On Fri, Apr 27, 2018 at 11:20 AM, Masahiro Yamada <yamada.masahiro@xxxxxxxxxxxxx> wrote: > Hi Rob, > > > 2018-04-26 0:21 GMT+09:00 Rob Herring <robh@xxxxxxxxxx>: > >>> diff --git a/Documentation/devicetree/bindings/usb/dwc3.txt b/Documentation/devicetree/bindings/usb/dwc3.txt >>> index 0dbd308..feb1cc33 100644 >>> --- a/Documentation/devicetree/bindings/usb/dwc3.txt >>> +++ b/Documentation/devicetree/bindings/usb/dwc3.txt >>> @@ -7,6 +7,27 @@ Required properties: >>> - compatible: must be "snps,dwc3" >>> - reg : Address and length of the register set for the device >>> - interrupts: Interrupts used by the dwc3 controller. >>> + - clock-names: should contain "ref", "bus_early", "suspend" >>> + - clocks: list of phandle and clock specifier pairs corresponding to >>> + entries in the clock-names property. >>> + - resets: a single pair of phandle and reset specifier >> >> This should be optional as some SoCs don't have separate, s/w controlled >> resets of modules. > > OK. I will move resets to optional property. > > > Please let ask a question. > > > The number of clocks should be the same across SoCs. > (Even if there is no s/w control for clocks, > we should input something such as clk-fixed-rate.) I guess if there's really not s/w control, then the number may vary, but let's worry about that when someone has that problem. > On the other hand, the number of resets can be different > across SoCs. If there is no s/w control for resets, > we can make it optional. (optional = 1 or 0 reset) > > Is this what you mean? Yes. I think it's just more common to not have a reset than not have clocks. > If we had something like reset-nop (or reset-dummy) > in case no s/w control, we would be able to input something. > I am not sure if this is the right thing, though. I don't think we should require things if they are not needed. Rob -- 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
