diff mbox series

[net] net/tcp: Don't consider TCP_CLOSE in TCP_AO_ESTABLISHED

Message ID 20240529-tcp_ao-sk_state-v1-1-d69b5d323c52@gmail.com (mailing list archive)
State Accepted
Commit 33700a0c9b562700c28d31360a5f04508f459a45
Delegated to: Netdev Maintainers
Headers show
Series [net] net/tcp: Don't consider TCP_CLOSE in TCP_AO_ESTABLISHED | expand

Checks

Context Check Description
netdev/series_format success Single patches do not need cover letters
netdev/tree_selection success Clearly marked for net
netdev/ynl success Generated files up to date; no warnings/errors; no diff in generated;
netdev/fixes_present success Fixes tag present in non-next series
netdev/header_inline success No static functions without inline keyword in header files
netdev/build_32bit success Errors and warnings before: 1805 this patch: 1805
netdev/build_tools success Errors and warnings before: 0 this patch: 0
netdev/cc_maintainers fail 2 blamed authors not CCed: noureddine@arista.com fruggeri@arista.com; 2 maintainers not CCed: noureddine@arista.com fruggeri@arista.com
netdev/build_clang success Errors and warnings before: 924 this patch: 924
netdev/verify_signedoff success Signed-off-by tag matches author and committer
netdev/deprecated_api success None detected
netdev/check_selftest success No net selftest shell script
netdev/verify_fixes success Fixes tag looks correct
netdev/build_allmodconfig_warn success Errors and warnings before: 1834 this patch: 1834
netdev/checkpatch success total: 0 errors, 0 warnings, 0 checks, 70 lines checked
netdev/build_clang_rust success No Rust files in patch. Skipping build
netdev/kdoc success Errors and warnings before: 0 this patch: 0
netdev/source_inline success Was 0 now: 0
netdev/contest success net-next-2024-05-30--06-00 (tests: 1042)

Commit Message

Dmitry Safonov via B4 Relay May 29, 2024, 5:29 p.m. UTC 
From: Dmitry Safonov <0x7f454c46@gmail.com>

TCP_CLOSE may or may not have current/rnext keys and should not be
considered "established". The fast-path for TCP_CLOSE is
SKB_DROP_REASON_TCP_CLOSE. This is what tcp_rcv_state_process() does
anyways. Add an early drop path to not spend any time verifying
segment signatures for sockets in TCP_CLOSE state.

Cc: stable@vger.kernel.org # v6.7
Fixes: 0a3a809089eb ("net/tcp: Verify inbound TCP-AO signed segments")
Signed-off-by: Dmitry Safonov <0x7f454c46@gmail.com>
---
 include/net/tcp_ao.h |  7 ++++---
 net/ipv4/tcp_ao.c    | 13 +++++++++----
 2 files changed, 13 insertions(+), 7 deletions(-)


---
base-commit: e0cce98fe279b64f4a7d81b7f5c3a23d80b92fbc
change-id: 20240529-tcp_ao-sk_state-4eb21b045001

Best regards,

Comments

Simon Horman June 1, 2024, 12:14 p.m. UTC | #1 
On Wed, May 29, 2024 at 06:29:32PM +0100, Dmitry Safonov via B4 Relay wrote:
> From: Dmitry Safonov <0x7f454c46@gmail.com>
> 
> TCP_CLOSE may or may not have current/rnext keys and should not be
> considered "established". The fast-path for TCP_CLOSE is
> SKB_DROP_REASON_TCP_CLOSE. This is what tcp_rcv_state_process() does
> anyways. Add an early drop path to not spend any time verifying
> segment signatures for sockets in TCP_CLOSE state.
> 
> Cc: stable@vger.kernel.org # v6.7
> Fixes: 0a3a809089eb ("net/tcp: Verify inbound TCP-AO signed segments")
> Signed-off-by: Dmitry Safonov <0x7f454c46@gmail.com>

...

> diff --git a/net/ipv4/tcp_ao.c b/net/ipv4/tcp_ao.c
> index 781b67a52571..37c42b63ff99 100644
> --- a/net/ipv4/tcp_ao.c
> +++ b/net/ipv4/tcp_ao.c
> @@ -933,6 +933,7 @@ tcp_inbound_ao_hash(struct sock *sk, const struct sk_buff *skb,
>  	struct tcp_ao_key *key;
>  	__be32 sisn, disn;
>  	u8 *traffic_key;
> +	int state;
>  	u32 sne = 0;

Hi Dimitry,

It's probably not a good reason to respon this patch, but if you do make a
v2 for some other reason, please consider reverse xmas tree order -
longest line to shortest for local variable declarations - here.

I'll leave actual review of this patch to others.

...
patchwork-bot+netdevbpf@kernel.org June 1, 2024, 11:40 p.m. UTC | #2 
Hello:

This patch was applied to netdev/net.git (main)
by Jakub Kicinski <kuba@kernel.org>:

On Wed, 29 May 2024 18:29:32 +0100 you wrote:
> From: Dmitry Safonov <0x7f454c46@gmail.com>
> 
> TCP_CLOSE may or may not have current/rnext keys and should not be
> considered "established". The fast-path for TCP_CLOSE is
> SKB_DROP_REASON_TCP_CLOSE. This is what tcp_rcv_state_process() does
> anyways. Add an early drop path to not spend any time verifying
> segment signatures for sockets in TCP_CLOSE state.
> 
> [...]

