Here are the pending networking bits.
There was going to be the addition of the ath9k wireless driver but
there was some fallout that is being worked on right now (linux/list.h
changes needed lkml review, ath9k driver triggered some gcc aborts,
all kinds of fun stuff :-) so hopefully it will make it in the next
pass. I tried to wait an extra day for it to be resolved, but that
was optimistic.
Highlights:
1) Intel driver guys triggered an OOPS when setting MTU during
a busy transmit stream, which turned out to be a qdisc locking
error.
Fix the bug, and in the next commit add some assertions to
qdisc_root_lock() to make sure it is only used in contexts where
it is safe (basically, it can only be used when RTNL is held)
2) Neighbour table procfs output can miss entries, fix from
Chris Larson.
3) When traffic classifier actions in the packet scheduler return
certain status values, this can end up corrupting root level
qdisc queue lengths and cause other weirdness. Based upon some
collaboration between Patrick McHardy, Jarek Poplawski, and
myself Jarek composed a patch which uses flag bits (which
get masked out before anything outside the qdisc layer can see
them) to indicate exactly which actions should be performed
on qdiscs leading up to the root.
4) Lennert Buytenhek did some really nice analysis on a network
device that cannot do TSO offloading in hardware. He checked
out what happens if you enable the software TSO mechanism fallback
we have in the kernel, and it improves CPU utilization tremendously.
It is safe to do this as long as the device in question can
support scatter-gather.
Herbert and I are discussing a way to do this even more efficiently
with some help from the device (currently the code has to allocate
extra sk_buff objects as we split up the TSO frame, and then do
a bunch of extra page ref counting, when all we need is some headers
and some way to say where the data ...
