[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <YLemjDc1hhuXUojl@smile.fi.intel.com>
Date: Wed, 2 Jun 2021 18:41:00 +0300
From: Andy Shevchenko <andriy.shevchenko@...ux.intel.com>
To: Serge Semin <fancer.lancer@...il.com>
Cc: Serge Semin <Sergey.Semin@...kalelectronics.ru>,
linux-gpio@...r.kernel.org, linux-kernel@...r.kernel.org,
Hoan Tran <hoan@...amperecomputing.com>,
Linus Walleij <linus.walleij@...aro.org>,
Bartosz Golaszewski <bgolaszewski@...libre.com>
Subject: Re: [PATCH v1 2/2] gpio: dwapb: Switch to use fwnode_irq_get()
On Wed, Jun 02, 2021 at 05:05:46PM +0300, Serge Semin wrote:
> On Tue, Jun 01, 2021 at 07:21:28PM +0300, Andy Shevchenko wrote:
> > We have open coded variant of fwnode_irq_get() in dwapb_get_irq().
> > Replace it with a simple call.
>
> Sometime ago I was trying to figure out a way to simplify this part of
> the driver by using the platform_get_irq_optional() method for both
> ACPI and OF cases. As you must have already found out by yourself it
> didn't work out because of DW APB GPIO DT-nodes are supposed to have
> sub-nodes with ports description. The OF-descriptors of these
> sub-nodes aren't touched by the platform_get_irq_optional() method, it
> just fails to detect IRQ-controller because it only works with the
> device OF-node. So I gave up and decided to leave the code as is. I
> can't remember now why I haven't used fwnode_irq_get() here. Most
> likely I just preferred a direct of_irq_get() invocation here just for
> clarity, since the only way we'd be calling fwnode_irq_get() here is
> to actually get IRQ number from the OF-node anyway, while the
> acpi_irq_get() method call made from the method fwnode_irq_get() will
> hardly ever be required here. If you think otherwise or I missing
> something please tell me.
>
> Anyway by applying your patch at least we'll save a few lines of the
> code and may in future have swnode-base IRQs support in the
> fwnode_irq_get() method. So it still worths merging in. Thanks for
> suggesting this change.
>
> Acked-by: Serge Semin <fancer.lancer@...il.com>
> Tested-by: Serge Semin <fancer.lancer@...il.com>
Thanks for testing!
--
With Best Regards,
Andy Shevchenko
Powered by blists - more mailing lists