cascardo/ovs.git
7 years agodatapath: compat: clear skb IPCB block.
Pravin B Shelar [Sun, 17 Jul 2016 16:52:01 +0000 (09:52 -0700)]
datapath: compat: clear skb IPCB block.

OVS compat ip_local_out/ip6_local_out clear out skb IPCB, we also
need to do it for non compat version.

Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agopackaging: Remove PyQt4 dependency.
Joe Stringer [Fri, 15 Jul 2016 00:53:46 +0000 (17:53 -0700)]
packaging: Remove PyQt4 dependency.

This doesn't seem to have been necessary since ovsdbmonitor was removed
in commit 7868fbc6c97c ("ovsdbmonitor: Remove."). Drop the dependency.

Signed-off-by: Joe Stringer <joe@ovn.org>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agodatapath-windows: Fixed packet limits on STT protocol
Paul Boca [Mon, 4 Jul 2016 09:53:23 +0000 (09:53 +0000)]
datapath-windows: Fixed packet limits on STT protocol

"If the length of the payload to be encapsulated exceeds 64KB, or if
the offset to the L4 header exceeds 255 bytes, then it will not be
possible to offload the packet to the NIC for segmentation." (STT RFC)
In this case the packet needs to be segmented by us, before sending it.

Signed-off-by: Paul-Daniel Boca <pboca@cloudbasesolutions.com>
Acked-by: Sairam Venugopal <vsairam@vmware.com>
Signed-off-by: Gurucharan Shetty <guru@ovn.org>
7 years agoovs-vtep: Handle tunnel key configuration in any order.
Darrell Ball [Thu, 14 Jul 2016 21:59:00 +0000 (14:59 -0700)]
ovs-vtep: Handle tunnel key configuration in any order.

Presently, ovs-vtep expects the datapath tunnel key to be available
in the VTEP DB at startup. This may not be the case which is also
observed as interrmittent unit test failures. This patch allows
for the tunnel key to later appear in the VTEP database.

Signed-off-by: Darrell Ball <dlu998@gmail.com>
Acked-by: Russell Bryant <russell@ovn.org>
Acked-by: Daniele Di Proietto <diproiettod@vmware.com>
7 years agoovn-controller-vtep: occasional failure in "binding 1" test case
Lance Richardson [Fri, 15 Jul 2016 13:15:05 +0000 (09:15 -0400)]
ovn-controller-vtep: occasional failure in "binding 1" test case

The ovn-controller-vtep "binding 1" test case fails occasionally
due to a race with the ovs-vtep daemon.  If ovs-vtep happens to
execute handle_physical() after "ovs-vsctl del-port p0", but before
the test script has executed "vtep-ctl del-port br-vtep p0", the
latter command will fail because ovs-vtep will have already deleted
p0 from the vtep db.

Eliminate this race while adding an additional check to verify that
ovs-vtep is working as expected by waiting for ovs-vtep to remove
entries for the deleted physical ports from the vtep db.

Signed-off-by: Lance Richardson <lrichard@redhat.com>
Suggested-by: Darrell Ball <dlu998@gmail.com>
Acked-by: Darrell Ball <dlu998@gmail.com>
Signed-off-by: Russell Bryant <russell@ovn.org>
7 years agosystem-traffic: Remove datapath specific tests and macro.
William Tu [Fri, 1 Jul 2016 16:45:52 +0000 (09:45 -0700)]
system-traffic: Remove datapath specific tests and macro.

We generally try to keep the testsuite independent of the underlying
datapath. This patch removes the datapath-specific tests and macros.

Tested-at: https://travis-ci.org/williamtu/ovs-travis/builds/141642065
Signed-off-by: William Tu <u9012063@gmail.com>
Signed-off-by: Joe Stringer <joe@ovn.org>
7 years agoovn-northd: Remove extraneous ip4 mulicast logical flow.
Ryan Moats [Thu, 14 Jul 2016 18:34:54 +0000 (13:34 -0500)]
ovn-northd: Remove extraneous ip4 mulicast logical flow.

IPv4 multicast is already dropped by a previous logical flow
written for the logical router, so there is no reason to add
a lower priority rule for it.

Signed-off-by: Ryan Moats <rmoats@us.ibm.com>
Acked-by: Flavio Fernandes <flavio@flaviof.com>
Signed-off-by: Russell Bryant <russell@ovn.org>
7 years agotests: daemon specific tests
Alin Serdean [Tue, 12 Jul 2016 15:36:02 +0000 (15:36 +0000)]
tests: daemon specific tests

Testing out the named pipe implementation revealed a problem in
"daemon --detach startup errors". If the daemon actually started nobody
is stopping it.
In the case of test failure kill the daemon.

Signed-off-by: Alin Gabriel Serdean <aserdean@cloudbasesolutions.com>
Acked-by: Paul Boca <pboca@cloudbasesolutions.com>
Signed-off-by: Gurucharan Shetty <guru@ovn.org>
7 years agoofp-actions: Remove XXX comment for write-actions in ofpacts_format().
Ben Pfaff [Wed, 13 Jul 2016 23:43:08 +0000 (16:43 -0700)]
ofp-actions: Remove XXX comment for write-actions in ofpacts_format().

I don't know why this was here.  The write_actions instruction works fine.

Signed-off-by: Ben Pfaff <blp@ovn.org>
Acked-by: Jarno Rajahalme <jarno@ovn.org>
7 years agoofp-actions: Translate OF1.1+ mod_nw_ttl action to OF1.0 properly.
Ben Pfaff [Wed, 13 Jul 2016 23:41:00 +0000 (16:41 -0700)]
ofp-actions: Translate OF1.1+ mod_nw_ttl action to OF1.0 properly.

Signed-off-by: Ben Pfaff <blp@ovn.org>
Acked-by: Jarno Rajahalme <jarno@ovn.org>
7 years agoofp-actions: Translate mod_nw_ecn action to OF1.1 properly.
Ben Pfaff [Wed, 13 Jul 2016 23:50:33 +0000 (16:50 -0700)]
ofp-actions: Translate mod_nw_ecn action to OF1.1 properly.

Also, translate OF1.2+ "set_field" on OXM_OF_IP_ECN properly to OF1.1
"mod_nw_ecn".

Signed-off-by: Ben Pfaff <blp@ovn.org>
Acked-by: Jarno Rajahalme <jarno@ovn.org>
7 years agoofp-actions: Translate OF1.0 "enqueue" actions for OF1.1+.
Ben Pfaff [Wed, 13 Jul 2016 22:53:20 +0000 (15:53 -0700)]
ofp-actions: Translate OF1.0 "enqueue" actions for OF1.1+.

Previously, the OF1.0 "enqueue" action was simply omitted when actions
were translated into OpenFlow 1.1 or later, which do not have a similar
action.  This commit translates this action into an equivalent sequence
of actions.

Signed-off-by: Ben Pfaff <blp@ovn.org>
Acked-by: Jarno Rajahalme <jarno@ovn.org>
7 years agoOVN-Tutorial: Update sample output.
Russell Bryant [Thu, 7 Jul 2016 19:03:41 +0000 (15:03 -0400)]
OVN-Tutorial: Update sample output.

A previous patch made some additions to section 1 of this tutorial.
This patch includes updates to the sample output that I got while
testing the additions to the tutorial.

Signed-off-by: Russell Bryant <russell@ovn.org>
Acked-by: Ryan Moats <rmoats@us.ibm.com>
7 years agoofp-actions: Add extension to support "group" action in OF1.0.
Ben Pfaff [Thu, 14 Jul 2016 15:27:21 +0000 (08:27 -0700)]
ofp-actions: Add extension to support "group" action in OF1.0.

