On Wed, 2012-01-18 at 14:54 +0100, Mike Galbraith wrote:

On Wed, 2012-01-18 at 12:11 +0100, Tim Sander wrote:

quoted

quoted

If that's it, you can apply the below, do the same edit, and see which
thread is grinding away.  From there, I'd set a trap.  Let sirq threads
detect that they are being awakened too fast (hey, I can't go to sleep,
the sirq I just processed is busy again, N times in a row) and leave a
note for wakeup_softirqd().  There, WARN_ON(ksoftirqd)[i].help_me) or
such, to see who is flogging which softirq mercilessly.

I didn't use this tricks, since top was already doing its job good enough :-).

Ok, you now know which softirq is being flogged, maybe that's enough for
someone who knows network goop, dunno.  If you set the trap, you'll get
an 8x10 color glossy of the slave driver, whip in hand ;-)

This could be another trylock spin loop. I'll have to have a deeper
look.

Thanks,

-- Steve