This patch extends the userspace datapaths support of tunnel tso from
only supporting VxLAN and Geneve to also supporting GRE tunnels. There
is also a software fallback for cases where the egress netdev does not
support this feature.
Reviewed-by: David Marchand <david.marchand@redhat.com>
Signed-off-by: Mike Pattrick <mkp@redhat.com>
Signed-off-by: Ilya Maximets <i.maximets@ovn.org>
For userspace datapath, this patch provides vxlan and geneve tunnel tso.
Only support userspace vxlan or geneve tunnel, meanwhile support
tunnel outter and inner csum offload. If netdev do not support offload
features, there is a software fallback.If netdev do not support vxlan
and geneve tso,packets will drop. Front-end devices can close offload
features by ethtool also.
Acked-by: Simon Horman <horms@ovn.org>
Signed-off-by: Dexia Li <dexia.li@jaguarmicro.com>
Co-authored-by: Mike Pattrick <mkp@redhat.com>
Signed-off-by: Mike Pattrick <mkp@redhat.com>
Signed-off-by: Ilya Maximets <i.maximets@ovn.org>
Currently, the netdev's speed is being calculated by taking the link's
feature bits (using netdev_get_features()) and transforming them into
bps.
This mechanism can be both inaccurate and difficult to maintain, mainly
because we currently use the feature bits supported by OpenFlow which
would have to be extended to support all new feature bits of all netdev
implementations while keeping the OpenFlow API intact.
In order to expose the link speed accurately for all current and future
hardware, add a new netdev API call that allows the implementations to
provide the current and maximum link speeds in Mbps.
Internally, the logic to get the maximum supported speed still relies on
feature bits so it might still get out of sync in the future. However,
the maximum configurable speed is not used as much as the current speed
and these feature bits are not exposed through the netdev interface so
it should be easier to add more.
Use this new function instead of netdev_get_features() where the link
speed is needed.
As a consequence of this patch, link speeds of cards is properly
reported (internally in OVSDB) even if not supported by OpenFlow.
A test verifies this behavior using a tap device.
Also, in order to avoid using the old, this patch adds a checkpatch.py
warning if the old API is used.
Reported-at: https://bugzilla.redhat.com/show_bug.cgi?id=2137567
Acked-by: Eelco Chaudron <echaudro@redhat.com>
Signed-off-by: Adrian Moreno <amorenoz@redhat.com>
Signed-off-by: Ilya Maximets <i.maximets@ovn.org>
OVS has support for using policing to enforce a rate limit in
kilobits per second. This is configured using OVSDB. f.e.
$ ovs-vsctl set interface tap0 ingress_policing_rate=1000
$ ovs-vsctl set interface tap0 ingress_policing_burst=100
This patch adds a related feature, allowing policing to enforce a rate
limit in kilo-packets per second. This is also configured using OVSDB.
$ ovs-vsctl set interface tap0 ingress_policing_kpkts_rate=1000
$ ovs-vsctl set interface tap0 ingress_policing_kpkts_burst=100
The kilo-bit and kilo-packet rate limits may be used separately or in
combination.
Add separate action for BPS and PPS in netlink message.
Revise code and change action result to pipe to allow
traffic pipe into second action.
This patch implements the feature for:
* OVSDB (northbound API)
* TC policer when used both with and without TC offload (kernel API)
Signed-off-by: Yong Xu <yong.xu@corigine.com>
Signed-off-by: Simon Horman <simon.horman@netronome.com>
Storing of the dpif type of the owning datapath interface will allow
us to easily distinguish, for example, userspace tunneling ports from
the system ones. This is required in terms of HW offloading to avoid
offloading of userspace flows to kernel interfaces that doesn't belong
to userspace datapath, but have same dpif_port names.
Acked-by: Eli Britstein <elibr@mellanox.com>
Acked-by: Roni Bar Yanai <roniba@mellanox.com>
Acked-by: Ophir Munk <ophirmu@mellanox.com>
Signed-off-by: Ilya Maximets <i.maximets@ovn.org>
The patch adds a new netdev class 'afxdp-nonpmd' to enable afxdp
interrupt mode. This is similar to 'type=afxdp', except that the
is_pmd field is set to false. As a result, the packet processing
is handled by main thread, not pmd thread. This avoids burning
the CPU to always 100% when there is no traffic.
