[PATCH] ARM: tegra: consolidate GIC extension usage

[ccross@google.com (Colin Cross)]

On Wed, Apr 27, 2011 at 10:38 AM, Will Deacon <will.deacon@arm.com> wrote:
> Hi Colin,
>
> On Wed, 2011-04-27 at 18:34 +0100, Colin Cross wrote:
>> On Wed, Apr 27, 2011 at 4:44 AM, Marc Zyngier <marc.zyngier@arm.com> wrote:
>> > On Wed, 27 Apr 2011 09:36:57 +0100
>> > Russell King - ARM Linux <linux@arm.linux.org.uk> wrote:
>> >
>> >> On Wed, Apr 20, 2011 at 11:58:24AM +0100, Marc Zyngier wrote:
>> >>
>> >> A good follow-on step would be...
>> >>
>> >> > ?static void tegra_mask(struct irq_data *d)
>> >> > ?{
>> >> > - ? tegra_gic_mask_irq(d);
>> >> > ? ? tegra_legacy_mask_irq(d->irq);
>> >> > ?}
>> >>
>> >> ... arranging tegra_legacy_mask_irq() etc to take struct irq_data directly.
>> >> These functions could then be removed entirely as they serve no useful
>> >> purpose other than transforming the new API back to the old API.
>> >
>> > How about the following patch? I haven't touched the gic_arch_extn part
>> > as it would break Will's fastEOI patch set (which depends on this one),
>> > but will prepare another patch cleaning up that part.
>>
>> I have a patch that fixes all of this and more as part of the Tegra PM
>> series, but I was sitting on it assuming Tegra PM would not go into
>> 2.6.40. ?Let me sort out the cleanups from the PM stuff, and I'll post
>> a patch that drops legacy_irq.c entirely.
>
> Ok. Where does that leave my GIC fasteoi stuff? I'm becoming more
> conscious of the chained_entry/exit functions sitting there unused in
> the kernel and it would be good to get the plumbed into all of the
> platforms sooner rather than later.
>
> Ideally your cleanup patch will mean that I don't have to touch the
> Tegra irq.c, leaving me with just the GPIO changes (which you can Ack if
> you like :).

The irq.c changes call your chained functions.  I have changes for
gpio.c that will conflict as well, and I haven't had a chance to test
yours and reconcile them with mine.  I'll try to get to it this week.