This commit changes the semantics of 'netdev_set_multiq()' to allow OVS
DPDK to run on device with limited multi queue support.
* If a netdev doesn't have the requested number of rxqs it can simply
inform the datapath without failing.
* If a netdev doesn't have the requested number of txqs it should try
to create as many as possible and use locking.
Signed-off-by: Daniele Di Proietto <diproiettod@vmware.com>
Signed-off-by: Ethan Jackson <ethan@nicira.com>
Acked-by: Ethan Jackson <ethan@nicira.com>
If we receive a packet with an invalid tunnel header, we
should drop the packet without further processing. Currently
we do this by removing any parsed tunnel metadata. However,
this is not sufficient to stop processing - this only results
in the packet getting dropped by chance when something
usually runs across part of the packet that does not make
sense. Since both the packet and its metadata are in an
inconsistent state, it's also possible that the result is
an ovs-vswitchd crash or forwarding of a mangled packet.
Rather than clear the metadata, an alternate solution is to
remove all of the packet data. This guarantees that the
packet gets dropped during the next round of processing.
Signed-off-by: Jesse Gross <jesse@nicira.com>
Acked-by: Pravin B Shelar <pshelar@nicira.com>
The userspace tunneling API for pushing and popping tunnel headers
is currently based on processing batches of packets. However, there
is no obvious way to take advantage of batching for these operations
and so each tunnel operation has a pair of loops to process the
batch. This changes the API to operate on single packets to enable
better code reuse.
Signed-off-by: Jesse Gross <jesse@nicira.com>
Acked-by: Pravin B Shelar <pshelar@nicira.com>
e.g. Set tunnel id for encapsulated VxLAN packet (out_key=flow):
ovs-vsctl add-port int-br vxlan0 -- set interface vxlan0 \
type=vxlan options:remote_ip=172.168.1.2 options:out_key=flow
ovs-ofctl add-flow int-br in_port=LOCAL, icmp,\
actions=set_tunnel:3, output:1 (1 is the port# of vxlan0)
Output tunnel ID should be modified to 3 with this patch.
Signed-off-by: Ricky Li <ricky.li@intel.com>
Acked-by: Pravin B Shelar <pshelar@nicira.com>
This patch adds support for a new port type to userspace datapath
called dpdkvhost. This allows KVM (QEMU) to offload the servicing
of virtio-net devices to its associated dpdkvhost port. Instructions
for use are in INSTALL.DPDK.
This has been tested on Intel multi-core platforms and with clients
that have virtio-net interfaces.
Signed-off-by: Ciara Loftus <ciara.loftus@intel.com>
Signed-off-by: Kevin Traynor <kevin.traynor@intel.com>
Signed-off-by: Maryam Tahhan <maryam.tahhan@intel.com>
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Currently dp-packet make use of ofpbuf for managing packet
buffers. That complicates ofpbuf, by making dp-packet
independent of ofpbuf both libraries can be optimized for
their own use case.
This avoids mapping operation between ofpbuf and dp_packet
in datapath upcalls.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Acked-by: Jarno Rajahalme <jrajahalme@nicira.com>
Acked-by: Ben Pfaff <blp@nicira.com>
dp_packet is short and better name for datapath packet
structure.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Acked-by: Jarno Rajahalme <jrajahalme@nicira.com>
A new function vlog_insert_module() is introduced to avoid using
list_insert() from the vlog.h header.
Signed-off-by: Thomas Graf <tgraf@noironetworks.com>
Acked-by: Ben Pfaff <blp@nicira.com>
struct list is a common name and can't be used in public headers.
Signed-off-by: Thomas Graf <tgraf@noironetworks.com>
Acked-by: Ben Pfaff <blp@nicira.com>
OVS router depends on tnl_conf_seq and all tunnel related
components should be initialized before registering dpif
implementations.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Acked-by: Jarno Rajahalme <jrajahalme@nicira.com>
Following patch adds support for userspace tunneling. Tunneling
needs three more component first is routing table which is configured by
caching kernel routes and second is ARP cache which build automatically
by snooping arp. And third is tunnel protocol table which list all
listening protocols which is populated by vswitchd as tunnel ports
are added. GRE and VXLAN protocol support is added in this patch.
