netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Juan Manuel Santos <godlike64@gmail.com>
To: davem@davemloft.net, yoshfuji@linux-ipv6.org, dsahern@kernel.org
Cc: netdev@vger.kernel.org
Subject: Fwd: WebRTC protocols broken when forwarding after a change to ip_dst_mtu_maybe_forward()
Date: Sun, 20 Jun 2021 19:58:36 -0300	[thread overview]
Message-ID: <CAGvuCg-FJM7LLcb5i6gfZLRWJBUyDrGyCyZ4xPVqek58-kAVbQ@mail.gmail.com> (raw)
In-Reply-To: <CAGvuCg_iLtzHr_rF0291oKAtYFCespoJ=dszFWjDftZd8EN6JA@mail.gmail.com>

Hello,

Apologies if somebody receives this email twice. My first email was
filtered by the mailing list because gmail defaults to HTML. me--

I am writing to this mailing list because I believe based on the
maintainer list[1] that this is the correct place to report issues
like these, especially when unsure. I am a Gentoo user and I already
reported this downstream[2] but in my troubleshooting I was able to
confirm that linux-stable is affected, and possibly others.

The issue happens whenever a kernel version with a certain patch is
used in a gateway device doing IP forwarding for a LAN, such as a
Linux box running iptables / firewalld+iptables / firewalld+nftables.
It does not seem to matter which of the three methods is used, all are
affected. Applications inside the LAN using WebRTC (such as Google
Meet, Discord, etc) are affected. In the case of Meet, no video of any
participant can be seen, although audio works. In the case of Discord,
neither audio nor video works. It does not matter whether the
conference is started or joined from a device within the LAN, it won't
work properly.

I was able to git-bisect this using linux-stable and found the
offending upstream commit[3]. In linux-stable this was backported
right after 5.4.72 so >=5.4.73 are affected, up to 5.4.126. I can
confirm that reverting the commit (even if it is just commenting those
4 lines that the commit adds) fixes the issue at least on 5.4.109. No
other protocol/connection type seems to be affected, and this only
seems to affect webrtc in the context of forwarding (i.e. when
started/joined from a device in the LAN, not the gateway itself).

I am unsure how to proceed, whether this requires an upstream bugzilla
to be opened (which I can gladly do) or not. I searched the archives
first but I could find no mention relating webrtc and the change to
this function (I only found the relationship when git-bisecting).

Thanks in advance.

Regards,



[1] https://www.kernel.org/doc/html/latest/process/maintainers.html#maintainers
[2] https://bugs.gentoo.org/797211
[3] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=02a1b175b0e92d9e0fa5df3957ade8d733ceb6a0

           reply	other threads:[~2021-06-20 22:58 UTC|newest]

Thread overview: expand[flat|nested]  mbox.gz  Atom feed
 [parent not found: <CAGvuCg_iLtzHr_rF0291oKAtYFCespoJ=dszFWjDftZd8EN6JA@mail.gmail.com>]

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=CAGvuCg-FJM7LLcb5i6gfZLRWJBUyDrGyCyZ4xPVqek58-kAVbQ@mail.gmail.com \
    --to=godlike64@gmail.com \
    --cc=davem@davemloft.net \
    --cc=dsahern@kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=yoshfuji@linux-ipv6.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).