Re: [PATCH] serial: 8250_ingenic: Add support for the JZ4725B SoC

[Rob Herring]

On Fri, Jul 20, 2018 at 12:58 PM Paul Cercueil <paul@xxxxxxxxxxxxxxx> wrote:
>
> Hi Rob,
>
> Le ven. 20 juil. 2018 Ã 17:30, Rob Herring <robh@xxxxxxxxxx> a Ãcrit :
> > On Fri, Jul 13, 2018 at 04:38:40PM +0200, Paul Cercueil wrote:
> >>  The UART in the jz4725b works just like in the other JZ SoCs, so
> >> this
> >>  commit simply adds a new compatible string.
> >>
> >>  Signed-off-by: Paul Cercueil <paul@xxxxxxxxxxxxxxx>
> >>  ---
> >>   Documentation/devicetree/bindings/serial/ingenic,uart.txt | 1 +
> >>   drivers/tty/serial/8250/8250_ingenic.c                    | 5 +++++
> >>   2 files changed, 6 insertions(+)
> >>
> >>  diff --git
> >> a/Documentation/devicetree/bindings/serial/ingenic,uart.txt
> >> b/Documentation/devicetree/bindings/serial/ingenic,uart.txt
> >>  index c3c6406d5cfe..a0b34fd3b602 100644
> >>  --- a/Documentation/devicetree/bindings/serial/ingenic,uart.txt
> >>  +++ b/Documentation/devicetree/bindings/serial/ingenic,uart.txt
> >>  @@ -3,6 +3,7 @@
> >>   Required properties:
> >>   - compatible : One of:
> >>     - "ingenic,jz4740-uart",
> >>  +  - "ingenic,jz4725b-uart",
> >>     - "ingenic,jz4760-uart",
> >>     - "ingenic,jz4770-uart",
> >>     - "ingenic,jz4775-uart",
> >>  diff --git a/drivers/tty/serial/8250/8250_ingenic.c
> >> b/drivers/tty/serial/8250/8250_ingenic.c
> >>  index 15a8c8dfa92b..760266559a5a 100644
> >>  --- a/drivers/tty/serial/8250/8250_ingenic.c
> >>  +++ b/drivers/tty/serial/8250/8250_ingenic.c
> >>  @@ -133,6 +133,10 @@ EARLYCON_DECLARE(jz4740_uart,
> >> ingenic_early_console_setup);
> >>   OF_EARLYCON_DECLARE(jz4740_uart, "ingenic,jz4740-uart",
> >>                  ingenic_early_console_setup);
> >>
> >>  +EARLYCON_DECLARE(jz4725b_uart, ingenic_early_console_setup);
> >>  +OF_EARLYCON_DECLARE(jz4725b_uart, "ingenic,jz4725b-uart",
> >>  +               ingenic_early_console_setup);
> >>  +
> >
> > This shouldn't be necessary. Looks like it should be compatible with
> > ingenic,jz4740-uart.
> >
> >>   EARLYCON_DECLARE(jz4770_uart, ingenic_early_console_setup);
> >>   OF_EARLYCON_DECLARE(jz4770_uart, "ingenic,jz4770-uart",
> >>                  ingenic_early_console_setup);
> >>  @@ -330,6 +334,7 @@ static const struct ingenic_uart_config
> >> jz4780_uart_config = {
> >>
> >>   static const struct of_device_id of_match[] = {
> >>      { .compatible = "ingenic,jz4740-uart", .data =
> >> &jz4740_uart_config },
> >>  +   { .compatible = "ingenic,jz4725b-uart", .data =
> >> &jz4740_uart_config },
> >
> > And this too.
>
> Well, I'm confused, the driver already uses multiple compatible strings
> for SoCs that
> work the exact same, so that was wrong?

Not wrong, but not necessary. Given that the data was the same was the
clue telling me that they should be compatible.

Rob