On Fri, Aug 29, 2014 at 1:58 AM, Laurent Pinchart <[email protected]> wrote: [...] > We could thus extend the DIV6 clocks with clock source selection using the > following layout > > Bits Name Function > ----------------------------------------------------------------------------- > 8 CKSTP Clock Stop > 7-6 EXSRC Source Selection > 5-0 DIV Division Ratio > > The clocks property would be a list of one, two, three or four clock > references, mapping to the 00, 01, 10 and 11 values of the EXSRC field. Empty > references (just a 0 value) would be used to denote invalid values of the > EXSRC field. > > A new compatible string would then be added for the VCLK clocks with the > following layout > > Bits Name Function > ----------------------------------------------------------------------------- > 14-12 EXSRC Source Selection (VCLK[123] clocks) > 8 CKSTP Clock Stop > 7-6 PDIV PLL Division Ratio (VCLK3 clock) > 5-0 DIV Division Ratio
I think even a new compatible string would not be necessary, given that all such clocks have more than four parents, which cannot be the case for clocks with two EXSRC bits. > The clocks property would be handled as for the common DIV6 clocks, and we > would need to define a way to describe the valid values of the PDIV field. As > far as I can see, the PDIV values are identical for all VCLK clocks when > implemented (00 -> 1/1, 01 -> 1/2, 10 -> Reserved, 11 -> 1/4), so a single > renesas,has-pdiv property should be enough. Sounds reasonable to me. I'll send a patch. CU Uli -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
