[PATCH 2/2] i2c: Add devtree-aware iic support for PPC4xx

[Grant Likely]

On 10/15/07, Scott Wood <scottwood at freescale.com> wrote:
> Grant Likely wrote:
> > On 10/15/07, Scott Wood <scottwood at freescale.com> wrote:
> >> On Mon, Oct 15, 2007 at 10:57:48AM -0600, Grant Likely wrote:
> >>> Segher is recommending that we use an aliases node as per the open
> >>> firmware example for this.  I think in this case it would look
> >>> something like this (but I'm not the expert):
> >>>
> >>> aliases {
> >>>      IIC0 = "/path/to/bus/iic at 0x2000";
> >>>      IIC1 = "/path/to/bus/iic at 0x2000";
> >>> };
> >> I think this is overly complicated; something like linux,i2c-index in the
> >> i2c adapter node would be simpler.
> >
> > But not perhaps better.  Enumeration is a system-wide thing.  It does
> > make sense to group all the device enumerations in one place.
>
> For purposes of generating a Linux bus number, having to scan all
> aliases for a matching path and turn IIC0/IIC1 into a number is a bit silly.
>
> For associating a device node with a human readable label, I'd prefer a
> "label" property in the device node, rather than doing it backwards with
> aliases.

Here the corresponding problem; having to scan every device node to
make sure you don't assign a number already selected by another node
(in the case where one node is assigned a number and another is not).

At some point you're going to need to reverse map from number to
device.  I'd rather scan a list of properties in aliases instead of
scanning the whole tree looking for all devices of the same type.

>
> > As per your point below; if all the i2c devices are children of the
> > adapter, then yes you are right that the bus number doesn't matter to
> > the user.  But it *does* matter for things like serial and ethernet
> > ports.
>
> And a label property would be great for that. :-)

Not really; if the user needs to renumber devices; you don't want him
fiddling around in the hardware description.  Just like the chosen
node; an aliases describes logical constructs, not physical ones.  I
don't think this is any different from the linux,stdout-path property
in chosen.

>
> >> Though, I don't see what the problem with the original approach is, as long
> >> as the numbers are chosen in the same way when registering i2c clients based
> >> on the children of the adapter node.  There's no concept in the hardware
> >> itself of a bus number.
> >
> > Even if you take this approach, the driver still need to know what bus
> > number to use (as per the i2c infrastructure).  It is not sane for an
> > i2c bus driver to attempt to assign the bus number itself because it
> > could conflict with another arbitrarily chosen bus number from another
> > driver.
>
> Hence why global bus numbers suck. :-)

I'm not going to disagree with you there; but it's unavoidable.

> However, statically chosen numbers should only be done for platform
> devices, not dynamic things such as PCI, so in practice I don't think
> it'd be a problem.

So... in the I2C case, as long as all the i2c devices are in the
device tree and the i2c layer is responsible for actually handing out
the bus numbers; then yes I agree.

Cheers,
g.

-- 
Grant Likely, B.Sc., P.Eng.
Secret Lab Technologies Ltd.
grant.likely at secretlab.ca
(403) 399-0195