diff options
| author | Eric Dumazet <edumazet@google.com> | 2015-11-01 15:36:55 -0800 |
|---|---|---|
| committer | David S. Miller <davem@davemloft.net> | 2015-11-02 16:28:49 -0500 |
| commit | 9e17f8a475fca81950fdddc08df428ed66cf441f (patch) | |
| tree | 417a79910266f08806f78be2dd970102afa3356f /net | |
| parent | eca1e006cf6f6f66a1a90c055a8a6d393475c3f9 (diff) | |
| download | op-kernel-dev-9e17f8a475fca81950fdddc08df428ed66cf441f.zip op-kernel-dev-9e17f8a475fca81950fdddc08df428ed66cf441f.tar.gz | |
net: make skb_set_owner_w() more robust
skb_set_owner_w() is called from various places that assume
skb->sk always point to a full blown socket (as it changes
sk->sk_wmem_alloc)
We'd like to attach skb to request sockets, and in the future
to timewait sockets as well. For these kind of pseudo sockets,
we need to take a traditional refcount and use sock_edemux()
as the destructor.
It is now time to un-inline skb_set_owner_w(), being too big.
Fixes: ca6fb0651883 ("tcp: attach SYNACK messages to request sockets instead of listener")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Bisected-by: Haiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'net')
| -rw-r--r-- | net/core/sock.c | 22 | ||||
| -rw-r--r-- | net/ipv4/tcp_output.c | 4 |
2 files changed, 23 insertions, 3 deletions
diff --git a/net/core/sock.c b/net/core/sock.c index 0ef30aa..7529eb9 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -1656,6 +1656,28 @@ void sock_wfree(struct sk_buff *skb) } EXPORT_SYMBOL(sock_wfree); +void skb_set_owner_w(struct sk_buff *skb, struct sock *sk) +{ + skb_orphan(skb); + skb->sk = sk; +#ifdef CONFIG_INET + if (unlikely(!sk_fullsock(sk))) { + skb->destructor = sock_edemux; + sock_hold(sk); + return; + } +#endif + skb->destructor = sock_wfree; + skb_set_hash_from_sk(skb, sk); + /* + * We used to take a refcount on sk, but following operation + * is enough to guarantee sk_free() wont free this sock until + * all in-flight packets are completed + */ + atomic_add(skb->truesize, &sk->sk_wmem_alloc); +} +EXPORT_SYMBOL(skb_set_owner_w); + void skb_orphan_partial(struct sk_buff *skb) { /* TCP stack sets skb->ooo_okay based on sk_wmem_alloc, diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index f4f9793..cb7ca56 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -2963,9 +2963,7 @@ struct sk_buff *tcp_make_synack(const struct sock *sk, struct dst_entry *dst, skb_reserve(skb, MAX_TCP_HEADER); if (attach_req) { - skb->destructor = sock_edemux; - sock_hold(req_to_sk(req)); - skb->sk = req_to_sk(req); + skb_set_owner_w(skb, req_to_sk(req)); } else { /* sk is a const pointer, because we want to express multiple * cpu might call us concurrently. |
