Hi,
I am encountering what I suspect to be a bug in the kernel processing of
tc netlink messages:
I am using kernel version 2.4.9 with iproute2-current-001007
package. Everything works fine in the kernel. Now I open a netlink
socket with a user process, read the tc netlink messages and parse them.
What I see is that the kernel processes the messages (adds some default
values for example) before making it available to my process.
Problem comes when I parse the tcindex messages. Here is an excerpt
from the examples in iproute2 diffserv distribution.
$TC qdisc add $EGDEV handle 2:0 root dsmark indices 64
#
# The class mapping
#
$TC filter add $EGDEV parent 2:0 protocol ip prio 1 \
handle 1 tcindex classid 2:1
$TC filter add $EGDEV parent 2:0 protocol ip prio 1 \
handle 2 tcindex classid 2:2
$TC filter add $EGDEV parent 2:0 protocol ip prio 1 \
handle 3 tcindex classid 2:3
$TC filter add $EGDEV parent 2:0 protocol ip prio 1 \
handle 4 tcindex classid 2:4
When I parse the tcindex filter messages I dont see the handle and the
classid information. I only see the default value that are set by the
kernel for mask, shift and fall_through.
I SEE >> hash 64 mask 0xffff shift 0 fall_through
But when I do a #### tc filter ls dev eth0 parent 2:0, I see all the
handle - classid pair I have instantiated. I feel there is something
wrong in the way the tcindex filter messages are being processed by the
kernel. If indeed does anyone know how to fix this and if not then can
someone please tell me how the read the handle and classid information
for these messages.
PS - My process parses all the other tc messages correctly.
Thanks,
Mohit Gupta.
-
: send the line "unsubscribe linux-net" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
