X-Git-Url: https://err.no/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=net%2Fdccp%2Fproto.c;h=18a0e69c9dc75f709e15be01e7ef6def17de3eb5;hb=f35279d3f713e5c97b98cbdbf47d98f79942c11f;hp=70284e6afe05cda3df8b83cd2dccd502f8b838af;hpb=7c657876b63cb1d8a2ec06f8fc6c37bb8412e66c;p=linux-2.6 diff --git a/net/dccp/proto.c b/net/dccp/proto.c index 70284e6afe..18a0e69c9d 100644 --- a/net/dccp/proto.c +++ b/net/dccp/proto.c @@ -39,7 +39,7 @@ #include "ccid.h" #include "dccp.h" -DEFINE_SNMP_STAT(struct dccp_mib, dccp_statistics); +DEFINE_SNMP_STAT(struct dccp_mib, dccp_statistics) __read_mostly; atomic_t dccp_orphan_count = ATOMIC_INIT(0); @@ -140,6 +140,62 @@ int dccp_disconnect(struct sock *sk, int flags) return err; } +/* + * Wait for a DCCP event. + * + * Note that we don't need to lock the socket, as the upper poll layers + * take care of normal races (between the test and the event) and we don't + * go look at any of the socket buffers directly. + */ +static unsigned int dccp_poll(struct file *file, struct socket *sock, + poll_table *wait) +{ + unsigned int mask; + struct sock *sk = sock->sk; + + poll_wait(file, sk->sk_sleep, wait); + if (sk->sk_state == DCCP_LISTEN) + return inet_csk_listen_poll(sk); + + /* Socket is not locked. We are protected from async events + by poll logic and correct handling of state changes + made by another threads is impossible in any case. + */ + + mask = 0; + if (sk->sk_err) + mask = POLLERR; + + if (sk->sk_shutdown == SHUTDOWN_MASK || sk->sk_state == DCCP_CLOSED) + mask |= POLLHUP; + if (sk->sk_shutdown & RCV_SHUTDOWN) + mask |= POLLIN | POLLRDNORM; + + /* Connected? */ + if ((1 << sk->sk_state) & ~(DCCPF_REQUESTING | DCCPF_RESPOND)) { + if (atomic_read(&sk->sk_rmem_alloc) > 0) + mask |= POLLIN | POLLRDNORM; + + if (!(sk->sk_shutdown & SEND_SHUTDOWN)) { + if (sk_stream_wspace(sk) >= sk_stream_min_wspace(sk)) { + mask |= POLLOUT | POLLWRNORM; + } else { /* send SIGIO later */ + set_bit(SOCK_ASYNC_NOSPACE, + &sk->sk_socket->flags); + set_bit(SOCK_NOSPACE, &sk->sk_socket->flags); + + /* Race breaker. If space is freed after + * wspace test but before the flags are set, + * IO signal will be lost. + */ + if (sk_stream_wspace(sk) >= sk_stream_min_wspace(sk)) + mask |= POLLOUT | POLLWRNORM; + } + } + } + return mask; +} + int dccp_ioctl(struct sock *sk, int cmd, unsigned long arg) { dccp_pr_debug("entry\n"); @@ -147,25 +203,69 @@ int dccp_ioctl(struct sock *sk, int cmd, unsigned long arg) } int dccp_setsockopt(struct sock *sk, int level, int optname, - char *optval, int optlen) + char __user *optval, int optlen) { - dccp_pr_debug("entry\n"); + struct dccp_sock *dp; + int err; + int val; if (level != SOL_DCCP) return ip_setsockopt(sk, level, optname, optval, optlen); - return -EOPNOTSUPP; + if (optlen < sizeof(int)) + return -EINVAL; + + if (get_user(val, (int __user *)optval)) + return -EFAULT; + + lock_sock(sk); + + dp = dccp_sk(sk); + err = 0; + + switch (optname) { + case DCCP_SOCKOPT_PACKET_SIZE: + dp->dccps_packet_size = val; + break; + default: + err = -ENOPROTOOPT; + break; + } + + release_sock(sk); + return err; } int