From c1207c049b204b0a96535dc5416aee331b51e0e1 Mon Sep 17 00:00:00 2001 From: Guenter Roeck Date: Sun, 2 Nov 2014 18:19:15 -0800 Subject: netfilter: nft_reject_bridge: Fix powerpc build error Fix: net/bridge/netfilter/nft_reject_bridge.c: In function 'nft_reject_br_send_v6_unreach': net/bridge/netfilter/nft_reject_bridge.c:240:3: error: implicit declaration of function 'csum_ipv6_magic' csum_ipv6_magic(&nip6h->saddr, &nip6h->daddr, ^ make[3]: *** [net/bridge/netfilter/nft_reject_bridge.o] Error 1 Seen with powerpc:allmodconfig. Fixes: 523b929d5446 ("netfilter: nft_reject_bridge: don't use IP stack to reject traffic") Cc: Pablo Neira Ayuso Signed-off-by: Guenter Roeck Signed-off-by: David S. Miller --- net/bridge/netfilter/nft_reject_bridge.c | 1 + 1 file changed, 1 insertion(+) (limited to 'net') diff --git a/net/bridge/netfilter/nft_reject_bridge.c b/net/bridge/netfilter/nft_reject_bridge.c index 654c9018e3e7..48da2c54a69e 100644 --- a/net/bridge/netfilter/nft_reject_bridge.c +++ b/net/bridge/netfilter/nft_reject_bridge.c @@ -18,6 +18,7 @@ #include #include #include +#include #include #include "../br_private.h" -- cgit v1.2.3-70-g09d2 From 6c6151daaf2d8dc2046d9926539feed5f66bf74e Mon Sep 17 00:00:00 2001 From: Steffen Klassert Date: Mon, 3 Nov 2014 09:19:27 +0100 Subject: ip6_tunnel: Use ip6_tnl_dev_init as the ndo_init function. ip6_tnl_dev_init() sets the dev->iflink via a call to ip6_tnl_link_config(). After that, register_netdevice() sets dev->iflink = -1. So we loose the iflink configuration for ipv6 tunnels. Fix this by using ip6_tnl_dev_init() as the ndo_init function. Then ip6_tnl_dev_init() is called after dev->iflink is set to -1 from register_netdevice(). Signed-off-by: Steffen Klassert Signed-off-by: David S. Miller --- net/ipv6/ip6_tunnel.c | 10 +--------- 1 file changed, 1 insertion(+), 9 deletions(-) (limited to 'net') diff --git a/net/ipv6/ip6_tunnel.c b/net/ipv6/ip6_tunnel.c index 9409887fb664..9cb94cfa0ae7 100644 --- a/net/ipv6/ip6_tunnel.c +++ b/net/ipv6/ip6_tunnel.c @@ -272,9 +272,6 @@ static int ip6_tnl_create2(struct net_device *dev) int err; t = netdev_priv(dev); - err = ip6_tnl_dev_init(dev); - if (err < 0) - goto out; err = register_netdevice(dev); if (err < 0) @@ -1462,6 +1459,7 @@ ip6_tnl_change_mtu(struct net_device *dev, int new_mtu) static const struct net_device_ops ip6_tnl_netdev_ops = { + .ndo_init = ip6_tnl_dev_init, .ndo_uninit = ip6_tnl_dev_uninit, .ndo_start_xmit = ip6_tnl_xmit, .ndo_do_ioctl = ip6_tnl_ioctl, @@ -1546,16 +1544,10 @@ static int __net_init ip6_fb_tnl_dev_init(struct net_device *dev) struct ip6_tnl *t = netdev_priv(dev); struct net *net = dev_net(dev); struct ip6_tnl_net *ip6n = net_generic(net, ip6_tnl_net_id); - int err = ip6_tnl_dev_init_gen(dev); - - if (err) - return err; t->parms.proto = IPPROTO_IPV6; dev_hold(dev); - ip6_tnl_link_config(t); - rcu_assign_pointer(ip6n->tnls_wc[0], t); return 0; } -- cgit v1.2.3-70-g09d2 From 16a0231bf7dc3fb37e9b1f1cb1a277dc220b5c5e Mon Sep 17 00:00:00 2001 From: Steffen Klassert Date: Mon, 3 Nov 2014 09:19:28 +0100 Subject: vti6: Use vti6_dev_init as the ndo_init function. vti6_dev_init() sets the dev->iflink via a call to vti6_link_config(). After that, register_netdevice() sets dev->iflink = -1. So we loose the iflink configuration for vti6 tunnels. Fix this by using vti6_dev_init() as the