linux/net/ipv4/tcp_ipv4.c

2438 lines
62 KiB
C
Raw Normal View History

/*
* INET An implementation of the TCP/IP protocol suite for the LINUX
* operating system. INET is implemented using the BSD Socket
* interface as the means of communication with the user level.
*
* Implementation of the Transmission Control Protocol(TCP).
*
* IPv4 specific functions
*
*
* code split from:
* linux/ipv4/tcp.c
* linux/ipv4/tcp_input.c
* linux/ipv4/tcp_output.c
*
* See tcp.c for author information
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation; either version
* 2 of the License, or (at your option) any later version.
*/
/*
* Changes:
* David S. Miller : New socket lookup architecture.
* This code is dedicated to John Dyson.
* David S. Miller : Change semantics of established hash,
* half is devoted to TIME_WAIT sockets
* and the rest go in the other half.
* Andi Kleen : Add support for syncookies and fixed
* some bugs: ip options weren't passed to
* the TCP layer, missed a check for an
* ACK bit.
* Andi Kleen : Implemented fast path mtu discovery.
* Fixed many serious bugs in the
* request_sock handling and moved
* most of it into the af independent code.
* Added tail drop and some other bugfixes.
* Added new listen semantics.
* Mike McLagan : Routing by source
* Juan Jose Ciarlante: ip_dynaddr bits
* Andi Kleen: various fixes.
* Vitaly E. Lavrov : Transparent proxy revived after year
* coma.
* Andi Kleen : Fix new listen.
* Andi Kleen : Fix accept error reporting.
* YOSHIFUJI Hideaki @USAGI and: Support IPV6_V6ONLY socket option, which
* Alexey Kuznetsov allow both IPv4 and IPv6 sockets to bind
* a single port at the same time.
*/
#define pr_fmt(fmt) "TCP: " fmt
#include <linux/bottom_half.h>
#include <linux/types.h>
#include <linux/fcntl.h>
#include <linux/module.h>
#include <linux/random.h>
#include <linux/cache.h>
#include <linux/jhash.h>
#include <linux/init.h>
#include <linux/times.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 08:04:11 +00:00
#include <linux/slab.h>
#include <net/net_namespace.h>
#include <net/icmp.h>
#include <net/inet_hashtables.h>
#include <net/tcp.h>
#include <net/transp_v6.h>
#include <net/ipv6.h>
#include <net/inet_common.h>
#include <net/timewait_sock.h>
#include <net/xfrm.h>
#include <net/secure_seq.h>
#include <net/busy_poll.h>
#include <linux/inet.h>
#include <linux/ipv6.h>
#include <linux/stddef.h>
#include <linux/proc_fs.h>
#include <linux/seq_file.h>
#include <crypto/hash.h>
#include <linux/scatterlist.h>
int sysctl_tcp_tw_reuse __read_mostly;
int sysctl_tcp_low_latency __read_mostly;
EXPORT_SYMBOL(sysctl_tcp_low_latency);
#ifdef CONFIG_TCP_MD5SIG
static int tcp_v4_md5_hash_hdr(char *md5_hash, const struct tcp_md5sig_key *key,
__be32 daddr, __be32 saddr, const struct tcphdr *th);
#endif
struct inet_hashinfo tcp_hashinfo;
EXPORT_SYMBOL(tcp_hashinfo);
static __u32 tcp_v4_init_sequence(const struct sk_buff *skb)
{
return secure_tcp_sequence_number(ip_hdr(skb)->daddr,
ip_hdr(skb)->saddr,
tcp_hdr(skb)->dest,
tcp_hdr(skb)->source);
}
int tcp_twsk_unique(struct sock *sk, struct sock *sktw, void *twp)
{
const struct tcp_timewait_sock *tcptw = tcp_twsk(sktw);
struct tcp_sock *tp = tcp_sk(sk);
/* With PAWS, it is safe from the viewpoint
of data integrity. Even without PAWS it is safe provided sequence
spaces do not overlap i.e. at data rates <= 80Mbit/sec.
Actually, the idea is close to VJ's one, only timestamp cache is
held not per host, but per port pair and TW bucket is used as state
holder.
If TW bucket has been already destroyed we fall back to VJ's scheme
and use initial timestamp retrieved from peer table.
*/
if (tcptw->tw_ts_recent_stamp &&
(!twp || (sysctl_tcp_tw_reuse &&
get_seconds() - tcptw->tw_ts_recent_stamp > 1))) {
tp->write_seq = tcptw->tw_snd_nxt + 65535 + 2;
if (tp->write_seq == 0)
tp->write_seq = 1;
tp->rx_opt.ts_recent = tcptw->tw_ts_recent;
tp->rx_opt.ts_recent_stamp = tcptw->tw_ts_recent_stamp;
sock_hold(sktw);
return 1;
}
return 0;
}
EXPORT_SYMBOL_GPL(tcp_twsk_unique);
/* This will initiate an outgoing connection. */
int tcp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len)
{
struct sockaddr_in *usin = (struct sockaddr_in *)uaddr;
struct inet_sock *inet = inet_sk(sk);
struct tcp_sock *tp = tcp_sk(sk);
__be16 orig_sport, orig_dport;
__be32 daddr, nexthop;
struct flowi4 *fl4;
struct rtable *rt;
int err;
struct ip_options_rcu *inet_opt;
if (addr_len < sizeof(struct sockaddr_in))
return -EINVAL;
if (usin->sin_family != AF_INET)
return -EAFNOSUPPORT;
nexthop = daddr = usin->sin_addr.s_addr;
inet_opt = rcu_dereference_protected(inet->inet_opt,
lockdep_sock_is_held(sk));
if (inet_opt && inet_opt->opt.srr) {
if (!daddr)
return -EINVAL;
nexthop = inet_opt->opt.faddr;
}
orig_sport = inet->inet_sport;
orig_dport = usin->sin_port;
fl4 = &inet->cork.fl.u.ip4;
rt = ip_route_connect(fl4, nexthop, inet->inet_saddr,
RT_CONN_FLAGS(sk), sk->sk_bound_dev_if,
IPPROTO_TCP,
orig_sport, orig_dport, sk);
if (IS_ERR(rt)) {
err = PTR_ERR(rt);
if (err == -ENETUNREACH)
IP_INC_STATS(sock_net(sk), IPSTATS_MIB_OUTNOROUTES);
return err;
}
if (rt->rt_flags & (RTCF_MULTICAST | RTCF_BROADCAST)) {
ip_rt_put(rt);
return -ENETUNREACH;
}
if (!inet_opt || !inet_opt->opt.srr)
daddr = fl4->daddr;
if (!inet->inet_saddr)
inet->inet_saddr = fl4->saddr;
sk_rcv_saddr_set(sk, inet->inet_saddr);
if (tp->rx_opt.ts_recent_stamp && inet->inet_daddr != daddr) {
/* Reset inherited state */
tp->rx_opt.ts_recent = 0;
tp->rx_opt.ts_recent_stamp = 0;
if (likely(!tp->repair))
tp->write_seq = 0;
}
if (tcp_death_row.sysctl_tw_recycle &&
!tp->rx_opt.ts_recent_stamp && fl4->daddr == daddr)
tcp_fetch_timewait_stamp(sk, &rt->dst);
inet->inet_dport = usin->sin_port;
sk_daddr_set(sk, daddr);
inet_csk(sk)->icsk_ext_hdr_len = 0;
if (inet_opt)
inet_csk(sk)->icsk_ext_hdr_len = inet_opt->opt.optlen;
tp->rx_opt.mss_clamp = TCP_MSS_DEFAULT;
/* Socket identity is still unknown (sport may be zero).
