diff --git a/Documentation/networking/scaling.txt b/Documentation/networking/scaling.txt index 8ce7c30e7230..fe67b5c79f0f 100644 --- a/Documentation/networking/scaling.txt +++ b/Documentation/networking/scaling.txt @@ -27,7 +27,7 @@ applying a filter to each packet that assigns it to one of a small number of logical flows. Packets for each flow are steered to a separate receive queue, which in turn can be processed by separate CPUs. This mechanism is generally known as “Receive-side Scaling” (RSS). The goal of RSS and -the other scaling techniques to increase performance uniformly. +the other scaling techniques is to increase performance uniformly. Multi-queue distribution can also be used for traffic prioritization, but that is not the focus of these techniques. @@ -186,10 +186,10 @@ are steered using plain RPS. Multiple table entries may point to the same CPU. Indeed, with many flows and few CPUs, it is very likely that a single application thread handles flows with many different flow hashes. -rps_sock_table is a global flow table that contains the *desired* CPU for -flows: the CPU that is currently processing the flow in userspace. Each -table value is a CPU index that is updated during calls to recvmsg and -sendmsg (specifically, inet_recvmsg(), inet_sendmsg(), inet_sendpage() +rps_sock_flow_table is a global flow table that contains the *desired* CPU +for flows: the CPU that is currently processing the flow in userspace. +Each table value is a CPU index that is updated during calls to recvmsg +and sendmsg (specifically, inet_recvmsg(), inet_sendmsg(), inet_sendpage() and tcp_splice_read()). When the scheduler moves a thread to a new CPU while it has outstanding diff --git a/drivers/net/macvlan.c b/drivers/net/macvlan.c index 05172c39a0ce..376e3e94bae0 100644 --- a/drivers/net/macvlan.c +++ b/drivers/net/macvlan.c @@ -239,7 +239,7 @@ static int macvlan_queue_xmit(struct sk_buff *skb, struct net_device *dev) dest = macvlan_hash_lookup(port, eth->h_dest); if (dest && dest->mode == MACVLAN_MODE_BRIDGE) { /* send to lowerdev first for its network taps */ - vlan->forward(vlan->lowerdev, skb); + dev_forward_skb(vlan->lowerdev, skb); return NET_XMIT_SUCCESS; } diff --git a/net/bridge/br_device.c b/net/bridge/br_device.c index 32b8f9f7f79e..ff3ed6086ce1 100644 --- a/net/bridge/br_device.c +++ b/net/bridge/br_device.c @@ -91,7 +91,6 @@ static int br_dev_open(struct net_device *dev) { struct net_bridge *br = netdev_priv(dev); - netif_carrier_off(dev); netdev_update_features(dev); netif_start_queue(dev); br_stp_enable_bridge(br); @@ -108,8 +107,6 @@ static int br_dev_stop(struct net_device *dev) { struct net_bridge *br = netdev_priv(dev); - netif_carrier_off(dev); - br_stp_disable_bridge(br); br_multicast_stop(br); diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 21fab3edb92c..d73aab3fbfc0 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -1389,9 +1389,7 @@ static int tcp_shifted_skb(struct sock *sk, struct sk_buff *skb, BUG_ON(!pcount); - /* Tweak before seqno plays */ - if (!tcp_is_fack(tp) && tcp_is_sack(tp) && tp->lost_skb_hint && - !before(TCP_SKB_CB(tp->lost_skb_hint)->seq, TCP_SKB_CB(skb)->seq)) + if (skb == tp->lost_skb_hint) tp->lost_cnt_hint += pcount; TCP_SKB_CB(prev)->end_seq += shifted; diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index c34f01513945..7963e03f1068 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -927,18 +927,21 @@ int tcp_v4_md5_do_add(struct sock *sk, __be32 addr, } sk_nocaps_add(sk, NETIF_F_GSO_MASK); } - if (tcp_alloc_md5sig_pool(sk) == NULL) { + + md5sig = tp->md5sig_info; + if (md5sig->entries4 == 0 && + tcp_alloc_md5sig_pool(sk) == NULL) { kfree(newkey); return -ENOMEM; } - md5sig = tp->md5sig_info; if (md5sig->alloced4 == md5sig->entries4) { keys = kmalloc((sizeof(*keys) * (md5sig->entries4 + 1)), GFP_ATOMIC); if (!keys) { kfree(newkey); - tcp_free_md5sig_pool(); + if (md5sig->entries4 == 0) + tcp_free_md5sig_pool(); return -ENOMEM; } @@ -982,6 +985,7 @@ int tcp_v4_md5_do_del(struct sock *sk, __be32 addr) kfree(tp->md5sig_info->keys4); tp->md5sig_info->keys4 = NULL; tp->md5sig_info->alloced4 = 0; + tcp_free_md5sig_pool(); } else if (tp->md5sig_info->entries4 != i) { /* Need to do some manipulation */ memmove(&tp->md5sig_info->keys4[i], @@ -989,7 +993,6 @@ int tcp_v4_md5_do_del(struct sock *sk, __be32 addr) (tp->md5sig_info->entries4 - i) * sizeof(struct tcp4_md5sig_key)); } - tcp_free_md5sig_pool(); return 0; } } diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 79cc6469508d..7b8fc5794352 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -591,7 +591,8 @@ static int tcp_v6_md5_do_add(struct sock *sk, const struct in6_addr *peer, } sk_nocaps_add(sk, NETIF_F_GSO_MASK); } - if (tcp_alloc_md5sig_pool(sk) == NULL) { + if (tp->md5sig_info->entries6 == 0 && + tcp_alloc_md5sig_pool(sk) == NULL) { kfree(newkey); return -ENOMEM; } @@ -600,8 +601,9 @@ static int tcp_v6_md5_do_add(struct sock *sk, const struct in6_addr *peer, (tp->md5sig_info->entries6 + 1)), GFP_ATOMIC); if (!keys) { - tcp_free_md5sig_pool(); kfree(newkey); + if (tp->md5sig_info->entries6 == 0) + tcp_free_md5sig_pool(); return -ENOMEM; } @@ -647,6 +649,7 @@ static int tcp_v6_md5_do_del(struct sock *sk, const struct in6_addr *peer) kfree(tp->md5sig_info->keys6); tp->md5sig_info->keys6 = NULL; tp->md5sig_info->alloced6 = 0; + tcp_free_md5sig_pool(); } else { /* shrink the database */ if (tp->md5sig_info->entries6 != i) @@ -655,7 +658,6 @@ static int tcp_v6_md5_do_del(struct sock *sk, const struct in6_addr *peer) (tp->md5sig_info->entries6 - i) * sizeof (tp->md5sig_info->keys6[0])); } - tcp_free_md5sig_pool(); return 0; } } diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c index 2b771dc708a3..5290ac353a5e 100644 --- a/net/netfilter/ipvs/ip_vs_ctl.c +++ b/net/netfilter/ipvs/ip_vs_ctl.c @@ -3679,7 +3679,7 @@ int __net_init ip_vs_control_net_init(struct net *net) int idx; struct netns_ipvs *ipvs = net_ipvs(net); - ipvs->rs_lock = __RW_LOCK_UNLOCKED(ipvs->rs_lock); + rwlock_init(&ipvs->rs_lock); /* Initialize rs_table */ for (idx = 0; idx < IP_VS_RTAB_SIZE; idx++)