Skip to contentSkip to conversation

ipv6: ensure ip6_dst_mtu_forward() returns at least IPV6_MIN_MTU

PATCH In review Maciej Żenczykowski 2 participants 2 messages Jun 7, 2021 – Jun 8, 2021

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 ipv6: ensure ip6_dst_mtu_forward() returns at least IPV6_MIN_MTU

From: Maciej Żenczykowski <[email protected]>

ip6_dst_mtu_forward() has just two call sites, one of which already enforces the minimum mtu (which we can now remove), but it does materially affect the other (presumably buggy) call site in:
net/netfilter/nf_flow_table_core.c flow_offload_fill_route()

Cc David S. Miller <[email protected]>
Cc Hideaki YOSHIFUJI <[email protected]>
Cc David Ahern <[email protected]>
Cc Jakub Kicinski <[email protected]>
---
 include/net/ip6_route.h | 4 ++--
 net/ipv6/ip6_output.c   | 2 --
 2 files changed, 2 insertions(+), 4 deletions(-)
include/net/ip6_route.h +2 -2
index 9dd3c75a4d62..e01364c0821d 100644
@@ -315,14 +315,14 @@ static inline unsigned int ip6_dst_mtu_forward(const struct dst_entry *dst)
 	unsigned int mtu = dst_metric_raw(dst, RTAX_MTU);
 
 	if (mtu)
-		return mtu;
+		return max_t(unsigned int, mtu, IPV6_MIN_MTU);
 
 	rcu_read_lock();
 	idev = __in6_dev_get(dst->dev);
 	mtu = idev ? idev->cnf.mtu6 : IPV6_MIN_MTU;
 	rcu_read_unlock();
 
-	return mtu;
+	return max_t(unsigned int, mtu, IPV6_MIN_MTU);
 }
 
 u32 ip6_mtu_from_fib6(const struct fib6_result *res,
net/ipv6/ip6_output.c +0 -2
index ff4f9ebcf7f6..8e56378713cd 100644
@@ -578,8 +578,6 @@ int ip6_forward(struct sk_buff *skb)
 	}
 
 	mtu = ip6_dst_mtu_forward(dst);
-	if (mtu < IPV6_MIN_MTU)
-		mtu = IPV6_MIN_MTU;
 
 	if (ip6_pkt_too_big(skb, mtu)) {
 		/* Again, force OUTPUT device used as source address */
Signature · 1 line
2.32.0.rc1.229.g3e70b5a671-goog
#1 lore raw
Quoted 23 lines · Maciej Żenczykowski

From: Maciej Żenczykowski <[email protected]>

ip6_dst_mtu_forward() has just two call sites, one of which already enforces the minimum mtu (which we can now remove), but it does materially affect the other (presumably buggy) call site in:
net/netfilter/nf_flow_table_core.c flow_offload_fill_route()

Cc David S. Miller <[email protected]>
Cc Hideaki YOSHIFUJI <[email protected]>
Cc David Ahern <[email protected]>
Cc Jakub Kicinski <[email protected]>
---
 include/net/ip6_route.h | 4 ++--
 net/ipv6/ip6_output.c   | 2 --
 2 files changed, 2 insertions(+), 4 deletions(-)
include/net/ip6_route.h +0 -0
index 9dd3c75a4d62..e01364c0821d 100644

include/net/ip6_route.h static inline unsigned int ip6_dst_mtu_forward(const struct dst_entry *dst) · line 315

Quoted 1 line
include/net/ip6_route.h +0 -0
@@ -315,14 +315,14 @@ static inline unsigned int ip6_dst_mtu_forward(const struct dst_entry *dst)
 	unsigned int mtu = dst_metric_raw(dst, RTAX_MTU);

we can prevent a route getting an mtu as a metric that is < IPV6_MIN_MTU.

include/net/ip6_route.h static inline unsigned int ip6_dst_mtu_forward(const struct dst_entry *dst) · line 322

Quoted 8 lines
include/net/ip6_route.h +1 -1
 	if (mtu)
-		return mtu;
+		return max_t(unsigned int, mtu, IPV6_MIN_MTU);
 
 	rcu_read_lock();
 	idev = __in6_dev_get(dst->dev);
 	mtu = idev ? idev->cnf.mtu6 : IPV6_MIN_MTU;

and we can prevent cnf.mtu6 getting set < IPV6_MIN_MTU:

addrconf_notify(), NETDEV_CHANGEMTU case. ipv6_add_dev() already requires dev->mtu to be > IPV6_MIN_MTU ndisc_router_discovery requires it.

so it seems implausible for the idev to get a value below the minimum.

the metric is less clear that it is checked, so it would be the one to make sure is meeting expectations.