On Thu, Apr 07, 2016 at 12:57:36PM -0500, Rob Herring wrote: > On Mon, Apr 04, 2016 at 02:08:10PM -0700, Sreedhar Sambangi wrote: > > This adds the regulator nodes to DK04 device tree to support > > > > Change-Id: I9c1df0e720a330bf6db1889fd2247f6a70ea6faa > > Signed-off-by: Sreedhar Sambangi <ssamb...@codeaurora.org> > > --- > > .../bindings/regulator/ipq4019-regulator.txt | 19 > > +++++++++++++++++++ > > arch/arm/boot/dts/qcom-ipq4019-ap.dk04.1.dtsi | 11 +++++++++++ > > 2 files changed, 30 insertions(+) > > create mode 100644 > > Documentation/devicetree/bindings/regulator/ipq4019-regulator.txt > > > > diff --git > > a/Documentation/devicetree/bindings/regulator/ipq4019-regulator.txt > > b/Documentation/devicetree/bindings/regulator/ipq4019-regulator.txt > > new file mode 100644 > > index 0000000..9d934a4 > > --- /dev/null > > +++ b/Documentation/devicetree/bindings/regulator/ipq4019-regulator.txt > > @@ -0,0 +1,19 @@ > > +* Qualcomm Technologies, Inc. IPQ4019 regulators > > + > > +Required properties: > > +- compatible : Must be "regulator-ipq4019". > > Doesn't match your example which correctly has vendor prefix. Also, the > preferred form is <vendor>,<soc>-<block>. > > > +- states : Selection of available voltages and corresponding > > values > > Too generic of a name. > > Is this not fixed by the h/w block?
I'd put this info in the driver itself. I believe it has 4 fixed states. > > > +- reg : Register address for controlling LDO > > +- mask : Mask value for switching voltages > > What else is in this register? If something unrelated, then this is all > wrong to the regulator. You can only define the register address in one > place in the DT. Otherwise, you should know the mask from the compatible > string. I agree. All of this needs to be done inside the code itself. All you need is the register. This is a TCSR register which means that it could have other bits doing other things. I would hope that the hw guys isolated only regulator bits in here. Regards, Andy Gross