* However we set state to SYN-SENT and not releasing socket
* lock select source port, enter ourselves into the hash tables and
* complete initialization after this.
*/
tcp_set_state(sk, TCP_SYN_SENT);
err = inet_hash_connect(&tcp_death_row, sk);
if (err)
goto failure;
sk_set_txhash(sk);
rt = ip_route_newports(fl4, rt, orig_sport, orig_dport,
inet->inet_sport, inet->inet_dport, sk);
if (IS_ERR(rt)) {
err = PTR_ERR(rt);
rt = NULL;
goto failure;
}
/* OK, now commit destination to socket. */
sk->sk_gso_type = SKB_GSO_TCPV4;
sk_setup_caps(sk, &rt->dst);
if (!tp->write_seq && likely(!tp->repair))
tp->write_seq = secure_tcp_sequence_number(inet->inet_saddr,
inet->inet_daddr,
inet->inet_sport,
usin->sin_port);
inet->inet_id = tp->write_seq ^ jiffies;
err = tcp_connect(sk);
rt = NULL;
if (err)
goto failure;
return 0;
failure:
/*
* This unhashes the socket and releases the local port,
* if necessary.
*/
tcp_set_state(sk, TCP_CLOSE);
ip_rt_put(rt);
sk->sk_route_caps = 0;
inet->inet_dport = 0;
return err;
}
EXPORT_SYMBOL(tcp_v4_connect);
/*
* This routine reacts to ICMP_FRAG_NEEDED mtu indications as defined in RFC1191.
* It can be called through tcp_release_cb() if socket was owned by user
* at the time tcp_v4_err() was called to handle ICMP message.
*/
void tcp_v4_mtu_reduced(struct sock *sk)
{
struct dst_entry *dst;
struct inet_sock *inet = inet_sk(sk);
u32 mtu = tcp_sk(sk)->mtu_info;
dst = inet_csk_update_pmtu(sk, mtu);
if (!dst)
return;
/* Something is about to be wrong... Remember soft error
* for the case, if this connection will not able to recover.
*/
if (mtu < dst_mtu(dst) && ip_dont_fragment(sk, dst))
sk->sk_err_soft = EMSGSIZE;
mtu = dst_mtu(dst);
if (inet->pmtudisc != IP_PMTUDISC_DONT &&
ipv4: introduce new IP_MTU_DISCOVER mode IP_PMTUDISC_INTERFACE Sockets marked with IP_PMTUDISC_INTERFACE won't do path mtu discovery, their sockets won't accept and install new path mtu information and they will always use the interface mtu for outgoing packets. It is guaranteed that the packet is not fragmented locally. But we won't set the DF-Flag on the outgoing frames. Florian Weimer had the idea to use this flag to ensure DNS servers are never generating outgoing fragments. They may well be fragmented on the path, but the server never stores or usees path mtu values, which could well be forged in an attack. (The root of the problem with path MTU discovery is that there is no reliable way to authenticate ICMP Fragmentation Needed But DF Set messages because they are sent from intermediate routers with their source addresses, and the IMCP payload will not always contain sufficient information to identify a flow.) Recent research in the DNS community showed that it is possible to implement an attack where DNS cache poisoning is feasible by spoofing fragments. This work was done by Amir Herzberg and Haya Shulman: <https://sites.google.com/site/hayashulman/files/fragmentation-poisoning.pdf> This issue was previously discussed among the DNS community, e.g. <http://www.ietf.org/mail-archive/web/dnsext/current/msg01204.html>, without leading to fixes. This patch depends on the patch "ipv4: fix DO and PROBE pmtu mode regarding local fragmentation with UFO/CORK" for the enforcement of the non-fragmentable checks. If other users than ip_append_page/data should use this semantic too, we have to add a new flag to IPCB(skb)->flags to suppress local fragmentation and check for this in ip_finish_output. Many thanks to Florian Weimer for the idea and feedback while implementing this patch. Cc: David S. Miller <davem@davemloft.net> Suggested-by: Florian Weimer <fweimer@redhat.com> Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-11-05 01:24:17 +00:00
ip_sk_accept_pmtu(sk) &&
inet_csk(sk)->icsk_pmtu_cookie > mtu) {
tcp_sync_mss(sk, mtu);
/* Resend the TCP packet because it's
* clear that the old packet has been
* dropped. This is the new "fast" path mtu
* discovery.
*/
tcp_simple_retransmit(sk);
} /* else let the usual retransmit timer handle it */
}
EXPORT_SYMBOL(tcp_v4_mtu_reduced);
static void do_redirect(struct sk_buff *skb, struct sock *sk)
{
struct dst_entry *dst = __sk_dst_check(sk, 0);
if (dst)
dst->ops->redirect(dst, sk, skb);
}
/* handle ICMP messages on TCP_NEW_SYN_RECV request sockets */
void tcp_req_err(struct sock *sk, u32 seq, bool abort)
{
struct request_sock *req = inet_reqsk(sk);
struct net *net = sock_net(sk);
/* ICMPs are not backlogged, hence we cannot get
* an established socket here.
*/
if (seq != tcp_rsk(req)->snt_isn) {
__NET_INC_STATS(net, LINUX_MIB_OUTOFWINDOWICMPS);
} else if (abort) {
/*
* Still in SYN_RECV, just remove it silently.
* There is no good way to pass the error to the newly
* created socket, and POSIX does not want network
* errors returned from accept().