From time to time it confuses users that "group" actions disappear when
using OpenFlow 1.0.

Signed-off-by: Ben Pfaff <blp@ovn.org>
Acked-by: Jarno Rajahalme <jarno@ovn.org>
7 years agoAUTHORS: Add Eziz Durdyyev.
Ben Pfaff [Thu, 14 Jul 2016 14:56:07 +0000 (07:56 -0700)]
AUTHORS: Add Eziz Durdyyev.

Signed-off-by: Ben Pfaff <blp@ovn.org>
7 years agotoken-bucket: Correctly truncate existing tokens.
Eziz Durdyyev [Tue, 5 Jul 2016 17:41:50 +0000 (22:41 +0500)]
token-bucket: Correctly truncate existing tokens.

Submitted-at: https://github.com/openvswitch/ovs/pull/141
Signed-off-by: Eziz Durdyyev <ezizdurdy@gmail.com>
Signed-off-by: Ben Pfaff <blp@ovn.org>
7 years agoWindows: Use NETLINK_NETFILTER instead of NETLINK_GENERIC
Sairam Venugopal [Mon, 11 Jul 2016 21:59:50 +0000 (14:59 -0700)]
Windows: Use NETLINK_NETFILTER instead of NETLINK_GENERIC

Windows datapath lacked support for different Netlink Family protocols.
Now that Windows supports different Netlink protocol, revert the change to
override NETLINK_NETFILTER to use NETLINK_GENERIC.

Signed-off-by: Sairam Venugopal <vsairam@vmware.com>
Acked-by: Nithin Raju <nithin@vmware.com>
Signed-off-by: Gurucharan Shetty <guru@ovn.org>
7 years agoWindows: Add support for handling protocol (netlink family)
Sairam Venugopal [Mon, 11 Jul 2016 21:59:49 +0000 (14:59 -0700)]
Windows: Add support for handling protocol (netlink family)

Windows datapath currently has no notion of netlink family.
It assumes all netlink messages to belong to NETLINK_GENERIC family.
This patch adds support for handling other protocols if the userspace sends it down to kernel.

This patch introduces a new NETLINK_CMD - OVS_CTRL_CMD_SOCK_PROP to manage
all properties associated with a socket. The properties are passed down as
netlink message attributes. This makes it easier to introduce other
properties in the future.

Signed-off-by: Sairam Venugopal <vsairam@vmware.com>
Acked-by: Nithin Raju <nithin@vmware.com>
Signed-off-by: Gurucharan Shetty <guru@ovn.org>
7 years agoovn-sbctl: Adjust formatting of lflow-list.
Russell Bryant [Sun, 10 Jul 2016 21:46:26 +0000 (17:46 -0400)]
ovn-sbctl: Adjust formatting of lflow-list.

Adjust the field width for a couple of elements in the output of the
lflow-list sub-command of ovn-sbctl. We now have a table ID of 10, so
set a field width of 2 to keep output aligned. Table names have grown to
a max of 19 characters, so update the field width to keep output
aligned.  Finally, left justify values in fields, as I find that to look
a little nicer.

Also update sample output in OVN-Tutorial.md to match the current state.

Signed-off-by: Russell Bryant <russell@ovn.org>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agoofproto: Fix consistent hashing
Liran Schour [Tue, 12 Jul 2016 06:15:40 +0000 (09:15 +0300)]
ofproto: Fix consistent hashing

Hashing will not be consistent as long as we use for hashing the index of the
bucket in the list (for remove/insert of buckets not from/to the end of the
bucket list).
Use bucket_id for hashing instead.

Signed-off-by: Liran Schour <lirans@il.ibm.com>
Signed-off-by: Ben Pfaff <blp@ovn.org>
Acked-by: Simon Horman <simon.horman@netronome.com>
7 years agosset: New function sset_from_delimited_string().
Ben Pfaff [Sat, 25 Jun 2016 04:30:39 +0000 (21:30 -0700)]
sset: New function sset_from_delimited_string().

This simplifies code in a couple of places.

Signed-off-by: Ben Pfaff <blp@ovn.org>
Acked-By: Ryan Moats <rmoats@us.ibm.com>
7 years agoovn-northd: Use ovs_be32 for an IP address in find_lrp_member_ip().
Ben Pfaff [Wed, 13 Jul 2016 20:02:59 +0000 (13:02 -0700)]
ovn-northd: Use ovs_be32 for an IP address in find_lrp_member_ip().

Fixes a number of warnings from "sparse".

Fixes: 4685e523695c ("ovn: Support multiple addresses on a single logical router port.")
Signed-off-by: Ben Pfaff <blp@ovn.org>
Acked-by: Gurucharan Shetty <guru@ovn.org>
7 years agosystem-common-macros.at: Add a iface-id to interfaces.
Gurucharan Shetty [Mon, 11 Jul 2016 10:06:21 +0000 (03:06 -0700)]
system-common-macros.at: Add a iface-id to interfaces.

IntegrationGuide.md explains the meaning of the key 'iface-id'. Adding
it as part of ADD_VETH is useful for OVN related tests. For non-OVN
tests, it should have no effect. Upcoming commits take advantage of this.

Signed-off-by: Gurucharan Shetty <guru@ovn.org>
Acked-by: Joe Stringer <joe@ovn.org>
7 years agosystem-common-macros: Enhance ADD_VETH to take mac and gateway.
Gurucharan Shetty [Mon, 11 Jul 2016 09:55:59 +0000 (02:55 -0700)]
system-common-macros: Enhance ADD_VETH to take mac and gateway.

It is useful to set the mac address and gateway while using ADD_VETH
to connect a namespace to a OVN logical topology. Upcoming commits
use this enhancement.

Signed-off-by: Gurucharan Shetty <guru@ovn.org>
Acked-by: Joe Stringer <joe@ovn.org>
7 years agoovsdb-client: Move iterator definition to prevent shadowing.
Justin Pettit [Tue, 12 Jul 2016 19:51:22 +0000 (12:51 -0700)]
ovsdb-client: Move iterator definition to prevent shadowing.

This doesn't fix a bug, but is more standard.

Signed-off-by: Justin Pettit <jpettit@ovn.org>
Acked-by: Daniele Di Proietto <diproiettod@vmware.com>
7 years agodpctl: Fix shadowed iterator in show_dpif().
Justin Pettit [Tue, 12 Jul 2016 19:46:20 +0000 (12:46 -0700)]
dpctl: Fix shadowed iterator in show_dpif().

Signed-off-by: Justin Pettit <jpettit@ovn.org>
Acked-by: Daniele Di Proietto <diproiettod@vmware.com>
7 years agoovn-northd: Fix shadowed iterators in port security functions.
Justin Pettit [Tue, 12 Jul 2016 18:15:00 +0000 (11:15 -0700)]
ovn-northd: Fix shadowed iterators in port security functions.

Signed-off-by: Justin Pettit <jpettit@ovn.org>
Acked-by: Daniele Di Proietto <diproiettod@vmware.com>
7 years agoovn-nbctl: Print router port networks in "show".
Justin Pettit [Mon, 11 Jul 2016 07:23:17 +0000 (00:23 -0700)]
ovn-nbctl: Print router port networks in "show".

Signed-off-by: Justin Pettit <jpettit@ovn.org>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agoovn: Support multiple addresses on a single logical router port.
Justin Pettit [Sat, 9 Jul 2016 16:55:20 +0000 (09:55 -0700)]
ovn: Support multiple addresses on a single logical router port.

