Bad interaction between sleep timeouts and Salut?

[Daniel Drake]

On 16 September 2010 18:11, John Gilmore <gnu at toad.com> wrote:
> It's pretty simple, actually.  When in "idle suspend", the system should
> remain fully functional, just burning fewer ergs.  It's an optimization,
> not a change of behavior.
>
> This means the system should wake up anytime it would've gotten an
> interrupt during normal operation.  Which means for any unicast
> packet, any ARP packet directed to it, and any multicast packet that
> it's listening for.

Right. The Marvell firmware already has a load of logic for deciding
whether to pass frames to the host when the host is powered up, i.e.
it only passes:
 1. Broadcast
 2. Multicast in a group that we're subscribed to
 3. Unicast to this particular XO
etc....

It should be a simple firmware tweak to make the same filter apply to
the "The host is asleep, I've received a frame, do I wake it up?"
question.

This will result in a lot of wakeups, but I think this is the only way
to reliably achieve this optimization without affecting the behaviour
of the system. And optimizations can happen afterwards to reduce the
wakeups.

Daniel