[ovs-dev,net-next,V15,3/3] 802.1AD: Flow handling, actions, vlan parsing and netlink attributes
diff mbox

Message ID 1444520433-1958-4-git-send-email-thomasfherbert@gmail.com
State Not Applicable
Headers show

Commit Message

Thomas F Herbert Oct. 10, 2015, 11:40 p.m. UTC
Add support for 802.1ad including the ability to push and pop double
tagged vlans. Add support for 802.1ad to netlink parsing and flow
conversion. Uses double nested encap attributes to represent double
tagged vlan. Inner TPID encoded along with ctci in nested attributes.

Signed-off-by: Thomas F Herbert <thomasfherbert@gmail.com>
---
 net/openvswitch/actions.c      |   6 +-
 net/openvswitch/flow.c         |  92 +++++++++++++++++++----
 net/openvswitch/flow.h         |  11 ++-
 net/openvswitch/flow_netlink.c | 166 +++++++++++++++++++++++++++++++++++++----
 net/openvswitch/vport-netdev.c |   4 +-
 5 files changed, 245 insertions(+), 34 deletions(-)

Comments

Pravin B Shelar Oct. 13, 2015, 6:47 a.m. UTC | #1
On Sat, Oct 10, 2015 at 4:40 PM, Thomas F Herbert
<thomasfherbert@gmail.com> wrote:
> Add support for 802.1ad including the ability to push and pop double
> tagged vlans. Add support for 802.1ad to netlink parsing and flow
> conversion. Uses double nested encap attributes to represent double
> tagged vlan. Inner TPID encoded along with ctci in nested attributes.
>
> Signed-off-by: Thomas F Herbert <thomasfherbert@gmail.com>
> ---
>  net/openvswitch/actions.c      |   6 +-
>  net/openvswitch/flow.c         |  92 +++++++++++++++++++----
>  net/openvswitch/flow.h         |  11 ++-
>  net/openvswitch/flow_netlink.c | 166 +++++++++++++++++++++++++++++++++++++----
>  net/openvswitch/vport-netdev.c |   4 +-
>  5 files changed, 245 insertions(+), 34 deletions(-)
>
...

> diff --git a/net/openvswitch/flow.c b/net/openvswitch/flow.c
> index c8db44a..0f9479c 100644
> --- a/net/openvswitch/flow.c
> +++ b/net/openvswitch/flow.c
> @@ -305,21 +305,83 @@ static bool icmp6hdr_ok(struct sk_buff *skb)
>  static int parse_vlan(struct sk_buff *skb, struct sw_flow_key *key)
>  {
>         struct qtag_prefix {
> -               __be16 eth_type; /* ETH_P_8021Q */
> +               __be16 eth_type; /* ETH_P_8021Q  or ETH_P_8021AD */
>                 __be16 tci;
>         };
> -       struct qtag_prefix *qp;
> +       struct qtag_prefix *qp = (struct qtag_prefix *)skb->data;
>
> -       if (unlikely(skb->len < sizeof(struct qtag_prefix) + sizeof(__be16)))
> +       if (likely(skb_vlan_tag_present(skb))) {
> +               key->eth.vlan.tci = htons(skb->vlan_tci);
> +               key->eth.vlan.tpid = skb->vlan_proto;
> +
> +               /* Case where upstream
> +                * processing has already stripped the outer vlan tag.
> +                */
> +               if (unlikely(skb->vlan_proto == htons(ETH_P_8021AD))) {
> +                       if (unlikely(skb->len < sizeof(struct qtag_prefix) +
> +                                       sizeof(__be16))) {
> +                               key->eth.vlan.tci = 0;
> +                               return 0;
> +                       }
> +
> +                       if (unlikely(!pskb_may_pull(skb,
> +                                                   sizeof(struct qtag_prefix) +
> +                                                   sizeof(__be16))))
> +                               return -ENOMEM;
> +
> +                       qp = (struct qtag_prefix *)skb->data;
> +                       key->eth.cvlan.tci =
> +                               qp->tci | htons(VLAN_TAG_PRESENT);
> +                       key->eth.cvlan.tpid = qp->eth_type;
> +
> +                       __skb_pull(skb, sizeof(struct qtag_prefix));
> +               }
>                 return 0;
>
> -       if (unlikely(!pskb_may_pull(skb, sizeof(struct qtag_prefix) +
> -                                        sizeof(__be16))))
> -               return -ENOMEM;
> +       } else if (qp->eth_type == htons(ETH_P_8021AD)) {
> +
> +               if (unlikely(skb->len < sizeof(struct qtag_prefix) +
> +                                       sizeof(__be16)))
> +                       return 0;
> +
> +               if (unlikely(!pskb_may_pull(skb, sizeof(struct qtag_prefix) +
> +                               sizeof(__be16))))
> +                       return -ENOMEM;
> +
> +               qp = (struct qtag_prefix *)skb->data;
> +               key->eth.vlan.tci = qp->tci | htons(VLAN_TAG_PRESENT);
> +               key->eth.vlan.tpid = qp->eth_type;
> +
> +               __skb_pull(skb, sizeof(struct qtag_prefix));
>
> -       qp = (struct qtag_prefix *) skb->data;
> -       key->eth.tci = qp->tci | htons(VLAN_TAG_PRESENT);
> -       __skb_pull(skb, sizeof(struct qtag_prefix));
> +               if (unlikely(skb->len < sizeof(struct qtag_prefix) +
> +                                       sizeof(__be16)))
> +                       return 0;
> +
> +               if (unlikely(!pskb_may_pull(skb, sizeof(struct qtag_prefix) +
> +                               sizeof(__be16))))
> +                       return -ENOMEM;
> +
There is no check for inner protocol in the packet.

> +               key->eth.cvlan.tci = qp->tci | htons(VLAN_TAG_PRESENT);
> +               key->eth.cvlan.tpid = qp->eth_type;
> +
> +               __skb_pull(skb, sizeof(struct qtag_prefix));
> +
> +               return 0;
> +
> +       } else if (qp->eth_type == htons(ETH_P_8021Q)) {
> +               if (unlikely(skb->len < sizeof(struct qtag_prefix) +
> +                                       sizeof(__be16)))
> +                       return 0;
> +
> +               if (unlikely(!pskb_may_pull(skb, sizeof(struct qtag_prefix) +
> +                               sizeof(__be16))))
> +                       return -ENOMEM;
> +               key->eth.vlan.tci = qp->tci | htons(VLAN_TAG_PRESENT);
> +               key->eth.vlan.tpid = qp->eth_type;
> +
> +               __skb_pull(skb, sizeof(struct qtag_prefix));
> +       }
>
>         return 0;
>  }
I see lot of duplicate code here. How about code below:

struct qtag_prefix {
        __be16 eth_type; /* ETH_P_8021Q  or ETH_P_8021AD */
        __be16 tci;
};

/* Return  < 0 on memory error
 * Return   == 0 on non vlan or incomplete packet packet
 * Return > 0 on successfully parsing vlan tag.
 */
static int parse_vlan_tag(__be16 vlan_proto, struct sk_buff *skb,
                          struct vlan_tag *cvlan)
{
        if (likely(!eth_type_vlan(skb->vlan_proto)))
                return 0;

        if (unlikely(skb->len < sizeof(struct qtag_prefix) + sizeof(__be16))) {
                vlan->tci = 0;
                return 0;
        }

        if (unlikely(!pskb_may_pull(skb, sizeof(struct qtag_prefix) +
sizeof(__be16))))
                        return -ENOMEM;

        qp = (struct qtag_prefix *)skb->data;
        key->eth.cvlan.tci = qp->tci | htons(VLAN_TAG_PRESENT);
        key->eth.cvlan.tpid = qp->eth_type;

