Add a const qualifier to several struct lustre_handle * parameters in
the LDLM interface.

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Reviewed-on: http://review.whamcloud.com/17071
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7403
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/llite: change it_data to it_request

Change the void *it_data member of struct lookup_intent to struct
ptlrpc_request *it_request.

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Reviewed-on: http://review.whamcloud.com/17070
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7403
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre: Inline Lustre intent disposition functions

They are just one-liners, so no point in having them exported
and called through a different module.

Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/llite: flatten struct lookup_intent

Replace the union in struct lookup_intent with the members of struct
lustre_indent_data. Remove the then unused struct lustre_intent_data.

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Reviewed-on: http://review.whamcloud.com/17069
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7403
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: Frank Zago <fzago@cray.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre: Add newline to LU_OBJECT_DEBUG() message

LU_OBJECT_DEBUG expects non \n terminated message from the caller,
so it should add it's own to keep debug logger happy.

Signed-off-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-on: http://review.whamcloud.com/19960
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-8094
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre: LDLM_DEBUG() shouldn't be passed \n

as it adds own \n, so any extra \n break log format.

Signed-off-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-on: http://review.whamcloud.com/17494
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7521
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/llite: take trunc_sem only at vvp layer

The lli_trunc_sem is taken in 'read' mode in both
ll_page_mkwrite and vvp_io_fault_start. This can lead to a
deadlock with another thread which asks for the semaphore
in write mode between thse two read calls.

Since all users of lli_trunc_sem are in the vvp layer, we
can satisfy the requirement to exclude truncate by taking
the semaphore only in vvp_io_fault_start.

Signed-off-by: Patrick Farrell <paf@cray.com>
Reviewed-on: http://review.whamcloud.com/19315
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7981
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Andriy Skulysh <andriy.skulysh@seagate.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/ptlrpc: lost bulk leads to a hang

The reverse order of request_out_callback() and reply_in_callback()
puts the RPC into UNREGISTERING state, which is waiting for RPC &
bulk md unlink, whereas only RPC md unlink has been called so far.
If bulk is lost, even expired_set does not check for UNREGISTERING
state.

The same for write if server returns an error.

This phase is ambiguous, split to UNREG_RPC and UNREG_BULK.

Signed-off-by: Vitaly Fertman <vitaly.fertman@seagate.com>
Seagate-bug-id:  MRP-2953, MRP-3206
Reviewed-by: Andriy Skulysh <andriy.skulysh@seagate.com>
Reviewed-by: Alexey Leonidovich Lyashkov <alexey.lyashkov@seagate.com>
Tested-by: Elena V. Gryaznova <elena.gryaznova@seagate.com>
Reviewed-on: http://review.whamcloud.com/19953
Reviewed-by: Chris Horn <hornc@cray.com>
Reviewed-by: Ann Koehler <amk@cray.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/ptlrpc: Remove __ptlrpc_request_bufs_pack

Combine __ptlrpc_request_bufs_pack into ptlrpc_request_bufs_pack
because it was an unnecessary wrapper otherwise.

Signed-off-by: Ben Evans <bevans@cray.com>
Reviewed-on: http://review.whamcloud.com/16765
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7269
Reviewed-by: Frank Zago <fzago@cray.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Chris Horn <hornc@cray.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/ptlrpc: Early Reply vs Reply MDunlink

A race between unregister_reply & early reply.
When buffers are busy for the early transfer, they cannon be unlinked
by unregister_reply, so the RPC gets into UNREGISTERING state. The
coming reply_in_callback for the early RPC already has unlinked flag
set due to previous mdunlink attempt, but we handle it properly only
for UNILNK event, whereas this is PUT in this case.

Signed-off-by: Vitaly Fertman <vitaly.fertman@seagate.com>
Seagate-bug-id: MRP-3323
Reviewed-by: Alexey Leonidovich Lyashkov <alexey.lyashkov@seagate.com>
Reviewed-by: Andriy Skulysh <andriy.skulysh@seagate.com>
Tested-by: Parinay Vijayprakash Kondekar <parinay.kondekar@seagate.com>
Reviewed-on: http://review.whamcloud.com/18934
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7434
Reviewed-by: Chris Horn <hornc@cray.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/ptlrpc: missing wakeup for ptlrpc_check_set

This patch changes a few things:

- There is no guarantee that request_out_callback will happen
  before reply_in_callback, if a request got reply and unlinked
  reply buffer before request_out_callback is called, then the
  thread waiting on ptlrpc_request_set will miss wakeup event.

  This may seriously impact performance of some IO workloads or
  result in RPC timeout

- To make code more easier to understand, this patch changes
  action-bits "rq_req_unlink" and "rq_reply_unlink" to
  status-bits "rq_req_unlinked" and "rq_reply_unlinked"

Signed-off-by: Liang Zhen <liang.zhen@intel.com>
Reviewed-on: http://review.whamcloud.com/12158
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-5696
Reviewed-by: Johann Lombardi <johann.lombardi@intel.com>
Reviewed-by: Li Wei <wei.g.li@intel.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/ptlrpc: reorganize ptlrpc_request

ptlrpc_request has some structure members are only for client side,
and some others are only for server side, this patch moved these
members to different structure then putting into an union.

