[dpdk-dev] [PATCH 00/13] add hairpin feature
orika at mellanox.com
Thu Sep 26 08:28:56 CEST 2019
This patch set implements the hairpin feature.
The hairpin feature was introduced in RFC
The hairpin feature (different name can be forward) acts as "bump on the wire",
meaning that a packet that is received from the wire can be modified using
offloaded action and then sent back to the wire without application intervention
which save CPU cycles.
The hairpin is the inverse function of loopback in which application
sends a packet then it is received again by the
application without being sent to the wire.
The hairpin can be used by a number of different NVF, for example load
balancer, gateway and so on.
As can be seen from the hairpin description, hairpin is basically RX queue
connected to TX queue.
During the design phase I was thinking of two ways to implement this
feature the first one is adding a new rte flow action. and the second
one is create a special kind of queue.
The advantages of using the queue approch:
1. More control for the application. queue depth (the memory size that
should be used).
2. Enable QoS. QoS is normaly a parametr of queue, so in this approch it
will be easy to integrate with such system.
3. Native integression with the rte flow API. Just setting the target
queue/rss to hairpin queue, will result that the traffic will be routed
to the hairpin queue.
4. Enable queue offloading.
Each hairpin Rxq can be connected Txq / number of Txqs which can belong to a
different ports assuming the PMD supports it. The same goes the other
way each hairpin Txq can be connected to one or more Rxqs.
This is the reason that both the Txq setup and Rxq setup are getting the
hairpin configuration structure.
>From PMD prespctive the number of Rxq/Txq is the total of standard
queues + hairpin queues.
To configure hairpin queue the user should call
rte_eth_rx_hairpin_queue_setup / rte_eth_tx_hairpin_queue_setup insteed
of the normal queue setup functions.
The hairpin queues are not part of the normal RSS functiosn.
To use the queues the user simply create a flow that points to RSS/queue
actions that are hairpin queues.
The reason for selecting 2 new functions for hairpin queue setup are:
1. avoid API break.
2. avoid extra and unused parameters.
This series must be applied after series
Cc: wenzhuo.lu at intel.com
Cc: bernard.iremonger at intel.com
Cc: thomas at monjalon.net
Cc: ferruh.yigit at intel.com
Cc: arybchenko at solarflare.com
Cc: viacheslavo at mellanox.com
Ori Kam (13):
ethdev: support setup function for hairpin queue
net/mlx5: query hca hairpin capabilities
net/mlx5: support Rx hairpin queues
net/mlx5: prepare txq to work with different types
net/mlx5: support Tx hairpin queues
app/testpmd: add hairpin support
net/mlx5: add hairpin binding function
net/mlx5: add support for hairpin hrxq
net/mlx5: add internal tag item and action
net/mlx5: add id generation function
net/mlx5: add default flows for hairpin
net/mlx5: split hairpin flows
doc: add hairpin feature
app/test-pmd/parameters.c | 12 +
app/test-pmd/testpmd.c | 59 ++++-
app/test-pmd/testpmd.h | 1 +
doc/guides/rel_notes/release_19_11.rst | 5 +
drivers/net/mlx5/mlx5.c | 160 ++++++++++++-
drivers/net/mlx5/mlx5.h | 65 ++++-
drivers/net/mlx5/mlx5_devx_cmds.c | 194 +++++++++++++++
drivers/net/mlx5/mlx5_flow.c | 393 ++++++++++++++++++++++++++++++-
drivers/net/mlx5/mlx5_flow.h | 73 +++++-
drivers/net/mlx5/mlx5_flow_dv.c | 231 +++++++++++++++++-
drivers/net/mlx5/mlx5_flow_verbs.c | 11 +-
drivers/net/mlx5/mlx5_prm.h | 127 +++++++++-
drivers/net/mlx5/mlx5_rxq.c | 323 ++++++++++++++++++++++---
drivers/net/mlx5/mlx5_rxtx.c | 2 +-
drivers/net/mlx5/mlx5_rxtx.h | 72 +++++-
drivers/net/mlx5/mlx5_trigger.c | 134 ++++++++++-
drivers/net/mlx5/mlx5_txq.c | 309 ++++++++++++++++++++----
lib/librte_ethdev/rte_ethdev.c | 213 +++++++++++++++++
lib/librte_ethdev/rte_ethdev.h | 145 ++++++++++++
lib/librte_ethdev/rte_ethdev_core.h | 18 ++
lib/librte_ethdev/rte_ethdev_version.map | 4 +
21 files changed, 2424 insertions(+), 127 deletions(-)
More information about the dev