Re: [PATCH v2] net: batch skb dequeueing from softnet input_pkt_queue

["Paul E. McKenney" <paulmck@linux.vnet.ibm.com>]

On Tue, Apr 13, 2010 at 05:50:29PM +0800, Changli Gao wrote:
> On Tue, Apr 13, 2010 at 4:08 PM, Eric Dumazet <eric.dumazet@gmail.com> wrote:
> >
> >        Probably not necessary.
> >
> >> +     volatile bool           flush_processing_queue;
> >
> > Use of 'volatile' is strongly discouraged, I would say, forbidden.
> 
> volatile is used to avoid compiler optimization.

Would it be reasonable to use ACCESS_ONCE() where this variable is used?

						Thanx, Paul

> > Its usually a sign of 'I dont exactly what memory ordering I need, so I
> > throw a volatile just in case'. We live in a world full of RCU, read ,
> > write, full barriers. And these apis are well documented.
> >
> 
> There isn't memory accessing order problem.
> 
> >> @@ -2803,6 +2808,7 @@ static void flush_backlog(void *arg)
> >>                       __skb_unlink(skb, &queue->input_pkt_queue);
> >>                       kfree_skb(skb);
> >>               }
> >> +     queue->flush_processing_queue = true;
> >
> >        Probably not necessary
> >
> 
> If flush_backlog() is called when there are still packets in
> processing_queue, there maybe some packets refer to the netdev gone,
> if we remove this line.
> 
> >>       rps_unlock(queue);
> >>  }
> >>
> >> @@ -3112,14 +3118,23 @@ static int process_backlog(struct napi_struct *napi, int quota)
> >>       struct softnet_data *queue = &__get_cpu_var(softnet_data);
> >>       unsigned long start_time = jiffies;
> >>
> >> +     if (queue->flush_processing_queue) {
> >
> > Really... this is bloat IMHO
> 
> 
> Any better idea?
> 
> >
> >>
> >
> > I advise to keep it simple.
> >
> > My suggestion would be to limit this patch only to process_backlog().
> >
> > Really if you touch other areas, there is too much risk.
> >
> > Perform sort of skb_queue_splice_tail_init() into a local (stack) queue,
> > but the trick is to not touch input_pkt_queue.qlen, so that we dont slow
> > down enqueue_to_backlog().
> >
> > Process at most 'quota' skbs (or jiffies limit).
> >
> > relock queue.
> > input_pkt_queue.qlen -= number_of_handled_skbs;
> >
> 
> Oh no, in order to let latter packets in as soon as possible, we have
> to update qlen immediately.
> 
> -- 
> Regards，
> Changli Gao(xiaosuo@gmail.com)
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html