On Fri, 3 May 2019 at 03:44, Florian Fainelli <f.fainelli@xxxxxxxxx> wrote: > > > > On 5/2/2019 5:26 AM, Jonas Gorski wrote: > > Add binding documentation for the gated clock controller found on MIPS > > based BCM63XX SoCs. > > > > Signed-off-by: Jonas Gorski <jonas.gorski@xxxxxxxxx> > > Reviewed-by: Florian Fainelli <f.fainelli@xxxxxxxxx> Thanks! > > > --- > > .../bindings/clock/brcm,bcm63xx-clocks.txt | 22 ++++++++++++++++++++++ > > 1 file changed, 22 insertions(+) > > create mode 100644 Documentation/devicetree/bindings/clock/brcm,bcm63xx-clocks.txt > > > > diff --git a/Documentation/devicetree/bindings/clock/brcm,bcm63xx-clocks.txt b/Documentation/devicetree/bindings/clock/brcm,bcm63xx-clocks.txt > > new file mode 100644 > > index 000000000000..3041657e2f96 > > --- /dev/null > > +++ b/Documentation/devicetree/bindings/clock/brcm,bcm63xx-clocks.txt > > @@ -0,0 +1,22 @@ > > +Gated Clock Controller Bindings for MIPS based BCM63XX SoCs > > + > > +Required properties: > > +- compatible: must be one of: > > + "brcm,bcm3368-clocks" > > + "brcm,bcm6328-clocks" > > + "brcm,bcm6358-clocks" > > + "brcm,bcm6362-clocks" > > + "brcm,bcm6368-clocks" > > + "brcm,bcm63268-clocks" > > We could always add 6348/6338 to that list later one. That's the plan*. But currently neither one is supported by BMIPS_GENERIC, which was my starting point. And making BCM63XX use the driver is ... complicated, due to (important) consumers not being platform drivers (the MPI/PCIe controller codes). And they can't be just converted to platform drivers, as they should then be non-legacy PCI controller drivers, so a full rewrite is needed. So let's stick with BMIPS_GENERIC first, else we never get anything done ;-). In the end BCM63XX should go away anyway, once BMIPS_GENERIC has reached feature parity. Regards Jonas * if it even makes sense, as these are quite old, and systems will quite struggle running a modern kernel/OS. Jonas