ndo_init function. Then vti6_dev_init() is called after dev->iflink is set to -1 from register_netdevice(). Signed-off-by: Steffen Klassert Signed-off-by: David S. Miller --- net/ipv6/ip6_vti.c | 11 +---------- 1 file changed, 1 insertion(+), 10 deletions(-) (limited to 'net') diff --git a/net/ipv6/ip6_vti.c b/net/ipv6/ip6_vti.c index d440bb585524..31089d153fd3 100644 --- a/net/ipv6/ip6_vti.c +++ b/net/ipv6/ip6_vti.c @@ -172,10 +172,6 @@ static int vti6_tnl_create2(struct net_device *dev) struct vti6_net *ip6n = net_generic(net, vti6_net_id); int err; - err = vti6_dev_init(dev); - if (err < 0) - goto out; - err = register_netdevice(dev); if (err < 0) goto out; @@ -783,6 +779,7 @@ static int vti6_change_mtu(struct net_device *dev, int new_mtu) } static const struct net_device_ops vti6_netdev_ops = { + .ndo_init = vti6_dev_init, .ndo_uninit = vti6_dev_uninit, .ndo_start_xmit = vti6_tnl_xmit, .ndo_do_ioctl = vti6_ioctl, @@ -852,16 +849,10 @@ static int __net_init vti6_fb_tnl_dev_init(struct net_device *dev) struct ip6_tnl *t = netdev_priv(dev); struct net *net = dev_net(dev); struct vti6_net *ip6n = net_generic(net, vti6_net_id); - int err = vti6_dev_init_gen(dev); - - if (err) - return err; t->parms.proto = IPPROTO_IPV6; dev_hold(dev); - vti6_link_config(t); - rcu_assign_pointer(ip6n->tnls_wc[0], t); return 0; } -- cgit v1.2.3-70-g09d2 From ebe084aafb7e93adf210e80043c9f69adf56820d Mon Sep 17 00:00:00 2001 From: Steffen Klassert Date: Mon, 3 Nov 2014 09:19:29 +0100 Subject: sit: Use ipip6_tunnel_init as the ndo_init function. ipip6_tunnel_init() sets the dev->iflink via a call to ipip6_tunnel_bind_dev(). After that, register_netdevice() sets dev->iflink = -1. So we loose the iflink configuration for ipv6 tunnels. Fix this by using ipip6_tunnel_init() as the ndo_init function. Then ipip6_tunnel_init() is called after dev->iflink is set to -1 from register_netdevice(). Signed-off-by: Steffen Klassert Signed-off-by: David S. Miller --- net/ipv6/sit.c | 15 ++++++--------- 1 file changed, 6 insertions(+), 9 deletions(-) (limited to 'net') diff --git a/net/ipv6/sit.c b/net/ipv6/sit.c index 58e5b4710127..a24557a1c1d8 100644 --- a/net/ipv6/sit.c +++ b/net/ipv6/sit.c @@ -195,10 +195,8 @@ static int ipip6_tunnel_create(struct net_device *dev) struct sit_net *sitn = net_generic(net, sit_net_id); int err; - err = ipip6_tunnel_init(dev); - if (err < 0) - goto out; - ipip6_tunnel_clone_6rd(dev, sitn); + memcpy(dev->dev_addr, &t->parms.iph.saddr, 4); + memcpy(dev->broadcast, &t->parms.iph.daddr, 4); if ((__force u16)t->parms.i_flags & SIT_ISATAP) dev->priv_flags |= IFF_ISATAP; @@ -207,7 +205,8 @@ static int ipip6_tunnel_create(struct net_device *dev) if (err < 0) goto out; - strcpy(t->parms.name, dev->name); + ipip6_tunnel_clone_6rd(dev, sitn); + dev->rtnl_link_ops = &sit_link_ops; dev_hold(dev); @@ -1330,6 +1329,7 @@ static int ipip6_tunnel_change_mtu(struct net_device *dev, int new_mtu) } static const struct net_device_ops ipip6_netdev_ops = { + .ndo_init = ipip6_tunnel_init, .ndo_uninit = ipip6_tunnel_uninit, .ndo_start_xmit = sit_tunnel_xmit, .ndo_do_ioctl = ipip6_tunnel_ioctl, @@ -1378,9 +1378,7 @@ static int ipip6_tunnel_init(struct net_device *dev) tunnel->dev = dev; tunnel->net = dev_net(dev); - - memcpy(dev->dev_addr, &tunnel->parms.iph.saddr, 4); - memcpy(dev->broadcast, &tunnel->parms.iph.daddr, 