Re: [alsa-devel] Fwd: [PATCH 29/46] Input: touchscreen: use wrapper for pxa2xx ac97 registers

From: Dmitry Torokhov <dmitry.torokhov@gmail.com>
To: Arnd Bergmann <arnd@arndb.de>
Cc: ALSA Development Mailing List <alsa-devel@alsa-project.org>,
	Linus Walleij <linus.walleij@linaro.org>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	Haojian Zhuang <haojian.zhuang@gmail.com>,
	Daniel Mack <daniel@zonque.org>,
	"open list:HID CORE LAYER" <linux-input@vger.kernel.org>,
	Robert Jarzmik <robert.jarzmik@free.fr>,
	Linux ARM <linux-arm-kernel@lists.infradead.org>
Subject: Re: [alsa-devel] Fwd: [PATCH 29/46] Input: touchscreen: use wrapper for pxa2xx ac97 registers
Date: Fri, 18 Oct 2019 13:39:02 -0700	[thread overview]
Message-ID: <20191018203902.GQ35946@dtor-ws> (raw)
In-Reply-To: <CAK8P3a0AP8q9Hk1DB+E=xFGV5RqXnkjkcZRobgU1NRDFMH00AA@mail.gmail.com>

On Fri, Oct 18, 2019 at 09:39:31PM +0200, Arnd Bergmann wrote:
> On Fri, Oct 18, 2019 at 8:48 PM Dmitry Torokhov
> <dmitry.torokhov@gmail.com> wrote:
> >
> > On Fri, Oct 18, 2019 at 05:41:44PM +0200, Arnd Bergmann wrote:
> > > To avoid a dependency on the pxa platform header files with
> > > hardcoded registers, change the driver to call a wrapper
> > > in the pxa2xx-ac97-lib that encapsulates all the other
> > > ac97 stuff.
> >
> > Not supper happy about adding module dependencies. Can we include
> > mach/regs-ac97.h from include/sound/pxa2xx-lib.h and use static inlines?
> > Someone needs to include mach/regs-ac97.h in the end...
> >
> > Or there is something later in the series that needs it?
> 
> One of the goals of the series is to completely remove all mach/*.h headers
> and place them somewhere else, ideally inaccessible to device drivers.
> 
> In case of mach/regs-ac97.h, the later patch "ASoC: pxa: ac97: use normal
> MMIO accessors" passes the physical register base address as a platform
> device resource that gets ioremapped in the ac97 driver, rather than
> hardcoding the virtual address in a global header.
> 
> I agree that the exported function is still ugly, but I hope it's enough of
> an improvement over the previous state that we can do it anyway.
> 
> If you have any better ideas though, I can give that a try, too.
> One possibility would be a higher-level interface exported on top
> of 'struct snd_ac97', but I could not figure out how to do this.

No, I do mot really have better ideas given your stated goals and I
guess there is not really much benefit on spending too much effort
polishing essentially one driver.

Acked-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

Thanks.

-- 
Dmitry
_______________________________________________
Alsa-devel mailing list
Alsa-devel@alsa-project.org
https://mailman.alsa-project.org/mailman/listinfo/alsa-devel