Signed-off-by: William Tu <u9012063@gmail.com>
Signed-off-by: Ilya Maximets <i.maximets@ovn.org>
Ideally SCTP checksum offload needs be advertised by the
NIC when userspace TSO is enabled. However, very few drivers
do that and it's not a widely used protocol. So, this patch
enables SCTP checksum offload if available, otherwise userspace
TSO can still be enabled but SCTP packets will be dropped on
NICs without support.
Fixes: 29cf9c1b3b9c ("userspace: Add TCP Segmentation Offload support")
Signed-off-by: Flavio Leitner <fbl@sysclose.org>
Signed-off-by: Ilya Maximets <i.maximets@ovn.org>
Virtio doesn't expose flags to control which protocols checksum
offload needs to be enabled or disabled. This patch checks if the
NIC supports UDP checksum offload and active it when TSO is enabled.
Reported-by: Ilya Maximets <i.maximets@ovn.org>
Fixes: 29cf9c1b3b9c ("userspace: Add TCP Segmentation Offload support")
Signed-off-by: Flavio Leitner <fbl@sysclose.org>
Signed-off-by: Ilya Maximets <i.maximets@ovn.org>
Abbreviated as TSO, TCP Segmentation Offload is a feature which enables
the network stack to delegate the TCP segmentation to the NIC reducing
the per packet CPU overhead.
A guest using vhostuser interface with TSO enabled can send TCP packets
much bigger than the MTU, which saves CPU cycles normally used to break
the packets down to MTU size and to calculate checksums.
It also saves CPU cycles used to parse multiple packets/headers during
the packet processing inside virtual switch.
If the destination of the packet is another guest in the same host, then
the same big packet can be sent through a vhostuser interface skipping
the segmentation completely. However, if the destination is not local,
the NIC hardware is instructed to do the TCP segmentation and checksum
calculation.
It is recommended to check if NIC hardware supports TSO before enabling
the feature, which is off by default. For additional information please
check the tso.rst document.
Signed-off-by: Flavio Leitner <fbl@sysclose.org>
Tested-by: Ciara Loftus <ciara.loftus.intel.com>
Signed-off-by: Ian Stokes <ian.stokes@intel.com>
"rxq_enabled" of netdev is writen in the vhost thread and read by pmd
thread once it observes 'change_seq' is updated. This patch is to keep
order on aarch64 or other weak memory model CPU to ensure 'rxq_enabled' is
observed before 'change_seq'.
Reviewed-by: Gavin Hu <Gavin.Hu@arm.com>
Signed-off-by: Yanqin Wei <Yanqin.Wei@arm.com>
Signed-off-by: Ben Pfaff <blp@ovn.org>
The patch introduces experimental AF_XDP support for OVS netdev.
AF_XDP, the Address Family of the eXpress Data Path, is a new Linux socket
type built upon the eBPF and XDP technology. It is aims to have comparable
performance to DPDK but cooperate better with existing kernel's networking
stack. An AF_XDP socket receives and sends packets from an eBPF/XDP program
attached to the netdev, by-passing a couple of Linux kernel's subsystems
As a result, AF_XDP socket shows much better performance than AF_PACKET
For more details about AF_XDP, please see linux kernel's
Documentation/networking/af_xdp.rst. Note that by default, this feature is
not compiled in.
Signed-off-by: William Tu <u9012063@gmail.com>
Signed-off-by: Ilya Maximets <i.maximets@samsung.com>
We currently poll all available queues based on the max queue count
exchanged with the vhost peer and rely on the vhost library in DPDK to
check the vring status beneath.
This can lead to some overhead when we have a lot of unused queues.
To enhance the situation, we can skip the disabled queues.
On rxq notifications, we make use of the netdev's change_seq number so
that the pmd thread main loop can cache the queue state periodically.
