diff options
Diffstat (limited to 'net/socket.c')
-rw-r--r-- | net/socket.c | 167 |
1 files changed, 110 insertions, 57 deletions
diff --git a/net/socket.c b/net/socket.c index 2b0e54b2405c..848116d06b51 100644 --- a/net/socket.c +++ b/net/socket.c @@ -136,9 +136,10 @@ static void sock_splice_eof(struct file *file); static void sock_show_fdinfo(struct seq_file *m, struct file *f) { struct socket *sock = f->private_data; + const struct proto_ops *ops = READ_ONCE(sock->ops); - if (sock->ops->show_fdinfo) - sock->ops->show_fdinfo(m, sock); + if (ops->show_fdinfo) + ops->show_fdinfo(m, sock); } #else #define sock_show_fdinfo NULL @@ -646,12 +647,14 @@ EXPORT_SYMBOL(sock_alloc); static void __sock_release(struct socket *sock, struct inode *inode) { - if (sock->ops) { - struct module *owner = sock->ops->owner; + const struct proto_ops *ops = READ_ONCE(sock->ops); + + if (ops) { + struct module *owner = ops->owner; if (inode) inode_lock(inode); - sock->ops->release(sock); + ops->release(sock); sock->sk = NULL; if (inode) inode_unlock(inode); @@ -722,7 +725,7 @@ static noinline void call_trace_sock_send_length(struct sock *sk, int ret, static inline int sock_sendmsg_nosec(struct socket *sock, struct msghdr *msg) { - int ret = INDIRECT_CALL_INET(sock->ops->sendmsg, inet6_sendmsg, + int ret = INDIRECT_CALL_INET(READ_ONCE(sock->ops)->sendmsg, inet6_sendmsg, inet_sendmsg, sock, msg, msg_data_left(msg)); BUG_ON(ret == -EIOCBQUEUED); @@ -786,13 +789,14 @@ int kernel_sendmsg_locked(struct sock *sk, struct msghdr *msg, struct kvec *vec, size_t num, size_t size) { struct socket *sock = sk->sk_socket; + const struct proto_ops *ops = READ_ONCE(sock->ops); - if (!sock->ops->sendmsg_locked) + if (!ops->sendmsg_locked) return sock_no_sendmsg_locked(sk, msg, size); iov_iter_kvec(&msg->msg_iter, ITER_SOURCE, vec, num, size); - return sock->ops->sendmsg_locked(sk, msg, msg_data_left(msg)); + return ops->sendmsg_locked(sk, msg, msg_data_left(msg)); } EXPORT_SYMBOL(kernel_sendmsg_locked); @@ -1017,7 +1021,8 @@ static noinline void call_trace_sock_recv_length(struct sock *sk, int ret, int f static inline int sock_recvmsg_nosec(struct socket *sock, struct msghdr *msg, int flags) { - int ret = INDIRECT_CALL_INET(sock->ops->recvmsg, inet6_recvmsg, + int ret = INDIRECT_CALL_INET(READ_ONCE(sock->ops)->recvmsg, + inet6_recvmsg, inet_recvmsg, sock, msg, msg_data_left(msg), flags); if (trace_sock_recv_length_enabled()) @@ -1072,19 +1077,23 @@ static ssize_t sock_splice_read(struct file *file, loff_t *ppos, unsigned int flags) { struct socket *sock = file->private_data; + const struct proto_ops *ops; - if (unlikely(!sock->ops->splice_read)) + ops = READ_ONCE(sock->ops); + if (unlikely(!ops->splice_read)) return copy_splice_read(file, ppos, pipe, len, flags); - return sock->ops->splice_read(sock, ppos, pipe, len, flags); + return ops->splice_read(sock, ppos, pipe, len, flags); } static void sock_splice_eof(struct file *file) { struct socket *sock = file->private_data; + const struct proto_ops *ops; - if (sock->ops->splice_eof) - sock->ops->splice_eof(sock); + ops = READ_ONCE(sock->ops); + if (ops->splice_eof) + ops->splice_eof(sock); } static ssize_t sock_read_iter(struct