Revert "netfilter: tproxy: do not assign timewait sockets to skb->sk"
Subsystems (2)
Florian Westphal
maintainerPATCH 1/1 Revert "netfilter: tproxy: do not assign timewait sockets to skb->sk"
This reverts commit d503b30bd648b3cb4e5f50b65d27e389960cc6d9.
Nowadays we have early socket demux; all the input code paths should handle "skb->sk points to timewait sock" properly.
--- include/net/netfilter/nf_tproxy_core.h | 12 +++++++++++- net/netfilter/nf_tproxy_core.c | 27 +++++++++++++++------------ net/netfilter/xt_TPROXY.c | 22 ++-------------------- net/netfilter/xt_socket.c | 13 ++----------- 4 files changed, 30 insertions(+), 44 deletions(-)
include/net/netfilter/nf_tproxy_core.h +11 -1
@@ -203,8 +203,18 @@ nf_tproxy_get_sock_v6(struct net *net, const u8 protocol, }#endif +static inline void +nf_tproxy_put_sock (struct sock *sk) +{ +/* TIME_WAIT inet sockets have to be handled differently */ +if ((sk->sk_protocol ==IPPROTO_TCP ) && (sk->sk_state ==TCP_TIME_WAIT )) +inet_twsk_put (inet_twsk (sk)); +else +sock_put (sk); +} +/* assign a socket to the skb -- consumes sk */ -void +int nf_tproxy_assign_sock (struct sk_buff *skb,struct sock *sk);#endif
net/netfilter/nf_tproxy_core.c +15 -12
@@ -28,23 +28,26 @@ nf_tproxy_destructor(struct sk_buff *skb) skb->destructor =NULL ;if (sk) -sock_put (sk); +nf_tproxy_put_sock (sk); }/* consumes sk */ -void +int nf_tproxy_assign_sock (struct sk_buff *skb,struct sock *sk) { -/* assigning tw sockets complicates things; most -* skb->sk->X checks would have to test sk->sk_state first */ -if (sk->sk_state ==TCP_TIME_WAIT ) { -inet_twsk_put (inet_twsk (sk)); -return ; - } - -skb_orphan (skb); - skb->sk = sk; - skb->destructor = nf_tproxy_destructor; +bool transparent = (sk->sk_state ==TCP_TIME_WAIT ) ? +inet_twsk (sk)->tw_transparent : +inet_sk (sk)->transparent; + +if (transparent) { +skb_orphan (skb); + skb->sk = sk; + skb->destructor = nf_tproxy_destructor; +return 1 ; + }else +nf_tproxy_put_sock (sk); + +return 0 ; }EXPORT_SYMBOL_GPL (nf_tproxy_assign_sock);
net/netfilter/xt_TPROXY.c +2 -20
@@ -33,20 +33,6 @@#include <net/netfilter/nf_tproxy_core.h> #include <linux/netfilter/xt_TPROXY.h> -static bool tproxy_sk_is_transparent (struct sock *sk) -{ -if (sk->sk_state !=TCP_TIME_WAIT ) { -if (inet_sk (sk)->transparent) -return true; -sock_put (sk); - }else { -if (inet_twsk (sk)->tw_transparent) -return true; -inet_twsk_put (inet_twsk (sk)); - } -return false; -} -static inline __be32 tproxy_laddr4 (struct sk_buff *skb,__be32 user_laddr,__be32 daddr) {
@@ -155,7 +141,7 @@ tproxy_tg4(struct sk_buff *skb, __be32 laddr, __be16 lport, skb->dev, NFT_LOOKUP_LISTENER);/* NOTE: assign_sock consumes our sk reference */ -if (sk &&tproxy_sk_is_transparent (sk)) { +if (sk &&nf_tproxy_assign_sock (skb, sk)) {/* This should be in a separate target, but we don't do multiple targets on the same rule yet */ skb->mark = (skb->mark & ~mark_mask) ^ mark_value;
@@ -163,8 +149,6 @@ tproxy_tg4(struct sk_buff *skb, __be32 laddr, __be16 lport,pr_debug ("redirecting: proto %hhu %pI4:%hu -> %pI4:%hu, mark: %x\n" , iph->protocol, &iph->daddr,ntohs (hp->dest), &laddr,ntohs (lport), skb->mark); - -nf_tproxy_assign_sock (skb, sk);return NF_ACCEPT ; }