4); + strcpy(tunnel->parms.name, dev->name); ipip6_tunnel_bind_dev(dev); dev->tstats = netdev_alloc_pcpu_stats(struct pcpu_sw_netstats); @@ -1405,7 +1403,6 @@ static int __net_init ipip6_fb_tunnel_init(struct net_device *dev) tunnel->dev = dev; tunnel->net = dev_net(dev); - strcpy(tunnel->parms.name, dev->name); iph->version = 4; iph->protocol = IPPROTO_IPV6; -- cgit v1.2.3-70-g09d2 From f03eb128e3f4276f46442d14f3b8f864f3775821 Mon Sep 17 00:00:00 2001 From: Steffen Klassert Date: Mon, 3 Nov 2014 09:19:30 +0100 Subject: gre6: Move the setting of dev->iflink into the ndo_init functions. Otherwise it gets overwritten by register_netdev(). Signed-off-by: Steffen Klassert Signed-off-by: David S. Miller --- net/ipv6/ip6_gre.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'net') diff --git a/net/ipv6/ip6_gre.c b/net/ipv6/ip6_gre.c index 12c3c8ef3849..4564e1fca3eb 100644 --- a/net/ipv6/ip6_gre.c +++ b/net/ipv6/ip6_gre.c @@ -961,8 +961,6 @@ static void ip6gre_tnl_link_config(struct ip6_tnl *t, int set_mtu) else dev->flags &= ~IFF_POINTOPOINT; - dev->iflink = p->link; - /* Precalculate GRE options length */ if (t->parms.o_flags&(GRE_CSUM|GRE_KEY|GRE_SEQ)) { if (t->parms.o_flags&GRE_CSUM) @@ -1272,6 +1270,7 @@ static int ip6gre_tunnel_init(struct net_device *dev) u64_stats_init(&ip6gre_tunnel_stats->syncp); } + dev->iflink = tunnel->parms.link; return 0; } @@ -1481,6 +1480,8 @@ static int ip6gre_tap_init(struct net_device *dev) if (!dev->tstats) return -ENOMEM; + dev->iflink = tunnel->parms.link; + return 0; } -- cgit v1.2.3-70-g09d2 From 45cac46e51da75628ac2a593c70f5144abb9b31d Mon Sep 17 00:00:00 2001 From: Jesse Gross Date: Mon, 3 Nov 2014 19:38:37 -0800 Subject: geneve: Set GSO type on transmit. Geneve does not currently set the inner protocol type when transmitting packets. This causes GSO segmentation to fail on NICs that do not support Geneve offloading. CC: Andy Zhou Signed-off-by: Jesse Gross Signed-off-by: David S. Miller --- net/ipv4/geneve.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'net') diff --git a/net/ipv4/geneve.c b/net/ipv4/geneve.c index 065cd94c640c..6e5266cf403d 100644 --- a/net/ipv4/geneve.c +++ b/net/ipv4/geneve.c @@ -144,6 +144,8 @@ int geneve_xmit_skb(struct geneve_sock *gs, struct rtable *rt, gnvh = (struct genevehdr *)__skb_push(skb, sizeof(*gnvh) + opt_len); geneve_build_header(gnvh, tun_flags, vni, opt_len, opt); + skb_set_inner_protocol(skb, htons(ETH_P_TEB)); + return udp_tunnel_xmit_skb(gs->sock, rt, skb, src, dst, tos, ttl, df, src_port, dst_port, xnet); } -- cgit v1.2.3-70-g09d2 From d3ca9eafc0ed97b8f56fdf23655cfece89c48354 Mon Sep 17 00:00:00 2001 From: Jesse Gross Date: Mon, 3 Nov 2014 19:38:38 -0800 Subject: geneve: Unregister pernet subsys on module unload. The pernet ops aren't ever unregistered, which causes a memory leak and an OOPs if the module is ever reinserted. Fixes: 0b5e8b8eeae4 ("net: Add Geneve tunneling protocol driver") CC: Andy Zhou Signed-off-by: Jesse Gross Acked-by: Nicolas Dichtel Signed-off-by: David S. Miller --- net/ipv4/geneve.c | 1 + 1 file changed, 1 insertion(+) (limited to 'net') diff --git a/net/ipv4/geneve.c b/net/ipv4/geneve.c index 6e5266cf403d..dedb21e99914 100644 --- a/net/ipv4/geneve.c +++ b/net/ipv4/geneve.c @@ -366,6 +366,7 @@ late_initcall(geneve_init_module); static void __exit