[PATCH v2 1/2] ARM: dts: orange-pi-zero: specify wifi interrupts

[Icenowy Zheng]

于 2017年5月25日 GMT+08:00 下午3:21:44, Chen-Yu Tsai <wens at csie.org> 写到:
>On Thu, May 25, 2017 at 3:10 PM, Sergey Matyukevich
><geomatsi at gmail.com> wrote:
>>> > diff --git a/arch/arm/boot/dts/sun8i-h2-plus-orangepi-zero.dts
>b/arch/arm/boot/dts/sun8i-h2-plus-orangepi-zero.dts
>>> > index 9e8b082c134f..e97dfd1775d1 100644
>>> > --- a/arch/arm/boot/dts/sun8i-h2-plus-orangepi-zero.dts
>>> > +++ b/arch/arm/boot/dts/sun8i-h2-plus-orangepi-zero.dts
>>> > @@ -128,6 +128,9 @@
>>> >          */
>>> >         xr819: sdio_wifi at 1 {
>>> >                 reg = <1>;
>>> > +               compatible = "xradio,xr819";
>>>
>>> You should submit device tree bindings first. See:
>>>
>>>
>https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/tree/Documentation/devicetree/bindings/net/wireless
>>>
>>> for examples.
>>
>> Does it make sense for the out-of-tree driver ?
>
>There is already a binding for ESP8089, which is supported out of tree.
>There are also bindings for Mali GPUs which are, again, supported out
>of tree.
>
>AFAIK there was some discussion about this during the submission of the
>Mali bindings. A set of stable bindings allows the vendor to create
>stable
>device trees. Given that the binding for this WiFi chip is relatively
>simple,
>and it might be compatible or similar to an existing chip already
>supported
>in kernel (but without a binding), it should be straight forward.
>
>>
>>> > +               interrupt-parent = <&pio>;
>>> > +               interrupts = <6 10 IRQ_TYPE_EDGE_RISING>;
>>>
>>> Are these optional or required properties? AFAIK SDIO supports
>in-band
>>> interrupts with a bit of overhead, so they should be optional.
>>
>> Those properties are optional. Driver works anyway, though it
>complains about
>> missing interrupts. IIUC xradio driver falls back to polling mode
>only if
>> xr819 interrupt configuraton is not specified.
>
>So it doesn't support in-band SDIO interrupts right now. Is that a
>hardware
>or driver limitation?

The original driver code modified by me to mainline uses in-band interrupts.

But dgp modified it to use out-of-band interrupts.

>
>ChenYu