On Tue, Nov 20, 2012 at 3:15 PM, Ben Pfaff <[email protected]> wrote:
> On Tue, Nov 20, 2012 at 03:05:24PM -0800, Ansis Atteka wrote:
>> This patch adds support for skb mark matching and set action.
>>
>> Signed-off-by: Ansis Atteka <[email protected]>
>
> I would increment FLOW_WC_SEQ to draw attention to some other places
> that might require updates.
I just incremented FLOW_WC_SEQ. It seems that nothing additional needs
to be done for skb marks.
But, what got my attention, is that match_format() function accepts
'priority' as argument. Nevertheless, it seems that two different
'priorities' are being used interchangeably as argument in this
function:
1. rule priority
2, skb/packet priority
1. example usage of rule priority (16-bits)
match_format(&fsr.match, string, OFP_DEFAULT_PRIORITY);
or
match_format(&fs->match, string, fs->priority);
2. example usage of skb priority (32-bits):
void
flow_format(struct ds *ds, const struct flow *flow)
{
struct match match;
match_wc_init(&match, flow);
match_format(&match, ds, flow->skb_priority);
}
Maybe someone can shed some light on this, if I am understanding the
code correctly?
>
> I worry that it will be very difficult to debug if set_mark() fails on
> some old kernel that doesn't have netfilter enabled. Maybe we don't
> care.
>
> Jesse certainly needs to review also.
_______________________________________________
dev mailing list
[email protected]
http://openvswitch.org/mailman/listinfo/dev