$ ovs-appctl dpif-netdev/pmd-rxq-show
pmd thread numa_id 0 core_id 1:
isolated : true
port: dpdk0 queue-id: 0 (enabled) pmd usage: 0 %
pmd thread numa_id 0 core_id 2:
isolated : true
port: vhost1 queue-id: 0 (enabled) pmd usage: 0 %
port: vhost3 queue-id: 0 (enabled) pmd usage: 0 %
pmd thread numa_id 0 core_id 15:
isolated : true
port: dpdk1 queue-id: 0 (enabled) pmd usage: 0 %
pmd thread numa_id 0 core_id 16:
isolated : true
port: vhost0 queue-id: 0 (enabled) pmd usage: 0 %
port: vhost2 queue-id: 0 (enabled) pmd usage: 0 %
$ while true; do
ovs-appctl dpif-netdev/pmd-rxq-show |awk '
/port: / {
tot++;
if ($5 == "(enabled)") {
en++;
}
}
END {
print "total: " tot ", enabled: " en
}'
sleep 1
done
total: 6, enabled: 2
total: 6, enabled: 2
...
# Started vm, virtio devices are bound to kernel driver which enables
# F_MQ + all queue pairs
total: 6, enabled: 2
total: 66, enabled: 66
...
# Unbound vhost0 and vhost1 from the kernel driver
total: 66, enabled: 66
total: 66, enabled: 34
...
# Configured kernel bound devices to use only 1 queue pair
total: 66, enabled: 34
total: 66, enabled: 19
total: 66, enabled: 4
...
# While rebooting the vm
total: 66, enabled: 4
total: 66, enabled: 2
...
total: 66, enabled: 66
...
# After shutting down the vm
total: 66, enabled: 66
total: 66, enabled: 2
Signed-off-by: David Marchand <david.marchand@redhat.com>
Acked-by: Ilya Maximets <i.maximets@samsung.com>
Signed-off-by: Ian Stokes <ian.stokes@intel.com>
New module 'netdev-offload' created to manage different flow API
implementations. All the generic and provider independent code moved
there from the 'netdev' module.
Flow API providers further encapsulated.
The only function that was changed is 'netdev_any_oor'.
Now it uses offloading related hmap instead of common 'netdev_shash'.
Signed-off-by: Ilya Maximets <i.maximets@samsung.com>
Acked-by: Ben Pfaff <blp@ovn.org>
Acked-by: Roi Dayan <roid@mellanox.com>
Current issues with Flow API:
* OVS calls offloading functions regardless of successful
flow API initialization. (ex. on init_flow_api failure)
* Static initilaization of Flow API for a netdev_class forbids
having different offloading types for different instances
of netdev with the same netdev_class. (ex. different vports in
'system' and 'netdev' datapaths at the same time)
Solution:
* Move Flow API from the netdev_class to netdev instance.
* Make Flow API dynamic, i.e. probe the APIs and choose the
suitable one.
Side effects:
* Flow API providers localized as possible in their modules.
* Now we have an ability to make runtime checks. For example,
we could check if particular device supports features we
need, like if dpdk device supports RSS+MARK action.
Signed-off-by: Ilya Maximets <i.maximets@samsung.com>
Acked-by: Roi Dayan <roid@mellanox.com>
This is the third patch in the patch-set to support dynamic rebalancing
of offloaded flows.
The dynamic rebalancing functionality is implemented in this patch. The
ukeys that are not scheduled for deletion are obtained and passed as input
to the rebalancing routine. The rebalancing is done in the context of
revalidation leader thread, after all other revalidator threads are
done with gathering rebalancing data for flows.
For each netdev that is in OOR state, a list of flows - both offloaded
and non-offloaded (pending) - is obtained using the ukeys. For each netdev
that is in OOR state, the flows are grouped and sorted into offloaded and
pending flows. The offloaded flows are sorted in descending order of
pps-rate, while pending flows are sorted in ascending order of pps-rate.
The rebalancing is done in two phases. In the first phase, we try to
offload all pending flows and if that succeeds, the OOR state on the device
is cleared. If some (or none) of the pending flows could not be offloaded,
then we start replacing an offloaded flow that has a lower pps-rate than
a pending flow, until there are no more pending flows with a higher rate
than an offloaded flow. The flows that are replaced from the device are
added into kernel datapath.
A new OVS configuration parameter "offload-rebalance", is added to ovsdb.
The default value of this is "false". To enable this feature, set the
value of this parameter to "true", which provides packets-per-second
rate based policy to dynamically offload and un-offload flows.
Note: This option can be enabled only when 'hw-offload' policy is enabled.
