Re: [PATCH] media: i2c: imx415: Drop redundant runtime PM callbacks

[Elgin Perumbilly]

Hi Sakari,

> Hi Elgin,
>
> On Sat, Mar 28, 2026 at 06:11:43AM +0000, Elgin Perumbilly wrote:
> >Hi Michael, Sakari,
> >Sorry for the late reply, and thanks for the feedback.
> >
> >>Hi Elgin, Sakari,
> >>
> >>On 3/25/26 13:28, Sakari Ailus wrote:
> >>> Hi Michael,
> >>>
> >>> On Tue, Mar 24, 2026 at 08:42:31PM +0100, Michael Riesch wrote:
> >>>> Hi Elgin,
> >>>>
> >>>> Thanks for the patch but...
> >>>>
> >>>> On 3/24/26 13:45, Elgin Perumbilly wrote:
> >>>>> Replace runtime_suspend/resume wrappers by using power helpers
> >>>>> directly with DEFINE_RUNTIME_DEV_PM_OPS().
> >>>>
> >>>> ...why? What advantage does this refactoring bring?
> >>>
> >>> It looks like patch removes two redundant functions, doesn't it? :-)
> >>
> >>Well yes. 10 lines saved. Not sure whether this was the biggest issue
> >>with this driver -- surely it was not a big issue for me.
> >>
> >>Elgin, I was wondering whether this change is within a certain context
> >>or whether there is something I fail to recognize.
> >>
> >>Anyway, if you absolutely want this change: I feel it would be nicer if
> >>you left the imx415_runtime_{resume,suspend} and moved the bit of code
> >>from imx415_power_{on,off} to them. Reasons for that may be mostly
> >>aesthetical -- I claim better readability that way.
> >
> >The intent of this change is to remove redundant runtime PM
> >wrappers and simplify the code, without altering behavior.
> >
> >If you’re okay with this approach, we can proceed as is. If you
> >feel it’s not necessary at this point, that’s fine as well.
>
> Can you address Michael's comments?
>
> Code tends to be copied to other drivers so cleaning up existing drivers is
> worthwhile.

As far as I understand, would it make sense to follow an approach
similar to t4ka3.c?

i.e. keep runtime_suspend() / runtime_resume() and move the power
on/off logic there. This would address Michael’s readability concern
while still achieving the cleanup.

Regards,
Elgin