From: Menglong Dong imagedong@tencent.com

The return value of BPF_CGROUP_RUN_PROG_INET{4,6}_POST_BIND() in __inet_bind() is not handled properly. While the return value is non-zero, it will set inet_saddr and inet_rcv_saddr to 0 and exit:

err = BPF_CGROUP_RUN_PROG_INET4_POST_BIND(sk); if (err) { inet->inet_saddr = inet->inet_rcv_saddr = 0; goto out_release_sock; }

Let's take UDP for example and see what will happen. For UDP socket, it will be added to 'udp_prot.h.udp_table->hash' and 'udp_prot.h.udp_table->hash2' after the sk->sk_prot->get_port() called success. If 'inet->inet_rcv_saddr' is specified here, then 'sk' will be in the 'hslot2' of 'hash2' that it don't belong to (because inet_saddr is changed to 0), and UDP packet received will not be passed to this sock. If 'inet->inet_rcv_saddr' is not specified here, the sock will work fine, as it can receive packet properly, which is wired, as the 'bind()' is already failed.

To undo the get_port() operation, introduce the 'put_port' field for 'struct proto'. For TCP proto, it is inet_put_port(); For UDP proto, it is udp_lib_unhash(); For icmp proto, it is ping_unhash().

Therefore, after sys_bind() fail caused by BPF_CGROUP_RUN_PROG_INET4_POST_BIND(), it will be unbinded, which means that it can try to be binded to another port.

Signed-off-by: Menglong Dong imagedong@tencent.com --- v3: - NULL check for sk->sk_prot->put_port

v2: - introduce 'put_port' field for 'struct proto' --- include/net/sock.h | 1 + net/ipv4/af_inet.c | 2 ++ net/ipv4/ping.c | 1 + net/ipv4/tcp_ipv4.c | 1 + net/ipv4/udp.c | 1 + net/ipv6/af_inet6.c | 2 ++ net/ipv6/ping.c | 1 + net/ipv6/tcp_ipv6.c | 1 + net/ipv6/udp.c | 1 + 9 files changed, 11 insertions(+)

diff --git a/include/net/sock.h b/include/net/sock.h index 7b4b4237e6e0..ff9b508d9c5f 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -1209,6 +1209,7 @@ struct proto { void (*unhash)(struct sock *sk); void (*rehash)(struct sock *sk); int (*get_port)(struct sock *sk, unsigned short snum); + void (*put_port)(struct sock *sk); #ifdef CONFIG_BPF_SYSCALL int (*psock_update_sk_prot)(struct sock *sk, struct sk_psock *psock, diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c index f53184767ee7..9c465bac1eb0 100644 --- a/net/ipv4/af_inet.c +++ b/net/ipv4/af_inet.c @@ -531,6 +531,8 @@ int __inet_bind(struct sock *sk, struct sockaddr *uaddr, int addr_len, err = BPF_CGROUP_RUN_PROG_INET4_POST_BIND(sk); if (err) { inet->inet_saddr = inet->inet_rcv_saddr = 0; + if (sk->sk_prot->put_port) + sk->sk_prot->put_port(sk); goto out_release_sock; } } diff --git a/net/ipv4/ping.c b/net/ipv4/ping.c index e540b0dcf085..0e56df3a45e2 100644 --- a/net/ipv4/ping.c +++ b/net/ipv4/ping.c @@ -994,6 +994,7 @@ struct proto ping_prot = { .hash = ping_hash, .unhash = ping_unhash, .get_port = ping_get_port, + .put_port = ping_unhash, .obj_size = sizeof(struct inet_sock), }; EXPORT_SYMBOL(ping_prot); diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index ac10e4cdd8d0..9861786b8336 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -3076,6 +3076,7 @@ struct proto tcp_prot = { .hash = inet_hash, .unhash = inet_unhash, .get_port = inet_csk_get_port, + .put_port = inet_put_port, #ifdef CONFIG_BPF_SYSCALL .psock_update_sk_prot = tcp_bpf_update_proto, #endif diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index 7b18a6f42f18..c2a4411d2b04 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -2927,6 +2927,7 @@ struct proto udp_prot = { .unhash = udp_lib_unhash, .rehash = udp_v4_rehash, .get_port = udp_v4_get_port, + .put_port = udp_lib_unhash, #ifdef CONFIG_BPF_SYSCALL .psock_update_sk_prot = udp_bpf_update_proto, #endif diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c index d1636425654e..8fe7900f1949 100644 --- a/net/ipv6/af_inet6.c +++ b/net/ipv6/af_inet6.c @@ -413,6 +413,8 @@ static int __inet6_bind(struct sock *sk, struct sockaddr *uaddr, int addr_len, if (err) { sk->sk_ipv6only = saved_ipv6only; inet_reset_saddr(sk); + if (sk->sk_prot->put_port) + sk->sk_prot->put_port(sk); goto out; } } diff --git a/net/ipv6/ping.c b/net/ipv6/ping.c index 6ac88fe24a8e..9256f6ba87ef 100644 --- a/net/ipv6/ping.c +++ b/net/ipv6/ping.c @@ -177,6 +177,7 @@ struct proto pingv6_prot = { .hash = ping_hash, .unhash = ping_unhash, .get_port = ping_get_port, + .put_port = ping_unhash, .obj_size = sizeof(struct raw6_sock), }; EXPORT_SYMBOL_GPL(pingv6_prot); diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 1ac243d18c2b..075ee8a2df3b 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -2181,6 +2181,7 @@ struct proto tcpv6_prot = { .hash = inet6_hash, .unhash = inet_unhash, .get_port = inet_csk_get_port, + .put_port = inet_put_port, #ifdef CONFIG_BPF_SYSCALL .psock_update_sk_prot = tcp_bpf_update_proto, #endif diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c index 1accc06abc54..90718a924ca8 100644 --- a/net/ipv6/udp.c +++ b/net/ipv6/udp.c @@ -1732,6 +1732,7 @@ struct proto udpv6_prot = { .unhash = udp_lib_unhash, .rehash = udp_v6_rehash, .get_port = udp_v6_get_port, + .put_port = udp_lib_unhash, #ifdef CONFIG_BPF_SYSCALL .psock_update_sk_prot = udp_bpf_update_proto, #endif

