Re: [PATCH] rtc: class: Do not inject sleep time without a suspend snapshot

[sashiko-bot@kernel.org]

> rtc_suspend() returns early when timekeeping_rtc_skipsuspend() is true,
> which is the case on every system with a persistent clock. old_rtc and
> old_system are then never written and stay zero.
> 
> rtc_resume() is skipped under a different condition,
> timekeeping_rtc_skipresume(), which is only true once
> timekeeping_resume() has injected sleep time. timekeeping_resume() does
> that if a nonstop clocksource reports elapsed time or if the persistent
> clock is ahead of timekeeping_suspend_time. If neither holds, it injects
> nothing and rtc_resume() runs. With old_rtc and old_system being zero,
> rtc_resume() computes
> 
> 	sleep_time = new_rtc - new_system
> 
> and injects it when it is not negative. For an RTC kept in UTC this is
> [ ... ]
> 
> Fixes: 0fa88cb4b82b ("time, drivers/rtc: Don't bother with rtc_resume() for the nonstop clocksource")
> Link: https://bugzilla.redhat.com/show_bug.cgi?id=2543517
> Cc: stable@vger.kernel.org
> Signed-off-by: Alperen Kılıç <sabri.alperen03@gmail.com>

Sashiko has reviewed this patch and found no issues. It looks great!

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261003163138.14221-1-sabri.alperen03@gmail.com?part=1