X-Git-Url: https://err.no/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=net%2Fipv4%2Ftcp.c;h=74998f250071c20c21549b06ae2cf20341bc50c0;hb=fc13dcae24bfd877aeff07774c035b2b896e05ca;hp=00aa80e9324323305fb72b365f75f96243390d1c;hpb=db9edfd7e339ca4113153d887e782dd05be5a9eb;p=linux-2.6 diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 00aa80e932..74998f2500 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -257,12 +257,13 @@ #include #include #include +#include #include #include #include #include - +#include #include #include @@ -275,9 +276,9 @@ atomic_t tcp_orphan_count = ATOMIC_INIT(0); EXPORT_SYMBOL_GPL(tcp_orphan_count); -int sysctl_tcp_mem[3]; -int sysctl_tcp_wmem[3] = { 4 * 1024, 16 * 1024, 128 * 1024 }; -int sysctl_tcp_rmem[3] = { 4 * 1024, 87380, 87380 * 2 }; +int sysctl_tcp_mem[3] __read_mostly; +int sysctl_tcp_wmem[3] __read_mostly; +int sysctl_tcp_rmem[3] __read_mostly; EXPORT_SYMBOL(sysctl_tcp_mem); EXPORT_SYMBOL(sysctl_tcp_rmem); @@ -365,7 +366,7 @@ unsigned int tcp_poll(struct file *file, struct socket *sock, poll_table *wait) if (sk->sk_shutdown == SHUTDOWN_MASK || sk->sk_state == TCP_CLOSE) mask |= POLLHUP; if (sk->sk_shutdown & RCV_SHUTDOWN) - mask |= POLLIN | POLLRDNORM; + mask |= POLLIN | POLLRDNORM | POLLRDHUP; /* Connected? */ if ((1 << sk->sk_state) & ~(TCPF_SYN_SENT | TCPF_SYN_RECV)) { @@ -621,14 +622,10 @@ ssize_t tcp_sendpage(struct socket *sock, struct page *page, int offset, ssize_t res; struct sock *sk = sock->sk; -#define TCP_ZC_CSUM_FLAGS (NETIF_F_IP_CSUM | NETIF_F_NO_CSUM | NETIF_F_HW_CSUM) - if (!(sk->sk_route_caps & NETIF_F_SG) || - !(sk->sk_route_caps & TCP_ZC_CSUM_FLAGS)) + !(sk->sk_route_caps & NETIF_F_ALL_CSUM)) return sock_no_sendpage(sock, page, offset, size, flags); -#undef TCP_ZC_CSUM_FLAGS - lock_sock(sk); TCP_CHECK_TIMER(sk); res = do_tcp_sendpages(sk, &page, offset, size, flags); @@ -725,9 +722,7 @@ new_segment: /* * Check whether we can use HW checksum. */ - if (sk->sk_route_caps & - (NETIF_F_IP_CSUM | NETIF_F_NO_CSUM | - NETIF_F_HW_CSUM)) + if (sk->sk_route_caps & NETIF_F_ALL_CSUM) skb->ip_summed = CHECKSUM_HW; skb_entail(sk, tp, skb); @@ -936,7 +931,7 @@ static int tcp_recv_urg(struct sock *sk, long timeo, * calculation of whether or not we must ACK for the sake of * a window update. */ -static void cleanup_rbuf(struct sock *sk, int copied) +void tcp_cleanup_rbuf(struct sock *sk, int copied) { struct tcp_sock *tp = tcp_sk(sk); int time_to_ack = 0; @@ -1071,11 +1066,11 @@ int tcp_read_sock(struct sock *sk, read_descriptor_t *desc, break; } if (skb->h.th->fin) { - sk_eat_skb(sk, skb); + sk_eat_skb(sk, skb, 0); ++seq; break; } - sk_eat_skb(sk, skb); + sk_eat_skb(sk, skb, 0); if (!desc->count) break; } @@ -1085,7 +1080,7 @@ int tcp_read_sock(struct sock *sk, read_descriptor_t *desc, /* Clean up data we have read: This will do ACK frames. */ if (copied) - cleanup_rbuf(sk, copied); + tcp_cleanup_rbuf(sk, copied); return copied; } @@ -1109,6 +1104,7 @@ int tcp_recvmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg, int target; /* Read at least this many bytes */ long timeo; struct task_struct *user_recv = NULL; + int copied_early = 0; lock_sock(sk); @@ -1132,6 +1128,17 @@ int tcp_recvmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg, target = sock_rcvlowat(sk, flags & MSG_WAITALL, len); +#ifdef CONFIG_NET_DMA + tp->ucopy.dma_chan = NULL; + preempt_disable(); + if ((len > sysctl_tcp_dma_copybreak) && !