[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <YzRfAE58ZMI/MC0K@smile.fi.intel.com>
Date: Wed, 28 Sep 2022 17:49:36 +0300
From: Andy Shevchenko <andriy.shevchenko@...ux.intel.com>
To: Thierry Reding <thierry.reding@...il.com>
Cc: Uwe Kleine-König
<u.kleine-koenig@...gutronix.de>, linux-pwm@...r.kernel.org,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH v1 1/1] pwm: core: Replace custom implementation of
device_match_fwnode()
On Wed, Sep 28, 2022 at 04:26:47PM +0200, Thierry Reding wrote:
> On Tue, Sep 27, 2022 at 08:22:58PM +0300, Andy Shevchenko wrote:
> > Replace custom implementation of the device_match_fwnode().
> I really don't see the point in having an exported symbol for this. It's
> a simple comparison and the result is even longer than the original.
Longer doesn't always mean worse.
> The
> *only* reason why this helper exists seems to be because it is getting
> used in *_find_device() callbacks.
Yes and no. Initially for the purpose to be a callback it can be reused.
The point is that it hides the dev_fwnode() machinery behind and taking
into account ongoing discussion about constification of the dev_fwnode()
we might need to touch this or similar places to avoid problems with
compiler.
> Honestly, I don't see a reason why this should be applied. And frankly,
> why bother making all these changes. It's a waste of time, in my
> opinion.
Obviously I will not do it if I be with you on the same page.
But okay, not a big deal in this case.
--
With Best Regards,
Andy Shevchenko
Powered by blists - more mailing lists