bpf: Add BPF_CGROUP_INET_SOCK_RELEASE hook

Sometimes it's handy to know when the socket gets freed. In
particular, we'd like to try to use a smarter allocation of
ports for bpf_bind and explore the possibility of limiting
the number of SOCK_DGRAM sockets the process can have.

Implement BPF_CGROUP_INET_SOCK_RELEASE hook that triggers on
inet socket release. It triggers only for userspace sockets
(not in-kernel ones) and therefore has the same semantics as
the existing BPF_CGROUP_INET_SOCK_CREATE.
Signed-off-by: Stanislav Fomichev <sdf@google.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Andrii Nakryiko <andriin@fb.com>
Link: https://lore.kernel.org/bpf/20200706230128.4073544-2-sdf@google.com

include/linux/bpf-cgroup.h

@@ -210,6 +210,9 @@ int bpf_percpu_cgroup_storage_update(struct bpf_map *map, void *key,
 #define BPF_CGROUP_RUN_PROG_INET_SOCK(sk)				       \
 	BPF_CGROUP_RUN_SK_PROG(sk, BPF_CGROUP_INET_SOCK_CREATE)
 
+#define BPF_CGROUP_RUN_PROG_INET_SOCK_RELEASE(sk)			       \
+	BPF_CGROUP_RUN_SK_PROG(sk, BPF_CGROUP_INET_SOCK_RELEASE)
+
 #define BPF_CGROUP_RUN_PROG_INET4_POST_BIND(sk)				       \
 	BPF_CGROUP_RUN_SK_PROG(sk, BPF_CGROUP_INET4_POST_BIND)
 
@@ -401,6 +404,7 @@ static inline int bpf_percpu_cgroup_storage_update(struct bpf_map *map,
 #define BPF_CGROUP_RUN_PROG_INET_INGRESS(sk,skb) ({ 0; })
 #define BPF_CGROUP_RUN_PROG_INET_EGRESS(sk,skb) ({ 0; })
 #define BPF_CGROUP_RUN_PROG_INET_SOCK(sk) ({ 0; })
+#define BPF_CGROUP_RUN_PROG_INET_SOCK_RELEASE(sk) ({ 0; })
 #define BPF_CGROUP_RUN_PROG_INET4_BIND(sk, uaddr) ({ 0; })
 #define BPF_CGROUP_RUN_PROG_INET6_BIND(sk, uaddr) ({ 0; })
 #define BPF_CGROUP_RUN_PROG_INET4_POST_BIND(sk) ({ 0; })

include/uapi/linux/bpf.h

@@ -226,6 +226,7 @@ enum bpf_attach_type {
 	BPF_CGROUP_INET4_GETSOCKNAME,
 	BPF_CGROUP_INET6_GETSOCKNAME,
 	BPF_XDP_DEVMAP,
+	BPF_CGROUP_INET_SOCK_RELEASE,
 	__MAX_BPF_ATTACH_TYPE
 };
 

kernel/bpf/syscall.c

@@ -1981,6 +1981,7 @@ bpf_prog_load_check_attach(enum bpf_prog_type prog_type,
 	case BPF_PROG_TYPE_CGROUP_SOCK:
 		switch (expected_attach_type) {
 		case BPF_CGROUP_INET_SOCK_CREATE:
+		case BPF_CGROUP_INET_SOCK_RELEASE:
 		case BPF_CGROUP_INET4_POST_BIND:
 		case BPF_CGROUP_INET6_POST_BIND:
 			return 0;
@@ -2779,6 +2780,7 @@ attach_type_to_prog_type(enum bpf_attach_type attach_type)
 		return BPF_PROG_TYPE_CGROUP_SKB;
 		break;
 	case BPF_CGROUP_INET_SOCK_CREATE:
+	case BPF_CGROUP_INET_SOCK_RELEASE:
 	case BPF_CGROUP_INET4_POST_BIND:
 	case BPF_CGROUP_INET6_POST_BIND:
 		return BPF_PROG_TYPE_CGROUP_SOCK;
@@ -2929,6 +2931,7 @@ static int bpf_prog_query(const union bpf_attr *attr,
 	case BPF_CGROUP_INET_INGRESS:
 	case BPF_CGROUP_INET_EGRESS:
 	case BPF_CGROUP_INET_SOCK_CREATE:
+	case BPF_CGROUP_INET_SOCK_RELEASE:
 	case BPF_CGROUP_INET4_BIND:
 	case BPF_CGROUP_INET6_BIND:
 	case BPF_CGROUP_INET4_POST_BIND:

net/core/filter.c

@@ -6890,6 +6890,7 @@ static bool __sock_filter_check_attach_type(int off,
 	case offsetof(struct bpf_sock, priority):
 		switch (attach_type) {
 		case BPF_CGROUP_INET_SOCK_CREATE:
+		case BPF_CGROUP_INET_SOCK_RELEASE:
 			goto full_access;
 		default:
 			return false;

net/ipv4/af_inet.c

@@ -411,6 +411,9 @@ int inet_release(struct socket *sock)
 	if (sk) {
 		long timeout;
 
+		if (!sk->sk_kern_sock)
+			BPF_CGROUP_RUN_PROG_INET_SOCK_RELEASE(sk);
+
 		/* Applications forget to leave groups before exiting */
 		ip_mc_drop_socket(sk);