(flags & MSG_PEEK) && + !sysctl_tcp_low_latency && __get_cpu_var(softnet_data.net_dma)) { + preempt_enable_no_resched(); + tp->ucopy.pinned_list = dma_pin_iovec_pages(msg->msg_iov, len); + } else + preempt_enable_no_resched(); +#endif + do { struct sk_buff *skb; u32 offset; @@ -1219,7 +1226,7 @@ int tcp_recvmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg, } } - cleanup_rbuf(sk, copied); + tcp_cleanup_rbuf(sk, copied); if (!sysctl_tcp_low_latency && tp->ucopy.task == user_recv) { /* Install new reader */ @@ -1273,6 +1280,10 @@ int tcp_recvmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg, } else sk_wait_data(sk, &timeo); +#ifdef CONFIG_NET_DMA + tp->ucopy.wakeup = 0; +#endif + if (user_recv) { int chunk; @@ -1328,13 +1339,39 @@ do_prequeue: } if (!(flags & MSG_TRUNC)) { - err = skb_copy_datagram_iovec(skb, offset, - msg->msg_iov, used); - if (err) { - /* Exception. Bailout! */ - if (!copied) - copied = -EFAULT; - break; +#ifdef CONFIG_NET_DMA + if (!tp->ucopy.dma_chan && tp->ucopy.pinned_list) + tp->ucopy.dma_chan = get_softnet_dma(); + + if (tp->ucopy.dma_chan) { + tp->ucopy.dma_cookie = dma_skb_copy_datagram_iovec( + tp->ucopy.dma_chan, skb, offset, + msg->msg_iov, used, + tp->ucopy.pinned_list); + + if (tp->ucopy.dma_cookie < 0) { + + printk(KERN_ALERT "dma_cookie < 0\n"); + + /* Exception. Bailout! */ + if (!copied) + copied = -EFAULT; + break; + } + if ((offset + used) == skb->len) + copied_early = 1; + + } else +#endif + { + err = skb_copy_datagram_iovec(skb, offset, + msg->msg_iov, used); + if (err) { + /* Exception. Bailout! */ + if (!copied) + copied = -EFAULT; + break; + } } } @@ -1354,15 +1391,19 @@ skip_copy: if (skb->h.th->fin) goto found_fin_ok; - if (!(flags & MSG_PEEK)) - sk_eat_skb(sk, skb); + if (!(flags & MSG_PEEK)) { + sk_eat_skb(sk, skb, copied_early); + copied_early = 0; + } continue; found_fin_ok: /* Process the FIN. */ ++*seq; - if (!(flags & MSG_PEEK)) - sk_eat_skb(sk, skb); + if (!(flags & MSG_PEEK)) { + sk_eat_skb(sk, skb, copied_early); + copied_early = 0; + } break; } while (len > 0); @@ -1385,12 +1426,42 @@ skip_copy: tp->ucopy.len = 0; } +#ifdef CONFIG_NET_DMA + if (tp->ucopy.dma_chan) { + struct sk_buff *skb; + dma_cookie_t done, used; + + dma_async_memcpy_issue_pending(tp->ucopy.dma_chan); + + while (dma_async_memcpy_complete(tp->ucopy.dma_chan, + tp->ucopy.dma_cookie, &done, + &used) == DMA_IN_PROGRESS) { + /* do partial cleanup of sk_async_wait_queue */ + while ((skb = skb_peek(&sk->sk_async_wait_queue)) && + (dma_async_is_complete(skb->dma_cookie, done, + used) == DMA_SUCCESS)) { + __skb_dequeue(&sk->sk_async_wait_queue); + kfree_skb(skb); + } + } + + /* Safe to free early-copied skbs now */ + __skb_queue_purge(&sk->sk_async_wait_queue); + dma_chan_put(tp->ucopy.dma_chan); + tp->ucopy.dma_chan = NULL; + } + if (tp->ucopy.pinned_list) { + dma_unpin_iovec_pages(tp->ucopy.pinned_list); + tp->ucopy.pinned_list = NULL; + } +#endif + /* 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. */ - cleanup_rbuf(sk, copied); + tcp_cleanup_rbuf(sk, copied); TCP_CHECK_TIMER(sk); release_sock(sk); @@ -1467,6 +1538,7 @@ void tcp_close(struct sock *sk, long timeout) { struct sk_buff *skb; int data_was_unread = 0; + int state; lock_sock(sk); sk->sk_shutdown = SHUTDOWN_MASK; @@ -1543,6 +1615,11 @@ void tcp_close(struct sock *sk, long timeout) sk_stream_wait_close(sk, timeout); adjudge_to_death: + state = sk->sk_state; + sock_hold(sk); + sock_orphan(sk); + atomic_inc(sk->sk_prot->orphan_count); + /* It is the last release_sock in its life. It will remove backlog. */ release_sock(sk); @@ -1554,8 +1631,9 @@ adjudge_to_death: bh_lock_sock(sk); BUG_TRAP(!sock_owned_by_user(sk)); - sock_hold(sk); - sock_orphan(sk); + /* Have we already been destroyed by a softirq or backlog? */ + if (state != TCP_CLOSE && sk->sk_state == TCP_CLOSE) + goto out; /* This is a (useful) BSD violating of the RFC. There is a * problem with TCP as specified in that the other end could @@ -1583,7 +1661,6 @@ adjudge_to_death: if (tmo > TCP_TIMEWAIT_LEN) { inet_csk_reset_keepalive_timer(sk, tcp_fin_time(sk)); } else { - atomic_inc(sk->sk_prot->orphan_count); tcp_time_wait(sk, TCP_FIN_WAIT2, tmo); goto out; } @@ -1602,7 +1679,6 @@ adjudge_to_death: NET_INC_STATS_BH(LINUX_MIB_TCPABORTONMEMORY); } } - atomic_inc(sk->sk_prot->orphan_count); if (sk->sk_state == TCP_CLOSE) inet_csk_destroy_sock(sk); @@ -1652,6 +1728,9 @@ int tcp_disconnect(struct sock *sk, int flags) __skb_queue_purge(&sk->sk_receive_queue); sk_stream_writequeue_purge(sk); __skb_queue_purge(&tp->out_of_order_queue); +#ifdef CONFIG_NET_DMA + __skb_queue_purge(&sk->sk_async_wait_queue); +#endif inet->dport = 0; @@ -1687,18 +1766,14 @@ int tcp_disconnect(struct sock *sk, int flags) /* * Socket option code for TCP. */ -int tcp_setsockopt(struct sock *sk, int level, int optname, char __user *optval, - int optlen) +static int do_tcp_setsockopt(struct sock *sk, int level, + int optname, char __user *optval, int optlen) { struct tcp_sock *tp = tcp_sk(sk); struct inet_connection_sock *icsk = inet_csk(sk); int val; int err = 0; - if (level != SOL_TCP) - return icsk->icsk_af_ops->setsockopt(sk, level, optname, - optval, optlen); - /* This is a string value all the others are int's */ if (optname == TCP_CONGESTION) { char name[TCP_CA_NAME_MAX]; @@ -1856,7 +1931,7 @@ int tcp_setsockopt(struct sock *sk, int level, int optname, char __user *optval, (TCPF_ESTABLISHED | TCPF_CLOSE_WAIT) && inet_csk_ack_scheduled(sk)) { icsk->icsk_ack.pending |= ICSK_ACK_PUSHED; - cleanup_rbuf(sk, 1); + tcp_cleanup_rbuf(sk, 1); if (!