netdev - Re: [PATCH] net: fec: select queue depending on VLAN priority

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]

Message-ID: <20170509121512.GA16665@lunn.ch>
Date:   Tue, 9 May 2017 14:15:12 +0200
From:   Andrew Lunn <andrew@...n.ch>
To:     Stefan Agner <stefan@...er.ch>
Cc:     fugang.duan@....com, festevam@...il.com, netdev@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH] net: fec: select queue depending on VLAN priority

On Mon, May 08, 2017 at 10:37:08PM -0700, Stefan Agner wrote:
> Since the addition of the multi queue code with commit 59d0f7465644
> ("net: fec: init multi queue date structure") the queue selection
> has been handelt by the default transmit queue selection
> implementation which tries to evenly distribute the traffic across
> all available queues. This selection presumes that the queues are
> using an equal priority, however, the queues 1 and 2 are actually
> of higher priority (the classification of the queues is enabled in
> fec_enet_enable_ring).
> 
> This can lead to net scheduler warnings and continuous TX ring
> dumps when exercising the system with iperf.
> 
> Use only queue 0 for all common traffic (no VLAN and P802.1p
> priority 0 and 1) and route level 2-7 through queue 1 and 2.

Hi Stefan

Did you try:

vconfig set_egress_map eth0.42 0 7
ip addr add 10.42.42.42/24 eth0.42
iperf -c 10.42.42.1

i.e. send a continuous stream on one of the higher priority queues.

>From what was said earlier in this thread, isn't queue 0 going to be
starved? As well as this patch, don't we also need some default
bandwidth allocations to the queues to ensure queue 0 does get some
bandwidth?

	Andrew