Tunneling works as follows:
On packet receive vswitchd check if this packet is targeted to tunnel
port. If it is then vswitchd inserts tunnel pop action which pops
header and sends packet to tunnel port.
On packet xmit rather than generating Set tunnel action it generate
tunnel push action which has tunnel header data. datapath can use
tunnel-push action data to generate header for each packet and
forward this packet to output port. Since tunnel-push action
contains most of packet header vswitchd needs to lookup routing
table and arp table to build this action.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Acked-by: Jarno Rajahalme <jrajahalme@nicira.com>
Acked-by: Thomas Graf <tgraf@noironetworks.com>
Acked-by: Ben Pfaff <blp@nicira.com>
In this patch, we add a lib/netdev-windows.c which mostly contains stub
code and in subsequent patches, would use the netlink interface to query
netdev information for a vport.
The code implements netdev functionality for "internal" and "system"
types of vports.
Signed-off-by: Nithin Raju <nithin@vmware.com>
Acked-by: Ankur Sharma <ankursharma@vmware.com>
Acked-by: Alin Gabriel Serdean <aserdean@cloudbasesolutions.com>
Tested-by: Alin Gabriel Serdean <aserdean@cloudbasesolutions.com>
Signed-off-by: Ben Pfaff <blp@nicira.com>
Reported-by: Daniel Badea <daniel.badea@windriver.com>
Signed-off-by: Alex Wang <alexw@nicira.com>
Acked-by: Daniele Di Proietto <ddiproietto@vmware.com>
This commit adds a new API to the 'struct netdev_class' which
allows user to configure the number of tx queues and rx queues
of 'netdev'. Upcoming patches will use this function to set
multiple tx/rx queues when adding the netdev to dpif-netdev.
Currently, only netdev-dpdk module implements this function.
Signed-off-by: Alex Wang <alexw@nicira.com>
Acked-by: Pravin B Shelar <pshelar@nicira.com>
There are couple of reasons to remove this support:
* This is used in very old OVS use-case. It is much better
to read stats directly from OVS.
* Forthcoming commit will remove support for setting stats
for vport. The stats update depends on stats-set.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Acked-by: Ben Pfaff <blp@nicira.com>
This commit adds new variable n_txq to 'struct netdev' for recording
the number of tx queues. Correspondingly, the send_*() functions are
extended to accept queue id as input argument.
All 'netdev-*' implementation will ignore the queue id since having
multiple tx queues is not supported. Upcomping patches will start
using it and create multiple tx queues for dpdk netdev.
Signed-off-by: Alex Wang <alexw@nicira.com>
Acked-by: Pravin B Shelar <pshelar@nicira.com>
This commit adds a new API to the 'struct netdev_class' which
allows user to query the numa node id the 'netdev' is on.
Currently, only netdev-dpdk module implements this function.
Signed-off-by: Alex Wang <alexw@nicira.com>
Acked-by: Pravin B Shelar <pshelar@nicira.com>
All access to struct netdev_registered_class ref_cnt member was done
with netdev_class_mutex held, so it does not need to be an atomic
variable.
Signed-off-by: Jarno Rajahalme <jrajahalme@nicira.com>
Acked-by: Ben Pfaff <blp@nicira.com>
Shared memory ring patch
This patch enables the client dpdk rings within the netdev-dpdk. It adds
a new dpdk device called dpdkr (other naming suggestions?). This allows
for the use of shared memory to communicate with other dpdk applications,
on the host or within a virtual machine. Instructions for use are in
INSTALL.DPDK.
This has been tested on Intel multi-core platforms and with the client
application within the host.
Signed-off-by: Gerald Rogers <gerald.rogers@intel.com>
Signed-off-by: Maryam Tahhan <maryam.tahhan@intel.com>
Acked-by: Pravin B Shelar <pshelar@nicira.com>
The netdev_send function has been modified to accept multiple packets, to
allow netdev providers to amortize locking and queuing costs.
