netdev - Re: [PATCH net-next] sch_htb: add HTB_LEAF

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: <1386337732.30495.260.camel@edumazet-glaptop2.roam.corp.google.com>
Date:	Fri, 06 Dec 2013 05:48:52 -0800
From:	Eric Dumazet <eric.dumazet@...il.com>
To:	Yang Yingliang <yangyingliang@...wei.com>
Cc:	davem@...emloft.net, netdev@...r.kernel.org, jhs@...atatu.com
Subject: Re: [PATCH net-next] sch_htb: add HTB_LEAF_LIMIT attribute

On Fri, 2013-12-06 at 17:32 +0800, Yang Yingliang wrote:
> HTB uses an internal pfifo queue, which limit is inherited from
> device tx_queue_len. But virtual device's tx_queue_len is 0, so
> the limit will be initialized to 1.
> 
> Introduce TCA_HTB_LEAF_LIMIT attribute to allow finer control.

Most users overcome this by explicitly attach pfifo or whatever qdisc
they want on the htb class.

As a bonus, this qdisc is visible and we can get a rate estimator, and
all the stats.

And if somebody attaches a HTB qdisc to a virtual device, he also
changes txqueuelen before doing so.

Anyway, I was working on being able to attach fq instead of pifo, with
a variant of fq sharing the hash table.

I cannot really understand how people can still use pfifo on htb classes
nowadays.

--
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