Re: [PATCH 1/3] clk: introduce omap_clk_associate

[Tony Lindgren <tony@atomide.com>]

* Felipe Balbi <me@felipebalbi.com> [081014 12:31]:
> On Tue, Oct 14, 2008 at 10:47:44AM -0700, David Brownell wrote:
> > On Tuesday 14 October 2008, Felipe Balbi wrote:
> > > I didn't quite get you here. The idea of mmc_fck is so that
> > > 
> > > clk_get(dev, "mmc_fck");
> > > 
> > > works fine and returns the correct clock. If we have several fck and ick
> > > function names, how will we clk_get() the right one ??
> > 
> > If "dev" is an MMC device, there's no way to confuse
> > its "fck" and "ick" with those for, say, I2C.   Right?
> > That's the whole point of associating logical clock
> > names with the device.
> > 
> > And as Paul noted, if a device has several such clocks,
> > then it needs several such names.
> 
> hmm... that's true. Forgot about matching dev as well :-p
> hehehe. Makes sense to me, let's use fclk and iclk then :-)
> 
> The main idea then would be that clk(dev, "iclk") translates to english
> into "get me the interface clock of mmc device" (when dev is an mmc
> device, of course).

And that we can use same naming in the driver no matter which omap :)

Tony