Re: [PATCH v5 5/8] dt-bindings: media: add bindings for TI DS90UB960

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi Tomi,

On Tue, Dec 13, 2022 at 04:25:46PM +0200, Tomi Valkeinen wrote:
> On 11/12/2022 19:58, Laurent Pinchart wrote:
> > On Thu, Dec 08, 2022 at 12:40:03PM +0200, Tomi Valkeinen wrote:
> >> Add DT bindings for TI DS90UB960 FPDLink-3 Deserializer.
> >>
> >> Signed-off-by: Tomi Valkeinen <tomi.valkeinen@xxxxxxxxxxxxxxxx>
> >> ---
> >>   .../bindings/media/i2c/ti,ds90ub960.yaml      | 358 ++++++++++++++++++
> >>   1 file changed, 358 insertions(+)
> >>   create mode 100644 Documentation/devicetree/bindings/media/i2c/ti,ds90ub960.yaml
> >>
> >> diff --git a/Documentation/devicetree/bindings/media/i2c/ti,ds90ub960.yaml b/Documentation/devicetree/bindings/media/i2c/ti,ds90ub960.yaml
> >> new file mode 100644
> >> index 000000000000..d8b5e219d420
> >> --- /dev/null
> >> +++ b/Documentation/devicetree/bindings/media/i2c/ti,ds90ub960.yaml
> >> @@ -0,0 +1,358 @@
> >> +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
> >> +%YAML 1.2
> >> +---
> >> +$id: http://devicetree.org/schemas/media/i2c/ti,ds90ub960.yaml#
> >> +$schema: http://devicetree.org/meta-schemas/core.yaml#
> >> +
> >> +title: Texas Instruments DS90UB9XX Family FPD-Link Deserializer Hubs
> >> +
> >> +maintainers:
> >> +  - Tomi Valkeinen <tomi.valkeinen@xxxxxxxxxxxxxxxx>
> >> +
> >> +description:
> >> +  The TI DS90UB9XX devices are FPD-Link video deserializers with I2C and GPIO
> >> +  forwarding.
> >> +
> >> +properties:
> >> +  compatible:
> >> +    enum:
> >> +      - ti,ds90ub960-q1
> >> +      - ti,ds90ub9702-q1
> >> +
> >> +  reg:
> >> +    maxItems: 1
> >> +    description:
> >> +      i2c addresses for the deserializer and the serializers
> > 
> > s/i2c/I2C/
> > 
> > Same below.
> > 
> > A bit more details would be nice, for instance the order in which
> > addresses should be specified should be documented. The example below
> > has one address only, so it's quite unclear. Or is this a left-over,
> > from before the i2c-alias-pool ?
> 
> That's a left over, but not related to i2c-alias-pool but the i2c-alias 
> for the serializers. It already says above 'maxItems: 1', so now it only 
> contains the deserializer address. I'll drop the desc.

Looks good to me.

> >> +
> >> +  clocks:
> >> +    maxItems: 1
> >> +    description:
> >> +      Reference clock connected to the REFCLK pin.
> >> +
> >> +  clock-names:
> >> +    items:
> >> +      - const: refclk
> >> +
> >> +  powerdown-gpios:
> >> +    maxItems: 1
> >> +    description:
> >> +      Specifier for the GPIO connected to the PDB pin.
> >> +
> >> +  i2c-alias-pool:
> >> +    $ref: /schemas/types.yaml#/definitions/uint16-array
> >> +    description:
> >> +      i2c alias pool is a pool of i2c addresses on the main i2c bus that can be
> >> +      used to access the remote peripherals. The addresses must be available,
> >> +      not used by any other peripheral. Each remote peripheral is assigned an
> >> +      alias from the pool, and transactions to that address will be forwarded
> >> +      to the remote peripheral, with the address translated to the remote
> >> +      peripheral's real address.
> > 
> > As this property is optional, should you describe what happens when it's
> > not specified ?
> > 
> > I would also indicate that the pool doesn't cover the serializers, only
> > the devices behind them.
> 
> Yep, I'll clarify these.
> 
> >> +
> >> +  links:
> >> +    type: object
> >> +    additionalProperties: false
> >> +
> >> +    properties:
> >> +      '#address-cells':
> >> +        const: 1
> >> +
> >> +      '#size-cells':
> >> +        const: 0
> >> +
> >> +      ti,manual-strobe:
> >> +        type: boolean
> >> +        description:
> >> +          Enable manual strobe position and EQ level
> >> +
> >> +    patternProperties:
> >> +      '^link@[0-9a-f]+$':
> > 
> > There can be up to 4 links only, right ? I would then use
> > 
> >        '^link@[0-3]$':
> 
> Yes, I'll change that.
> 
> >> +        type: object
> >> +        additionalProperties: false
> >> +        properties:
> >> +          reg:
> >> +            description: The link number
> >> +            maxItems: 1
> >> +
> >> +          i2c-alias:
> >> +            description:
> >> +              The i2c address used for the serializer. Transactions to this
> >> +              address on the i2c bus where the deserializer resides are
> >> +              forwarded to the serializer.
> >> +
> >> +          ti,rx-mode:
> >> +            $ref: /schemas/types.yaml#/definitions/uint32
> >> +            enum:
> >> +              - 0 # RAW10
> >> +              - 1 # RAW12 HF
> >> +              - 2 # RAW12 LF
> >> +              - 3 # CSI2 SYNC
> >> +              - 4 # CSI2 NON-SYNC
> >> +            description: FPD-Link Input Mode
> > 
> > Are there use cases for controlling this dynamically (in particular the
> > sync/non-sync modes) ? Is there anything that could be queried at
> > runtime from the serializers instead of being specified in DT ?
> 
> We need a link to the serializer before we can query anything from the 
> serializer.