kiocb *iocb, struct iov_iter *to) @@ -1181,13 +1190,14 @@ EXPORT_SYMBOL(vlan_ioctl_set); static long sock_do_ioctl(struct net *net, struct socket *sock, unsigned int cmd, unsigned long arg) { + const struct proto_ops *ops = READ_ONCE(sock->ops); struct ifreq ifr; bool need_copyout; int err; void __user *argp = (void __user *)arg; void __user *data; - err = sock->ops->ioctl(sock, cmd, arg); + err = ops->ioctl(sock, cmd, arg); /* * If this ioctl is unknown try to hand it down @@ -1216,6 +1226,7 @@ static long sock_do_ioctl(struct net *net, struct socket *sock, static long sock_ioctl(struct file *file, unsigned cmd, unsigned long arg) { + const struct proto_ops *ops; struct socket *sock; struct sock *sk; void __user *argp = (void __user *)arg; @@ -1223,6 +1234,7 @@ static long sock_ioctl(struct file *file, unsigned cmd, unsigned long arg) struct net *net; sock = file->private_data; + ops = READ_ONCE(sock->ops); sk = sock->sk; net = sock_net(sk); if (unlikely(cmd >= SIOCDEVPRIVATE && cmd <= (SIOCDEVPRIVATE + 15))) { @@ -1280,23 +1292,23 @@ static long sock_ioctl(struct file *file, unsigned cmd, unsigned long arg) break; case SIOCGSTAMP_OLD: case SIOCGSTAMPNS_OLD: - if (!sock->ops->gettstamp) { + if (!ops->gettstamp) { err = -ENOIOCTLCMD; break; } - err = sock->ops->gettstamp(sock, argp, - cmd == SIOCGSTAMP_OLD, - !IS_ENABLED(CONFIG_64BIT)); + err = ops->gettstamp(sock, argp, + cmd == SIOCGSTAMP_OLD, + !IS_ENABLED(CONFIG_64BIT)); break; case SIOCGSTAMP_NEW: case SIOCGSTAMPNS_NEW: - if (!sock->ops->gettstamp) { + if (!ops->gettstamp) { err = -ENOIOCTLCMD; break; } - err = sock->ops->gettstamp(sock, argp, - cmd == SIOCGSTAMP_NEW, - false); + err = ops->gettstamp(sock, argp, + cmd == SIOCGSTAMP_NEW, + false); break; case SIOCGIFCONF: @@ -1357,9 +1369,10 @@ EXPORT_SYMBOL(sock_create_lite); static __poll_t sock_poll(struct file *file, poll_table *wait) { struct socket *sock = file->private_data; + const struct proto_ops *ops = READ_ONCE(sock->ops); __poll_t events = poll_requested_events(wait), flag = 0; - if (!sock->ops->poll) + if (!ops->poll) return 0; if (sk_can_busy_loop(sock->sk)) { @@ -1371,14 +1384,14 @@ static __poll_t sock_poll(struct file *file, poll_table *wait) flag = POLL_BUSY_LOOP; } - return sock->ops->poll(file, sock, wait) | flag; + return ops->poll(file, sock, wait) | flag; } static int sock_mmap(struct file *file, struct vm_area_struct *vma) { struct socket *sock = file->private_data; - return sock->ops->mmap(file, sock, vma); + return READ_ONCE(sock->ops)->mmap(file, sock, vma); } static int sock_close(struct inode *inode, struct file *filp) @@ -1644,12 +1657,36 @@ struct file *__sys_socket_file(int family, int type, int protocol) return sock_alloc_file(sock, flags, NULL); } +/* A hook for bpf progs to attach to and update socket protocol. + * + * A static noinline declaration here could cause the compiler to + * optimize away the function. A global noinline declaration will + * keep the definition, but may optimize away the callsite. + * Therefore, __weak is needed to ensure that the call is still + * emitted, by telling the compiler that we don't know what the + * function might eventually be. + * + * __diag_* below are needed