Ben Pfaff [Mon, 10 Jan 2011 00:57:45 +0000 (16:57 -0800)]
netlink-socket: Add functions for joining and leaving multicast groups.
When this library was originally implemented, support for Linux 2.4 was
important. The Netlink implementation in Linux only added support for
joining and leaving multicast groups after a socket is bound as of Linux
2.6.14, so the library did not support it either. But the current version
of Open vSwitch targets Linux 2.6.18 and over, so it's fine to add this
support now, and this commit does so.
This will be used more extensively in upcoming commits.
Ben Pfaff [Wed, 19 Jan 2011 21:17:22 +0000 (13:17 -0800)]
netlink: Add nl_attr_type constants for big-endian values.
These are semantically identical, so they might as well share existing
values, but they help to document the endianness of attributes in Netlink
attribute policies.
Ben Pfaff [Thu, 13 Jan 2011 20:36:06 +0000 (12:36 -0800)]
netlink: New functions for finding a specific attribute.
These functions are useful in the occasional case where a piece of code
only cares about one or a few attributes, probably knows that the format
is correct, and doesn't want to go to the trouble of doing a full parse.
Upcoming commits will add a user.
Ben Pfaff [Mon, 27 Dec 2010 22:32:45 +0000 (14:32 -0800)]
ofpbuf: Enable ofpbuf_push() to expand headroom.
Until now, ofpbuf_put() has been able to reallocate an ofpbuf to
expand the tailroom, but ofpbuf_push() has not been able to expand
the headroom in an analogous way. This omission is simply because
it has never been useful to do this. However, an upcoming change
will make the amount of headroom in an ofpbuf passed up from a
dpif to ofproto harder to predict. It seems that we might as well
simply implement expanding headroom, so this commit does that.
Ben Pfaff [Tue, 25 Jan 2011 23:23:31 +0000 (15:23 -0800)]
stream-ssl: Only cache SSL sessions after they shut down.
A cached SSL session may only be used for new connections after the initial
connection has shut down. As far as I can tell, nothing in the OpenSSL
documentation actually comes out and says this, but it is implied by
various examples found around the web and doing it this way makes caching
work much more reliably in my testing.
Ben Pfaff [Tue, 25 Jan 2011 23:17:32 +0000 (15:17 -0800)]
stream-ssl: Set a session context ID string on our SSL_CTX.
In the finest OpenSSL tradition of putting important documentation only in
code comments, ssl/ssl_sess.c in the OpenSSL tree has the following comment
inside ssl_get_prev_session():
/* We can't be sure if this session is being used out of
* context, which is especially important for SSL_VERIFY_PEER.
* The application should have used SSL[_CTX]_set_session_id_context.
*
* For this error case, we generate an error instead of treating
* the event like a cache miss (otherwise it would be easy for
* applications to effectively disable the session cache by
* accident without anyone noticing).
*/
This meant that ovs-controller couldn't effectively cache SSL server
sessions and we got a weird error whenever ovs-vswitchd tried.
Bug #4448. CC: David Tsai <dtsai@nicira.com> CC: Jeremy Stribling <strib@nicira.com>
Ethan Jackson [Tue, 25 Jan 2011 18:58:47 +0000 (10:58 -0800)]
ovs-bugtool: Indiscriminately capture all openvswitch log files.
ovs-bugtool was only capturing openvswitch core files. This commit
causes it to capture all files under /var/log/openvswitch and
/var/log/core. It also moves the ovs-bugtool output to
/var/log/ovs-bugtool.
Reported-by: Ben Pfaff <blp@nicira.com>
Bug #4481.
Ben Pfaff [Mon, 24 Jan 2011 22:03:57 +0000 (14:03 -0800)]
datapath: Move rcu_read_lock_held(), rcu_dereference_*() to rcupdate.h.
<linux/rcupdate.h> is the normal location of these functions, so we should
not put compat code for them in <linux/rtnetlink.h>.
This caused build failures for datapath/table.c, which in later commits
uses rcu_dereference_check() and rcu_dereference_protected() but does not
need RTNL.
Signed-off-by: Ben Pfaff <blp@nicira.com> Acked-by: Jesse Gross <jesse@nicira.com>
Ethan Jackson [Mon, 24 Jan 2011 18:54:17 +0000 (10:54 -0800)]
xenserver: Change all bridge's fail_mode when updating.
The ovs plugin's update command was only setting fail_mode for
bridges attached to pifs. This patch applies the fail_mode
setting to all bridges on the XenServer system.