It also requires 'tc-policy' to be set to 'skip_sw'; otherwise, flow
offload errors (specifically ENOSPC error this feature depends on) reported
by an offloaded device are supressed by TC-Flower kernel module.
Signed-off-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
Co-authored-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Reviewed-by: Sathya Perla <sathya.perla@broadcom.com>
Reviewed-by: Ben Pfaff <blp@ovn.org>
Signed-off-by: Simon Horman <simon.horman@netronome.com>
This is the first patch in the patch-set to support dynamic rebalancing
of offloaded flows.
The patch detects OOR condition on a netdev port when ENOSPC error is
returned by TC-Flower while adding a flow rule. A new structure is added
to the netdev called "netdev_hw_info", to store OOR related information
required to perform dynamic offload-rebalancing.
Signed-off-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
Co-authored-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Reviewed-by: Sathya Perla <sathya.perla@broadcom.com>
Reviewed-by: Ben Pfaff <blp@ovn.org>
Signed-off-by: Simon Horman <simon.horman@netronome.com>
The macros are hard to read. This makes it a little more readable.
Signed-off-by: Ben Pfaff <blp@ovn.org>
Signed-off-by: Ian Stokes <ian.stokes@intel.com>
Add a new class op for netdevs to get the block_id if one exists. The
block_id is used in offload ops to group multiple qdiscs together.
Stub calls are made to the new class op (implementation to follow in
further patches). The default block_id of 0 (no block) will be used in
these cases.
Signed-off-by: John Hurley <john.hurley@netronome.com>
Reviewed-by: Dirk van der Merwe <dirk.vandermerwe@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Simon Horman <simon.horman@netronome.com>
Previously, any rule that is offloaded via a netdev, not necessarily
to the HW, would be reported as "offloaded". This patch fixes this
misalignment, and introduces the 'dp' state, as follows:
rule is in HW via TC offload -> offloaded=yes dp:tc
rule is in not HW over TC DP -> offloaded=no dp:tc
rule is in not HW over OVS DP -> offloaded=no dp:ovs
To achieve this, the flows's 'offloaded' flag was encapsulated in a new
attrs struct, which contains the offloaded state of the flow and the
DP layer the flow is handled in, and instead of setting the flow's
'offloaded' state based solely on the type of dump it was acquired
via, for netdev flows it now sends the new attrs struct to be
collected along with the rest of the flow via the netdev, allowing
it to be set per flow.
For TC offloads, the offloaded state is set based on the 'in_hw' and
'not_in_hw' flags received from the TC as part of the flower. If no
such flag was received, due to lack of kernel support, it defaults
to true.
Signed-off-by: Gavi Teitz <gavi@mellanox.com>
Acked-by: Roi Dayan <roid@mellanox.com>
[simon: resolved conflict in lib/dpctl.man]
Signed-off-by: Simon Horman <simon.horman@netronome.com>
The patch adds additional 'struct netdev *' to the
native tunnel's push_header() interface. This is used
for later GRE sequence number support.
Signed-off-by: William Tu <u9012063@gmail.com>
Signed-off-by: Ben Pfaff <blp@ovn.org>
If the caller provides a non-NULL qfill pointer and the netdev
implemementation supports reading the rx queue fill level, the rxq_recv()
function returns the remaining number of packets in the rx queue after
reception of the packet burst to the caller. If the implementation does
not support this, it returns -ENOTSUP instead. Reading the remaining queue
fill level should not substantilly slow down the recv() operation.
A first implementation is provided for ethernet and vhostuser DPDK ports
in netdev-dpdk.c.
This output parameter will be used in the upcoming commit for PMD
performance metrics to supervise the rx queue fill level for DPDK
vhostuser ports.
Signed-off-by: Jan Scheurich <jan.scheurich@ericsson.com>
Acked-by: Billy O'Mahony <billy.o.mahony@intel.com>
Signed-off-by: Ian Stokes <ian.stokes@intel.com>
- New get_custom_stats interface function is added to netdev. It
allows particular netdev implementation to expose custom
counters in dictionary format (counter name/counter value).
- New statistics are retrieved using experimenter code and
are printed as a result to ofctl dump-ports.
- New counters are available for OpenFlow 1.4+.
