Just some minor grammar/spelling fixes of things I noticed while reading
the code.
Signed-off-by: NMartin Kelly <mkelly@xevo.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

When in spi mode, we are setting i2c disable bit in user_ctrl
register. But the register is overwritten after when turning fifo
on. So save user_ctrl init value and always use it when updating
the register.
Signed-off-by: NJean-Baptiste Maneyrol <jmaneyrol@invensense.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

Implement generic skip first samples mechanism and use it to
filter out first sample when gyro is on.

The problem for these chips is that the first sample of the gyro
is out of specs, because gyro is not completely stabilized. To
ensure all data are within sensor specs, we just skip the first
sample when turning gyro on.
Signed-off-by: NJean-Baptiste Maneyrol <jmaneyrol@invensense.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

Detected by kernel circular locking dependency checker.

We are locking iio mutex (iio_device_claim_direct_mode) after
locking our internal mutex. But when the buffer starts, iio first
locks its mutex and then we lock our internal one.

To avoid possible deadlock, we need to use the same order
everwhere. So we change the ordering by locking first iio mutex,
then our internal mutex.

Fixes: 68cd6e5b ("iio: imu: inv_mpu6050: fix lock issues by using our own mutex")
Cc: stable@vger.kernel.org
Signed-off-by: NJean-Baptiste Maneyrol <jmaneyrol@invensense.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

icm20608 does not have i2c aux bus, so it does not make sense to
allocate an i2c mux for this chip.
Signed-off-by: NJean-Baptiste Maneyrol <jmaneyrol@invensense.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

The iio resolver drivers in staging use angle channels. This patch
add missing documentation for this type of channel.

As was discussed in [1], radians is chosen as the unit, to match the
unit of angular velocity.

[1] https://marc.info/?l=linux-driver-devel&m=152190078308330&w=2Signed-off-by: NDavid Veenstra <davidjulianveenstra@gmail.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

The manual states that the data is contained in the upper 12 bits
of the 16 bits read by spi. The code that extracts these 12 bits
is correct for both be and le machines, but this is not clear
from a first glance.

To improve readability the relevant expressions are replaced
with equivalent expressions that use be16_to_cpup.
Signed-off-by: NDavid Veenstra <davidjulianveenstra@gmail.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

If an ADC channel measures the midpoint of a voltage divider, the
interesting voltage is often the voltage over the full resistance.
E.g. if the full voltage is too big for the ADC to handle.
Likewise, if an ADC channel measures the voltage across a shunt
resistor, with or without amplification, the interesting value is
often the current through the resistor.

This driver solves these problems by allowing to linearly scale a channel
and/or by allowing changes to the type of the channel.
Signed-off-by: NPeter Rosin <peda@axentia.se>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

Similar to current sense shunts, but an amplifier enables the use
of a smaller sense resistance.
Signed-off-by: NPeter Rosin <peda@axentia.se>
Reviewed-by: NRob Herring <robh@kernel.org>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

An ADC is often used to measure other quantities indirectly. This
binding describe one cases, a "big" voltage measured with the help
of a voltage divider.
Signed-off-by: NPeter Rosin <peda@axentia.se>
Reviewed-by: NRob Herring <robh@kernel.org>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

An ADC is often used to measure other quantities indirectly. This
binding describe one cases, a current through a shunt resistor
measured by the voltage over it.
Signed-off-by: NPeter Rosin <peda@axentia.se>
Reviewed-by: NRob Herring <robh@kernel.org>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

The regmap core now handles splitting up transactions according to
max_raw_read, so this code is no longer required in client drivers.
Signed-off-by: NCharles Keepax <ckeepax@opensource.cirrus.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

Delete the useless ored result and give a second chance to turn
the chip back off at the end.
Signed-off-by: NJean-Baptiste Maneyrol <jmaneyrol@invensense.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

Factorize reading channel data in its own function.
Signed-off-by: NJean-Baptiste Maneyrol <jmaneyrol@invensense.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

Use devm_* for iio_triggered_buffer_setup, iio_device_register,
iio_trigger_register. Delete unneeded inv_mpu6050_remove_trigger,
inv_mpu_core_remove, and inv_mpu_remove for spi driver.
Signed-off-by: NJean-Baptiste Maneyrol <jmaneyrol@invensense.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

Some functions are turning the chip on and not back off in error
path. With set_power function using a reference counter that
would keep the chip on forever.
Signed-off-by: NJean-Baptiste Maneyrol <jmaneyrol@invensense.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

spi-max-frequency is requested for SPI master mode (only), to tune output
clock. It may happen requested frequency isn't reachable.
Add explicit check, so probe fails with error in this case. Otherwise,
output clock may simply be silently turned off (conversions fail).
Signed-off-by: NFabrice Gasnier <fabrice.gasnier@st.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

reg property should be checked against number of available filters.
BTW, dfsdm->num_fls wasn't used. But it can be used for this purpose.
This prevents using data out of allocated dfsdm->fl_list array.
Signed-off-by: NFabrice Gasnier <fabrice.gasnier@st.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

Misc fixes & style improvements:
- checkpatch warns about line over 80 characters.
- remove extra spaces and a blank line (e.g. checkpatch --strict)
- remove bad error message always printed in probe routine.
Signed-off-by: NFabrice Gasnier <fabrice.gasnier@st.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

Add blank lines to improve readability.
Signed-off-by: NDavid Veenstra <davidjulianveenstra@gmail.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

Reorders the variable declarations to prefer a reverse Christmas tree
order to improve readability.
Signed-off-by: NDavid Veenstra <davidjulianveenstra@gmail.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

