Ben Pfaff [Wed, 6 Aug 2008 21:12:22 +0000 (14:12 -0700)]
Fix typo in comment.
Justin Pettit [Mon, 18 Aug 2008 06:17:07 +0000 (23:17 -0700)]
Fix dump-flows to be usable by add-flows.
Some of the fields when printing a flow are hex, but they don't have a
preceeding "0x". When one tries to import them, they're treated as decimal
and weird results occur. Bug 27.
Justin Pettit [Mon, 18 Aug 2008 06:15:21 +0000 (23:15 -0700)]
Use a consistent naming convention for counts.
Justin Pettit [Sat, 16 Aug 2008 16:43:05 +0000 (09:43 -0700)]
Add stats message that allows querying description of components in OpenFlow.
Justin Pettit [Fri, 15 Aug 2008 06:12:37 +0000 (23:12 -0700)]
Added new interface statistics.
Ben Pfaff [Thu, 14 Aug 2008 23:22:00 +0000 (16:22 -0700)]
Make hwtable_dummy compile again.
Bug #109.
Ben Pfaff [Thu, 14 Aug 2008 23:18:46 +0000 (16:18 -0700)]
Fix vlogconf EINTR error returns.
Bug #111.
Ben Pfaff [Thu, 14 Aug 2008 23:10:59 +0000 (16:10 -0700)]
Use OpenFlow OUI for Ethernet address of local ports.
Ben Pfaff [Thu, 14 Aug 2008 18:05:50 +0000 (11:05 -0700)]
Use proper byte order for dl_type of 802.2, non-SNAP frames.
Justin Pettit [Thu, 14 Aug 2008 07:02:14 +0000 (00:02 -0700)]
Switch to new packet-out format and add OFPP_IN_PORT.
The original packet-out format allowed multiple actions to be specified for
buffered packets, but only a destination port for messages including data.
This change makes packet-out more consistent by allowing multiple actions
regardless of how the packet is stored.
This change also disallows sending packets through the incoming port without
explicitly using the OFPP_IN_PORT virtual port.
Ben Pfaff [Wed, 13 Aug 2008 21:37:36 +0000 (14:37 -0700)]
Print OpenFlow ref. impl. and protocol versions at secchan startup time.
Bug #83.
Ben Pfaff [Wed, 13 Aug 2008 21:29:33 +0000 (14:29 -0700)]
Describe how to build and use Debian packages in INSTALL.
Ben Pfaff [Wed, 13 Aug 2008 21:29:18 +0000 (14:29 -0700)]
Add missing prerequisite to debian/control.
Ben Pfaff [Wed, 13 Aug 2008 18:30:31 +0000 (11:30 -0700)]
rconn: Wake up immediately if we drain the send queue.
It is possible that some task is waiting for space to open up in the
send queue, so we need to give it a chance to run.
Ben Pfaff [Wed, 13 Aug 2008 18:26:26 +0000 (11:26 -0700)]
rconn: Use rconn_send() to send echo requests.
There's no point in sending echo requests by hand.
Ben Pfaff [Wed, 13 Aug 2008 18:21:49 +0000 (11:21 -0700)]
rconn: Queue packets for tx only if connected.
Until now, the rconn code would queue up packets not just while connected
but also while connecting. This is not just unnecessary, however, it
actually causes a problem in secchan: if the secchan receives packets
from nl:0 and tries to transmit them, then they will quickly fill up
the rconn's transmit buffer (it uses a 1-packet buffer), which causes
secchan to stop reading packets from nl:0 until the buffer frees up.
That cannot happen until the connection completes. With in-band control,
however, the connection cannot complete until we receive and process
packet_in messages in our in-band hook. Thus, we have a deadlock.
Fixes bug #90, "Sometimes secchan has to go into fail-open mode to connect."
Ben Pfaff [Tue, 12 Aug 2008 23:21:45 +0000 (16:21 -0700)]
Prevent the secchan from dying due to SIGPIPE.
Fix for bug #88.
Ben Pfaff [Tue, 12 Aug 2008 23:15:42 +0000 (16:15 -0700)]
Reduce default burst limit to rate limit / 4.