From: Menglong Dong imagedong@tencent.com

With previous patch, kernel is able to 'put_port' after sys_bind() fails. Add the test for that case: rebind another port after sys_bind() fails. If the bind success, it means previous bind operation is already undoed.

Signed-off-by: Menglong Dong imagedong@tencent.com --- tools/testing/selftests/bpf/test_sock.c | 150 ++++++++++++++++++++---- 1 file changed, 130 insertions(+), 20 deletions(-)

diff --git a/tools/testing/selftests/bpf/test_sock.c b/tools/testing/selftests/bpf/test_sock.c index 94f9b126f5ed..fe10f8134278 100644 --- a/tools/testing/selftests/bpf/test_sock.c +++ b/tools/testing/selftests/bpf/test_sock.c @@ -35,12 +35,15 @@ struct sock_test { /* Endpoint to bind() to */ const char *ip; unsigned short port; + unsigned short port_retry; /* Expected test result */ enum { LOAD_REJECT, ATTACH_REJECT, BIND_REJECT, SUCCESS, + RETRY_SUCCESS, + RETRY_REJECT } result; };

@@ -251,6 +254,99 @@ static struct sock_test tests[] = { .port = 4098, .result = SUCCESS, }, + { + .descr = "bind4 deny specific IP & port of TCP, and retry", + .insns = { + BPF_MOV64_REG(BPF_REG_6, BPF_REG_1), + + /* if (ip == expected && port == expected) */ + BPF_LDX_MEM(BPF_W, BPF_REG_7, BPF_REG_6, + offsetof(struct bpf_sock, src_ip4)), + BPF_JMP_IMM(BPF_JNE, BPF_REG_7, + __bpf_constant_ntohl(0x7F000001), 4), + BPF_LDX_MEM(BPF_W, BPF_REG_7, BPF_REG_6, + offsetof(struct bpf_sock, src_port)), + BPF_JMP_IMM(BPF_JNE, BPF_REG_7, 0x1002, 2), + + /* return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_JMP_A(1), + + /* else return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .expected_attach_type = BPF_CGROUP_INET4_POST_BIND, + .attach_type = BPF_CGROUP_INET4_POST_BIND, + .domain = AF_INET, + .type = SOCK_STREAM, + .ip = "127.0.0.1", + .port = 4098, + .port_retry = 5000, + .result = RETRY_SUCCESS, + }, + { + .descr = "bind4 deny specific IP & port of UDP, and retry", + .insns = { + BPF_MOV64_REG(BPF_REG_6, BPF_REG_1), + + /* if (ip == expected && port == expected) */ + BPF_LDX_MEM(BPF_W, BPF_REG_7, BPF_REG_6, + offsetof(struct bpf_sock, src_ip4)), + BPF_JMP_IMM(BPF_JNE, BPF_REG_7, + __bpf_constant_ntohl(0x7F000001), 4), + BPF_LDX_MEM(BPF_W, BPF_REG_7, BPF_REG_6, + offsetof(struct bpf_sock, src_port)), + BPF_JMP_IMM(BPF_JNE, BPF_REG_7, 0x1002, 2), + + /* return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_JMP_A(1), + + /* else return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .expected_attach_type = BPF_CGROUP_INET4_POST_BIND, + .attach_type = BPF_CGROUP_INET4_POST_BIND, + .domain = AF_INET, + .type = SOCK_DGRAM, + .ip = "127.0.0.1", + .port = 4098, + .port_retry = 5000, + .result = RETRY_SUCCESS, + }, + { + .descr = "bind6 deny specific IP & port, and retry", + .insns = { + BPF_MOV64_REG(BPF_REG_6, BPF_REG_1), + + /* if (ip == expected && port == expected) */ + BPF_LDX_MEM(BPF_W, BPF_REG_7, BPF_REG_6, + offsetof(struct