[PATCH net-next] tcp: remove redundant checks

[PATCH net-next] tcp: remove redundant checks

[Eric Dumazet]

From: Eric Dumazet  <edumazet@google.com>

tcp_v4_rcv() checks the following before calling tcp_v4_do_rcv():

if (th->doff < sizeof(struct tcphdr) / 4)
    goto bad_packet;
if (!pskb_may_pull(skb, th->doff * 4))
    goto discard_it;

So following check in tcp_v4_do_rcv() is redundant
and "goto csum_err;" is wrong anyway.

if (skb->len < tcp_hdrlen(skb) || ...)
	goto csum_err;

A second check can be removed after no_tcp_socket label for same reason.

Same tests can be removed in tcp_v6_do_rcv()

Note : short tcp frames are not properly accounted in tcpInErrs MIB,
because pskb_may_pull() failure simply drops incoming skb, we might
fix this in a separate patch.

Signed-off-by: Eric Dumazet  <edumazet@google.com>
---
 net/ipv4/tcp_ipv4.c |    4 ++--
 net/ipv6/tcp_ipv6.c |    4 ++--
 2 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index feb875769b8d57dcdb85e12b782f3f5e0fb6193a..3130c42240c86545f1037648635a9e7de0876095 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -1400,7 +1400,7 @@ int tcp_v4_do_rcv(struct sock *sk, struct sk_buff *skb)
 		return 0;
 	}
 
-	if (skb->len < tcp_hdrlen(skb) || tcp_checksum_complete(skb))
+	if (tcp_checksum_complete(skb))
 		goto csum_err;
 
 	if (sk->sk_state == TCP_LISTEN) {
@@ -1647,7 +1647,7 @@ no_tcp_socket:
 	if (!xfrm4_policy_check(NULL, XFRM_POLICY_IN, skb))
 		goto discard_it;
 
-	if (skb->len < (th->doff << 2) || tcp_checksum_complete(skb)) {
+	if (tcp_checksum_complete(skb)) {
 csum_error:
 		TCP_INC_STATS_BH(net, TCP_MIB_CSUMERRORS);
 bad_packet:
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index 7be3d858cbf0a4f5b7f4ecce8788aa8e4136d6fa..b72e1ffc6bb40f486764874e6a167b72d899605b 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -1250,7 +1250,7 @@ static int tcp_v6_do_rcv(struct sock *sk, struct sk_buff *skb)
 		return 0;
 	}
 
-	if (skb->len < tcp_hdrlen(skb) || tcp_checksum_complete(skb))
+	if (tcp_checksum_complete(skb))
 		goto csum_err;
 
 	if (sk->sk_state == TCP_LISTEN) {
@@ -1442,7 +1442,7 @@ no_tcp_socket:
 
 	tcp_v6_fill_cb(skb, hdr, th);
 
-	if (skb->len < (th->doff<<2) || tcp_checksum_complete(skb)) {
+	if (tcp_checksum_complete(skb)) {
 csum_error:
 		TCP_INC_STATS_BH(net, TCP_MIB_CSUMERRORS);
 bad_packet:

Re: [PATCH net-next] tcp: remove redundant checks

[David Miller]

From: Eric Dumazet <eric.dumazet@gmail.com>
Date: Wed, 03 Jun 2015 23:49:21 -0700

> From: Eric Dumazet  <edumazet@google.com>
> 
> tcp_v4_rcv() checks the following before calling tcp_v4_do_rcv():
> 
> if (th->doff < sizeof(struct tcphdr) / 4)
>     goto bad_packet;
> if (!pskb_may_pull(skb, th->doff * 4))
>     goto discard_it;
> 
> So following check in tcp_v4_do_rcv() is redundant
> and "goto csum_err;" is wrong anyway.
> 
> if (skb->len < tcp_hdrlen(skb) || ...)
> 	goto csum_err;
> 
> A second check can be removed after no_tcp_socket label for same reason.
> 
> Same tests can be removed in tcp_v6_do_rcv()
> 
> Note : short tcp frames are not properly accounted in tcpInErrs MIB,
> because pskb_may_pull() failure simply drops incoming skb, we might
> fix this in a separate patch.
> 
> Signed-off-by: Eric Dumazet  <edumazet@google.com>

Looks sane, applied, thanks Eric!

[PATCH net-next] tcp: remove redundant checks II

[Eric Dumazet]

From: Eric Dumazet <edumazet@google.com>

For same reasons than in commit 12e25e1041d0 ("tcp: remove redundant
checks"), we can remove redundant checks done for timewait sockets.

Signed-off-by: Eric Dumazet <edumazet@google.com>
---
 net/ipv4/tcp_ipv4.c |    4 ----
 net/ipv6/tcp_ipv6.c |    4 ----
 2 files changed, 8 deletions(-)

diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 3130c42240c8..d7d4c2b79cf2 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -1671,10 +1671,6 @@ do_time_wait:
 		goto discard_it;
 	}
 
-	if (skb->len < (th->doff << 2)) {
-		inet_twsk_put(inet_twsk(sk));
-		goto bad_packet;
-	}
 	if (tcp_checksum_complete(skb)) {
 		inet_twsk_put(inet_twsk(sk));
 		goto csum_error;
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index b72e1ffc6bb4..45a7176ed460 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -1467,10 +1467,6 @@ do_time_wait:
 
 	tcp_v6_fill_cb(skb, hdr, th);
 
-	if (skb->len < (th->doff<<2)) {
-		inet_twsk_put(inet_twsk(sk));
-		goto bad_packet;
-	}
 	if (tcp_checksum_complete(skb)) {
 		inet_twsk_put(inet_twsk(sk));
 		goto csum_error;

Re: [PATCH net-next] tcp: remove redundant checks II

[David Miller]

From: Eric Dumazet <eric.dumazet@gmail.com>
Date: Thu, 04 Jun 2015 08:01:12 -0700

> From: Eric Dumazet <edumazet@google.com>
> 
> For same reasons than in commit 12e25e1041d0 ("tcp: remove redundant
> checks"), we can remove redundant checks done for timewait sockets.
> 
> Signed-off-by: Eric Dumazet <edumazet@google.com>

Applied, thanks Eric.