Re: [PATCH 7/8] ASoC: atmel: document clock properties of the wm8904 driver

[Mark Rutland]

On Mon, Mar 17, 2014 at 11:24:41AM +0000, Boris BREZILLON wrote:
> Le 17/03/2014 11:48, Mark Rutland a Ãcrit :
> > On Mon, Mar 17, 2014 at 10:18:08AM +0000, boris brezillon wrote:
> >> Hello Mark,
> >>
> >> Le 17/03/2014 11:02, Mark Rutland a Ãcrit :
> >>> On Mon, Mar 17, 2014 at 09:45:40AM +0000, Bo Shen wrote:
> >>>> From: Boris BREZILLON <b.brezillon@xxxxxxxxxxx>
> >>>>
> >>>> Signed-off-by: Boris BREZILLON <b.brezillon@xxxxxxxxxxx>
> >>>> Signed-off-by: Bo Shen <voice.shen@xxxxxxxxx>
> >>>> ---
> >>>>
> >>>>    Documentation/devicetree/bindings/sound/atmel-wm8904.txt | 2 ++
> >>>>    1 file changed, 2 insertions(+)
> >>>>
> >>>> diff --git a/Documentation/devicetree/bindings/sound/atmel-wm8904.txt b/Documentation/devicetree/bindings/sound/atmel-wm8904.txt
> >>>> index 8bbe50c..aca341c 100644
> >>>> --- a/Documentation/devicetree/bindings/sound/atmel-wm8904.txt
> >>>> +++ b/Documentation/devicetree/bindings/sound/atmel-wm8904.txt
> >>>> @@ -3,6 +3,8 @@ Atmel ASoC driver with wm8904 audio codec complex
> >>>>    Required properties:
> >>>>      - compatible: "atmel,asoc-wm8904"
> >>>>      - atmel,model: The user-visible name of this sound complex.
> >>>> +  - clocks: A list of clocks needed by the wm8904 chip.
> >>>> +  - clock-output-names: Driver related clock names. Shall contain "pck0".
> >>> The word "driver"  doesn't need to appear in biding documents, and this
> >>> fails to describe what it sets out to. How about the following:
> >>>
> >>> - clocks: a list of phandle + clock-specifier pairs, one for each entry
> >>>     in clock-names.
> >> Agreed.
> >>> - clock-names: Should contain "pck0"
> >> After thinking a bit more about it, this can be any programmable clk (pckX).
> > Huh?
> >
> > The clock-names property describes the names of the inputs to the
> > device, from the view of the device, not the names of the clocks fed
> > into those inputs.
> >
> > What are the names of the clock input lines on the wm8904?
> 
> mclk, and you're right, this is how we should name the clk (I just got
> influenced by the existing driver, which was requesting pck0).
> 
> This gives the following:
> 
> - clock-names: Should contain "mclk"
> 
> Do you agree ?

That sounds right to me.

Cheers,
Mark.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/