This is based on Willy Tarreau's patch from 2008[1]. The goal is to close a corner-case of TCP that isn't used and poses a small DoS risk. For systems that do not want to take any risk at all, this is a desirable configuration knob.
It is possible for two clients to connect with crossed SYNs without checking sequence numbers. As such, it might be possible to guess a source port number to block a system from making connections to well-known ports and IP addresses (e.g. auto-update checks) without requiring a MiTM position. The feature can now be disabled via sysctl: $ echo 0 > /proc/sys/net/ipv4/tcp_simult_connect $ echo ohai | nc -w 1 -p 50000 localhost 50000 -v -v -v nc: connect to localhost port 50000 (tcp) timed out: Operation now in progress nc: connect to localhost port 50000 (tcp) timed out: Operation now in progress $ echo 1 > /proc/sys/net/ipv4/tcp_simult_connect $ echo ohai | nc -w 1 -p 50000 localhost 50000 -v -v -v Connection to localhost 50000 port [tcp/*] succeeded! ohai [1] http://thread.gmane.org/gmane.linux.network/107971 Cc: Willy Tarreau <w...@1wt.eu> Signed-off-by: Kees Cook <keesc...@chromium.org> --- Documentation/networking/ip-sysctl.txt | 17 +++++++++++++++++ include/net/tcp.h | 1 + net/ipv4/sysctl_net_ipv4.c | 9 +++++++++ net/ipv4/tcp_input.c | 3 ++- 4 files changed, 29 insertions(+), 1 deletion(-) diff --git a/Documentation/networking/ip-sysctl.txt b/Documentation/networking/ip-sysctl.txt index dbca661..2e5bd51 100644 --- a/Documentation/networking/ip-sysctl.txt +++ b/Documentation/networking/ip-sysctl.txt @@ -431,6 +431,23 @@ tcp_rmem - vector of 3 INTEGERs: min, default, max tcp_sack - BOOLEAN Enable select acknowledgments (SACKS). +tcp_simult_connect - BOOLEAN + Enables TCP simultaneous connect feature conforming to RFC793. + Strict implementation of RFC793 (TCP) requires support for a + feature called "simultaneous connect", which allows two clients to + connect to each other without anyone entering a listening state. + While almost never used, and supported by few OSes, Linux supports + this feature. + + However, it introduces a weakness in the protocol which makes it + very easy for an attacker to prevent a client from connecting to + a known server. The attacker only has to guess the source port + to shut down the client connection during its establishment. The + impact is limited, but it may be used to prevent an antivirus + or IPS from fetching updates and not detecting an attack, or to + prevent an SSL gateway or browser from fetching a CRL. + Default: TRUE + tcp_slow_start_after_idle - BOOLEAN If set, provide RFC2861 behavior and time out the congestion window after an idle period. An idle period is defined at diff --git a/include/net/tcp.h b/include/net/tcp.h index aed42c7..ecd55d0 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -292,6 +292,7 @@ extern int sysctl_tcp_thin_dupack; extern int sysctl_tcp_early_retrans; extern int sysctl_tcp_limit_output_bytes; extern int sysctl_tcp_challenge_ack_limit; +extern int sysctl_tcp_simult_connect; extern atomic_long_t tcp_memory_allocated; extern struct percpu_counter tcp_sockets_allocated; diff --git a/net/ipv4/sysctl_net_ipv4.c b/net/ipv4/sysctl_net_ipv4.c index d84400b..01e475f 100644 --- a/net/ipv4/sysctl_net_ipv4.c +++ b/net/ipv4/sysctl_net_ipv4.c @@ -774,6 +774,15 @@ static struct ctl_table ipv4_table[] = { .extra2 = &two, }, { + .procname = "tcp_simult_connect", + .data = &sysctl_tcp_simult_connect, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec_minmax, + .extra1 = &zero, + .extra2 = &two, + }, + { .procname = "udp_mem", .data = &sysctl_udp_mem, .maxlen = sizeof(sysctl_udp_mem), diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 18f97ca..c71f8bb 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -102,6 +102,7 @@ int sysctl_tcp_thin_dupack __read_mostly; int sysctl_tcp_moderate_rcvbuf __read_mostly = 1; int sysctl_tcp_abc __read_mostly; int sysctl_tcp_early_retrans __read_mostly = 2; +int sysctl_tcp_simult_connect __read_mostly = 1; #define FLAG_DATA 0x01 /* Incoming frame contained data. */ #define FLAG_WIN_UPDATE 0x02 /* Incoming ACK was a window update. */ @@ -5846,7 +5847,7 @@ discard: tcp_paws_reject(&tp->rx_opt, 0)) goto discard_and_undo; - if (th->syn) { + if (th->syn && sysctl_tcp_simult_connect) { /* We see SYN without ACK. It is attempt of * simultaneous connect with crossed SYNs. * Particularly, it can be connect to self. -- 1.7.9.5 -- Kees Cook Chrome OS Security -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
