Several LRO fixes

[kbowling]

Summary

This changeset contains a few tightly intermingled and additive bugfixes from Netflix and LLNW:

1. rrs - Lets make the LRO code look for true dup-acks and window update acks fly on through and combine.
2. rrs - Make the LRO engine a bit more aware of ack-only seq space. Lets not have it incorrectly wipe out newer acks for older acks when we have out-of-order acks (common in wifi environments).
3. jeggleston - LRO eating window updates

Based on all of the above I think we are RFC compliant doing it this way:

https://tools.ietf.org/html/rfc1122

section 4.2.2.16

"Note that TCP has a heuristic to select the latest window update despite possible datagram reordering; as a result, it may ignore a window update with a smaller window than previously offered if neither the sequence number nor the acknowledgment number is increased."

Test Plan

These changes are in use at LLNW and Netflix

[rstone]

I've written some unit tests that covering these cases here:

https://github.com/rysto32/sysunit/blob/master/netinet/tcp_lro.gtest.cpp#L850

I can confirm that the new behaviour is much more sane than the older behaviour.

sys/netinet/tcp_lro.c

826 ↗

(On Diff #39819)

I think that you mean "old ack" here. As far as I can tell, the only way to reach this point is to have a packet with no payload and th_ack < le->ack_seq.

[jason_eggnet.com]

sys/netinet/tcp_lro.c

826 ↗

(On Diff #39819)

Agreed.

[kbowling]

Fix comment per rstone and jegg

[rstone]

sys/netinet/tcp_lro.c

819 ↗

(On Diff #40065)

Technically we should only do this (or look at th->th_ack at all) if TH_ACK is set

[jason_eggnet.com]

sys/netinet/tcp_lro.c

819 ↗

(On Diff #40065)

On the transport call we discussed that this should probably be handled in another patch as it is a generic issue with the LRO code.

[sbruno]

I'll grab this and shovel it in after builds are done.