Kernel PLL, from IRC

[Hal Murray]

gemiller
  I have a hard time believing a PLL
  running at user timer interrupt time
  can be anywhere as good as a kerenl
  PLL running at TICK intervals.

There are two parts to this discussion.  There is the PPS time stamp and the 
PLL.

The PPS time stamp gives you the data to work with.  The work can be done in 
user space or in the kernel.  New samples arrive every second.  Suppose I use 
the same algorithm.  Running the code in user space adds a few sub-ms kernel 
calls.  (I'm assuming the scheduler cooperates.)  I don't think that minor 
delay will be significant.

The kernel PLL is probably using different parameters than ntpd.


gemiller
  Is there a simple switch to select
  between the two?  I could run some
  tests then.

It's a flag on the PPS driver.  But you will probably have to build a kernel. 
 Most distros ship with the magic feature disabled.

-------

gemiller
  yes, best to measure

Check your clockstats on your minpoll-0 runs and see if it really was running 
at 0.  On my quick tries, it was actually running at 1 (2 samples per batch)  
I didn't track it down.


-- 
These are my opinions.  I hate spam.