svn commit: r225892 - head/sys/mips/mips

[Alexander Motin]

Adrian Chadd wrote:
> On 1 October 2011 15:21, Jayachandran C. <jchandra at freebsd.org> wrote:
> 
>> I would like to understand this, reverting r216862 would take out the
>> critical_enter() and cpu_idleclock() which would keep the timer
>> interrupts coming is as usual, this should not affect the event timer.
> 
> I haven't verified that mav's timer stuff does the correct thing by
> configuring the clock timer to occur every 1000hz in this instance. I
> kinda hope so.

When setting kern.eventtimer.idletick=1, you were effectively disabling
cpu_idleclock()/cpu_activeclock() and you've properly received about
1127 timer interrupts per second (mix of 1000 of hz and 127 of stathz).
Also when CPU is busy or has high interrupt/context switch rate, that
happen automatically, so it definitely works.

If you are talking about programming hardware timer for periodic 1000Hz
rate, it doesn't happen automatically, because I doubt benefits it gives
worth strict aliasing of hardclock() and statclock() it causes. If you
like, you can do it manually via kern.eventtimer.periodic=1.

-- 
Alexander Motin