netdev - Re: [PATCH net-next 3/3] tcp: invoke pkts

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <CA++eYdsdW9emVJbiWnCwK5a7STPLsun2X-2edZkpYHEt6KpA6A@mail.gmail.com>
Date:	Fri, 1 May 2015 00:20:32 +0200
From:	Kenneth Klette Jonassen <kennetkl@....uio.no>
To:	Yuchung Cheng <ycheng@...gle.com>
Cc:	netdev <netdev@...r.kernel.org>,
	Eric Dumazet <edumazet@...gle.com>,
	Neal Cardwell <ncardwell@...gle.com>
Subject: Re: [PATCH net-next 3/3] tcp: invoke pkts_acked hook on every ACK

> There might be congestion control that assumes pkts_acked > 0 or data
> is cumulatively when this is called. Did you audit that?

HTCP, Illinois, Yeah:
These modules save the pkts_acked count internally and use it to grow
snd_cwnd_cnt in cong_avoid.
This patch ensures they grow by the same count that is passed to cong_avoid().

BIC:
Works well with pkts_acked = 0.

These modules have pkts_acked functions for RTT, but are not using the
acked count:
Cubic, LP, Vegas, Veno, Westwood.

PS on BIC:
Impossibly high counts can potentially overflow ca->delayed_ack to 0
and cause division by zero, e.g. (u32) -30.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html