[ovs-dev] [PATCH] Documentation: add notes for TSO & i40e
david.marchand at redhat.com
Mon Jan 13 15:57:43 UTC 2020
On Mon, Jan 13, 2020 at 4:49 PM Ciara Loftus <ciara.loftus at intel.com> wrote:
> When using TSO in OVS-DPDK with an i40e device, the following
> commit is required for DPDK, which fixes an issue on the TSO path:
Please avoid referencing dpdk-next-net sha1's, as this tree is rebased
on dpdk master quite frequently.
This commit will reach the master branch in a few days hopefully
(20.02-rc1 is expected this week).
In URLs, prefer https://.
> Document this as a limitation until a DPDK release with the fix
> included is supported by OVS.
> Also, document best known methods for performance tuning when
> testing TSO with the tool iperf.
> Signed-off-by: Ciara Loftus <ciara.loftus at intel.com>
> Documentation/topics/dpdk/tso.rst | 27 +++++++++++++++++++++++++++
> 1 file changed, 27 insertions(+)
> diff --git a/Documentation/topics/dpdk/tso.rst b/Documentation/topics/dpdk/tso.rst
> index 189c86480..5c56c458c 100644
> --- a/Documentation/topics/dpdk/tso.rst
> +++ b/Documentation/topics/dpdk/tso.rst
> @@ -94,3 +94,30 @@ datapath must support TSO or packets using that feature will be dropped
> on ports without TSO support. That also means guests using vhost-user
> in client mode will receive TSO packet regardless of TSO being enabled
> or disabled within the guest.
> +When the NIC performing the segmentation is using the i40e DPDK PMD, a fix
> +must be included in the DPDK build, otherwise TSO will not work. The fix can
> +be found on `the dpdk.org net-next tree`__.
> +__ http://git.dpdk.org/next/dpdk-next-net/commit/?id=e975e3720ba16278c6b43f8938b75710573bbd6a
> +This fix is expected to be included in the 19.11.1 release. When OVS migrates
> +to this DPDK release, this limitation can be removed.
> +Performance Tuning
> +iperf is often used to test TSO performance. Care needs to be taken when
> +configuring the environment in which the iperf server process is being run.
> +Since the iperf server uses the NIC's kernel driver, IRQs will be generated.
> +By default with some NICs eg. i40e, the IRQs will land on the same core as that
> +which is being used by the server process, provided the number of NIC queues is
> +greater or equal to that lcoreid. This causes contention between the iperf
> +server process and the IRQs. For optimal performance, it is suggested to pin
> +the IRQs to their own core. To change the affinity associated with a given IRQ
> +number, you can 'echo' the desired coremask to the file
> +For more on SMP affinity, refer to the `Linux kernel documentation`__.
> +__ https://www.kernel.org/doc/Documentation/IRQ-affinity.txt
> dev mailing list
> dev at openvswitch.org
More information about the dev