This is especially true for netdev-dpdk.
Later commits exploit the new API.
Signed-off-by: Daniele Di Proietto <ddiproietto@vmware.com>
Acked-by: Pravin B Shelar <pshelar@nicira.com>
This commit introduces a new data structure used for receiving packets from
netdevs and passing them to dpifs.
The purpose of this change is to allow storing some private data for each
packet. The subsequent commits make use of it.
Signed-off-by: Daniele Di Proietto <ddiproietto@vmware.com>
Acked-by: Pravin B Shelar <pshelar@nicira.com>
In most cases, tunnel ports specify a dpif name to act as the backing
port in the datapath. However, in the case of UDP tunnels the type is
used with the port number appended. This is potentially a problem for
IPsec tunnels because they have different types but should have the
same backing port. The hasn't been a problem in practice though because
no UDP tunnels are currently used with IPsec.
This switches to use the dpif_port in all cases plus a port number if
necessary. It does this by making the names short enough to accomodate
ports, which also makes the naming more consistent.
Signed-off-by: Jesse Gross <jesse@nicira.com>
Acked-by: Ben Pfaff <blp@nicira.com>
Store the error condition of a failed port configuration in a new
column 'error' in the Interface table.
Example:
$ ovs-vsctl add-port br0 test -- \
set Interface test type=vxlan options:unknown=1
ovs-vsctl: Error detected while setting up 'test'. [...]
$ ovs-vsctl list Interface test | grep error
error : "test: could not set configuration (Invalid argument)"
Fixing the error will clear the error column:
$ ovs-vsctl set Interface test options:remote_ip=1.1.1.1
$ ovs-vsctl list Interface test | grep error
error : []
$
For now, the high level error messages when opening and configuring
the netdev are used. Further patches can extend passing the error
pointer into the individual netdev implementations to allow for more
fine grained error messages to be stored.
Signed-off-by: Thomas Graf <tgraf@redhat.com>
Signed-off-by: Ben Pfaff <blp@nicira.com>
When the user changes port type (i.e. changing p0 from type 'internal' to
'gre'), the netdev must first be deleted, then re-created with the new type.
Deleting the netdev requires there exist no more references to the netdev.
However, the xlate cache holds references to netdevs and the cache is only
invalidated by revalidator threads. Thus, if cache is not invalidated prior to
the netdev being re-created, the netdev will not be able to be re-created and
the configuration change will fail.
This patch always removes the netdev from the global netdev shash when the
user changes port type. This ensures that the new netdev can always be created
while handler and revalidator threads can retain references to the old netdev
until they are finished.
Signed-off-by: Ryan Wilson <wryan@nicira.com>
Signed-off-by: Ben Pfaff <blp@nicira.com>
netdev_open() would previously increment a netdev's refcount without
holding a lock for it. This commit shifts the locking to protect it.
Found by inspection.
Signed-off-by: Joe Stringer <joestringer@nicira.com>
Reviewed-by: YAMAMOTO Takashi <yamamoto@valinux.co.jp>
Acked-by: Ben Pfaff <blp@nicira.com>
netdev_rxq_open() open-codes much of netdev_ref(), so re-use that
function instead.
Signed-off-by: Joe Stringer <joestringer@nicira.com>
Reviewed-by: YAMAMOTO Takashi <yamamoto@valinux.co.jp>
Acked-by: Ben Pfaff <blp@nicira.com>
Commit 3e912ffcbb (netdev: Add 'change_seq' back to netdev.) added per-
netdev change number for indicating status change. Future commits used
this change number to optimize the netdev status update to database.
However, the work also introduced the bug in the following scenario:
- assume interface eth0 has address 1.2.3.4, eth1 has adddress 10.0.0.1.
- assume tunnel port p1 is set with remote_ip=10.0.0.5.
