Conntrack TW->SS in Reply direction

[Marcelo Ricardo Leitner]

Hi,

Any reason on why we don't allow TW->SS state change in reply direction? It is
currently marked as (since ever it seems) in a comment:
* sTW -> sIV Reopened connection, but server may not do it.

That is true but just if the server role is never swapped between hosts and I
don't see it violating any spec by allowing it..

Thing is, there is this application that needs a connection as a callback on a
specified port range. This application is used between 2 servers, using
conntrack and only allowing in new connections and its (related) packets.
There will be a moment that the port that once was server for the callback,
will now originate a callback to the other server.

As Linux stack uses TIME_WAIT timeout as 1min and conntrack as 2mins, there
will be a moment on which the port selection allows using a port that, for
conntrack, is still not allowed...

Reducing conntrack's TIME_WAIT timeout is more complicated because it may
restrain the compatibility with other implementations, but allowing TW->SS
imposes no harm?

Thanks,
Marcelo
--
To unsubscribe from this list: send the line "unsubscribe netfilter-devel" in
the body of a message to ***@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html

[Marcelo Ricardo Leitner]

When a port that was used to listen for inbound connections gets closed
and reused for outgoing connections (like rsh ends up doing for stderr
flow), current we may reject the SYN/ACK packet for the new connection
because tcp_conntracks states forbirds a port to become a client while
there is still a TIME_WAIT entry in there for it.

As TCP may expire the TIME_WAIT socket in 60s and conntrack's timeout
for it is 120s, there is a ~60s window that the application can end up
opening a port that conntrack will end up blocking.

This patch fixes this by simply allowing such state transition: if we
see a SYN, in TIME_WAIT state, on REPLY direction, move it to sSS. Note
that the rest of the code already handles this situation, more
specificly in tcp_packet(), first switch clause.

Signed-off-by: Marcelo Ricardo Leitner <***@redhat.com>
---
net/netfilter/nf_conntrack_proto_tcp.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/net/netfilter/nf_conntrack_proto_tcp.c b/net/netfilter/nf_conntrack_proto_tcp.c
index 44d1ea32570a07338dc39f34624bd823b6f76916..d87b6423ffb21e0f8f9b6ef25ef51c1cb5f54ad6 100644
--- a/net/netfilter/nf_conntrack_proto_tcp.c
+++ b/net/netfilter/nf_conntrack_proto_tcp.c
@@ -213,7 +213,7 @@ static const u8 tcp_conntracks[2][6][TCP_CONNTRACK_MAX] = {
{
/* REPLY */
/* sNO, sSS, sSR, sES, sFW, sCW, sLA, sTW, sCL, sS2 */
-/*syn*/ { sIV, sS2, sIV, sIV, sIV, sIV, sIV, sIV, sIV, sS2 },
+/*syn*/ { sIV, sS2, sIV, sIV, sIV, sIV, sIV, sSS, sIV, sS2 },
/*
* sNO -> sIV Never reached.
* sSS -> sS2 Simultaneous open
@@ -223,7 +223,7 @@ static const u8 tcp_conntracks[2][6][TCP_CONNTRACK_MAX] = {
* sFW -> sIV
* sCW -> sIV
* sLA -> sIV
- * sTW -> sIV Reopened connection, but server may not do it.
+ * sTW -> sSS Reopened connection, but server may have switched role
* sCL -> sIV
*/
/* sNO, sSS, sSR, sES, sFW, sCW, sLA, sTW, sCL, sS2 */

[Jozsef Kadlecsik]

Post by Marcelo Ricardo Leitner
When a port that was used to listen for inbound connections gets closed
and reused for outgoing connections (like rsh ends up doing for stderr
flow), current we may reject the SYN/ACK packet for the new connection
because tcp_conntracks states forbirds a port to become a client while
there is still a TIME_WAIT entry in there for it.
As TCP may expire the TIME_WAIT socket in 60s and conntrack's timeout
for it is 120s, there is a ~60s window that the application can end up
opening a port that conntrack will end up blocking.
This patch fixes this by simply allowing such state transition: if we
see a SYN, in TIME_WAIT state, on REPLY direction, move it to sSS. Note
that the rest of the code already handles this situation, more
specificly in tcp_packet(), first switch clause.