to dismiss the missing prototype warning. + */ + +__diag_push(); +__diag_ignore_all("-Wmissing-prototypes", + "A fmod_ret entry point for BPF programs"); + +__weak noinline int update_socket_protocol(int family, int type, int protocol) +{ + return protocol; +} + +__diag_pop(); + int __sys_socket(int family, int type, int protocol) { struct socket *sock; int flags; - sock = __sys_socket_create(family, type, protocol); + sock = __sys_socket_create(family, type, + update_socket_protocol(family, type, protocol)); if (IS_ERR(sock)) return PTR_ERR(sock); @@ -1728,7 +1765,7 @@ int __sys_socketpair(int family, int type, int protocol, int __user *usockvec) goto out; } - err = sock1->ops->socketpair(sock1, sock2); + err = READ_ONCE(sock1->ops)->socketpair(sock1, sock2); if (unlikely(err < 0)) { sock_release(sock2); sock_release(sock1); @@ -1789,7 +1826,7 @@ int __sys_bind(int fd, struct sockaddr __user *umyaddr, int addrlen) (struct sockaddr *)&address, addrlen); if (!err) - err = sock->ops->bind(sock, + err = READ_ONCE(sock->ops)->bind(sock, (struct sockaddr *) &address, addrlen); } @@ -1823,7 +1860,7 @@ int __sys_listen(int fd, int backlog) err = security_socket_listen(sock, backlog); if (!err) - err = sock->ops->listen(sock, backlog); + err = READ_ONCE(sock->ops)->listen(sock, backlog); fput_light(sock->file, fput_needed); } @@ -1843,6 +1880,7 @@ struct file *do_accept(struct file *file, unsigned file_flags, struct file *newfile; int err, len; struct sockaddr_storage address; + const struct proto_ops *ops; sock = sock_from_file(file); if (!sock) @@ -1851,15 +1889,16 @@ struct file *do_accept(struct file *file, unsigned file_flags, newsock = sock_alloc(); if (!newsock) return ERR_PTR(-ENFILE); + ops = READ_ONCE(sock->ops); newsock->type = sock->type; - newsock->ops = sock->ops; + newsock->ops = ops; /* * We don't need try_module_get here, as the listening socket (sock) * has the protocol module (sock->ops->owner) held. */ - __module_get(newsock->ops->owner); + __module_get(ops->owner); newfile = sock_alloc_file(newsock, flags, sock->sk->sk_prot_creator->name); if (IS_ERR(newfile)) @@ -1869,14 +1908,13 @@ struct file *do_accept(struct file *file, unsigned file_flags, if (err) goto out_fd; - err = sock->ops->accept(sock, newsock, sock->file->f_flags | file_flags, + err = ops->accept(sock, newsock, sock->file->f_flags | file_flags, false); if (err < 0) goto out_fd; if (upeer_sockaddr) { - len = newsock->ops->getname(newsock, - (struct sockaddr *)&address, 2); + len = ops->getname(newsock, (struct sockaddr *)&address, 2); if (len < 0) { err = -ECONNABORTED; goto out_fd; @@ -1989,8 +2027,8 @@ int __sys_connect_file(struct file *file, struct sockaddr_storage *address, if (err) goto out; - err = sock->ops->connect(sock, (struct sockaddr *)address, addrlen, - sock->file->f_flags | file_flags); + err = READ_ONCE(sock->ops)->connect(sock, (struct sockaddr *)address, + addrlen, sock->file->f_flags | file_flags); out: return err; } @@ -2039,7 +2077,7 @@ int __sys_getsockname(int fd, struct sockaddr __user *usockaddr, if (err) goto out_put; - err = sock->ops->getname(sock, (struct sockaddr *)&address, 0); + err = READ_ONCE(sock->ops)->getname(sock, (struct