Re: 2.6.11-rc5 and 2.6.12: cannot transmit anything - more info

[Tommy Christensen <tommy.christensen@tpack.net>]

Denis Vlasenko wrote:
> Hi,
>
> As reported earlier, sometimes my home box don't want
> to send anything.
>
> # ip r
> 1.1.5.5 dev tun0  proto kernel  scope link  src 1.1.5.6
> 1.1.4.0/24 dev if  proto kernel  scope link  src 1.1.4.6
> default via 1.1.5.5 dev tun0
> # ping 1.1.4.1 -i 0.01

> 2005-08-02_19:12:18.19551 kern.info: qdisc_restart: start, q->dequeue=c03e8662
> 2005-08-02_19:12:19.19536 kern.info: qdisc_restart: start, q->dequeue=c03e8662
>
> System.map:
> c03e8662 t noop_dequeue
>
> I guess this explains why I do not see calls to pfifo_fast_dequeue! :)
> But how come my interface is using noop queue, is a mystery to me.

Because link is down.  Or at least the kernel thinks so.

> # ip l
> 1: if: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 1000
>     link/ether 00:0a:e6:7c:dd:79 brd ff:ff:ff:ff:ff:ff
> 2: lo: <LOOPBACK,UP> mtu 16436 qdisc noqueue
>     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
> 17: tun0: <POINTOPOINT,MULTICAST,NOARP,UP> mtu 1464 qdisc pfifo_fast qlen 100
>     link/[65534]
>
> As you can see, ip l reports that iface 'if' uses pfifo_fast, not noop...

Yeah, a bit confusing.  pfifo_fast is the *configured* qdisc, but in this
case it is not the *active* qdisc.  The qdisc is set to noop when carrier
is lost.

> Any ideas?

Try tracking the calls to netif_carrier_on/off.


-Tommy
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/