netdev - Re: [PATCH 2/2 v7] xps: Transmit Packet Steering

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:	Mon, 22 Nov 2010 19:42:31 +0800
From:	Changli Gao <xiaosuo@...il.com>
To:	Tom Herbert <therbert@...gle.com>
Cc:	davem@...emloft.net, netdev@...r.kernel.org, eric.dumazet@...il.com
Subject: Re: [PATCH 2/2 v7] xps: Transmit Packet Steering

On Mon, Nov 22, 2010 at 7:17 AM, Tom Herbert <therbert@...gle.com> wrote:
> This patch implements transmit packet steering (XPS) for multiqueue
> devices.  XPS selects a transmit queue during packet transmission based
> on configuration.  This is done by mapping the CPU transmitting the
> packet to a queue.  This is the transmit side analogue to RPS-- where
> RPS is selecting a CPU based on receive queue, XPS selects a queue
> based on the CPU (previously there was an XPS patch from Eric
> Dumazet, but that might more appropriately be called transmit completion
> steering).
>
> Each transmit queue can be associated with a number of CPUs which will
> use the queue to send packets.  This is configured as a CPU mask on a
> per queue basis in:
>
> /sys/class/net/eth<n>/queues/tx-<n>/xps_cpus
>
> The mappings are stored per device in an inverted data structure that
> maps CPUs to queues.  In the netdevice structure this is an array of
> num_possible_cpu structures where each structure holds and array of
> queue_indexes for queues which that CPU can use.
>
> The benefits of XPS are improved locality in the per queue data
> structures.  Also, transmit completions are more likely to be done
> nearer to the sending thread, so this should promote locality back
> to the socket on free (e.g. UDP).  The benefits of XPS are dependent on
> cache hierarchy, application load, and other factors.  XPS would
> nominally be configured so that a queue would only be shared by CPUs
> which are sharing a cache, the degenerative configuration woud be that
> each CPU has it's own queue.
>
> Below are some benchmark results which show the potential benfit of
> this patch.  The netperf test has 500 instances of netperf TCP_RR test
> with 1 byte req. and resp.
>
> bnx2x on 16 core AMD
>   XPS (16 queues, 1 TX queue per CPU)  1234K at 100% CPU
>   No XPS (16 queues)                   996K at 100% CPU
>
> Signed-off-by: Tom Herbert <therbert@...gle.com>
> ---
>  include/linux/netdevice.h |   30 ++++
>  net/core/dev.c            |   53 ++++++-
>  net/core/net-sysfs.c      |  369 ++++++++++++++++++++++++++++++++++++++++++++-
>  net/core/net-sysfs.h      |    3 +
>  4 files changed, 447 insertions(+), 8 deletions(-)
>
> diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
> index b45c1b8..badf928 100644
> --- a/include/linux/netdevice.h
> +++ b/include/linux/netdevice.h
> @@ -503,6 +503,10 @@ struct netdev_queue {
>        struct Qdisc            *qdisc;
>        unsigned long           state;
>        struct Qdisc            *qdisc_sleeping;
> +#ifdef CONFIG_RPS
> +       struct kobject          kobj;
> +#endif
> +

Why do you reuse CONFIG_RPS? I think it is confusing, as the code
enclosed is for XPS not RPS.

-- 
Regards,
Changli Gao(xiaosuo@...il.com)
--
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