(val & 1)) icsk->icsk_ack.pingpong = 1; } @@ -1871,6 +1946,30 @@ int tcp_setsockopt(struct sock *sk, int level, int optname, char __user *optval, return err; } +int tcp_setsockopt(struct sock *sk, int level, int optname, char __user *optval, + int optlen) +{ + struct inet_connection_sock *icsk = inet_csk(sk); + + if (level != SOL_TCP) + return icsk->icsk_af_ops->setsockopt(sk, level, optname, + optval, optlen); + return do_tcp_setsockopt(sk, level, optname, optval, optlen); +} + +#ifdef CONFIG_COMPAT +int compat_tcp_setsockopt(struct sock *sk, int level, int optname, + char __user *optval, int optlen) +{ + if (level != SOL_TCP) + return inet_csk_compat_setsockopt(sk, level, optname, + optval, optlen); + return do_tcp_setsockopt(sk, level, optname, optval, optlen); +} + +EXPORT_SYMBOL(compat_tcp_setsockopt); +#endif + /* Return information about state of tcp endpoint in API format. */ void tcp_get_info(struct sock *sk, struct tcp_info *info) { @@ -1931,17 +2030,13 @@ void tcp_get_info(struct sock *sk, struct tcp_info *info) EXPORT_SYMBOL_GPL(tcp_get_info); -int tcp_getsockopt(struct sock *sk, int level, int optname, char __user *optval, - int __user *optlen) +static int do_tcp_getsockopt(struct sock *sk, int level, + int optname, char __user *optval, int __user *optlen) { struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); int val, len; - if (level != SOL_TCP) - return icsk->icsk_af_ops->getsockopt(sk, level, optname, - optval, optlen); - if (get_user(len, optlen)) return -EFAULT; @@ -2025,6 +2120,29 @@ int tcp_getsockopt(struct sock *sk, int level, int optname, char __user *optval, return 0; } +int tcp_getsockopt(struct sock *sk, int level, int optname, char __user *optval, + int __user *optlen) +{ + struct inet_connection_sock *icsk = inet_csk(sk); + + if (level != SOL_TCP) + return icsk->icsk_af_ops->getsockopt(sk, level, optname, + optval, optlen); + return do_tcp_getsockopt(sk, level, optname, optval, optlen); +} + +#ifdef CONFIG_COMPAT +int compat_tcp_getsockopt(struct sock *sk, int level, int optname, + char __user *optval, int __user *optlen) +{ + if (level != SOL_TCP) + return inet_csk_compat_getsockopt(sk, level, optname, + optval, optlen); + return do_tcp_getsockopt(sk, level, optname, optval, optlen); +} + +EXPORT_SYMBOL(compat_tcp_getsockopt); +#endif extern void __skb_cb_too_small_for_tcp(int, int); extern struct tcp_congestion_ops tcp_reno; @@ -2042,7 +2160,8 @@ __setup("thash_entries=", set_thash_entries); void __init tcp_init(void) { struct sk_buff *skb = NULL; - int order, i; + unsigned long limit; + int order, i, max_share; if (sizeof(struct tcp_skb_cb) > sizeof(skb->cb)) __skb_cb_too_small_for_tcp(sizeof(struct tcp_skb_cb), @@ -2116,12 +2235,16 @@ void __init tcp_init(void) sysctl_tcp_mem[1] = 1024 << order; sysctl_tcp_mem[2] = 1536 << order; - if (order < 3) { - sysctl_tcp_wmem[2] = 64 * 1024; - sysctl_tcp_rmem[0] = PAGE_SIZE; - sysctl_tcp_rmem[1] = 43689; - sysctl_tcp_rmem[2] = 2 * 43689; - } + limit = ((unsigned long)sysctl_tcp_mem[1]) << (PAGE_SHIFT - 7); + max_share = min(4UL*1024*1024, limit); + + sysctl_tcp_wmem[0] = SK_STREAM_MEM_QUANTUM; + sysctl_tcp_wmem[1] = 16*1024; + sysctl_tcp_wmem[2] = max(64*1024, max_share); + + sysctl_tcp_rmem[0] = SK_STREAM_MEM_QUANTUM; + sysctl_tcp_rmem[1] = 87380; + sysctl_tcp_rmem[2] = max(87380, max_share); printk(KERN_INFO "TCP: Hash tables configured " "(established %d bind %d)\n",