Fix collisions in socket cookie generation
Subsystems (4)
Daniel Borkmann
maintainerv2 cover Fix collisions in socket cookie generation
This change makes the socket cookie generator as a global counter instead of per netns in order to fix cookie collisions for BPF use cases we ran into. See main patch #1 for more details.
Given the change is small/trivial and fixes an issue we're seeing my preference would be net tree (though it cleanly applies to net-next as well). Went for net tree instead of bpf tree here given the main change is in net/core/sock_diag.c, but either way would be fine with me.
Thanks a lot!
v1 -> v2:
- Fix up commit description in patch #1, thanks Eric!
Daniel Borkmann (2):
sock: make cookie generation global instead of per netns
bpf: sync bpf.h to tools infrastructure
include/net/net_namespace.h | 1 - include/uapi/linux/bpf.h | 4 ++-- net/core/sock_diag.c | 3 ++- tools/include/uapi/linux/bpf.h | 11 +++++++---- 4 files changed, 11 insertions(+), 8 deletions(-)
Signature · 1 line
2.17.1
Daniel Borkmann
maintainerPATCH v2 1/2 sock: make cookie generation global instead of per netns
Generating and retrieving socket cookies are a useful feature that is exposed to BPF for various program types through bpf_get_socket_cookie() helper.
The fact that the cookie counter is per netns is quite a limitation for BPF in practice in particular for programs in host namespace that use socket cookies as part of a map lookup key since they will be causing socket cookie collisions e.g. when attached to BPF cgroup hooks or cls_bpf on tc egress in host namespace handling container traffic from veth or ipvlan devices with peer in different netns. Change the counter to be global instead.
Socket cookie consumers must assume the value as opqaue in any case. Not every socket must have a cookie generated and knowledge of the counter value itself does not provide much value either way hence conversion to global is fine.
--- include/net/net_namespace.h | 1 - include/uapi/linux/bpf.h | 4 ++-- net/core/sock_diag.c | 3 ++- 3 files changed, 4 insertions(+), 4 deletions(-)
include/net/net_namespace.h +0 -1
@@ -61,7 +61,6 @@ struct net {spinlock_t rules_mod_lock;u32 hash_mix; -atomic64_t cookie_gen;struct list_head list;/* list of network namespaces */ struct list_head exit_list;/* To linked to call pernet exit
include/uapi/linux/bpf.h +2 -2
@@ -1466,8 +1466,8 @@ union bpf_attr {* If no cookie has been set yet, generate a new cookie. Once * generated, the socket cookie remains stable for the life of the * socket. This helper can be useful for monitoring per socket -* networking traffic statistics as it provides a unique socket -* identifier per namespace. +* networking traffic statistics as it provides a global socket +* identifier that can be assumed unique. * Return * A 8-byte long non-decreasing number on success, or 0 if the * socket field is missing inside *skb*.
net/core/sock_diag.c +2 -1
@@ -19,6 +19,7 @@ static const struct sock_diag_handler *sock_diag_handlers[AF_MAX];static int (*inet_rcv_compat)(struct sk_buff *skb,struct nlmsghdr *nlh);static DEFINE_MUTEX (sock_diag_table_mutex);static struct workqueue_struct *broadcast_wq; +static atomic64_t cookie_gen;u64 sock_gen_cookie (struct sock *sk) {
@@ -27,7 +28,7 @@ u64 sock_gen_cookie(struct sock *sk)if (res)return res; - res =atomic64_inc_return (&sock_net (sk)->cookie_gen); + res =atomic64_inc_return (&cookie_gen);atomic64_cmpxchg (&sk->sk_cookie,0 , res); } }
Signature · 1 line
2.17.1
Daniel Borkmann
maintainerPATCH v2 2/2 bpf: sync bpf.h to tools infrastructure
In reply to Daniel BorkmannPull in updates in BPF helper function description.
--- tools/include/uapi/linux/bpf.h | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-)
tools/include/uapi/linux/bpf.h +7 -4
@@ -1466,8 +1466,8 @@ union bpf_attr {* If no cookie has been set yet, generate a new cookie. Once * generated, the socket cookie remains stable for the life of the * socket. This helper can be useful for monitoring per socket -* networking traffic statistics as it provides a unique socket -* identifier per namespace. +* networking traffic statistics as it provides a global socket +* identifier that can be assumed unique. * Return * A 8-byte long non-decreasing number on success, or 0 if the * socket field is missing inside *skb*.
@@ -1571,8 +1571,11 @@ union bpf_attr {* but this is only implemented for native XDP (with driver * support) as of this writing). * -* All values for *flags* are reserved for future usage, and must -* be left at zero. +* The lower two bits of *flags* are used as the return code if +* the map lookup fails. This is so that the return value can be +* one of the XDP program return codes up to XDP_TX, as chosen by +* the caller. Any higher bits in the *flags* argument must be +* unset. * * When used to redirect packets to net devices, this helper * provides a high performance increase over **bpf_redirect**\ ().
Signature · 1 line
2.17.1
David Miller
maintainerFrom: Daniel Borkmann <[email protected]>
Date: Thu, 8 Aug 2019 13:57:24 +0200
Quoted 10 lines
This change makes the socket cookie generator as a global counter instead of per netns in order to fix cookie collisions for BPF use cases we ran into. See main patch #1 for more details.
Given the change is small/trivial and fixes an issue we're seeing my preference would be net tree (though it cleanly applies to net-next as well). Went for net tree instead of bpf tree here given the main change is in
net/core/sock_diag.c, but either way would be fine with me.
Quoted 4 lines
Thanks a lot!
v1 -> v2:
- Fix up commit description in patch #1, thanks Eric!
Series applied.