        __skb_pull(skb, sizeof(struct qtag_prefix));
        return 1;
}

static int parse_vlan(struct sk_buff *skb, struct sw_flow_key *key)
{
        struct qtag_prefix *qp = (struct qtag_prefix *)skb->data;
        int res;

        if (likely(skb_vlan_tag_present(skb))) {
                key->eth.vlan.tci = htons(skb->vlan_tci);
                key->eth.vlan.tpid = skb->vlan_proto;

                /* Case where upstream
                 * processing has already stripped the outer vlan tag.
                 */
                res = parse_vlan_tag(skb->vlan_proto, skb, &key->eth.cvlan);
                if (res < 0)
                        return res;
                /* Since this was inner tag, return zero in either
success or failure
                 * in parsing inner tag.
                 */
                return 0;
        }
        res = parse_vlan_tag(qp->eth_type, skb, &key->eth.vlan);
        if (res <= 0)
                return res;

        res = parse_vlan_tag(key->eth.vlan.tpid, skb, &key->eth.vlan);
        if (res <= 0)
                return res;

        return 0;
}

> diff --git a/net/openvswitch/flow.h b/net/openvswitch/flow.h
> index fe527d2..539494e 100644
> --- a/net/openvswitch/flow.h
> +++ b/net/openvswitch/flow.h
> @@ -68,7 +68,16 @@ struct sw_flow_key {
>         struct {
>                 u8     src[ETH_ALEN];   /* Ethernet source address. */
>                 u8     dst[ETH_ALEN];   /* Ethernet destination address. */
> -               __be16 tci;             /* 0 if no VLAN, VLAN_TAG_PRESENT set otherwise. */
> +               struct {
> +                       __be16 tpid;    /* Outer Vlan type 802.1q or 802.1ad.*/
> +                       __be16 tci;     /* 0 if no VLAN, VLAN_TAG_PRESENT */
> +                                       /* set otherwise. */
> +               } vlan;
> +               struct {
> +                       __be16 tpid;    /* Inner Vlan DL_type 802.1q.*/
> +                       __be16 tci;     /* 0 if no CVLAN, VLAN_TAG_PRESENT */
> +                                       /* set otherwise. */
> +               } cvlan;

We need to define structure for vlan tag key here for the pseudo code to work.

>                 __be16 type;            /* Ethernet frame type. */
>         } eth;
>         union {
> diff --git a/net/openvswitch/flow_netlink.c b/net/openvswitch/flow_netlink.c
> index c92d6a2..af06683 100644
> --- a/net/openvswitch/flow_netlink.c
> +++ b/net/openvswitch/flow_netlink.c
> @@ -811,6 +811,33 @@ static int metadata_from_nlattrs(struct net *net, struct sw_flow_match *match,
>         return 0;
>  }
>
>  static int ovs_key_from_nlattrs(struct net *net, struct sw_flow_match *match,
>                                 u64 attrs, const struct nlattr **a,
>                                 bool is_mask, bool log)
> @@ -845,7 +872,7 @@ static int ovs_key_from_nlattrs(struct net *net, struct sw_flow_match *match,
>                         return -EINVAL;
>                 }
>
> -               SW_FLOW_KEY_PUT(match, eth.tci, tci, is_mask);
> +               SW_FLOW_KEY_PUT(match, eth.vlan.tci, tci, is_mask);
>                 attrs &= ~(1 << OVS_KEY_ATTR_VLAN);
>         }
>
> @@ -1064,6 +1091,86 @@ static void mask_set_nlattr(struct nlattr *attr, u8 val)
>         nlattr_set(attr, val, ovs_key_lens);
>  }
>
> +static int parse_vlan_from_nlattrs(const struct nlattr **nla,
> +                                  struct sw_flow_match *match,
> +                                  u64 *key_attrs, bool *ie_valid,
> +                                  const struct nlattr **a, bool is_mask,
> +                                  bool log)
> +{
> +       int err;
> +       const struct nlattr *encap;
> +
...

> +               u64 mask_v_attrs = 0;
> +
> +               err = parse_flow_mask_nlattrs(*nla, a, &mask_v_attrs, log);
> +               if (err)
> +                       return err;
> +
> +               if (mask_v_attrs & 1 << OVS_KEY_ATTR_ENCAP) {
> +                       if (!*ie_valid) {
> +                               OVS_NLERR(log, "Encap mask attribute is set for non-CVLAN frame.");
> +                               err = -EINVAL;
> +                               return err;
> +                       }
> +                       encap = a[OVS_KEY_ATTR_ENCAP];
> +
> +                       err = cust_vlan_from_nlattrs(match, a, is_mask, log);
> +                       if (err)
> +                               return err;
> +                       *nla = encap;
> +
There is no checking for ATTR_VLAN or ATTR_ETHERTYPE. This can result
in null pointer deference in cust_vlan_from_nlattrs().

> +                       mask_v_attrs &= ~(1 << OVS_KEY_ATTR_ENCAP);
> +                       mask_v_attrs &= ~(1ULL << OVS_KEY_ATTR_VLAN);
> +                       mask_v_attrs &= ~(1ULL << OVS_KEY_ATTR_ETHERTYPE);
> +
> +                       *key_attrs |= mask_v_attrs;
> +                       err = parse_flow_mask_nlattrs(*nla, a, key_attrs, log);
> +                       if (err)
> +                               return err;
> +               }
> +       }
> +       return 0;
> +}
> +
...

>
>         if (swkey->eth.type == htons(ETH_P_802_2)) {
>                 /*
> @@ -1525,6 +1659,8 @@ static int __ovs_nla_put_key(const struct sw_flow_key *swkey,
>  unencap:
>         if (encap)
>                 nla_nest_end(skb, encap);
> +       if (in_encap)
> +               nla_nest_end(skb, in_encap);
>

As pointed in last review, inner encap attribute should be terminated first.
Thomas F Herbert Oct. 13, 2015, 5:39 p.m. UTC | #2
Pravin,

Thanks for the review.

