Skip to contentSkip to conversation

fix null-deref in ipv4_link_failure

PATCH In review Kyle Zeng 4 participants 4 messages Sep 8, 2023 – Sep 11, 2023

Subsystems (3)
lore
Share & tools
lore

Kyle Zeng

author
#0 lore raw

PATCH 1/1 fix null-deref in ipv4_link_failure

Currently, we assume the skb is associated with a device before calling __ip_options_compile, which is not always the case if it is re-routed by ipvs.
When skb->dev is NULL, dev_net(skb->dev) will become null-dereference. This patch adds a check for the edge case and switch to use the net_device from the rtable when skb->dev is NULL.

---
 net/ipv4/route.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
net/ipv4/route.c +3 -1
index d8c99bdc617..1be34e5eea1 100644
@@ -1214,6 +1214,7 @@ EXPORT_INDIRECT_CALLABLE(ipv4_dst_check);
 static void ipv4_send_dest_unreach(struct sk_buff *skb)
 {
 	struct ip_options opt;
+	struct net_device *dev;
 	int res;
 
 	/* Recompile ip options since IPCB may not be valid anymore.
@@ -1230,7 +1231,8 @@ static void ipv4_send_dest_unreach(struct sk_buff *skb)
 		opt.optlen = ip_hdr(skb)->ihl * 4 - sizeof(struct iphdr);
 
 		rcu_read_lock();
-		res = __ip_options_compile(dev_net(skb->dev), &opt, skb, NULL);
+		dev = skb->dev ? skb->dev : skb_rtable(skb)->dst.dev;
+		res = __ip_options_compile(dev_net(dev), &opt, skb, NULL);
 		rcu_read_unlock();
 
 		if (res)
Signature · 1 line
2.34.1
#1 lore raw
Quoted 20 lines · Kyle Zeng

Currently, we assume the skb is associated with a device before calling __ip_options_compile, which is not always the case if it is re-routed by ipvs.
When skb->dev is NULL, dev_net(skb->dev) will become null-dereference. This patch adds a check for the edge case and switch to use the net_device from the rtable when skb->dev is NULL.

---
  net/ipv4/route.c | 4 +++-
  1 file changed, 3 insertions(+), 1 deletion(-)
net/ipv4/route.c +0 -0
index d8c99bdc617..1be34e5eea1 100644

net/ipv4/route.c EXPORT_INDIRECT_CALLABLE(ipv4_dst_check); · line 1217

Quoted 4 lines
net/ipv4/route.c +1 -0
@@ -1214,6 +1214,7 @@ EXPORT_INDIRECT_CALLABLE(ipv4_dst_check);
  static void ipv4_send_dest_unreach(struct sk_buff *skb)
  {
  	struct ip_options opt;
+	struct net_device *dev;

Please, use reverse x-mas tree order for local variables

Quoted 13 lines
net/ipv4/route.c +2 -1
  	int res;
 
  	/* Recompile ip options since IPCB may not be valid anymore.
@@ -1230,7 +1231,8 @@ static void ipv4_send_dest_unreach(struct sk_buff *skb)
  		opt.optlen = ip_hdr(skb)->ihl * 4 - sizeof(struct iphdr);
 
  		rcu_read_lock();
-		res = __ip_options_compile(dev_net(skb->dev), &opt, skb, NULL);
+		dev = skb->dev ? skb->dev : skb_rtable(skb)->dst.dev;
+		res = __ip_options_compile(dev_net(dev), &opt, skb, NULL);
  		rcu_read_unlock();
 
  		if (res)
maintainer Reviewed-by on v1 1/1
#2 lore raw
In reply to Kyle Zeng
Quoted 7 lines · Kyle Zeng

Currently, we assume the skb is associated with a device before calling __ip_options_compile, which is not always the case if it is re-routed by ipvs.
When skb->dev is NULL, dev_net(skb->dev) will become null-dereference. This patch adds a check for the edge case and switch to use the net_device from the rtable when skb->dev is NULL.

Quoted 8 lines
---
 net/ipv4/route.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
maintainer
#3 lore raw
In reply to Kyle Zeng
Quoted 28 lines · Kyle Zeng

Currently, we assume the skb is associated with a device before calling __ip_options_compile, which is not always the case if it is re-routed by ipvs.
When skb->dev is NULL, dev_net(skb->dev) will become null-dereference. This patch adds a check for the edge case and switch to use the net_device from the rtable when skb->dev is NULL.

---
 net/ipv4/route.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
net/ipv4/route.c +1 -0
index d8c99bdc617..1be34e5eea1 100644
@@ -1214,6 +1214,7 @@ EXPORT_INDIRECT_CALLABLE(ipv4_dst_check);
 static void ipv4_send_dest_unreach(struct sk_buff *skb)
 {
 	struct ip_options opt;
+	struct net_device *dev;
 	int res;
 
 	/* Recompile ip options since IPCB may not be valid anymore.

net/ipv4/route.c static void ipv4_send_dest_unreach(struct sk_buff *skb) · line 1238

Quoted 9 lines
net/ipv4/route.c +2 -1
@@ -1230,7 +1231,8 @@ static void ipv4_send_dest_unreach(struct sk_buff *skb)
 		opt.optlen = ip_hdr(skb)->ihl * 4 - sizeof(struct iphdr);
 
 		rcu_read_lock();
-		res = __ip_options_compile(dev_net(skb->dev), &opt, skb, NULL);
+		dev = skb->dev ? skb->dev : skb_rtable(skb)->dst.dev;
+		res = __ip_options_compile(dev_net(dev), &opt, skb, NULL);
 		rcu_read_unlock();
 
 		if (res)

The patch LGTM, but if you are going to re-post to address Vadim's feedback, please additionally drop my Suggested-by. I only reviewed a previous revision.

Thanks!

Paolo