*/
inet: fix double request socket freeing Eric Hugne reported following error : I'm hitting this warning on latest net-next when i try to SSH into a machine with eth0 added to a bridge (but i think the problem is older than that) Steps to reproduce: node2 ~ # brctl addif br0 eth0 [ 223.758785] device eth0 entered promiscuous mode node2 ~ # ip link set br0 up [ 244.503614] br0: port 1(eth0) entered forwarding state [ 244.505108] br0: port 1(eth0) entered forwarding state node2 ~ # [ 251.160159] ------------[ cut here ]------------ [ 251.160831] WARNING: CPU: 0 PID: 3 at include/net/request_sock.h:102 tcp_v4_err+0x6b1/0x720() [ 251.162077] Modules linked in: [ 251.162496] CPU: 0 PID: 3 Comm: ksoftirqd/0 Not tainted 4.0.0-rc3+ #18 [ 251.163334] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011 [ 251.164078] ffffffff81a8365c ffff880038a6ba18 ffffffff8162ace4 0000000000009898 [ 251.165084] 0000000000000000 ffff880038a6ba58 ffffffff8104da85 ffff88003fa437c0 [ 251.166195] ffff88003fa437c0 ffff88003fa74e00 ffff88003fa43bb8 ffff88003fad99a0 [ 251.167203] Call Trace: [ 251.167533] [<ffffffff8162ace4>] dump_stack+0x45/0x57 [ 251.168206] [<ffffffff8104da85>] warn_slowpath_common+0x85/0xc0 [ 251.169239] [<ffffffff8104db65>] warn_slowpath_null+0x15/0x20 [ 251.170271] [<ffffffff81559d51>] tcp_v4_err+0x6b1/0x720 [ 251.171408] [<ffffffff81630d03>] ? _raw_read_lock_irq+0x3/0x10 [ 251.172589] [<ffffffff81534e20>] ? inet_del_offload+0x40/0x40 [ 251.173366] [<ffffffff81569295>] icmp_socket_deliver+0x65/0xb0 [ 251.174134] [<ffffffff815693a2>] icmp_unreach+0xc2/0x280 [ 251.174820] [<ffffffff8156a82d>] icmp_rcv+0x2bd/0x3a0 [ 251.175473] [<ffffffff81534ea2>] ip_local_deliver_finish+0x82/0x1e0 [ 251.176282] [<ffffffff815354d8>] ip_local_deliver+0x88/0x90 [ 251.177004] [<ffffffff815350f0>] ip_rcv_finish+0xf0/0x310 [ 251.177693] [<ffffffff815357bc>] ip_rcv+0x2dc/0x390 [ 251.178336] [<ffffffff814f5da3>] __netif_receive_skb_core+0x713/0xa20 [ 251.179170] [<ffffffff814f7fca>] __netif_receive_skb+0x1a/0x80 [ 251.179922] [<ffffffff814f97d4>] process_backlog+0x94/0x120 [ 251.180639] [<ffffffff814f9612>] net_rx_action+0x1e2/0x310 [ 251.181356] [<ffffffff81051267>] __do_softirq+0xa7/0x290 [ 251.182046] [<ffffffff81051469>] run_ksoftirqd+0x19/0x30 [ 251.182726] [<ffffffff8106cc23>] smpboot_thread_fn+0x153/0x1d0 [ 251.183485] [<ffffffff8106cad0>] ? SyS_setgroups+0x130/0x130 [ 251.184228] [<ffffffff8106935e>] kthread+0xee/0x110 [ 251.184871] [<ffffffff81069270>] ? kthread_create_on_node+0x1b0/0x1b0 [ 251.185690] [<ffffffff81631108>] ret_from_fork+0x58/0x90 [ 251.186385] [<ffffffff81069270>] ? kthread_create_on_node+0x1b0/0x1b0 [ 251.187216] ---[ end trace c947fc7b24e42ea1 ]--- [ 259.542268] br0: port 1(eth0) entered forwarding state Remove the double calls to reqsk_put() [edumazet] : I got confused because reqsk_timer_handler() _has_ to call reqsk_put(req) after calling inet_csk_reqsk_queue_drop(), as the timer handler holds a reference on req. Signed-off-by: Fan Du <fan.du@intel.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: Erik Hugne <erik.hugne@ericsson.com> Fixes: fa76ce7328b2 ("inet: get rid of central tcp/dccp listener timer") Signed-off-by: David S. Miller <davem@davemloft.net>
2015-03-23 22:00:41 +00:00
inet_csk_reqsk_queue_drop(req->rsk_listener, req);
tcp_listendrop(req->rsk_listener);
}
reqsk_put(req);
}
EXPORT_SYMBOL(tcp_req_err);
/*
* This routine is called by the ICMP module when it gets some
* sort of error condition. If err < 0 then the socket should
* be closed and the error returned to the user. If err > 0
* it's just the icmp type << 8 | icmp code. After adjustment
* header points to the first 8 bytes of the tcp header. We need
* to find the appropriate port.
*
* The locking strategy used here is very "optimistic". When
* someone else accesses the socket the ICMP is just dropped
* and for some paths there is no check at all.
* A more general error queue to queue errors for later handling
* is probably better.
*
*/
void tcp_v4_err(struct sk_buff *icmp_skb, u32 info)
{
const struct iphdr *iph = (const struct iphdr *)icmp_skb->data;
struct tcphdr *th = (struct tcphdr *)(icmp_skb->data + (iph->ihl << 2));
struct inet_connection_sock *icsk;
struct tcp_sock *tp;
struct inet_sock *inet;
const int type = icmp_hdr(icmp_skb)->type;
const int code = icmp_hdr(icmp_skb)->code;
struct sock *sk;
struct sk_buff *skb;
struct request_sock *fastopen;
__u32 seq, snd_una;
__u32 remaining;
int err;
struct net *net = dev_net(icmp_skb->dev);
sk = __inet_lookup_established(net, &tcp_hashinfo, iph->daddr,
th->dest, iph->saddr, ntohs(th->source),
inet_iif(icmp_skb));
if (!sk) {
__ICMP_INC_STATS(net, ICMP_MIB_INERRORS);
return;
}
if (sk->sk_state == TCP_TIME_WAIT) {
inet_twsk_put(inet_twsk(sk));
return;
}
seq = ntohl(th->seq);
if (sk->sk_state == TCP_NEW_SYN_RECV)
return tcp_req_err(sk, seq,
type == ICMP_PARAMETERPROB ||
type == ICMP_TIME_EXCEEDED ||
(type == ICMP_DEST_UNREACH &&
(code == ICMP_NET_UNREACH ||
code == ICMP_HOST_UNREACH)));
bh_lock_sock(sk);
/* If too many ICMPs get dropped on busy
* servers this needs to be solved differently.
* We do take care of PMTU discovery (RFC1191) special case :
* we can receive locally generated ICMP messages while socket is held.
*/
if (sock_owned_by_user(sk)) {
if (!(type == ICMP_DEST_UNREACH && code == ICMP_FRAG_NEEDED))
__NET_INC_STATS(net, LINUX_MIB_LOCKDROPPEDICMPS);
}
if (sk->sk_state == TCP_CLOSE)
goto out;
if (unlikely(iph->ttl < inet_sk(sk)->min_ttl)) {
__NET_INC_STATS(net, LINUX_MIB_TCPMINTTLDROP);
goto out;
}
icsk = inet_csk(sk);
tp = tcp_sk(sk);
/* XXX (TFO) - tp->snd_una should be ISN (tcp_create_openreq_child() */
fastopen = tp->fastopen_rsk;
snd_una = fastopen ? tcp_rsk(fastopen)->snt_isn : tp->snd_una;
if (sk->sk_state != TCP_LISTEN &&
!between(seq, snd_una, tp->snd_nxt)) {
__NET_INC_STATS(net, LINUX_MIB_OUTOFWINDOWICMPS);
goto out;
}
switch (type) {
case ICMP_REDIRECT:
do_redirect(icmp_skb, sk);
goto out;
case ICMP_SOURCE_QUENCH:
/* Just silently ignore these. */
goto out;
case ICMP_PARAMETERPROB:
err = EPROTO;
break;
case ICMP_DEST_UNREACH:
if (code > NR_ICMP_UNREACH)
goto out;
if (code == ICMP_FRAG_NEEDED) { /* PMTU discovery (RFC1191) */
/* We are not interested in TCP_LISTEN and open_requests
* (SYN-ACKs send out by Linux are always <576bytes so
* they should go through unfragmented).
