Re: [PATCH 5/5] dt-bindings: iio: Add adis16475 documentation

["Sa, Nuno" <Nuno.Sa@analog.com>]

On Thu, 2020-03-05 at 11:34 +0100, Lars-Peter Clausen wrote:
> On 3/4/20 7:00 PM, Sa, Nuno wrote:
> > On Tue, 2020-03-03 at 21:10 +0000, Jonathan Cameron wrote:
> > > 
> > > On Tue, 25 Feb 2020 13:41:52 +0100
> > > Nuno Sá <nuno.sa@analog.com> wrote:
> > > 
> > > > Document the ADIS16475 device devicetree bindings.
> > > > 
> > > > Signed-off-by: Nuno Sá <nuno.sa@analog.com>
> > > One thing inline on the burst mode stuff.
> > > 
> > > Thanks,
> > > 
> > > Jonathan
> > > 
> > > > ---
> > > >   .../bindings/iio/imu/adi,adis16475.yaml       | 130
> > > > ++++++++++++++++++
> > > >   MAINTAINERS                                   |   1 +
> > > >   2 files changed, 131 insertions(+)
> > > >   create mode 100644
> > > > Documentation/devicetree/bindings/iio/imu/adi,adis16475.yaml
> > > > 
> > > > diff --git
> > > > a/Documentation/devicetree/bindings/iio/imu/adi,adis16475.yaml
> > > > b/Documentation/devicetree/bindings/iio/imu/adi,adis16475.yaml
> > > > new file mode 100644
> > > > index 000000000000..c0f2146e000c
> > > > --- /dev/null
> > > > +++
> > > > b/Documentation/devicetree/bindings/iio/imu/adi,adis16475.yaml
> > > > @@ -0,0 +1,130 @@
> > > > +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
> > > > +%YAML 1.2
> > > > +---
> > > > +$id: http://devicetree.org/schemas/iio/imu/adi,adis16475.yaml#
> > > > +$schema: http://devicetree.org/meta-schemas/core.yaml#
> > > > +
> > > > +title: Analog Devices ADIS16475 and similar IMUs
> > > > +
> > > > +maintainers:
> > > > +  - Nuno Sá <nuno.sa@analog.com>
> > > > +
> > > > +description: |
> > > > +  Analog Devices ADIS16475 and similar IMUs
> > > > +
> > > > https://www.analog.com/media/en/technical-documentation/data-sheets/ADIS16475.pdf
> > > > +
> > > > +properties:
> > > > +  compatible:
> > > > +    enum:
> > > > +      - adi,adis16475-1
> > > > +      - adi,adis16475-2
> > > > +      - adi,adis16475-3
> > > > +      - adi,adis16477-1
> > > > +      - adi,adis16477-2
> > > > +      - adi,adis16477-3
> > > > +      - adi,adis16470
> > > > +      - adi,adis16465-1
> > > > +      - adi,adis16465-2
> > > > +      - adi,adis16465-3
> > > > +      - adi,adis16467-1
> > > > +      - adi,adis16467-2
> > > > +      - adi,adis16467-3
> > > > +      - adi,adis16500
> > > > +      - adi,adis16505-1
> > > > +      - adi,adis16505-2
> > > > +      - adi,adis16505-3
> > > > +      - adi,adis16507-1
> > > > +      - adi,adis16507-2
> > > > +      - adi,adis16507-3
> > > > +
> > > > +  reg:
> > > > +    maxItems: 1
> > > > +
> > > > +  spi-cpha: true
> > > > +
> > > > +  spi-cpol: true
> > > > +
> > > > +  spi-max-frequency:
> > > > +    maximum: 2000000
> > > > +
> > > > +  interrupts:
> > > > +    maxItems: 1
> > > > +
> > > > +  clocks:
> > > > +    maxItems: 1
> > > > +
> > > > +  clock-names:
> > > > +    oneOf:
> > > > +      - const: sync
> > > > +      - const: direct-sync
> > > > +      - const: pulse-sync
> > > > +      - const: scaled-sync
> > > > +
> > > > +  reset-gpios:
> > > > +    description:
> > > > +      Must be the device tree identifier of the RESET pin. If
> > > > specified,
> > > > +      it will be asserted during driver probe. As the line is
> > > > active low,
> > > > +      it should be marked GPIO_ACTIVE_LOW.
> > > > +    maxItems: 1
> > > > +
> > > > +  adi,scaled-output-hz:
> > > > +    description:
> > > > +      This property must be present if the clock mode is
> > > > scaled-
> > > > sync through
> > > > +      clock-names property. In this mode, the input clock can
> > > > have
> > > > a range
> > > > +      of 1Hz to 128HZ which must be scaled to originate an
> > > > allowable sample