dccp_getsockopt(struct sock *sk, int level, int optname, - char *optval, int *optlen) + char __user *optval, int __user *optlen) { - dccp_pr_debug("entry\n"); + struct dccp_sock *dp; + int val, len; if (level != SOL_DCCP) return ip_getsockopt(sk, level, optname, optval, optlen); - return -EOPNOTSUPP; + if (get_user(len, optlen)) + return -EFAULT; + + len = min_t(unsigned int, len, sizeof(int)); + if (len < 0) + return -EINVAL; + + dp = dccp_sk(sk); + + switch (optname) { + case DCCP_SOCKOPT_PACKET_SIZE: + val = dp->dccps_packet_size; + break; + default: + return -ENOPROTOOPT; + } + + if (put_user(len, optlen) || copy_to_user(optval, &val, len)) + return -EFAULT; + + return 0; } int dccp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg, @@ -182,8 +282,7 @@ int dccp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg, return -EMSGSIZE; lock_sock(sk); - - timeo = sock_sndtimeo(sk, flags & MSG_DONTWAIT); + timeo = sock_sndtimeo(sk, noblock); /* * We have to use sk_stream_wait_connect here to set sk_write_pending, @@ -192,265 +291,136 @@ int dccp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg, /* Wait for a connection to finish. */ if ((1 << sk->sk_state) & ~(DCCPF_OPEN | DCCPF_PARTOPEN | DCCPF_CLOSING)) if ((rc = sk_stream_wait_connect(sk, &timeo)) != 0) - goto out_err; + goto out_release; size = sk->sk_prot->max_header + len; release_sock(sk); skb = sock_alloc_send_skb(sk, size, noblock, &rc); lock_sock(sk); - if (skb == NULL) goto out_release; skb_reserve(skb, sk->sk_prot->max_header); rc = memcpy_fromiovec(skb_put(skb, len), msg->msg_iov, len); - if (rc == 0) { - struct dccp_skb_cb *dcb = DCCP_SKB_CB(skb); - const struct dccp_ackpkts *ap = dp->dccps_hc_rx_ackpkts; - long delay; - - /* - * XXX: This is just to match the Waikato tree CA interaction - * points, after the CCID3 code is stable and I have a better - * understanding of behaviour I'll change this to look more like - * TCP. - */ - while (1) { - rc = ccid_hc_tx_send_packet(dp->dccps_hc_tx_ccid, sk, - skb, len, &delay); - if (rc == 0) - break; - if (rc != -EAGAIN) - goto out_discard; - if (delay > timeo) - goto out_discard; - release_sock(sk); - delay = schedule_timeout(delay); - lock_sock(sk); - timeo -= delay; - if (signal_pending(current)) - goto out_interrupted; - rc = -EPIPE; - if (!(sk->sk_state == DCCP_PARTOPEN || sk->sk_state == DCCP_OPEN)) - goto out_discard; - } + if (rc != 0) + goto out_discard; - if (sk->sk_state == DCCP_PARTOPEN) { - /* See 8.1.5. Handshake Completion */ - inet_csk_schedule_ack(sk); - inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK, inet_csk(sk)->icsk_rto, TCP_RTO_MAX); - dcb->dccpd_type = DCCP_PKT_DATAACK; - /* FIXME: we really should have a dccps_ack_pending or use icsk */ - } else if (inet_csk_ack_scheduled(sk) || - (dp->dccps_options.dccpo_send_ack_vector && - ap->dccpap_buf_ackno != DCCP_MAX_SEQNO + 1 && - ap->dccpap_ack_seqno == DCCP_MAX_SEQNO + 1)) - dcb->dccpd_type = DCCP_PKT_DATAACK; - else - dcb->dccpd_type = DCCP_PKT_DATA; - dccp_transmit_skb(sk, skb); - ccid_hc_tx_packet_sent(dp->dccps_hc_tx_ccid, sk, 0, len); - } else { -out_discard: - kfree_skb(skb); - } + rc = dccp_write_xmit(sk, skb, &timeo); + /* + * XXX we don't use sk_write_queue, so just discard the packet. + * Current plan however is to _use_ sk_write_queue with + * an algorith similar to tcp_sendmsg, where the main difference + * is that in DCCP we have to respect packet boundaries, so + * no coalescing of skbs. + * + * This bug was _quickly_ found & fixed by just looking at an OSTRA + * generated callgraph 8) -acme + */ + if (rc != 0) + goto out_discard; out_release: release_sock(sk); return rc ? : len; -out_err: - rc = sk_stream_error(sk, flags, rc); +out_discard: + kfree_skb(skb); goto out_release; -out_interrupted: - rc = sock_intr_errno(timeo); - goto out_discard; } -EXPORT_SYMBOL(dccp_sendmsg); - int dccp_recvmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg, size_t len, int nonblock, int flags, int *addr_len) { const struct dccp_hdr *dh; - int copied = 0; - unsigned long used; - int err; - int target; /* Read at least this many bytes */ long timeo; lock_sock(sk); - err = -ENOTCONN; - if (sk->sk_state == DCCP_LISTEN) + if (sk->sk_state == DCCP_LISTEN) { + len = -ENOTCONN; goto out; - - timeo = sock_rcvtimeo(sk, nonblock); - - /* Urgent data needs to be handled specially. */ - if (flags & MSG_OOB) - goto recv_urg; - - /* FIXME */ -#if 0 - seq = &tp->copied_seq; - if (flags & MSG_PEEK) { - peek_seq = tp->copied_seq; - seq = &peek_seq; } -#endif - target = sock_rcvlowat(sk, flags & MSG_WAITALL, len); + timeo = sock_rcvtimeo(sk, nonblock); do { - struct sk_buff *skb; - u32 offset; - - /* FIXME */ -#if 0 - /* Are we at urgent data? Stop if we have read anything or have SIGURG pending. */ - if (tp->urg_data && tp->urg_seq == *seq) { - if (copied) - break; - if (signal_pending(current)) { - copied = timeo ? sock_intr_errno(timeo) : -EAGAIN; - break; - } - } -#endif - - /* Next get a buffer. */ + struct sk_buff *skb = skb_peek(&sk->sk_receive_queue); - skb = skb_peek(&sk->sk_receive_queue); - do { - if (!skb) - break; - - offset = 0; - dh = dccp_hdr(skb); + if (skb == NULL) + goto verify_sock_status; - if (dh->dccph_type == DCCP_PKT_DATA || - dh->dccph_type == DCCP_PKT_DATAACK) - goto found_ok_skb; + dh = dccp_hdr(skb); - if (dh->dccph_type == DCCP_PKT_RESET || - dh->dccph_type == DCCP_PKT_CLOSE) { - dccp_pr_debug("found fin ok!\n"); - goto found_fin_ok; - } - dccp_pr_debug("packet_type=%s\n", dccp_packet_name(dh->dccph_type)); - BUG_TRAP(flags & MSG_PEEK); - skb = skb->next; - } while (skb != (struct sk_buff *)&sk->sk_receive_queue); + if (dh->dccph_type == DCCP_PKT_DATA || + dh->dccph_type == DCCP_PKT_DATAACK) + goto found_ok_skb; - /* Well, if we have backlog, try to process it now yet. */ - if (copied >= target && !sk->sk_backlog.tail) + if (dh->dccph_type == DCCP_PKT_RESET || + dh->dccph_type == DCCP_PKT_CLOSE) { + dccp_pr_debug("found fin ok!