On Tue, 10 Apr 2007 11:36:27 -0400 Chuck Ebbert <[EMAIL PROTECTED]> wrote:
> Can irqpoll be enhanced to show which interrupts are being misrouted? > It looks like it should be possible to show the actual interrupt, > the driver that is handling it, and what interrupt it's expecting > in a /proc file -- just what happened last time irqpoll resolved > the interrupt would be better than nothing. You can take a guess at this, and Ingo did propose that it ought to be possible to dynamically deduce all the IRQ routing this way. There are problems however. - It is possible a given IRQ simply vanished. In this scenario we pick up the fact work needed doing from the timer tick - We will get a lot of false positives as we scan. You could certainly collate statistical data and build counts of which stuck IRQ caused which IRQ events to be handled and which IRQ they are supposedly bound to. You can also do similar things to count work done on a timer tick, which may help. It won't be perfect data - and with the timer stuff may be horribly inadequate, but it should give you a good guess. - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
