[PATCH 1/8] devicetree: binding: R-car Gen3 CMT0 and CMT1 bindings

Bui Duc Phuc bd-phuc at jinso.co.jp
Mon Sep 12 08:12:08 PDT 2016


Dear Geert,

> I think the plan was to get rid of the renesas,cmt-48-* bindings, as they do
> not allow to differentiate between CMT0 and CMT1, which have different feature
> sets.
>
> Cfr. Magnus' series "clocksource: sh_cmt: DT binding rework V4"
> (https://lkml.org/lkml/2016/3/14/433).
>
> Magnus: What's the status of your series?
Thank for your comments. I will update in V2.
I am still waiting for Magnus's comments, but from my point of 
view,Magnus's patches seem
support 32 bit counter only, they are not suitable for 48bit counter.
According to hardware manual, there has to be registers ( CMCSRH, 
CMCNTH,CMCORH) set for 48 bit counter,
but I do not see the registers ( CMCSRH, CMCNTH,CMCORH) in Magnus's patches.


-- 
Regards,

Bui Duc Phuc




More information about the linux-arm-kernel mailing list