[asterisk-dev] Re: Kernel modules => mainline kernel

[Steven Critchfield]

On Wed, 2007-02-14 at 11:15 -0800, Paul Cadach wrote:
> Hi,
> 
> Andrew Kohlsmith wrote:
> > On Tuesday 13 February 2007 7:26 am, Tzafrir Cohen wrote:
> >> Still not grand.
> >> But if your main usage is PRI => VoIP, the extra 10ms would not be that
> >> critical. Chances are you have longer delays anyway.
> >
> > True; perhaps it's time to dig out some code and see what happens if you
> > disable interrupts on Zaptel and use ztdummy with hires timers to poll 
> > real
> > hardware.  :-)
> 
> And to update hardware PLLs to synchronize boards from zaptel clock.... ;-)
> 
> Polling is Ok if you have synchronized (or just single board) hardware so 
> you should not wake-up every 1/10000 sec to check data on next board.

Let me back this argument up for a moment. Maybe instead of worrying
about the interupt level for a machine that would need 8 or 9 PCI slots,
we should be looking at more dense spans per PCI cards for those
deployments. 

Seems to me that you are much more likely to get a more dense interface
card than a driver change to make it possible to put twice the normal
number of cards in a machine.
-- 
Steven Critchfield <critch at basesys.com>