- New statistics are printed to output via ofctl only if those
are present in reply message.
- New statistics definition is added to include/openflow/intel-ext.h.
- Custom statistics are implemented only for dpdk-physical
port type.
- DPDK-physical implementation uses xstats to collect statistics.
Only dropped and error counters are exposed.
Co-authored-by: Ben Pfaff <blp@ovn.org>
Signed-off-by: Ben Pfaff <blp@ovn.org>
Signed-off-by: Michal Weglicki <michalx.weglicki@intel.com>
Signed-off-by: Ben Pfaff <blp@ovn.org>
Not needed anymore because 'may_steal' already handled on
dpif-netdev layer and always true.
Acked-by: Eelco Chaudron <echaudro@redhat.com>
Signed-off-by: Ilya Maximets <i.maximets@samsung.com>
Signed-off-by: Ian Stokes <ian.stokes@intel.com
By reordering elements in netdev structure, holes can be removed.
Before: structure size: 88, sum holes: 10, cachelines:2
After : structure size: 80, sum holes: 2, cachelines:2
Signed-off-by: Bhanuprakash Bodireddy <bhanuprakash.bodireddy@intel.com>
Signed-off-by: Ben Pfaff <blp@ovn.org>
By reordering bool in netdev_flow_dump structure, pad bytes can be
reduced.
Before: structure size: 32, sum holes: 4, sum padbytes:7, cachelines:1
After : structure size: 24, sum holes: 3, sum padbytes:0, cachelines:1
Signed-off-by: Bhanuprakash Bodireddy <bhanuprakash.bodireddy@intel.com>
Signed-off-by: Ben Pfaff <blp@ovn.org>
Due to commit 67ac844 an existing issue with OVS persisten ports
surfaced. If we revert the commit we no longer get the error, and
basic traffic will flow. However the wrong netdev class is used, hence
the wrong callbacks get called.
The main issue is with netdev_open() being called with type = NULL
before the interface is actually configured in the system. This patch
tracks these "auto" generated interfaces, and once netdev_open() gets
called with a valid type, re-configures (re-create) it.
Signed-off-by: Eelco Chaudron <echaudro@redhat.com>
Signed-off-by: Ben Pfaff <blp@ovn.org>
In netdev_gre_build_header(), GRE protocol and VXLAN next_potocol is set based
on packet_type of flow. If it's about an Ethernet packet, it is set to
ETP_TYPE_TEB. Otherwise, if the name space is OFPHTN_ETHERNET, it is set
according to the name space type.
Signed-off-by: Jan Scheurich <jan.scheurich@ericsson.com>
Signed-off-by: Ben Pfaff <blp@ovn.org>
Add a new API interface for offloading dpif flows to netdev.
The API consist on the following:
flow_put - offload a new flow
flow_get - query an offloaded flow
flow_del - delete an offloaded flow
flow_flush - flush all offloaded flows
flow_dump_* - dump all offloaded flows
In upcoming commits we will introduce an implementation of this
API for netdev-linux.
Signed-off-by: Paul Blakey <paulb@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Acked-by: Flavio Leitner <fbl@sysclose.org>
Signed-off-by: Simon Horman <simon.horman@netronome.com>
Since 55e075e65ef9("netdev-dpdk: Arbitrary 'dpdk' port naming"),
set_config() is used to identify a DPDK device, so it's better to report
its detailed error message to the user. Tunnel devices and patch ports
rely a lot on set_config() as well.
This commit adds a param to set_config() that can be used to return
an error message and makes use of that in netdev-dpdk and netdev-vport.
Before this patch:
$ ovs-vsctl add-port br0 dpdk0 -- set Interface dpdk0 type=dpdk
ovs-vsctl: Error detected while setting up 'dpdk0': dpdk0: could not set
configuration (Invalid argument). See ovs-vswitchd log for details.
ovs-vsctl: The default log directory is "/var/log/openvswitch/".
$ ovs-vsctl add-port br0 p+ -- set Interface p+ type=patch
ovs-vsctl: Error detected while setting up 'p+': p+: could not set
configuration (Invalid argument). See ovs-vswitchd log for details.
ovs-vsctl: The default log directory is "/var/log/openvswitch/".
