linux-kernel - Re: [PATCH] rtc: Do not sync CMOS clock when CONFIG_RTC

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]

Date:   Mon, 27 Apr 2020 17:36:11 +0300
From:   Leonid Bloch <lb.workbox@...il.com>
To:     Thomas Gleixner <tglx@...utronix.de>
Cc:     linux-kernel@...r.kernel.org, John Stultz <john.stultz@...aro.org>,
        Stephen Boyd <sboyd@...nel.org>
Subject: Re: [PATCH] rtc: Do not sync CMOS clock when CONFIG_RTC_SYSTOHC is
 not set

On Mon, Apr 27, 2020 at 12:30 PM Thomas Gleixner <tglx@...utronix.de> wrote:
>
> Leonid Bloch <lb.workbox@...il.com> writes:
> > According to documentation in 'drivers/rtc/Kconfig', if
> > 'CONFIG_RTC_SYSTOHC' is set, then:
> >
> > '''
> > The system time (wall clock) will be stored in the RTC specified by
> > RTC_HCTOSYS_DEVICE approximately every 11 minutes if userspace reports
> > synchronized NTP status.
> > '''
> >
> > However in reality, even if 'CONFIG_RTC_SYSTOHC' is not set, the RTC
> > is still sometimes synced with the system time: at least when the RTC
> > driver is 'rtc_cmos', in certain situations. This commit prevents
> > that.
>
> But in reality sync_cmos_clock() depends on CONFIG_GENERIC_CMOS_UPDATE
> and has nothing to do with CONFIG_RTC_SYSTOHC.
>
> That's a historical leftover from the days where RTCs were strictly a
> platform/architecture specific issue.
>
> Your change would break all architectures which still depend on that.

Thanks for your clarification, Thomas!

Then maybe it's worth to reword the help string in
drivers/rtc/Kconfig? Because one can unset 'CONFIG_RTC_SYSTOHC' and
expect that the RTC clock will not be updated by NTP syncs, while in
reality it will be, unless 'CONFIG_GENERIC_CMOS_UPDATE' is unset, and
it is selected automatically by X86.

Regards,
Leonid.
___

>
>
> Thanks,
>
>         tglx