On 10/13/15 7:47 AM, Pravin Shelar wrote:
> On Sat, Oct 10, 2015 at 4:40 PM, Thomas F Herbert
> <thomasfherbert@gmail.com> wrote:
>> Add support for 802.1ad including the ability to push and pop double
>> tagged vlans. Add support for 802.1ad to netlink parsing and flow
>> conversion. Uses double nested encap attributes to represent double
>> tagged vlan. Inner TPID encoded along with ctci in nested attributes.
>>
>> Signed-off-by: Thomas F Herbert <thomasfherbert@gmail.com>
>> ---
>>   net/openvswitch/actions.c      |   6 +-
>>   net/openvswitch/flow.c         |  92 +++++++++++++++++++----
>>   net/openvswitch/flow.h         |  11 ++-
>>   net/openvswitch/flow_netlink.c | 166 +++++++++++++++++++++++++++++++++++++----
>>   net/openvswitch/vport-netdev.c |   4 +-
>>   5 files changed, 245 insertions(+), 34 deletions(-)
>>
> ...
>
>> diff --git a/net/openvswitch/flow.c b/net/openvswitch/flow.c
>> index c8db44a..0f9479c 100644
>> --- a/net/openvswitch/flow.c
>> +++ b/net/openvswitch/flow.c
>> @@ -305,21 +305,83 @@ static bool icmp6hdr_ok(struct sk_buff *skb)
>>   static int parse_vlan(struct sk_buff *skb, struct sw_flow_key *key)
>>   {
>>          struct qtag_prefix {
>> -               __be16 eth_type; /* ETH_P_8021Q */
>> +               __be16 eth_type; /* ETH_P_8021Q  or ETH_P_8021AD */
>>                  __be16 tci;
>>          };
>> -       struct qtag_prefix *qp;
>> +       struct qtag_prefix *qp = (struct qtag_prefix *)skb->data;
>>
>> -       if (unlikely(skb->len < sizeof(struct qtag_prefix) + sizeof(__be16)))
>> +       if (likely(skb_vlan_tag_present(skb))) {
>> +               key->eth.vlan.tci = htons(skb->vlan_tci);
>> +               key->eth.vlan.tpid = skb->vlan_proto;
>> +
>> +               /* Case where upstream
>> +                * processing has already stripped the outer vlan tag.
>> +                */
>> +               if (unlikely(skb->vlan_proto == htons(ETH_P_8021AD))) {
>> +                       if (unlikely(skb->len < sizeof(struct qtag_prefix) +
>> +                                       sizeof(__be16))) {
>> +                               key->eth.vlan.tci = 0;
>> +                               return 0;
>> +                       }
>> +
>> +                       if (unlikely(!pskb_may_pull(skb,
>> +                                                   sizeof(struct qtag_prefix) +
>> +                                                   sizeof(__be16))))
>> +                               return -ENOMEM;
>> +
>> +                       qp = (struct qtag_prefix *)skb->data;
>> +                       key->eth.cvlan.tci =
>> +                               qp->tci | htons(VLAN_TAG_PRESENT);
>> +                       key->eth.cvlan.tpid = qp->eth_type;
>> +
>> +                       __skb_pull(skb, sizeof(struct qtag_prefix));
>> +               }
>>                  return 0;
>>
>> -       if (unlikely(!pskb_may_pull(skb, sizeof(struct qtag_prefix) +
>> -                                        sizeof(__be16))))
>> -               return -ENOMEM;
>> +       } else if (qp->eth_type == htons(ETH_P_8021AD)) {
>> +
>> +               if (unlikely(skb->len < sizeof(struct qtag_prefix) +
>> +                                       sizeof(__be16)))
>> +                       return 0;
>> +
>> +               if (unlikely(!pskb_may_pull(skb, sizeof(struct qtag_prefix) +
>> +                               sizeof(__be16))))
>> +                       return -ENOMEM;
>> +
>> +               qp = (struct qtag_prefix *)skb->data;
>> +               key->eth.vlan.tci = qp->tci | htons(VLAN_TAG_PRESENT);
>> +               key->eth.vlan.tpid = qp->eth_type;
>> +
>> +               __skb_pull(skb, sizeof(struct qtag_prefix));
>>
>> -       qp = (struct qtag_prefix *) skb->data;
>> -       key->eth.tci = qp->tci | htons(VLAN_TAG_PRESENT);
>> -       __skb_pull(skb, sizeof(struct qtag_prefix));
>> +               if (unlikely(skb->len < sizeof(struct qtag_prefix) +
>> +                                       sizeof(__be16)))
>> +                       return 0;
>> +
>> +               if (unlikely(!pskb_may_pull(skb, sizeof(struct qtag_prefix) +
>> +                               sizeof(__be16))))
>> +                       return -ENOMEM;
>> +
> There is no check for inner protocol in the packet.
Yes, and the code should be generic to use eth_type_vlan() for both 
inner and outer TPIDs.
I think your pseudo code below fixes that.
>
>> +               key->eth.cvlan.tci = qp->tci | htons(VLAN_TAG_PRESENT);
>> +               key->eth.cvlan.tpid = qp->eth_type;
>> +
>> +               __skb_pull(skb, sizeof(struct qtag_prefix));
>> +
>> +               return 0;
>> +
>> +       } else if (qp->eth_type == htons(ETH_P_8021Q)) {
>> +               if (unlikely(skb->len < sizeof(struct qtag_prefix) +
>> +                                       sizeof(__be16)))
>> +                       return 0;
>> +
>> +               if (unlikely(!pskb_may_pull(skb, sizeof(struct qtag_prefix) +
>> +                               sizeof(__be16))))
>> +                       return -ENOMEM;
>> +               key->eth.vlan.tci = qp->tci | htons(VLAN_TAG_PRESENT);
>> +               key->eth.vlan.tpid = qp->eth_type;
>> +
>> +               __skb_pull(skb, sizeof(struct qtag_prefix));
>> +       }
>>
>>          return 0;
>>   }
> I see lot of duplicate code here. How about code below:
>
> struct qtag_prefix {
>          __be16 eth_type; /* ETH_P_8021Q  or ETH_P_8021AD */
>          __be16 tci;
> };
>
> /* Return  < 0 on memory error
>   * Return   == 0 on non vlan or incomplete packet packet
>   * Return > 0 on successfully parsing vlan tag.
>   */
> static int parse_vlan_tag(__be16 vlan_proto, struct sk_buff *skb,
>                            struct vlan_tag *cvlan)
> {
>          if (likely(!eth_type_vlan(skb->vlan_proto)))
>                  return 0;
>
>          if (unlikely(skb->len < sizeof(struct qtag_prefix) + sizeof(__be16))) {
>                  vlan->tci = 0;
>                  return 0;
>          }
>
>          if (unlikely(!pskb_may_pull(skb, sizeof(struct qtag_prefix) +
> sizeof(__be16))))
>                          return -ENOMEM;
>
>          qp = (struct qtag_prefix *)skb->data;
>          key->eth.cvlan.tci = qp->tci | htons(VLAN_TAG_PRESENT);
>          key->eth.cvlan.tpid = qp->eth_type;
>
>          __skb_pull(skb, sizeof(struct qtag_prefix));
>          return 1;
> }
This makes for cleaner code and certainly better for maintainability so 
I have just implemented it for this next revision. However, note that 
with this change, we incur the overhead of an additional function call 
for single tagged vlan packets.
>
> static int parse_vlan(struct sk_buff *skb, struct sw_flow_key *key)
> {
>          struct qtag_prefix *qp = (struct qtag_prefix *)skb->data;
>          int res;
>
>          if (likely(skb_vlan_tag_present(skb))) {
>                  key->eth.vlan.tci = htons(skb->vlan_tci);
>                  key->eth.vlan.tpid = skb->vlan_proto;
>
>                  /* Case where upstream
>                   * processing has already stripped the outer vlan tag.
>                   */
>                  res = parse_vlan_tag(skb->vlan_proto, skb, &key->eth.cvlan);
>                  if (res < 0)
>                          return res;
>                  /* Since this was inner tag, return zero in either
> success or failure
>                   * in parsing inner tag.
>                   */
>                  return 0;
>          }
>          res = parse_vlan_tag(qp->eth_type, skb, &key->eth.vlan);
>          if (res <= 0)
>                  return res;
>
>          res = parse_vlan_tag(key->eth.vlan.tpid, skb, &key->eth.vlan);
>          if (res <= 0)
>                  return res;
>
>          return 0;
> }
>
>> diff --git a/net/openvswitch/flow.h b/net/openvswitch/flow.h
>> index fe527d2..539494e 100644
>> --- a/net/openvswitch/flow.h
>> +++ b/net/openvswitch/flow.h
>> @@ -68,7 +68,16 @@ struct sw_flow_key {
>>          struct {
>>                  u8     src[ETH_ALEN];   /* Ethernet source address. */
>>                  u8     dst[ETH_ALEN];   /* Ethernet destination address. */
>> -               __be16 tci;             /* 0 if no VLAN, VLAN_TAG_PRESENT set otherwise. */
>> +               struct {
>> +                       __be16 tpid;    /* Outer Vlan type 802.1q or 802.1ad.*/
>> +                       __be16 tci;     /* 0 if no VLAN, VLAN_TAG_PRESENT */
>> +                                       /* set otherwise. */
>> +               } vlan;
>> +               struct {
>> +                       __be16 tpid;    /* Inner Vlan DL_type 802.1q.*/
>> +                       __be16 tci;     /* 0 if no CVLAN, VLAN_TAG_PRESENT */
>> +                                       /* set otherwise. */
>> +               } cvlan;
> We need to define structure for vlan tag key here for the pseudo code to work.
>
>>                  __be16 type;            /* Ethernet frame type. */
>>          } eth;
>>          union {
>> diff --git a/net/openvswitch/flow_netlink.c b/net/openvswitch/flow_netlink.c
>> index c92d6a2..af06683 100644
>> --- a/net/openvswitch/flow_netlink.c
>> +++ b/net/openvswitch/flow_netlink.c
>> @@ -811,6 +811,33 @@ static int metadata_from_nlattrs(struct net *net, struct sw_flow_match *match,
>>          return 0;
>>   }
>>
>>   static int ovs_key_from_nlattrs(struct net *net, struct sw_flow_match *match,
>>                                  u64 attrs, const struct nlattr **a,
>>                                  bool is_mask, bool log)
>> @@ -845,7 +872,7 @@ static int ovs_key_from_nlattrs(struct net *net, struct sw_flow_match *match,
>>                          return -EINVAL;
>>                  }
>>
>> -               SW_FLOW_KEY_PUT(match, eth.tci, tci, is_mask);
>> +               SW_FLOW_KEY_PUT(match, eth.vlan.tci, tci, is_mask);
>>                  attrs &= ~(1 << OVS_KEY_ATTR_VLAN);
>>          }
>>
>> @@ -1064,6 +1091,86 @@ static void mask_set_nlattr(struct nlattr *attr, u8 val)
>>          nlattr_set(attr, val, ovs_key_lens);
>>   }
>>
>> +static int parse_vlan_from_nlattrs(const struct nlattr **nla,
>> +                                  struct sw_flow_match *match,
>> +                                  u64 *key_attrs, bool *ie_valid,
>> +                                  const struct nlattr **a, bool is_mask,
>> +                                  bool log)
>> +{
>> +       int err;
>> +       const struct nlattr *encap;
>> +
> ...
>
>> +               u64 mask_v_attrs = 0;
>> +
>> +               err = parse_flow_mask_nlattrs(*nla, a, &mask_v_attrs, log);
>> +               if (err)
>> +                       return err;
>> +
>> +               if (mask_v_attrs & 1 << OVS_KEY_ATTR_ENCAP) {
>> +                       if (!*ie_valid) {
>> +                               OVS_NLERR(log, "Encap mask attribute is set for non-CVLAN frame.");
>> +                               err = -EINVAL;
>> +                               return err;
>> +                       }
>> +                       encap = a[OVS_KEY_ATTR_ENCAP];
>> +
>> +                       err = cust_vlan_from_nlattrs(match, a, is_mask, log);
>> +                       if (err)
>> +                               return err;
>> +                       *nla = encap;
>> +
> There is no checking for ATTR_VLAN or ATTR_ETHERTYPE. This can result
> in null pointer deference in cust_vlan_from_nlattrs().
The original vlan code does not check for these attribs in the masked 
case. It does check for them in the non-masked case and then sets a 
boolean and checks it in the masked case. I do the same thing for the 
inner vlan. I check for the attributes in the non-masked case and set a 
boolean and check the boolean in the masked case. Why is this not 
sufficient?
>
>> +                       mask_v_attrs &= ~(1 << OVS_KEY_ATTR_ENCAP);
>> +                       mask_v_attrs &= ~(1ULL << OVS_KEY_ATTR_VLAN);
>> +                       mask_v_attrs &= ~(1ULL << OVS_KEY_ATTR_ETHERTYPE);
>> +
>> +                       *key_attrs |= mask_v_attrs;
>> +                       err = parse_flow_mask_nlattrs(*nla, a, key_attrs, log);
>> +                       if (err)
>> +                               return err;
>> +               }
>> +       }
>> +       return 0;
>> +}
>> +
> ...
>
>>          if (swkey->eth.type == htons(ETH_P_802_2)) {
>>                  /*
>> @@ -1525,6 +1659,8 @@ static int __ovs_nla_put_key(const struct sw_flow_key *swkey,
>>   unencap:
>>          if (encap)
>>                  nla_nest_end(skb, encap);
>> +       if (in_encap)
>> +               nla_nest_end(skb, in_encap);
>>
> As pointed in last review, inner encap attribute should be terminated first.
Yes, I agree that inner should be first. Sorry I forgot to change that 
after the last review.
Pravin B Shelar Oct. 13, 2015, 6:14 p.m. UTC | #3
On Tue, Oct 13, 2015 at 10:39 AM, Thomas F Herbert
<thomasfherbert@gmail.com> wrote:
> Pravin,
>
> Thanks for the review.
>
>
> On 10/13/15 7:47 AM, Pravin Shelar wrote:
>>
>> On Sat, Oct 10, 2015 at 4:40 PM, Thomas F Herbert
>> <thomasfherbert@gmail.com> wrote:
>>>
>>> Add support for 802.1ad including the ability to push and pop double
>>> tagged vlans. Add support for 802.1ad to netlink parsing and flow
>>> conversion. Uses double nested encap attributes to represent double
>>> tagged vlan. Inner TPID encoded along with ctci in nested attributes.
>>>
>>> Signed-off-by: Thomas F Herbert <thomasfherbert@gmail.com>
>>> ---
>>>   net/openvswitch/actions.c      |   6 +-
>>>   net/openvswitch/flow.c         |  92 +++++++++++++++++++----
>>>   net/openvswitch/flow.h         |  11 ++-
>>>   net/openvswitch/flow_netlink.c | 166
>>> +++++++++++++++++++++++++++++++++++++----
>>>   net/openvswitch/vport-netdev.c |   4 +-
>>>   5 files changed, 245 insertions(+), 34 deletions(-)
>>>
>> ...
>>
...

