aboutsummaryrefslogtreecommitdiffstats
path: root/net/ipv4
diff options
context:
space:
mode:
Diffstat (limited to 'net/ipv4')
-rw-r--r--net/ipv4/af_inet.c8
-rw-r--r--net/ipv4/devinet.c21
-rw-r--r--net/ipv4/ip_output.c12
-rw-r--r--net/ipv4/ip_tunnel.c43
-rw-r--r--net/ipv4/tcp.c32
-rw-r--r--net/ipv4/tcp_input.c4
-rw-r--r--net/ipv4/udp.c4
7 files changed, 87 insertions, 37 deletions
diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c
index a3f77ac173b5..d7ebee3c048d 100644
--- a/net/ipv4/af_inet.c
+++ b/net/ipv4/af_inet.c
@@ -876,7 +876,7 @@ int inet_shutdown(struct socket *sock, int how)
EPOLLHUP, even on eg. unconnected UDP sockets -- RR */
/* fall through */
default:
- sk->sk_shutdown |= how;
+ WRITE_ONCE(sk->sk_shutdown, sk->sk_shutdown | how);
if (sk->sk_prot->shutdown)
sk->sk_prot->shutdown(sk, how);
break;
@@ -1572,10 +1572,12 @@ EXPORT_SYMBOL(inet_current_timestamp);
int inet_recv_error(struct sock *sk, struct msghdr *msg, int len, int *addr_len)
{
- if (sk->sk_family == AF_INET)
+ unsigned int family = READ_ONCE(sk->sk_family);
+
+ if (family == AF_INET)
return ip_recv_error(sk, msg, len, addr_len);
#if IS_ENABLED(CONFIG_IPV6)
- if (sk->sk_family == AF_INET6)
+ if (family == AF_INET6)
return pingv6_ops.ipv6_recv_error(sk, msg, len, addr_len);
#endif
return -EINVAL;
diff --git a/net/ipv4/devinet.c b/net/ipv4/devinet.c
index 4c013f8800f0..ed00b233cee2 100644
--- a/net/ipv4/devinet.c
+++ b/net/ipv4/devinet.c
@@ -1798,6 +1798,21 @@ done:
return err;
}
+/* Combine dev_addr_genid and dev_base_seq to detect changes.
+ */
+static u32 inet_base_seq(const struct net *net)
+{
+ u32 res = atomic_read(&net->ipv4.dev_addr_genid) +
+ net->dev_base_seq;
+
+ /* Must not return 0 (see nl_dump_check_consistent()).
+ * Chose a value far away from 0.
+ */
+ if (!res)
+ res = 0x80000000;
+ return res;
+}
+
static int inet_dump_ifaddr(struct sk_buff *skb, struct netlink_callback *cb)
{
const struct nlmsghdr *nlh = cb->nlh;
@@ -1849,8 +1864,7 @@ static int inet_dump_ifaddr(struct sk_buff *skb, struct netlink_callback *cb)
idx = 0;
head = &tgt_net->dev_index_head[h];
rcu_read_lock();
- cb->seq = atomic_read(&tgt_net->ipv4.dev_addr_genid) ^
- tgt_net->dev_base_seq;
+ cb->seq = inet_base_seq(tgt_net);
hlist_for_each_entry_rcu(dev, head, index_hlist) {
if (idx < s_idx)
goto cont;
@@ -2249,8 +2263,7 @@ static int inet_netconf_dump_devconf(struct sk_buff *skb,
idx = 0;
head = &net->dev_index_head[h];
rcu_read_lock();
- cb->seq = atomic_read(&net->ipv4.dev_addr_genid) ^
- net->dev_base_seq;
+ cb->seq = inet_base_seq(net);
hlist_for_each_entry_rcu(dev, head, index_hlist) {
if (idx < s_idx)
goto cont;
diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c
index 0f70c2dbbe5b..d57d484a929f 100644
--- a/net/ipv4/ip_output.c
+++ b/net/ipv4/ip_output.c
@@ -1255,6 +1255,12 @@ static int ip_setup_cork(struct sock *sk, struct inet_cork *cork,
if (unlikely(!rt))
return -EFAULT;
+ cork->fragsize = ip_sk_use_pmtu(sk) ?
+ dst_mtu(&rt->dst) : READ_ONCE(rt->dst.dev->mtu);
+
+ if (!inetdev_valid_mtu(cork->fragsize))
+ return -ENETUNREACH;
+
/*
* setup for corking.
*/
@@ -1271,12 +1277,6 @@ static int ip_setup_cork(struct sock *sk, struct inet_cork *cork,
cork->addr = ipc->addr;
}
- cork->fragsize = ip_sk_use_pmtu(sk) ?
- dst_mtu(&rt->dst) : READ_ONCE(rt->dst.dev->mtu);
-
- if (!inetdev_valid_mtu(cork->fragsize))
- return -ENETUNREACH;
-
cork->gso_size = ipc->gso_size;
cork->dst = &rt->dst;
diff --git a/net/ipv4/ip_tunnel.c b/net/ipv4/ip_tunnel.c
index 4559edad8cec..906c37c7f80d 100644
--- a/net/ipv4/ip_tunnel.c
+++ b/net/ipv4/ip_tunnel.c
@@ -365,7 +365,7 @@ int ip_tunnel_rcv(struct ip_tunnel *tunnel, struct sk_buff *skb,
{
struct pcpu_sw_netstats *tstats;
const struct iphdr *iph = ip_hdr(skb);
- int err;
+ int nh, err;
#ifdef CONFIG_NET_IPGRE_BROADCAST
if (ipv4_is_multicast(iph->daddr)) {
@@ -391,8 +391,21 @@ int ip_tunnel_rcv(struct ip_tunnel *tunnel, struct sk_buff *skb,
tunnel->i_seqno = ntohl(tpi->seq) + 1;
}
+ /* Save offset of outer header relative to skb->head,
+ * because we are going to reset the network header to the inner header
+ * and might change skb->head.
+ */
+ nh = skb_network_header(skb) - skb->head;
+
skb_set_network_header(skb, (tunnel->dev->type == ARPHRD_ETHER) ? ETH_HLEN : 0);
+ if (!pskb_inet_may_pull(skb)) {
+ DEV_STATS_INC(tunnel->dev, rx_length_errors);
+ DEV_STATS_INC(tunnel->dev, rx_errors);
+ goto drop;
+ }
+ iph = (struct iphdr *)(skb->head + nh);
+
err = IP_ECN_decapsulate(iph, skb);
if (unlikely(err)) {
if (log_ecn_error)
@@ -547,6 +560,20 @@ static int tnl_update_pmtu(struct net_device *dev, struct sk_buff *skb,
return 0;
}
+static void ip_tunnel_adj_headroom(struct net_device *dev, unsigned int headroom)
+{
+ /* we must cap headroom to some upperlimit, else pskb_expand_head
+ * will overflow header offsets in skb_headers_offset_update().
+ */
+ static const unsigned int max_allowed = 512;
+
+ if (headroom > max_allowed)
+ headroom = max_allowed;
+
+ if (headroom > READ_ONCE(dev->needed_headroom))
+ WRITE_ONCE(dev->needed_headroom, headroom);
+}
+
void ip_md_tunnel_xmit(struct sk_buff *skb, struct net_device *dev,
u8 proto, int tunnel_hlen)
{
@@ -620,13 +647,13 @@ void ip_md_tunnel_xmit(struct sk_buff *skb, struct net_device *dev,
}
headroom += LL_RESERVED_SPACE(rt->dst.dev) + rt->dst.header_len;
- if (headroom > READ_ONCE(dev->needed_headroom))
- WRITE_ONCE(dev->needed_headroom, headroom);
-
- if (skb_cow_head(skb, READ_ONCE(dev->needed_headroom))) {
+ if (skb_cow_head(skb, headroom)) {
ip_rt_put(rt);
goto tx_dropped;
}
+
+ ip_tunnel_adj_headroom(dev, headroom);
+
iptunnel_xmit(NULL, rt, skb, fl4.saddr, fl4.daddr, proto, tos, ttl,
df, !net_eq(tunnel->net, dev_net(dev)));
return;
@@ -804,16 +831,16 @@ void ip_tunnel_xmit(struct sk_buff *skb, struct net_device *dev,
max_headroom = LL_RESERVED_SPACE(rt->dst.dev) + sizeof(struct iphdr)
+ rt->dst.header_len + ip_encap_hlen(&tunnel->encap);
- if (max_headroom > READ_ONCE(dev->needed_headroom))
- WRITE_ONCE(dev->needed_headroom, max_headroom);
- if (skb_cow_head(skb, READ_ONCE(dev->needed_headroom))) {
+ if (skb_cow_head(skb, max_headroom)) {
ip_rt_put(rt);
dev->stats.tx_dropped++;
kfree_skb(skb);
return;
}
+ ip_tunnel_adj_headroom(dev, max_headroom);
+
iptunnel_xmit(NULL, rt, skb, fl4.saddr, fl4.daddr, protocol, tos, ttl,
df, !net_eq(tunnel->net, dev_net(dev)));
return;
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 647cb664c2ad..8ebcff40bc5a 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -505,6 +505,7 @@ __poll_t tcp_poll(struct file *file, struct socket *sock, poll_table *wait)
__poll_t mask;
struct sock *sk = sock->sk;
const struct tcp_sock *tp = tcp_sk(sk);
+ u8 shutdown;
int state;
sock_poll_wait(file, sock, wait);
@@ -547,9 +548,10 @@ __poll_t tcp_poll(struct file *file, struct socket *sock, poll_table *wait)
* NOTE. Check for TCP_CLOSE is added. The goal is to prevent
* blocking on fresh not-connected or disconnected socket. --ANK
*/
- if (sk->sk_shutdown == SHUTDOWN_MASK || state == TCP_CLOSE)
+ shutdown = READ_ONCE(sk->sk_shutdown);
+ if (shutdown == SHUTDOWN_MASK || state == TCP_CLOSE)
mask |= EPOLLHUP;
- if (sk->sk_shutdown & RCV_SHUTDOWN)
+ if (shutdown & RCV_SHUTDOWN)
mask |= EPOLLIN | EPOLLRDNORM | EPOLLRDHUP;
/* Connected or passive Fast Open socket? */
@@ -565,8 +567,8 @@ __poll_t tcp_poll(struct file *file, struct socket *sock, poll_table *wait)
if (tcp_stream_is_readable(tp, target, sk))
mask |= EPOLLIN | EPOLLRDNORM;
- if (!(sk->sk_shutdown & SEND_SHUTDOWN)) {
- if (sk_stream_is_writeable(sk)) {
+ if (!(shutdown & SEND_SHUTDOWN)) {
+ if (__sk_stream_is_writeable(sk, 1)) {
mask |= EPOLLOUT | EPOLLWRNORM;
} else { /* send SIGIO later */
sk_set_bit(SOCKWQ_ASYNC_NOSPACE, sk);
@@ -578,7 +580,7 @@ __poll_t tcp_poll(struct file *file, struct socket *sock, poll_table *wait)
* pairs with the input side.