\n"); + len = 0; + goto found_fin_ok; + } + dccp_pr_debug("packet_type=%s\n", + dccp_packet_name(dh->dccph_type)); + sk_eat_skb(sk, skb); +verify_sock_status: + if (sock_flag(sk, SOCK_DONE)) { + len = 0; break; + } - if (copied) { - if (sk->sk_err || - sk->sk_state == DCCP_CLOSED || - (sk->sk_shutdown & RCV_SHUTDOWN) || - !timeo || - signal_pending(current) || - (flags & MSG_PEEK)) - break; - } else { - if (sock_flag(sk, SOCK_DONE)) - break; - - if (sk->sk_err) { - copied = sock_error(sk); - break; - } - - if (sk->sk_shutdown & RCV_SHUTDOWN) - break; - - if (sk->sk_state == DCCP_CLOSED) { - if (!sock_flag(sk, SOCK_DONE)) { - /* This occurs when user tries to read - * from never connected socket. - */ - copied = -ENOTCONN; - break; - } - break; - } + if (sk->sk_err) { + len = sock_error(sk); + break; + } - if (!timeo) { - copied = -EAGAIN; - break; - } + if (sk->sk_shutdown & RCV_SHUTDOWN) { + len = 0; + break; + } - if (signal_pending(current)) { - copied = sock_intr_errno(timeo); + if (sk->sk_state == DCCP_CLOSED) { + if (!sock_flag(sk, SOCK_DONE)) { + /* This occurs when user tries to read + * from never connected socket. + */ + len = -ENOTCONN; break; } + len = 0; + break; } - /* FIXME: cleanup_rbuf(sk, copied); */ + if (!timeo) { + len = -EAGAIN; + break; + } - if (copied >= target) { - /* Do not sleep, just process backlog. */ - release_sock(sk); - lock_sock(sk); - } else - sk_wait_data(sk, &timeo); + if (signal_pending(current)) { + len = sock_intr_errno(timeo); + break; + } + sk_wait_data(sk, &timeo); continue; - found_ok_skb: - /* Ok so how much can we use? */ - used = skb->len - offset; - if (len < used) - used = len; - - if (!(flags & MSG_TRUNC)) { - err = skb_copy_datagram_iovec(skb, offset, - msg->msg_iov, used); - if (err) { - /* Exception. Bailout! */ - if (!copied) - copied = -EFAULT; - break; - } + if (len > skb->len) + len = skb->len; + else if (len < skb->len) + msg->msg_flags |= MSG_TRUNC; + + if (skb_copy_datagram_iovec(skb, 0, msg->msg_iov, len)) { + /* Exception. Bailout! */ + len = -EFAULT; + break; } - - copied += used; - len -= used; - - /* FIXME: tcp_rcv_space_adjust(sk); */ - -//skip_copy: - if (used + offset < skb->len) - continue; - - if (!(flags & MSG_PEEK)) - sk_eat_skb(sk, skb); - continue; found_fin_ok: if (!(flags & MSG_PEEK)) sk_eat_skb(sk, skb); break; - - } while (len > 0); - - /* According to UNIX98, msg_name/msg_namelen are ignored - * on connected socket. I was just happy when found this 8) --ANK - */ - - /* Clean up data we have read: This will do ACK frames. */ - /* FIXME: cleanup_rbuf(sk, copied); */ - - release_sock(sk); - return copied; - + } while (1); out: release_sock(sk); - return err; - -recv_urg: - /* FIXME: err = tcp_recv_urg(sk, timeo, msg, len, flags, addr_len); */ - goto out; + return len; } static int inet_dccp_listen(struct socket *sock, int backlog) @@ -490,16 +460,16 @@ out: } static const unsigned char dccp_new_state[] = { - /* current state: new state: action: */ - [0] = DCCP_CLOSED, - [DCCP_OPEN] = DCCP_CLOSING | DCCP_ACTION_FIN, - [DCCP_REQUESTING] = DCCP_CLOSED, - [DCCP_PARTOPEN] = DCCP_CLOSING | DCCP_ACTION_FIN, - [DCCP_LISTEN] = DCCP_CLOSED, - [DCCP_RESPOND] = DCCP_CLOSED, - [DCCP_CLOSING] = DCCP_CLOSED, - [DCCP_TIME_WAIT] = DCCP_CLOSED, - [DCCP_CLOSED] = DCCP_CLOSED, + /* current state: new state: action: */ + [0] = DCCP_CLOSED, + [DCCP_OPEN] = DCCP_CLOSING | DCCP_ACTION_FIN, + [DCCP_REQUESTING] = DCCP_CLOSED, + [DCCP_PARTOPEN] = DCCP_CLOSING | DCCP_ACTION_FIN, + [DCCP_LISTEN] = DCCP_CLOSED, + [DCCP_RESPOND] = DCCP_CLOSED, + [DCCP_CLOSING] = DCCP_CLOSED, + [DCCP_TIME_WAIT] = DCCP_CLOSED, + [DCCP_CLOSED] = DCCP_CLOSED, }; static int dccp_close_state(struct sock *sk) @@ -544,12 +514,15 @@ void dccp_close(struct sock *sk, long timeout) /* Check zero linger _after_ checking for unread data. */ sk->sk_prot->disconnect(sk, 0); } else if (dccp_close_state(sk)) { - dccp_send_close(sk); + dccp_send_close(sk, 1); } sk_stream_wait_close(sk, timeout); adjudge_to_death: + /* + * It is the last release_sock in its life. It will remove backlog. + */ release_sock(sk); /* * Now socket is owned by kernel and we acquire BH lock @@ -561,11 +534,26 @@ adjudge_to_death: sock_hold(sk); sock_orphan(sk); - - if (sk->sk_state != DCCP_CLOSED) + + /* + * The last release_sock may have processed the CLOSE or RESET + * packet moving sock to CLOSED state, if not we have to fire + * the CLOSE/CLOSEREQ retransmission timer, see "8.3. Termination" + * in draft-ietf-dccp-spec-11. -acme + */ + if (sk->sk_state == DCCP_CLOSING) { + /* FIXME: should start at 2 * RTT */ + /* Timer for repeating the CLOSE/CLOSEREQ until an answer. */ + inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS, + inet_csk(sk)->icsk_rto, + DCCP_RTO_MAX); +#if 0 + /* Yeah, we should use sk->sk_prot->orphan_count, etc */ dccp_set_state(sk, DCCP_CLOSED); +#endif + } - atomic_inc(&dccp_orphan_count); + atomic_inc(sk->sk_prot->orphan_count); if (sk->sk_state == DCCP_CLOSED) inet_csk_destroy_sock(sk); @@ -581,7 +569,7 @@ void dccp_shutdown(struct sock *sk, int how) dccp_pr_debug("entry\n"); } -struct proto_ops inet_dccp_ops = { +static struct proto_ops inet_dccp_ops = { .family = PF_INET, .owner = THIS_MODULE, .release = inet_release, @@ -590,9 +578,11 @@ struct proto_ops inet_dccp_ops = { .socketpair = sock_no_socketpair, .accept = inet_accept, .getname = inet_getname, - .poll = sock_no_poll, + /* FIXME: work on tcp_poll to rename it to inet_csk_poll */ + .poll = dccp_poll, .ioctl = inet_ioctl, - .listen = inet_dccp_listen, /* FIXME: work on inet_listen to rename it to sock_common_listen */ + /* FIXME: work on inet_listen to rename it to sock_common_listen */ + .listen = inet_dccp_listen, .shutdown = inet_shutdown, .setsockopt = sock_common_setsockopt, .getsockopt = sock_common_getsockopt, @@ -644,12 +634,18 @@ static int __init dccp_ctl_sock_init(void) return rc; } -static void __exit dccp_ctl_sock_exit(void) +#ifdef CONFIG_IP_DCCP_UNLOAD_HACK +void dccp_ctl_sock_exit(void) { - if (dccp_ctl_socket != NULL) + if (dccp_ctl_socket != NULL) { sock_release(dccp_ctl_socket); + dccp_ctl_socket = NULL; + } } +EXPORT_SYMBOL_GPL(dccp_ctl_sock_exit); +#endif + static int __init init_dccp_v4_mibs(void) { int rc = -ENOMEM; @@ -676,9 +672,11 @@ static int thash_entries; module_param(thash_entries, int, 0444); MODULE_PARM_DESC(thash_entries, "Number of ehash buckets"); +#ifdef CONFIG_IP_DCCP_DEBUG int dccp_debug; module_param(dccp_debug, int, 0444); MODULE_PARM_DESC(dccp_debug, "Enable