In those code branch if there was a valid FIN in either direction, we
destroy the old connection and a new will be created. That way the rules
about NEW connections will be applied, so the policies are not bypassed.
Otherwise we just ignore the SYN packet, so if it's invalid, we'll catch
the RST from the other side and destroy the conntrack entry. The event
flow looks OK to me.
Acked-by: Jozsef Kadlecsik <***@blackhole.kfki.hu>

Best regards,
Jozsef

Post by Marcelo Ricardo Leitner
---
net/netfilter/nf_conntrack_proto_tcp.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/net/netfilter/nf_conntrack_proto_tcp.c b/net/netfilter/nf_conntrack_proto_tcp.c
index 44d1ea32570a07338dc39f34624bd823b6f76916..d87b6423ffb21e0f8f9b6ef25ef51c1cb5f54ad6 100644
--- a/net/netfilter/nf_conntrack_proto_tcp.c
+++ b/net/netfilter/nf_conntrack_proto_tcp.c
@@ -213,7 +213,7 @@ static const u8 tcp_conntracks[2][6][TCP_CONNTRACK_MAX] = {
{
/* REPLY */
/* sNO, sSS, sSR, sES, sFW, sCW, sLA, sTW, sCL, sS2 */
-/*syn*/ { sIV, sS2, sIV, sIV, sIV, sIV, sIV, sIV, sIV, sS2 },
+/*syn*/ { sIV, sS2, sIV, sIV, sIV, sIV, sIV, sSS, sIV, sS2 },
/*
* sNO -> sIV Never reached.
* sSS -> sS2 Simultaneous open
@@ -223,7 +223,7 @@ static const u8 tcp_conntracks[2][6][TCP_CONNTRACK_MAX] = {
* sFW -> sIV
* sCW -> sIV
* sLA -> sIV
- * sTW -> sIV Reopened connection, but server may not do it.
+ * sTW -> sSS Reopened connection, but server may have switched role
* sCL -> sIV
*/
/* sNO, sSS, sSR, sES, sFW, sCW, sLA, sTW, sCL, sS2 */
--
1.9.3
--
To unsubscribe from this list: send the line "unsubscribe netfilter-devel" in
More majordomo info at http://vger.kernel.org/majordomo-info.html

-
E-mail : ***@blackhole.kfki.hu, ***@wigner.mta.hu
PGP key : http://www.kfki.hu/~kadlec/pgp_public_key.txt
Address : Wigner Research Centre for Physics, Hungarian Academy of Sciences
H-1525 Budapest 114, POB. 49, Hungary

[Pablo Neira Ayuso]

Post by Jozsef Kadlecsik

Post by Marcelo Ricardo Leitner
When a port that was used to listen for inbound connections gets closed
and reused for outgoing connections (like rsh ends up doing for stderr
flow), current we may reject the SYN/ACK packet for the new connection
because tcp_conntracks states forbirds a port to become a client while
there is still a TIME_WAIT entry in there for it.
As TCP may expire the TIME_WAIT socket in 60s and conntrack's timeout
for it is 120s, there is a ~60s window that the application can end up
opening a port that conntrack will end up blocking.
This patch fixes this by simply allowing such state transition: if we
see a SYN, in TIME_WAIT state, on REPLY direction, move it to sSS. Note
that the rest of the code already handles this situation, more
specificly in tcp_packet(), first switch clause.

In those code branch if there was a valid FIN in either direction, we
destroy the old connection and a new will be created. That way the rules
about NEW connections will be applied, so the policies are not bypassed.
Otherwise we just ignore the SYN packet, so if it's invalid, we'll catch
the RST from the other side and destroy the conntrack entry. The event
flow looks OK to me.

Applied, thanks.