*/
smp_mb__after_atomic();
- if (sk_stream_is_writeable(sk))
+ if (__sk_stream_is_writeable(sk, 1))
mask |= EPOLLOUT | EPOLLWRNORM;
}
} else
@@ -723,6 +725,7 @@ static void tcp_push(struct sock *sk, int flags, int mss_now,
if (!test_bit(TSQ_THROTTLED, &sk->sk_tsq_flags)) {
NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPAUTOCORKING);
set_bit(TSQ_THROTTLED, &sk->sk_tsq_flags);
+ smp_mb__after_atomic();
}
/* It is possible TX completion already happened
* before we set TSQ_THROTTLED.
@@ -2350,14 +2353,13 @@ bool tcp_check_oom(struct sock *sk, int shift)
return too_many_orphans || out_of_socket_memory;
}
-void tcp_close(struct sock *sk, long timeout)
+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;
+ WRITE_ONCE(sk->sk_shutdown, SHUTDOWN_MASK);
if (sk->sk_state == TCP_LISTEN) {
tcp_set_state(sk, TCP_CLOSE);
@@ -2520,6 +2522,12 @@ adjudge_to_death:
out:
bh_unlock_sock(sk);
local_bh_enable();
+}
+
+void tcp_close(struct sock *sk, long timeout)
+{
+ lock_sock(sk);
+ __tcp_close(sk, timeout);
release_sock(sk);
sock_put(sk);
}
@@ -2623,7 +2631,7 @@ int tcp_disconnect(struct sock *sk, int flags)
if (!(sk->sk_userlocks & SOCK_BINDADDR_LOCK))
inet_reset_saddr(sk);
- sk->sk_shutdown = 0;
+ WRITE_ONCE(sk->sk_shutdown, 0);
sock_reset_flag(sk, SOCK_DONE);
tp->srtt_us = 0;
tp->mdev_us = jiffies_to_usecs(TCP_TIMEOUT_INIT);
@@ -3443,11 +3451,11 @@ static int do_tcp_getsockopt(struct sock *sk, int level,
if (get_user(len, optlen))
return -EFAULT;
- len = min_t(unsigned int, len, sizeof(int));
-
if (len < 0)
return -EINVAL;
+ len = min_t(unsigned int, len, sizeof(int));
+
switch (optname) {
case TCP_MAXSEG:
val = tp->mss_cache;
@@ -3899,7 +3907,7 @@ void tcp_done(struct sock *sk)
if (req)
reqsk_fastopen_remove(sk, req, false);
- sk->sk_shutdown = SHUTDOWN_MASK;
+ WRITE_ONCE(sk->sk_shutdown, SHUTDOWN_MASK);
if (!sock_flag(sk, SOCK_DEAD))
sk->sk_state_change(sk);
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 982fe464156a..61243531a7f4 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -4216,7 +4216,7 @@ void tcp_fin(struct sock *sk)
inet_csk_schedule_ack(sk);
- sk->sk_shutdown |= RCV_SHUTDOWN;
+ WRITE_ONCE(sk->sk_shutdown, sk->sk_shutdown | RCV_SHUTDOWN);
sock_set_flag(sk, SOCK_DONE);
switch (sk->sk_state) {
@@ -6354,7 +6354,7 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb)
break;
tcp_set_state(sk, TCP_FIN_WAIT2);
- sk->sk_shutdown |= SEND_SHUTDOWN;
+ WRITE_ONCE(sk->sk_shutdown, sk->sk_shutdown | SEND_SHUTDOWN);
sk_dst_confirm(sk);
diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
index a6f982b2d32a..3b3f94479885 100644
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -2679,11 +2679,11 @@ int udp_lib_getsockopt(struct sock *sk, int level, int optname,
if (get_user(len, optlen))
return -EFAULT;
- len = min_t(unsigned int, len, sizeof(int));
-
if (len < 0)
return -EINVAL;
+ len = min_t(unsigned int, len, sizeof(int));
+
switch (optname) {
case UDP_CORK:
val = READ_ONCE(up->corkflag);
generated by cgit 1.2.3-korg (git 2.39.0) at 2024-04-27 11:08:22 +0000