Re: [PATCH 6/6] [net-next]net: i40e: Enable cloud filters in i40e via tc/flower classifier

On 8/1/2017 3:56 AM, Jamal Hadi Salim wrote:

On 17-07-31 08:38 PM, Amritha Nambiar wrote:

quoted

This patch enables tc-flower based hardware offloads. tc/flower
filter provided by the kernel is configured as driver specific
cloud filter. The patch implements functions and admin queue
commands needed to support cloud filters in the driver and
adds cloud filters to configure these tc-flower filters.

The only action supported is to redirect packets to a traffic class
on the same device.

# tc qdisc add dev eth0 ingress
# ethtool -K eth0 hw-tc-offload on

# tc filter add dev eth0 protocol ip parent ffff:\
   prio 1 flower dst_mac 3c:fd:fe:a0:d6:70 skip_sw indev eth0\
   action mirred ingress redirect dev eth0 tc 0

Out of curiosity - did you need to say "indev eth0" there?

It looks like I don't need to specify "indev eth0". I will need to look
up how this part is offloaded and probably validate in the driver when
this is specified.

Also: Is it possible to add an skbmark? Example something like
these that directs two flows to the same queue but different
skb marks:

# tc filter add dev eth0 protocol ip parent ffff: \
   prio 2 flower dst_ip 192.168.3.5/32 \
   ip_proto udp dst_port 2a skip_sw \
   action skbedit mark 11 \
   action mirred ingress redirect dev eth0 tcqueue 1

# tc filter add dev eth0 protocol ip parent ffff: \
     prio 1 flower dst_mac 3c:fd:fe:a0:d6:70 skip_sw \
     action skbedit mark 12 \
     action mirred ingress redirect dev eth0 tcqueue 1

It is possible to support the skbedit mark action for the first rule
here (L3 and L4) which I can take up in a subsequent patch, but this
cannot be supported on our device for L2 based match in the second rule.

cheers,
jamal