*/
if (sk->sk_state == TCP_LISTEN)
goto out;
tp->mtu_info = info;
tcp: fix possible socket refcount problem Commit 6f458dfb40 (tcp: improve latencies of timer triggered events) added bug leading to following trace : [ 2866.131281] IPv4: Attempt to release TCP socket in state 1 ffff880019ec0000 [ 2866.131726] [ 2866.132188] ========================= [ 2866.132281] [ BUG: held lock freed! ] [ 2866.132281] 3.6.0-rc1+ #622 Not tainted [ 2866.132281] ------------------------- [ 2866.132281] kworker/0:1/652 is freeing memory ffff880019ec0000-ffff880019ec0a1f, with a lock still held there! [ 2866.132281] (sk_lock-AF_INET-RPC){+.+...}, at: [<ffffffff81903619>] tcp_sendmsg+0x29/0xcc6 [ 2866.132281] 4 locks held by kworker/0:1/652: [ 2866.132281] #0: (rpciod){.+.+.+}, at: [<ffffffff81083567>] process_one_work+0x1de/0x47f [ 2866.132281] #1: ((&task->u.tk_work)){+.+.+.}, at: [<ffffffff81083567>] process_one_work+0x1de/0x47f [ 2866.132281] #2: (sk_lock-AF_INET-RPC){+.+...}, at: [<ffffffff81903619>] tcp_sendmsg+0x29/0xcc6 [ 2866.132281] #3: (&icsk->icsk_retransmit_timer){+.-...}, at: [<ffffffff81078017>] run_timer_softirq+0x1ad/0x35f [ 2866.132281] [ 2866.132281] stack backtrace: [ 2866.132281] Pid: 652, comm: kworker/0:1 Not tainted 3.6.0-rc1+ #622 [ 2866.132281] Call Trace: [ 2866.132281] <IRQ> [<ffffffff810bc527>] debug_check_no_locks_freed+0x112/0x159 [ 2866.132281] [<ffffffff818a0839>] ? __sk_free+0xfd/0x114 [ 2866.132281] [<ffffffff811549fa>] kmem_cache_free+0x6b/0x13a [ 2866.132281] [<ffffffff818a0839>] __sk_free+0xfd/0x114 [ 2866.132281] [<ffffffff818a08c0>] sk_free+0x1c/0x1e [ 2866.132281] [<ffffffff81911e1c>] tcp_write_timer+0x51/0x56 [ 2866.132281] [<ffffffff81078082>] run_timer_softirq+0x218/0x35f [ 2866.132281] [<ffffffff81078017>] ? run_timer_softirq+0x1ad/0x35f [ 2866.132281] [<ffffffff810f5831>] ? rb_commit+0x58/0x85 [ 2866.132281] [<ffffffff81911dcb>] ? tcp_write_timer_handler+0x148/0x148 [ 2866.132281] [<ffffffff81070bd6>] __do_softirq+0xcb/0x1f9 [ 2866.132281] [<ffffffff81a0a00c>] ? _raw_spin_unlock+0x29/0x2e [ 2866.132281] [<ffffffff81a1227c>] call_softirq+0x1c/0x30 [ 2866.132281] [<ffffffff81039f38>] do_softirq+0x4a/0xa6 [ 2866.132281] [<ffffffff81070f2b>] irq_exit+0x51/0xad [ 2866.132281] [<ffffffff81a129cd>] do_IRQ+0x9d/0xb4 [ 2866.132281] [<ffffffff81a0a3ef>] common_interrupt+0x6f/0x6f [ 2866.132281] <EOI> [<ffffffff8109d006>] ? sched_clock_cpu+0x58/0xd1 [ 2866.132281] [<ffffffff81a0a172>] ? _raw_spin_unlock_irqrestore+0x4c/0x56 [ 2866.132281] [<ffffffff81078692>] mod_timer+0x178/0x1a9 [ 2866.132281] [<ffffffff818a00aa>] sk_reset_timer+0x19/0x26 [ 2866.132281] [<ffffffff8190b2cc>] tcp_rearm_rto+0x99/0xa4 [ 2866.132281] [<ffffffff8190dfba>] tcp_event_new_data_sent+0x6e/0x70 [ 2866.132281] [<ffffffff8190f7ea>] tcp_write_xmit+0x7de/0x8e4 [ 2866.132281] [<ffffffff818a565d>] ? __alloc_skb+0xa0/0x1a1 [ 2866.132281] [<ffffffff8190f952>] __tcp_push_pending_frames+0x2e/0x8a [ 2866.132281] [<ffffffff81904122>] tcp_sendmsg+0xb32/0xcc6 [ 2866.132281] [<ffffffff819229c2>] inet_sendmsg+0xaa/0xd5 [ 2866.132281] [<ffffffff81922918>] ? inet_autobind+0x5f/0x5f [ 2866.132281] [<ffffffff810ee7f1>] ? trace_clock_local+0x9/0xb [ 2866.132281] [<ffffffff8189adab>] sock_sendmsg+0xa3/0xc4 [ 2866.132281] [<ffffffff810f5de6>] ? rb_reserve_next_event+0x26f/0x2d5 [ 2866.132281] [<ffffffff8103e6a9>] ? native_sched_clock+0x29/0x6f [ 2866.132281] [<ffffffff8103e6f8>] ? sched_clock+0x9/0xd [ 2866.132281] [<ffffffff810ee7f1>] ? trace_clock_local+0x9/0xb [ 2866.132281] [<ffffffff8189ae03>] kernel_sendmsg+0x37/0x43 [ 2866.132281] [<ffffffff8199ce49>] xs_send_kvec+0x77/0x80 [ 2866.132281] [<ffffffff8199cec1>] xs_sendpages+0x6f/0x1a0 [ 2866.132281] [<ffffffff8107826d>] ? try_to_del_timer_sync+0x55/0x61 [ 2866.132281] [<ffffffff8199d0d2>] xs_tcp_send_request+0x55/0xf1 [ 2866.132281] [<ffffffff8199bb90>] xprt_transmit+0x89/0x1db [ 2866.132281] [<ffffffff81999bcd>] ? call_connect+0x3c/0x3c [ 2866.132281] [<ffffffff81999d92>] call_transmit+0x1c5/0x20e [ 2866.132281] [<ffffffff819a0d55>] __rpc_execute+0x6f/0x225 [ 2866.132281] [<ffffffff81999bcd>] ? call_connect+0x3c/0x3c [ 2866.132281] [<ffffffff819a0f33>] rpc_async_schedule+0x28/0x34 [ 2866.132281] [<ffffffff810835d6>] process_one_work+0x24d/0x47f [ 2866.132281] [<ffffffff81083567>] ? process_one_work+0x1de/0x47f [ 2866.132281] [<ffffffff819a0f0b>] ? __rpc_execute+0x225/0x225 [ 2866.132281] [<ffffffff81083a6d>] worker_thread+0x236/0x317 [ 2866.132281] [<ffffffff81083837>] ? process_scheduled_works+0x2f/0x2f [ 2866.132281] [<ffffffff8108b7b8>] kthread+0x9a/0xa2 [ 2866.132281] [<ffffffff81a12184>] kernel_thread_helper+0x4/0x10 [ 2866.132281] [<ffffffff81a0a4b0>] ? retint_restore_args+0x13/0x13 [ 2866.132281] [<ffffffff8108b71e>] ? __init_kthread_worker+0x5a/0x5a [ 2866.132281] [<ffffffff81a12180>] ? gs_change+0x13/0x13 [ 2866.308506] IPv4: Attempt to release TCP socket in state 1 ffff880019ec0000 [ 2866.309689] ============================================================================= [ 2866.310254] BUG TCP (Not tainted): Object already free [ 2866.310254] ----------------------------------------------------------------------------- [ 2866.310254] The bug comes from the fact that timer set in sk_reset_timer() can run before we actually do the sock_hold(). socket refcount reaches zero and we free the socket too soon. timer handler is not allowed to reduce socket refcnt if socket is owned by the user, or we need to change sk_reset_timer() implementation. We should take a reference on the socket in case TCP_DELACK_TIMER_DEFERRED or TCP_DELACK_TIMER_DEFERRED bit are set in tsq_flags Also fix a typo in tcp_delack_timer(), where TCP_WRITE_TIMER_DEFERRED was used instead of TCP_DELACK_TIMER_DEFERRED. For consistency, use same socket refcount change for TCP_MTU_REDUCED_DEFERRED, even if not fired from a timer. Reported-by: Fengguang Wu <fengguang.wu@intel.com> Tested-by: Fengguang Wu <fengguang.wu@intel.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-08-20 00:22:46 +00:00
if (!sock_owned_by_user(sk)) {
tcp_v4_mtu_reduced(sk);
