Re: [PATCH v2] rtc: rzn1: implement one-second accuracy for alarms

[Alexandre Belloni <alexandre.belloni@bootlin.com>]

On 10/03/2025 10:07:05+0100, Wolfram Sang wrote:
> Hi Alexandre,
> 
> On Mon, Mar 10, 2025 at 09:16:58AM +0100, Wolfram Sang wrote:
> > 
> > > Yes, you followed the code correctly, I have a series that is removing
> > > RTC_UF that I didn't send yet.
> > 
> > Please CC me when you send this. I am interested.
> 
> Do you also have a series pending simplifying handling of
> 'max_user_freq'? AFAICS this is totally HW independent now, meaning we
> can just deal with the constant max value in the core and remove messing
> with with it in the drivers. If you don't have such a series, I am
> willing to work on this.
> 

Yes, I have something I worked on a few years ago now. I was wondering
about the proper default policy which is 64 Hz right now and which max
value we should allow, this is 4096Hz but we have one RTC setting 8192.

I don't this it matters too much because there seem to be very few
userspace programs using RTC_IRQP_SET and RTC_PIE_ON, muse, tutka,
twclock, tvtime and mplayer. They are all either very old or this is an
optional feature with a better replacement.

While I'm reviewing all your other series, do you mind having a look at
https://lore.kernel.org/all/20250205173918.600037-1-herve.codina@bootlin.com/
It has been submitted a while ago now.


-- 
Alexandre Belloni, co-owner and COO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com