Supporting multiple addresses is only mildly interesting for IPv4.
However, it is a requirement for IPv6, which will arrive in a future
commit.

This commit introduces the extract_lrp_networks() function to ovn-util.[ch].

Signed-off-by: Justin Pettit <jpettit@ovn.org>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agoovn-northd: Use strings from extract_lsp_addresses().
Justin Pettit [Sat, 4 Jun 2016 18:49:46 +0000 (11:49 -0700)]
ovn-northd: Use strings from extract_lsp_addresses().

Extract port security and logical switch port addresses once and store
them as part of the ovn_port structure.  Use the string representations
from the extracted addresses.

Signed-off-by: Justin Pettit <jpettit@ovn.org>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agoovn-util: Remove 'store_ipv6' argument from extract_lsp_addresses().
Justin Pettit [Fri, 3 Jun 2016 05:37:31 +0000 (22:37 -0700)]
ovn-util: Remove 'store_ipv6' argument from extract_lsp_addresses().

With the addition of IPv6 routing, there won't be much need to
special-case not parsing IPv6 addresses.

Signed-off-by: Justin Pettit <jpettit@ovn.org>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agoovn-util: Add string representations to 'lport_addresses'.
Justin Pettit [Fri, 3 Jun 2016 04:44:38 +0000 (21:44 -0700)]
ovn-util: Add string representations to 'lport_addresses'.

A future commit will reduce the amount of conversions used by the
existing users of 'lport_addresses'.  This change will also make it
possible to use this structure for logical router port networks.

Signed-off-by: Justin Pettit <jpettit@ovn.org>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agoovn-northd: Use strings for IP addresses in router instances of "ovn_port".
Justin Pettit [Mon, 26 Oct 2015 03:35:58 +0000 (20:35 -0700)]
ovn-northd: Use strings for IP addresses in router instances of "ovn_port".

Reduce the amount of string conversions necessary.  This will be
convenient when we want to start using IPv6 addresses, too.

Signed-off-by: Justin Pettit <jpettit@ovn.org>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agoovn-northd: Use dynamic strings when building router and switch flows.
Justin Pettit [Fri, 8 Jul 2016 23:25:08 +0000 (16:25 -0700)]
ovn-northd: Use dynamic strings when building router and switch flows.

Reduce the number of memory allocations and risk of introducing shadow
variables.

Signed-off-by: Justin Pettit <jpettit@ovn.org>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agoovn: Remove 'default_gw' from logical router table.
Justin Pettit [Tue, 17 May 2016 13:02:53 +0000 (06:02 -0700)]
ovn: Remove 'default_gw' from logical router table.

With static routes, it's not necessary to have a separate default
gateway parameter.  This also makes configuring router ports clearer
when IPv6 and IPv4 addresses may be assigned to the same port.

Signed-off-by: Justin Pettit <jpettit@ovn.org>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agoovn: Add xxreg[01] symbols.
Justin Pettit [Tue, 12 Jul 2016 01:25:24 +0000 (18:25 -0700)]
ovn: Add xxreg[01] symbols.

Signed-off-by: Justin Pettit <jpettit@ovn.org>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agoovn: Renumber logical field registers to the newly extended registers.
Justin Pettit [Fri, 24 Jun 2016 03:39:18 +0000 (20:39 -0700)]
ovn: Renumber logical field registers to the newly extended registers.

IPv6 addresses use four standard OVS registers, so move the existing
named registers back to make room for using more logical registers.

Signed-off-by: Justin Pettit <jpettit@ovn.org>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agoIntroduce 128-bit xxregs.
Justin Pettit [Sat, 31 Oct 2015 11:45:28 +0000 (04:45 -0700)]
Introduce 128-bit xxregs.

These are needed to handle IPv6 addresses.

Signed-off-by: Justin Pettit <jpettit@ovn.org>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agoIncrease number of registers to 16.
Justin Pettit [Fri, 24 Jun 2016 00:54:26 +0000 (17:54 -0700)]
Increase number of registers to 16.

With eight 32-bit registers, we can only store two IPv6 addresses, which is
pretty tight.

Signed-off-by: Justin Pettit <jpettit@ovn.org>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agoovn-northd: Fix comments about the flows.
Justin Pettit [Tue, 12 Jul 2016 06:21:29 +0000 (23:21 -0700)]
ovn-northd: Fix comments about the flows.

References to the specifc tables should probably be dropped, since
they'll continue to drift towards wrong.  In the meantime, correct the
ones that are there.

Signed-off-by: Justin Pettit <jpettit@ovn.org>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agoovn-controller: Log OpenFlow errors at "info" level.
Justin Pettit [Tue, 21 Jun 2016 21:06:29 +0000 (14:06 -0700)]
ovn-controller: Log OpenFlow errors at "info" level.

Otherwise, errors are logged at "debug" level.  Errors when pushing
flows can then seemingly be silently lost.

Signed-off-by: Justin Pettit <jpettit@ovn.org>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agoofctrl: Fix non-sensical comment.
Justin Pettit [Sun, 3 Jul 2016 12:30:06 +0000 (05:30 -0700)]
ofctrl: Fix non-sensical comment.

Signed-off-by: Justin Pettit <jpettit@ovn.org>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agoovs-bugtool: Port to python3.
Joe Stringer [Fri, 24 Jun 2016 21:15:21 +0000 (14:15 -0700)]
ovs-bugtool: Port to python3.

Fix python2-specific code in ovs-bugtool:
* python2 long() is the same as python2 int() and python3 int(). Convert
  the long() to int().
* raw_input() was renamed to input(). Use python-six's input() on python2.
* Drop lambda tuple unpacking, we can go back to regular lambda syntax.
* file() can be replaced with open().

Signed-off-by: Joe Stringer <joe@ovn.org>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agosystem-traffic: Use NC_EOF_OPT in truncate tests.
Joe Stringer [Mon, 11 Jul 2016 17:29:18 +0000 (10:29 -0700)]
system-traffic: Use NC_EOF_OPT in truncate tests.

NC_EOF_OPT should always be passed to netcat in system-traffic tests
when invoking netcat to send a single packet that does not expect a
response. While on typical fedora/RH based distributions the default
behaviour is to send the packet then return, there are multiple other
implementations of netcat that do not do this (for example, those used
by Debian and Ubuntu by default). For these alternative implementations,
we provide $NC_EOF_OPT to ensure that netcat simply sends the packet
then returns immediately.

Signed-off-by: Joe Stringer <joe@ovn.org>
Acked-by: Andy Zhou <azhou@ovn.org>
7 years agodatapath: Fix ip tunnel compilation for newer kernel.
Pravin B Shelar [Mon, 11 Jul 2016 20:06:10 +0000 (13:06 -0700)]
datapath: Fix ip tunnel compilation for newer kernel.

compat iptunnel_xmit is used in backported tunnel code. but
it was only defined for kernel older than 3.18, This patch fixes
it by compiling it for all kernel which needs to use backported
tunnel implementation.

Reported-by: Justin Pettit <jpettit@ovn.org>
Reported-by: Joe Stringer <joe@ovn.org>
Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath-windows: remove unused 'ovsUserTimestampDelta'
Nithin Raju [Sat, 9 Jul 2016 00:45:40 +0000 (17:45 -0700)]
datapath-windows: remove unused 'ovsUserTimestampDelta'

Signed-off-by: Nithin Raju <nithin@vmware.com>
Acked-by: Sairam Venugopal <vsairam@vmware.com>
Signed-off-by: Gurucharan Shetty <guru@ovn.org>
7 years agoovn-controller: Change strategy for gateway conntrack zone allocation.
Gurucharan Shetty [Fri, 8 Jul 2016 07:15:49 +0000 (00:15 -0700)]
ovn-controller: Change strategy for gateway conntrack zone allocation.

