Re: [PATCH RFC] gpio: brcmstb: Properly account for empty banks

[Gregory Fong <gregory.0xf0@xxxxxxxxx>]

Hi Justin,

I got this same patch 3 times in a row in one day, not sure why. :-)

On Wed, Jul 18, 2018 at 2:12 PM,  <justinpopo6@xxxxxxxxx> wrote:
> From: Justin Chen <justinpopo6@xxxxxxxxx>
>
> On some chips we have empty banks or a hole in the register space.
> The driver assumes all banks are consecutive and the same size.

It does assume both of these, but if this workaround works, then only
the former is a problem, right?  Consider omitting the mention of it
being the same size if it's not relevant for the change.

> To work around this, we create a fake bank where there is a hole
> and keep it out of the linked list so it won't be called or initialized

The idea seems OK, and it looks like this should already even be
handled properly by brcmstb_gpio_irq_map().  I think the main issue is
informational; the info line printed at the end of probe would give
the appearance of the driver registering more GPIOs than it actually
does.  Please update that to be clearer about what is actually
registered.

Best regards,
Gregory
--
To unsubscribe from this list: send the line "unsubscribe linux-gpio" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html