debug messages"); +#endif static int __init dccp_init(void) { @@ -689,10 +687,10 @@ static int __init dccp_init(void) if (rc) goto out; - dccp_hashinfo.bind_bucket_cachep = kmem_cache_create("dccp_bind_bucket", - sizeof(struct inet_bind_bucket), - 0, SLAB_HWCACHE_ALIGN, - NULL, NULL); + dccp_hashinfo.bind_bucket_cachep = + kmem_cache_create("dccp_bind_bucket", + sizeof(struct inet_bind_bucket), 0, + SLAB_HWCACHE_ALIGN, NULL, NULL); if (!dccp_hashinfo.bind_bucket_cachep) goto out_proto_unregister; @@ -708,14 +706,16 @@ static int __init dccp_init(void) goal = num_physpages >> (23 - PAGE_SHIFT); if (thash_entries) - goal = (thash_entries * sizeof(struct inet_ehash_bucket)) >> PAGE_SHIFT; + goal = (thash_entries * + sizeof(struct inet_ehash_bucket)) >> PAGE_SHIFT; for (ehash_order = 0; (1UL << ehash_order) < goal; ehash_order++) ; do { dccp_hashinfo.ehash_size = (1UL << ehash_order) * PAGE_SIZE / sizeof(struct inet_ehash_bucket); dccp_hashinfo.ehash_size >>= 1; - while (dccp_hashinfo.ehash_size & (dccp_hashinfo.ehash_size - 1)) + while (dccp_hashinfo.ehash_size & + (dccp_hashinfo.ehash_size - 1)) dccp_hashinfo.ehash_size--; dccp_hashinfo.ehash = (struct inet_ehash_bucket *) __get_free_pages(GFP_ATOMIC, ehash_order); @@ -737,7 +737,8 @@ static int __init dccp_init(void) do { dccp_hashinfo.bhash_size = (1UL << bhash_order) * PAGE_SIZE / sizeof(struct inet_bind_hashbucket); - if ((dccp_hashinfo.bhash_size > (64 * 1024)) && bhash_order > 0) + if ((dccp_hashinfo.bhash_size > (64 * 1024)) && + bhash_order > 0) continue; dccp_hashinfo.bhash = (struct inet_bind_hashbucket *) __get_free_pages(GFP_ATOMIC, bhash_order); @@ -793,26 +794,33 @@ static const char dccp_del_proto_err_msg[] __exitdata = static void __exit dccp_fini(void) { - dccp_ctl_sock_exit(); - inet_unregister_protosw(&dccp_v4_protosw); if (inet_del_protocol(&dccp_protocol, IPPROTO_DCCP) < 0) printk(dccp_del_proto_err_msg); - /* Free the control endpoint. */ - sock_release(dccp_ctl_socket); - - proto_unregister(&dccp_v4_prot); - + free_percpu(dccp_statistics[0]); + free_percpu(dccp_statistics[1]); + free_pages((unsigned long)dccp_hashinfo.bhash, + get_order(dccp_hashinfo.bhash_size * + sizeof(struct inet_bind_hashbucket))); + free_pages((unsigned long)dccp_hashinfo.ehash, + get_order(dccp_hashinfo.ehash_size * + sizeof(struct inet_ehash_bucket))); kmem_cache_destroy(dccp_hashinfo.bind_bucket_cachep); + proto_unregister(&dccp_v4_prot); } module_init(dccp_init); module_exit(dccp_fini); -/* __stringify doesn't likes enums, so use SOCK_DCCP (6) value directly */ -MODULE_ALIAS("net-pf-" __stringify(PF_INET) "-6"); +/* + * __stringify doesn't likes enums, so use SOCK_DCCP (6) and IPPROTO_DCCP (33) + * values directly, Also cover the case where the protocol is not specified, + * i.e. net-pf-PF_INET-proto-0-type-SOCK_DCCP + */ +MODULE_ALIAS("net-pf-" __stringify(PF_INET) "-proto-33-type-6"); +MODULE_ALIAS("net-pf-" __stringify(PF_INET) "-proto-0-type-6"); MODULE_LICENSE("GPL"); MODULE_AUTHOR("Arnaldo Carvalho de Melo "); MODULE_DESCRIPTION("DCCP - Datagram Congestion Controlled Protocol");