Commit 263064aeaa31e7 (Convert binding_run to incremental processing.)
changed the way patched_datapaths were handled. Previously we would
destroy the datastructure in every run and re-create it fresh. The new
way causes problems with the way conntrack zones are allocated as now
we can have stale port_binding entries causing segmentation faults.

With this commit, we simply don't depend on port_binding records in
conntrack zone allocation and instead store the UUID as a string in
the patch_datapath datastructure.

(The test enhanced with this commit would fail without the changes
in the commit. i.e. ovn-controller would crash. )

Signed-off-by: Gurucharan Shetty <guru@ovn.org>
Acked-by: Ryan Moats <rmoats@us.ibm.com>
7 years agoovn: Remove unreferenced patched datapaths.
Darrell Ball [Fri, 8 Jul 2016 02:26:06 +0000 (19:26 -0700)]
ovn: Remove unreferenced patched datapaths.

Patched datapaths that are no longer referenced should be removed from
the patched_datapaths map; otherwise incorrect state references for a
patched datapath may be used and also datapaths that are absent will be
interpreted as present.

Signed-off-by: Darrell Ball <dlu998@gmail.com>
Acked-by: Ryan Moats <rmoats@us.ibm.com>
Signed-off-by: Gurucharan Shetty <guru@ovn.org>
7 years agodatapath: backport: iptunnel: make rx/tx bytes counters consistent
Pravin B Shelar [Fri, 8 Jul 2016 23:37:34 +0000 (16:37 -0700)]
datapath: backport: iptunnel: make rx/tx bytes counters consistent

Upstream commit:
commit bc22a0e2ea03b75b51a1f722f93821744b5b5ff1
Author: Nicolas Dichtel <nicolas.dichtel@6wind.com>

    iptunnel: make rx/tx bytes counters consistent

    This was already done a long time ago in
    commit 64194c31a0b6 ("inet: Make tunnel RX/TX byte counters more consistent")
    but tx path was broken (at least since 3.10).

    Before the patch the gre header was included on tx.

    After the patch:
    $ ping -c1 192.168.0.121 ; ip -s l ls dev gre1
    PING 192.168.0.121 (192.168.0.121) 56(84) bytes of data.
    64 bytes from 192.168.0.121: icmp_req=1 ttl=64 time=2.95 ms

    --- 192.168.0.121 ping statistics ---
    1 packets transmitted, 1 received, 0% packet loss, time 0ms
    rtt min/avg/max/mdev = 2.955/2.955/2.955/0.000 ms
    7: gre1@NONE: <POINTOPOINT,NOARP,UP,LOWER_UP> mtu 1468 qdisc noqueue state UNKNOWN mode DEFAULT group default
        link/gre 10.16.0.249 peer 10.16.0.121
        RX: bytes  packets  errors  dropped overrun mcast
        84         1        0       0       0       0
        TX: bytes  packets  errors  dropped carrier collsns
        84         1        0       0       0       0

Reported-by: Julien Meunier <julien.meunier@6wind.com>
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: backport: ovs/vxlan: fix rtnl notifications on iface deletion
Pravin B Shelar [Fri, 8 Jul 2016 23:37:30 +0000 (16:37 -0700)]
datapath: backport: ovs/vxlan: fix rtnl notifications on iface deletion

At this point OVS tunneling is insync with upstream net
branch (commit 9a0fee2b552b1).

upstream commit:
    commit cf5da330bbdd0c06b05c525a3d1d58ccd82c87a6
    Author: Nicolas Dichtel <nicolas.dichtel@6wind.com>

    ovs/vxlan: fix rtnl notifications on iface deletion

    The function vxlan_dev_create() (only used by ovs) never calls
    rtnl_configure_link(). The consequence is that dev->rtnl_link_stat is
    never set to RTNL_LINK_INITIALIZED.
    During the deletion phase, the function rollback_registered_many() sends
    a RTM_DELLINK only if dev->rtnl_link_state is set to RTNL_LINK_INITIALIZED.

    Note that the function vxlan_dev_create() is moved after the rtnl stuff so
    that vxlan_dellink() can be called in this function.

    Fixes: dcc38c033b32 ("openvswitch: Re-add CONFIG_OPENVSWITCH_VXLAN")
CC: Thomas Graf <tgraf@suug.ch>
CC: Pravin B Shelar <pshelar@ovn.org>
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: backport: ovs/geneve: fix rtnl notifications on iface deletion
Pravin B Shelar [Fri, 8 Jul 2016 23:37:27 +0000 (16:37 -0700)]
datapath: backport: ovs/geneve: fix rtnl notifications on iface deletion

Upstream Commit:
    commit 41009481b690493c169ce85f591b9d32c6fd9422
    Author: Nicolas Dichtel <nicolas.dichtel@6wind.com>

    ovs/geneve: fix rtnl notifications on iface deletion

    The function geneve_dev_create_fb() (only used by ovs) never calls
    rtnl_configure_link(). The consequence is that dev->rtnl_link_state is
    never set to RTNL_LINK_INITIALIZED.
    During the deletion phase, the function rollback_registered_many() sends
    a RTM_DELLINK only if dev->rtnl_link_state is set to RTNL_LINK_INITIALIZED.

    Fixes: e305ac6cf5a1 ("geneve: Add support to collect tunnel metadata.")
CC: Pravin B Shelar <pshelar@ovn.org>
CC: Jesse Gross <jesse@ovn.org>
CC: Thomas Graf <tgraf@suug.ch>
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: backport: ovs/gre,geneve: fix error path when creating an iface
Pravin B Shelar [Fri, 8 Jul 2016 23:37:23 +0000 (16:37 -0700)]
datapath: backport: ovs/gre,geneve: fix error path when creating an iface

Upstream commit:
    commit 106da663ff495e0aea3ac15b8317aa410754fcac
    Author: Nicolas Dichtel <nicolas.dichtel@6wind.com>

    ovs/gre,geneve: fix error path when creating an iface

    After ipgre_newlink()/geneve_configure() call, the netdev is registered.

    Fixes: 7e059158d57b ("vxlan, gre, geneve: Set a large MTU on ovs-created tunnel devices")
CC: David Wragg <david@weave.works>
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: backport: geneve: fix tx_errors statistics
Pravin B Shelar [Fri, 8 Jul 2016 23:36:17 +0000 (16:36 -0700)]
datapath: backport: geneve: fix tx_errors statistics

Upstream commit:
    commit efeb2267bba8aa893afdadfc9bae4790777c600c
    Author: Haishuang Yan <yanhaishuang@cmss.chinamobile.com>

    geneve: fix tx_errors statistics

    Tx errors present summation of errors encountered while transmitting
    packets.

Signed-off-by: Haishuang Yan <yanhaishuang@cmss.chinamobile.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: backport: udp: prevent skbs lingering in tunnel socket queues
Pravin B Shelar [Fri, 8 Jul 2016 23:36:14 +0000 (16:36 -0700)]
datapath: backport: udp: prevent skbs lingering in tunnel socket queues

Upstream commit:
    commit e5aed006be918af163eb397e45aa5ea6cefd5e01
    Author: Hannes Frederic Sowa <hannes@stressinduktion.org>

    udp: prevent skbs lingering in tunnel socket queues

    In case we find a socket with encapsulation enabled we should call
    the encap_recv function even if just a udp header without payload is
    available. The callbacks are responsible for correctly verifying and
    dropping the packets.

