Re: [PATCH v4 00/13] Add ACPI _DSD and unified device properties support

From: Mark Rutland <mark.rutland@arm.com>
To: Darren Hart <dvhart@linux.intel.com>
Cc: David Woodhouse <dwmw2@infradead.org>,
	"Rafael J. Wysocki" <rjw@rjwysocki.net>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	Mika Westerberg <mika.westerberg@linux.intel.com>,
	ACPI Devel Maling List <linux-acpi@vger.kernel.org>,
	Aaron Lu <aaron.lu@intel.com>,
	"devicetree@vger.kernel.org" <devicetree@vger.kernel.org>,
	Linus Walleij <linus.walleij@linaro.org>,
	Alexandre Courbot <gnurou@gmail.com>,
	Dmitry Torokhov <dmitry.torokhov@gmail.com>,
	Bryan Wu <cooloney@gmail.com>,
	"grant.likely@linaro.org" <grant.likely@linaro.org>,
	Arnd Bergmann <arnd@arndb.de>,
	"dvhart@infradead.org" <dvhart@infradead.org>
Subject: Re: [PATCH v4 00/13] Add ACPI _DSD and unified device properties support
Date: Wed, 15 Oct 2014 16:17:02 +0100	[thread overview]
Message-ID: <20141015151702.GG20034@leverpostej> (raw)
In-Reply-To: <543E88CF.5060504@linux.intel.com>

On Wed, Oct 15, 2014 at 03:46:39PM +0100, Darren Hart wrote:
> 
> 
> On 10/15/14 16:08, David Woodhouse wrote:
> > 
> >> We have been checking for all DT platforms, and that's a bug for DT.
> >> Copying that bug to ACPI is inexcusable given we know it's a bug to do
> >> so.
> > 
> > We'll, perhaps it should be named 'used-by-firmware' and actually it's
> > just as valid under ACPI as it is on RTAS systems. All it does is stop the
> > OS from using the port.
> > 
> >> I understand that. However, where a binding doesn't make sense (as in
> >> this case), it shouldn't be enabled for ACPI as it provides a larger
> >> surface area for misuse, for no benefit.
> > 
> > These are *optional* properties. They were optional precisely *because*
> > they only make sense in some cases. I don't know that it makes sense to
> > take them away. The benefit we get is *consistency*. For example if
> > someone *does* use the property in question as 'used-by-firmware' and
> > expects the OS not to touch it, we don't want that to change behaviour
> > between ACPI and fdt boots.
> 
> My comment was going to be along the same lines. It is an optional
> parameter, which is what I would expect for a firmware-specific type of
> property.
> 
> I also don't agree that this is "copying that bug to ACPI". This line of
> code has no impact to ACPI. No ACPI implementation should add this,
> certainly not if it was actually tested as it would not run if it was
> present in the _DSD. So... what's the problem exactly? Or perhaps more
> specifically:
> 
> Mark, what would you propose we do differently to enable this driver to
> be firmware-type agnostic?

For this particular driver, all I'm asking for is that the
"used-by-rtas" property is not moved over from of_find_property to
device_get_property. It is irrelevant for all ACPI systems. Evidently my
comment was unclear; I apologise for that.

We have status = "disabled" as a less specific mechanism for telling the
OS to ignore a node in DT. I was under the impression that ACPI already
had a mechanism for marking devices to be ignored, but perhaps I am
mistaken.

The concerns I mentioned at the end of my original reply were of a more
general nature than this particular device description.

Thanks,
Mark.