geneve_cleanup_module(void) { destroy_workqueue(geneve_wq); + unregister_pernet_subsys(&geneve_net_ops); } module_exit(geneve_cleanup_module); -- cgit v1.2.3-70-g09d2 From 1f37bf87aa7523d28e7e4c4f7bb5dba98faa3e00 Mon Sep 17 00:00:00 2001 From: Marcelo Leitner Date: Tue, 4 Nov 2014 17:15:08 -0200 Subject: tcp: zero retrans_stamp if all retrans were acked Ueki Kohei reported that when we are using NewReno with connections that have a very low traffic, we may timeout the connection too early if a second loss occurs after the first one was successfully acked but no data was transfered later. Below is his description of it: When SACK is disabled, and a socket suffers multiple separate TCP retransmissions, that socket's ETIMEDOUT value is calculated from the time of the *first* retransmission instead of the *latest* retransmission. This happens because the tcp_sock's retrans_stamp is set once then never cleared. Take the following connection: Linux remote-machine | | send#1---->(*1)|--------> data#1 --------->| | | | RTO : : | | | ---(*2)|----> data#1(retrans) ---->| | (*3)|<---------- ACK <----------| | | | | : : | : : | : : 16 minutes (or more) : | : : | : : | : : | | | send#2---->(*4)|--------> data#2 --------->| | | | RTO : : | | | ---(*5)|----> data#2(retrans) ---->| | | | | | | RTO*2 : : | | | | | | ETIMEDOUT<----(*6)| | (*1) One data packet sent. (*2) Because no ACK packet is received, the packet is retransmitted. (*3) The ACK packet is received. The transmitted packet is acknowledged. At this point the first "retransmission event" has passed and been recovered from. Any future retransmission is a completely new "event". (*4) After 16 minutes (to correspond with retries2=15), a new data packet is sent. Note: No data is transmitted between (*3) and (*4). The socket's timeout SHOULD be calculated from this point in time, but instead it's calculated from the prior "event" 16 minutes ago. (*5) Because no ACK packet is received, the packet is retransmitted. (*6) At the time of the 2nd retransmission, the socket returns ETIMEDOUT. Therefore, now we clear retrans_stamp as soon as all data during the loss window is fully acked. Reported-by: Ueki Kohei Cc: Neal Cardwell Cc: Yuchung Cheng Signed-off-by: Marcelo Ricardo Leitner Acked-by: Neal Cardwell Tested-by: Neal Cardwell Signed-off-by: David S. Miller --- net/ipv4/tcp_input.c | 60 +++++++++++++++++++++++++++------------------------- 1 file changed, 31 insertions(+), 29 deletions(-) (limited to 'net') diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index a12b455928e5..88fa2d160685 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -2315,6 +2315,35 @@ static inline bool tcp_packet_delayed(const struct tcp_sock *tp) /* Undo procedures. */ +/* We can clear retrans_stamp when there are no retransmissions in the + * window. It would seem that it is trivially available for us in + * tp->retrans_out, however, that kind of assumptions doesn't consider + * what will happen if errors occur when sending retransmission for the + * second time. ...It could the that such segment has only + * TCPCB_EVER_RETRANS set at the present time. It seems that checking + * the head skb is enough except for some reneging corner cases that + * are not worth the effort. + * + * Main reason for all this complexity is the fact that connection dying + * time now depends