netdev - Re: CBQ broken in 2.6

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: <1264767932.4412.7.camel@mare-infinitum.sigsegv.cx>
Date:	Fri, 29 Jan 2010 12:25:31 +0000
From:	Anton Ivanov <anton.ivanov@...-begemot.co.uk>
To:	Eric Dumazet <eric.dumazet@...il.com>
Cc:	Jarek Poplawski <jarkao2@...il.com>,
	David Miller <davem@...emloft.net>, netdev@...r.kernel.org
Subject: Re: CBQ broken in 2.6

[snip]

> > Thanks for all the help. 
> 
> I am using CBQ myself with recent kernels and never found it
> 'borrowing', could you post a copy of your rules, or better, a subset of
> them desmonstrating the problem ?

Actually after going through it several times and looking at the code
Jarek pointed out it doesn't. 

Just the stats are very confusing and precision is not particularly
great.

It says "borrowed" while actually it is the child which has borrowed
from this class, not the class itself borrowing. The class has been
borrowed from, not it itself borrowed. 

As nobody else has complained about the stats they should be probably
left as is according to the "least surprise" principle.

As far as the precision becoming worse over the last 15 or so minor
revision 2.6.9 to 2.6.26 that cannot be helped. On my hardware it was so
bad that I had it confused with not working at all at some point. I
ended up sticking aggressive RED leafs on the biggest "offenders" and
this has gotten my config to a working state for now. I am not happy
with it, but it is better than no QoS at all.

> 
> 
> 
> 
-- 
   Understanding is a three-edged sword:
            your side, their side, and the truth. --Kosh Naranek

A. R. Ivanov
E-mail:  anton.ivanov@...-begemot.co.uk
WWW:     http://www.kot-begemot.co.uk/

--
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