Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 19 Aug 2020 09:49:46 +0000
From:      "Scheffenegger, Richard" <Richard.Scheffenegger@netapp.com>
To:        "rgrimes@FreeBSD.org" <rgrimes@FreeBSD.org>, Michael Tuexen <tuexen@freebsd.org>
Cc:        FreeBSD Transport <freebsd-transport@freebsd.org>
Subject:   ECN backports
Message-ID:  <SN4PR0601MB37284A9FA6E22D547A621455865D0@SN4PR0601MB3728.namprd06.prod.outlook.com>

next in thread | raw e-mail | index | archive | help
Hi guys,

I'm currently reviewing the pending backports from all the Diffs I have sup=
plied so far.

One major one is D22497, where all the ECN-related flags got moved from TF_=
 (t_flags) to TF2_ (t_flags2), and subsequently, all other ECN-related patc=
hes will have merge conflicts.

However, since there was a major interop issue around CWR flag not being se=
t according to RFC (D23364) which I backported out-of-line all the way to r=
eleng 11.4, stable/11 and stable/12 (manually adjusting the proper flag ref=
erences), I am now wondering:

Do we want to backport that (major) change into stable/12, where ECN-flags =
move between the tcpcb fields?

Or shall we retain the old TF flags for ECN in stable/12, manually patching=
 any backports from TF2 to TF, for binary compatibility reasons (although I=
 guess there aren't that many people who go debugging the TCPCB without rec=
ompiling the correctly matching tools)?

Best regards,

Richard Scheffenegger




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?SN4PR0601MB37284A9FA6E22D547A621455865D0>