on the validity of the retrans_stamp, in particular, + * that successive retransmissions of a segment must not advance + * retrans_stamp under any conditions. + */ +static bool tcp_any_retrans_done(const struct sock *sk) +{ + const struct tcp_sock *tp = tcp_sk(sk); + struct sk_buff *skb; + + if (tp->retrans_out) + return true; + + skb = tcp_write_queue_head(sk); + if (unlikely(skb && TCP_SKB_CB(skb)->sacked & TCPCB_EVER_RETRANS)) + return true; + + return false; +} + #if FASTRETRANS_DEBUG > 1 static void DBGUNDO(struct sock *sk, const char *msg) { @@ -2410,6 +2439,8 @@ static bool tcp_try_undo_recovery(struct sock *sk) * is ACKed. For Reno it is MUST to prevent false * fast retransmits (RFC2582). SACK TCP is safe. */ tcp_moderate_cwnd(tp); + if (!tcp_any_retrans_done(sk)) + tp->retrans_stamp = 0; return true; } tcp_set_ca_state(sk, TCP_CA_Open); @@ -2430,35 +2461,6 @@ static bool tcp_try_undo_dsack(struct sock *sk) return false; } -/* We can clear retrans_stamp when there are no retransmissions in the - * window. It would seem that it is trivially available for us in - * tp->retrans_out, however, that kind of assumptions doesn't consider - * what will happen if errors occur when sending retransmission for the - * second time. ...It could the that such segment has only - * TCPCB_EVER_RETRANS set at the present time. It seems that checking - * the head skb is enough except for some reneging corner cases that - * are not worth the effort. - * - * Main reason for all this complexity is the fact that connection dying - * time now depends on the validity of the retrans_stamp, in particular, - * that successive retransmissions of a segment must not advance - * retrans_stamp under any conditions. - */ -static bool tcp_any_retrans_done(const struct sock *sk) -{ - const struct tcp_sock *tp = tcp_sk(sk); - struct sk_buff *skb; - - if (tp->retrans_out) - return true; - - skb = tcp_write_queue_head(sk); - if (unlikely(skb && TCP_SKB_CB(skb)->sacked & TCPCB_EVER_RETRANS)) - return true; - - return false; -} - /* Undo during loss recovery after partial ACK or using F-RTO. */ static bool tcp_try_undo_loss(struct sock *sk, bool frto_undo) { -- cgit v1.2.3-70-g09d2 From b31f65fb4383a49bdcfa465176754b37e44e1e17 Mon Sep 17 00:00:00 2001 From: Andrew Lunn Date: Wed, 5 Nov 2014 19:47:28 +0100 Subject: net: dsa: slave: Fix autoneg for phys on switch MDIO bus When the ports phys are connected to the switches internal MDIO bus, we need to connect the phy to the slave netdev, otherwise auto-negotiation etc, does not work. Signed-off-by: Andrew Lunn Signed-off-by: David S. Miller --- net/dsa/slave.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) (limited to 'net') diff --git a/net/dsa/slave.c b/net/dsa/slave.c index 6d1817449c36..ab03e00ffe8f 100644 --- a/net/dsa/slave.c +++ b/net/dsa/slave.c @@ -489,11 +489,14 @@ static void dsa_slave_phy_setup(struct dsa_slave_priv *p, /* We could not connect to a designated PHY, so use the switch internal * MDIO bus instead */ - if (!p->phy) + if (!p->phy) { p->phy = ds->slave_mii_bus->phy_map[p->port]; - else + phy_connect_direct(slave_dev, p->phy, dsa_slave_adjust_link, + p->phy_interface); + } else { pr_info("attached PHY at address %d [%s]\n", p->phy->addr, p->phy->drv->name); + } } int dsa_slave_suspend(struct net_device *slave_dev) -- cgit v1.2.3-70-g09d2