Roland> Hmm, looking at kernel/irq/spurious.c I think I see a
Roland> bug... I'll take this up on lkml.
Atually I misread the code. But I still don't understand how we could
ever return IRQ_NONE for 99900 out of 100000 calls to our interrupt
handler -- that would seem to require the HCA to generate many
interrupts in a row with all EQs empty.
- R.
_______________________________________________
openib-general mailing list
[email protected]
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
