Re: [RFC/RFT PATCH 2/2] gpiolib: irqchip: get rid of irq_base

[Linus Walleij <linus.walleij@xxxxxxxxxx>]

On Fri, Sep 26, 2014 at 12:44 PM, Grygorii Strashko
<grygorii.strashko@xxxxxx> wrote:
> On 09/26/2014 11:44 AM, Linus Walleij wrote:
>>
>> On Thu, Sep 25, 2014 at 6:09 PM, Grygorii Strashko
>> <grygorii.strashko@xxxxxx> wrote:
>>
>>> Remove irq_base from struct gpio_chip, as it is seems to
>>> be unused.
>>> Aslo, using this field by drivers is unsafe because it's
>>> uncompatible with Sparse IRQ feature.
>>>
>>> Signed-off-by: Grygorii Strashko <grygorii.strashko@xxxxxx>
>>> ---
>>> Hi Linus,
>>>
>>> I've not found users of this field in drivers/gpio/ folder,
>>> so I've decided to created this patch to get more comments.
>>
>>
>> Wut?
>>
>> git grep '>irq_base'
>> (...)
>> drivers/gpio/gpio-max732x.c:    return chip->irq_base + off;
>
> ^struct max732x_chip *chip;
>
>> drivers/gpio/gpio-ml-ioh.c:     return chip->irq_base + offset;
>
> ^struct ioh_gpio *chip
>
>> drivers/gpio/gpio-pch.c:        chip->irq_base = irq_base;
>
> ^struct pch_gpio *chip
>
>> etc etc
>
> etc ;)

Aha now the variable names are confusing me at no end too :-)

> I've spent some time checking it, but it's possible
> that I missed smth or it's used outside gpio directory.

It'd be good if we could estimate this, I'd prefer if we could do
some semantic grep like cocinelle does to see if the (foo)->irq_base
affects a case where (foo) is struct gpio_chip...

But I guess I can also just apply the patch and throw it at
the autobuilders. I'm just worried about cases the autobuilder
would miss.

Yours,
Linus Walleij
--
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