That's why when you get a real time you adjust the times on your logged events. There's the time you got the time fix, everything else is N microseconds before that back to when you started recording. So you record back to <time fix time> minus <recording duration>.
On Sun, Feb 21, 2021, 12:47 PM Reco <[email protected]> wrote: > On Sun, Feb 21, 2021 at 12:19:31PM -0500, Alan Corey wrote: > > I think it's unreasonable to expect that kind of time accuracy from the > > first microsecond of bootup. Relative accuracy maybe, by counting cycles > > of a crystal oscillator and storing events in some buffer. Then once you > > have a good time reference write them all out to permanent storage by > doing > > the arithmetic to assign real times to those events. > > The kernel itself does exactly this. > But what does it tell the kernel about the "right" time in the rest of > the world (i.e. - any other host)? Exactly nothing. > > End result - you can measure whatever happens during the boot just fine, > but you start 1st Jan 1970 0:00 each boot. > > You see, the problem is not the timekeeping, it's the setting > more-or-less the same time on different systems. > > Reco > >