tcp: fix possible socket refcount problem Commit 6f458dfb40 (tcp: improve latencies of timer triggered events) added bug leading to following trace : [ 2866.131281] IPv4: Attempt to release TCP socket in state 1 ffff880019ec0000 [ 2866.131726] [ 2866.132188] ========================= [ 2866.132281] [ BUG: held lock freed! ] [ 2866.132281] 3.6.0-rc1+ #622 Not tainted [ 2866.132281] ------------------------- [ 2866.132281] kworker/0:1/652 is freeing memory ffff880019ec0000-ffff880019ec0a1f, with a lock still held there! [ 2866.132281] (sk_lock-AF_INET-RPC){+.+...}, at: [<ffffffff81903619>] tcp_sendmsg+0x29/0xcc6 [ 2866.132281] 4 locks held by kworker/0:1/652: [ 2866.132281] #0: (rpciod){.+.+.+}, at: [<ffffffff81083567>] process_one_work+0x1de/0x47f [ 2866.132281] #1: ((&task->u.tk_work)){+.+.+.}, at: [<ffffffff81083567>] process_one_work+0x1de/0x47f [ 2866.132281] #2: (sk_lock-AF_INET-RPC){+.+...}, at: [<ffffffff81903619>] tcp_sendmsg+0x29/0xcc6 [ 2866.132281] #3: (&icsk->icsk_retransmit_timer){+.-...}, at: [<ffffffff81078017>] run_timer_softirq+0x1ad/0x35f [ 2866.132281] [ 2866.132281] stack backtrace: [ 2866.132281] Pid: 652, comm: kworker/0:1 Not tainted 3.6.0-rc1+ #622 [ 2866.132281] Call Trace: [ 2866.132281] <IRQ> [<ffffffff810bc527>] debug_check_no_locks_freed+0x112/0x159 [ 2866.132281] [<ffffffff818a0839>] ? __sk_free+0xfd/0x114 [ 2866.132281] [<ffffffff811549fa>] kmem_cache_free+0x6b/0x13a [ 2866.132281] [<ffffffff818a0839>] __sk_free+0xfd/0x114 [ 2866.132281] [<ffffffff818a08c0>] sk_free+0x1c/0x1e [ 2866.132281] [<ffffffff81911e1c>] tcp_write_timer+0x51/0x56 [ 2866.132281] [<ffffffff81078082>] run_timer_softirq+0x218/0x35f [ 2866.132281] [<ffffffff81078017>] ? run_timer_softirq+0x1ad/0x35f [ 2866.132281] [<ffffffff810f5831>] ? rb_commit+0x58/0x85 [ 2866.132281] [<ffffffff81911dcb>] ? tcp_write_timer_handler+0x148/0x148 [ 2866.132281] [<ffffffff81070bd6>] __do_softirq+0xcb/0x1f9 [ 2866.132281] [<ffffffff81a0a00c>] ? _raw_spin_unlock+0x29/0x2e [ 2866.132281] [<ffffffff81a1227c>] call_softirq+0x1c/0x30 [ 2866.132281] [<ffffffff81039f38>] do_softirq+0x4a/0xa6 [ 2866.132281] [<ffffffff81070f2b>] irq_exit+0x51/0xad [ 2866.132281] [<ffffffff81a129cd>] do_IRQ+0x9d/0xb4 [ 2866.132281] [<ffffffff81a0a3ef>] common_interrupt+0x6f/0x6f [ 2866.132281] <EOI> [<ffffffff8109d006>] ? sched_clock_cpu+0x58/0xd1 [ 2866.132281] [<ffffffff81a0a172>] ? _raw_spin_unlock_irqrestore+0x4c/0x56 [ 2866.132281] [<ffffffff81078692>] mod_timer+0x178/0x1a9 [ 2866.132281] [<ffffffff818a00aa>] sk_reset_timer+0x19/0x26 [ 2866.132281] [<ffffffff8190b2cc>] tcp_rearm_rto+0x99/0xa4 [ 2866.132281] [<ffffffff8190dfba>] tcp_event_new_data_sent+0x6e/0x70 [ 2866.132281] [<ffffffff8190f7ea>] tcp_write_xmit+0x7de/0x8e4 [ 2866.132281] [<ffffffff818a565d>] ? __alloc_skb+0xa0/0x1a1 [ 2866.132281] [<ffffffff8190f952>] __tcp_push_pending_frames+0x2e/0x8a [ 2866.132281] [<ffffffff81904122>] tcp_sendmsg+0xb32/0xcc6 [ 2866.132281] [<ffffffff819229c2>] inet_sendmsg+0xaa/0xd5 [ 2866.132281] [<ffffffff81922918>] ? inet_autobind+0x5f/0x5f [ 2866.132281] [<ffffffff810ee7f1>] ? trace_clock_local+0x9/0xb [ 2866.132281] [<ffffffff8189adab>] sock_sendmsg+0xa3/0xc4 [ 2866.132281] [<ffffffff810f5de6>] ? rb_reserve_next_event+0x26f/0x2d5 [ 2866.132281] [<ffffffff8103e6a9>] ? native_sched_clock+0x29/0x6f [ 2866.132281] [<ffffffff8103e6f8>] ? sched_clock+0x9/0xd [ 2866.132281] [<ffffffff810ee7f1>] ? trace_clock_local+0x9/0xb [ 2866.132281] [<ffffffff8189ae03>] kernel_sendmsg+0x37/0x43 [ 2866.132281] [<ffffffff8199ce49>] xs_send_kvec+0x77/0x80 [ 2866.132281] [<ffffffff8199cec1>] xs_sendpages+0x6f/0x1a0 [ 2866.132281] [<ffffffff8107826d>] ? try_to_del_timer_sync+0x55/0x61 [ 2866.132281] [<ffffffff8199d0d2>] xs_tcp_send_request+0x55/0xf1 [ 2866.132281] [<ffffffff8199bb90>] xprt_transmit+0x89/0x1db [ 2866.132281] [<ffffffff81999bcd>] ? call_connect+0x3c/0x3c [ 2866.132281] [<ffffffff81999d92>] call_transmit+0x1c5/0x20e [ 2866.132281] [<ffffffff819a0d55>] __rpc_execute+0x6f/0x225 [ 2866.132281] [<ffffffff81999bcd>] ? call_connect+0x3c/0x3c [ 2866.132281] [<ffffffff819a0f33>] rpc_async_schedule+0x28/0x34 [ 2866.132281] [<ffffffff810835d6>] process_one_work+0x24d/0x47f [ 2866.132281] [<ffffffff81083567>] ? process_one_work+0x1de/0x47f [ 2866.132281] [<ffffffff819a0f0b>] ? __rpc_execute+0x225/0x225 [ 2866.132281] [<ffffffff81083a6d>] worker_thread+0x236/0x317 [ 2866.132281] [<ffffffff81083837>] ? process_scheduled_works+0x2f/0x2f [ 2866.132281] [<ffffffff8108b7b8>] kthread+0x9a/0xa2 [ 2866.132281] [<ffffffff81a12184>] kernel_thread_helper+0x4/0x10 [ 2866.132281] [<ffffffff81a0a4b0>] ? retint_restore_args+0x13/0x13 [ 2866.132281] [<ffffffff8108b71e>] ? __init_kthread_worker+0x5a/0x5a [ 2866.132281] [<ffffffff81a12180>] ? gs_change+0x13/0x13 [ 2866.308506] IPv4: Attempt to release TCP socket in state 1 ffff880019ec0000 [ 2866.309689] ============================================================================= [ 2866.310254] BUG TCP (Not tainted): Object already free [ 2866.310254] ----------------------------------------------------------------------------- [ 2866.310254] The bug comes from the fact that timer set in sk_reset_timer() can run before we actually do the sock_hold(). socket refcount reaches zero and we free the socket too soon. timer handler is not allowed to reduce socket refcnt if socket is owned by the user, or we need to change sk_reset_timer() implementation. We should take a reference on the socket in case TCP_DELACK_TIMER_DEFERRED or TCP_DELACK_TIMER_DEFERRED bit are set in tsq_flags Also fix a typo in tcp_delack_timer(), where TCP_WRITE_TIMER_DEFERRED was used instead of TCP_DELACK_TIMER_DEFERRED. For consistency, use same socket refcount change for TCP_MTU_REDUCED_DEFERRED, even if not fired from a timer. Reported-by: Fengguang Wu <fengguang.wu@intel.com> Tested-by: Fengguang Wu <fengguang.wu@intel.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-08-20 00:22:46 +00:00
} else {
if (!test_and_set_bit(TCP_MTU_REDUCED_DEFERRED, &tp->tsq_flags))
sock_hold(sk);
}
goto out;
}
err = icmp_err_convert[code].errno;
/* check if icmp_skb allows revert of backoff
* (see draft-zimmermann-tcp-lcd) */
if (code != ICMP_NET_UNREACH && code != ICMP_HOST_UNREACH)
break;
if (seq != tp->snd_una || !icsk->icsk_retransmits ||
!icsk->icsk_backoff || fastopen)
break;
if (sock_owned_by_user(sk))
break;
icsk->icsk_backoff--;
icsk->icsk_rto = tp->srtt_us ? __tcp_set_rto(tp) :
TCP_TIMEOUT_INIT;
icsk->icsk_rto = inet_csk_rto_backoff(icsk, TCP_RTO_MAX);
skb = tcp_write_queue_head(sk);
BUG_ON(!skb);
remaining = icsk->icsk_rto -
min(icsk->icsk_rto,
tcp_time_stamp - tcp_skb_timestamp(skb));
if (remaining) {
inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS,
remaining, TCP_RTO_MAX);
} else {
/* RTO revert clocked out retransmission.