- after setup, 'ovs-vsctl list interface p1 status' should show the
'tunnel_egress_iface="eth1"'.
- now if the address of eth1 is change to 0 via 'ifconfig eth1 0'.
- expectedly, after change, 'ovs-vsctl list interface p1 status' should
show the 'tunnel_egress_iface="eth0"'
However, 'tunnel_egress_iface' will not be updated on current master.
This is in that, the 'netdev-vport' module corresponding to p1 does
not react to routing related changes.
To fix the bug, this commit adds a change sequence number in the route-
table module and makes netdev-vport check the sequence number for
tunnel status update.
Bug #1240626
Signed-off-by: Alex Wang <alexw@nicira.com>
Acked-by: Ben Pfaff <blp@nicira.com>
Commit 05bf6d3c62e1d (ovs-thread: Add checking for mutex and
rwlock initialization.) helps find an use of uninitialized
mutex (netdev_class_mutex) during upgrade. The assertion
check aborts the ovs.
This commit fixes the issue by adding the proper initialization.
Bug #1239914.
Bug #1240598.
Bug #1240626.
Signed-off-by: Alex Wang <alexw@nicira.com>
Acked-by: Ben Pfaff <blp@nicira.com>
This code path currently does not initialize
netdev_class_mutex.
dummy_enable
->netdev_dummy_register
->netdev_register_provider
->ovs_mutex_lock(&netdev_class_mutex)
ovsdb-server on windows crashes without it.
This commit adds a new initialization function.
Signed-off-by: Gurucharan Shetty <gshetty@nicira.com>
Acked-by: Ben Pfaff <blp@nicira.com>
This commit can be seen as a partial revert of commit
da4a619179d (netdev: Globally track port status changes)
by adding the 'change_seq' to 'struct netdev'.
Signed-off-by: Alex Wang <alexw@nicira.com>
Acked-by: Ben Pfaff <blp@nicira.com>
Following patch adds DPDK netdev-class to userspace datapath. Now
OVS can use DPDK port for IO by just configuring DPDK port and then
adding dpdk type port to userspace datapath.
Refer to INSTALL.DPDK doc for further info.
This is based a patch from Gerald Rogers.
Signed-off-by: Gerald Rogers <gerald.rogers@intel.com>
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Acked-by: Thomas Graf <tgraf@redhat.com>
new netdev type like DPDK can support multi-queue IO. Following
patch Adds support for same.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Acked-by: Thomas Graf <tgraf@redhat.com>
Preparation for multi queue netdev IO. There are no functional changes
in this patch.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Acked-by: Thomas Graf <tgraf@redhat.com>
This patch adds PMD type netdev for netdevice with poll-mode
drivers. Since there is no way to get signal on a packet recv
from these devices we need to poll them in busy loop. So minimize
system call overhead this patch uses dpif-thread exclusively
for PMD devices and rest of devices which needs system calls to
do IO are moved to dpif-netdev-run().
PMD device like DPDK work in userspace so there is no system call
overhead for them.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Acked-by: Thomas Graf <tgraf@redhat.com>
DPDK netdev need to access ofpbuf while sending buffer. Following
patch changes netdev_send accordingly.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Acked-by: Thomas Graf <tgraf@redhat.com>
DPDK can receive multiple packets but current netdev API does
not allow that. Following patch allows dpif-netdev receive batch
of packet in a rx_recv() call for any netdev port. This will be
used by dpdk-netdev.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
None of the atomic implementations need a destroy function anymore, so it's
"more standard" and more convenient for users to get rid of them.
Signed-off-by: Ben Pfaff <blp@nicira.com>
Acked-by: Andy Zhou <azhou@nicira.com>
The LINUX_DATAPATH C preprocessor symbol was originally meant to be used as
a signal for whether the Linux datapath module could be used, but it was
used as a proxy for a lot of other stuff that is really just Linux
specific. This commit switches all of these users to just test for
__linux__, which is more straightforward and should have the same result.
CC: Luigi Rizzo <rizzo@iet.unipi.it>
Signed-off-by: Ben Pfaff <blp@nicira.com>
With glibc, rwlocks by default allow recursive read-locking even if a
thread is blocked waiting for the write-lock. POSIX allows such attempts
to deadlock, and it appears that the libc used in NetBSD, at least, does
deadlock. The netdev_class_rwlock is in fact acquired recursively in this
way, which is a bug. This commit fixes the problem by switching to a
recursive mutex. This allows for less parallelism, but according to an
existing comment that doesn't matter here anyway.
