Re: [PATCH v4 1/1] printk: fix zero-valued printk timestamps in early boot

"Roberto A. Foglietta" <[email protected]> Tue, 21 Apr 2026 11:51:44 +0200
Newsgroups org.kernel.vger.linux-embedded,org.kernel.vger.linux-kernel
Message-ID <CAJGKYO6c1g0WCeQmZmbKkVgDgnSyOF4oCw5JUuxAdoCdH8nHtg@mail.gmail.com>
On Tue, 21 Apr 2026 at 10:36, Petr Mladek <[email protected]> wrote:
>
> On Mon 2026-04-20 22:18:29, Bird, Tim wrote:
> > > From: Petr Mladek <[email protected]>
> > >    2. The early timestamps provided by the bogo cycles
> > >       are not synchronized with timestamps from
> > >       the proper time keeping.
> > >
> > >       Would it help to print a disclaimer, similar to,
> > >       for example, trace_printk() first use?

Every timestamp before the timekeeper init is arbitrary wrong or
unreliable, or it is a hardware value from a registry. As simple as
that.

> > >       Something like:
> > >
> > > [    0.002912] **********************************************************
> > > [    0.002917] ****   NOTICE NOTICE NOTICE NOTICE NOTICE NOTICE NOTICE
> > > [    0.002921] **
> > > [    0.002935] ** Using BOGO early timestamps
> > > [    0.002939] **
> > > [    0.002943] ** They are not properly calibrated and might use a source
> > > [    0.002949] ** with an unstable frequency.
> > > [    0.002953] **
> > > [    0.002957] ** They are not comparable with timestamps after
> > > [    0.002961] ** the timekeeping is initialized.
> > > [    0.002966] **
> > > [    0.002968] ****   NOTICE NOTICE NOTICE NOTICE NOTICE NOTICE NOTICE
> > > [    0.002971] *******************************************************
> >

Considering that it is supposed to be a debug feature, a large banner
like timesquare end of WW2 seems the most appropriate way to "not
interfere with the process under debug". Can we also have a playable
ASCII doom version of it?

Use a negative number to indicate something that precedes the time. A
negative number where a timestamp is expected should puzzle every
reasonable mind, plus this feature is for debugging and it is not
supposed to be seen in production. Or even simpler printing a registry
value and let the "expert senior" that activated that hacking dealing
with it. The kernel has no time to print? Then, don't mimic time.

Hence patch V6

Best regards, R-