On Mon, 9 Sep 2002, Ingo Molnar wrote:
>
> There's also the following effect that could generate additional
> interrupts: the *same* IRQ source that is currently executing might
> generate a (spurious but otherwise harmless) interrupt if we first ACK the
> card then ACK the APIC and then do processing. Our current way of masking
> interrupts in the IO-APIC at least leaves them pending there until the
> handler's main work loop is finished and mitigates irqs.
I agree with you, but that is only true for edge-triggered APIC
interrupts, though - for level-triggered ones we will just re-take the
interrupt when we unmask it again.
Which is kind of sad. Is there some fast way to read the status of a
level-trigger irq off the IO-APIC in case it is still pending, and to do
the mitigation even for level-triggered?
(Btw, if there is, that would also allow us to notice the "constantly
screaming PCI interrupt" without help from the low-level isrs)
Linus
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Sun Sep 15 2002 - 22:00:17 EST