$ ovs-vsctl add-port br0 gnv0 -- set Interface gnv0 type=geneve
ovs-vsctl: Error detected while setting up 'gnv0': gnv0: could not set
configuration (Invalid argument). See ovs-vswitchd log for details.
ovs-vsctl: The default log directory is "/var/log/openvswitch/".
After this patch:
$ ovs-vsctl add-port br0 dpdk0 -- set Interface dpdk0 type=dpdk
ovs-vsctl: Error detected while setting up 'dpdk0': 'dpdk0' is missing
'options:dpdk-devargs'. The old 'dpdk<port_id>' names are not
supported. See ovs-vswitchd log for details.
ovs-vsctl: The default log directory is "/var/log/openvswitch/".
$ ovs-vsctl add-port br0 p+ -- set Interface p+ type=patch
ovs-vsctl: Error detected while setting up 'p+': p+: patch type requires
valid 'peer' argument. See ovs-vswitchd log for details.
ovs-vsctl: The default log directory is "/var/log/openvswitch/".
$ ovs-vsctl add-port br0 gnv0 -- set Interface gnv0 type=geneve
ovs-vsctl: Error detected while setting up 'gnv0': gnv0: geneve type
requires valid 'remote_ip' argument. See ovs-vswitchd log for
details.
ovs-vsctl: The default log directory is "/var/log/openvswitch/".
CC: Ciara Loftus <ciara.loftus@intel.com>
CC: Kevin Traynor <ktraynor@redhat.com>
Signed-off-by: Daniele Di Proietto <diproiettod@vmware.com>
Acked-by: Kevin Traynor <ktraynor@redhat.com>
Tested-by: Ciara Loftus <ciara.loftus@intel.com>
By default Open vSwitch tries to configure internal interfaces MTU to
match the bridge minimum, overriding any attempt by the user to
configure it through standard system tools, or the database.
While this works in many simple cases (there are probably many users
that rely on this) it may create problems for more advanced use cases
(like any overlay networks).
This commit allows the user to override the default behavior by
providing an explict MTU in the mtu_request column in the Interface
table.
This means that Open vSwitch will now treat differently database MTU
requests from standard system tools MTU requests (coming from `ip link`
or `ifconfig`), but this seems the best way to remain compatible with
old users while providing a more powerful interface.
Suggested-by: Darrell Ball <dlu998@gmail.com>
Signed-off-by: Daniele Di Proietto <diproiettod@vmware.com>
Acked-by: Ben Pfaff <blp@ovn.org>
Tested-by: Joe Stringer <joe@ovn.org>
This will allow run() and wait() methods to be shared between different
classes and still perform class-specific work.
Signed-off-by: Daniele Di Proietto <diproiettod@vmware.com>
Acked-by: Ben Pfaff <blp@ovn.org>
Every provider silently drops the const attribute when converting the
parameter to the appropriate subclass. Might as well drop the const
attribute from the parameter, since this is a "set" function.
Signed-off-by: Daniele Di Proietto <diproiettod@vmware.com>
Acked-by: Ilya Maximets <i.maximets@samsung.com>
Commit 64839cf43 applies batch objects to netdev-providers, but
some comments were not updated accordingly. Fix these:
- replace 'pkts' with 'batch'
- replace '*cnt' with 'batch->count'
- replace MAX_RX_BATCH with NETDEV_MAX_BURST
- remove superfluous whitespace
Signed-off-by: Mark Kavanagh <mark.b.kavanagh@intel.com>
Acked-by: William Tu <u9012063@gmail.com>
Acked-by: Daniele Di Proietto <diproiettod@vmware.com>
If CPU number in pmd-cpu-mask is not divisible by the number of queues and
in a few more complex situations there may be unfair distribution of TX
queue-ids between PMD threads.
For example, if we have 2 ports with 4 queues and 6 CPUs in pmd-cpu-mask
such distribution is possible:
<------------------------------------------------------------------------>
pmd thread numa_id 0 core_id 13:
port: vhost-user1 queue-id: 1
port: dpdk0 queue-id: 3
pmd thread numa_id 0 core_id 14:
port: vhost-user1 queue-id: 2
pmd thread numa_id 0 core_id 16:
port: dpdk0 queue-id: 0
pmd thread numa_id 0 core_id 17:
port: dpdk0 queue-id: 1
pmd thread numa_id 0 core_id 12:
port: vhost-user1 queue-id: 0
port: dpdk0 queue-id: 2
pmd thread numa_id 0 core_id 15:
port: vhost-user1 queue-id: 3
<------------------------------------------------------------------------>
As we can see above dpdk0 port polled by threads on cores:
12, 13, 16 and 17.
