openvswitch: always update flow key after nat
Subsystems (3)
Aaron Conole
maintainerPATCH v2 1/1 openvswitch: always update flow key after nat
During NAT, a tuple collision may occur. When this happens, openvswitch will make a second pass through NAT which will perform additional packet modification. This will update the skb data, but not the flow key that OVS uses. This means that future flow lookups, and packet matches will have incorrect data. This has been supported since 5d50aa83e2c8 ("openvswitch: support asymmetric conntrack").
That commit failed to properly update the sw_flow_key attributes, since it only called the ovs_ct_nat_update_key once, rather than each time ovs_ct_nat_execute was called. As these two operations are linked, the ovs_ct_nat_execute() function should always make sure that the sw_flow_key is updated after a successful call through NAT infrastructure.
---v1->v2: removed forward decl., moved the ovs_nat_update_key function
made sure it compiles with NF_NAT disabled and enabled
net/openvswitch/conntrack.c | 118 ++++++++++++++++++------------------ 1 file changed, 59 insertions(+), 59 deletions(-)
net/openvswitch/conntrack.c +59 -59
@@ -734,6 +734,57 @@ static bool skb_nfct_cached(struct net *net, }#if IS_ENABLED (CONFIG_NF_NAT ) +static void ovs_nat_update_key (struct sw_flow_key *key, +const struct sk_buff *skb, +enum nf_nat_manip_type maniptype) +{ +if (maniptype ==NF_NAT_MANIP_SRC ) { +__be16 src; + + key->ct_state |=OVS_CS_F_SRC_NAT ; +if (key->eth.type ==htons (ETH_P_IP )) + key->ipv4.addr.src =ip_hdr (skb)->saddr; +else if (key->eth.type ==htons (ETH_P_IPV6 )) +memcpy (&key->ipv6.addr.src, &ipv6_hdr (skb)->saddr, +sizeof (key->ipv6.addr.src)); +else +return ; + +if (key->ip.proto ==IPPROTO_UDP ) + src =udp_hdr (skb)->source; +else if (key->ip.proto ==IPPROTO_TCP ) + src =tcp_hdr (skb)->source; +else if (key->ip.proto ==IPPROTO_SCTP ) + src =sctp_hdr (skb)->source; +else +return ; + + key->tp.src = src; + }else { +__be16 dst; + + key->ct_state |=OVS_CS_F_DST_NAT ; +if (key->eth.type ==htons (ETH_P_IP )) + key->ipv4.addr.dst =ip_hdr (skb)->daddr; +else if (key->eth.type ==htons (ETH_P_IPV6 )) +memcpy (&key->ipv6.addr.dst, &ipv6_hdr (skb)->daddr, +sizeof (key->ipv6.addr.dst)); +else +return ; + +if (key->ip.proto ==IPPROTO_UDP ) + dst =udp_hdr (skb)->dest; +else if (key->ip.proto ==IPPROTO_TCP ) + dst =tcp_hdr (skb)->dest; +else if (key->ip.proto ==IPPROTO_SCTP ) + dst =sctp_hdr (skb)->dest; +else +return ; + + key->tp.dst = dst; + } +} +/* Modelled after nf_nat_ipv[46]_fn(). * range is only used for new, uninitialized NAT state. * Returns either NF_ACCEPT or NF_DROP.