Reported-by: David Tsai <dtsai@nicira.com> Signed-off-by: Ethan Jackson <ethan@nicira.com>
Ben Pfaff [Sat, 22 Jan 2011 01:09:21 +0000 (17:09 -0800)]
multipath: Fix off-by-one in hash_threshold multipath calculation.
0xffffffff / (0xffffffff / n) can have a value as large as n (consider the
n == 1 case), but we need a value no bigger than n-1. So add 1 before
dividing to fix the problem.
This caused a test failure on Debian "lenny" amd64 when apparently
unrelated code changed.
Ben Pfaff [Fri, 21 Jan 2011 22:49:01 +0000 (14:49 -0800)]
ofproto: Correctly compute length of flow stats in multipart flow stats.
In a flow table dump, the length of the first nx_flow_stats in the second
and subsequent OpenFlow messages was being miscalculated, because
append_nxstats_reply() changed the length of the message in such cases.
So we need to load the starting length after calling append_nxstats_reply()
instead of before.
Ben Pfaff [Fri, 21 Jan 2011 00:34:15 +0000 (16:34 -0800)]
datapath: Remove vport_del_all() because it is now a no-op.
vport_del_all() was created when vports could exist without being attached
to any datapath. Now, a vport is always attached to a datapath. This
function was only called on module unload, but the module can't be unloaded
if any datapath exists, so it won't ever have any work to do, and we might
as well delete it entirely.
Justin Pettit [Tue, 18 Jan 2011 08:01:24 +0000 (00:01 -0800)]
ovs-monitor-ipsec: Use "require" when adding SPD entries.
Previously, SPD entries were added with a level of "default". This uses
the system-wide default for the protocol when processing a matching
packet. Switch the level to "require" so that a SA is always used when
sending the packet.
Ethan Jackson [Wed, 19 Jan 2011 22:46:57 +0000 (14:46 -0800)]
xenserver: Allow NULL pool in configuration cache.
InterfaceReconfigure assumes the pool is in the configuration
cache. This is always true except when upgrading from an older
version of openvswitch which does record the pool. This can cause
upgrades to lock up.
Ethan Jackson [Mon, 17 Jan 2011 19:31:46 +0000 (11:31 -0800)]
xenserver: Allow unknown other-config attributes.
When saving the database cache, InterfaceReconfigure can crash if
unexpected attributes are in an object's other-config column. This
commit causes it to skip that attribute and log a warning.
Signed-off-by: Ethan Jackson <ethan@nicira.com> CC: Ian Campbell <Ian.Campbell@citrix.com> CC: Dominic Curran <Dominic.Curran@citrix.com>
Ethan Jackson [Fri, 14 Jan 2011 00:29:31 +0000 (16:29 -0800)]
route-table: Handle route updates more robustly.
The kernel does not broadcast rtnetlink route messages in all cases
one would expect. This can cause stale entires to end up in the
route table which may cause incorrect results for
route_table_get_ifindex() queries. This commit causes rtnetlink
route messages to dump the entire route table on the next
route_table_get_ifindex() query.
Ethan Jackson [Wed, 12 Jan 2011 22:55:18 +0000 (14:55 -0800)]
lib: Simplify rtnetlink routing functionality.
This commit removes the rtnetlink-route module and replaces it with
a much simpler to use route-table module. The route-table uses
rtnetlink to maintain a routing table which may be used to query
the egress interface of particular addresses.
This commit also converts netdev-vport to use the new route-table
module.
Ben Pfaff [Wed, 12 Jan 2011 18:44:43 +0000 (10:44 -0800)]
datapath: Get rid of compat.h, compat26.h in favor of modern approach.
I had completely forgotten that we had a top-level compat.h and compat26.h.
It's better to distribute their contents to individual compat headers, so
this commit does so and deletes them.
Signed-off-by: Ben Pfaff <blp@nicira.com> Acked-by: Jesse Gross <jesse@nicira.com>
Sajjad Lateef [Thu, 13 Jan 2011 01:07:58 +0000 (17:07 -0800)]
debian: Modify ovs-bugtool to include collectd rrd logs
collectd is a utility that stores rrd logs in the directory
/var/lib/collectd/rrd/ by default. rrd logs are created when
collectd is configured to output logs in Round Robin Database
format.
With this change, ovs-bugtool will include collectd logs in the
output bundle from /var/lib/collectd/rrd. The output bundle will
include all sub-directories of the specified collectd directory.
ovs-bugtool checks if a directory exists before it attempts to
include the directory contents in the output bundle. So, running
ovs-bugtool on a system that does not have collectd logs should
not negatively affect the behavior of ovs-bugtool.
Ben Pfaff [Wed, 12 Jan 2011 21:57:53 +0000 (13:57 -0800)]
ofp-print: Add tests for OpenFlow message formatting.