Here is the summary with links:
  - [net] net/tcp: Don't consider TCP_CLOSE in TCP_AO_ESTABLISHED
    https://git.kernel.org/netdev/net/c/33700a0c9b56

You are awesome, thank you!
diff mbox series

Patch

diff --git a/include/net/tcp_ao.h b/include/net/tcp_ao.h
index 471e177362b4..5d8e9ed2c005 100644
--- a/include/net/tcp_ao.h
+++ b/include/net/tcp_ao.h
@@ -86,7 +86,8 @@  static inline int tcp_ao_sizeof_key(const struct tcp_ao_key *key)
 struct tcp_ao_info {
 	/* List of tcp_ao_key's */
 	struct hlist_head	head;
-	/* current_key and rnext_key aren't maintained on listen sockets.
+	/* current_key and rnext_key are maintained on sockets
+	 * in TCP_AO_ESTABLISHED states.
 	 * Their purpose is to cache keys on established connections,
 	 * saving needless lookups. Never dereference any of them from
 	 * listen sockets.
@@ -201,9 +202,9 @@  struct tcp6_ao_context {
 };
 
 struct tcp_sigpool;
+/* Established states are fast-path and there always is current_key/rnext_key */
 #define TCP_AO_ESTABLISHED (TCPF_ESTABLISHED | TCPF_FIN_WAIT1 | TCPF_FIN_WAIT2 | \
-			    TCPF_CLOSE | TCPF_CLOSE_WAIT | \
-			    TCPF_LAST_ACK | TCPF_CLOSING)
+			    TCPF_CLOSE_WAIT | TCPF_LAST_ACK | TCPF_CLOSING)
 
 int tcp_ao_transmit_skb(struct sock *sk, struct sk_buff *skb,
 			struct tcp_ao_key *key, struct tcphdr *th,
diff --git a/net/ipv4/tcp_ao.c b/net/ipv4/tcp_ao.c
index 781b67a52571..37c42b63ff99 100644
--- a/net/ipv4/tcp_ao.c
+++ b/net/ipv4/tcp_ao.c
@@ -933,6 +933,7 @@  tcp_inbound_ao_hash(struct sock *sk, const struct sk_buff *skb,
 	struct tcp_ao_key *key;
 	__be32 sisn, disn;
 	u8 *traffic_key;
+	int state;
 	u32 sne = 0;
 
 	info = rcu_dereference(tcp_sk(sk)->ao_info);
@@ -948,8 +949,9 @@  tcp_inbound_ao_hash(struct sock *sk, const struct sk_buff *skb,
 		disn = 0;
 	}
 
+	state = READ_ONCE(sk->sk_state);
 	/* Fast-path */
-	if (likely((1 << sk->sk_state) & TCP_AO_ESTABLISHED)) {
+	if (likely((1 << state) & TCP_AO_ESTABLISHED)) {
 		enum skb_drop_reason err;
 		struct tcp_ao_key *current_key;
 
@@ -988,6 +990,9 @@  tcp_inbound_ao_hash(struct sock *sk, const struct sk_buff *skb,
 		return SKB_NOT_DROPPED_YET;
 	}
 
+	if (unlikely(state == TCP_CLOSE))
+		return SKB_DROP_REASON_TCP_CLOSE;
+
 	/* Lookup key based on peer address and keyid.
 	 * current_key and rnext_key must not be used on tcp listen
 	 * sockets as otherwise:
@@ -1001,7 +1006,7 @@  tcp_inbound_ao_hash(struct sock *sk, const struct sk_buff *skb,
 	if (th->syn && !th->ack)
 		goto verify_hash;
 
-	if ((1 << sk->sk_state) & (TCPF_LISTEN | TCPF_NEW_SYN_RECV)) {
+	if ((1 << state) & (TCPF_LISTEN | TCPF_NEW_SYN_RECV)) {
 		/* Make the initial syn the likely case here */
 		if (unlikely(req)) {
 			sne = tcp_ao_compute_sne(0, tcp_rsk(req)->rcv_isn,
@@ -1018,14 +1023,14 @@  tcp_inbound_ao_hash(struct sock *sk, const struct sk_buff *skb,
 			/* no way to figure out initial sisn/disn - drop */
 			return SKB_DROP_REASON_TCP_FLAGS;
 		}
-	} else if ((1 << sk->sk_state) & (TCPF_SYN_SENT | TCPF_SYN_RECV)) {
+	} else if ((1 << state) & (TCPF_SYN_SENT | TCPF_SYN_RECV)) {
 		disn = info->lisn;
 		if (th->syn || th->rst)
 			sisn = th->seq;
 		else
 			sisn = info->risn;
 	} else {
-		WARN_ONCE(1, "TCP-AO: Unexpected sk_state %d", sk->sk_state);
+		WARN_ONCE(1, "TCP-AO: Unexpected sk_state %d", state);
 		return SKB_DROP_REASON_TCP_AOFAILURE;
 	}
 verify_hash: