Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 18 Dec 2018 09:53:04 +0000
From:      bugzilla-noreply@freebsd.org
To:        net@FreeBSD.org
Subject:   [Bug 221919] ixl: TX queue hang when using TSO and having a high and mixed network load
Message-ID:  <bug-221919-7501-ugE9A8Kiy4@https.bugs.freebsd.org/bugzilla/>
In-Reply-To: <bug-221919-7501@https.bugs.freebsd.org/bugzilla/>
References:  <bug-221919-7501@https.bugs.freebsd.org/bugzilla/>

next in thread | previous in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D221919

--- Comment #25 from Peter Eriksson <peter.x.eriksson@liu.se> ---
> Do you have any scripts that reconfigure the interface (ifconfig ixl0 dow=
n=20
> up/mtu/tso) while it receives/transmits traffic?

Hmm... Aha! Bingo!

When testing on the production servers (that always receive SMB/NFS/SSH
traffic) I just did an "ifconfig lagg0 tso" to enable it and then started my
iperf3 testing (and some off and on to get test data).

I can now reliably reproduce this if I start an "iperf3" test session betwe=
en
two servers and while it is running disable / enable tso "on the fly".

I can now provoke the "hang" on:
FreeBSD 11.2-RELEASE-p6 with ixl driver 1.9.9-k and firmware 5.60
FreeBSD 11.2-RELEASE-p6 with ixl driver 1.10.4 and firmware 6.80

I've not (so far) been able to provoke it to occur on:
FreeBSD 12.0-RELEASE-p0 with ixl driver 2.1.0-k and firmware 6.80

Sometimes it self-heals after a while, but most often I have to do an "ifco=
nfig
lagg0 down ; ifconfig lagg0 up" to get it to recover.

--=20
You are receiving this mail because:
You are on the CC list for the bug.=



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-221919-7501-ugE9A8Kiy4>