Kernel PLL

[Hal Murray]

> I'd like to hear more about this.  It sounds like a separate issue from the
> damon split. 

> Can't really respond to this as I don't understand the kernel PLL.

Mills wrote some kernel code that did a PLL off of a PPS signal.  It, or its 
descendents. is available in Linux, FreeBSD, and NetBSD.  It is not configured 
on any distros that I know about. When enabled, it works much better than ntpd.

Our current SHM or PPS processing is polled.  I don't know if the "much 
better" is because he got to pick better PLL parameters knowing that the 
signal was a PPS rather than potentially junk data from network timing or if 
the key is quick response time due to avoiding any polling delays.

So I'd like a way to avoid the polling delays.

One of these days, I'll make enough time to explore this area.


-- 
These are my opinions.  I hate spam.