Hi Rob,

On 11/22/2015 2:31 PM, Rob Herring wrote:

On Fri, Nov 20, 2015 at 05:17:15PM -0700, Joshua Henderson wrote:

quoted

From: Purna Chandra Mandal <redacted>

Document the devicetree bindings for the clock driver found on Microchip
PIC32 class devices.

Signed-off-by: Purna Chandra Mandal <redacted>
Signed-off-by: Joshua Henderson <redacted>
---
 .../devicetree/bindings/clock/microchip,pic32.txt  |  263 ++++++++++++++++++++
 1 file changed, 263 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/clock/microchip,pic32.txt

diff --git a/Documentation/devicetree/bindings/clock/microchip,pic32.txt b/Documentation/devicetree/bindings/clock/microchip,pic32.txt
new file mode 100644
index 0000000..4cef72d
--- /dev/null
+++ b/Documentation/devicetree/bindings/clock/microchip,pic32.txt

@@ -0,0 +1,263 @@
+Binding for a Clock hardware block found on
+certain Microchip PIC32 MCU devices.
+
+Microchip SoC clocks-node consists of few oscillators, PLL, multiplexer
+and few divider nodes.

[...]

quoted

+Required properties:
+- compatible : should have "microchip,pic32-clk".
+- reg : A Base address and length of the register set.
+- interrupts : source of interrupt.
+
+Optional properties (for subnodes):
+- #clock-cells: From common clock binding, should be 0.
+
+- microchip,clock-indices: in multiplexer node clock sources always aren't linear
+    and contiguous. This property helps define clock-sources with respect to
+    the mux clock node.
+
+- microchip,ignore-unused : ignore gate request even if the gated clock is unused.

There is some discussion about this upstream with "critical-clocks" 
binding. Can you use and wait for that?

The way this is going, we might not have to wait. :)  Is there a patch available yet to try it out?  

quoted

+- microchip,status-bit-mask: bitmask for status check. This will be used to confirm
+    particular operation by clock sub-node is completed. It is dependent sub-node.
+- microchip,bit-mask: enable mask, similar to microchip,status-bit-mask.

We've generally decided not to describe clocks at this level of detail 
in DT. It's fine though for simple clock trees. This one seems to be 
borderline IMO.

The binding example is the entire clock tree.  These masks are right from the datasheet.  For reference, do you have an example of a better alternative?

quoted

+- microchip,slew-step: enable frequency slewing(stepping) during rate change;
+    applicable only to sys-clock subnode.

Thanks,
Josh

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html