On Tue, 14 Nov 2017, Linus Torvalds wrote:
> On Tue, Nov 14, 2017 at 12:21 PM, Thomas Gleixner <t...@linutronix.de> wrote:
> >
> > First of all sched_clock() and CLOCK_MONOTONIC can drift apart frequency
> > wise pretty bad.
>
> Yes. And I would actuall ynot object to just making the printk buffers
> use CLOCK_MONOTONIC.
>
> The reason we avoided that was we didn't want to take the locks, though.
I know, but today we can use the NMI safe accessor which avoids that.
> sched_clock() is used not because it's the right clock, or
> particularly good, but because it is fast and doesn't have locking
> issues.
>
> But all of this is entirely separate from the "changing from one clock
> that has small issues for good reasons to FOUR DIFFERENT CLOCKS THAT
> THE USER CHOOSES RANDOMLY".
>
> See my objection?
I understand that and I certainly was not trying to argue about that.
Thanks,
tglx