This patches sorts all the includes in alphabetic order.
Signed-off-by: NDavid Veenstra <davidjulianveenstra@gmail.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

This patches removes unneeded slab.h header.
Signed-off-by: NDavid Veenstra <davidjulianveenstra@gmail.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

The configuration register on the device is represented with the
prox_config member on the tsl2x7x_settings structure. According to the
TSL2772 data sheet, this register can hold: 1) the proximity drive
level, 2) ALS/Proximity long wait, and 3) the ALS gain level. This
patch renames prox_config to als_prox_config since ALS settings can
be stored here as well.
Signed-off-by: NBrian Masney <masneyb@onstation.org>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

This patch removes several unnecessary comments, changes some comments
so that the use as much of the allowable 80 characters as possible, adds
the proper whitespace, removes some structure members from the kernel
docs that are no longer present, and improves the existing kernel doc
information for some existing structure members.
Signed-off-by: NBrian Masney <masneyb@onstation.org>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

This patch changes the defaults of the als_time, prox_time and
wait_time to match the defaults according to the TSL2772 datasheet.
Signed-off-by: NBrian Masney <masneyb@onstation.org>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

The driver mostly uses the 'prox' naming convention for most of the
proximity settings, however prx_time and tsl2x7x_prx_gain was present.
This patch renames these to prox_time and tsl2x7x_prox_gain for
consistency with everything else in the driver.

The kernel documentation for prx_gain is corrected to prox_gain so that
it matches what is actually in the structure.
Signed-off-by: NBrian Masney <masneyb@onstation.org>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

The power and diode defines are needed for the platform data so this
patch moves the defines out of the .c file and into the header file. A
comment for the diode is also cleaned up while this code is touched.
Signed-off-by: NBrian Masney <masneyb@onstation.org>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

This patch renames tsl2x7x_device_id() to tsl2x7x_device_id_verif(),
removes the unnecessary pointer on the id parameter, and only calls
the verification function once.
Signed-off-by: NBrian Masney <masneyb@onstation.org>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

tsl2x7x_write_interrupt_config() has an unnecessary return value check
at the end of the function. This patch changes the function to just
return the value from the call to tsl2x7x_invoke_change().
Signed-off-by: NBrian Masney <masneyb@onstation.org>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

tsl2x7x_suspend() and tsl2x7x_resume() both check to see what the
current chip status is. These checks are not necessary so this patch
removes those checks.
Signed-off-by: NBrian Masney <masneyb@onstation.org>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

tsl2x7x_clear_interrupts() takes a reg argument but there are only
two callers to this function and both callers pass the same value.
Since this function was introduced, interrupts are now working
properly for this driver, and several unnecessary calls to
tsl2x7x_clear_interrupts() were removed. This patch removes the
tsl2x7x_clear_interrupts() function and replaces the two callers
with the i2c_smbus_write_byte() call instead.
Signed-off-by: NBrian Masney <masneyb@onstation.org>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

tsl2x7x_event_handler() could return an error and this could cause the
interrupt to remain masked. We shouldn't return an error in the
interrupt handler so this patch always returns IRQ_HANDLED. An error
will be logged if one occurs.
Signed-off-by: NBrian Masney <masneyb@onstation.org>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

The integration_time* attributes are currently associated with the
IIO_LIGHT channel but should be associated with the IIO_INTENSITY
channel. Directory listing of the sysfs attributes for a TSL2772
with this patch applied:

dev
events
in_illuminance0_calibrate
in_illuminance0_calibscale_available
in_illuminance0_input
in_illuminance0_lux_table
in_illuminance0_target_input
in_intensity0_calibbias
in_intensity0_calibscale
in_intensity0_integration_time
in_intensity0_integration_time_available
in_intensity0_raw
in_intensity1_raw
in_proximity0_calibrate
in_proximity0_calibscale
in_proximity0_calibscale_available
in_proximity0_raw
name
of_node
power
subsystem
uevent
Signed-off-by: NBrian Masney <masneyb@onstation.org>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

We should get drvdata from struct device directly. Going via
platform_device is an unneeded step back and forth.
Signed-off-by: NWolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

We should get drvdata from struct device directly. Going via
platform_device is an unneeded step back and forth.
Signed-off-by: NWolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

We should get drvdata from struct device directly. Going via
platform_device is an unneeded step back and forth.
Signed-off-by: NWolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

The Lenovo Ideapad Miix 320 uses an st accelerometer with an ACPI id
of SMO8840, add support for this.
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

Document that the hardware supports falling edge, rising edge, level
low, and level high interrupt types, rather than just rising edge.

The language used is the same as that in st_lsm6dsx.txt.
Signed-off-by: NMartin Kelly <mkelly@xevo.com>
Reviewed-by: NRob Herring <robh@kernel.org>
Acked-by: NJean-Baptiste Maneyrol <jmaneyrol@invensense.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>

Currently, we support only rising edge interrupts, and in fact we assume
that the interrupt we're given is rising edge (and things won't work if
it's not). However, the device supports rising edge, falling edge, level
low, and level high interrupts.

Empirically, on my system, switching to level interrupts has fixed a
problem I had with significant (~40%) interrupt loss with edge
interrupts. This issue is likely related to the SoC I'm using (Allwinner
H3), but being able to switch the interrupt type is still a very useful
workaround.

I tested this with each interrupt type and verified correct behavior in
a logic analyzer.

Add support for these interrupt types while also eliminating the error
case of the device tree and driver using different interrupt types.
Signed-off-by: NMartin Kelly <mkelly@xevo.com>
Acked-by: NJean-Baptiste Maneyrol <jmaneyrol@invensense.com>
Signed-off-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>