I meant querying it from the serializer driver, not the serializer
hardware. This being said, it would likely be difficult to do so, as the
serializer driver would need to probe first. I think I'm thus fine
selecting the mode in DT on the deserializer side.

> To have a link, we need the mode... So, as I mentioned in 
> the other reply, we could define these in some way in the serializer's 
> properties instead of here, but I'm not sure if that's a good change.
> 
> The driver can change the mode at runtime (say, from sync to non-sync 
> mode, if the HW supports that). But I think this property should reflect 
> the HW strapped configuration of the serializer.

That would possibly work for the DS90UB953, but the DS90UB913 has no
strapped mode selected at boot time but is instead configured
automatically through the back-channel (see my last reply to patch 3/8).

When connecting a DS90UB913 to a DS90UB914 deserializer, we can probably
start without mode selection in software, as the MODE pin is meant to
bootstrap that to a correct value which is then automatically
transmitted to the serializer (hardware designs where the mode would
need to be overridden should be rate). However, when connecting multiple
DS90UB913 to a DS90UB960, I can imagine connecting different types of
cameras on the four input ports, so the need to specify the mode
per-port in DT would be more common.

For these reasons, I don't think the ti,rx-mode property can be defined
as reflecting the hardware MODE strap with the DS90UB913. I also think
it would be quite confusing to define it as the desired runtime
configuration for the DS90UB913 and as the hardware MODE strap for the
DS90UB953. Could it be (explicitly) defined as the desired runtime
configuration in all cases ?

> > Same question for the parameters below. Additionally, are there any
> > parameters that need to be identical for all links ?
> 
> The same answer to the cdr-mode. No need to be identical.
> 
> The strobe-pos and eq-level are unrelated to this topic.
> 
> >> +
> >> +          ti,cdr-mode:
> >> +            $ref: /schemas/types.yaml#/definitions/uint32
> >> +            enum:
> >> +              - 0 # FPD3
> >> +              - 1 # FPD4
> >> +            description: FPD-Link CDR Mode
> >> +
> >> +          ti,strobe-pos:
> >> +            $ref: /schemas/types.yaml#/definitions/int32
> >> +            minimum: -13
> >> +            maximum: 13
> >> +            description: Manual strobe position
> >> +
> >> +          ti,eq-level:
> >> +            $ref: /schemas/types.yaml#/definitions/uint32
> >> +            maximum: 14
> >> +            description: Manual EQ level
> >> +
> >> +          serializer:
> >> +            type: object
> >> +            description: FPD-Link Serializer node
> >> +
> >> +        required:
> >> +          - reg
> >> +          - i2c-alias
> >> +          - ti,rx-mode
> >> +          - serializer
> >> +
> >> +  ports:
> >> +    $ref: /schemas/graph.yaml#/properties/ports
> >> +
> >> +    properties:
> >> +      port@0:
> >> +        $ref: /schemas/graph.yaml#/properties/port
> >> +        unevaluatedProperties: false
> >> +        description: FPD-Link input 0
> >> +
> >> +      port@1:
> >> +        $ref: /schemas/graph.yaml#/properties/port
> >> +        unevaluatedProperties: false
> >> +        description: FPD-Link input 1
> >> +
> >> +      port@2:
> >> +        $ref: /schemas/graph.yaml#/properties/port
> >> +        unevaluatedProperties: false
> >> +        description: FPD-Link input 2
> >> +
> >> +      port@3:
> >> +        $ref: /schemas/graph.yaml#/properties/port
> >> +        unevaluatedProperties: false
> >> +        description: FPD-Link input 3
> >> +
> >> +      port@4:
> >> +        $ref: /schemas/graph.yaml#/$defs/port-base
> >> +        unevaluatedProperties: false
> >> +        description: CSI-2 Output 0
> >> +
> >> +        properties:
> >> +          endpoint:
> >> +            $ref: /schemas/media/video-interfaces.yaml#
> >> +            unevaluatedProperties: false
> >> +
> >> +            properties:
> >> +              data-lanes:
> >> +                minItems: 1
> >> +                maxItems: 4
> >> +
> >> +      port@5:
> >> +        $ref: /schemas/graph.yaml#/$defs/port-base
> >> +        unevaluatedProperties: false
> >> +        description: CSI-2 Output 1
> >> +
> >> +        properties:
> >> +          endpoint:
> >> +            $ref: /schemas/media/video-interfaces.yaml#
> >> +            unevaluatedProperties: false
> >> +
> >> +            properties:
> >> +              data-lanes:
> >> +                minItems: 1
> >> +                maxItems: 4
> > 
> > The ports should be mandatory, shouldn't they ?
> 
> Did you mean data-lanes? Yes, data-lanes should be mandatory.

Yes that's what I meant, sorry.

-- 
Regards,

Laurent Pinchart



[Index of Archives]     [Linux GPIO]     [Linux SPI]     [Linux Hardward Monitoring]     [LM Sensors]     [Linux USB Devel]     [Linux Media]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux