Shan Wei [Fri, 16 Nov 2012 21:21:36 +0000 (13:21 -0800)]
datapath: use this_cpu_ptr per-cpu helper
just use more faster this_cpu_ptr instead of per_cpu_ptr(p, smp_processor_id());
Signed-off-by: Shan Wei <davidshan@tencent.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Jesse Gross <jesse@nicira.com>
Jesse Gross [Fri, 16 Nov 2012 21:20:58 +0000 (13:20 -0800)]
datapath: Backport this_cpu_ptr to kernels < 2.6.33
An upstream commit uses this_cpu_ptr for a small performance benefit,
so this provides an equivalent to kernels that don't have that
function.
Signed-off-by: Jesse Gross <jesse@nicira.com>
Justin Pettit [Fri, 16 Nov 2012 06:30:41 +0000 (22:30 -0800)]
ofproto-dpif: Remove datapath flows of deleted ports.
Commit acf608 (ofproto-dpif: Use a single underlying datapath across
multiple bridges.) causes datapath flows from deleted ports to not be
removed. The issue is that the code that bulk deletes old flows doesn't
know the datapath port number that makes up the datapath flow
definition. This commit keeps track of the datapath port in the facet
for use when the datapath flow eventually needs to be removed.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Thu, 15 Nov 2012 22:42:10 +0000 (14:42 -0800)]
netdev-linux: Don't log vport warnings when kernel datapath not loaded.
The *_get_stats functions call get_stats_via_vport(), which tries to get
information about ports attached to the kernel datapath. When a pure
userspace switch is used (eg, the OVS kernel module isn't loaded), ports
are not attached to a kernel datapath, so warnings get logged. This
commit handles that case and doesn't log a warning. However, if the
kernel datapath is loaded, ports attached to a userspace datapath will
still log a warning.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Thu, 15 Nov 2012 22:27:28 +0000 (14:27 -0800)]
netdev-linux: "Down" tap devices before setting hw addr.
On Linux, it is not possible to set the mac address on "up" tap
interfaces. This commit temporarily "down"s the interface so the
address can be set for the netdev_linux_set_etheraddr() call.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Thu, 15 Nov 2012 21:55:06 +0000 (13:55 -0800)]
bridge: Always "up" internal devices.
The kernel datapath automatically "up"s internal devices, but this
wasn't happening for the userspace datapath. This change has the bridge
module always "up" them.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Thu, 15 Nov 2012 21:53:16 +0000 (13:53 -0800)]
bridge: Introduce iface_is_internal() function.
This will have an additional caller in the future.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Fri, 16 Nov 2012 07:00:39 +0000 (23:00 -0800)]
ofproto-dpif: Query port existence by name to prevent warnings.
The port_destruct() function checks if the port still exists in the
datapath to see if additional cleanup is necessary. It used
dpif_port_query_by_number(), since the datapath port number is readily
available and cheap to use as a lookup handle. Unfortunately, that
function logs a warning message if a port with that number doesn't exist
(because the datapath did remove it), which is confusing. The
dpif_port_query_by_name() function doesn't log such an warning, so use
it instead.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Thu, 15 Nov 2012 21:05:58 +0000 (13:05 -0800)]
dpif: Don't log warning for ENOENT with dpif_port_exists().
The caller wants to know whether 'devname' is attached to 'dpif', and
ENOENT is a legitimate response to that not being the case.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Tue, 13 Nov 2012 02:57:42 +0000 (18:57 -0800)]
ofproto-dpif: Correct in_port on send_packet().
The switch to a single datapath uncovered a long-standing bug in the
send_packet() function. The code had used an (invalid) ingress OpenFlow
port of 0, which happened to get translated to the datapath's local
port, which was always 0. With the single datapath, this invalid
OpenFlow port number no longer maps properly, so switch-generated
packets could not be transmitted. This patch properly sets the ingress
port to OFPP_LOCAL.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Wed, 14 Nov 2012 23:49:47 +0000 (15:49 -0800)]
bridge: Open internal ports with the correct type.
Use the new ofproto function that returns how a port should be opened
based on the datapath type. This fixes a problem with internal ports
being attached to userspace datapaths.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Wed, 14 Nov 2012 23:50:20 +0000 (15:50 -0800)]
Add functions to determine how port should be opened based on type.
Depending on the port and type of datapath, a port may need to be opened
as a different type of device than it's configured. For example, an
"internal" port on a "dummy" datapath should opened as a "dummy" port.
This commit adds the ability for a dpif to provide this information to a
caller. It will be used in a future commit.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Wed, 14 Nov 2012 23:58:19 +0000 (15:58 -0800)]
dpif-linux.c: Let the kernel pick a port number if one not requested.
With the single datapath change, we no longer depend on the kernel to
make sure that we don't reuse OpenFlow port numbers, since the ofproto
library now picks them. Remove the code that contained that logic.
Suggested-by: Ben Pfaff <blp@nicira.com>
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Wed, 14 Nov 2012 01:57:11 +0000 (17:57 -0800)]
dpif-netdev: Don't special-case "br" bridges in create_dp_netdev().
With a single datapath, we no longer need to special case bridges
beginning with "br*" for testing, since all netdev-based bridges
are named "ovs-netdev". This removes that unnecessary code.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Ben Pfaff [Fri, 16 Nov 2012 06:10:56 +0000 (22:10 -0800)]
nicira-ext: Remove NXFF_OPENFLOW12.
This was meant to splice OXM support into OpenFlow 1.0, in place of NXM,
but I no longer see any value in it.
Signed-off-by: Ben Pfaff <blp@nicira.com>
Reviewed-by: Simon Horman <horms@verge.net.au>
Ben Pfaff [Fri, 16 Nov 2012 06:09:07 +0000 (22:09 -0800)]
ofp-util: Make ofputil_encode_set_protocol() able to return failure.
Soon, it's not going to be possible to switch between every possible
protocol on an established OpenFlow connection, yet
ofputil_encode_set_protocol() didn't have a documented way to report such
a problem. This commit adds a means for reporting and makes its callers
able to handle the problem.
Also, initially make ofputil_encode_set_protocol() fail when the current
and requested protocols are for different OpenFlow versions.
Signed-off-by: Ben Pfaff <blp@nicira.com>
Reviewed-by: Simon Horman <horms@verge.net.au>
Ben Pfaff [Fri, 16 Nov 2012 06:36:15 +0000 (22:36 -0800)]
ofp-util: Allow use of OpenFlow 12 flow format
This enables the use of the OpenFlow 12 flow format.
Signed-off-by: Simon Horman <horms@verge.net.au>
Signed-off-by: Ben Pfaff <blp@nicira.com>
Ben Pfaff [Fri, 16 Nov 2012 05:01:13 +0000 (21:01 -0800)]
ofp-util: Rename OFPUTIL_P_* constants and update comments for clarity.
It wasn't clear to me, at least, whether an OFPUTIL_P_* constant
indicated an OpenFlow version and a flow format or just a flow format.
After some reflection, I think it's more useful if it indicates both,
because otherwise it might be necessary to pass both an OpenFlow version
and an OFPUTIL_P_* constant in some contexts, but this way only the latter
is needed.
Signed-off-by: Ben Pfaff <blp@nicira.com>
Reviewed-by: Simon Horman <horms@verge.net.au>
Ben Pfaff [Fri, 16 Nov 2012 05:28:56 +0000 (21:28 -0800)]
ovs-ofctl: Style fix.
Signed-off-by: Ben Pfaff <blp@nicira.com>
Reviewed-by: Simon Horman <horms@verge.net.au>
Pravin B Shelar [Thu, 15 Nov 2012 05:10:54 +0000 (21:10 -0800)]
odp-utils: Print human readable ipv4-tunnel-key flags.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Pravin B Shelar [Thu, 15 Nov 2012 05:10:22 +0000 (21:10 -0800)]
odp-utils: Refactor slow_path_reason parse and format functions.
These functions are used in next patch.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Pravin B Shelar [Wed, 14 Nov 2012 01:20:22 +0000 (17:20 -0800)]
vswitchd: Fix function prototype of packet_set_ipv6()
Follwoing patch fixes sparse error:
lib/packets.c:643:1: error: symbol 'packet_set_ipv6' redeclared
with different type (originally declared at lib/packets.h:493)
- incompatible argument 6 (different base types)
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Ansis Atteka [Mon, 5 Nov 2012 13:53:32 +0000 (15:53 +0200)]
datapath: add ipv6 'set' action
This patch adds ipv6 set action functionality. It allows to change
traffic class, flow label, hop-limit, ipv6 source and destination
address fields.
Acked-by: Jesse Gross <jesse@nicira.com>
Signed-off-by: Ansis Atteka <aatteka@nicira.com>
Ansis Atteka [Wed, 7 Nov 2012 17:14:34 +0000 (19:14 +0200)]
sparse: Add ip6_rthdr struct to the ip6.h
Acked-by: Ben Pfaff <blp@nicira.com>
Signed-off-by: Ansis Atteka <aatteka@nicira.com>
Simon Horman [Wed, 7 Nov 2012 08:03:02 +0000 (17:03 +0900)]
ofproto, connmgr: Parameterise OpenFlow versions for adding controller
Allow allowed Open Flow versions to be passed rather
than relying on hard-coded defaults.
This is in preparation for allowing configuration of the
allowed OpenFlow versions.
Signed-off-by: Simon Horman <horms@verge.net.au>
Signed-off-by: Ben Pfaff <blp@nicira.com>
Simon Horman [Wed, 7 Nov 2012 08:03:01 +0000 (17:03 +0900)]
vonn: Allow snoops to use the OpenFlow version of the controller connection
Override the allowed versions of the snoop vconn so that
only the version of the controller connection is allowed.
This is because the snoop will see the same messages as the
controller.
Without this change the snoop will try to negotiate a connection
using the default allowed versions, or in other words only
allow OpenFlow 1.0. This breaks snoops for controller connections
using other OpenFlow versions, that is OpenFlow 1.2.
Signed-off-by: Simon Horman <horms@verge.net.au>
Signed-off-by: Ben Pfaff <blp@nicira.com>
Simon Horman [Wed, 7 Nov 2012 08:03:00 +0000 (17:03 +0900)]
rconn: Add allowed OpenFlow versions
Add allowed OpenFlow versions to struct rconn to allow
reconnect to use these parameters rather than hard-coded defaults.
This is in preparation for allowing configuration of the
allowed OpenFlow versions.
Signed-off-by: Simon Horman <horms@verge.net.au>
Signed-off-by: Ben Pfaff <blp@nicira.com>
Ben Pfaff [Fri, 9 Nov 2012 18:00:22 +0000 (10:00 -0800)]
tests: Add tests for encoding and decoding OpenFlow hello messages.
These tests helped me find the problems that I fixed in previous commits.
Signed-off-by: Ben Pfaff <blp@nicira.com>
Ben Pfaff [Fri, 9 Nov 2012 17:59:12 +0000 (09:59 -0800)]
ofp-util: Make ofputil_encode_hello() return a message with correct length.
This doesn't fix a visible bug, because code to send OpenFlow messages
updates the length itself, but it still seems like the right thing to do.
Signed-off-by: Ben Pfaff <blp@nicira.com>
Ben Pfaff [Fri, 9 Nov 2012 17:57:57 +0000 (09:57 -0800)]
ofp-util: Ignore invalid bitmaps in ofputil_decode_hello_bitmap().
This code has, until now, parsed and accepted invalid bitmaps. It seems
better to simply ignore them, leaving the original set of allowed versions
from the version field in the ofp_header.
Signed-off-by: Ben Pfaff <blp@nicira.com>
Ben Pfaff [Fri, 9 Nov 2012 17:55:12 +0000 (09:55 -0800)]
ofp-msgs: Accept all versions of OpenFlow "hello" messages.
Since the ofp-msgs code was introduced, only "hello" messages for versions
of OpenFlow actually understood (to some extent) by Open vSwitch could be
decoded correctly. This commit corrects that, by making every version of
OpenFlow immutable messages acceptable.
Signed-off-by: Ben Pfaff <blp@nicira.com>
Simon Horman [Wed, 7 Nov 2012 08:02:59 +0000 (17:02 +0900)]
ofp-util: Correct length of hello version bitmap
The length of the hello version bitmap is measured in bytes
so the return value of sizeof is sufficient. Dividing
this by CHAR_BIT (=8) seems to be an artifact of reworking
a previous version of this code.
Signed-off-by: Simon Horman <horms@verge.net.au>
Signed-off-by: Ben Pfaff <blp@nicira.com>
Jesse Gross [Wed, 7 Nov 2012 15:57:22 +0000 (07:57 -0800)]
datapath: Move tun_key out of struct phy and union.
Tunneling metadata is important enough to move out of struct phy
and handled on its own. This makes it somewhat easier to tell
how well the other structures are packed and also the name shorter.
This also drops the union since it's not needed quite yet. We
can introduce it back when we have support for IPv6 tunneling.
Signed-off-by: Jesse Gross <jesse@nicira.com>
Acked-by: Kyle Mestery <kmestery@cisco.com>
Jesse Gross [Wed, 7 Nov 2012 15:35:14 +0000 (07:35 -0800)]
datapath: ipv6_skip_exthdr() is now fully upstream.
OVS required some extensions to ipv6_skip_exthdr() which have
since been accepted upstream. On kernels with these changes we can
now use the upstream version instead of our compatibility code.
Signed-off-by: Jesse Gross <jesse@nicira.com>
Acked-by: Kyle Mestery <kmestery@cisco.com>
Jesse Gross [Wed, 7 Nov 2012 01:55:43 +0000 (17:55 -0800)]
tunnel: Zero keys if marked as not present.
If a key used to find a port then we mark it as being not present
(previously this was accomplished by setting it to zero). However,
we also still pass the key to userspace which could lead to an
inconsistent state. This also zeros out the key as before in order
to make our reported results consistent.
Signed-off-by: Jesse Gross <jesse@nicira.com>
Acked-by: Kyle Mestery <kmestery@cisco.com>
Jesse Gross [Tue, 6 Nov 2012 21:22:28 +0000 (13:22 -0800)]
datapath: Allow GRE64 to use flow based tunneling.
There's no reason why GRE64 can't be configured using flow based
tunneling. Since that's the direction we are moving, everything
should support that. Once this special casing is removed, it's
possible to unify the tunnel parameter parsing code since both
GRE and CAPWAP are the same.
Signed-off-by: Jesse Gross <jesse@nicira.com>
Acked-by: Kyle Mestery <kmestery@cisco.com>
Jesse Gross [Tue, 6 Nov 2012 20:45:19 +0000 (12:45 -0800)]
datapath: Shorten flow tunneling flags.
The names for the flags used by flow based tunneling are pretty long.
This shortens them a little by removing the word FLOW, which is a
distinction that won't be meaningful in the near future.
Signed-off-by: Jesse Gross <jesse@nicira.com>
Acked-by: Kyle Mestery <kmestery@cisco.com>
Jesse Gross [Mon, 5 Nov 2012 23:47:41 +0000 (15:47 -0800)]
datapath: Don't allow flows installed with only TUN_ID.
Tunnel ports now always include full outer IP information, even if
userspace can't understand it. Since our flows our exact match this
information must also be provided when setting up flows. Since flows
with only OVS_KEY_ATTR_TUN_ID keys don't contain all of this information
they can never be hit and we should just reject them at setup time.
Signed-off-by: Jesse Gross <jesse@nicira.com>
Acked-by: Kyle Mestery <kmestery@cisco.com>
Jesse Gross [Mon, 5 Nov 2012 19:30:35 +0000 (11:30 -0800)]
datapath: Don't write into IPV4_TUNNEL data when using TUN_ID.
When the IPV4_TUNNEL action is executed, a pointer in the skb is
directly assigned the address of the action, which is protected by
RCU. If a TUN_ID action is later executed it will write into the
action, which is not allowed by RCU. This avoids the problem by
making a copy of the data and writing into the copy.
Signed-off-by: Jesse Gross <jesse@nicira.com>
Acked-by: Kyle Mestery <kmestery@cisco.com>
Jesse Gross [Mon, 5 Nov 2012 19:08:47 +0000 (11:08 -0800)]
datapath: Include OVS_KEY_ATTR_IPV4_TUNNEL in upstream range.
During development it was preferable to keep OVS_KEY_ATTR_IPV4_TUNNEL
in the non-upstream range of identifiers to avoid conflicts or
compatibility issues as it evolved. However, since the intention is
to get it upstream, it makes sense to move it down now to avoid issues
with compatibility when upgrading.
Signed-off-by: Jesse Gross <jesse@nicira.com>
Acked-by: Kyle Mestery <kmestery@cisco.com>
Ben Pfaff [Wed, 7 Nov 2012 20:55:53 +0000 (12:55 -0800)]
socket-util: Report fd of -1 on error in inet_open_active().
inet_open_active() is documented to report a fd of -1 when an error occurs.
All three of its callers rely on this, by checking only the fd to determine
whether there was an error. This means that if the call to
set_nonblocking() or set_dscp() or connect() failed, then the callers would
try to use a fd that had already been closed, wreaking havoc.
This fixes a bug introduced in commit
a4efa3fc5d (socket-util: Close socket
on failed dscp modification.)
Bug #13750.
Reported-by: Scott Hendricks <shendricks@vmware.com>
Signed-off-by: Ben Pfaff <blp@nicira.com>
Acked-by: Ethan Jackson <ethan@nicira.com>
Ethan Jackson [Tue, 6 Nov 2012 03:00:07 +0000 (19:00 -0800)]
bridge: Fix a segmentation fault in bridge_init_ofproto().
When the database is initially created there may no be rows in the
Open_vSwitch table. In this case, the ovsrec_open_vswitch passed
to bridge_init_ofproto() is NULL and causes a segmentation fault.
Signed-off-by: Ethan Jackson <ethan@nicira.com>
Ethan Jackson [Tue, 6 Nov 2012 02:26:52 +0000 (18:26 -0800)]
netdev-vport: Fix warning due to new GRE type.
This patch fixes the following warning caused by a switch case
which was not handled.
lib/netdev-vport.c:144:5: error: enumeration value
‘OVS_VPORT_TYPE_FT_GRE’ not handled in switch
Signed-off-by: Ethan Jackson <ethan@nicira.com>
Ansis Atteka [Tue, 6 Nov 2012 15:12:38 +0000 (17:12 +0200)]
datapath: improve ipv6_find_hdr() function for IPv6 'set' action
This patch prepares ipv6_find_hdr() function so that it could be
able to skip routing headers, where segements_left is 0. This is
required for us to handle multiple routing header case correctly.
Signed-off-by: Ansis Atteka <aatteka@nicira.com>
Ansis Atteka [Tue, 6 Nov 2012 11:39:58 +0000 (13:39 +0200)]
datapath: Add ipv6_find_hdr() compatibility function.
Implementation of the IPv6 'set' action depends on ipv6_find_exthdr()
function to find routing header. By looking at the routing headers,
it is possible to tell, whether L4 checksums will need to be
recalculated, whenever the destination address is changed in
the main IPv6 header.
This function will need to be customized so that it would skip all
routing headers, where segements_left is equal to 0.
Signed-off-by: Ansis Atteka <aatteka@nicira.com>
Jesse Gross [Tue, 30 Oct 2012 23:34:50 +0000 (16:34 -0700)]
datapath: Remove flow refcount functionality.
Header caching previously required the ability to maintain the lifetime
of flows across RCU boundaries. However, now that header caching is
gone we can simplfy the code and make it match the upstream version.
Signed-off-by: Jesse Gross <jesse@nicira.com>
Acked-by: Pravin B Shelar <pshelar@nicira.com>
Pravin B Shelar [Mon, 5 Nov 2012 23:12:59 +0000 (15:12 -0800)]
datapath: Add missing bug.h file
Commit
51f4701be1 (datapath: Fix sparse warning for symbol
'BUILD_BUG_ON_NOT_POWER_OF_2') missed file bug.h. Commiting it now.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Pravin B Shelar [Mon, 5 Nov 2012 21:44:23 +0000 (13:44 -0800)]
datapath: Add support for 3.6 kernel.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Acked-by: Jesse Gross <jesse@nicira.com>
Pravin B Shelar [Mon, 5 Nov 2012 21:44:00 +0000 (13:44 -0800)]
datapath: Remove tunnel header caching.
Tunnel caching was added to reduce CPU utilization on TX path
by caching packet header, So performance gain is directly proportional
to number of skbs transferred. But with help of offloads skb are getting
larger. So there are less number of skbs. Therefore header caching does
not shows similar gains we seen in past. And now kernel 3.6 has removed
dst caching from networking which makes header caching even more tricky.
So this commit removes header caching from OVS tunnelling.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Acked-by: Jesse Gross <jesse@nicira.com>
Pravin B Shelar [Mon, 5 Nov 2012 21:41:02 +0000 (13:41 -0800)]
datapath: Fix sparse warning for symbol 'BUILD_BUG_ON_NOT_POWER_OF_2'
BUILD_BUG_ON_NOT_POWER_OF_2 symbol is moved from kernel.h to
bug.h in 3.4. Therefore sparse is giving warning:
include/linux/bug.h:15:9: warning: preprocessor token
BUILD_BUG_ON_NOT_POWER_OF_2 redefined
ovs/datapath/linux/compat/include/linux/kernel.h:44:9:
this was the original definition
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Acked-by: Jesse Gross <jesse@nicira.com>
Pravin B Shelar [Mon, 5 Nov 2012 21:40:17 +0000 (13:40 -0800)]
netdev: Deprecate CAPWAP support.
The CAPWAP implementation is just the encapsulation format and
therefore really not the full protocol. While there were some
uses of it (primarily hardware support and UDP transport). But
these are most likely better provided by VXLAN. As a result,
CAPWAP will be removed no earlier than February 2013.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Pravin B Shelar [Mon, 5 Nov 2012 21:39:00 +0000 (13:39 -0800)]
datapath: enable encap for capwap.
kernel 3.5 added a switch to turn on UDP encap, capwap needs
to enable it.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Acked-by: Jesse Gross <jesse@nicira.com>
Pravin B Shelar [Mon, 5 Nov 2012 21:38:34 +0000 (13:38 -0800)]
datapath: Add Upstream id for GRE type.
Once GRE is upstream it will have new type to have continuous sequence
of ids for vport type. Following patch adds this ID to have
compatibility with it.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Acked-by: Jesse Gross <jesse@nicira.com>
Mehak Mahajan [Mon, 5 Nov 2012 19:14:36 +0000 (11:14 -0800)]
Correct the warning for format specifier.
ofputil_format_version uses type 'size_t' to print version whereas
expected type is 'unsigned int'.
Signed-off-by: Mehak Mahajan <mmahajan@nicira.com>
Mehak Mahajan [Mon, 5 Nov 2012 19:05:18 +0000 (11:05 -0800)]
Update master to reflect updates on branch-1.9
This commit updates NEWS and debian/changelog to reflect that
commit
8087f5ff (Process RARP packets with ethertype 0x8035
similar to ARP packets.) is going to be backported intto 1.9.0, so
that they should no longer be listed as post-1.9 features.
Signed-off-by: Mehak Mahajan <mmahajan@nicira.com>
Simon Horman [Fri, 26 Oct 2012 02:35:52 +0000 (11:35 +0900)]
connmgr: Parameterize ofservice_create() over allowed OpenFlow versions
This is in preparation for allowing configuration of the
allowed OpenFlow versions.
Signed-off-by: Simon Horman <horms@verge.net.au>
Signed-off-by: Ben Pfaff <blp@nicira.com>
Simon Horman [Mon, 5 Nov 2012 06:04:55 +0000 (22:04 -0800)]
ofp-util: Add version bitmap support to hello messages.
Allow encoding and decoding of version bitmap in hello messages
as specified in Open Flow 1.3.1.
Signed-off-by: Simon Horman <horms@verge.net.au>
[blp@nicira.com simplified and generalized decode/encode functions]
Signed-off-by: Ben Pfaff <blp@nicira.com>
Simon Horman [Mon, 5 Nov 2012 05:41:02 +0000 (21:41 -0800)]
vconn: Replace minimum version with bitmap of allowed versions.
This is in preparation for allowing the range of allowed OpenFlow versions
to be configured.
As part of this change pvconn_open() is now paramatised over the allowed
versions. this is to avoid avoids needing to provide version information
as a parameter to pvconn_accept(). This will in turn avoid the need to
pass version information to connmgr_run().
Signed-off-by: Simon Horman <horms@verge.net.au>
[blp@nicira.com simplified slightly and generalize log messages]
Signed-off-by: Ben Pfaff <blp@nicira.com>
Simon Horman [Mon, 5 Nov 2012 05:40:54 +0000 (21:40 -0800)]
ofp-util: Add functions to support version number bitmaps.
The bitmap is simple a uint32_t, allowing for versions up until
wire-protocol number 31.
Bit offsets correspond to ofp_version numbers which in turn
correspond to wire-protocol numbers for Open Flow versions..
E.g. (1 << OFP10_VERSION) is the mask for Open Flow 1.1.
If the bit for a version is set then it is allowed, otherwise it is
disallowed.
This is in preparation for allowing the range of allowed OpenFlow
versions to be configured.
Signed-off-by: Simon Horman <horms@verge.net.au>
[blp@nicira.com simplified some functions to constants, and rewrote
a few functions in terms of utility bitwise functions]
Signed-off-by: Ben Pfaff <blp@nicira.com>
Ben Pfaff [Thu, 1 Nov 2012 00:13:27 +0000 (17:13 -0700)]
util: New functions for the index of the leftmost or rightmost 1-bit.
These will acquire a user in an upcoming commit.
Signed-off-by: Ben Pfaff <blp@nicira.com>
Ben Pfaff [Thu, 1 Nov 2012 00:12:38 +0000 (17:12 -0700)]
util: Group functions for bitwise tests.
This only moves code around for more logical grouping.
Signed-off-by: Ben Pfaff <blp@nicira.com>
Ben Pfaff [Sun, 4 Nov 2012 01:00:39 +0000 (18:00 -0700)]
ofproto: Report 0 Mbps when speed not available instead of 100 Mbps.
When a link is down, or when a link has no speed because it is not a
physical interface, Open vSwitch previously reported that its rate is 100
Mbps as a default. This is counterintuitive, however, so this commit
changes Open vSwitch behavior to report 0 Mbps when a link is down or its
speed is otherwise unavailable.
Bug #13388.
Reported-by: Hiroshi Tanaka <htanaka@nicira.com>
Signed-off-by: Ben Pfaff <blp@nicira.com>
Mehak Mahajan [Fri, 2 Nov 2012 18:43:46 +0000 (11:43 -0700)]
Process RARP packets with ethertype 0x8035 similar to ARP packets.
With this commit, OVS will match the data in the RARP packets having
ethertype 0x8035, in the same way as the data in the ARP packets.
Signed-off-by: Mehak Mahajan <mmahajan@nicira.com>
Ben Pfaff [Fri, 2 Nov 2012 17:37:59 +0000 (10:37 -0700)]
ofproto: Better hide hidden tables.
ofproto has a concept of "hidden" OpenFlow tables. Currently these are
used internally only for ofproto-dpif for a couple of unimportant
purposes. However, hidden tables were not hidden well enough, because
OFTest was able to spot ofproto-dpif's hidden table and, seeing that it
had a couple of flows in it even after OFTest had tried to delete all
flows, failed at least one test.
This commit hides the tables better:
- The number of tables reported in a feature reply no longer counts
hidden tables.
- Table stats replies omit hidden tables.
This commit introduces the requirement that hidden tables, if any, be the
highest-numbered tables. This is because it's not clear to me that
OpenFlow intends to allow tables to be numbered noncontiguously.
We could take this further, by not exposing hidden tables in any way, but
I have this pet theory that being able to get the statistics for these
tables will come in handy for debugging someday.
Found by openflow_protocol_messages.ModifyStateDelete in OFTest.
Signed-off-by: Ben Pfaff <blp@nicira.com>
Acked-by: Kyle Mestery <kmestery@cisco.com>
Ben Pfaff [Fri, 2 Nov 2012 17:21:08 +0000 (10:21 -0700)]
FAQ: Explain the difference between a switch and a bridge.
(There isn't any difference.)
Based on a conversation in #openvswitch.
Signed-off-by: Ben Pfaff <blp@nicira.com>
Acked-by: Kyle Mestery <kmestery@cisco.com>
Justin Pettit [Wed, 31 Oct 2012 00:41:22 +0000 (17:41 -0700)]
ofproto-dpif: Use a single underlying datapath across multiple bridges.
This commit switches to using a single backing datapath (called
"ovs-datapath") for all bridges of that datapath's type. Previously,
resources couldn't be shared across bridges, since each was in its own
datapath. This change will allow sharing of tunnels and cheaper patch
ports to be added in the future.
Since bridges share a common datapath, the ovs-dpctl commands won't
provide bridge-specific information. Users wishing to have that
information should use the new "ovs-appctl dpif/*" commands as
documented in ovs-vswitchd(8).
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Thu, 1 Nov 2012 23:16:16 +0000 (16:16 -0700)]
dpif-sflow: Remove unused parameter from dpif_sflow_create().
Suggested-by: Ben Pfaff <blp@nicira.com>
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Thu, 1 Nov 2012 23:04:06 +0000 (16:04 -0700)]
dpif: Add function to get the dpif type.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Thu, 18 Oct 2012 06:11:53 +0000 (23:11 -0700)]
dpif: Add new dpif_port_exists() function.
Provide the ability to determine whether a port exists in a datapath
without having to deal with a "dpif_port" structure as with
dpif_port_query_by_name(). A future patch will use this function.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Fri, 12 Oct 2012 00:17:40 +0000 (17:17 -0700)]
ofproto: Add type "run", "run_fast", and "wait" provider methods.
Add the ability for ofproto providers to have top-level "run",
"run_fast", and "wait" methods, similar to the current ofproto ones.
There are no current users, but this will be useful in a future commit.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Sat, 13 Oct 2012 00:46:06 +0000 (17:46 -0700)]
ovs-dpctl: Make specifying datapath optional for some commands.
A future commit will make all bridges use a single backing datapath.
This commit makes the "dp" argument for "dump-flows" and "del-flows"
optional, since there will typically only be one actual datapath.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Wed, 17 Oct 2012 02:21:52 +0000 (19:21 -0700)]
ofproto-dpif: Introduce subfacet_destroy_batch().
A future commit will introduce another caller that wants to destroy
batches of subfacets. Repurpose expire_batch() to be a more generic
method for destroying batches of subfacets.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Wed, 26 Sep 2012 01:47:02 +0000 (18:47 -0700)]
ofproto-dpif: Add ovs-appctl commands for ovs-dpctl functions.
These commands will be useful in a future commit that makes multiple
bridges share a single backing datapath. The ovs-dpctl commands will
show information about the backing datapath, so it will be difficult to
determine which information belongs to which bridge. The new "dpif/*"
ovs-appctl commands return information about the bridge--regardless of
how the backing datapath is configured.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Sun, 14 Oct 2012 00:45:00 +0000 (17:45 -0700)]
tests: Rewrite unit tests to not expect bridge with odp zero.
A future commit will make all bridges of a particular type share a
single backing datapath. That backing datapath will have a datapath
port number of zero and bridges will be assigned other numbers. This
commit modifies the tests so that they don't expect port zero.
It adopts the convention that bridges of type "dummy" with a name of the
form "br<n>" will be assigned a port number of 100+<n>.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Sat, 29 Sep 2012 00:56:07 +0000 (17:56 -0700)]
Separate OpenFlow port numbers from datapath ones.
In a future commit, we will make multiple bridges share a single backing
datapath. Our simple mapping from datapath to OpenFlow port numbers
won't work, since we'll want the same OpenFlow port numbers on different
bridges. For example, the OFPP_LOCAL port must be the same on all
bridges, but will have to be a different datapath port on the converged
datapath.
This commit makes it the responsibility of ofproto to assign the
OpenFlow port numbers instead of doing a simple translation from the
datapath ones.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Thu, 4 Oct 2012 07:51:41 +0000 (00:51 -0700)]
tests: Define new ADD_OF_PORTS macro for ofproto tests.
A future commit will break the relation between OpenFlow and datapath
port numbers. The new ADD_OF_PORTS macro adds an interface to a bridge
and configures it such that both the OpenFlow and datapath port numbers
are the same. This is important in tests that deal with both OpenFlow
and datapath port numbers (e.g., ones that use ofproto/trace).
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Tue, 9 Oct 2012 00:57:22 +0000 (17:57 -0700)]
ofproto: Add initialization function.
A future commit will make all bridges of a particular dpif share a
single backing datapath. In order to handle restart, the datapath will
need to have some idea of what the initial state looks like. Otherwise,
it won't know which ports belong to which bridges and orphaned ports may
never be cleaned up.
This commit introduces an initialization method to ofproto, which takes
as an argument a high-level description of the bridges and ports. An
ofproto provider can then use this information to initialize its state.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Wed, 3 Oct 2012 05:25:51 +0000 (22:25 -0700)]
Allow the OpenFlow port to be requested for a port.
A new "ofport_request" column makes it possible to request the OpenFlow
port number when adding a port.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Tue, 25 Sep 2012 22:25:51 +0000 (15:25 -0700)]
Use ODP ports in dpif layer and below.
The current code has a simple mapping between datapath and OpenFlow port
numbers (the port numbers were the same other than OFPP_LOCAL which maps
to datapath port 0). Since the translation was know at compile time,
this allowed different layers to easily translate between the two, so
the translation often occurred late.
A future commit will break this simple mapping, so this commit draws a
line between where datapath and OpenFlow port numbers are used. The
ofproto-dpif layer will be responsible for the translations. Callers
above will use OpenFlow port numbers. Providers below will use
datapath port numbers.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Wed, 26 Sep 2012 23:22:47 +0000 (16:22 -0700)]
Always treat datapath ports as 32 bits.
Most of the code referred to datapath ports as 32-bit values, but a few
places still used 16-bit references.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Thu, 1 Nov 2012 06:08:32 +0000 (23:08 -0700)]
sset: Add sset_at_position() function.
This will be useful in a future commit.
Suggested-by: Ben Pfaff <blp@nicira.com>
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Justin Pettit [Tue, 25 Sep 2012 23:42:45 +0000 (16:42 -0700)]
dpif: Fix minor typo in comment.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Jesse Gross [Thu, 1 Nov 2012 22:50:16 +0000 (15:50 -0700)]
odp: Enable parsing ipv4_tunnel key.
We can format ipv4_tunnel keys from the kernel but can't currently
parse them. Userspace doesn't know how to do anything with this
information but this support enables utilities like ovs-dpctl and
ovs-appctl ofproto/trace to show useful information.
Bug #13785
Signed-off-by: Jesse Gross <jesse@nicira.com>
Ben Pfaff [Thu, 1 Nov 2012 20:54:20 +0000 (13:54 -0700)]
Makefile: Allow "make modules_install" to work from top-level directory.
INSTALL recommends running "make modules_install" but it didn't mention
that this has to be done from datapath/linux. This commit makes this work
from the top-level build directory.
(Another solution would be to update INSTALL, but this approach seems a
little easier for the user.)
Signed-off-by: Ben Pfaff <blp@nicira.com>
Ansis Atteka [Mon, 29 Oct 2012 16:12:03 +0000 (18:12 +0200)]
netdev-vport: Warn users that pmtud is deprecated
This is a complementary patch to the other one I sent out previously.
The purpose of it is to make it more obvious that tunnel pmtud is
deprecated and will be removed soon.
Requested-by: Jesse Gross <jesse@nicira.com>
Signed-off-by: Ansis Atteka <aatteka@nicira.com>
Ben Pfaff [Wed, 15 Aug 2012 00:14:43 +0000 (17:14 -0700)]
ofproto-dpif: Fix "ofproto/trace" of flow for local port.
The 'in_port' variable is an OpenFlow port, not an ODP port, so we should
not translate it to ODP.
Signed-off-by: Ben Pfaff <blp@nicira.com>
Acked-by: Justin Pettit <jpettit@nicira.com>
Gurucharan Shetty [Mon, 29 Oct 2012 19:25:09 +0000 (12:25 -0700)]
ovs-ctl.in: Do not fail 'restart'.
ovs-ctl restart is called by the debian package upgrade.
We do not want to fail the package upgrade just because
restoring of flows failed.
The error message will still be printed on the console.
Bug #13730.