It's become clear over the past few weeks that some people really depend
on the formatting of ofp-print output. It has also become clear that OVS
doesn't have anything that actually checks or enforces the formatting of
the output. The former is not something I want to encourage, but the
latter is somewhat worrisome because the formatting does change from time
to time and I want to make sure that that is intentional.
So this commit adds at least one test for each supported OpenFlow and
Nicira extension message, so that we can better keep track of changes.
Ethan Jackson [Wed, 5 Jan 2011 19:51:15 +0000 (11:51 -0800)]
bridge: Move tunnel_egress_iface to status column.
This commit removes the tunnel_egress_iface column from the
interface table and moves it's data to the status column. In the
process it reverts the database to version 1.0.0.
Justin Pettit [Tue, 11 Jan 2011 05:32:09 +0000 (21:32 -0800)]
xenserver: Remove "Required-" parameters from "openvswitch" init script.
Commit 72dfb3b (xenserver: Add LSB init info to scripts.) added LSB init
info to the XenServer scripts. This was verified to work on XenServer
5.5, but newer versions of XenServer interpret these values and alter
when "openvswitch" gets called. This causes XenServer to not come up
with its networking configuration properly.
By removing the "Required-" parameters, OVS will properly come up. This
is just a temporary fix, and a more proper one will be forthcoming.
Ben Pfaff [Mon, 10 Jan 2011 20:54:26 +0000 (12:54 -0800)]
tests: Fix Y2011 bug in testsuite.
The tests have been failing for a few days now, because the PKI expired a
few days into 2011. This commit instead generates the PKI at "make check"
time, which has the additional benefit of getting some test exposure for
the ovs-pki program.
Reported-by: Aaron M. Ucko <ucko@debian.org> CC: 609506@bugs.debian.org
Patrick Mullaney [Wed, 15 Dec 2010 19:07:10 +0000 (11:07 -0800)]
xenserver: Remove Linux bridge on startup if no bridges exist.
Otherwise, if the bridge module is loaded but no bridges exist, on all
but the most recent Linux kernels this prevents the Open vSwitch kernel
module from loading.
Patrick Mullaney [Wed, 15 Dec 2010 18:39:08 +0000 (10:39 -0800)]
xenserver: Don't require /etc/xensource/network.conf if not on XenServer.
On XenServer systems, the "network mode" (that is, whether the network
uses OpenvSwitch or the Linux bridge) is stored in
/etc/xensource/network.conf. On other systems, we can't expect it to be
there. This changes the init script to tolerate its absence, treating that
as meaning that Open vSwitch should be used (otherwise, why did the admin
install it?).
NETWORK_MODE can always be overridden in /etc/sysconfig/openvswitch anyway.
Signed-off-by: Patrick Mullaney <pm.mullaney@gmail.com>
[changed back to read /etc/xensource/network.conf if present] Signed-off-by: Ben Pfaff <blp@nicira.com>
Ben Pfaff [Wed, 5 Jan 2011 19:08:25 +0000 (11:08 -0800)]
xenserver: Generate system UUID for non-XenServer systems.
Currently the scripts in xensever/ are intended specifically for XenServer,
but supporting other distros seems like a worthy goal, so this is a step
in the right direction.
It would be good to be able to determine a suitable system-type and
system-version, but those are less important than system-id.
Signed-off-by: Patrick Mullaney <pm.mullaney@gmail.com>
[changed not to set xs-system-uuid on non-XenServer systems]
[whitespace changes]
[changed to handle missing uuidgen, suggested by Justin Pettit] Signed-off-by: Ben Pfaff <blp@nicira.com>
Ben Pfaff [Wed, 5 Jan 2011 19:00:41 +0000 (11:00 -0800)]
xenserver: Be compatible with non-XenServer func libraries in init script.
XenServer has its function library in /etc/init.d/functions but other
distros have it in different places. Currently this init script is
specifically intended for XenServer but adding compatibility with other
distros seems like a worthwhile goal.
Also, SuSE does not have the "action" function, so test whether it is
implemented and when not supply a fallback.
Signed-off-by: Patrick Mullaney <pm.mullaney@gmail.com>
[changed to only define action() if not already defined] Signed-off-by: Ben Pfaff <blp@nicira.com>
Ben Pfaff [Wed, 5 Jan 2011 18:52:36 +0000 (10:52 -0800)]
xenserver: Add LSB init info to scripts.
XenServer itself does not use this information as part of startup, but
other distributions that may wish to derive their startup scripts from
these may need it.
