Re: [PATCH] ASoC: fsl-asrc: Convert to use regmap framework's endianness method.

[Mark Rutland <mark.rutland@arm.com>]

On Mon, Aug 18, 2014 at 01:03:14PM +0100, Li.Xiubo@freescale.com wrote:
> >>  Documentation/devicetree/bindings/sound/fsl,asrc.txt | 10 +++++++---
> >>  sound/soc/fsl/fsl_asrc.c                             |  6 +-----
> >>  2 files changed, 8 insertions(+), 8 deletions(-)
> >>
> >> diff --git a/Documentation/devicetree/bindings/sound/fsl,asrc.txt b/Documentation/devicetree/bindings/sound/fsl,asrc.txt
> >> index b93362a..791f372 100644
> >> --- a/Documentation/devicetree/bindings/sound/fsl,asrc.txt
> >> +++ b/Documentation/devicetree/bindings/sound/fsl,asrc.txt
> >> @@ -26,9 +26,12 @@ Required properties:
> >>       "ipg"             Peripheral clock to driver module.
> >>       "asrck_<0-f>"     Clock sources for input and output clock.
> >>
> >> -   - big-endian              : If this property is absent, the little endian mode
> >>-                       will be in use as default. Otherwise, the big endian
> >> -                       mode will be in use for all the device registers.
> >> +   - big-endian              : If this property is absent, the native endian mode
> >> +                       (same with CPU) will be in use as default. Otherwise,
> >> +                       the big endian mode will be in use for all the device
> >> +                       registers.
> >> +                       See Documentation/devicetree/bindings/regmap/regmap.txt
> >> +                       for more detail.
> >
> >Why does this have to change the semantics of the DT binding?
> >
> 
> I'm thinking that maybe in the late fulture, this device will be
> applied to some PowerPC SoC, from the regmap framework code, we can
> see that the 'big-endian' property could be ignored.
> 
> So,in this case,  if it is absent, the default endian mode should be
> used as defualt or native as the regmap framework said.

As I have mentioned in the past w.r.t. endianness bindings, there is no
such thing as a "default endianness" or "native endianness".

PowerPC and ARM can be Bi-endian, configured by the kernel.

The hardware's registers have a fixed endianness regardless of this
runtime configuration.

So describe that fixed property, as that does not vary with kernel
configuration (and is therefore a property of the HW rather than the
combination of HW + kernel).

Thanks,
Mark.