@@ -741,7 +792,7 @@ static bool skb_nfct_cached(struct net *net,static int ovs_ct_nat_execute (struct sk_buff *skb,struct nf_conn *ct,enum ip_conntrack_info ctinfo,const struct nf_nat_range2 *range, -enum nf_nat_manip_type maniptype) +enum nf_nat_manip_type maniptype,struct sw_flow_key *key) {int hooknum, nh_off, err =NF_ACCEPT ;
@@ -813,58 +864,11 @@ static int ovs_ct_nat_execute(struct sk_buff *skb, struct nf_conn *ct, push:skb_push_rcsum (skb, nh_off); -return err; -} - -static void ovs_nat_update_key (struct sw_flow_key *key, -const struct sk_buff *skb, -enum nf_nat_manip_type maniptype) -{ -if (maniptype ==NF_NAT_MANIP_SRC ) { -__be16 src; - - key->ct_state |=OVS_CS_F_SRC_NAT ; -if (key->eth.type ==htons (ETH_P_IP )) - key->ipv4.addr.src =ip_hdr (skb)->saddr; -else if (key->eth.type ==htons (ETH_P_IPV6 )) -memcpy (&key->ipv6.addr.src, &ipv6_hdr (skb)->saddr, -sizeof (key->ipv6.addr.src)); -else -return ; - -if (key->ip.proto ==IPPROTO_UDP ) - src =udp_hdr (skb)->source; -else if (key->ip.proto ==IPPROTO_TCP ) - src =tcp_hdr (skb)->source; -else if (key->ip.proto ==IPPROTO_SCTP ) - src =sctp_hdr (skb)->source; -else -return ; - - key->tp.src = src; - }else { -__be16 dst; - - key->ct_state |=OVS_CS_F_DST_NAT ; -if (key->eth.type ==htons (ETH_P_IP )) - key->ipv4.addr.dst =ip_hdr (skb)->daddr; -else if (key->eth.type ==htons (ETH_P_IPV6 )) -memcpy (&key->ipv6.addr.dst, &ipv6_hdr (skb)->daddr, -sizeof (key->ipv6.addr.dst)); -else -return ; - -if (key->ip.proto ==IPPROTO_UDP ) - dst =udp_hdr (skb)->dest; -else if (key->ip.proto ==IPPROTO_TCP ) - dst =tcp_hdr (skb)->dest; -else if (key->ip.proto ==IPPROTO_SCTP ) - dst =sctp_hdr (skb)->dest; -else -return ; +/* Update the flow key if NAT successful. */ +if (err ==NF_ACCEPT ) +ovs_nat_update_key (key, skb, maniptype); - key->tp.dst = dst; - } +return err; }/* Returns NF_DROP if the packet should be dropped, NF_ACCEPT otherwise. */
@@ -906,7 +910,7 @@ static int ovs_ct_nat(struct net *net, struct sw_flow_key *key, }else {return NF_ACCEPT ;/* Connection is not NATed. */ } - err =ovs_ct_nat_execute (skb, ct, ctinfo, &info->range, maniptype); + err =ovs_ct_nat_execute (skb, ct, ctinfo, &info->range, maniptype, key);if (err ==NF_ACCEPT && ct->status &IPS_DST_NAT ) {if (ct->status &IPS_SRC_NAT ) {
@@ -916,17 +920,13 @@ static int ovs_ct_nat(struct net *net, struct sw_flow_key *key, maniptype =NF_NAT_MANIP_SRC ; err =ovs_ct_nat_execute (skb, ct, ctinfo, &info->range, - maniptype); + maniptype, key); }else if (CTINFO2DIR (ctinfo) ==IP_CT_DIR_ORIGINAL ) { err =ovs_ct_nat_execute (skb, ct, ctinfo,NULL , -NF_NAT_MANIP_SRC ); +NF_NAT_MANIP_SRC , key); } } -/* Mark NAT done if successful and update the flow key. */ -if (err ==NF_ACCEPT ) -ovs_nat_update_key (key, skb, maniptype); -return err; }#else /* !CONFIG_NF_NAT */
Signature · 1 line
2.31.1
Quoted 13 lines · Aaron Conole
During NAT, a tuple collision may occur. When this happens, openvswitch will make a second pass through NAT which will perform additional packet modification. This will update the skb data, but not the flow key that OVS uses. This means that future flow lookups, and packet matches will have incorrect data. This has been supported since
5d50aa83e2c8("openvswitch: support asymmetric conntrack").That commit failed to properly update the
sw_flow_keyattributes, since it only called theovs_ct_nat_update_keyonce, rather than each timeovs_ct_nat_executewas called. As these two operations are linked, theovs_ct_nat_execute()function should always make sure that thesw_flow_keyis updated after a successful call through NAT infrastructure.