>>
>> I see lot of duplicate code here. How about code below:
>>
>> struct qtag_prefix {
>>          __be16 eth_type; /* ETH_P_8021Q  or ETH_P_8021AD */
>>          __be16 tci;
>> };
>>
>> /* Return  < 0 on memory error
>>   * Return   == 0 on non vlan or incomplete packet packet
>>   * Return > 0 on successfully parsing vlan tag.
>>   */
>> static int parse_vlan_tag(__be16 vlan_proto, struct sk_buff *skb,
>>                            struct vlan_tag *cvlan)
>> {
>>          if (likely(!eth_type_vlan(skb->vlan_proto)))
>>                  return 0;
>>
>>          if (unlikely(skb->len < sizeof(struct qtag_prefix) +
>> sizeof(__be16))) {
>>                  vlan->tci = 0;
>>                  return 0;
>>          }
>>
>>          if (unlikely(!pskb_may_pull(skb, sizeof(struct qtag_prefix) +
>> sizeof(__be16))))
>>                          return -ENOMEM;
>>
>>          qp = (struct qtag_prefix *)skb->data;
>>          key->eth.cvlan.tci = qp->tci | htons(VLAN_TAG_PRESENT);
>>          key->eth.cvlan.tpid = qp->eth_type;
>>
>>          __skb_pull(skb, sizeof(struct qtag_prefix));
>>          return 1;
>> }
>
> This makes for cleaner code and certainly better for maintainability so I
> have just implemented it for this next revision. However, note that with
> this change, we incur the overhead of an additional function call for single
> tagged vlan packets.
>
If there is any performance issue we can fix the code later.

