netdev - Re: Fix FRTO+NewReno problem

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

Date:	Tue, 13 May 2008 12:42:03 +0300 (EEST)
From:	"Ilpo Järvinen" <ilpo.jarvinen@...sinki.fi>
To:	David Miller <davem@...emloft.net>
cc:	damon@...tek.com, Netdev <netdev@...r.kernel.org>
Subject: Re: Fix FRTO+NewReno problem

On Mon, 12 May 2008, David Miller wrote:

> From: "Ilpo_Järvinen" <ilpo.jarvinen@...sinki.fi>
> Date: Mon, 12 May 2008 22:12:47 +0300 (EEST)
> 
> > Then one question for DaveM:
> > 
> > What I'm not fully sure of, is do we want this workaround to be a sysctl 
> > or unconditionally enabled which causes potentially up to two unnecessary 
> > retransmissions? With SACK one or both of them will get SACKed before they 
> > get retransmitted (both cases have common scenarios). (I made that 
> > workaround patch for 2.6.24.1, so YMMV if you just plainly try to apply it 
> > to net-2.6).
> 
> I think we will have to handle this behavior, unconditionally, all
> the time.  Don't even add the sysctl.

Yes, that's what I was thinking in the second option.

> If I understand correctly, it is only non-SACK newreno case that can
> get the unnecessary retransmissions, right?  If so, it's not a big
> deal at all.

No, also SACK could get at least one quite easily if TCP has a small 
window, to get two of them with SACK one needs to have a stranger case
unless I missed something. If I just remove that forward transmission LOST 
marking avoidance unconditionally, it fixes that SACKFRTO window of 
failure too which was a pending thing to fix :-).

Btw, the first two fixes (workaround was not included) also fixed bugzilla 
#10063, I just got a confirmation about that as well.

I'll send the two patches still pending separately in a minute.

-- 
 i.