    Also, in case the header validation fails for geneve and vxlan we
    shouldn't put the skb back into the socket queue, no one will pick
    them up there.  Instead we can simply discard them in the respective
    encap_recv functions.

Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: backport: udp_offload: Set encapsulation before inner completes.
Pravin B Shelar [Fri, 8 Jul 2016 23:36:11 +0000 (16:36 -0700)]
datapath: backport: udp_offload: Set encapsulation before inner completes.

Upstream commit:
    commit 229740c63169462a838a8b8e16391ed000934631
    Author: Jarno Rajahalme <jarno@ovn.org>

    udp_offload: Set encapsulation before inner completes.

    UDP tunnel segmentation code relies on the inner offsets being set for
    an UDP tunnel GSO packet, but the inner *_complete() functions will
    set the inner offsets only if 'encapsulation' is set before calling
    them.  Currently, udp_gro_complete() sets 'encapsulation' only after
    the inner *_complete() functions are done.  This causes the inner
    offsets having invalid values after udp_gro_complete() returns, which
    in turn will make it impossible to properly segment the packet in case
    it needs to be forwarded, which would be visible to the user either as
    invalid packets being sent or as packet loss.

    This patch fixes this by setting skb's 'encapsulation' in
    udp_gro_complete() before calling into the inner complete functions,
    and by making each possible UDP tunnel gro_complete() callback set the
    inner_mac_header to the beginning of the tunnel payload.

Signed-off-by: Jarno Rajahalme <jarno@ovn.org>
Reviewed-by: Alexander Duyck <aduyck@mirantis.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: backport: geneve: add IPv6 bits to geneve_fill_metadata_dst
Pravin B Shelar [Fri, 8 Jul 2016 23:36:08 +0000 (16:36 -0700)]
datapath: backport: geneve: add IPv6 bits to geneve_fill_metadata_dst

Upstream commit:
    commit b8812fa88371ae567c907448d9a7ba62d09b90c9
    Author: John W. Linville <linville@tuxdriver.com>

    geneve: add IPv6 bits to geneve_fill_metadata_dst

Signed-off-by: John W. Linville <linville@tuxdriver.com>
Reviewed-by: Jesse Gross <jesse@ovn.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: compat: get rid of OVS_CB inner header offsets.
Pravin B Shelar [Sat, 9 Jul 2016 01:25:55 +0000 (18:25 -0700)]
datapath: compat: get rid of OVS_CB inner header offsets.

OVS has GSO compat functionality which needs inner offset
of the packet to segment a packet. older kernel did not
include these offsets in skb, therefore these were stored
in OVS_GSO_CB. Now OVS has dropped support for these
old kernel, So none of the supported kernel needs this
comapt code. Following patch removes it.

Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: backport: openvswitch: correct encoding of set tunnel action attributes
Pravin B Shelar [Fri, 8 Jul 2016 23:24:29 +0000 (16:24 -0700)]
datapath: backport: openvswitch: correct encoding of set tunnel action attributes

upstream commit:
    commit e905eabc90a5b787d8708df164543ee295bea5f2
    Author: Simon Horman <simon.horman@netronome.com>

    openvswitch: correct encoding of set tunnel action attributes

    In a set action tunnel attributes should be encoded in a
    nested action.

    I noticed this because ovs-dpctl was reporting an error
    when dumping flows due to the incorrect encoding of tunnel attributes
    in a set action.

    Fixes: fc4099f17240 ("openvswitch: Fix egress tunnel info.")
Signed-off-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: compat: Use dst-cache for Geneve and VxLAN tunnels.
Pravin B Shelar [Fri, 8 Jul 2016 23:24:24 +0000 (16:24 -0700)]
datapath: compat: Use dst-cache for Geneve and VxLAN tunnels.

It partialy backport commit:
    commit d71785ffc7e7cae3fbdc4ea8a9d05b7a1c59f7b8
    Author: Paolo Abeni <pabeni@redhat.com>

    net: add dst_cache to ovs vxlan lwtunnel

    In case of UDP traffic with datagram length
    below MTU this give about 2% performance increase
    when tunneling over ipv4 and about 60% when tunneling
    over ipv6

Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Suggested-and-acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Bug fix commit db3c6139e6e ("bpf, vxlan, geneve, gre: fix usage of
dst_cache on xmit"). is also included. Geneve changes
were added in 468dfffcd762cbb2777ec5a76bc21e3748ebf47e ("geneve: add
dst caching support")

Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: Add support for IPv6 tunnels.
Pravin B Shelar [Fri, 8 Jul 2016 23:24:19 +0000 (16:24 -0700)]
datapath: Add support for IPv6 tunnels.

Mostly backports upstream commit along with other pieces
to make IPv6 tunneling work.

    commit 6b26ba3a7d952e611dcde1f3f77ce63bcc70540a
    Author: Jiri Benc <jbenc@redhat.com>

    openvswitch: netlink attributes for IPv6 tunneling

    Add netlink attributes for IPv6 tunnel addresses. This enables IPv6 support
    for tunnels.

Signed-off-by: Jiri Benc <jbenc@redhat.com>
Acked-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: compat: Update Geneve and VxLAN modules.
Pravin B Shelar [Fri, 8 Jul 2016 23:24:02 +0000 (16:24 -0700)]
datapath: compat: Update Geneve and VxLAN modules.

This patch brings in various updates to upstream Geneve and VxLAN
modules. For geneve this patch adds IPv6 support, for vxlan it adds
VXLAN GPE is the major feature.
This should make OVS compat tunnel implementation in sync upto
current net branch.

Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: compat: Add support for IPv6 UDP tunnel segmentation.
Pravin B Shelar [Sat, 9 Jul 2016 01:18:39 +0000 (18:18 -0700)]
datapath: compat: Add support for IPv6 UDP tunnel segmentation.

Next patch adds support for IPV6 Geneve and VXLAN, But support for UDP
segmentation is available on all supported kernel.
Following patch adds support for UDP tunnels over IPv6 for such kernels.

Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: compat: Prepare tnl-segmentation for ipv6.
Pravin B Shelar [Fri, 8 Jul 2016 04:52:34 +0000 (21:52 -0700)]
datapath: compat: Prepare tnl-segmentation for ipv6.

Current tnl_skb_gso_segment() is written for ipv4 tunnels. Following
patch make it generic so that it can also handle ipv6 tunnels.

Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: backport: ip_tunnel: Move stats update to iptunnel_xmit()
Pravin B Shelar [Fri, 8 Jul 2016 04:52:29 +0000 (21:52 -0700)]
datapath: backport: ip_tunnel: Move stats update to iptunnel_xmit()

Upstream commit:
    commit 039f50629b7f860f36644ed1f34b27da9aa62f43
    Author: Pravin B Shelar <pshelar@ovn.org>

    ip_tunnel: Move stats update to iptunnel_xmit()

    By moving stats update into iptunnel_xmit(), we can simplify
    iptunnel_xmit() usage. With this change there is no need to
    call another function (iptunnel_xmit_stats()) to update stats
    in tunnel xmit code path.

Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: backport: udp: Add socket based GRO and config
Pravin B Shelar [Fri, 8 Jul 2016 04:52:24 +0000 (21:52 -0700)]
datapath: backport: udp: Add socket based GRO and config

Upstream commit:
    commit 38fd2af24fcfda93f9fea3e53f26e48775ae9e09
    Author: Tom Herbert <tom@herbertland.com>

    udp: Add socket based GRO and config

    Add gro_receive and  gro_complete to struct udp_tunnel_sock_cfg.

