First set of IIO and counter new device and feature support for the 5.16 cycle

This has been a busy cycle in both subsystems, so lots here and a few
more stragglers to come next week.

Counter
======

Most interesting element this time is the new chrdev based interface
for the counter subsystem.  Affects all drivers. Some minor precursor
patches.

Major parts:
* Bring all the sysfs attribute setup into the counter core rather than
  leaving it to individual drivers.  Docs updates accompany these changes.
* Move various definitions to a uapi header as now needed from userspace.
* Add the chardev interface + extensive documentation and example tool
* Add new ABI needed to identify indexes needed for chrdev interface
* Implement new interface for the 104-quad-8
* Follow up deals with wrong path for documentation build
* Various trivial cleanups and missing feature additions related to this
  series

IIO
===

Big new feature for IIO in this cycle is probably output buffer support.
This has been in the works for a very long time so it's great to see
Mihail pick up the challenge and build upon his predecessors work to finally
bring this feature to mainline.

New device support
------------------

* adi,adxl313
  - New driver and dt bindings for this low power accelerometer.
* adi,adxl355
  - New driver and dt bindings for this accelerometer.
  - Later series adds buffer support.
* asahi-kasei,ak8975
  - Minor additions to driver to support ak09916
* aspeed,aspeed-adc
  - Substantial rework plus feature additions to add support for the
    ast2600 including a new dt bindings doc.
* atmel,at91_sama5d2
  - Rework and support introduced for the sama7g5 parts.
* maxim,max31865
  - New driver and bindings for this RTD temperature sensor chip.
* nxp,imx8qxp
  - New driver and bindings for the ADC found on the i.MX 8QuadXPlus Soc.
* senseair,sunrise
  - New driver and bindings for this family of carbon dioxide gas sensors.
* sensiron,scd4x
  - New driver and bindings for this carbon dioxide gas sensor.

New features
------------

* Output buffer support.  Works in a similar fashion to input buffers, but
  in this case userspace pushes data into the kfifo which is then drained
  to the device when a trigger occurs.  Support added to the ad5766 DAC
  driver.
* Core, devm_iio_map_array_register() to avoid need for
  devm_add_action_or_reset() based cleanup in fully managed allocation
  drivers.
* Core iio_push_to_buffers_with_ts_unaligned() function to safely handle a
  few drivers where it really hard to ensure the correct data alignment in
  an iio_push_to_buffers_with_timestamp() call. Note this uses a bounce
  buffer so should be avoided whenever possible.  Used in the ti,adc108s102,
  invense,mpu3050 and adi,adis16400.  This closes the last   known set
  of drivers with alignment issues at this interface.
* maxim,max1027
  - Substantial rework to this driver main target of which was supporting
    use of other triggers than it's own EOC interrupt.
  - Transfer optimization.
* nxp,fxls8962af
  - Threshold even support including using it as a wakeup source.

Cleanups, minor fixes etc
-------------------------

Chances of a common type to multiple drivers:

* devm_ conversion and drop of .remove() callbacks in:
  - adi,ad5064
  - adi,ad7291
  - adi,ad7303
  - adi,ad7746
  - adi,ad9832
  - adi,adis16080
  - dialog,da9150-gpadc
  - intel,mrfld_adc
  - marvell,berlin2
  - maxim,max1363
  - maxim,max44000
  - nuvoton,nau7802
  - st_sensors (includes a lot of rework!)
  - ti,ads8344
  - ti,lp8788

* devm_platform_ioremap_resource() used to reduce boilerplate
  - cirrus,ep93xx
  - rockchip,saradc
  - stm,stm32-dac

* Use dev_err_probe() in more places to both not print on deferred probe and
  ensure a reason for the deferral is available for debug purposes.
  - adi,ad8801
  - capella,cm36651
  - linear,ltc1660
  - maxim,ds4424
  - maxim,max5821
  - microchip,mcp4922
  - nxp,lpc18xx
  - onnn,noa1305
  - st,lsm9ds0
  - st,st_sensors
  - st,stm32-dac
  - ti,afe4403
  - ti,afe4404
  - ti,dac7311

* Drop error returns in SPI and I2C remove() functions as they are ignored and
  long term plan is to change these all over to returning void. In some cases
  these patches just make it 'obvious' they always return 0 where it was the
  case before but not easy to tell.
  - adi,ad5380
  - adi,ad5446
  - adi,ad5686
  - adi,ad5592r
  - bosch,bma400
  - bosch,bmc150
  - fsl,mma7455
  - honeywell,hmc5843
  - kionix,kxsd9
  - maxim,max5487
  - meas,ms5611
  - ti,afe4403

Driver specific changes

* adi,ad5770r
  - Bring driver inline with documented bindings.
* adi,ad7746
  - Trivial style fix
* adi,ad7949
  - Express some magic values as the underlying parts via new #defines.
  - Make it work with SPI controllers that don't support 14 or 16 bit messages
  - Support selection of voltage reference from dt including expanding the
    dt-bindings to cover this new functionality.
* adi,ad799x
  - Implement selection of external reference voltage on AD7991, AD7995 and
    AD7999.
  - Add missing dt-bindings doc for devices supported by this driver.
* adi,adislib
  - Move interrupt startup to better location in startup flow.
  - Handle devices that cannot mask/unmask the drdy pin and must instead mask
    at the interrupt controller.  Applies to the adis16460 and adis16475 from
    which we then drop equivalent code.
* adi,ltc2983
  - Add support for optional reset pin.
  - Fail to probe if no channels specified in dt binding.
* asahi-kasei,ak8975
  - dt-binding additions of missing vid-supply regulator.
* aspeed,aspeed-adc
  - Typo fix.
* fsl,mma7660
  - Mark acpi_device_id table __maybe_unused to avoid build warning.
* fsl,imx25-gcq
  - Avoid initializing regulators that aren't used.
* invensense,mpu3050
  - Drop a dead protection against a clash with the old input driver.
* invensense,mpu6050
  - Rework code to not use strcpy() and hence avoid possibility of wrong sized
    buffers. Note this wasn't a bug, but the new code is a lot more readable.
  - Mark acpi_device_id table __maybe_unused to avoid build warning.
* kionix,kxcjk1013
  - dt-binding addition to note it supports interrupts.
* marvell,berlin2-adc
  - Enable COMPILE_TEST building.
* maxim,max1027
  - Avoid returning success in an error path.
* nxp,imx8qxp
  - Fix warning when runtime pm not enabled via __maybe_unused.
* ricoh,rn5t618
  - Use the new devm_iio_map_array_register() instead of open coding the same.
* samsung,exynos_adc
  - Improve kconfig help text.
* st,lsm6dsx
  - Move max_fifo_size into the fifo_ops structure where the other configuration
    parameters are found.
* st,st_sensors:
  - Reorder to ensure we turn the power off after removing userspace interfaces.
* senseair,sunrise
  - Add missing I2C dependency.
* ti,twl6030
  - Small code tidy up.
-----BEGIN PGP SIGNATURE-----
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=rEdi
-----END PGP SIGNATURE-----