sockaddr *)&address, 0); if (err < 0) goto out_put; /* "err" is actually length in this case */ @@ -2071,13 +2109,15 @@ int __sys_getpeername(int fd, struct sockaddr __user *usockaddr, sock = sockfd_lookup_light(fd, &err, &fput_needed); if (sock != NULL) { + const struct proto_ops *ops = READ_ONCE(sock->ops); + err = security_socket_getpeername(sock); if (err) { fput_light(sock->file, fput_needed); return err; } - err = sock->ops->getname(sock, (struct sockaddr *)&address, 1); + err = ops->getname(sock, (struct sockaddr *)&address, 1); if (err >= 0) /* "err" is actually length in this case */ err = move_addr_to_user(&address, err, usockaddr, @@ -2227,6 +2267,7 @@ int __sys_setsockopt(int fd, int level, int optname, char __user *user_optval, int optlen) { sockptr_t optval = USER_SOCKPTR(user_optval); + const struct proto_ops *ops; char *kernel_optval = NULL; int err, fput_needed; struct socket *sock; @@ -2255,12 +2296,13 @@ int __sys_setsockopt(int fd, int level, int optname, char __user *user_optval, if (kernel_optval) optval = KERNEL_SOCKPTR(kernel_optval); + ops = READ_ONCE(sock->ops); if (level == SOL_SOCKET && !sock_use_custom_sol_socket(sock)) err = sock_setsockopt(sock, level, optname, optval, optlen); - else if (unlikely(!sock->ops->setsockopt)) + else if (unlikely(!ops->setsockopt)) err = -EOPNOTSUPP; else - err = sock->ops->setsockopt(sock, level, optname, optval, + err = ops->setsockopt(sock, level, optname, optval, optlen); kfree(kernel_optval); out_put: @@ -2285,6 +2327,7 @@ int __sys_getsockopt(int fd, int level, int optname, char __user *optval, int __user *optlen) { int max_optlen __maybe_unused; + const struct proto_ops *ops; int err, fput_needed; struct socket *sock; @@ -2299,12 +2342,13 @@ int __sys_getsockopt(int fd, int level, int optname, char __user *optval, if (!in_compat_syscall()) max_optlen = BPF_CGROUP_GETSOCKOPT_MAX_OPTLEN(optlen); + ops = READ_ONCE(sock->ops); if (level == SOL_SOCKET) err = sock_getsockopt(sock, level, optname, optval, optlen); - else if (unlikely(!sock->ops->getsockopt)) + else if (unlikely(!ops->getsockopt)) err = -EOPNOTSUPP; else - err = sock->ops->getsockopt(sock, level, optname, optval, + err = ops->getsockopt(sock, level, optname, optval, optlen); if (!in_compat_syscall()) @@ -2332,7 +2376,7 @@ int __sys_shutdown_sock(struct socket *sock, int how) err = security_socket_shutdown(sock, how); if (!err) - err = sock->ops->shutdown(sock, how); + err = READ_ONCE(sock->ops)->shutdown(sock, how); return err; } @@ -3324,6 +3368,7 @@ static int compat_sock_ioctl_trans(struct file *file, struct socket *sock, void __user *argp = compat_ptr(arg); struct sock *sk = sock->sk; struct net *net = sock_net(sk); + const struct proto_ops *ops; if (cmd >= SIOCDEVPRIVATE && cmd <= (SIOCDEVPRIVATE + 15)) return sock_ioctl(file, cmd, (unsigned long)argp); @@ -3333,10 +3378,11 @@ static int compat_sock_ioctl_trans(struct file *file, struct socket *sock, return compat_siocwandev(net, argp); case SIOCGSTAMP_OLD: case SIOCGSTAMPNS_OLD: - if (!sock->ops->gettstamp) + ops = READ_ONCE(sock->ops); + if (!ops->gettstamp) return -ENOIOCTLCMD; - return sock->ops->gettstamp(sock, argp, cmd == SIOCGSTAMP_OLD, - !COMPAT_USE_64BIT_TIME); + return ops->gettstamp(sock, argp, cmd == SIOCGSTAMP_OLD, + !COMPAT_USE_64BIT_TIME); case SIOCETHTOOL: case SIOCBONDSLAVEINFOQUERY: @@ -3417,6 +3463,7 @@ static long compat_sock_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { struct socket *sock = file->private_data; + const struct proto_ops *ops = READ_ONCE(sock->ops); int ret = -ENOIOCTLCMD; struct sock *sk; struct net *net; @@ -3424,8 +3471,8 @@ static long compat_sock_ioctl(struct file *file, unsigned int cmd, sk = sock->sk; net = sock_net(sk); - if (sock->ops->compat_ioctl) - ret = sock->ops->compat_ioctl(sock, cmd, arg); + if (ops->compat_ioctl) + ret = ops->compat_ioctl(sock, cmd, arg); if (ret == -ENOIOCTLCMD && (cmd >= SIOCIWFIRST && cmd <= SIOCIWLAST)) @@ -3449,7 +3496,7 @@ static long compat_sock_ioctl(struct file *file, unsigned int cmd, int kernel_bind(struct socket *sock, struct sockaddr *addr, int addrlen) { - return sock->ops->bind(sock, addr, addrlen); + return READ_ONCE(sock->ops)->bind(sock, addr, addrlen); } EXPORT_SYMBOL(kernel_bind); @@ -3463,7 +3510,7 @@ EXPORT_SYMBOL(kernel_bind); int kernel_listen(struct socket *sock, int backlog) { - return sock->ops->listen(sock, backlog); + return READ_ONCE(sock->ops)->listen(sock, backlog); } EXPORT_SYMBOL(kernel_listen); @@ -3481,6 +3528,7 @@ EXPORT_SYMBOL(kernel_listen); int kernel_accept(struct socket *sock, struct socket **newsock, int flags) { struct sock *sk = sock->sk; + const struct proto_ops *ops = READ_ONCE(sock->ops); int err; err = sock_create_lite(sk->sk_family, sk->sk_type, sk->sk_protocol, @@ -3488,15 +3536,15 @@ int kernel_accept(struct socket *sock, struct socket **newsock, int flags) if (err < 0) goto done; - err = sock->ops->accept(sock, *newsock, flags, true); + err = ops->accept(sock, *newsock, flags, true); if (err < 0) { sock_release(*newsock); *newsock = NULL; goto done; } - (*newsock)->ops = sock->ops; - __module_get((*newsock)->ops->owner); + (*newsock)->ops = ops; + __module_get(ops->owner); done: return err; @@ -3519,7 +3567,12 @@ EXPORT_SYMBOL(kernel_accept); int kernel_connect(struct socket *sock, struct sockaddr *addr, int addrlen, int flags) { - return sock->ops->connect(sock, addr, addrlen, flags); + struct sockaddr_storage address; + + memcpy(&address, addr, addrlen); + + return READ_ONCE(sock->ops)->connect(sock, (struct sockaddr *)&address, + addrlen, flags); } EXPORT_SYMBOL(kernel_connect); @@ -3534,7 +3587,7 @@ EXPORT_SYMBOL(kernel_connect); int kernel_getsockname(struct socket *sock, struct sockaddr *addr) { - return sock->ops->getname(sock, addr, 0); + return READ_ONCE(sock->ops)->getname(sock, addr, 0); } EXPORT_SYMBOL(kernel_getsockname); @@ -3549,7 +3602,7 @@ EXPORT_SYMBOL(kernel_getsockname); int kernel_getpeername(struct socket *sock, struct sockaddr *addr) { - return sock->ops->getname(sock, addr, 1); + return READ_ONCE(sock->ops)->getname(sock, addr, 1); } EXPORT_SYMBOL(kernel_getpeername); @@ -3563,7 +3616,7 @@ EXPORT_SYMBOL(kernel_getpeername); int kernel_sock_shutdown(struct socket *sock, enum sock_shutdown_cmd how) { - return sock->ops->shutdown(sock, how); + return READ_ONCE(sock->ops)->shutdown(sock, how); } EXPORT_SYMBOL(kernel_sock_shutdown); |