ipvs: fix oops in backup for fwmark conn templates
Fixes bug http://bugzilla.kernel.org/show_bug.cgi?id=10556
where conn templates with protocol=IPPROTO_IP can oops backup box.
Result from ip_vs_proto_get() should be checked because
protocol value can be invalid or unsupported in backup. But
for valid message we should not fail for templates which use
IPPROTO_IP. Also, add checks to validate message limits and
connection state. Show state NONE for templates using IPPROTO_IP.
Signed-off-by: Julian Anastasov <ja@ssi.bg>
Signed-off-by: David S. Miller <davem@davemloft.net>
diff --git a/net/ipv4/ipvs/ip_vs_proto.c b/net/ipv4/ipvs/ip_vs_proto.c
index dde28a2..4b1c16c 100644
--- a/net/ipv4/ipvs/ip_vs_proto.c
+++ b/net/ipv4/ipvs/ip_vs_proto.c
@@ -148,7 +148,7 @@
struct ip_vs_protocol *pp = ip_vs_proto_get(proto);
if (pp == NULL || pp->state_name == NULL)
- return "ERR!";
+ return (IPPROTO_IP == proto) ? "NONE" : "ERR!";
return pp->state_name(state);
}
diff --git a/net/ipv4/ipvs/ip_vs_proto_ah.c b/net/ipv4/ipvs/ip_vs_proto_ah.c
index a842676..4bf835e 100644
--- a/net/ipv4/ipvs/ip_vs_proto_ah.c
+++ b/net/ipv4/ipvs/ip_vs_proto_ah.c
@@ -160,6 +160,7 @@
struct ip_vs_protocol ip_vs_protocol_ah = {
.name = "AH",
.protocol = IPPROTO_AH,
+ .num_states = 1,
.dont_defrag = 1,
.init = ah_init,
.exit = ah_exit,
diff --git a/net/ipv4/ipvs/ip_vs_proto_esp.c b/net/ipv4/ipvs/ip_vs_proto_esp.c
index aef0d3e..db6a6b7 100644
--- a/net/ipv4/ipvs/ip_vs_proto_esp.c
+++ b/net/ipv4/ipvs/ip_vs_proto_esp.c
@@ -159,6 +159,7 @@
struct ip_vs_protocol ip_vs_protocol_esp = {
.name = "ESP",
.protocol = IPPROTO_ESP,
+ .num_states = 1,
.dont_defrag = 1,
.init = esp_init,
.exit = esp_exit,
diff --git a/net/ipv4/ipvs/ip_vs_proto_tcp.c b/net/ipv4/ipvs/ip_vs_proto_tcp.c
index 620e40f..b83dc14 100644
--- a/net/ipv4/ipvs/ip_vs_proto_tcp.c
+++ b/net/ipv4/ipvs/ip_vs_proto_tcp.c
@@ -594,6 +594,7 @@
struct ip_vs_protocol ip_vs_protocol_tcp = {
.name = "TCP",
.protocol = IPPROTO_TCP,
+ .num_states = IP_VS_TCP_S_LAST,
.dont_defrag = 0,
.appcnt = ATOMIC_INIT(0),
.init = ip_vs_tcp_init,
diff --git a/net/ipv4/ipvs/ip_vs_proto_udp.c b/net/ipv4/ipvs/ip_vs_proto_udp.c
index 1caa290..75771cb 100644
--- a/net/ipv4/ipvs/ip_vs_proto_udp.c
+++ b/net/ipv4/ipvs/ip_vs_proto_udp.c
@@ -409,6 +409,7 @@
struct ip_vs_protocol ip_vs_protocol_udp = {
.name = "UDP",
.protocol = IPPROTO_UDP,
+ .num_states = IP_VS_UDP_S_LAST,
.dont_defrag = 0,
.init = udp_init,
.exit = udp_exit,
diff --git a/net/ipv4/ipvs/ip_vs_sync.c b/net/ipv4/ipvs/ip_vs_sync.c
index 69c5666..eff54ef 100644
--- a/net/ipv4/ipvs/ip_vs_sync.c
+++ b/net/ipv4/ipvs/ip_vs_sync.c
@@ -288,11 +288,16 @@
char *p;
int i;
+ if (buflen < sizeof(struct ip_vs_sync_mesg)) {
