1-4 look good.
Caveat to note Dave: both old TC_STAT as well as new TC_STAT2 will be
generated. Newer software can ignore TC_STAT and older s/ware can ignore
TC_STAT2.
patch 5 is a fix thats independent to 1-4 and should be applied
regardless of the status of 1-4
cheers,
jamal, semi-alive
On Wed, 2004-10-06 at 21:08, Thomas Graf wrote:
> hfsc and htb qdisc are not supposed to copy TCA_STATS
> on their own and queue length statistic is already
> updated in generic code part.
>
> Signed-off-by: Thomas Graf <tgraf@xxxxxxx>
>
> diff -Nru linux-2.6.9-rc3-bk6.orig/net/sched/sch_hfsc.c
> linux-2.6.9-rc3-bk6/net/sched/sch_hfsc.c
> --- linux-2.6.9-rc3-bk6.orig/net/sched/sch_hfsc.c 2004-10-07
> 00:32:25.000000000 +0200
> +++ linux-2.6.9-rc3-bk6/net/sched/sch_hfsc.c 2004-10-07 00:55:35.000000000
> +0200
> @@ -1653,11 +1653,6 @@
>
> qopt.defcls = q->defcls;
> RTA_PUT(skb, TCA_OPTIONS, sizeof(qopt), &qopt);
> -
> - sch->stats.qlen = sch->q.qlen;
> - if (qdisc_copy_stats(skb, &sch->stats, sch->stats_lock) < 0)
> - goto rtattr_failure;
> -
> return skb->len;
>
> rtattr_failure:
> diff -Nru linux-2.6.9-rc3-bk6.orig/net/sched/sch_htb.c
> linux-2.6.9-rc3-bk6/net/sched/sch_htb.c
> --- linux-2.6.9-rc3-bk6.orig/net/sched/sch_htb.c 2004-10-07
> 00:32:25.000000000 +0200
> +++ linux-2.6.9-rc3-bk6/net/sched/sch_htb.c 2004-10-07 00:55:22.000000000
> +0200
> @@ -1332,8 +1332,6 @@
> RTA_PUT(skb, TCA_OPTIONS, 0, NULL);
> RTA_PUT(skb, TCA_HTB_INIT, sizeof(gopt), &gopt);
> rta->rta_len = skb->tail - b;
> - sch->stats.qlen = sch->q.qlen;
> - RTA_PUT(skb, TCA_STATS, sizeof(sch->stats), &sch->stats);
> HTB_QUNLOCK(sch);
> return skb->len;
> rtattr_failure:
>
>
|