[PATCH 3.19.y-ckt 192/210] net: preserve IP control block during GSO segmentation

[Kamal Mostafa]

3.19.8-ckt14 -stable review patch.  If anyone has any objections, please let me know.

---8<------------------------------------------------------------

From: Konstantin Khlebnikov <koct9i@xxxxxxxxx>

[ Upstream commit 9207f9d45b0ad071baa128e846d7e7ed85016df3 ]

Skb_gso_segment() uses skb control block during segmentation.
This patch adds 32-bytes room for previous control block which
will be copied into all resulting segments.

This patch fixes kernel crash during fragmenting forwarded packets.
Fragmentation requires valid IP CB in skb for clearing ip options.
Also patch removes custom save/restore in ovs code, now it's redundant.

Signed-off-by: Konstantin Khlebnikov <koct9i@xxxxxxxxx>
Link: http://lkml.kernel.org/r/CALYGNiP-0MZ-FExV2HutTvE9U-QQtkKSoE--KN=JQE5STYsjAA@xxxxxxxxxxxxxx
Signed-off-by: David S. Miller <davem@xxxxxxxxxxxxx>
Signed-off-by: Kamal Mostafa <kamal@xxxxxxxxxxxxx>
---
 include/linux/skbuff.h     | 3 ++-
 net/core/dev.c             | 5 +++++
 net/ipv4/ip_output.c       | 1 +
 net/openvswitch/datapath.c | 5 +----
 net/xfrm/xfrm_output.c     | 2 ++
 5 files changed, 11 insertions(+), 5 deletions(-)

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index b0e064e..85b1df30 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -3271,7 +3271,8 @@ struct skb_gso_cb {
 	int	encap_level;
 	__u16	csum_start;
 };
-#define SKB_GSO_CB(skb) ((struct skb_gso_cb *)(skb)->cb)
+#define SKB_SGO_CB_OFFSET	32
+#define SKB_GSO_CB(skb) ((struct skb_gso_cb *)((skb)->cb + SKB_SGO_CB_OFFSET))
 
 static inline int skb_tnl_header_len(const struct sk_buff *inner_skb)
 {
diff --git a/net/core/dev.c b/net/core/dev.c
index 9c14e87..a0d1188 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -2422,6 +2422,8 @@ static inline bool skb_needs_check(struct sk_buff *skb, bool tx_path)
  *
  *	It may return NULL if the skb requires no segmentation.  This is
  *	only possible when GSO is used for verifying header integrity.
+ *
+ *	Segmentation preserves SKB_SGO_CB_OFFSET bytes of previous skb cb.
  */
 struct sk_buff *__skb_gso_segment(struct sk_buff *skb,
 				  netdev_features_t features, bool tx_path)
@@ -2436,6 +2438,9 @@ struct sk_buff *__skb_gso_segment(struct sk_buff *skb,
 			return ERR_PTR(err);
 	}
 
+	BUILD_BUG_ON(SKB_SGO_CB_OFFSET +
+		     sizeof(*SKB_GSO_CB(skb)) > sizeof(skb->cb));
+
 	SKB_GSO_CB(skb)->mac_offset = skb_headroom(skb);
 	SKB_GSO_CB(skb)->encap_level = 0;
 
diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c
index b70d9ae..4c8d4d5f 100644
--- a/net/ipv4/ip_output.c
+++ b/net/ipv4/ip_output.c
@@ -230,6 +230,7 @@ static int ip_finish_output_gso(struct sk_buff *skb)
 	 * from host network stack.
 	 */
 	features = netif_skb_features(skb);
+	BUILD_BUG_ON(sizeof(*IPCB(skb)) > SKB_SGO_CB_OFFSET);
 	segs = skb_gso_segment(skb, features & ~NETIF_F_GSO_MASK);
 	if (IS_ERR_OR_NULL(segs)) {
 		kfree_skb(skb);
diff --git a/net/openvswitch/datapath.c b/net/openvswitch/datapath.c
index 34d2643..23cea2a 100644
--- a/net/openvswitch/datapath.c
+++ b/net/openvswitch/datapath.c
@@ -336,12 +336,10 @@ static int queue_gso_packets(struct datapath *dp, struct sk_buff *skb,
 	unsigned short gso_type = skb_shinfo(skb)->gso_type;
 	struct sw_flow_key later_key;
 	struct sk_buff *segs, *nskb;
-	struct ovs_skb_cb ovs_cb;
 	int err;
 
-	ovs_cb = *OVS_CB(skb);
+	BUILD_BUG_ON(sizeof(*OVS_CB(skb)) > SKB_SGO_CB_OFFSET);
 	segs = __skb_gso_segment(skb, NETIF_F_SG, false);
-	*OVS_CB(skb) = ovs_cb;
 	if (IS_ERR(segs))
 		return PTR_ERR(segs);
 	if (segs == NULL)
@@ -359,7 +357,6 @@ static int queue_gso_packets(struct datapath *dp, struct sk_buff *skb,
 	/* Queue all of the segments. */
 	skb = segs;
 	do {
-		*OVS_CB(skb) = ovs_cb;
 		if (gso_type & SKB_GSO_UDP && skb != segs)
 			key = &later_key;
 
diff --git a/net/xfrm/xfrm_output.c b/net/xfrm/xfrm_output.c
index 7c53285..55f18c5 100644
--- a/net/xfrm/xfrm_output.c
+++ b/net/xfrm/xfrm_output.c
@@ -153,6 +153,8 @@ static int xfrm_output_gso(struct sk_buff *skb)
 {
 	struct sk_buff *segs;
 
+	BUILD_BUG_ON(sizeof(*IPCB(skb)) > SKB_SGO_CB_OFFSET);
+	BUILD_BUG_ON(sizeof(*IP6CB(skb)) > SKB_SGO_CB_OFFSET);
 	segs = skb_gso_segment(skb, 0);
 	kfree_skb(skb);
 	if (IS_ERR(segs))
-- 
1.9.1