[PATCH 5/7] ipv4: Correct comparisons and calculations using skb->tail and skb-transport_header
flat view
STALE4887d
SUPERSEDED
From: Simon Horman <horms@verge.net.au>
Date: 2013-05-29 05:27:10
Subsystem:
networking [general], networking [ipv4/ipv6], networking [tcp], the rest · Maintainers:
"David S. Miller", Eric Dumazet, Jakub Kicinski, Paolo Abeni, David Ahern, Ido Schimmel, Neal Cardwell, Linus Torvalds
Revision v1 of 2 in this series.
Revisions (2)
- v1 current
- v2 [diff vs current]
This corrects an regression introduced by "net: Use 16bits for *_headers fields of struct skbuff" when NET_SKBUFF_DATA_USES_OFFSET is not set. In that case skb->tail will be a pointer whereas skb->transport_header will be an offset from head. This is corrected by using wrappers that ensure that comparisons and calculations are always made using pointers. Signed-off-by: Simon Horman <horms@verge.net.au> --- net/ipv4/icmp.c | 3 ++- net/ipv4/igmp.c | 2 +- net/ipv4/tcp.c | 3 ++- 3 files changed, 5 insertions(+), 3 deletions(-)
diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c
index 562efd9..5d0d379 100644
--- a/net/ipv4/icmp.c
+++ b/net/ipv4/icmp.c@@ -503,7 +503,8 @@ void icmp_send(struct sk_buff *skb_in, int type, int code, __be32 info) iph = ip_hdr(skb_in); if ((u8 *)iph < skb_in->head || - (skb_in->network_header + sizeof(*iph)) > skb_in->tail) + (skb_network_header(skb_in) + sizeof(*iph)) > + skb_tail_pointer(skb_in)) goto out; /*
diff --git a/net/ipv4/igmp.c b/net/ipv4/igmp.c
index d8c2327..450f625 100644
--- a/net/ipv4/igmp.c
+++ b/net/ipv4/igmp.c@@ -363,7 +363,7 @@ static struct sk_buff *igmpv3_newpack(struct net_device *dev, int size) static int igmpv3_sendpack(struct sk_buff *skb) { struct igmphdr *pig = igmp_hdr(skb); - const int igmplen = skb->tail - skb->transport_header; + const int igmplen = skb_tail_pointer(skb) - skb_transport_header(skb); pig->csum = ip_compute_csum(igmp_hdr(skb), igmplen);
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index ba4186e..1f58594 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c@@ -2989,7 +2989,8 @@ struct sk_buff *tcp_tso_segment(struct sk_buff *skb, swap(gso_skb->truesize, skb->truesize); } - delta = htonl(oldlen + (skb->tail - skb->transport_header) + + delta = htonl(oldlen + (skb_tail_pointer(skb) - + skb_transport_header(skb)) + skb->data_len); th->check = ~csum_fold((__force __wsum)((__force u32)th->check + (__force u32)delta));
--
1.7.10.4