bpf_sock, src_ip6[3])), + BPF_JMP_IMM(BPF_JNE, BPF_REG_7, + __bpf_constant_ntohl(0x00000001), 4), + BPF_LDX_MEM(BPF_W, BPF_REG_7, BPF_REG_6, + offsetof(struct bpf_sock, src_port)), + BPF_JMP_IMM(BPF_JNE, BPF_REG_7, 0x2001, 2), + + /* return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_JMP_A(1), + + /* else return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .expected_attach_type = BPF_CGROUP_INET6_POST_BIND, + .attach_type = BPF_CGROUP_INET6_POST_BIND, + .domain = AF_INET6, + .type = SOCK_STREAM, + .ip = "::1", + .port = 8193, + .port_retry = 9000, + .result = RETRY_SUCCESS, + }, { .descr = "bind4 allow all", .insns = { @@ -315,14 +411,15 @@ static int attach_sock_prog(int cgfd, int progfd, return bpf_prog_attach(progfd, cgfd, attach_type, BPF_F_ALLOW_OVERRIDE); }

-static int bind_sock(int domain, int type, const char *ip, unsigned short port) +static int bind_sock(int domain, int type, const char *ip, + unsigned short port, unsigned short port_retry) { struct sockaddr_storage addr; struct sockaddr_in6 *addr6; struct sockaddr_in *addr4; int sockfd = -1; socklen_t len; - int err = 0; + int res = SUCCESS;

sockfd = socket(domain, type, 0); if (sockfd < 0) @@ -348,21 +445,44 @@ static int bind_sock(int domain, int type, const char *ip, unsigned short port) goto err; }

- if (bind(sockfd, (const struct sockaddr *)&addr, len) == -1) - goto err; + if (bind(sockfd, (const struct sockaddr *)&addr, len) == -1) { + /* sys_bind() may fail for different reasons, errno has to be + * checked to confirm that BPF program rejected it. + */ + if (errno != EPERM) + goto err; + if (port_retry) + goto retry; + res = BIND_REJECT; + goto out; + }

+ goto out; +retry: + if (domain == AF_INET) + addr4->sin_port = htons(port_retry); + else + addr6->sin6_port = htons(port_retry); + if (bind(sockfd, (const struct sockaddr *)&addr, len) == -1) { + if (errno != EPERM) + goto err; + res = RETRY_REJECT; + } else { + res = RETRY_SUCCESS; + } goto out; err: - err = -1; + res = -1; out: close(sockfd); - return err; + return res; }

static int run_test_case(int cgfd, const struct sock_test *test) { int progfd = -1; int err = 0; + int res;

printf("Test case: %s .. ", test->descr); progfd = load_sock_prog(test->insns, test->expected_attach_type); @@ -380,21 +500,11 @@ static int run_test_case(int cgfd, const struct sock_test *test) goto err; }

- if (bind_sock(test->domain, test->type, test->ip, test->port) == -1) { - /* sys_bind() may fail for different reasons, errno has to be - * checked to confirm that BPF program rejected it. - */ - if (test->result == BIND_REJECT && errno == EPERM) - goto out; - else - goto err; - } - + res = bind_sock(test->domain, test->type, test->ip, test->port, + test->port_retry); + if (res > 0 && test->result == res) + goto out;

- if (test->result != SUCCESS) - goto err; - - goto out; err: err = -1; out: