Hi Geert,

On Thursday, February 16, 2017, Geert Uytterhoeven wrote:

On Wed, Feb 15, 2017 at 6:33 PM, Chris Brandt [off-list ref]
wrote:

quoted

On Tuesday, February 14, 2017, Geert Uytterhoeven wrote:

quoted

quoted

+Required properties:
+  - compatible: must be one or more of the following:
+    - "renesas,r7s72100-reset" for the r7s72100 SoC
+    - "renesas,wdt-reset"
+                This is a fallback for the above renesas,*-reset
+entries

Please use "renesas,r7s72100-wdt". DT describes the hardware
(watchdog), not the software policy (use it for reset only, as a max.
timeout of 125 ms is too short for a usable watchdog).

I had a look at:

  Documentation/devicetree/bindings/watchdog/renesas-wdt.txt

Which has:

Required properties:
- compatible : Should be "renesas,<soctype>-wdt", and
               "renesas,rcar-gen3-wdt" as fallback.
               Examples with soctypes are:
                 - "renesas,r8a7795-wdt" (R-Car H3)
                 - "renesas,r8a7796-wdt" (R-Car M3-W)


So in my 'renesas-reset.txt' should I do:

As "reset" is software policy, perhaps you should instead extend the
existing binding document
Documentation/devicetree/bindings/watchdog/renesas-wdt.txt?

Nothing says you have to use the same Linux driver for R-Car Gen3 and
RZ/A1.

That is what I was thinking as well. If I'm just supposed to document hardware,
I might as well group it in the same file...but then in SW use it for something
completely different (reset instead of its WDT functionality)

quoted

A. "renesas,r7s72100-wdt", "renesas,rz-wdt"

Please don't use plain "rz", as RZ/A, RZ/G, and RZ/T are completely
different beasts.

Ya, I'm stuck on a name here. Internally, they don't really have defined IP block
name for the "8-bit WDT that can also be used as a simple timer". I know this
WDT was also used in SH4A devices like SH7757 and SH7758.

quoted

  or just:

B. "renesas,r7s72100-wdt"   (no fallback, change the driver to add new

SoCs)

Do you know if future RZ/A SoCs will use the exact same type of WDT?

Yes, RZ/A2 will for sure. I assume it would also be used for any RZ/A3+ because
it's easy to hook up from a HW perspective (if I remember correctly). Although,
I want them to extend the counter from 8-bit to 16-bit. 125ms is ridiculously
too short!

If yes, you can document both "renesas,r7s72100-wdt" and "renesas,rza-wdt",
and let the driver match against the latter.

I'm going to go with that.
If the timeout does ever get extended and it does become useful to be used
as an actual watchdog, I'll create a new driver or something.


As always, thank you for your replies!


Chris