By design of dpif-netdev, there is only one TX queue-id assigned to each
pmd thread. This queue-id's are sequential similar to core-id's. And
thread will send packets to queue with exact this queue-id regardless
of port.
In previous example:
pmd thread on core 12 will send packets to tx queue 0
pmd thread on core 13 will send packets to tx queue 1
...
pmd thread on core 17 will send packets to tx queue 5
So, for dpdk0 port after truncating in netdev-dpdk:
core 12 --> TX queue-id 0 % 4 == 0
core 13 --> TX queue-id 1 % 4 == 1
core 16 --> TX queue-id 4 % 4 == 0
core 17 --> TX queue-id 5 % 4 == 1
As a result only 2 of 4 queues used.
To fix this issue some kind of XPS implemented in following way:
* TX queue-ids are allocated dynamically.
* When PMD thread first time tries to send packets to new port
it allocates less used TX queue for this port.
* PMD threads periodically performes revalidation of
allocated TX queue-ids. If queue wasn't used in last
XPS_TIMEOUT_MS milliseconds it will be freed while revalidation.
* XPS is not working if we have enough TX queues.
Reported-by: Zhihong Wang <zhihong.wang@intel.com>
Signed-off-by: Ilya Maximets <i.maximets@samsung.com>
Signed-off-by: Daniele Di Proietto <diproiettod@vmware.com>
To easily allow both in- and out-of-tree building of the Python
wrapper for the OVS JSON parser (e.g. w/ pip), move json.h to
include/openvswitch. This also requires moving lib/{hmap,shash}.h.
Both hmap.h and shash.h were #include-ing "util.h" even though the
headers themselves did not use anything from there, but rather from
include/openvswitch/util.h. Fixing that required including util.h
in several C files mostly due to OVS_NOT_REACHED and things like
xmalloc.
Signed-off-by: Terry Wilson <twilson@redhat.com>
Signed-off-by: Ben Pfaff <blp@ovn.org>
Commit 1895cc8dbb64 ("dpif-netdev: create batch object") introduces
batch process functions and 'struct dp_packet_batch' to associate with
batch-level metadata. This patch applies the packet batch object to
the netdev provider interface (dummy, Linux, BSD, and DPDK) so that
batch APIs can be used in providers. With batch metadata visible in
providers, optimizations can be introduced at per-batch level instead
of per-packet.
Tested-at: https://travis-ci.org/williamtu/ovs-travis/builds/145694197
Signed-off-by: William Tu <u9012063@gmail.com>
Signed-off-by: Daniele Di Proietto <diproiettod@vmware.com>
The native tunneling build tunnel header code is spread across
two different modules, it makes pretty hard to follow the code.
Following patch refactors the code to move all code to
netdev-ative-tnl module.
Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
This introduces in dpif-netdev and netdev-dpdk the first use for the
newly introduce reconfigure netdev call.
When a request to change the number of queues comes, netdev-dpdk will
remember this and notify the upper layer via
netdev_request_reconfigure().
The datapath, instead of periodically calling netdev_set_multiq(), can
detect this and call reconfigure().
This mechanism can also be used to:
* Automatically match the number of rxq with the one provided by qemu
via the new_device callback.
* Provide a way to change the MTU of dpdk devices at runtime.
* Move a DPDK vhost device to the proper NUMA socket.
Signed-off-by: Daniele Di Proietto <diproiettod@vmware.com>
Tested-by: Ilya Maximets <i.maximets@samsung.com>
Acked-by: Ilya Maximets <i.maximets@samsung.com>
A netdev provider, especially a PMD provider (like netdev DPDK) might
not be able to change some of its parameters (such as MTU, or number of
queues) without stopping everything and restarting.
