LKM Timemark Driver
Stromeko at nexgo.de
Tue Aug 28 17:53:26 UTC 2018
MLewis via devel writes:
> "Strictly speaking, the PPS pulse isnï¿½t even necessary for this process. One could
> simply generate a pulse on the external interrupt line once per second at a known
> time. "
Yes, and if you use both measurements to eliminate the interrupt
latency, then random error propagation (assuming jitter is not
correlated between input and output) ensures that your measurement error
goes up by a factor of sqrt(2). So you're better of just ignoring the
PPS kernel timestamp even from a theoretical standpoint.
> I'm currently using REALTIME_CLOCK. There's also the thread clock and
> the cpu clock.
Principally it should be better to use an unslewed clock source (one of
the hardware timers). That's what the RADclock was all about, but the
API sadly never went into the mainline kernel. But without a proper API
these are tricky to use portably and you're left with various
interference from e.g. power saving.
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk Blofeld]>+
SD adaptation for Waldorf rackAttack V1.04R1:
More information about the devel