Quoted 4 lines
Fixes 5d50aa83e2c8 ("openvswitch: support asymmetric conntrack")Cc Dumitru Ceara <[email protected]>Cc Numan Siddique <[email protected]>Signed-off-by Aaron Conole <[email protected]>
You were right about the diff, it really looks messy and I had to apply it to review it :)
The patch looks fine to me!!
Quoted 172 lines
---
v1->v2: removed forward decl., moved theovs_nat_update_keyfunction
made sure it compiles withNF_NATdisabled and enablednet/openvswitch/conntrack.c | 118 ++++++++++++++++++------------------ 1 file changed, 59 insertions(+), 59 deletions(-)net/openvswitch/conntrack.c +59
-59@@ -734,6 +734,57 @@ static bool skb_nfct_cached(struct net *net, }#if IS_ENABLED (CONFIG_NF_NAT ) +static void ovs_nat_update_key (struct sw_flow_key *key, +const struct sk_buff *skb, +enum nf_nat_manip_type maniptype) +{ +if (maniptype ==NF_NAT_MANIP_SRC ) { +__be16 src; + + key->ct_state |=OVS_CS_F_SRC_NAT ; +if (key->eth.type ==htons (ETH_P_IP )) + key->ipv4.addr.src =ip_hdr (skb)->saddr; +else if (key->eth.type ==htons (ETH_P_IPV6 )) +memcpy (&key->ipv6.addr.src, &ipv6_hdr (skb)->saddr, +sizeof (key->ipv6.addr.src)); +else +return ; + +if (key->ip.proto ==IPPROTO_UDP ) + src =udp_hdr (skb)->source; +else if (key->ip.proto ==IPPROTO_TCP ) + src =tcp_hdr (skb)->source; +else if (key->ip.proto ==IPPROTO_SCTP ) + src =sctp_hdr (skb)->source; +else +return ; + + key->tp.src = src; + }else { +__be16 dst; + + key->ct_state |=OVS_CS_F_DST_NAT ; +if (key->eth.type ==htons (ETH_P_IP )) + key->ipv4.addr.dst =ip_hdr (skb)->daddr; +else if (key->eth.type ==htons (ETH_P_IPV6 )) +memcpy (&key->ipv6.addr.dst, &ipv6_hdr (skb)->daddr, +sizeof (key->ipv6.addr.dst)); +else +return ; + +if (key->ip.proto ==IPPROTO_UDP ) + dst =udp_hdr (skb)->dest; +else if (key->ip.proto ==IPPROTO_TCP ) + dst =tcp_hdr (skb)->dest; +else if (key->ip.proto ==IPPROTO_SCTP ) + dst =sctp_hdr (skb)->dest; +else +return ; + + key->tp.dst = dst; + } +} +/* Modelled after nf_nat_ipv[46]_fn(). * range is only used for new, uninitialized NAT state. * Returns either NF_ACCEPT or NF_DROP. @@ -741,7 +792,7 @@ static bool skb_nfct_cached(struct net *net,static int ovs_ct_nat_execute (struct sk_buff *skb,struct nf_conn *ct,enum ip_conntrack_info ctinfo,const struct nf_nat_range2 *range, -enum nf_nat_manip_type maniptype) +enum nf_nat_manip_type maniptype,struct sw_flow_key *key) {int hooknum, nh_off, err =NF_ACCEPT ;@@ -813,58 +864,11 @@ static int ovs_ct_nat_execute(struct sk_buff *skb, struct nf_conn *ct, push:skb_push_rcsum (skb, nh_off); -return err; -} - -static void ovs_nat_update_key (struct sw_flow_key *key, -const struct sk_buff *skb, -enum nf_nat_manip_type maniptype) -{ -if (maniptype ==NF_NAT_MANIP_SRC ) { -__be16 src; - - key->ct_state |=OVS_CS_F_SRC_NAT ; -if (key->eth.type ==htons (ETH_P_IP )) - key->ipv4.addr.src =ip_hdr (skb)->saddr; -else if (key->eth.type ==htons (ETH_P_IPV6 )) -memcpy (&key->ipv6.addr.src, &ipv6_hdr (skb)->saddr, -sizeof (key->ipv6.addr.src)); -else -return ; - -if (key->ip.proto ==IPPROTO_UDP ) - src =udp_hdr (skb)->source; -else if (key->ip.proto ==IPPROTO_TCP ) - src =tcp_hdr (skb)->source; -else if (key->ip.proto ==IPPROTO_SCTP ) - src =sctp_hdr (skb)->source; -else -return ; - - key->tp.src = src; - }else { -__be16 dst; - - key->ct_state |=OVS_CS_F_DST_NAT ; -if (key->eth.type ==htons (ETH_P_IP )) - key->ipv4.addr.dst =ip_hdr (skb)->daddr; -else if (key->eth.type ==htons (ETH_P_IPV6 )) -memcpy (&key->ipv6.addr.dst, &ipv6_hdr (skb)->daddr, -sizeof (key->ipv6.addr.dst)); -else -return ; - -if (key->ip.proto ==IPPROTO_UDP ) - dst =udp_hdr (skb)->dest; -else if (key->ip.proto ==IPPROTO_TCP ) - dst =tcp_hdr (skb)->dest; -else if (key->ip.proto ==IPPROTO_SCTP ) - dst =sctp_hdr (skb)->dest; -else -return ; +/* Update the flow key if NAT successful. */ +if (err ==NF_ACCEPT ) +ovs_nat_update_key (key, skb, maniptype); - key->tp.dst = dst; - } +return err; }/* Returns NF_DROP if the packet should be dropped, NF_ACCEPT otherwise. */ @@ -906,7 +910,7 @@ static int ovs_ct_nat(struct net *net, struct sw_flow_key *key, }else {return NF_ACCEPT ;/* Connection is not NATed. */ } - err =ovs_ct_nat_execute (skb, ct, ctinfo, &info->range, maniptype); + err =ovs_ct_nat_execute (skb, ct, ctinfo, &info->range, maniptype, key);if (err ==NF_ACCEPT && ct->status &IPS_DST_NAT ) {if (ct->status &IPS_SRC_NAT ) {@@ -916,17 +920,13 @@ static int ovs_ct_nat(struct net *net, struct sw_flow_key *key, maniptype =NF_NAT_MANIP_SRC ; err =ovs_ct_nat_execute (skb, ct, ctinfo, &info->range, - maniptype); + maniptype, key); }else if (CTINFO2DIR (ctinfo) ==IP_CT_DIR_ORIGINAL ) { err =ovs_ct_nat_execute (skb, ct, ctinfo,NULL , -NF_NAT_MANIP_SRC ); +NF_NAT_MANIP_SRC , key); } } -/* Mark NAT done if successful and update the flow key. */ -if (err ==NF_ACCEPT ) -ovs_nat_update_key (key, skb, maniptype); -return err; }#else /* !CONFIG_NF_NAT */ Signature · 1 line
2.31.1
This patch was applied to netdev/net.git (master) by Jakub Kicinski <[email protected]>:
Hello:
This patch was applied to netdev/net.git (master)
by Jakub Kicinski <[email protected]>:
Quoted 8 lines · On Fri, 18 Mar 2022 08:43:19 -0400 you wrote:
During NAT, a tuple collision may occur. When this happens, openvswitch will make a second pass through NAT which will perform additional packet modification. This will update the skb data, but not the flow key that OVS uses. This means that future flow lookups, and packet matches will have incorrect data. This has been supported since
5d50aa83e2c8("openvswitch: support asymmetric conntrack").[...]
Here is the summary with links:
- [net,v2] openvswitch: always update flow key after nat
https://git.kernel.org/netdev/net/c/60b44ca6bd75
You are awesome, thank you!
Signature · 2 lines
Deet-doot-dot, I am a bot. https://korg.docs.kernel.org/patchwork/pwbot.html