Re: Weird clock behaviour with current (amd64) kernel

[Robert Elz <kre%munnari.OZ.AU@localhost>]

    Date:        Thu, 4 Aug 2022 12:49:35 -0000 (UTC)
    From:        mlelstv%serpens.de@localhost (Michael van Elst)
    Message-ID:  <tcgf8u$icc$1%serpens.de@localhost>

  | The measurement runs with enabled interrupts. If you have lots of interrupts
  | or interrupts that take some time, the measurement is biased.
  |
  | Console output can do this.

That is what I suspected.   A normal boot dmesg on this system is about
50KB.   With PCI_CONFIG_DUMP it is about 1MB (just a bit over).  That's
a lot of work for wscons scrolling (via the BIOS the whole time - the
dump all happens before the console switches to graphics mode) a fairly
large screen.

The effect is that when I do this, the TSC seems to be running about 3
times faster than it actually does, and consequently, timekeeping is
horrid (as long as it remains the clock source).  Switching away from it
causes time to flow at the correct rate - but things like "sleep 1" still
take 3 seconds to complete.

This is not important - PCI_CONFIG_DUMP isn't something that it is rational
to leave turned on in an everyday kernel.   I'll probably stop even bothering
to build that one now.

kre