* Will retransmit now */
tcp_retransmit_timer(sk);
}
break;
case ICMP_TIME_EXCEEDED:
err = EHOSTUNREACH;
break;
default:
goto out;
}
switch (sk->sk_state) {
case TCP_SYN_SENT:
case TCP_SYN_RECV:
/* Only in fast or simultaneous open. If a fast open socket is
* is already accepted it is treated as a connected one below.
*/
if (fastopen && !fastopen->sk)
break;
if (!sock_owned_by_user(sk)) {
sk->sk_err = err;
sk->sk_error_report(sk);
tcp_done(sk);
} else {
sk->sk_err_soft = err;
}
goto out;
}
/* If we've already connected we will keep trying
* until we time out, or the user gives up.
*
* rfc1122 4.2.3.9 allows to consider as hard errors
* only PROTO_UNREACH and PORT_UNREACH (well, FRAG_FAILED too,
* but it is obsoleted by pmtu discovery).
*
* Note, that in modern internet, where routing is unreliable
* and in each dark corner broken firewalls sit, sending random
* errors ordered by their masters even this two messages finally lose
* their original sense (even Linux sends invalid PORT_UNREACHs)
*
* Now we are in compliance with RFCs.
* --ANK (980905)
*/
inet = inet_sk(sk);
if (!sock_owned_by_user(sk) && inet->recverr) {
sk->sk_err = err;
sk->sk_error_report(sk);
} else { /* Only an error on timeout */
sk->sk_err_soft = err;
}
out:
bh_unlock_sock(sk);
sock_put(sk);
}
void __tcp_v4_send_check(struct sk_buff *skb, __be32 saddr, __be32 daddr)
{
struct tcphdr *th = tcp_hdr(skb);
if (skb->ip_summed == CHECKSUM_PARTIAL) {
th->check = ~tcp_v4_check(skb->len, saddr, daddr, 0);
skb->csum_start = skb_transport_header(skb) - skb->head;
skb->csum_offset = offsetof(struct tcphdr, check);
} else {
th->check = tcp_v4_check(skb->len, saddr, daddr,
csum_partial(th,
th->doff << 2,
skb->csum));
}
}
/* This routine computes an IPv4 TCP checksum. */
void tcp_v4_send_check(struct sock *sk, struct sk_buff *skb)
{
const struct inet_sock *inet = inet_sk(sk);
__tcp_v4_send_check(skb, inet->inet_saddr, inet->inet_daddr);
}
EXPORT_SYMBOL(tcp_v4_send_check);
/*
* This routine will send an RST to the other tcp.
*
* Someone asks: why I NEVER use socket parameters (TOS, TTL etc.)
* for reset.
* Answer: if a packet caused RST, it is not for a socket
* existing in our system, if it is matched to a socket,
* it is just duplicate segment or bug in other side's TCP.
* So that we build reply only basing on parameters
* arrived with segment.
* Exception: precedence violation. We do not implement it in any case.
*/
static void tcp_v4_send_reset(const struct sock *sk, struct sk_buff *skb)
{
const struct tcphdr *th = tcp_hdr(skb);
struct {
struct tcphdr th;
#ifdef CONFIG_TCP_MD5SIG
__be32 opt[(TCPOLEN_MD5SIG_ALIGNED >> 2)];
#endif
} rep;
struct ip_reply_arg arg;
#ifdef CONFIG_TCP_MD5SIG
struct tcp_md5sig_key *key = NULL;
const __u8 *hash_location = NULL;
unsigned char newhash[16];
int genhash;
struct sock *sk1 = NULL;
#endif
struct net *net;
/* Never send a reset in response to a reset. */
if (th->rst)
return;
/* If sk not NULL, it means we did a successful lookup and incoming
* route had to be correct. prequeue might have dropped our dst.
*/
if (!sk && skb_rtable(skb)->rt_type != RTN_LOCAL)
return;
/* Swap the send and the receive. */
memset(&rep, 0, sizeof(rep));
rep.th.dest = th->source;
rep.th.source = th->dest;
rep.th.doff = sizeof(struct tcphdr) / 4;
rep.th.rst = 1;
if (th->ack) {
rep.th.seq = th->ack_seq;
} else {
rep.th.ack = 1;
rep.th.ack_seq = htonl(ntohl(th->seq) + th->syn + th->fin +
skb->len - (th->doff << 2));
}
memset(&arg, 0, sizeof(arg));
arg.iov[0].iov_base = (unsigned char *)&rep;
arg.iov[0].iov_len = sizeof(rep.th);
net = sk ? sock_net(sk) : dev_net(skb_dst(skb)->dev);
#ifdef CONFIG_TCP_MD5SIG
rcu_read_lock();
hash_location = tcp_parse_md5sig_option(th);
if (sk && sk_fullsock(sk)) {
key = tcp_md5_do_lookup(sk, (union tcp_md5_addr *)
&ip_hdr(skb)->saddr, AF_INET);
} else if (hash_location) {
/*
* active side is lost. Try to find listening socket through
* source port, and then find md5 key through listening socket.
* we are not loose security here:
* Incoming packet is checked with md5 hash with finding key,
* no RST generated if md5 hash doesn't match.
