On Thu, Aug 1, 2013 at 10:49 AM, Andy Zhou <az...@nicira.com> wrote:
> Pre mega flow, netlink allows the in_port key attribute
> to be missing. Missing in_port is interpreted as DP_MAX_PORTS.
>
> For backward compatibility, mega flow implementation will always allow
> the mask of in_port to be specified, as if the in_port key attribute
> is always specified.
>
> To prevent accidental match of the DP_MAX_PORTS, which value is opaque to
> the user space, we will always force the mask to be exact match,
> regardless of the value supplied by the netline message. Missing
> in_port mask continue to mean wildcarded match, same as other masks.
>
> Signed-off-by: Andy Zhou <az...@nicira.com>
> ---
> datapath/flow.c | 14 +++++++-------
> 1 file changed, 7 insertions(+), 7 deletions(-)
>
> diff --git a/datapath/flow.c b/datapath/flow.c
> index 29d3062..5aac1f9 100644
> --- a/datapath/flow.c
> +++ b/datapath/flow.c
> @@ -135,12 +135,9 @@ static bool ovs_match_validate(const struct
> sw_flow_match *match,
> | (1ULL << OVS_KEY_ATTR_ARP)
> | (1ULL << OVS_KEY_ATTR_ND));
>
> - /* Tunnel mask is always allowed. */
> - mask_allowed |= (1ULL << OVS_KEY_ATTR_TUNNEL);
> -
> - if (match->key->phy.in_port == DP_MAX_PORTS &&
> - match->mask && (match->mask->key.phy.in_port == 0xffff))
> - mask_allowed |= (1ULL << OVS_KEY_ATTR_IN_PORT);
> + /* Always allowed mask fields. */
> + mask_allowed |= ((1ULL << OVS_KEY_ATTR_TUNNEL)
> + | (1ULL << OVS_KEY_ATTR_IN_PORT));
>
> if (match->key->eth.type == htons(ETH_P_802_2) &&
> match->mask && (match->mask->key.eth.type == htons(0xffff)))
> @@ -1315,8 +1312,11 @@ static int metadata_from_nlattrs(struct sw_flow_match
> *match, u64 *attrs,
> if (*attrs & (1ULL << OVS_KEY_ATTR_IN_PORT)) {
> u32 in_port = nla_get_u32(a[OVS_KEY_ATTR_IN_PORT]);
>
> - if (!is_mask && in_port >= DP_MAX_PORTS)
> + if (is_mask)
> + in_port = 0xffff; /* Always exact match in_port.
> */
I applied this but I widened the constant to be 32 bits since that's
the nominal width of the field, even if it will end up being truncated
later.
_______________________________________________
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