@@ -322,7 +306,7 @@ tproxy_tg6_v1(struct sk_buff *skb, const struct xt_action_param *par) par->in, NFT_LOOKUP_LISTENER);/* NOTE: assign_sock consumes our sk reference */ -if (sk &&tproxy_sk_is_transparent (sk)) { +if (sk &&nf_tproxy_assign_sock (skb, sk)) {/* This should be in a separate target, but we don't do multiple targets on the same rule yet */ skb->mark = (skb->mark & ~tgi->mark_mask) ^ tgi->mark_value;
@@ -330,8 +314,6 @@ tproxy_tg6_v1(struct sk_buff *skb, const struct xt_action_param *par)pr_debug ("redirecting: proto %hhu %pI6:%hu -> %pI6:%hu, mark: %x\n" , tproto, &iph->saddr,ntohs (hp->source), laddr,ntohs (lport), skb->mark); - -nf_tproxy_assign_sock (skb, sk);return NF_ACCEPT ; }
net/netfilter/xt_socket.c +2 -11
@@ -35,15 +35,6 @@#include <net/netfilter/nf_conntrack.h> #endif -static void -xt_socket_put_sk (struct sock *sk) -{ -if (sk->sk_state ==TCP_TIME_WAIT ) -inet_twsk_put (inet_twsk (sk)); -else -sock_put (sk); -} - static int extract_icmp4_fields(const struct sk_buff *skb, u8 *protocol,
@@ -176,7 +167,7 @@ socket_match(const struct sk_buff *skb, struct xt_action_param *par, inet_twsk(sk)->tw_transparent));if (sk != skb->sk) -xt_socket_put_sk (sk); +nf_tproxy_put_sock (sk);if (wildcard || !transparent) sk =NULL ;
@@ -315,7 +306,7 @@ socket_mt6_v1(const struct sk_buff *skb, struct xt_action_param *par) inet_twsk(sk)->tw_transparent));if (sk != skb->sk) -xt_socket_put_sk (sk); +nf_tproxy_put_sock (sk);if (wildcard || !transparent) sk =NULL ;
Signature · 1 line
1.7.8.6
Quoted 5 lines · Florian Westphal
This reverts commit
d503b30bd648b3cb4e5f50b65d27e389960cc6d9.Nowadays we have early socket demux; all the input code paths should handle "
skb->skpoints to timewait sock" properly.
Quoted 7 lines
Signed-off-by Florian Westphal <[email protected]>--- include/net/netfilter/nf_tproxy_core.h | 12 +++++++++++- net/netfilter/nf_tproxy_core.c | 27 +++++++++++++++------------ net/netfilter/xt_TPROXY.c | 22 ++-------------------- net/netfilter/xt_socket.c | 13 ++----------- 4 files changed, 30 insertions(+), 44 deletions(-)
BTW, why xt_socket ignores listener bound on INADDR_ANY ?
Pablo Neira Ayuso
maintainerQuoted 4 lines · Florian Westphal
This reverts commit
d503b30bd648b3cb4e5f50b65d27e389960cc6d9.Nowadays we have early socket demux; all the input code paths should handle "
skb->skpoints to timewait sock" properly.
I'm fine with this but I think that we should have some nf_sock_put definition somewhere else. Using the tproxy namespace from xt_socket doesn't seem nice to me.
Florian Westphal
maintainerQuoted 9 lines · Pablo Neira Ayuso
Quoted 4 lines · Florian Westphal
This reverts commit
d503b30bd648b3cb4e5f50b65d27e389960cc6d9.Nowadays we have early socket demux; all the input code paths should handle "
skb->skpoints to timewait sock" properly.I'm fine with this but I think that we should have some
nf_sock_putdefinition somewhere else. Using the tproxy namespace fromxt_socketdoesn't seem nice to me.
I'll respin this revert to use sock_edemux() in its place instead, afaics we should be able to use that as replacement.
Thanks.