> > > > +      rate. This property specifies that rate.
> > > > +    minimum: 1900
> > > > +    maximum: 2100
> > > > +
> > > > +required:
> > > > +  - compatible
> > > > +  - reg
> > > > +  - interrupts
> > > > +  - spi-cpha
> > > > +  - spi-cpol
> > > > +
> > > > +if:
> > > > +  properties:
> > > > +    compatible:
> > > > +      contains:
> > > > +        enum:
> > > > +          - adi,adis16500
> > > > +          - adi,adis16505-1
> > > > +          - adi,adis16505-2
> > > > +          - adi,adis16505-3
> > > > +          - adi,adis16507-1
> > > > +          - adi,adis16507-2
> > > > +          - adi,adis16507-3
> > > > +
> > > > +then:
> > > > +  properties:
> > > > +    clock-names:
> > > > +      oneOf:
> > > > +        - const: sync
> > > > +        - const: direct-sync
> > > > +        - const: scaled-sync
> > > > +
> > > > +    adi,burst32-enable:
> > > > +      description:
> > > > +        Enable burst32 mode. In this mode, a burst reading
> > > > contains calibrated
> > > > +        gyroscope and accelerometer data in 32-bit format.
> > > Why is this in DT?  Is it not a runtime decision
> > > (ideally automatically selected)
> > So, you mean just have this mode by default on parts that support
> > it?
> 
> Maybe lets start with explaining what burst32 mode is, so everybody
> is 
> on the same page.
> 
> The way registers are usually accessed for this chip is that you
> first 
> write the address you want to read on the SPI bus and then read the 
> selected register. This can be quite slow though if you want to read 
> multiple registers and is too slow to be able to read all the data
> at 
> full data rate. So there is a special burst mode which allows to
> read 
> all the data registers in one go.
> 
> Now by default the data registers are 16-bit. But there is an
> internal 
> decimation filter and the extra bits produced by the decimation
> filter 
> go into additional data registers making the data 32-bit wide. The
> chip 
> allows to configure whether to read the only 16-bit MSBs or the full 
> 32-bit register.
> 
> So the decision whether a user wants to use 32-bit or 16-bit depends
> on 
> whether the extra 16-bit are needed or if they are even available.
> E.g. 
> if the decimation filter is off there wont be any extra bits.
> 
> This means ideally it would be user configurable whether to use 16-
> bit 
> or 32-bit burst mode, since it is application specific. The problem
> is 
> we don't have an interface for changing the bit width of a buffer 
> channel. Adding such an interface would require quite a bit of
> effort 
> since the assumption currently is that the bit width does not
> chance. 
> E.g. libiio assumes this and would stop working if it did change.
> 
> Maybe as a compromise for now. Use 32-bit burst when there is
> actually 
> meaningful data is the LSBs, i.e. the decimation filter is used and 
> disable it otherwise. And then think about how to make it
> configurable 
> as a follow up action.

I do agree with that. Just to add that I think we also need to take
into account the FIR filter which can also be responsible for bit
growth. I will cache these values and if one of them is != 0 than burst
32 will be used...

> In my opinion there is should not be a difference in the userspace 
> interface for chips that do support 32-bit burst and those that
> don't. 
> For devices that don't support 32-bit burst it should be emulated by 
> reading the LSB bits registers manually.

Hmm. In terms of interface I think there is no difference. We always
report 32bits channels (for accel and gyro). However, what we do right
know is just to set the LSB to 0 if burst32 is not supported. So, we
can be just ignoring the LSB bits if they are being used...

- Nuno Sá
> - Lars
>