On 26 November 2021 10:23, Nikita Shubin wrote: > > Thanks for the detailed response; it helped a lot. Having reviewed > > the core code along with your description I know understand what's > > happening here. Basically marking as 'wakeup-source' is simply a > > means to expose the sysfs attribute to user-space. > > > > Yes I think in the commit message you should be clear that there's a > > need to access the sys attribute 'wakealarm' in the RTC core and > > clarify exactly why there is that need. Your commit log should be > > good enough so that if anyone else needs to look at this later they > > completely understand the intention behind the change. > > > > By the way, I assume the functionality you're looking for could also > > have been achieved through using the /dev/rtcX instance for DA9063? > > Thank you for pointing this out, indeed i missed that obvious thing. > > We can also simply set alarm via rtcwake, even if CONFIG_PM is off: > > rtcwake -m no -s 60 > > Now i am not sure we should make changes to da9063-rtc driver - what do > you think ? I think the change is still valid from what I can tell. Just be clear on intent in the commit log of the patch. :)
