netfilter: nft_exthdr: call ipv6_find_hdr() with explicitly initialized offset

In nft's nft_exthdr_eval() routine we process IPv6 extension header
through invoking ipv6_find_hdr(), but we call it with an uninitialized
offset variable that contains some stack value. In ipv6_find_hdr()
we then test if the value of offset != 0 and call skb_header_pointer()
on that offset in order to map struct ipv6hdr into it. Fix it up by
initializing offset to 0 as it was probably intended to be.

Fixes: 96518518cc41 ("netfilter: add nftables")
Signed-off-by: Daniel Borkmann <dborkman@redhat.com>
Cc: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

diff --git a/net/netfilter/nft_exthdr.c b/net/netfilter/nft_exthdr.c
index 8e0bb75e7c51e61092fb904dc1d93befe1e7320d..55c939f5371fabf35ad996efb51ea52a29630703 100644 (file)
--- a/net/netfilter/nft_exthdr.c
+++ b/net/netfilter/nft_exthdr.c
@@ -31,7 +31,7 @@ static void nft_exthdr_eval(const struct nft_expr *expr,
 {
        struct nft_exthdr *priv = nft_expr_priv(expr);
        struct nft_data *dest = &data[priv->dreg];
-       unsigned int offset;
+       unsigned int offset = 0;
        int err;
 
        err = ipv6_find_hdr(pkt->skb, &offset, priv->type, NULL, NULL);