Re: [PATCH 6/6] Clocklib: use correct name for 3,6MHz clock

[Dmitry Baryshkov]

Hi,

On Tue, Apr 08, 2008 at 09:07:21PM +0100, Russell King wrote:
> On Tue, Apr 08, 2008 at 11:58:11PM +0400, Dmitry wrote:
> > 2008/4/8, Russell King <rmk+lkml@xxxxxxxxxxxxxxxx>:
> > > So... what is the correct name.  Bear in mind what I said in the previous
> > >  reply this evening - which says that it should be the name used by the
> > >  SA1111.  Look in the data sheet - the pin itself to which the 3.6MHz
> > >  clock is supplised will have a name.  That's the name which should be
> > >  used.
> > 
> > I use the same pin/clock for the tc6393xb driver. And I'm pretty sure
> > the datasheets won't agree on the name of the pin. Which name should I
> > use?
> 
> You missed the fundamental issue about the clock API - the _name_ is
> not the clock name defined by the host.  It's the _device_ clock name.
> 
> So, you shouldn't be using the SA1111 clock name for the tc6393xb driver.
> You should be using its own name.  The platform specific bit of the clock
> API is then supposed to return you the struct clk corresponding with
> that input, by using the platform knowledge that it's connected to GPIO27
> or whatever.

BTW: Who should know that there should be a clock for a device?
E.g. on pxa, 3.6MHz clock is registered in clock.c, various CKEN-based
clocks are registered from pxa*.c. Which file should contain
host_name -> device_name allocation board config?


And BTW2: The sa1111 spec names the clock simply as "CLK". Should we use
this name?


-- 
With best wishes
Dmitry

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