> Hi, all. Just to let you know, I'm still getting lost interrupts
> with my NE2000 PCI clone. Motherboard is a Tyan Titan Pro (dual CPU).
> This is with 2.1.99-SMP.
"Lost interrupt _condition_" has sense, but "lost interrupts" hasn't, IMO.
If the driver is expecting that the number of interrupt handler calls
always matches the number of expected events from the device, then
the driver is broken. I cannot imagine the driver be broken this way.
So, I would think that the board is perhaps using edge-triggered
interrupt signaling or it does not sustain the signal level of the
level-sensitive interrupt line as it should.
Using edge-triggered interrupt when level-sensitive is expected leads
IMO to interrupt condition being lost and donnot allow sharing IRQ
as you know.
> I note that pre-patch-2.1.100-1 doesn't have any fixes for the IRQ
> code: does that mean nobody has any ideas on how to fix it?
Yes. A bug in the kernel is also candidate.
Regards,
Gerard.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu