[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Wed, 8 Aug 2007 16:14:02 +0400
From: Evgeniy Polyakov <johnpol@....mipt.ru>
To: Krishna Kumar <krkumar2@...ibm.com>
Cc: davem@...emloft.net, shemminger@...ux-foundation.org,
kaber@...sh.net, sri@...ibm.com, jagana@...ibm.com,
Robert.Olsson@...a.slu.se, peter.p.waskiewicz.jr@...el.com,
herbert@...dor.apana.org.au, gaagaan@...il.com,
kumarkr@...ux.ibm.com, rdreier@...co.com, rick.jones2@...com,
mcarlson@...adcom.com, jeff@...zik.org,
general@...ts.openfabrics.org, mchan@...adcom.com, tgraf@...g.ch,
hadi@...erus.ca, netdev@...r.kernel.org, xma@...ibm.com
Subject: Re: [PATCH 3/9 Rev3] [sched] Modify qdisc_run to support batching
On Wed, Aug 08, 2007 at 03:01:45PM +0530, Krishna Kumar (krkumar2@...ibm.com) wrote:
> +static inline int get_skb(struct net_device *dev, struct Qdisc *q,
> + struct sk_buff_head *blist, struct sk_buff **skbp)
> +{
> + if (likely(!blist || (!skb_queue_len(blist) && qdisc_qlen(q) <= 1))) {
> + return likely((*skbp = dev_dequeue_skb(dev, q)) != NULL);
> + } else {
> + int max = dev->tx_queue_len - skb_queue_len(blist);
> + struct sk_buff *skb;
> +
> + while (max > 0 && (skb = dev_dequeue_skb(dev, q)) != NULL)
> + max -= dev_add_skb_to_blist(skb, dev);
> +
> + *skbp = NULL;
> + return 1; /* we have atleast one skb in blist */
> + }
> +}
Same here - is it possible to get a list in one go instead of pulling
one-by-one, since it forces quite a few additional unneded lock
get/releases. What about dev_dequeue_number_skb(dev, q, num), which will
grab the lock and move a list of skbs from one queue to provided head.
> @@ -158,7 +198,10 @@ static inline int qdisc_restart(struct n
> /* And release queue */
> spin_unlock(&dev->queue_lock);
>
> - ret = dev_hard_start_xmit(skb, dev);
> + if (likely(skb))
> + ret = dev_hard_start_xmit(skb, dev);
> + else
> + ret = dev->hard_start_xmit_batch(dev);
Perfectionism says that having array of two functions and calling one of
them via array_func_pointer[!!skb] will be much faster. Just a though.
It is actually much faster than if/else on x86 at least.
--
Evgeniy Polyakov
-
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
Powered by blists - more mailing lists