Signed-off-by: Tom Herbert <tom@herbertland.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: compat: Update IPv6 in setup_udp_tunnel_sock()
Pravin B Shelar [Fri, 8 Jul 2016 04:52:20 +0000 (21:52 -0700)]
datapath: compat: Update IPv6 in setup_udp_tunnel_sock()

Update setup_udp_tunnel_sock() to handle ipv6 sockets.

Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: compat: Update udp_sock_create
Pravin B Shelar [Fri, 8 Jul 2016 04:51:17 +0000 (21:51 -0700)]
datapath: compat: Update udp_sock_create

Update udp-socket-create to create ipv6 socket currectly.

Partially backports commit fd384412e199b ("udp_tunnel: Seperate ipv6
functions into its own file.")

Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: compat: rename HAVE_METADATA_DST to USE_UPSTREAM_TUNNEL
Pravin B Shelar [Fri, 8 Jul 2016 04:49:20 +0000 (21:49 -0700)]
datapath: compat: rename HAVE_METADATA_DST to USE_UPSTREAM_TUNNEL

To better represent the meaning of symbol.

Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: backport: ip_tunnel: add support for setting flow label via collect metadata
Pravin B Shelar [Fri, 8 Jul 2016 04:49:10 +0000 (21:49 -0700)]
datapath: backport: ip_tunnel: add support for setting flow label via collect metadata

Update udp_tunnel6_xmit_skb(). Specificaly changes are
related to setting ipv6 label.

Upstream commit:
    commit 134611446dc657e1bbc73ca0e4e6b599df687db0
    Author: Daniel Borkmann <daniel@iogearbox.net>

    ip_tunnel: add support for setting flow label via collect metadata

    This patch extends udp_tunnel6_xmit_skb() to pass in the IPv6 flow label
    from call sites. Currently, there's no such option and it's always set to
    zero when writing ip6_flow_hdr(). Add a label member to ip_tunnel_key, so
    that flow-based tunnels via collect metadata frontends can make use of it.
    vxlan and geneve will be converted to add flow label support separately.

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: compat: Remove unnecessary iptunnel_xmit() declaration.
Pravin B Shelar [Fri, 8 Jul 2016 04:49:07 +0000 (21:49 -0700)]
datapath: compat: Remove unnecessary iptunnel_xmit() declaration.

Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: backport: tunnel: introduce ipv6_tun_rx_dst()
Pravin B Shelar [Fri, 8 Jul 2016 04:49:02 +0000 (21:49 -0700)]
datapath: backport: tunnel: introduce ipv6_tun_rx_dst()

Update ovs_udp_tun_rx_dst() to handle ipv6 tunnels.

This commit partially backports c29a70d2c ("tunnel: introduce udp_tun_rx_dst()")

Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: backport: ip_tunnel_core: iptunnel_handle_offloads returns int and doesn...
Pravin B Shelar [Fri, 8 Jul 2016 04:48:57 +0000 (21:48 -0700)]
datapath: backport: ip_tunnel_core: iptunnel_handle_offloads returns int and doesn't free skb

There is return type change in upstream handle-offload functions.
Following patch brings these changes in.
This is backport of aed069df ("ip_tunnel_core:
iptunnel_handle_offloads returns int and doesn't free skb")
I have also removed duplicate definitions of tunnel_handle_offloads()
from ip-tunnel header.

Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: backport: net: add dst_cache support
Pravin B Shelar [Fri, 8 Jul 2016 04:46:19 +0000 (21:46 -0700)]
datapath: backport: net: add dst_cache support

This backports dst-cache implementation from upstream implementation.

    commit 911362c70df5b766c243dc297fadeaced786ffd8
    Author: Paolo Abeni <pabeni@redhat.com>

    net: add dst_cache support
    This patch add a generic, lockless dst cache implementation.
    The need for lock is avoided updating the dst cache fields
    only in per cpu scope, and requiring that the cache manipulation
    functions are invoked with the local bh disabled.

    The refresh_ts and reset_ts fields are used to ensure the cache
    consistency in case of cuncurrent cache update (dst_cache_set*) and
    reset operation (dst_cache_reset).

    Consider the following scenario:

    CPU1:                                       CPU2:
      <cache lookup with emtpy cache: it fails>
      <get dst via uncached route lookup>
                                                <related configuration changes>
                                                dst_cache_reset()
      dst_cache_set()

    The dst entry set passed to dst_cache_set() should not be used
    for later dst cache lookup, because it's obtained using old
    configuration values.

    Since the refresh_ts is updated only on dst_cache lookup, the
    cached value in the above scenario will be discarded on the next
    lookup.

Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Suggested-and-acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: backport: tunnels: Remove encapsulation offloads on decap.
Pravin B Shelar [Fri, 8 Jul 2016 04:46:11 +0000 (21:46 -0700)]
datapath: backport: tunnels: Remove encapsulation offloads on decap.

Following patch backports updated iptunnel pull function.
Also brings in following upstream fix:

    commit a09a4c8dd1ec7f830e1fb9e59eb72bddc965d168
    Author: Jesse Gross <jesse@kernel.org>

    tunnels: Remove encapsulation offloads on decap.

    If a packet is either locally encapsulated or processed through GRO
    it is marked with the offloads that it requires. However, when it is
    decapsulated these tunnel offload indications are not removed. This
    means that if we receive an encapsulated TCP packet, aggregate it with
    GRO, decapsulate, and retransmit the resulting frame on a NIC that does
    not support encapsulation, we won't be able to take advantage of hardware
    offloads even though it is just a simple TCP packet at this point.

    This fixes the problem by stripping off encapsulation offload indications
    when packets are decapsulated.

    The performance impacts of this bug are significant. In a test where a
    Geneve encapsulated TCP stream is sent to a hypervisor, GRO'ed, decapsulated,
    and bridged to a VM performance is improved by 60% (5Gbps->8Gbps) as a
    result of avoiding unnecessary segmentation at the VM tap interface.

Reported-by: Ramu Ramamurthy <sramamur@linux.vnet.ibm.com>
    Fixes: 68c33163 ("v4 GRE: Add TCP segmentation offload for GRE")
Signed-off-by: Jesse Gross <jesse@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: backport: iptunnel: scrub packet in iptunnel_pull_header
Pravin B Shelar [Fri, 8 Jul 2016 04:46:01 +0000 (21:46 -0700)]
datapath: backport: iptunnel: scrub packet in iptunnel_pull_header

Upstream Commit:
    commit 7f290c94352e59b1d720055fce760a69a63bd0a1
    Author: Jiri Benc <jbenc@redhat.com>

    iptunnel: scrub packet in iptunnel_pull_header

    Part of skb_scrub_packet was open coded in iptunnel_pull_header. Let it call
    skb_scrub_packet directly instead.

Signed-off-by: Jiri Benc <jbenc@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodatapath: compat: Refactor egress tunnel info
Pravin B Shelar [Fri, 8 Jul 2016 02:35:33 +0000 (19:35 -0700)]
datapath: compat: Refactor egress tunnel info

upstream tunnel egress info is retrieved using ndo_fill_metadata_dst.
Since we do not have it on older kernel we need to keep vport operation
to do same on these kernels.
Following patch try to merge these to operations into one to avoid code
duplication.
This commit backports fc4099f1 ("openvswitch:
Fix egress tunnel info.")

Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
7 years agodpif-netdev: Remove PMD latency on seq_mutex
Flavio Leitner [Tue, 5 Jul 2016 13:33:38 +0000 (10:33 -0300)]
dpif-netdev: Remove PMD latency on seq_mutex

The PMD thread needs to keep processing RX queues in order
to achieve maximum throughput. It also needs to sweep emc
cache and quiesce which use seq_mutex. That mutex can
eventually block the PMD thread causing latency spikes and
affecting the throughput.