*/
sk1 = __inet_lookup_listener(net, &tcp_hashinfo, NULL, 0,
ip_hdr(skb)->saddr,
th->source, ip_hdr(skb)->daddr,
ntohs(th->source), inet_iif(skb));
/* don't send rst if it can't find key */
if (!sk1)
goto out;
key = tcp_md5_do_lookup(sk1, (union tcp_md5_addr *)
&ip_hdr(skb)->saddr, AF_INET);
if (!key)
goto out;
genhash = tcp_v4_md5_hash_skb(newhash, key, NULL, skb);
if (genhash || memcmp(hash_location, newhash, 16) != 0)
goto out;
}
if (key) {
rep.opt[0] = htonl((TCPOPT_NOP << 24) |
(TCPOPT_NOP << 16) |
(TCPOPT_MD5SIG << 8) |
TCPOLEN_MD5SIG);
/* Update length and the length the header thinks exists */
arg.iov[0].iov_len += TCPOLEN_MD5SIG_ALIGNED;
rep.th.doff = arg.iov[0].iov_len / 4;
tcp_v4_md5_hash_hdr((__u8 *) &rep.opt[1],
key, ip_hdr(skb)->saddr,
ip_hdr(skb)->daddr, &rep.th);
}
#endif
arg.csum = csum_tcpudp_nofold(ip_hdr(skb)->daddr,
ip_hdr(skb)->saddr, /* XXX */
arg.iov[0].iov_len, IPPROTO_TCP, 0);
arg.csumoffset = offsetof(struct tcphdr, check) / 2;
arg.flags = (sk && inet_sk_transparent(sk)) ? IP_REPLY_ARG_NOSRCCHECK : 0;
/* When socket is gone, all binding information is lost.
* routing might fail in this case. No choice here, if we choose to force
* input interface, we will misroute in case of asymmetric route.
*/
if (sk)
arg.bound_dev_if = sk->sk_bound_dev_if;
BUILD_BUG_ON(offsetof(struct sock, sk_bound_dev_if) !=
offsetof(struct inet_timewait_sock, tw_bound_dev_if));
arg.tos = ip_hdr(skb)->tos;
preempt_disable();
ip_send_unicast_reply(*this_cpu_ptr(net->ipv4.tcp_sk),
skb, &TCP_SKB_CB(skb)->header.h4.opt,
ip_hdr(skb)->saddr, ip_hdr(skb)->daddr,
&arg, arg.iov[0].iov_len);
__TCP_INC_STATS(net, TCP_MIB_OUTSEGS);
__TCP_INC_STATS(net, TCP_MIB_OUTRSTS);
preempt_enable();
#ifdef CONFIG_TCP_MD5SIG
out:
rcu_read_unlock();
#endif
}
/* The code following below sending ACKs in SYN-RECV and TIME-WAIT states
outside socket context is ugly, certainly. What can I do?
*/
tcp: fix NULL deref in tcp_v4_send_ack() Neal reported crashes with this stack trace : RIP: 0010:[<ffffffff8c57231b>] tcp_v4_send_ack+0x41/0x20f ... CR2: 0000000000000018 CR3: 000000044005c000 CR4: 00000000001427e0 ... [<ffffffff8c57258e>] tcp_v4_reqsk_send_ack+0xa5/0xb4 [<ffffffff8c1a7caa>] tcp_check_req+0x2ea/0x3e0 [<ffffffff8c19e420>] tcp_rcv_state_process+0x850/0x2500 [<ffffffff8c1a6d21>] tcp_v4_do_rcv+0x141/0x330 [<ffffffff8c56cdb2>] sk_backlog_rcv+0x21/0x30 [<ffffffff8c098bbd>] tcp_recvmsg+0x75d/0xf90 [<ffffffff8c0a8700>] inet_recvmsg+0x80/0xa0 [<ffffffff8c17623e>] sock_aio_read+0xee/0x110 [<ffffffff8c066fcf>] do_sync_read+0x6f/0xa0 [<ffffffff8c0673a1>] SyS_read+0x1e1/0x290 [<ffffffff8c5ca262>] system_call_fastpath+0x16/0x1b The problem here is the skb we provide to tcp_v4_send_ack() had to be parked in the backlog of a new TCP fastopen child because this child was owned by the user at the time an out of window packet arrived. Before queuing a packet, TCP has to set skb->dev to NULL as the device could disappear before packet is removed from the queue. Fix this issue by using the net pointer provided by the socket (being a timewait or a request socket). IPv6 is immune to the bug : tcp_v6_send_response() already gets the net pointer from the socket if provided. Fixes: 168a8f58059a ("tcp: TCP Fast Open Server - main code path") Reported-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Jerry Chu <hkchu@google.com> Cc: Yuchung Cheng <ycheng@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-01-21 16:02:54 +00:00
static void tcp_v4_send_ack(struct net *net,
struct sk_buff *skb, u32 seq, u32 ack,
u32 win, u32 tsval, u32 tsecr, int oif,
struct tcp_md5sig_key *key,
int reply_flags, u8 tos)
{
const struct tcphdr *th = tcp_hdr(skb);
struct {
struct tcphdr th;
__be32 opt[(TCPOLEN_TSTAMP_ALIGNED >> 2)
#ifdef CONFIG_TCP_MD5SIG
+ (TCPOLEN_MD5SIG_ALIGNED >> 2)
#endif
];
} rep;
struct ip_reply_arg arg;
memset(&rep.th, 0, sizeof(struct tcphdr));
memset(&arg, 0, sizeof(arg));
arg.iov[0].iov_base = (unsigned char *)&rep;
arg.iov[0].iov_len = sizeof(rep.th);
if (tsecr) {
rep.opt[0] = htonl((TCPOPT_NOP << 24) | (TCPOPT_NOP << 16) |
(TCPOPT_TIMESTAMP << 8) |
TCPOLEN_TIMESTAMP);
rep.opt[1] = htonl(tsval);
rep.opt[2] = htonl(tsecr);
arg.iov[0].iov_len += TCPOLEN_TSTAMP_ALIGNED;
}
/* Swap the send and the receive. */
rep.th.dest = th->source;
rep.th.source = th->dest;
rep.th.doff = arg.iov[0].iov_len / 4;
rep.th.seq = htonl(seq);
rep.th.ack_seq = htonl(ack);
rep.th.ack = 1;
rep.th.window = htons(win);
#ifdef CONFIG_TCP_MD5SIG
if (key) {
int offset = (tsecr) ? 3 : 0;
rep.opt[offset++] = htonl((TCPOPT_NOP << 24) |
(TCPOPT_NOP << 16) |
(TCPOPT_MD5SIG << 8) |
TCPOLEN_MD5SIG);
arg.iov[0].iov_len += TCPOLEN_MD5SIG_ALIGNED;
rep.th.doff = arg.iov[0].iov_len/4;
tcp_v4_md5_hash_hdr((__u8 *) &rep.opt[offset],
key, ip_hdr(skb)->saddr,
ip_hdr(skb)->daddr, &rep.th);
}
#endif
arg.flags = reply_flags;
arg.csum = csum_tcpudp_nofold(ip_hdr(skb)->daddr,
ip_hdr(skb)->saddr, /* XXX */
arg.iov[0].iov_len, IPPROTO_TCP, 0);
arg.csumoffset = offsetof(struct tcphdr, check) / 2;
if (oif)
arg.bound_dev_if = oif;
arg.tos = tos;
preempt_disable();
ip_send_unicast_reply(*this_cpu_ptr(net->ipv4.tcp_sk),
skb, &TCP_SKB_CB(skb)->header.h4.opt,
ip_hdr(skb)->saddr, ip_hdr(skb)->daddr,
&arg, arg.iov[0].iov_len);
__TCP_INC_STATS(net, TCP_MIB_OUTSEGS);
preempt_enable();
}
static void tcp_v4_timewait_ack(struct sock *sk, struct sk_buff *skb)
{
struct inet_timewait_sock *tw = inet_twsk(sk);
struct tcp_timewait_sock *tcptw = tcp_twsk(sk);
