Hi Linus! Thank you very much for your feedback - and sorry about the delay in getting back to you. I will send you a v2 updated version addressing you comments. Thank you! See below for comments ---Lars > From: Linus Walleij <linus.walleij@xxxxxxxxxx> > Date: Fri, 7 Dec 2018 10:43:38 +0100 > Message-ID: <CACRpkdZ47tELuJR3DG- > d8tsvz4ez_j6TxCYiiGT3L01CJ1LSTA@xxxxxxxxxxxxxx> > Subject: Re: [PATCH] mips: gpio: mscc_sgpio: Add DT bindings > documentation > To: lars.povlsen@xxxxxxxxxxxxx > Cc: "open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS" > <devicetree@xxxxxxxxxxxxxxx>, > Gregory Clement <gregory.clement@xxxxxxxxxxx> > > Hi Lars, > > thanks for your patch! > > On Mon, Nov 26, 2018 at 1:22 PM Lars Povlsen > <lars.povlsen@xxxxxxxxxxxxx> wrote: > > > +MSCC Serial GPIO driver > > Expand the acronym please, if IIUC this is > "Microsemi Corporation (MSCC) Serial GPIO driver" Will do. > > > +- mscc,sgpio-frequency: The frequency at which the serial bitstream > is > > + generated and sampled. Default: 12500000 (Hz). > > Uhm what is a serial bitstream in this context, I am lost. > Please elaborate. I added some background details to the description. > > > +- mscc,sgpio-bitcount: The number of bits enabled per > > + port. (1-4). Default: 2. > > Can't you just use the standard "ngpios" property for this? > See > Documentation/devicetree/bindings/gpio/gpio.txt > I dropped this property in favor of deriving from gpio-ranges and (optionally) ngpios as proposed. > Why is this a global setting for all ports and not per-port? > > > +- mscc,sgpio-ports: A bitmask (32 bits) of which ports are enabled in > > + the serialized gpio stream. Default: 0xFFFFFFFF. > > I don't understand this either. Are there up to 32 ports? > (Also referred to as "banks") with 1-4 GPIOs each? Yes each port represent 1-4 GPIOs. Changed to: - mscc,sgpio-ports: A bitmask (32 bits) of which ports are enabled in the serialized gpio stream. One 'port' will transport from 1 to 4 gpio bits. Default: 0xFFFFFFFF. Again, thank you very much for your comments - they are highly appreciated. ---Lars > > Yours, > Linus Walleij > -------------------- End of forwarded message -------------------- > > -- > Gregory Clement, Bootlin > Embedded Linux and Kernel engineering > http://bootlin.com