Since there is no requirement for running those tasks at a
specific time, this patch extend seq API to allow tentative
locking instead.

Reported-by: Karl Rister <krister@redhat.com>
Co-authored-by: Karl Rister <krister@redhat.com>
Signed-off-by: Flavio Leitner <fbl@redhat.com>
Signed-off-by: Daniele Di Proietto <diproiettod@vmware.com>
7 years agonetdev-dummy: Add n_txq option.
Ilya Maximets [Fri, 8 Jul 2016 13:52:38 +0000 (16:52 +0300)]
netdev-dummy: Add n_txq option.

Will be used for testing with different numbers of TX queues.

Signed-off-by: Ilya Maximets <i.maximets@samsung.com>
Acked-by: Daniele Di Proietto <diproiettod@vmware.com>
7 years agonetdev-dpdk: Obtain number of queues for vhost ports from attached virtio.
Ilya Maximets [Fri, 8 Jul 2016 13:52:37 +0000 (16:52 +0300)]
netdev-dpdk: Obtain number of queues for vhost ports from attached virtio.

Currently, there are few inconsistencies in ways to configure number of
queues for netdev device:

* dpif-netdev can't know about exact number of queues
  allocated inside netdev.
  This leads to constant mapping of queue-ids to 'real' ones.

* We are able to configure 'n_rxq' for vhost-user devices, but
  there is only one sane number of rx queues which must be used
  and configured manually (number of queues that allocated
  in QEMU).

This patch disables configuration of 'n_rxq' for DPDK vHost devices.
Configuration of rx and tx queues now automatically applied from
connected virtio device. Standard reconfiguration mechanism was used to
apply this changes.

Also, now 'n_txq' and 'n_rxq' are always the real numbers of queues
in the device.

Signed-off-by: Ilya Maximets <i.maximets@samsung.com>
Signed-off-by: Daniele Di Proietto <diproiettod@vmware.com>
7 years agoovn: Support l2gateway-chassis option in "l2gateway" logical ports
Numan Siddique [Fri, 8 Jul 2016 11:37:37 +0000 (17:07 +0530)]
ovn: Support l2gateway-chassis option in "l2gateway" logical ports

ovn-controller will now bind the l2gateway logical ports.

Signed-Off-by: Numan Siddique <nusiddiq@redhat.com>
Signed-off-by: Russell Bryant <russell@ovn.org>
7 years agoovn test: Skip "send gratuitous arp on localnet" if no Python installed
Paul Boca [Mon, 4 Jul 2016 10:00:47 +0000 (10:00 +0000)]
ovn test: Skip "send gratuitous arp on localnet" if no Python installed

This test needs Python in order to run.

Signed-off-by: Paul-Daniel Boca <pboca@cloudbasesolutions.com>
Acked-by: Alin Gabriel Serdean <aserdean@cloudbasesolutions.com>
Signed-off-by: Russell Bryant <russell@ovn.org>
7 years agoINSTALL.Docker.md: Clarify OVS python library path.
Gurucharan Shetty [Thu, 7 Jul 2016 08:28:54 +0000 (01:28 -0700)]
INSTALL.Docker.md: Clarify OVS python library path.

Reported-by: Taekho Nam <thnam@smartx.kr>
Signed-off-by: Gurucharan Shetty <guru@ovn.org>
Acked-by: Russell Bryant <russell@ovn.org>
7 years agoovn-sbctl: Change lport-(un)bind to lsp-(un)bind.
Russell Bryant [Fri, 24 Jun 2016 20:36:36 +0000 (16:36 -0400)]
ovn-sbctl: Change lport-(un)bind to lsp-(un)bind.

A previous commit changed the command names in ovn-nbctl from lport-* to
lsp-*.  Change lport-bind and lport-unbind in ovn-sbctl to match.

Signed-off-by: Russell Bryant <russell@ovn.org>
Acked-by: Amitabha Biswas <abiswas@us.ibm.com>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agoovn: improve OVN tutorial.
nickcooper-zhangtonghao [Mon, 4 Jul 2016 15:03:49 +0000 (08:03 -0700)]
ovn: improve OVN tutorial.

Improve the tutorial of the basic OVN features. The addresses and port_security
columns of the logical port is described in more detail.

Signed-off-by: nickcooper-zhangtonghao <nickcooper-zhangtonghao@opencloud.tech>
Signed-off-by: Russell Bryant <russell@ovn.org>
7 years agoovn-controller: Remove old address set after change.
Ryan Moats [Thu, 7 Jul 2016 18:37:04 +0000 (13:37 -0500)]
ovn-controller: Remove old address set after change.

Currently, when address set value changes, ovn controller
doesn't remove the old entry from the tracking hash, it
just adds the new one, leading to multiple entries for the
same symbol.

Fix this behavior and add a smoke test to avoid a regression
in the future.

Signed-off-by: Ryan Moats <rmoats@us.ibm.com>
Acked-by: Flavio Fernandes <flavio@flaviof.com>
Signed-off-by: Russell Bryant <russell@ovn.org>
7 years agoMakefile.am: Add INSTALL.DPDK-ADVANCED.md to EXTRA_DIST.
Daniele Di Proietto [Wed, 6 Jul 2016 23:07:56 +0000 (16:07 -0700)]
Makefile.am: Add INSTALL.DPDK-ADVANCED.md to EXTRA_DIST.

Fixes: c9b9d6dfc06c("INSTALL.DPDK: Refactor DPDK install guide, add
ADVANCED doc")

Signed-off-by: Daniele Di Proietto <diproiettod@vmware.com>
7 years agonetdev-linux: Do not log a warning if the device is down.
Daniele Di Proietto [Thu, 7 Jul 2016 01:09:12 +0000 (18:09 -0700)]
netdev-linux: Do not log a warning if the device is down.

In the userspace datapath we use tap devices as internal netdev.  The
datapath doesn't consider whether a device is up or down before sending
to it, and so far this hasn't been a problem.

Since Linux upstream commit 1bd4978a88ac("tun: honor IFF_UP in
tun_get_user()"), included in 4.4, writing to a tap device that is not
up sets errno to EIO.  This commit avoids printing a warning in this
case.

This fixes a failures in the system-userspace-testsuites.

Reported-by: Joe Stringer <joe@ovn.org>
Signed-off-by: Daniele Di Proietto <diproiettod@vmware.com>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agovlog.py: Remove redundant setLevel() if "/dev/log" doesn't exist.
Daniele Di Proietto [Wed, 6 Jul 2016 23:39:15 +0000 (16:39 -0700)]
vlog.py: Remove redundant setLevel() if "/dev/log" doesn't exist.

Also update a comment.

Signed-off-by: Daniele Di Proietto <diproiettod@vmware.com>
Acked-by: Gurucharan Shetty <guru@ovn.org>
7 years agonetdev-dpdk: Use instant sending instead of queueing of packets.
Ilya Maximets [Mon, 27 Jun 2016 13:28:16 +0000 (16:28 +0300)]
netdev-dpdk: Use instant sending instead of queueing of packets.

Current implementarion of TX packet's queueing is broken in several ways:

* TX queue flushing implemented on receive assumes that all
  core_id-s are sequential and starts from zero. This may lead
  to situation when packets will stuck in queue forever and,
  also, this influences on latency.

* For a long time flushing logic depends on uninitialized
  'txq_needs_locking', because it usually calculated after
  'netdev_dpdk_alloc_txq' but used inside of this function
  for initialization of 'flush_tx'.

