Message-Id: <201207242134.00769.rjw@sisk.pl>
Date:	Tue, 24 Jul 2012 21:34:00 +0200
From:	"Rafael J. Wysocki" <rjw@...k.pl>
To:	Arnd Bergmann <arnd@...db.de>
Cc:	Linux PM list <linux-pm@...r.kernel.org>,
	Mark Brown <broonie@...nsource.wolfsonmicro.com>,
	LKML <linux-kernel@...r.kernel.org>,
	Matthew Garrett <mjg59@...f.ucam.org>,
	Magnus Damm <magnus.damm@...il.com>,
	Grant Likely <grant.likely@...retlab.ca>,
	"Linux-sh list" <linux-sh@...r.kernel.org>
Subject: Re: [RFC][PATCH 0/14] PM / shmobile: Pass power domain information via DT (was: Re: [RFD] PM: Device tree representation of power domains)

On Tuesday, July 24, 2012, Arnd Bergmann wrote:
> On Saturday 21 July 2012, Rafael J. Wysocki wrote:
> > On Monday, July 16, 2012, Rafael J. Wysocki wrote:
> > > On Thursday, July 05, 2012, Rafael J. Wysocki wrote:
> > > > On Wednesday, July 04, 2012, Mark Brown wrote:
> > > > > I guess the OMAP hwmod stuff is the closest thing we've got at the
> > > > > minute (I don't recall seeing any other implementations in mainline) but
> > > > > the hwmods themselves don't appear in the DTS right now.  They have a
> > > > > ti,hwmods property on each device naming the hwmod it's in, something
> > > > > like that seems like a reasonable approach, possibly a reference to
> > > > > another DT node rather than or as well as a string?  That seems fairly
> > > > > easy.
> > > > 
> > > > Well, it looks like (and please tell me if I'm wrong) the hwmons are just
> > > > string attributes that are parsed by the platform-specific code through
> > > > a platform bus type notifier.
> > > > 
> > > > We could do that for power domains too, but then each platform wanting to
> > > > use them would need to implement such a notifier and add its own routine
> > > > for parsing those strings.  Would that be acceptable to everyone concerned?
> > > 
> > > I tried to follow the above suggestion and prepared the following patchset
> > > that allows power domain information for Renesas platforms to be passed as
> > > "renesas,pmdomain" string attribute of device nodes.  It adds functions
> > > allowing the generic PM domains framework to use names for domain
> > > identification in various situations and reworks the ARM/shmobile power domains
> > > support code to used those functions instead of the "raw" ones that take
> > > domain pointers as their arguments.  Finally, it defines a platform bus type
> > > notifier that will add devices whose DT nodes contain the "renesas,pmdomain"
> > > attribute to the power domains indicated by it (the value of that attribute
> > > should be the name of the PM domain to add the device to after it's been
> > > registered).  All of this should allow platform devices to be added to
> > > appropriate power domains automatically based on the information read from
> > > a DT.
> > > 
> > > The patches are on top of the current linux-next tree.
> > > 
> > > I've tested the patches that could be tested on the Mackerel board, except
> > > for the last one (I'm still working on testing it).
> > 
> > Well, no comments, no objections.  Good!
> > 
> > I've just tested [14/14] too and it works as expected.
> 
> Sorry for taking so long to reply. I am really not that familiar with the
> power domain requirements, but I do have two comments on your approach:
> 
> * I think when we want to add a generic concept to the device tree such
>   as power domains, we should always make it specified in a generic way.

Do we really want that?  I'm a bit skeptical, because apparently nobody
cares, as the (zero) response to this patchset evidently indicates and
since nobody cares, it's probably better not to add such "generic" things
just yet.

>   You have used the "renesas,pmdomain" attribute, which is specific to
>   one vendor and requires platform specific code to read it.
>   Is there any reason not to put the code into the generic
>   drivers/base/power/domain.c file (or near it) and make a binding that
>   works for everyone?

Yes, there is.  A couple of them in fact.

First off, power domains will always need platform-specific code to support
them, no matter what.  The problem is that they tend to require special
handling, even within the same SoC, let alone different SoCs, and that handling
has to be implemented in an SoC-specific way, because it has to know various
things about the platform (like what to write to what register(s) at what time
and so on).  Of course, that platform-specific code needs to know which
domain the given description corresponds to and there doesn't seem to be any
useful way to specify that through DTs.

Second, the generic code needed to support such a binding would have to be
quite complex and I don't see any advantage from adding it.

Finally, even if I did that, I seriously doubt it would really work for
everyone.  Unfortunately, I have no idea what _everyone's_ needs are,
I only know what the needs of Renesas platforms are, so I went for an
approach limited to those platforms.

> * Mark suggested two options (string and phandle), and (you guessed it),

To me, the Mark's suggestion was to follow the example of TI hwmods, which
I did.  In fact, the power domains on Renesas platforms are an analogous
concept, so in my opinion handling them in analogy with hwmods makes a lot of
sense.

>   I would much prefer the other one.

However, I wouldn't. :-)

>   IMHO using phandle makes it much easier to do this in a generic way,
>   without having to document every possible string that this can be
>   set to in the binding document. Obviously the phandle requires having
>   a node in the device tree for each domain, but I think having that
>   node is actually an advantage because it lets you describe the
>   hierarchy of the domains there.

I don't quite see how it is better.  I could (and should in fact) represent
that hierarchy as an array of pairs of strings without adding any special
parsing code to the kernel (except for a simple routine walking that table
and calling pm_genpd_add_subdomain_names() for every pair).

The only disadvantage of the current approach I can see is that whoever
creates a dts for a board based on the given SoC has to know the names of the
power domains to use in there.  I don't think this is an overly burdensome
requirement.

The other way around, the platform code supposed to handle the power domains
would need a way to match DT nodes against the domains, which also would
require some string-based identification and that would have to be documented
as well.

So no, I don't think using phandles would buy us anything, as of today at least.

Thanks,
Rafael
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Powered by blists - more mailing lists