>>
>> static int parse_vlan(struct sk_buff *skb, struct sw_flow_key *key)
>> {
>>          struct qtag_prefix *qp = (struct qtag_prefix *)skb->data;
...

>>> +               u64 mask_v_attrs = 0;
>>> +
>>> +               err = parse_flow_mask_nlattrs(*nla, a, &mask_v_attrs,
>>> log);
>>> +               if (err)
>>> +                       return err;
>>> +
>>> +               if (mask_v_attrs & 1 << OVS_KEY_ATTR_ENCAP) {
>>> +                       if (!*ie_valid) {
>>> +                               OVS_NLERR(log, "Encap mask attribute is
>>> set for non-CVLAN frame.");
>>> +                               err = -EINVAL;
>>> +                               return err;
>>> +                       }
>>> +                       encap = a[OVS_KEY_ATTR_ENCAP];
>>> +
>>> +                       err = cust_vlan_from_nlattrs(match, a, is_mask,
>>> log);
>>> +                       if (err)
>>> +                               return err;
>>> +                       *nla = encap;
>>> +
>>
>> There is no checking for ATTR_VLAN or ATTR_ETHERTYPE. This can result
>> in null pointer deference in cust_vlan_from_nlattrs().
>
> The original vlan code does not check for these attribs in the masked case.
> It does check for them in the non-masked case and then sets a boolean and
> checks it in the masked case. I do the same thing for the inner vlan. I
> check for the attributes in the non-masked case and set a boolean and check
> the boolean in the masked case. Why is this not sufficient?

Original code is checking for attributes before referencing them. For
example  in function ovs_nla_get_match() before extracting eth_type,
it does check a[OVS_KEY_ATTR_ETHERTYPE]. But If you spot bug in
current code please send fix for net tree.
Regarding the Boolean, it is for presence of inner vlan for key
attribute, mask attribute still could be missing vlan attribute.
For vlan mask, we can keep check sanity check as outer vlan. It means
eth_type must be specified and should be 0xffff, and tci mask is
optional and by default initialized to 0xffff.
Thomas F Herbert Oct. 15, 2015, 12:50 p.m. UTC | #4
On 10/13/15 2:14 PM, Pravin Shelar wrote:
> On Tue, Oct 13, 2015 at 10:39 AM, Thomas F Herbert
> <thomasfherbert@gmail.com>  wrote:
>> Pravin,
>>
>> Thanks for the review.
>>
>>
>> On 10/13/15 7:47 AM, Pravin Shelar wrote:
>>> On Sat, Oct 10, 2015 at 4:40 PM, Thomas F Herbert
>>> <thomasfherbert@gmail.com>  wrote:
>>>> Add support for 802.1ad including the ability to push and pop double
>>>> tagged vlans. Add support for 802.1ad to netlink parsing and flow
>>>> conversion. Uses double nested encap attributes to represent double
>>>> tagged vlan. Inner TPID encoded along with ctci in nested attributes.
>>>>
>>>> Signed-off-by: Thomas F Herbert<thomasfherbert@gmail.com>
>>>> ---
>>>>    net/openvswitch/actions.c      |   6 +-
>>>>    net/openvswitch/flow.c         |  92 +++++++++++++++++++----
>>>>    net/openvswitch/flow.h         |  11 ++-
>>>>    net/openvswitch/flow_netlink.c | 166
>>>> +++++++++++++++++++++++++++++++++++++----
>>>>    net/openvswitch/vport-netdev.c |   4 +-
>>>>    5 files changed, 245 insertions(+), 34 deletions(-)
>>>>
>>> ...
>>>
> ...
>
>>> I see lot of duplicate code here. How about code below:
>>>
>>> struct qtag_prefix {
>>>           __be16 eth_type; /* ETH_P_8021Q  or ETH_P_8021AD */
>>>           __be16 tci;
>>> };
>>>
>>> /* Return  < 0 on memory error
>>>    * Return   == 0 on non vlan or incomplete packet packet
>>>    * Return > 0 on successfully parsing vlan tag.
>>>    */
>>> static int parse_vlan_tag(__be16 vlan_proto, struct sk_buff *skb,
>>>                             struct vlan_tag *cvlan)
>>> {
>>>           if (likely(!eth_type_vlan(skb->vlan_proto)))
>>>                   return 0;
>>>
>>>           if (unlikely(skb->len < sizeof(struct qtag_prefix) +
>>> sizeof(__be16))) {
>>>                   vlan->tci = 0;
>>>                   return 0;
>>>           }
>>>
>>>           if (unlikely(!pskb_may_pull(skb, sizeof(struct qtag_prefix) +
>>> sizeof(__be16))))
>>>                           return -ENOMEM;
>>>
>>>           qp = (struct qtag_prefix *)skb->data;
>>>           key->eth.cvlan.tci = qp->tci | htons(VLAN_TAG_PRESENT);
>>>           key->eth.cvlan.tpid = qp->eth_type;
>>>
>>>           __skb_pull(skb, sizeof(struct qtag_prefix));
>>>           return 1;
>>> }
>> This makes for cleaner code and certainly better for maintainability so I
>> have just implemented it for this next revision. However, note that with
>> this change, we incur the overhead of an additional function call for single
>> tagged vlan packets.
>>
> If there is any performance issue we can fix the code later.
>
>>> static int parse_vlan(struct sk_buff *skb, struct sw_flow_key *key)
>>> {
>>>           struct qtag_prefix *qp = (struct qtag_prefix *)skb->data;
> ...
>
>>>> +               u64 mask_v_attrs = 0;
>>>> +
>>>> +               err = parse_flow_mask_nlattrs(*nla, a, &mask_v_attrs,
>>>> log);
>>>> +               if (err)
>>>> +                       return err;
>>>> +
>>>> +               if (mask_v_attrs & 1 << OVS_KEY_ATTR_ENCAP) {
>>>> +                       if (!*ie_valid) {
>>>> +                               OVS_NLERR(log, "Encap mask attribute is
>>>> set for non-CVLAN frame.");
>>>> +                               err = -EINVAL;
>>>> +                               return err;
>>>> +                       }
>>>> +                       encap = a[OVS_KEY_ATTR_ENCAP];
>>>> +
>>>> +                       err = cust_vlan_from_nlattrs(match, a, is_mask,
>>>> log);
>>>> +                       if (err)
>>>> +                               return err;
>>>> +                       *nla = encap;
>>>> +
>>> There is no checking for ATTR_VLAN or ATTR_ETHERTYPE. This can result
>>> in null pointer deference in cust_vlan_from_nlattrs().
>> The original vlan code does not check for these attribs in the masked case.
>> It does check for them in the non-masked case and then sets a boolean and
>> checks it in the masked case. I do the same thing for the inner vlan. I
>> check for the attributes in the non-masked case and set a boolean and check
>> the boolean in the masked case. Why is this not sufficient?
> Original code is checking for attributes before referencing them. For
> example  in function ovs_nla_get_match() before extracting eth_type,
> it does check a[OVS_KEY_ATTR_ETHERTYPE]. But If you spot bug in
> current code please send fix for net tree.
> Regarding the Boolean, it is for presence of inner vlan for key
> attribute, mask attribute still could be missing vlan attribute.
> For vlan mask, we can keep check sanity check as outer vlan. It means
> eth_type must be specified and should be 0xffff, and tci mask is
> optional and by default initialized to 0xffff.
You are correct. I was thinking of something else. I had this but must 
have lost it one of the patch revisions. Fixed it in V16.

