Re: Kernel 2.4.25 + VLAN + CBQ disc broken ?

From: Patrick McHardy
Date: Tue Mar 23 2004 - 17:56:57 EST


Gianfranco Delli Carri wrote:
Greetigs,

seems that in a 2.4.25 kernel, with VLAN configured, if you try tu use CBQ disc on VLAN subIf no more traffic can be passed. (NO ARP, NO IP)

Just after: /sbin/tc qdisc add dev eth1.10 root handle 1 cbq bandwidth 100Mbit avpkt 1000 cell 8

traffic was dropped.

Seems that in the same server enviroment the CBQ disc attached to the untagged interface can work without problems.

Please try this patch.

Dave, the txqueuelen=0 fix for pfifo_fast apparently didn't went in
2.4, this is the patch from 2.6, it applies with minor offset.

Best regards
Patrick
ChangeSet 1.1561.2.5, 2004/02/18 13:18:53-08:00, kaber@xxxxxxxxx

[PKTSCHED]: Use queue limit of 1 when tx_queue_len is zero.


# This patch includes the following deltas:
# ChangeSet 1.1561.2.4 -> 1.1561.2.5
# net/sched/sch_fifo.c 1.7 -> 1.8
# net/sched/sch_gred.c 1.13 -> 1.14
#

sch_fifo.c | 6 ++++--
sch_gred.c | 2 +-
2 files changed, 5 insertions(+), 3 deletions(-)


diff -Nru a/net/sched/sch_fifo.c b/net/sched/sch_fifo.c
--- a/net/sched/sch_fifo.c Wed Feb 18 18:09:39 2004
+++ b/net/sched/sch_fifo.c Wed Feb 18 18:09:39 2004
@@ -141,10 +141,12 @@
struct fifo_sched_data *q = (void*)sch->data;

if (opt == NULL) {
+ unsigned int limit = sch->dev->tx_queue_len ? : 1;
+
if (sch->ops == &bfifo_qdisc_ops)
- q->limit = sch->dev->tx_queue_len*sch->dev->mtu;
+ q->limit = limit*sch->dev->mtu;
else
- q->limit = sch->dev->tx_queue_len;
+ q->limit = limit;
} else {
struct tc_fifo_qopt *ctl = RTA_DATA(opt);
if (opt->rta_len < RTA_LENGTH(sizeof(*ctl)))
diff -Nru a/net/sched/sch_gred.c b/net/sched/sch_gred.c
--- a/net/sched/sch_gred.c Wed Feb 18 18:09:39 2004
+++ b/net/sched/sch_gred.c Wed Feb 18 18:09:39 2004
@@ -110,7 +110,7 @@
unsigned long qave=0;
int i=0;

- if (!t->initd && skb_queue_len(&sch->q) < sch->dev->tx_queue_len) {
+ if (!t->initd && skb_queue_len(&sch->q) < (sch->dev->tx_queue_len ? : 1)) {
D2PRINTK("NO GRED Queues setup yet! Enqueued anyway\n");
goto do_enqueue;
}