Testing shows no performance difference with and without queueing.
Lets remove queueing at all because it doesn't work properly now and
also does not increase performance.

Signed-off-by: Ilya Maximets <i.maximets@samsung.com>
Acked-by: Daniele Di Proietto <diproiettod@vmware.com>
7 years agoINSTALL.DPDK: Refactor DPDK install guide, add ADVANCED doc
Bhanuprakash Bodireddy [Wed, 6 Jul 2016 13:06:39 +0000 (14:06 +0100)]
INSTALL.DPDK: Refactor DPDK install guide, add ADVANCED doc

Add INSTALL.DPDK-ADVANCED document that is forked off from original
INSTALL.DPDK guide. This document is targeted at users looking for
optimum performance on OVS using dpdk datapath.

Signed-off-by: Bhanuprakash Bodireddy <bhanuprakash.bodireddy@intel.com>
Acked-by: Flavio Leitner <fbl@sysclose.org>
Acked-by: Daniele Di Proietto <diproiettod@vmware.com>
7 years agoINSTALL.DPDK: Refactor DPDK install documentation
Bhanuprakash Bodireddy [Wed, 6 Jul 2016 13:06:38 +0000 (14:06 +0100)]
INSTALL.DPDK: Refactor DPDK install documentation

Refactor the INSTALL.DPDK in to two documents named INSTALL.DPDK and
INSTALL.DPDK-ADVANCED. While INSTALL.DPDK document shall facilitate the
novice user in setting up the OVS DPDK and running it out of box, the
ADVANCED document is targeted at expert users looking for the optimum
performance running dpdk datapath.

This commit updates INSTALL.DPDK.md document.

Signed-off-by: Bhanuprakash Bodireddy <bhanuprakash.bodireddy@intel.com>
Acked-by: Flavio Leitner <fbl@sysclose.org>
Acked-by: Daniele Di Proietto <diproiettod@vmware.com>
7 years agovlog test: Disable default syslog logger
Paul Boca [Wed, 6 Jul 2016 12:38:32 +0000 (12:38 +0000)]
vlog test: Disable default syslog logger

Disable the syslog logger in case on Windows, '/dev/log' doesn't exist.
Seems like on Python34 a default handler is added to the logger and it prints
even if no handler is set by us.

Signed-off-by: Paul-Daniel Boca <pboca@cloudbasesolutions.com>
Acked-by: Daniele Di Proietto <diproiettod@vmware.com>
7 years agobridge: open_type should be used for netdev_open
Thadeu Lima de Souza Cascardo [Mon, 4 Jul 2016 18:19:56 +0000 (15:19 -0300)]
bridge: open_type should be used for netdev_open

ofproto_port_open_type should be used for netdev_open, but not for other tests.
For example, STP/RSTP check for interfaces of internal type, but that check will
fail when the netdev datapath is used.

The same thing goes for setting MAC address of internal Interfaces. That fails
for the netdev datapath because the interface type is set to "tap", but they are
still interfaces of type "internal", just their netdev implementation is
different.

Use a netdev_type for the type that needs to be used for netdev_open and
ofproto_port, while we still keep the type as the normalized configured type in
the database.

Signed-off-by: Thadeu Lima de Souza Cascardo <cascardo@redhat.com>
Signed-off-by: Jesse Gross <jesse@kernel.org>
7 years agoovs-ofctl: Clarify CT action documentation.
Jarno Rajahalme [Wed, 6 Jul 2016 07:47:20 +0000 (00:47 -0700)]
ovs-ofctl: Clarify CT action documentation.

Since the 'commit' flag is required to set the mark and/or labels, the
set values are always available for following lookups.

Signed-off-by: Jarno Rajahalme <jarno@ovn.org>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agoovn.at: Use = instead of == for test
YAMAMOTO Takashi [Tue, 5 Jul 2016 08:04:47 +0000 (08:04 +0000)]
ovn.at: Use = instead of == for test

== is a GNU extension which might not be available.

Signed-off-by: YAMAMOTO Takashi <yamamoto@ovn.org>
Acked-By: Ryan Moats <rmoats@us.ibm.com>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agolib: Remove extra API dependency for ovs_thread_create()
Andy Zhou [Fri, 17 Jun 2016 22:41:26 +0000 (15:41 -0700)]
lib: Remove extra API dependency for ovs_thread_create()

When calling ovs_thread_create() without calling fatal_signal_init()
first, ovs_thread_create() some times asserts. This dependency is
subtle and not very obvious.

The root cause seems to be that, within ovs_thread_create(), the
multi-threaded state is declared before all initializations are done.

Signed-off-by: Andy Zhou <azhou@ovn.org>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agonetlink-notifier: Avoid valgrind possible leak warning.
Ben Pfaff [Tue, 5 Jul 2016 15:33:05 +0000 (08:33 -0700)]
netlink-notifier: Avoid valgrind possible leak warning.

This ensures that pointers to nln_notifiers are to the beginning of the
structs instead of to the middle, meaning that valgrind does not consider
them "possible" leaks.

Reported-by: William Tu <u9012063@gmail.com>
Tested-by: William Tu <u9012063@gmail.com>
Signed-off-by: Ben Pfaff <blp@ovn.org>
7 years agobridge: Add assertion to document an invariant in find_local_hw_addr().
Ben Pfaff [Sun, 3 Jul 2016 04:16:55 +0000 (21:16 -0700)]
bridge: Add assertion to document an invariant in find_local_hw_addr().

Avoids a possible null pointer dereference report from Clang.

Reported-at: http://openvswitch.org/pipermail/dev/2016-June/073967.html
Signed-off-by: Ben Pfaff <blp@ovn.org>
Tested-by: William Tu <u9012063@gmail.com>
7 years agoovn: Add support for Load balancers.
Gurucharan Shetty [Wed, 15 Jun 2016 13:24:30 +0000 (06:24 -0700)]
ovn: Add support for Load balancers.

This commit adds schema changes to the OVN_Northbound database to support
Load balancers.

In ovn-northd, it adds two logical tables to program logical flows.
It adds a 'pre_lb' table that sits before 'pre_stateful' table.
For packets that need to be load balanced, this table sets reg0[0]
to act as a hint for the pre-stateful table to send the packet to
the conntrack table for defragmentation.

It also adds a 'lb' table that sits before 'stateful' table.
For packets from established connections, this table sets reg0[2] to
indicate to the 'stateful' table that the packet needs to be sent to
connection tracking table to just do NAT.

In stateful table, packet for a new connection that needs to be load balanced
is given a ct_lb($IP_LIST) action.

Signed-off-by: Gurucharan Shetty <guru@ovn.org>
Acked-by: Ben Pfaff <blp@ovn.org>
7 years agoovn-controller: Add support for load balancing.
Gurucharan Shetty [Sun, 3 Jul 2016 12:31:37 +0000 (05:31 -0700)]
ovn-controller: Add support for load balancing.

ovn-controller now supports 2 new logical actions.

1. ct_lb;
Sends the packet through the conntrack zone to NAT
packets. Packets that are part of established connection
will automatically get NATed based on the NAT arguments
supplied to conntrack when the first packet was committed.

2. ct_lb(192.168.1.2, 192.168.1.3);
   ct_lb(192.168.1.2:80, 192.168.1.3:80);
Creates an OpenFlow group with multiple buckets and equal weights
that changes the destination IP address (and port number) of the packet
statefully to one of the options provided inside the parenthesis.

Signed-off-by: Gurucharan Shetty <guru@ovn.org>
Signed-off-by: Ben Pfaff <blp@ovn.org>