This commit introduces a mechanism that allows a netdev provider to
request a restart (netdev_request_reconfigure()). The upper layer can
be notified via netdev_wait_reconf_required() and
netdev_is_reconf_required(). After closing all the rxqs the upper layer
can finally call netdev_reconfigure(), to make sure that the new
configuration is in place.
This will be used by next commit to reconfigure rx and tx queues in
netdev-dpdk.
Signed-off-by: Daniele Di Proietto <diproiettod@vmware.com>
Tested-by: Ilya Maximets <i.maximets@samsung.com>
Acked-by: Ilya Maximets <i.maximets@samsung.com>
Acked-by: Mark Kavanagh <mark.b.kavanagh@intel.com>
Current tunnel-pop API does not allow the netdev implementation
retain a packet but STT can keep a packet from batch of packets
during TCP reassembly processing. To return exact count of
valid packet STT need to pass this number of packet parameter
as a reference.
Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
It is better to move tunnel push-pop action specific functions into
separate module.
Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>
All code is now in include/openvswitch/list.h.
Signed-off-by: Ben Warren <ben@skyportsystems.com>
Acked-by: Ryan Moats <rmoats@us.ibm.com>
Signed-off-by: Ben Pfaff <blp@ovn.org>
Since netdev can have multiple IP address use
generic api netdev_get_addr_list(). This also make it
easier to handle IPv4 and IPv6 address across vswitchd
layers.
Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Ben Pfaff <blp@ovn.org>
Device can have multiple IP address but netdev_get_in4/6()
returns only one configured IPv6 address. Following
patch fixes it.
OVS router is also updated to return source ip address for
given destination, This is required when interface has multiple
IP address configured.
Signed-off-by: Pravin B Shelar <pshelar@ovn.org>
Acked-by: Ben Pfaff <blp@ovn.org>
Made to simplify creation of derived classes.
Signed-off-by: Ilya Maximets <i.maximets@samsung.com>
Acked-by: Daniele Di Proietto <diproiettod@vmware.com>
The comment was incomplete in some ways and simply wrong in others.
Also ensure that *cnt is set to 0 if an error is encountered. It's nice
when callers can rely on this.
Signed-off-by: Ben Pfaff <blp@ovn.org>
Acked-by: Justin Pettit <jpettit@ovn.org>
This manifested as a memory leak in test 898 "ofproto-dpif - sFlow packet
sampling - tunnel set", which included an output to a tunnel vport that
doesn't have an implementation of netdev_send().
Reported-by: William Tu <u9012063@gmail.com>
Reported-at: http://openvswitch.org/pipermail/dev/2016-February/065873.html
Signed-off-by: Ben Pfaff <blp@ovn.org>
Currently, all of the PMD netdevs can only have the same number of
rx queues, which is specified in other_config:n-dpdk-rxqs.
Fix that by introducing of new option for PMD interfaces: 'n_rxq', which
specifies the maximum number of rx queues to be created for this
interface.
Example:
ovs-vsctl set Interface dpdk0 options:n_rxq=8
Old 'other_config:n-dpdk-rxqs' deleted.
Signed-off-by: Ilya Maximets <i.maximets@samsung.com>
Acked-by: Ben Pfaff <blp@ovn.org>
Acked-by: Flavio Leitner <fbl@sysclose.org>
Signed-off-by: Daniele Di Proietto <diproiettod@vmware.com>
Define struct eth_addr and use it instead of a uint8_t array for all
ethernet addresses in OVS userspace. The struct is always the right
size, and it can be assigned without an explicit memcpy, which makes
code more readable.
"struct eth_addr" is a good type name for this as many utility
functions are already named accordingly.
struct eth_addr can be accessed as bytes as well as ovs_be16's, which
makes the struct 16-bit aligned. All use seems to be 16-bit aligned,
so some algorithms on the ethernet addresses can be made a bit more
efficient making use of this fact.
As the struct fits into a register (in 64-bit systems) we pass it by
value when possible.
This patch also changes the few uses of Linux specific ETH_ALEN to
OVS's own ETH_ADDR_LEN, and removes the OFP_ETH_ALEN, as it is no
longer needed.
This work stemmed from a desire to make all struct flow members
assignable for unrelated exploration purposes. However, I think this
might be a nice code readability improvement by itself.
Signed-off-by: Jarno Rajahalme <jrajahalme@nicira.com>