Patch
diff mbox

diff --git a/net/openvswitch/actions.c b/net/openvswitch/actions.c
index 315f533..09cc1c9 100644
--- a/net/openvswitch/actions.c
+++ b/net/openvswitch/actions.c
@@ -236,7 +236,8 @@  static int pop_vlan(struct sk_buff *skb, struct sw_flow_key *key)
 	if (skb_vlan_tag_present(skb))
 		invalidate_flow_key(key);
 	else
-		key->eth.tci = 0;
+		key->eth.vlan.tci = 0;
+		key->eth.vlan.tpid = 0;
 	return err;
 }
 
@@ -246,7 +247,8 @@  static int push_vlan(struct sk_buff *skb, struct sw_flow_key *key,
 	if (skb_vlan_tag_present(skb))
 		invalidate_flow_key(key);
 	else
-		key->eth.tci = vlan->vlan_tci;
+		key->eth.vlan.tci = vlan->vlan_tci;
+		key->eth.vlan.tpid = vlan->vlan_tpid;
 	return skb_vlan_push(skb, vlan->vlan_tpid,
 			     ntohs(vlan->vlan_tci) & ~VLAN_TAG_PRESENT);
 }
diff --git a/net/openvswitch/flow.c b/net/openvswitch/flow.c
index c8db44a..0f9479c 100644
--- a/net/openvswitch/flow.c
+++ b/net/openvswitch/flow.c
@@ -305,21 +305,83 @@  static bool icmp6hdr_ok(struct sk_buff *skb)
 static int parse_vlan(struct sk_buff *skb, struct sw_flow_key *key)
 {
 	struct qtag_prefix {
-		__be16 eth_type; /* ETH_P_8021Q */
+		__be16 eth_type; /* ETH_P_8021Q  or ETH_P_8021AD */
 		__be16 tci;
 	};
-	struct qtag_prefix *qp;
+	struct qtag_prefix *qp = (struct qtag_prefix *)skb->data;
 
-	if (unlikely(skb->len < sizeof(struct qtag_prefix) + sizeof(__be16)))
+	if (likely(skb_vlan_tag_present(skb))) {
+		key->eth.vlan.tci = htons(skb->vlan_tci);
+		key->eth.vlan.tpid = skb->vlan_proto;
+
+		/* Case where upstream
+		 * processing has already stripped the outer vlan tag.
+		 */
+		if (unlikely(skb->vlan_proto == htons(ETH_P_8021AD))) {
+			if (unlikely(skb->len < sizeof(struct qtag_prefix) +
+					sizeof(__be16))) {
+				key->eth.vlan.tci = 0;
+				return 0;
+			}
+
+			if (unlikely(!pskb_may_pull(skb,
+						    sizeof(struct qtag_prefix) +
+						    sizeof(__be16))))
+				return -ENOMEM;
+
+			qp = (struct qtag_prefix *)skb->data;
+			key->eth.cvlan.tci =
+				qp->tci | htons(VLAN_TAG_PRESENT);
+			key->eth.cvlan.tpid = qp->eth_type;
+
+			__skb_pull(skb, sizeof(struct qtag_prefix));
+		}
 		return 0;
 
-	if (unlikely(!pskb_may_pull(skb, sizeof(struct qtag_prefix) +
-					 sizeof(__be16))))
-		return -ENOMEM;
+	} else if (qp->eth_type == htons(ETH_P_8021AD)) {
+
+		if (unlikely(skb->len < sizeof(struct qtag_prefix) +
+					sizeof(__be16)))
+			return 0;
+
+		if (unlikely(!pskb_may_pull(skb, sizeof(struct qtag_prefix) +
+				sizeof(__be16))))
+			return -ENOMEM;
+
+		qp = (struct qtag_prefix *)skb->data;
+		key->eth.vlan.tci = qp->tci | htons(VLAN_TAG_PRESENT);
+		key->eth.vlan.tpid = qp->eth_type;
+
+		__skb_pull(skb, sizeof(struct qtag_prefix));
 
-	qp = (struct qtag_prefix *) skb->data;
-	key->eth.tci = qp->tci | htons(VLAN_TAG_PRESENT);
-	__skb_pull(skb, sizeof(struct qtag_prefix));
+		if (unlikely(skb->len < sizeof(struct qtag_prefix) +
+					sizeof(__be16)))
+			return 0;
+
+		if (unlikely(!pskb_may_pull(skb, sizeof(struct qtag_prefix) +
+				sizeof(__be16))))
+			return -ENOMEM;
+
+		key->eth.cvlan.tci = qp->tci | htons(VLAN_TAG_PRESENT);
+		key->eth.cvlan.tpid = qp->eth_type;
+
+		__skb_pull(skb, sizeof(struct qtag_prefix));
+
+		return 0;
+
+	} else if (qp->eth_type == htons(ETH_P_8021Q)) {
+		if (unlikely(skb->len < sizeof(struct qtag_prefix) +
+					sizeof(__be16)))
+			return 0;
+
+		if (unlikely(!pskb_may_pull(skb, sizeof(struct qtag_prefix) +
+				sizeof(__be16))))
+			return -ENOMEM;
+		key->eth.vlan.tci = qp->tci | htons(VLAN_TAG_PRESENT);
+		key->eth.vlan.tpid = qp->eth_type;
+
+		__skb_pull(skb, sizeof(struct qtag_prefix));
+	}
 
 	return 0;
 }
@@ -480,12 +542,12 @@  static int key_extract(struct sk_buff *skb, struct sw_flow_key *key)
 	 * update skb->csum here.
 	 */
 
