Thank Werner,
  that's helps alot :)
i managed to figure out that the following does the trick for me :

tc qdisc add dev eth1 handle 1:0 root dsmark indices 64
tc class change dev eth1 classid 1:11 dsmark mask 0x3 value 0x28
tc class change dev eth1 classid 1:21 dsmark mask 0x3 value 0x48
tc filter add dev eth1 parent 1:0 protocol all prio 1 handle 2 \
              tcindex classid 1:11
tc filter add dev eth1 parent 1:0 protocol all prio 1 handle 1 \
              tcindex classid 1:21

#tc qdisc add dev eth1 parent 1:0 handle 2:0 dsmark indices 64 set_tc_index
#tc filter add dev eth1 parent 2:0 protocol all prio 1 tcindex mask 0xfc
shift 2 pass_on
#tc filter add dev eth1 parent 2:0 prio 1 handle 0x22 tcindex classid 2:111
#tc filter add dev eth1 parent 2:0 prio 1 handle 0x12 tcindex classid 2:191

tc qdisc add dev eth1 parent 1:0 handle 2:0 cbq bandwidth \
                      10Mbit cell 8 avpkt 1000 mpu 64
tc filter add dev eth1 parent 2:0 protocol all prio 1 \
                       tcindex mask 0xf0 shift 4 pass_on
tc class add dev eth1 parent 2:0 classid 2:99 cbq bandwidth \
                      10Mbit rate 9Mbit avpkt 1000 prio 2   \
                      allot 1514 weight 9 maxburst 21 defmap 0
tc class add dev eth1 parent 2:0 classid 2:44 cbq bandwidth \
                      10Mbit rate 1Mbit avpkt 1000 prio 2   \
                      allot 1514 weight 9 maxburst 21 defmap 0

tc filter add dev eth1 parent 2:0 prio 1 handle 1 tcindex \
                       classid 2:99
tc filter add dev eth1 parent 2:0 prio 1 handle 2 tcindex \
                       classid 2:44

tc qdisc add dev eth1 parent 2:99 handle 4:0 pfifo limit 10
tc qdisc add dev eth1 parent 2:44 handle 5:0 pfifo limit 10

  note that after i delete the lines begin with notation mark "#",
all things seems to be back to normal. :)
  will someone give some comment on this ?

>From: Werner Almesberger <wa@almesberger.net>
>To: Folke Aeon <aeon_folke@hotmail.com>
>CC: Subject: Re: [LARTC] does any body ever met this strange problem?
>Date: Thu, 24 Oct 2002 03:19:51 -0300
>
>Folke Aeon wrote:
> > hi, does any body ever met this strange problem?
>
>No, I haven't, because I usually don't nest dsmark ;-)
>
>dsmark changes the DS field according to the value in skb->tc_index
>when the packet _leaves_ the qdisc, and your inner dsmark changes
>skb->tc_index.
>
>I suggest you have a look at figures 6 and 8 (the latter is on page
>6) of ftp://icaftp.epfl.ch/pub/linux/diffserv/misc/dsid-01.ps.gz
>
>- Werner
>
>--
>
>_________________________________________________________________________
>  / Werner Almesberger, Buenos Aires, Argentina         wa@almesberger.net
>/
>/_http://www.almesberger.net/____________________________________________/


_________________________________________________________________
Unlimited Internet access -- and 2 months free!  Try MSN.
http://resourcecenter.msn.com/access/plans/2monthsfree.asp