Signed-off-by: Gurucharan Shetty <gshetty@nicira.com>
Justin Pettit [Sat, 27 Oct 2012 06:05:55 +0000 (15:05 +0900)]
flow: Set ttl in flow_compose().
Thanks to Ben Pfaff for immediately pinpointing the likely location of
the issue.
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Signed-off-by: Simon Horman <horms@verge.net.au>
Signed-off-by: Ben Pfaff <blp@nicira.com>
Gurucharan Shetty [Tue, 4 Sep 2012 19:40:27 +0000 (12:40 -0700)]
debian: Automatically up the slave interfaces of a bond.
When physical interfaces are added to a bond through the
/etc/network/interfaces script, administratively 'up' the
slave interfaces.
Signed-off-by: Gurucharan Shetty <gshetty@nicira.com>
Ethan Jackson [Fri, 26 Oct 2012 19:33:01 +0000 (12:33 -0700)]
timeval: Fix occasional backtrace() related deadlock.
Occasionally, backtrace() will deadlock in the signal handler
because it does some non signal safe initialization. Specifically,
it opens a shared object. As a work around, this patch forces
backtrace() to run outside of a signal handler, so that future
calls will perform as expected.
Signed-off-by: Ethan Jackson <ethan@nicira.com>
Gurucharan Shetty [Fri, 26 Oct 2012 18:17:17 +0000 (11:17 -0700)]
ovs-save: Use timeout option for ovs-vsctl
Running ovs-save directly to save datapath configuration can
wait forever if ovsdb-server is not running. Use a timeout.
Signed-off-by: Gurucharan Shetty <gshetty@nicira.com>
Ansis Atteka [Mon, 22 Oct 2012 07:04:28 +0000 (10:04 +0300)]
netdev-vport: Disable Path MTU Discovery by default
In near future we will switch to the flow-based tunneling and drop path
MTU discovery support. This patch prepares for that by changing the default
PMTUD value from 'enabled' to 'disabled'.
Signed-off-by: Ansis Atteka <aatteka@nicira.com>
Isaku Yamahata [Fri, 26 Oct 2012 04:43:19 +0000 (13:43 +0900)]
OF11: push_vlan support
This implementes push_vlan with 802.1Q.
NOTE: 802.1AD (QinQ) is not supported. It requires another effort.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Ben Pfaff <blp@nicira.com>
Ethan Jackson [Mon, 22 Oct 2012 23:53:54 +0000 (16:53 -0700)]
poll-loop: Log backtraces when CPU usage is high.
Often when debugging Open vSwitch, one will see in the logs that
CPU usage has been high for some period of time, but it's totally
unclear why. In an attempt to remedy the situation, this patch
logs backtraces taken at regular intervals as a poor man's
profiling alternative.
Signed-off-by: Ethan Jackson <ethan@nicira.com>
Ethan Jackson [Mon, 22 Oct 2012 22:42:31 +0000 (15:42 -0700)]
timeval: Coalesce backtraces with counts.
With this patch, `ovs-appctl backtrace` will return a unique list
of backtraces and a count of how many times it has been recorded.
This work had previously been done by ovs-parse-backtrace. However,
in future patches poll-loop will begin logging backtraces as a
matter of course. At this point, coalescing the backtraces will
help keep these log messages brief.
Signed-off-by: Ethan Jackson <ethan@nicira.com>
Ethan Jackson [Wed, 17 Oct 2012 23:43:38 +0000 (16:43 -0700)]
timeval: Take a backtrace on each SIGALRM.
With this patch, timeval will take a backtrace with each SIGALRM
allowing it to retrieve a profiling snapshot instantly. This will
be useful in future patches when backtraces are logged.
Signed-off-by: Ethan Jackson <ethan@nicira.com>
Gurucharan Shetty [Wed, 24 Oct 2012 20:05:01 +0000 (13:05 -0700)]
debian: Save openflow flows during package upgrade.
When debian package for openvswitch-switch is upgraded,
export a variable, OVS_RESTART_SAVE_FLOWS=yes.
This will save the openflow flows in vswitchd and
re-apply it after the upgrade.
Feature #13555.
Signed-off-by: Gurucharan Shetty <gshetty@nicira.com>
Gurucharan Shetty [Wed, 24 Oct 2012 20:02:22 +0000 (13:02 -0700)]
xenserver, rhel, debian: Use ovs-ctl restart.
ovs-ctl has a new command called "restart" which
saves and restores the openflow flows on bridges.
Use that command from the init scripts when doing
a "restart --save-flows=yes".
Also, the debian package postinst script can
set the variable OVS_RESTART_SAVE_FLOWS to "yes"
to ask for save and restore of flows.
Feature #13555.
Signed-off-by: Gurucharan Shetty <gshetty@nicira.com>
Gurucharan Shetty [Wed, 24 Oct 2012 20:19:24 +0000 (13:19 -0700)]
ovs-ctl.in: Don't save flows if the daemons are not running.
When a 'ovs-ctl restart' is executed and the userspace daemons
like ovsdb-server and ovs-vswitchd are not running, attempt to
save flows can wait forever. This also results in the daemons
from not getting started.
Signed-off-by: Gurucharan Shetty <gshetty@nicira.com>
Ben Pfaff [Wed, 24 Oct 2012 23:13:31 +0000 (16:13 -0700)]
Prepare for post-1.9.0 (1.9.90).
Signed-off-by: Ben Pfaff <blp@nicira.com>