Re: regressions in linux-next?

[Nishanth Menon <nm@ti.com>]

On 04/23/2014 08:01 AM, Linus Walleij wrote:
> On Wed, Apr 23, 2014 at 9:24 AM, Javier Martinez Canillas
> <javier@dowhile0.org> wrote:
> 
>> Linus, what do you think of the following patch?
>>
>> From ede333e85e0320d32e8c2d123560808ed7e43ece Mon Sep 17 00:00:00 2001
>> From: Javier Martinez Canillas <javier.martinez@collabora.co.uk>
>> Date: Wed, 23 Apr 2014 09:13:54 +0200
>> Subject: [PATCH 1/1] gpio: don't call irq_set_irq_type() on IRQ domain map
>>  function
> (...)
> 
> So no setting a default type in the mapping function...
> 
>> - irq_set_irq_type(irq, chip->irq_default_type);
> 
> But there are drivers exploiting this to set up the hardware to some
> default state :-(
> 
> What about this:
> 
> if (chip->irq_default_type != IRQ_TYPE_NONE)
>     irq_set_irq_type(irq, chip->irq_default_type);
> 
> This way you can pass IRQ_TYPE_NONE and nothing happens in
> the mapping.

What if these drivers depend on IRQ_TYPE_NONE to do something for the
GPIO pins?

would you expect these drivers to pass IRQ_TYPE_DEFAULT? OR I wonder
if we could pass some flag like -1 for platforms that dont care?


-- 
Regards,
Nishanth Menon