[PATCH V2] clk: scmi: add is_prepared hook

Dhruva Gole d-gole at ti.com
Thu Aug 1 23:15:26 PDT 2024


Hi Peng,

On Aug 01, 2024 at 03:35:37 +0000, Peng Fan wrote:
> Hi Sudeep, Dhruva,
> 
> > Subject: Re: [PATCH V2] clk: scmi: add is_prepared hook
> > 
> > On Fri, Jul 26, 2024 at 07:22:31PM +0530, Dhruva Gole wrote:
> > > On Jul 26, 2024 at 21:10:07 +0800, Peng Fan (OSS) wrote:
> > > > From: Peng Fan <peng.fan at nxp.com>
> > > >
> > > > Some clks maybe default enabled by hardware, so add is_prepared
> > hook
> > > > for non-atomic clk_ops to get the status of the clk. Then when
> > > > disabling unused clks, those unused clks but default hardware on
> > > > clks could be in off state to save power.
> > >
> > > Just a nit - reword the commit message as:
> > > Then when disabling the unused clocks, they can be simply turned
> > OFF
> > > to save power.
> > >
> > 
> > Ah this was what it meant. I couldn't parse the original text and was
> > about to ask.
> > 
> > > Also if you can make it still verbose, explain when you expect this
> > > disabling of unused clks to take place exactly? During boot?  Driver
> > probe sequence?
> > > or By some user commands?
> > >
> > 
> > Agreed. Being little more verbose here would be beneficial IMO.
> > 
> 
> I will use below in V3:
> "
> Some clocks maybe default enabled by hardwar. For clocks that not 

s/not/don't

> have users, they will be left in hardware default state, because prepare

s/they/that, will be left in default hardware state.

> count and enable count is zero,if there is no is_prepared hook to get
> the hardware state. So add is_prepared hook to detect the hardware
> state. Then when disabling the unused clocks, they can be simply
> turned OFF to save power during kernel boot.
> "

Thanks, rest looks better.

-- 
Best regards,
Dhruva Gole <d-gole at ti.com>



More information about the linux-arm-kernel mailing list