Re: [RFC PATCH] Add rtc server, rtc-read and rtc-set operations

[Zhaoming Luo]

On 11/19/24 5:08 PM, Samuel Thibault wrote:
> Zhaoming Luo, le mar. 19 nov. 2024 16:54:38 +0800, a ecrit:
> > On 11/17/24 9:04 PM, Sergey Bugaev wrote:
> > I don't think I need to do anything before exit. However, maybe I need to
> > ensure the rtc operations is thread-safe; I haven't done it.
>
> Ah, yes. Since it's not a performance-critical service, you can as well
> use ports_manage_port_operations_one_thread instead, and get done with
> it :)
Indeed, I will employ ports_manage_port_operations_one_thread.
> > I will do some experiments to see whether two threads setting rtc causes
> > data race
>
> It will since cmos access requires two outb calls, but it'll be very
> hard to reproduce.
OK, remove it from my TODO list.
> > > > +  The following flags have to be released exactly in this order,
> > > > +  otherwise the DS12887 (popular MC146818A clone with integrated
> > > > +  battery and quartz) will not reset the oscillator and will not
> > > > +  update precisely 500 ms later. You won't find this mentioned in
> > > > +  the Dallas Semiconductor data sheets, but who believes data
> > > > +  sheets anyway ... -- Markus Kuhn */
> > > > +  cmos_write(11, save_control);
> > > > +  cmos_write(10, save_freq_select);
> > >
> > > Can these writes somehow fail, and if so, can you detect it and report
> > > the error? (I've no idea.)
> > I think the only way to check is reading rtc between `cmos_write(9,
> > tm.tm_year);` and `cmos_write(11, save_control);`; then compare the values
> > we just read with the values for setting.
>
> I don't think it is useful to do this. Nowadays RTC is not even a
> separate chip, we won't have any issue.
OK, remove it from my TODO list.

--
Zhaoming Luo