+ IP_VS_ERR_RL("sync message header too short\n");
+ return;
+ }
+
/* Convert size back to host byte order */
m->size = ntohs(m->size);
if (buflen != m->size) {
- IP_VS_ERR("bogus message\n");
+ IP_VS_ERR_RL("bogus sync message size\n");
return;
}
@@ -307,9 +312,48 @@
for (i=0; i<m->nr_conns; i++) {
unsigned flags, state;
- s = (struct ip_vs_sync_conn *)p;
+ if (p + SIMPLE_CONN_SIZE > buffer+buflen) {
+ IP_VS_ERR_RL("bogus conn in sync message\n");
+ return;
+ }
+ s = (struct ip_vs_sync_conn *) p;
flags = ntohs(s->flags) | IP_VS_CONN_F_SYNC;
+ flags &= ~IP_VS_CONN_F_HASHED;
+ if (flags & IP_VS_CONN_F_SEQ_MASK) {
+ opt = (struct ip_vs_sync_conn_options *)&s[1];
+ p += FULL_CONN_SIZE;
+ if (p > buffer+buflen) {
+ IP_VS_ERR_RL("bogus conn options in sync message\n");
+ return;
+ }
+ } else {
+ opt = NULL;
+ p += SIMPLE_CONN_SIZE;
+ }
+
state = ntohs(s->state);
+ if (!(flags & IP_VS_CONN_F_TEMPLATE)) {
+ pp = ip_vs_proto_get(s->protocol);
+ if (!pp) {
+ IP_VS_ERR_RL("Unsupported protocol %u in sync msg\n",
+ s->protocol);
+ continue;
+ }
+ if (state >= pp->num_states) {
+ IP_VS_DBG(2, "Invalid %s state %u in sync msg\n",
+ pp->name, state);
+ continue;
+ }
+ } else {
+ /* protocol in templates is not used for state/timeout */
+ pp = NULL;
+ if (state > 0) {
+ IP_VS_DBG(2, "Invalid template state %u in sync msg\n",
+ state);
+ state = 0;
+ }
+ }
+
if (!(flags & IP_VS_CONN_F_TEMPLATE))
cp = ip_vs_conn_in_get(s->protocol,
s->caddr, s->cport,
@@ -345,14 +389,9 @@
IP_VS_ERR("ip_vs_conn_new failed\n");
return;
}
- cp->state = state;
} else if (!cp->dest) {
dest = ip_vs_try_bind_dest(cp);
- if (!dest) {
- /* it is an unbound entry created by
- * synchronization */
- cp->flags = flags | IP_VS_CONN_F_HASHED;
- } else
+ if (dest)
atomic_dec(&dest->refcnt);
} else if ((cp->dest) && (cp->protocol == IPPROTO_TCP) &&
(cp->state != state)) {
@@ -371,23 +410,22 @@
}
}
- if (flags & IP_VS_CONN_F_SEQ_MASK) {
- opt = (struct ip_vs_sync_conn_options *)&s[1];
+ if (opt)
memcpy(&cp->in_seq, opt, sizeof(*opt));
- p += FULL_CONN_SIZE;
- } else
- p += SIMPLE_CONN_SIZE;
-
atomic_set(&cp->in_pkts, sysctl_ip_vs_sync_threshold[0]);
cp->state = state;
- pp = ip_vs_proto_get(s->protocol);
- cp->timeout = pp->timeout_table[cp->state];
+ cp->old_state = cp->state;
+ /*
+ * We can not recover the right timeout for templates
+ * in all cases, we can not find the right fwmark
+ * virtual service. If needed, we can do it for
+ * non-fwmark persistent services.
+ */
+ if (!(flags & IP_VS_CONN_F_TEMPLATE) && pp->timeout_table)
+ cp->timeout = pp->timeout_table[state];
+ else
+ cp->timeout = (3*60*HZ);
ip_vs_conn_put(cp);
-
- if (p > buffer+buflen) {
- IP_VS_ERR("bogus message\n");
- return;
- }
}
}