On Mon, Nov 3, 2014 at 6:24 PM, Geert Uytterhoeven <geert@xxxxxxxxxxxxxx> wrote: > On Mon, Nov 3, 2014 at 5:01 PM, Ulrich Hecht > <ulrich.hecht+renesas@xxxxxxxxx> wrote: [...] >> - - clocks: Reference to the parent clock >> + - clocks: Reference to the parent clock(s); if there are multiple parent >> + clocks, one must be specified for each possible parent clock setting >> + in the clock register. Invalid settings must be specified as "<0>". >> + Trailing invalid settings may be omitted. > > Is there a possibility that omitting trailing invalid settings will cause 4 > or less entries for a clock with 8 parents? That would change its class, > and the corresponding src_width and src_shift. I checked that, and all these clocks actually have more than four possible parents. However, ... > So I'm inclined to say the number of parent clocks must be one of 1, 4, or 8, > i.e. no omissions, to remove this ambiguity. ... considering that we don't know what future SoCs may bring, that indeed looks like the better option to me. CU Uli -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