By doing this, size of ptlrpc_request is decreased about 300 bytes,
besides saving memory, it also can reduce memory footprint while
processing.

Signed-off-by: Liang Zhen <liang.zhen@intel.com>
Reviewed-on: http://review.whamcloud.com/8806
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-181
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Bobi Jam <bobijam@gmail.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/osc: Fix reverted condition in osc_lock_weight

When imprting clio simplification patch, the check for
pbject got reversed by mistake when converting from
if (obj == NULL) it somehow became (if (obj) which is obviously wrong,
and so when it does hit, a crash was happening as result.

Fix the condition and all if fine now.

Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/osc: osc_lock_weight endless loop fix

With huge number of pages to scan by osc_lock_weight() it is likely
CLP_GANG_RESCHED is returned from osc_page_gang_lookup() and the scan
will be repeated again from the start. To be sure that the scan is
progressing across those restarts, next scan should be started from
the last scanned page index plus one.

Xyratex-bug-id: MRP-2145
Signed-off-by: Alexander Zarochentsev <alexander.zarochentsev@seagate.com>
Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-on: http://review.whamcloud.com/12362
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-5781
Reviewed-by: Bobi Jam <bobijam@hotmail.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/llite: lock i_lock before __d_drop()

There has been several Lustre Client crashes reported by sites
running with Lustre versions 2.1/2.5, all showing the same
dentry->d_hash->next corrupted pointer cause.

This patch fixes a regression that has been introduced since a
long time by commit :
(LU-506 kernel: FC15 - support dcache scalability changes.)

where i_lock protection usage has been removed and
that is likely to cause racy condition during dentry [un]hashing
and to be the root cause of these crashes.

Signed-off-by: Bruno Faccini <bruno.faccini@intel.com>
Reviewed-on: http://review.whamcloud.com/19287
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7973
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Yang Sheng <yang.sheng@intel.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/llite: Get rid of ll_lock_dcache/ll_unlock_dcache

These are just doing spin_lock/unlock on inode's i_lock,
so just do the spinlock directly to make the code more clear

Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/llite: correct request handling after ll_lookup_it()

In the FIFO cases of ll_atomic_open() and ll_lookup_nd() remove
spurious calls to ptlrpc_req_finished(). Explain that these cases are
unreachable in practice anyway.

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Reviewed-on: http://review.whamcloud.com/17068
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7402
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging/lustre/llite: allocate and free client cache asynchronously

Since the inflight request holds import refcount as well as export,
sometimes obd_disconnect() in client_common_put_super() can't put
the last refcount of OSC import (e.g. due to network disconnection),
this will cause cl_cache being accessed after free.

To fix this issue, ccc_users is used as cl_cache refcount, and
lov/llite/osc all hold one cl_cache refcount respectively, to avoid
the race that a new OST is being added into the system when the client
is mounted.
The following cl_cache functions are added:
- cl_cache_init(): allocate and initialize cl_cache
- cl_cache_incref(): increase cl_cache refcount
- cl_cache_decref(): decrease cl_cache refcount and free the cache
  if refcount=0.

Signed-off-by: Emoly Liu <emoly.liu@intel.com>
Reviewed-on: http://review.whamcloud.com/13746
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-6173
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: lustre: remove remote client support

There are several obsolete sub commands for lfs to work with
remote client. We do not support that anymore, and should be
deleted along with any kernel code related to remote client.

Signed-off-by: Fan Yong <fan.yong@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-6971
Reviewed-on: http://review.whamcloud.com/19789
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Merge 4.7-rc4 into staging-next

We want the fixes in here, and we can resolve a merge issue in
drivers/iio/industrialio-trigger.c

Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Linux 4.7-rc4

Merge branch 'for_linus' of git://git./linux/kernel/git/jack/linux-fs

Pull UDF fixes and a reiserfs fix from Jan Kara:
 "A couple of udf fixes (most notably a bug in parsing UDF partitions
  which led to inability to mount recent Windows installation media) and
  a reiserfs fix for handling kstrdup failure"

* 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs:
  reiserfs: check kstrdup failure
  udf: Use correct partition reference number for metadata
  udf: Use IS_ERR when loading metadata mirror file entry
  udf: Don't BUG on missing metadata partition descriptor

Merge tag 'dmaengine-fix-4.7-rc4' of git://git.infradead.org/users/vkoul/slave-dma

Pull dmaengine fixes from Vinod Koul:
 "Some fixes has piled up, so time to send them upstream.

  These fixes include:
   - at_xdmac fixes for residue and other stuff
   - update MAINTAINERS for dma dt bindings
   - mv_xor fix for incorrect offset"

* tag 'dmaengine-fix-4.7-rc4' of git://git.infradead.org/users/vkoul/slave-dma:
  dmaengine: mv_xor: Fix incorrect offset in dma_map_page()
  dmaengine: at_xdmac: double FIFO flush needed to compute residue
  dmaengine: at_xdmac: fix residue corruption
  dmaengine: at_xdmac: align descriptors on 64 bits
  MAINTAINERS: Add file patterns for dma device tree bindings