Hi Linus,
On Fri, May 12, 2017 at 11:04 AM, Linus Walleij
<linus.walleij@xxxxxxxxxx> wrote:
> On Tue, May 9, 2017 at 12:54 PM, Geert Uytterhoeven
> <geert@xxxxxxxxxxxxxx> wrote:
>
> Oops missed this:
>
>> Hence I think we should not use generic pin properties, but consider these
>> settings to be part of pinmux configuration.
>> As having large tables in the driver is undesirable, I think storing the
>> settings in the "pinmux" property (by encoding them as flags passed to the
>> RZA1_PINMUX() macro) is our best option.
>
> I think it is better to have large tables in the driver in this case.
Jacopo, Chris: Would two bits per pin/function (none, input, output, bidir)
be sufficient?
That makes one u16 per pin. So roughtly 12 ports x 16 pins => 384 bytes.
Plus code to handle it. After all not that bad...
> It is the lesser evil.
>
> Having unintelligible and hard to grasp stuff in the device tree that
> no user will understand or dare to touch is not good, then it is better
> to have it with the code, where it is being used, so the developers of
> the driver can see it when they are dealing with this (quirky) hardware.
>
> As you say this is actually fixing hardware bugs, we can expect these
> quirky tables to be gone in the next hardware generation, right?
Let's hope so. Chris has a better crystal ball than I have ;-)
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@xxxxxxxxxxxxxx
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
