Re: [PATCH net-next v4 07/15] net: ravb: Move reference clock enable/disable on runtime PM APIs

[Sergey Shtylyov]

On 1/29/24 4:53 PM, claudiu beznea wrote:

[...]

>>> From: Claudiu Beznea <claudiu.beznea.uj@xxxxxxxxxxxxxx>
>>>
>>> Reference clock could be or not part of the power domain. If it is part of
>>
>>    Could be or not be, perhaps?
>>
>>> the power domain, the power domain takes care of propertly setting it. In
>>
>>    Properly. :-)
>>
>>> case it is not part of the power domain and full runtime PM support is
>>> available in driver the clock will not be propertly disabled/enabled at
>>> runtime. For this, keep the prepare/unprepare operations in the driver's
>>> probe()/remove() functions and move the enable/disable in runtime PM
>>> functions.
>>>
>>> Along with it, the other clock request operations were moved close to
>>> reference clock request and prepare to have all the clock requests
>>> specific code grouped together.
>>>
>>> Signed-off-by: Claudiu Beznea <claudiu.beznea.uj@xxxxxxxxxxxxxx>
>> [...]
>>
>>> diff --git a/drivers/net/ethernet/renesas/ravb_main.c b/drivers/net/ethernet/renesas/ravb_main.c
>>> index 9fc0e39e33c2..4673cc2faec0 100644
>>> --- a/drivers/net/ethernet/renesas/ravb_main.c
>>> +++ b/drivers/net/ethernet/renesas/ravb_main.c
>> [...]
>>> @@ -3060,21 +3058,27 @@ static int ravb_resume(struct device *dev)
>>>  	return ret;
>>>  }
>>>  
>>> -static int ravb_runtime_nop(struct device *dev)
>>> +static int ravb_runtime_suspend(struct device *dev)
>>>  {
>>> -	/* Runtime PM callback shared between ->runtime_suspend()
>>> -	 * and ->runtime_resume(). Simply returns success.
>>> -	 *
>>> -	 * This driver re-initializes all registers after
>>> -	 * pm_runtime_get_sync() anyway so there is no need
>>> -	 * to save and restore registers here.
>>> -	 */
>>
>>    I want to pull out the dummy {ravb|sh_eth}_runtime_nop() funcs --
>> they don't seem to be necessary... Then we can implement your clock

   The need to have the dummy RPM suspend/resume methods is gone since:

https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=63d00be69348fda431ae59aba6af268a5cf5058e

>> dance with freshly added ravb_runtime_{suspend|resume}()...
> 
> For this series, does it worth having a patch that removes ravb runtime
> suspend/resume ops to then add a new patch that add it it again?

    Probably not, indeed... I just wanted to have 2 symmetric patches
for sh_eth and ravb removing the dummy methods...

> I can do it but it I see no reason in doing it in this series...
> 
> The dummy functions were there and the commit description explains the
> reason they were updated.

   Yet you don't say a word about the big comment in ravb_runtime_nop()
that you remove. This comment doesn't really make much sense as this
driver currently has the RPM calls and ndo_{open|stop}() methods decoupled...
This stuff was copied from sh_eth.c verbatim -- I clearly overlooked it when
prepping this driver for upstream... :-<
   You can keep this patch as is (but not its description!) or have a separate
patch that removes just the big comment not making much sense, both options
would be fine by me. I will take care of sh_eth.c myself (not really sure
whether you have targets having this IP)...

> Thank you,
> Claudiu Beznea

MBR, Sergey