Signed-off-by: Ben Pfaff <blp@nicira.com>
Acked-by: Joe Stringer <joestringer@nicira.com>
If VLAN acceleration is used when the kernel receives a packet
then the outer-most VLAN tag will not be present in the packet
when it is received by netdev-linux. Rather, it will be present
in auxdata.
This patch uses recvmsg() instead of recv() to read auxdata for
each packet and if the vlan_tid is set then it is added to the packet.
Adding the vlan_tid makes use of headroom available
in the buffer parameter of rx_recv.
Signed-off-by: Simon Horman <horms@verge.net.au>
Co-authored-by: Ben Pfaff <blp@nicira.com>
Signed-off-by: Ben Pfaff <blp@nicira.com>
Update the netdev_class so that struct ofpbuf * is passed to rx_recv()
to provide both the data and size of the data to read a packet into.
On success, update struct ofpbuf size inside netdev_class rx_recv
implementation and return 0. This moves logic from the caller.
On error a positive error code is returned, whereas previously
a negative error code was returned. This is a more common convention.
This patch should not have any behavioural changes.
This patch is in preparation for the netdev-linux variant of rx_recv()
making use of headroom in the struct ofpbuf * parameter to push a VLAN tag
obtained from auxdata.
Signed-off-by: Simon Horman <horms@verge.net.au>
Co-authored-by: Ben Pfaff <blp@nicira.com>
Signed-off-by: Ben Pfaff <blp@nicira.com>
Previously, we tracked status changes for ofports on a per-device basis.
Each time in the main thread's loop, we would inspect every ofport
to determine whether the status had changed for corresponding devices.
This patch replaces the per-netdev change_seq with a global 'struct seq'
which tracks status change for all ports. In the average case where
ports are not constantly going up or down, this allows us to check the
sequence once per main loop and not poll any ports. In the worst case,
execution is expected to be similar to how it is currently.
In a test environment of 5000 internal ports and 50 tunnel ports with
bfd, this reduces average CPU usage of the main thread from about 40% to
about 35%.
Signed-off-by: Joe Stringer <joestringer@nicira.com>
Signed-off-by: Ethan Jackson <ethan@nicira.com>
Acked-by: Ethan Jackson <ethan@nicira.com>
We have a call chain like this:
iface_configure_qos() calls
netdev_dump_queues(), which calls
netdev_linux_dump_queues(), which calls back through 'cb' to
qos_unixctl_show_cb(), which calls
netdev_delete_queue(), which calls
netdev_linux_delete_queue().
Both netdev_dump_queues() and netdev_linux_delete_queue() take the same
mutex in the same netdev, which deadlocks.
This commit fixes the problem by getting rid of the callback.
netdev_linux_dump_queue_stats() would benefit from the same treatment but
it's less urgent because I don't see any callbacks from that function that
call back into a netdev function.
Bug #19319.
Reported-by: Scott Hendricks <shendricks@vmware.com>
Signed-off-by: Ben Pfaff <blp@nicira.com>
This is the same lifecycle used in the ofproto provider interface.
Compared to the previous netdev provider interface, it has the
advantage that the netdev top layer can control when any given
netdev becomes visible to the outside world.
Signed-off-by: Ben Pfaff <blp@nicira.com>
This API change is necessary for thread safety, to be added in an upcoming
commit. Otherwise, the client would not be able to safely use the returned
netdev because it could already have been destroyed.
Signed-off-by: Ben Pfaff <blp@nicira.com>
Acked-by: Andy Zhou <azhou@nicira.com>
