lkml.org 
[lkml]   [2013]   [Feb]   [8]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH 09/14] pinctrl/abx500: use direct IRQ defines
On Thu, 07 Feb 2013, Stephen Warren wrote:

> On 02/07/2013 02:01 AM, Lee Jones wrote:
> > I don't see myself on cc. Was that intentional?
>
> The original patch was that way; I assume git send-email only CC'd you
> on patches written by you.

No, I didn't send this patch at all.

I was asking Linus if he ment to CC me, as I thought I should have been.

> > I quite like the idea of this.
> >
> > Stephen,
> >
> > It doesn't mean the other patch was wrong, it just transfers the math.
>
> Ah, I see. The issue is that the code below clearly calculates the hwirq
> differently, and it wasn't immediately obvious that this part of the
> patch for example:
>
> > struct abx500_gpio_irq_cluster ab8500_gpio_irq_cluster[] = {
> > - GPIO_IRQ_CLUSTER(6, 13, 34),
> > - GPIO_IRQ_CLUSTER(24, 25, 24),
> > - GPIO_IRQ_CLUSTER(36, 41, 14),
> > + GPIO_IRQ_CLUSTER(6, 13, AB8500_INT_GPIO6R),
> > + GPIO_IRQ_CLUSTER(24, 25, AB8500_INT_GPIO24R),
> > + GPIO_IRQ_CLUSTER(36, 41, AB8500_INT_GPIO36R),
> > };
>
> ... actually changes the values in the table (AB8500_INT_GPIO6R is 40,
> so when using that value, you need to subtract of the value 6 for the
> base to get the original 34).

Yes, I see how that may of looked if you didn't see the other change.

So you're happy?

> > I wouldn't squash it into mine. I like the transition and the
> > possibility to revert it if there's been some mistake.
> >
> > (not to say there is one, but just in case.)
> >
> > Sent from my mobile Linux device.
> >
> > On Feb 7, 2013 12:14 AM, "Stephen Warren" <swarren@wwwdotorg.org
> > <mailto:swarren@wwwdotorg.org>> wrote:
> >
> > On 02/05/2013 12:48 PM, Linus Walleij wrote:
> > > From: Linus Walleij <linus.walleij@linaro.org
> > <mailto:linus.walleij@linaro.org>>
> > >
> > > Make it harder to do mistakes by introducing the actual
> > > defined ABx500 IRQ number into the IRQ cluster definitions.
> > > Deduct cluster offset from the GPIO offset to make each
> > > cluster coherent.
> >
> > Shouldn't this patch be squashed into the previous patch to avoid churn?
> >
> > > static struct abx500_pinctrl_soc_data ab9540_soc = {
> >
> > > @@ -273,8 +273,7 @@ static int abx500_gpio_to_irq(struct gpio_chip
> > *chip, unsigned offset)
> >
> > > - hwirq = gpio + cluster->to_irq;
> > > -
> > > + hwirq = gpio - cluster->start + cluster->to_irq;
> > > return
> > irq_create_mapping(pct->parent->domain, hwirq);
> >
> > In particular, this change implies that the previous patch was simply
> > incorrect, although I haven't really thought about it in detail.
> >
>

--
Lee Jones
Linaro ST-Ericsson Landing Team Lead
Linaro.org │ Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2013-02-08 10:01    [W:0.041 / U:0.840 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site