With the previous default burst limit of rate limit * 2, we would queue
up 2 seconds worth of packet_in messages. This is not only much more
than actually needed, it causes an actual problem: the datapath only
retains buffered packets for up to 1 second, by default, so that flow
setups sent in response have no packet to work with.
Ben Pfaff [Tue, 12 Aug 2008 23:09:31 +0000 (16:09 -0700)]
Make discovery happen faster.
This reduces the initial delay before sending DHCPDISCOVER from
1...10 seconds to 1...3 seconds.
Ben Pfaff [Tue, 12 Aug 2008 22:34:47 +0000 (15:34 -0700)]
Implement rate limiting in secchan.
Ben Pfaff [Mon, 11 Aug 2008 23:24:24 +0000 (16:24 -0700)]
Drop rconn's responsibility for limiting the tx queue.
Now it helps clients track the number of in-flight messages, but lets the
clients do the limiting themselves. This will come in handy for packet-in
rate limiting (in an upcoming commit), in which we want to track in-flight
packet-in messages separately from other in-flight messages.
Ben Pfaff [Mon, 11 Aug 2008 20:44:29 +0000 (13:44 -0700)]
Use signal-based timekeeping functions throughout the source base.
Ben Pfaff [Tue, 12 Aug 2008 20:13:16 +0000 (13:13 -0700)]
Initialize the timetracking module in each userspace program.
Also, use time_alarm() instead of the POSIX alarm() function, because
alarm() interferes with time_alarm().
Ben Pfaff [Tue, 12 Aug 2008 21:41:48 +0000 (14:41 -0700)]
Implement signal-driven time tracking.
Checking the current time only when a periodic timer expires allows us
to reduce the number of system calls to obtain the current time to 10 per
second (from an arbitrary number) plus one per poll loop.
This fixes poll_block() to properly handle being interrupted.
Otherwise, a timeout longer than the interval between signals will never
complete (because it will restart from the beginning after every
signal).
Ben Pfaff [Tue, 12 Aug 2008 21:24:10 +0000 (14:24 -0700)]
Make fatal-signal more willing to share signals with other handlers.
Ben Pfaff [Fri, 8 Aug 2008 22:39:20 +0000 (15:39 -0700)]
Set up flow to drop packets that come in on their output port.
Commit
0a3d4ba3a, "Don't allow the learning switch to send packets back
out the incoming port," drops packets that come in on their incoming port,
but it does not set up a flow to drop those packets. If we are connected
to a hub, then this can be very expensive, since we can continue to get
many packets in that flow and have to drop them in secchan one by one. So
now we set up a flow to drop them for us.
Ben Pfaff [Fri, 8 Aug 2008 18:43:41 +0000 (11:43 -0700)]
Fix handling of IP but non-TCP, non-UDP packets in kernel flow_extract().
In particular, ICMP was broken.
Ben Pfaff [Thu, 7 Aug 2008 22:37:44 +0000 (15:37 -0700)]
Make vconn_open() consistently set *vconnp to NULL on failure.
Fixes issue #80.
Ben Pfaff [Thu, 7 Aug 2008 22:19:31 +0000 (15:19 -0700)]
Implement OpenFlow hard timeouts.
This updates the OpenFlow protocol version and changes the names of
structure members, so any software that uses OpenFlow will need to be
updated to match.
Ben Pfaff [Wed, 6 Aug 2008 18:39:51 +0000 (11:39 -0700)]
Fix formatting of flow matches in ofp-print:
Print a comma after wildcard matches as well as exact matches.
Ben Pfaff [Wed, 6 Aug 2008 18:40:14 +0000 (11:40 -0700)]
Fix uninitialized struct member in userspace table-linear.
Ben Pfaff [Thu, 7 Aug 2008 20:19:09 +0000 (13:19 -0700)]
Fix bug in clearing unsupported switch flags.
We need to keep the flags we support, not drop them.
Ben Pfaff [Thu, 7 Aug 2008 18:46:22 +0000 (11:46 -0700)]
Don't call kfree_skb() with interrupts disabled.
Freeing an skb that has a destructor may require interrupts to be enabled.
This can happen when netfilter is performing NAT, for example.
Discovered by Murphy McCauley.
Ben Pfaff [Tue, 5 Aug 2008 20:51:29 +0000 (13:51 -0700)]
Don't allow unsupported flags to be set in the switch.
Ben Pfaff [Mon, 4 Aug 2008 22:34:26 +0000 (15:34 -0700)]
Implement OFPC_FRAG_DROP fragment handling policy.
Ben Pfaff [Tue, 5 Aug 2008 20:58:12 +0000 (13:58 -0700)]
Reduce redundancy in datapath code.
Ben Pfaff [Tue, 5 Aug 2008 21:13:54 +0000 (14:13 -0700)]
Maintain userspace switch configuration in host byte order.
This makes it easier to work with and makes the code more like the
kernel switch.
Ben Pfaff [Tue, 5 Aug 2008 21:35:26 +0000 (14:35 -0700)]
Fix memory leak when OFPP_TABLE is used for a packet that matches no flow.
Ben Pfaff [Tue, 5 Aug 2008 20:05:56 +0000 (13:05 -0700)]
Drop controller-bound traffic that arrives on the controller's port.
Before, if a hub connected a number of OpenFlow switches and the controller,
then in-band control traffic from one of the OpenFlow switches would be
seen by each of the other switches and forwarded up to the controller as
an ofp_packet_in message. That message would then be seen by all of the
other OpenFlow switches and also forwarded, and so on in an infinite loop.
This change prevents this situation by keeping secchan from forwarding
ofp_packet_in messages for a packet destined to the controller that
arrives on the port where the controller is located.
This code has at least two weaknesses. First, if the controller's port
changes, then the flows set up to drop packets will not be deleted. This
should not be a major problem: if this inadvertently kills a switch's
connection to the controller, then the switch will realize it after it
stops receiving data and re-connect. Its new connection will have new
flow data and therefore its packets will not be dropped.
Second, the notion of the "controller's port" does not take into account
the possibility of loops in the network topology. We need spanning tree
protocol for that.
Justin Pettit [Wed, 6 Aug 2008 08:58:15 +0000 (01:58 -0700)]
Properly track table match counts.
Ben Pfaff [Mon, 4 Aug 2008 22:30:19 +0000 (15:30 -0700)]
Fix typo in comment.
Ben Pfaff [Mon, 4 Aug 2008 21:47:27 +0000 (14:47 -0700)]
Mark functions not used in other source files static.
Ben Pfaff [Mon, 4 Aug 2008 21:21:34 +0000 (14:21 -0700)]
Make it easier for the compiler to optimize IP_IS_FRAGMENT.
Ben Pfaff [Mon, 4 Aug 2008 21:20:26 +0000 (14:20 -0700)]
Fix bug in flow_extract() for extracting VLAN from short packet.
Ben Pfaff [Mon, 4 Aug 2008 21:18:24 +0000 (14:18 -0700)]
Don't accept incomplete TCP headers when extracting flows in kernel.
This makes the kernel and the userspace implementations of flow
extraction behave the same way regarding headers.
Ben Pfaff [Mon, 4 Aug 2008 21:45:59 +0000 (14:45 -0700)]
Don't try to extract UDP/TCP port numbers from IP fragments.
Ben Pfaff [Mon, 4 Aug 2008 21:18:36 +0000 (14:18 -0700)]
Fix typo in comment.
Ben Pfaff [Wed, 30 Jul 2008 22:39:03 +0000 (15:39 -0700)]
dhcp: Make client signal a change if renewing yields different options.
This allows the secchan to connect to a new controller if the DHCP server
starts announcing a new vconn without invalidating old leases.
Ben Pfaff [Wed, 30 Jul 2008 22:45:50 +0000 (15:45 -0700)]
dhcp: Make dhcp_option_to_string() act sensibly with null or empty options.
Also, update ofp-switch-setup to parse the new syntax.
Ben Pfaff [Wed, 30 Jul 2008 22:45:38 +0000 (15:45 -0700)]
dhcp: New function dhcp_option_equals().
Ben Pfaff [Wed, 30 Jul 2008 22:00:01 +0000 (15:00 -0700)]
Clean up secchan code.
Ben Pfaff [Wed, 30 Jul 2008 22:45:33 +0000 (15:45 -0700)]
netdev: Let netdev_get_in4(), netdev_get_in6() take null pointer argument.
Some callers want to just test whether the device has an address assigned
and don't care about the particular address, so this simplifies those
callers slightly.
Ben Pfaff [Wed, 30 Jul 2008 22:46:36 +0000 (15:46 -0700)]
Debian: Listen on Unix domain socket for management connections.
Ben Pfaff [Wed, 30 Jul 2008 22:45:21 +0000 (15:45 -0700)]
vconn: Implement Unix domain socket vconn.
These are useful for local management connections because, unlike TCP
sockets, they are subject to regular file system permissions.
Ben Pfaff [Tue, 29 Jul 2008 21:38:00 +0000 (14:38 -0700)]
Move unix socket helper functions into socket-util.
This makes them available for use by the upcoming vconn_unix.
Ben Pfaff [Wed, 30 Jul 2008 22:45:04 +0000 (15:45 -0700)]
vconn: Use vconn_stream to factor out code from vconn_tcp.
Ben Pfaff [Wed, 30 Jul 2008 22:44:22 +0000 (15:44 -0700)]
vconn: Introduce infrastructure for stream socket-based vconns.
Ben Pfaff [Wed, 30 Jul 2008 22:43:57 +0000 (15:43 -0700)]
vconn: Allow vconns to delegate to underlying implementations.
Ben Pfaff [Tue, 29 Jul 2008 22:03:22 +0000 (15:03 -0700)]
Fix typo in comment.
Natasha Gude [Wed, 30 Jul 2008 21:41:47 +0000 (14:41 -0700)]
IP_ARGS takes address of nw_addr
Justin Pettit [Wed, 30 Jul 2008 06:43:58 +0000 (23:43 -0700)]
Don't allow the learning switch to send packets back out the incoming port.
The learning switch functionality included in OpenFlow will send packets
out the port they arrived on if the learned destination is from that port.
This causes problems when newer switches flood packets for destinations they
don't know and OpenFlow reinjects packets back into the network. The correct
behavior is to just drop the packet.
Ben Pfaff [Tue, 29 Jul 2008 00:29:26 +0000 (17:29 -0700)]
rconn: Reconnect reliably when underlying vconn reports error.
When a vconn reports an error, the rconn would not reliably reconnect.
In particular, if the error was reported after the call to rconn_run()
but before rconn_run_wait() was called, then the state's "run" routine
would not set min_timeout properly, leading to a potentially arbitrarily
long wait (depending on what other events were going on in) until the
state's "run" routine was called again.
The fix is to have a separate per-state "timeout" routine to compute
when the state needs to be re-entered.
This commit was tested using the following change to randomly inject
errors:
@@ -554,11 +554,16 @@
static int
try_send(struct rconn *rc)
{
int retval = 0;
struct buffer *next = rc->txq.head->next;
- retval = vconn_send(rc->vconn, rc->txq.head);
+ if (!random_range(1000)) {
+ fprintf(stderr, "injecting ECONNRESET\n");
+ retval = ECONNRESET;
+ } else {
+ retval = vconn_send(rc->vconn, rc->txq.head);
+ }
if (retval) {
if (retval != EAGAIN) {
disconnect(rc, retval);
}
return retval;
Justin Pettit [Mon, 28 Jul 2008 23:08:37 +0000 (16:08 -0700)]
Export the "dp_mutex" symbol, since it's needed by the hardware tables.
Ben Pfaff [Mon, 28 Jul 2008 22:07:24 +0000 (15:07 -0700)]
Drop unnecessary conditional and incorrect comment.
'controller_relay' is always non-null (but in an earlier revision of
the code that was never pushed to the repository, this was not the case.)
Ben Pfaff [Mon, 28 Jul 2008 14:08:38 +0000 (07:08 -0700)]
Be slightly pickier about going into fail-open mode.
Before, the amount of time disconnected was measured relative to the
last time we connected, that is, the last time connect(2) succeeded.
Thus, if we were connected for a long time, and then disconnected,
we would immediately go into fail-open mode.
This change make the disconnected time relative to the last OpenFlow
message received. Thus, if we are connected for a long time and
receive plenty of packets, and then disconnect, there will still be
an opportunity to reconnect before failing open.
Ben Pfaff [Mon, 28 Jul 2008 17:11:51 +0000 (10:11 -0700)]
Avert GCC false-positive warning.
Ben Pfaff [Thu, 24 Jul 2008 23:07:32 +0000 (16:07 -0700)]
Merge branch 'locking'
Ben Pfaff [Thu, 24 Jul 2008 00:21:58 +0000 (17:21 -0700)]
Don't invoke Debconf from openflow-switch postinst.
This was in there because at one time it seemed desirable to configure
from the postinst, but that idea was abandoned. Now Debconf seems to
be screwing something up somehow, so we might as well not run it at all,
since it wasn't doing anything useful.
Ben Pfaff [Thu, 24 Jul 2008 00:04:05 +0000 (17:04 -0700)]
vlog: Reduce syslog level of VLL_EMER messages to LOG_ALERT.
The LOG_EMERG log level sprayed these messages across every xterm and
console in the system, which was excessive.
Ben Pfaff [Thu, 24 Jul 2008 00:04:21 +0000 (17:04 -0700)]
vlog: Send multi-line log messages to syslog() one line at a time.
syslogd swallows new-lines, which makes multi-line messages otherwise
difficult to read.
Ben Pfaff [Thu, 24 Jul 2008 00:03:03 +0000 (17:03 -0700)]
Verify OpenFlow version number in vconn_recv().
Ben Pfaff [Wed, 23 Jul 2008 23:42:05 +0000 (16:42 -0700)]
Fix kdist_clean target in debian/rules.
This fixes "module-assistant auto-install openflow-datapath" for
bootstrapping from a clean environment. (Use "m-a -f purge
openflow-datapath" to test that it works starting from an un-clean
environment.)
Ben Pfaff [Wed, 23 Jul 2008 21:57:22 +0000 (14:57 -0700)]
Update Debconf templates PO file.
Ben Pfaff [Wed, 23 Jul 2008 21:47:37 +0000 (14:47 -0700)]
Don't log messages to the console by default in Debian package.
Ben Pfaff [Wed, 23 Jul 2008 21:54:19 +0000 (14:54 -0700)]
Be less picky about precise name of process in init script.
The previous version always printed ERROR because for whatever reason
we were comparing "/usr/sbin/secchan" against "secchan", which of
course failed.
Ben Pfaff [Wed, 23 Jul 2008 21:53:06 +0000 (14:53 -0700)]
Only generate private key and certificate request if ssl enabled.
Generating them unconditionally caused a problem with the init script:
the script required there to be a certificate for the private key if
the private key existed, but we do want to allow TCP-only configurations
and always having a private key prevented that.
Also, give the user advice on how to deal with this problem when it
arises in the init script.
Ben Pfaff [Wed, 23 Jul 2008 21:30:59 +0000 (14:30 -0700)]
Enable secchan, ofp-discover to update /etc/resolv.conf.
This way it becomes possible to more reliably refer to the controller
and the PKI server using hostnames.
Ben Pfaff [Wed, 23 Jul 2008 20:11:00 +0000 (13:11 -0700)]
Support controller discovery in Debian packages.
Ben Pfaff [Wed, 23 Jul 2008 20:09:25 +0000 (13:09 -0700)]
New utility ofp-kill.
Needed for controller discovery in upcoming revision of ofp-switch-setup.
Ben Pfaff [Wed, 23 Jul 2008 20:08:31 +0000 (13:08 -0700)]
New function make_pidfile_name().
The upcoming ofp-kill utility wants to use this.
Ben Pfaff [Wed, 23 Jul 2008 20:07:40 +0000 (13:07 -0700)]
secchan: Improve logging when rejecting a controller vconn.
Ben Pfaff [Wed, 23 Jul 2008 20:07:02 +0000 (13:07 -0700)]
Lock pidfiles with fcntl and create them atomically.
This makes it possible to verify that the program that created the
pidfile is still running.
Ben Pfaff [Tue, 22 Jul 2008 21:44:10 +0000 (14:44 -0700)]
New program ofp-discover.
Ben Pfaff [Tue, 22 Jul 2008 20:26:03 +0000 (13:26 -0700)]
Add SIGALRM to blockable fatal signals.
We use SIGALRM to limit runtime of dpctl and ofp-discover. There is no
reason that we should not clean up after it in the same way as any other
catchable signal. In particular, ofp-discover wants to restore network
device flags on timeout.
Ben Pfaff [Tue, 22 Jul 2008 21:01:25 +0000 (14:01 -0700)]
dhcp: Always build test-dhcp-client.
Putting test-dhcp-client in check_PROGRAMS instead of noinst_PROGRAMS
makes it too easy to break test-dhcp-client without noticing it.
Ben Pfaff [Tue, 22 Jul 2008 21:01:17 +0000 (14:01 -0700)]
dhcp: Fix race condition in test-dhcp-client.
dhclient_wait() might also access dhcp state.
Ben Pfaff [Tue, 22 Jul 2008 20:32:42 +0000 (13:32 -0700)]
dhcp: Only print "entering" message if state actually changes.
Ben Pfaff [Tue, 22 Jul 2008 20:16:08 +0000 (13:16 -0700)]
dhcp: Don't attempt to drain netdev if we ran out of packets.
Fixes a theoretical race condition that would have caused more
retransmission than otherwise necessary.
Ben Pfaff [Wed, 23 Jul 2008 20:12:14 +0000 (13:12 -0700)]
dhcp: Make output of dhcp_option_to_string() more uniform, to ease parsing.
(The ofp-switch-setup program will want to parse its output.)
Ben Pfaff [Tue, 22 Jul 2008 21:01:10 +0000 (14:01 -0700)]
dhcp: Make dhcp_msg_to_string() support a multiline format also.
The upcoming ofp-discover program wants to print out the binding
information, and dhcp_msg_to_string() is pretty close to what it
wants. But one-field-per-line is easier for other programs to parse.
Ben Pfaff [Tue, 22 Jul 2008 21:01:01 +0000 (14:01 -0700)]
dhcp: Break out netdev configuration from DHCP binding.
This allows us to have DHCP clients that don't actually bind or unbind
a network device's IP address. This is useful for doing controller
discovery without actually reconfiguring a network device.
Ben Pfaff [Tue, 22 Jul 2008 21:00:03 +0000 (14:00 -0700)]
dhcp: Be more precise about updating 'changed' flag.
There's no need to set changed unconditionally in dhclient_init(), and
it can make extra work for the users of the dhclient, by causing them
to think that something changed when it didn't.
Ben Pfaff [Tue, 22 Jul 2008 20:53:20 +0000 (13:53 -0700)]
Make set_pidfile() easier to use.
Ben Pfaff [Tue, 22 Jul 2008 20:33:02 +0000 (13:33 -0700)]
dpctl: Get rid of effectively unused variable.
Ben Pfaff [Mon, 21 Jul 2008 23:30:53 +0000 (16:30 -0700)]
Don't include unneeded header file.
Ben Pfaff [Mon, 21 Jul 2008 22:44:18 +0000 (15:44 -0700)]
Do the Debian package build in a _debian subdirectory.
This makes it easier to do a Debian build in the same source tree
also being used for other development.
Ben Pfaff [Mon, 21 Jul 2008 22:42:31 +0000 (15:42 -0700)]
Don't unnecessarily link against -lresolv and -ldl.
Eliminates warnings given by dpkg-shlibdeps when building Debian
packages.
Ben Pfaff [Mon, 21 Jul 2008 22:41:38 +0000 (15:41 -0700)]
Distribute new rculist.h headers.
Ben Pfaff [Mon, 21 Jul 2008 22:04:54 +0000 (15:04 -0700)]
Move Autoconf's macro definitions into config.h.
This makes "make" output slightly more readable.
Ben Pfaff [Mon, 21 Jul 2008 21:48:53 +0000 (14:48 -0700)]
Drop use of libtool for libraries.
libtool was needed when NOX was directly linking against libopenflow.
Since it doesn't do that anymore, we can drop its use (and that makes
OpenFlow configure and build slightly faster).