9.x vs 10.x PF performance using ALTQ

[Nick Rogers]

Hello,

I have a question about the performance changes to PF introduced in 10.0. I
understand that PF is now able to take advantage of multiple CPUs / SMP
better because of finer grain locking. Does this also apply to using ALTQ
w/ HFSC? Can I expect packet forwarding performance to increase when
upgrading from 9.x to 10.x when using ALTQ? My understanding is that ALTQ
is still locked to a single core, or is this no longer true?

Also, I compile my igb driver with the LEGACY_IGB_TX to be able to use
igb(4) interfaces with ALTQ. Can I still expect to see a performance
increase with the PF changes in 10.x?

Thanks,
-Nick