DPDK patches and discussions
 help / color / mirror / Atom feed
From: Fried Onions <yummy.fried.onions@gmail.com>
To: dev@dpdk.org
Subject: [dpdk-dev] i40e PMD Linux bonding issue with VFs
Date: Fri, 28 May 2021 18:29:00 +0800	[thread overview]
Message-ID: <CAMfW34QFeXFu-DDJbMjCK5zafp5uZU8svgjGPRktdn+9bEHwTw@mail.gmail.com> (raw)

I came across the patch for the i40e PMD that allows VF to set the MAC
address even if the PF had already done so in
https://mails.dpdk.org/archives/dev/2020-March/160631.html
<https://urldefense.com/v3/__https://mails.dpdk.org/archives/dev/2020-March/160631.html__;!!JYsgTRAg6ZQ!dfFxEpVAzspiJP_FtoL7Uqk4LEkqFvvgp-_FlJ9zsPvyjhnnka0JM9_QdscYtStjkg$>

I tried backporting it to DPDK 18.11's i40e PMD and it fixes the issue of
allowing the VF to set the MAC address.

However, Linux bonding is still not working because the values don't seem
to be updated at host level.

Would appreciate it if anyone has any idea what other patches are needed
for Linux bonding with VFs to work on the i40e PMD in DPDK18.11 in terms of
patches to the PMD or firmware upgrades on the XL710 card? Or if Linux
bonding can be supported at all with the i40e PMD with VFs?

The kernel's i40e driver has the following known issue with Linux bonding
with VFs (copied below https://github.com/dmarion/i40e/blob/master/README),
but that seems to have been fixed with the 2.15.9 release, and the patch
above seems to resolve the issue of setting the MAC address by the VF for
the PMD, but looks like it's insufficient for it to work when backported to
the DPDK18.11's i40e PMD.

Linux bonding fails with Virtual Functions bound to an Intel(R) Ethernet

Controller 700 series based device

------------------------------------------------------------------------

If you bind Virtual Functions (VFs) to an Intel(R) Ethernet Controller 700

series based device, the VF slaves may fail when they become the active
slave.

*If the MAC address of the VF is set by the PF (Physical Function) of the*

*device, when you add a slave, or change the active-backup slave, Linux
bonding*

*tries to sync the backup slave's MAC address to the same MAC address as
the*

*active slave. Linux bonding will fail at this point. *This issue will not
occur

if the VF's MAC address is not set by the PF.

                 reply	other threads:[~2021-05-28 10:29 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=CAMfW34QFeXFu-DDJbMjCK5zafp5uZU8svgjGPRktdn+9bEHwTw@mail.gmail.com \
    --to=yummy.fried.onions@gmail.com \
    --cc=dev@dpdk.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).