Re: Linux timekeeping plans

Colin Plumb (colin@nyx.net)
Thu, 3 Dec 1998 08:43:11 -0700 (MST)


>> Currently, as I understand it, external interrupts are delivered to one
>> processor, but how that processor is chosen is undefined. It may or

> Its chosen by the APIC hardware, you can control this in great detail
> if you desire (see the 82489 manual). The APICs are also a source of
> time interrupts of their own.

I guess I wasn't clear - I can look up the x86-specific hardware, but I
wonder about other architectures. Including future architectures for
which there currently is no code. What is the Linux kernel SMP model?

I'd like to avoid adding restrictions that aren't already there, so I
want to understand what assumptions *are* already present.

>> Can I create a broadcast interrupt which all processors use for
>> timekeeping?

> Yes for 16 or less processor machines

So not on the Fujitsu 64-processor SPARC? But I can on a 12-processor SGI
machine? (Yes, I'm being snide, as I know that's not what you're saying.
But it's the question I was trying to ask.)

-- 
	-Colin

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/