-	key->eth.tci = 0;
-	if (skb_vlan_tag_present(skb))
-		key->eth.tci = htons(skb->vlan_tci);
-	else if (eth->h_proto == htons(ETH_P_8021Q))
-		if (unlikely(parse_vlan(skb, key)))
-			return -ENOMEM;
+	key->eth.vlan.tci = 0;
+	key->eth.vlan.tpid = 0;
+	key->eth.cvlan.tci = 0;
+	key->eth.cvlan.tpid = 0;
+	if (unlikely(parse_vlan(skb, key)))
+		return -ENOMEM;
 
 	key->eth.type = parse_ethertype(skb);
 	if (unlikely(key->eth.type == htons(0)))
diff --git a/net/openvswitch/flow.h b/net/openvswitch/flow.h
index fe527d2..539494e 100644
--- a/net/openvswitch/flow.h
+++ b/net/openvswitch/flow.h
@@ -68,7 +68,16 @@  struct sw_flow_key {
 	struct {
 		u8     src[ETH_ALEN];	/* Ethernet source address. */
 		u8     dst[ETH_ALEN];	/* Ethernet destination address. */
-		__be16 tci;		/* 0 if no VLAN, VLAN_TAG_PRESENT set otherwise. */
+		struct {
+			__be16 tpid;	/* Outer Vlan type 802.1q or 802.1ad.*/
+			__be16 tci;	/* 0 if no VLAN, VLAN_TAG_PRESENT */
+					/* set otherwise. */
+		} vlan;
+		struct {
+			__be16 tpid;	/* Inner Vlan DL_type 802.1q.*/
+			__be16 tci;	/* 0 if no CVLAN, VLAN_TAG_PRESENT */
+					/* set otherwise. */
+		} cvlan;
 		__be16 type;		/* Ethernet frame type. */
 	} eth;
 	union {
diff --git a/net/openvswitch/flow_netlink.c b/net/openvswitch/flow_netlink.c
index c92d6a2..af06683 100644
--- a/net/openvswitch/flow_netlink.c
+++ b/net/openvswitch/flow_netlink.c
@@ -811,6 +811,33 @@  static int metadata_from_nlattrs(struct net *net, struct sw_flow_match *match,
 	return 0;
 }
 
+static int cust_vlan_from_nlattrs(struct sw_flow_match *match,
+				  const struct nlattr *a[],
+				  bool is_mask, bool log)
+{
+	__be16 ctci = 0;
+	__be16 c_tpid = 0;
+
+	ctci = nla_get_be16(a[OVS_KEY_ATTR_VLAN]);
+	c_tpid = nla_get_be16(a[OVS_KEY_ATTR_ETHERTYPE]);
+
+	if (is_mask && c_tpid != htons(0xffff)) {
+		OVS_NLERR(log, "VLAN frames must have an exact match on the CTPID (mask=%x).",
+			  ntohs(c_tpid));
+		return -EINVAL;
+	}
+	if (!(ctci & htons(VLAN_TAG_PRESENT))) {
+		if (is_mask)
+			OVS_NLERR(log, "VLAN CTCI mask does not have exact match for VLAN_TAG_PRESENT bit.");
+		else
+			OVS_NLERR(log, "VLAN CTCI does not have VLAN_TAG_PRESENT bit set.");
+		return -EINVAL;
+	}
+	SW_FLOW_KEY_PUT(match, eth.cvlan.tpid, c_tpid, is_mask);
+	SW_FLOW_KEY_PUT(match, eth.cvlan.tci, ctci, is_mask);
+	return 0;
+}
+
 static int ovs_key_from_nlattrs(struct net *net, struct sw_flow_match *match,
 				u64 attrs, const struct nlattr **a,
 				bool is_mask, bool log)
@@ -845,7 +872,7 @@  static int ovs_key_from_nlattrs(struct net *net, struct sw_flow_match *match,
 			return -EINVAL;
 		}
 
-		SW_FLOW_KEY_PUT(match, eth.tci, tci, is_mask);
+		SW_FLOW_KEY_PUT(match, eth.vlan.tci, tci, is_mask);
 		attrs &= ~(1 << OVS_KEY_ATTR_VLAN);
 	}
 
@@ -1064,6 +1091,86 @@  static void mask_set_nlattr(struct nlattr *attr, u8 val)
 	nlattr_set(attr, val, ovs_key_lens);
 }
 
+static int parse_vlan_from_nlattrs(const struct nlattr **nla,
+				   struct sw_flow_match *match,
+				   u64 *key_attrs, bool *ie_valid,
+				   const struct nlattr **a, bool is_mask,
+				   bool log)
+{
+	int err;
+	const struct nlattr *encap;
+
+	if (!is_mask) {
+		u64 v_attrs = 0;
+
+		err = parse_flow_nlattrs(*nla, a, &v_attrs, log);
+		if (err)
+			return err;
+		/* Another encap attribute here indicates
+		 * the presence of a double tagged vlan.
+		 */
+		if ((v_attrs & (1 << OVS_KEY_ATTR_ETHERTYPE)) &&
+		    eth_type_vlan(nla_get_be16(a[OVS_KEY_ATTR_ETHERTYPE]))) {
+			if (!((v_attrs & (1ULL << OVS_KEY_ATTR_VLAN)) &&
+			      (v_attrs & (1ULL << OVS_KEY_ATTR_ENCAP)))) {
+				OVS_NLERR(log, "Invalid Inner VLAN frame");
+				return -EINVAL;
+			}
+			encap = a[OVS_KEY_ATTR_ENCAP];
+
+			err = cust_vlan_from_nlattrs(match, a, is_mask, log);
+			if (err)
+				return err;
+			*ie_valid = true;
+			*nla = encap;
+
+			/* Insure that tci key attribute isn't
+			 * overwritten by encapsulated customer tci.
+			 * Ethertype is cleared because it is c_tpid.
+			 */
+			v_attrs &= ~(1 << OVS_KEY_ATTR_ENCAP);
+			v_attrs &= ~(1 << OVS_KEY_ATTR_VLAN);
+			v_attrs &= ~(1 << OVS_KEY_ATTR_ETHERTYPE);
+
+			*key_attrs |= v_attrs;
+			err = parse_flow_nlattrs(*nla, a, key_attrs, log);
+			if (err)
+				return err;
+		}
+
+	} else {
+		u64 mask_v_attrs = 0;
+
+		err = parse_flow_mask_nlattrs(*nla, a, &mask_v_attrs, log);
+		if (err)
+			return err;
+
+		if (mask_v_attrs & 1 << OVS_KEY_ATTR_ENCAP) {
+			if (!*ie_valid) {
+				OVS_NLERR(log, "Encap mask attribute is set for non-CVLAN frame.");
+				err = -EINVAL;
+				return err;
+			}
+			encap = a[OVS_KEY_ATTR_ENCAP];
+
+			err = cust_vlan_from_nlattrs(match, a, is_mask, log);
+			if (err)
+				return err;
+			*nla = encap;
+
+			mask_v_attrs &= ~(1 << OVS_KEY_ATTR_ENCAP);
+			mask_v_attrs &= ~(1ULL << OVS_KEY_ATTR_VLAN);
+			mask_v_attrs &= ~(1ULL << OVS_KEY_ATTR_ETHERTYPE);
+
+			*key_attrs |= mask_v_attrs;
+			err = parse_flow_mask_nlattrs(*nla, a, key_attrs, log);
+			if (err)
+				return err;
+		}
+	}
+	return 0;
+}
+
 /**
  * ovs_nla_get_match - parses Netlink attributes into a flow key and
  * mask. In case the 'mask' is NULL, the flow is treated as exact match
@@ -1091,6 +1198,7 @@  int ovs_nla_get_match(struct net *net, struct sw_flow_match *match,
 	u64 key_attrs = 0;
 	u64 mask_attrs = 0;
 	bool encap_valid = false;
+	bool i_encap_valid = false;
 	int err;
 
 	err = parse_flow_nlattrs(nla_key, a, &key_attrs, log);
@@ -1099,11 +1207,11 @@  int ovs_nla_get_match(struct net *net, struct sw_flow_match *match,
 
 	if ((key_attrs & (1 << OVS_KEY_ATTR_ETHERNET)) &&
 	    (key_attrs & (1 << OVS_KEY_ATTR_ETHERTYPE)) &&
-	    (nla_get_be16(a[OVS_KEY_ATTR_ETHERTYPE]) == htons(ETH_P_8021Q))) {
+	    eth_type_vlan(nla_get_be16(a[OVS_KEY_ATTR_ETHERTYPE]))) {
 		__be16 tci;
 
