[Asterisk-Users] Generic X100P's

[Steven Critchfield]

On Tue, 2004-10-12 at 17:44 +0200, Peter Svensson wrote:
> On Tue, 12 Oct 2004, Steven Critchfield wrote:
> 
> > Funny since there is only 4 real IRQ lines on a PCI bus. They are A, B,
> > C, and D. If you have more than 4 slots on a PCI bus, then you are most
> > definately reusing a real IRQ wire. 
> > 
> > As for if PPC could handle it, I haven't seen any drivers. Could be me
> > ignoring those threads.  
> 
> This is not correct. Each pci slot has four physical interrupt lines, A-D. 
> The implementation is free to supply four separate interrupt lines to each 
> card, i.e. the interupt lines are different for each slot on the bus.

Unless some PCI chipset is now dolling out 4 IRQ lines per slot, they
have to merge somewhere and at that merge, they become shared. 

http://old.lwn.net/1998/0205/io-apic.html

> On the other hand, most of the comments about the problems with interrupt 
> sharing on the pci bus are not very informed either. The interrupt handles 
> are normally very fast and the pci specification definitly allows for 
> shared interrupts. I suspect either a hardware problem or that the dirvers 
> are somehow incorrect. 
> 
> Does anyone have any hard numbers on the interrupt latencies tolerated by 
> the digium cards? I would assume 100-200us to be acceptable. 

The facts about Zapata hardware is that there is no buffering. If you
miss an interupt, you miss data. Interupts are fired 1000 times a second
or as soon as 8 bits are collected on the 8000hz phone lines. 

-- 
Steven Critchfield <critch at basesys.com>