diff mbox series

[bpf] bpf: Fix use of sk->sk_reuseport from sk_assign

Message ID 20200408033540.10339-1-joe@wand.net.nz
State Accepted
Delegated to: BPF Maintainers
Headers show
Series [bpf] bpf: Fix use of sk->sk_reuseport from sk_assign | expand

Commit Message

Joe Stringer April 8, 2020, 3:35 a.m. UTC
In testing, we found that for request sockets the sk->sk_reuseport field
may yet be uninitialized, which caused bpf_sk_assign() to randomly
succeed or return -ESOCKTNOSUPPORT when handling the forward ACK in a
three-way handshake.

Fix it by only applying the reuseport check for full sockets.

Fixes: cf7fbe660f2d ("bpf: Add socket assign support")
Signed-off-by: Joe Stringer <joe@wand.net.nz>
---
 net/core/filter.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Comments

Martin KaFai Lau April 8, 2020, 6:12 p.m. UTC | #1
On Tue, Apr 07, 2020 at 08:35:40PM -0700, Joe Stringer wrote:
> In testing, we found that for request sockets the sk->sk_reuseport field
> may yet be uninitialized, which caused bpf_sk_assign() to randomly
> succeed or return -ESOCKTNOSUPPORT when handling the forward ACK in a
> three-way handshake.
> 
> Fix it by only applying the reuseport check for full sockets.
Acked-by: Martin KaFai Lau <kafai@fb.com>
Daniel Borkmann April 8, 2020, 11:14 p.m. UTC | #2
On 4/8/20 5:35 AM, Joe Stringer wrote:
> In testing, we found that for request sockets the sk->sk_reuseport field
> may yet be uninitialized, which caused bpf_sk_assign() to randomly
> succeed or return -ESOCKTNOSUPPORT when handling the forward ACK in a
> three-way handshake.
> 
> Fix it by only applying the reuseport check for full sockets.
> 
> Fixes: cf7fbe660f2d ("bpf: Add socket assign support")
> Signed-off-by: Joe Stringer <joe@wand.net.nz>

Applied, thanks!
diff mbox series

Patch

diff --git a/net/core/filter.c b/net/core/filter.c
index 7628b947dbc3..7d6ceaa54d21 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -5925,7 +5925,7 @@  BPF_CALL_3(bpf_sk_assign, struct sk_buff *, skb, struct sock *, sk, u64, flags)
 		return -EOPNOTSUPP;
 	if (unlikely(dev_net(skb->dev) != sock_net(sk)))
 		return -ENETUNREACH;
-	if (unlikely(sk->sk_reuseport))
+	if (unlikely(sk_fullsock(sk) && sk->sk_reuseport))
 		return -ESOCKTNOSUPPORT;
 	if (sk_is_refcounted(sk) &&
 	    unlikely(!refcount_inc_not_zero(&sk->sk_refcnt)))