Signed-off-by: Patrick Mullaney <pm.mullaney@gmail.com>
[updated some info in xenserver/etc_init.d_openvswitch-xapi-update]
[used Debian LSB info for xenserver/etc_init.d/openvswitch] Signed-off-by: Ben Pfaff <blp@nicira.com>
Ben Pfaff [Fri, 7 Jan 2011 17:35:19 +0000 (09:35 -0800)]
debian: Update check for Debian package version.
A version number like "1.1.0pre2" is properly translated into a Debian
version number as "1.1.0~pre2", so we need to make that translation. Also,
the Debian version number will often have suffixes on it, so don't check
for a trailing parenthesis.
Both tunnel.c and vport-patch.c have set_config() functions. They are
static so this does not confuse the compiler, but it does (slightly)
confuse me from time to time, so this commit changes their names to
tnl_set_config() and patch_set_config(), respectively.
Signed-off-by: Ben Pfaff <blp@nicira.com> Acked-by: Jesse Gross <jesse@nicira.com>
Ben Pfaff [Tue, 4 Jan 2011 20:36:07 +0000 (12:36 -0800)]
datapath: Improve vport-patch names that are too generic.
"device_config" and "devconf" always sound to me like something that would
be part of the kernel's core device layer. Give them the more-specific
names "patch_config" and "patchconf".
Signed-off-by: Ben Pfaff <blp@nicira.com> Acked-by: Jesse Gross <jesse@nicira.com>
Ben Pfaff [Wed, 5 Jan 2011 18:17:17 +0000 (10:17 -0800)]
datapath: Fix vport_get_stats() in !VPORT_F_GEN_STATS case.
When VPORT_F_GEN_STATS was not set, vport_get_stats() would always return
an error (either an error returned by ->get_stats(), otherwise
-EOPNOTSUPP). This fixes the problem.
Found by inspection.
Signed-off-by: Ben Pfaff <blp@nicira.com> Acked-by: Jesse Gross <jesse@nicira.com>
Simon Horman [Thu, 6 Jan 2011 01:26:42 +0000 (10:26 +0900)]
datapath: Treat GSO skbs as if they were fragments
In dp_output_control() UDP GSO skbs are split into fragments which are
passed to userspace. So the resulting flow set-up by the controller
(I am using ovs-vswitchd) is created based on a fragment. This means
that the UDP source and destination port of the flow is zero.
In order for the datapath to match the resulting flow flow_extract() needs
to treat UDP GSO skbs as if they are fragments. That is, set the UDP
source and destination port to 0.
A flow established for a UDP GSO skb with this change won't match any
subsequent non-GSO skbs, they will need to be passed to the controller and
a new flow established. But without this change no UDP GSO skbs will ever
match any flow.
I noticed this while using KVM using virtio with VhostNet and netperf's
UDP_STREAM test. The result was that the test sent ~5Gbit/s but only a
small fraction of that was received by the other side. Much less than the
1Gbit/s available on the physical link between the host (and guest) and the
machine running netserver. 100% of one of the host's CPUs was consumed, 50%
for the host and 50% for the guest. The host consumption was contributed
to largely by ovs-vswitchd.
With this change I get a much nicer result of a fraction under 1Gbit/s sent
and almost all packets ending up at the other end.
Signed-off-by: Simon Horman <horms@verge.net.au> Signed-off-by: Jesse Gross <jesse@nicira.com>
Jesse Gross [Wed, 5 Jan 2011 15:25:41 +0000 (07:25 -0800)]
datapath: Fix double counting of packet stats for Linux devices.
The kernel augments stats for Linux devices that only provide 32-bit stats
with its own internal 64-bit counters. When doing this it takes the error
stats from the device but uses the packet and byte values from its local
counters. However, we were also taking the packet and byte counts from
the device, leading to double counting.
Jesse Gross [Wed, 5 Jan 2011 20:39:57 +0000 (12:39 -0800)]
datapath: Use call_rcu() when deleting a datapath.
When deleting a datapath, we remove all of the vports and then immediately
free the datapath data structures. Since the vports are allowed to use
call_rcu() to free their data, it's possible for them to return immediately
while packet processing is still taking place. This breaks apart the dropping
of references and the freeing of the data using call_rcu() for protection.
This race cannot actually occur in practice since the last port to be
deleted is an internal device, which uses synchronize_rcu() itself
(implicitly through unregister_netdevice()). However, there is no
requirement that it must do this nor should there be.
Reported-by: Ben Pfaff <blp@nicira.com> Signed-off-by: Jesse Gross <jesse@nicira.com> Acked-by: Ben Pfaff <blp@nicira.com>
projects / openvswitch / log