netdev - Re: [PATCH 1/1] sched: add head drop fifo queue

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 for Android: free password hash cracker in your pocket

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

Date:	Mon, 18 Jan 2010 20:25:17 +0100
From:	Jarek Poplawski <jarkao2@...il.com>
To:	Hagen Paul Pfeifer <hagen@...u.net>
CC:	netdev@...r.kernel.org, kaber@...sh.net, davem@...emloft.net
Subject: Re: [PATCH 1/1] sched: add head drop fifo queue

Hagen Paul Pfeifer wrote, On 01/18/2010 05:44 PM:

> This add an additional queuing strategy, called pfifo_head_drop,

"This adds"...

> to remove the oldest skb in the case of an overflow within the queue -
> the head element - instead of the last skb (tail). To remove the oldest
> skb in a congested situations is useful for sensor network environments

"skb in congested situations"...

> where newer packets reflects the superior information.

"where newer packets reflect"...

> 
> Reviewed-by: Florian Westphal <fw@...len.de>
> Acked-by: Patrick McHardy <kaber@...sh.net>
> Signed-off-by: Hagen Paul Pfeifer <hagen@...u.net>
> ---
>  include/net/pkt_sched.h |    1 +
>  net/sched/sch_api.c     |    1 +
>  net/sched/sch_fifo.c    |   36 ++++++++++++++++++++++++++++++++++++
>  3 files changed, 38 insertions(+), 0 deletions(-)

...

> +static int pfifo_front_enqueue(struct sk_buff *skb, struct Qdisc* sch)

Why do you call it "front_enqueue" if always does "enqueue_tail"?
Btw, usually these methods take name from the qdisc.

> +{
> +	struct sk_buff *skb_head;
> +	struct fifo_sched_data *q = qdisc_priv(sch);
> +
> +	if (likely(skb_queue_len(&sch->q) < q->limit))
> +		return qdisc_enqueue_tail(skb, sch);
> +
> +	/* queue full, remove one skb to fulfill the limit */
> +	skb_head = qdisc_dequeue_head(sch);
> +	sch->bstats.bytes -= qdisc_pkt_len(skb_head);
> +	sch->bstats.packets--;
> +	sch->q.qlen--;

I don't think this is needed; qdisc_dequeue_head() updated it already.

> +	sch->qstats.drops++;
> +	kfree_skb(skb_head);
> +
> +	qdisc_enqueue_tail(skb, sch);
> +
> +	return NET_XMIT_CN;
> +}
> +
> +
>  static int fifo_init(struct Qdisc *sch, struct nlattr *opt)
>  {
>  	struct fifo_sched_data *q = qdisc_priv(sch);
> @@ -108,6 +130,20 @@ struct Qdisc_ops bfifo_qdisc_ops __read_mostly = {
>  };
>  EXPORT_SYMBOL(bfifo_qdisc_ops);
>  
> +struct Qdisc_ops pfifo_head_drop_qdisc_ops __read_mostly = {
> +	.id		=	"pfifo_head_drop",
> +	.priv_size	=	sizeof(struct fifo_sched_data),
> +	.enqueue	=	pfifo_front_enqueue,
> +	.dequeue	=	qdisc_dequeue_head,
> +	.peek		=	qdisc_peek_head,
> +	.drop		=	qdisc_queue_drop,

Probably it isn't used much, but it seems for consistency this drop
should be implemented as a head drop.

Jarek P.
--
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