-		if (!((key_attrs & (1 << OVS_KEY_ATTR_VLAN)) &&
-		      (key_attrs & (1 << OVS_KEY_ATTR_ENCAP)))) {
+		if (!((key_attrs & (1ULL << OVS_KEY_ATTR_VLAN)) &&
+		      (key_attrs & (1ULL << OVS_KEY_ATTR_ENCAP)))) {
 			OVS_NLERR(log, "Invalid Vlan frame.");
 			return -EINVAL;
 		}
@@ -1115,9 +1223,12 @@  int ovs_nla_get_match(struct net *net, struct sw_flow_match *match,
 		encap_valid = true;
 
 		if (tci & htons(VLAN_TAG_PRESENT)) {
-			err = parse_flow_nlattrs(encap, a, &key_attrs, log);
+			err = parse_vlan_from_nlattrs(&encap, match, &key_attrs,
+						      &i_encap_valid, a, false,
+						      log);
 			if (err)
 				return err;
+
 		} else if (!tci) {
 			/* Corner case for truncated 802.1Q header. */
 			if (nla_len(encap)) {
@@ -1169,7 +1280,7 @@  int ovs_nla_get_match(struct net *net, struct sw_flow_match *match,
 			goto free_newmask;
 
 		/* Always match on tci. */
-		SW_FLOW_KEY_PUT(match, eth.tci, htons(0xffff), true);
+		SW_FLOW_KEY_PUT(match, eth.vlan.tci, htons(0xffff), true);
 
 		if (mask_attrs & 1 << OVS_KEY_ATTR_ENCAP) {
 			__be16 eth_type = 0;
@@ -1188,10 +1299,13 @@  int ovs_nla_get_match(struct net *net, struct sw_flow_match *match,
 			if (eth_type == htons(0xffff)) {
 				mask_attrs &= ~(1 << OVS_KEY_ATTR_ETHERTYPE);
 				encap = a[OVS_KEY_ATTR_ENCAP];
-				err = parse_flow_mask_nlattrs(encap, a,
-							      &mask_attrs, log);
+				err = parse_vlan_from_nlattrs(&encap, match,
+							      &mask_attrs,
+							      &i_encap_valid,
+							      a, true, log);
 				if (err)
 					goto free_newmask;
+
 			} else {
 				OVS_NLERR(log, "VLAN frames must have an exact match on the TPID (mask=%x).",
 					  ntohs(eth_type));
@@ -1320,6 +1434,7 @@  static int __ovs_nla_put_key(const struct sw_flow_key *swkey,
 {
 	struct ovs_key_ethernet *eth_key;
 	struct nlattr *nla, *encap;
+	struct nlattr *in_encap = NULL;
 
 	if (nla_put_u32(skb, OVS_KEY_ATTR_RECIRC_ID, output->recirc_id))
 		goto nla_put_failure;
@@ -1368,17 +1483,36 @@  static int __ovs_nla_put_key(const struct sw_flow_key *swkey,
 	ether_addr_copy(eth_key->eth_src, output->eth.src);
 	ether_addr_copy(eth_key->eth_dst, output->eth.dst);
 
-	if (swkey->eth.tci || swkey->eth.type == htons(ETH_P_8021Q)) {
+	if (swkey->eth.vlan.tci || eth_type_vlan(swkey->eth.type)) {
 		__be16 eth_type;
-		eth_type = !is_mask ? htons(ETH_P_8021Q) : htons(0xffff);
+
+		eth_type = !is_mask ? output->eth.vlan.tpid : htons(0xffff);
+
 		if (nla_put_be16(skb, OVS_KEY_ATTR_ETHERTYPE, eth_type) ||
-		    nla_put_be16(skb, OVS_KEY_ATTR_VLAN, output->eth.tci))
+		    nla_put_be16(skb, OVS_KEY_ATTR_VLAN, output->eth.vlan.tci))
 			goto nla_put_failure;
 		encap = nla_nest_start(skb, OVS_KEY_ATTR_ENCAP);
-		if (!swkey->eth.tci)
+		if (!swkey->eth.vlan.tci)
 			goto unencap;
-	} else
+		if (swkey->eth.cvlan.tci) {
+			__be16 eth_type;
+
+			/* Customer tci is nested but uses same key attribute.
+			 */
+			eth_type = !is_mask ? output->eth.cvlan.tpid :
+					      htons(0xffff);
+			if (nla_put_be16(skb, OVS_KEY_ATTR_ETHERTYPE,
+					 eth_type) ||
+			    nla_put_be16(skb, OVS_KEY_ATTR_VLAN,
+					 output->eth.cvlan.tci))
+				goto nla_put_failure;
+			in_encap = nla_nest_start(skb, OVS_KEY_ATTR_ENCAP);
+			if (!swkey->eth.cvlan.tci)
+				goto unencap;
+		}
+	} else {
 		encap = NULL;
+	}
 
 	if (swkey->eth.type == htons(ETH_P_802_2)) {
 		/*
@@ -1525,6 +1659,8 @@  static int __ovs_nla_put_key(const struct sw_flow_key *swkey,
 unencap:
 	if (encap)
 		nla_nest_end(skb, encap);
+	if (in_encap)
+		nla_nest_end(skb, in_encap);
 
 	return 0;
 
@@ -2174,7 +2310,7 @@  static int __ovs_nla_copy_actions(struct net *net, const struct nlattr *attr,
 
 		case OVS_ACTION_ATTR_PUSH_VLAN:
 			vlan = nla_data(a);
-			if (vlan->vlan_tpid != htons(ETH_P_8021Q))
+			if (!eth_type_vlan(vlan->vlan_tpid))
 				return -EINVAL;
 			if (!(vlan->vlan_tci & htons(VLAN_TAG_PRESENT)))
 				return -EINVAL;
@@ -2279,7 +2415,7 @@  int ovs_nla_copy_actions(struct net *net, const struct nlattr *attr,
 
 	(*sfa)->orig_len = nla_len(attr);
 	err = __ovs_nla_copy_actions(net, attr, key, 0, sfa, key->eth.type,
-				     key->eth.tci, log);
+				     key->eth.vlan.tci, log);
 	if (err)
 		ovs_nla_free_flow_actions(*sfa);
 
diff --git a/net/openvswitch/vport-netdev.c b/net/openvswitch/vport-netdev.c
index f7e8dcc..d2581b7 100644
--- a/net/openvswitch/vport-netdev.c
+++ b/net/openvswitch/vport-netdev.c
@@ -194,7 +194,9 @@  static unsigned int packet_length(const struct sk_buff *skb)
 {
 	unsigned int length = skb->len - ETH_HLEN;
 
-	if (skb->protocol == htons(ETH_P_8021Q))
+	if (eth_type_vlan(skb->protocol))
+		length -= VLAN_HLEN;
+	if (skb->protocol == htons(ETH_P_8021AD))
 		length -= VLAN_HLEN;
 
 	return length;