linux-kernel - Re: [PATCH] ntp: Make sure RTC is synchronized for any time jump

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <gyumkaygfe3jow65bpzm2bgj2tgipbxuj7xa4oxwngkq6xfvyh@ojhlg4oaaaii>
Date: Sun, 27 Oct 2024 18:42:03 +0100
From: Benjamin ROBIN <dev@...jarobin.fr>
To: Thomas Gleixner <tglx@...utronix.de>
Cc: jstultz@...gle.com, sboyd@...nel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] ntp: Make sure RTC is synchronized for any time jump

On Wed, Oct 16, 2024 at 12:28:31AM +0200, Thomas Gleixner wrote:

Sorry for replying with such delay, I was not really available...

> > @@ -2610,7 +2611,7 @@ int do_adjtimex(struct __kernel_timex *txc)
> >  	if (clock_set)
> >  		clock_was_set(CLOCK_SET_WALL);
> >  
> > -	ntp_notify_cmos_timer(offset_set);
> > +	ntp_notify_cmos_timer(false);
> 
> Can you please add a comment why this does not need to cancel the
> hrtimer? It took me a few looks to validate that this is correct. Six
> months down the road we both forgot about it. :)

I tried to add a comment that explain it in the v2.
I hope this is "good enough". From my understanding of why this is not
needed to cancel NTP sync hrtimer here: Because NTP sync flag was not cleared
since (either):
 - The frequency was adjusted, which has no impact on NTP sync hrtimer.
 - TAI offset was adjusted, and it should not have any impact on REALTIME.
 - Leap second was handled, which does not require to restart NTP sync hrtimer.

> Also can you please rebase this on top of
> 
>    git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git timers/core
> 
> which has a larger change of the NTP code.

I did rebase on top of this branch.
I will send the v2 patch in a separate email.

Thanks, Benjamin