[PATCH 1/6] gpio/mxs: adopt irq_domain support for mxs gpio driver

From: shawn.guo@linaro.org (Shawn Guo)
To: linux-arm-kernel@lists.infradead.org
Subject: [PATCH 1/6] gpio/mxs: adopt irq_domain support for mxs gpio driver
Date: Wed, 22 Aug 2012 10:14:27 +0800	[thread overview]
Message-ID: <20120822021424.GO24242@S2101-09.ap.freescale.net> (raw)
In-Reply-To: <50338A35.9070803@gmail.com>

On Tue, Aug 21, 2012 at 08:16:37AM -0500, Rob Herring wrote:
> On 08/21/2012 08:01 AM, Linus Walleij wrote:
> > On Mon, Aug 20, 2012 at 4:19 PM, Shawn Guo <shawn.guo@linaro.org> wrote:
> > 
> >> Remove irq_domain_add_legacy call from mach-mxs.c and have the gpio
> >> driver adopt irqdomain support, so that we can have the mapping
> >> between gpio and irq number available without using virtual_irq_start
> >> and MXS_GPIO_IRQ_START.
> > 
> > Seems like a good thing to do, but with Grant and Rob being better at
> > irqdomain than me I'd prefer one of them to have a look at this too.
> > 
> >> +       irq_base = irq_alloc_descs(-1, 0, 32, numa_node_id());
> >> +       if (irq_base < 0)
> >> +               return irq_base;
> >> +
> >> +       port->domain = irq_domain_add_legacy(np, 32, irq_base, 0,
> >> +                                            &irq_domain_simple_ops, NULL);
> > 
> > For example I want to understand if it'd be possible to use a
> > linear domain here.
> 
> Unfortunately, not if you care about your irq base and need sequential
> irq numbers which is the case for non-DT.
> 
Right.  For for this particular case, it's more because generic-chip
currently does not cope with the driver using linear domain.

-- 
Regards,
Shawn