Merge git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf-next

Pablo Neira Ayuso says:

====================
Netfilter/IPVS updates for net-next

The following patchset contains Netfilter/IPVS updates for your net-next tree
in this 4.4 development cycle, they are:

1) Schedule ICMP traffic to IPVS instances, this introduces a new schedule_icmp
   proc knob to enable/disable it. By default is off to retain the old
   behaviour. Patchset from Alex Gartrell.

I'm also including what Alex originally said for the record:

"The configuration of ipvs at Facebook is relatively straightforward.  All
ipvs instances bgp advertise a set of VIPs and the network prefers the
nearest one or uses ECMP in the event of a tie.  For the uninitiated, ECMP
deterministically and statelessly load balances by hashing the packet
(usually a 5-tuple of protocol, saddr, daddr, sport, and dport) and using
that number as an index (basic hash table type logic).

The problem is that ICMP packets (which contain really important
information like whether or not an MTU has been exceeded) will get a
different hash value and may end up at a different ipvs instance.  With no
information about where to route these packets, they are dropped, creating
ICMP black holes and breaking Path MTU discovery.  Suddenly, my mom's
pictures can't load and I'm fielding midday calls that I want nothing to do
with.

To address this, this patch set introduces the ability to schedule icmp
packets which is gated by a sysctl net.ipv4.vs.schedule_icmp.  If set to 0,
the old behavior is maintained -- otherwise ICMP packets are scheduled."

2) Add another proc entry to ignore tunneled packets to avoid routing loops
   from IPVS, also from Alex.

3) Fifteen patches from Eric Biederman to:

* Stop passing nf_hook_ops as parameter to the hook and use the state hook
  object instead all around the netfilter code, so only the private data
  pointer is passed to the registered hook function.

* Now that we've got state->net, propagate the netns pointer to netfilter hook
  clients to avoid its computation over and over again. A good example of how
  this has been simplified is the former TEE target (now nf_dup infrastructure)
  since it has killed the ugly pick_net() function.

There's another round of netns updates from Eric Biederman making the line. To
avoid the patchbomb again to almost all the networking mailing list (that is 84
patches) I'd suggest we send you a pull request with no patches or let me know
if you prefer a better way.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

1 files changed, 10 insertions, 0 deletions

diff --git a/Documentation/networking/ipvs-sysctl.txt b/Documentation/networking/ipvs-sysctl.txt
index 3ba7095..e6b1c02 100644
--- a/Documentation/networking/ipvs-sysctl.txt
+++ b/Documentation/networking/ipvs-sysctl.txt
@@ -157,6 +157,16 @@ expire_quiescent_template - BOOLEAN
 	persistence template if it is to be used to schedule a new
 	connection and the destination server is quiescent.
 
+ignore_tunneled - BOOLEAN
+	0 - disabled (default)
+	not 0 - enabled
+
+	If set, ipvs will set the ipvs_property on all packets which are of
+	unrecognized protocols.  This prevents us from routing tunneled
+	protocols like ipip, which is useful to prevent rescheduling
+	packets that have been tunneled to the ipvs host (i.e. to prevent
+	ipvs routing loops when ipvs is also acting as a real server).
+
 nat_icmp_send - BOOLEAN
         0 - disabled (default)
         not 0 - enabled