tcp: fix NULL deref in tcp_v4_send_ack() Neal reported crashes with this stack trace : RIP: 0010:[<ffffffff8c57231b>] tcp_v4_send_ack+0x41/0x20f ... CR2: 0000000000000018 CR3: 000000044005c000 CR4: 00000000001427e0 ... [<ffffffff8c57258e>] tcp_v4_reqsk_send_ack+0xa5/0xb4 [<ffffffff8c1a7caa>] tcp_check_req+0x2ea/0x3e0 [<ffffffff8c19e420>] tcp_rcv_state_process+0x850/0x2500 [<ffffffff8c1a6d21>] tcp_v4_do_rcv+0x141/0x330 [<ffffffff8c56cdb2>] sk_backlog_rcv+0x21/0x30 [<ffffffff8c098bbd>] tcp_recvmsg+0x75d/0xf90 [<ffffffff8c0a8700>] inet_recvmsg+0x80/0xa0 [<ffffffff8c17623e>] sock_aio_read+0xee/0x110 [<ffffffff8c066fcf>] do_sync_read+0x6f/0xa0 [<ffffffff8c0673a1>] SyS_read+0x1e1/0x290 [<ffffffff8c5ca262>] system_call_fastpath+0x16/0x1b The problem here is the skb we provide to tcp_v4_send_ack() had to be parked in the backlog of a new TCP fastopen child because this child was owned by the user at the time an out of window packet arrived. Before queuing a packet, TCP has to set skb->dev to NULL as the device could disappear before packet is removed from the queue. Fix this issue by using the net pointer provided by the socket (being a timewait or a request socket). IPv6 is immune to the bug : tcp_v6_send_response() already gets the net pointer from the socket if provided. Fixes: 168a8f58059a ("tcp: TCP Fast Open Server - main code path") Reported-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Jerry Chu <hkchu@google.com> Cc: Yuchung Cheng <ycheng@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-01-21 16:02:54 +00:00
tcp_v4_send_ack(sock_net(sk), skb,
tcptw->tw_snd_nxt, tcptw->tw_rcv_nxt,
tcptw->tw_rcv_wnd >> tw->tw_rcv_wscale,
tcp_time_stamp + tcptw->tw_ts_offset,
tcptw->tw_ts_recent,
tw->tw_bound_dev_if,
tcp_twsk_md5_key(tcptw),
tw->tw_transparent ? IP_REPLY_ARG_NOSRCCHECK : 0,
tw->tw_tos
);
inet_twsk_put(tw);
}
static void tcp_v4_reqsk_send_ack(const struct sock *sk, struct sk_buff *skb,
struct request_sock *req)
{
/* sk->sk_state == TCP_LISTEN -> for regular TCP_SYN_RECV
* sk->sk_state == TCP_SYN_RECV -> for Fast Open.
*/
tcp: fix NULL deref in tcp_v4_send_ack() Neal reported crashes with this stack trace : RIP: 0010:[<ffffffff8c57231b>] tcp_v4_send_ack+0x41/0x20f ... CR2: 0000000000000018 CR3: 000000044005c000 CR4: 00000000001427e0 ... [<ffffffff8c57258e>] tcp_v4_reqsk_send_ack+0xa5/0xb4 [<ffffffff8c1a7caa>] tcp_check_req+0x2ea/0x3e0 [<ffffffff8c19e420>] tcp_rcv_state_process+0x850/0x2500 [<ffffffff8c1a6d21>] tcp_v4_do_rcv+0x141/0x330 [<ffffffff8c56cdb2>] sk_backlog_rcv+0x21/0x30 [<ffffffff8c098bbd>] tcp_recvmsg+0x75d/0xf90 [<ffffffff8c0a8700>] inet_recvmsg+0x80/0xa0 [<ffffffff8c17623e>] sock_aio_read+0xee/0x110 [<ffffffff8c066fcf>] do_sync_read+0x6f/0xa0 [<ffffffff8c0673a1>] SyS_read+0x1e1/0x290 [<ffffffff8c5ca262>] system_call_fastpath+0x16/0x1b The problem here is the skb we provide to tcp_v4_send_ack() had to be parked in the backlog of a new TCP fastopen child because this child was owned by the user at the time an out of window packet arrived. Before queuing a packet, TCP has to set skb->dev to NULL as the device could disappear before packet is removed from the queue. Fix this issue by using the net pointer provided by the socket (being a timewait or a request socket). IPv6 is immune to the bug : tcp_v6_send_response() already gets the net pointer from the socket if provided. Fixes: 168a8f58059a ("tcp: TCP Fast Open Server - main code path") Reported-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Jerry Chu <hkchu@google.com> Cc: Yuchung Cheng <ycheng@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-01-21 16:02:54 +00:00
u32 seq = (sk->sk_state == TCP_LISTEN) ? tcp_rsk(req)->snt_isn + 1 :
tcp_sk(sk)->snd_nxt;
tcp_v4_send_ack(sock_net(sk), skb, seq,
tcp_rsk(req)->rcv_nxt, req->rsk_rcv_wnd,
tcp_time_stamp,
req->ts_recent,
0,
tcp_md5_do_lookup(sk, (union tcp_md5_addr *)&ip_hdr(skb)->daddr,
AF_INET),
inet_rsk(req)->no_srccheck ? IP_REPLY_ARG_NOSRCCHECK : 0,
ip_hdr(skb)->tos);
}
/*
* Send a SYN-ACK after having received a SYN.
* This still operates on a request_sock only, not on a big
* socket.
*/
static int tcp_v4_send_synack(const struct sock *sk, struct dst_entry *dst,
struct flowi *fl,
struct request_sock *req,
struct tcp_fastopen_cookie *foc,
enum tcp_synack_type synack_type)
{
const struct inet_request_sock *ireq = inet_rsk(req);
struct flowi4 fl4;
int err = -1;
struct sk_buff *skb;
/* First, grab a route. */
if (!dst && (dst = inet_csk_route_req(sk, &fl4, req)) == NULL)
return -1;
skb = tcp_make_synack(sk, dst, req, foc, synack_type);
if (skb) {
__tcp_v4_send_check(skb, ireq->ir_loc_addr, ireq->ir_rmt_addr);
err = ip_build_and_send_pkt(skb, sk, ireq->ir_loc_addr,
ireq->ir_rmt_addr,
ireq->opt);
err = net_xmit_eval(err);
}
return err;
}
/*
* IPv4 request_sock destructor.
*/
static void tcp_v4_reqsk_destructor(struct request_sock *req)
{
kfree(inet_rsk(req)->opt);
}
#ifdef CONFIG_TCP_MD5SIG
/*
* RFC2385 MD5 checksumming requires a mapping of
* IP address->MD5 Key.
* We need to maintain these in the sk structure.
*/
/* Find the Key structure for an address. */
struct tcp_md5sig_key *tcp_md5_do_lookup(const struct sock *sk,
const union tcp_md5_addr *addr,
int family)
{
const struct tcp_sock *tp = tcp_sk(sk);
struct tcp_md5sig_key *key;
unsigned int size = sizeof(struct in_addr);
const struct tcp_md5sig_info *md5sig;
/* caller either holds rcu_read_lock() or socket lock */
md5sig = rcu_dereference_check(tp->md5sig_info,
lockdep_sock_is_held(sk));
if (!md5sig)
return NULL;
#if IS_ENABLED(CONFIG_IPV6)
if (family == AF_INET6)
size = sizeof(struct in6_addr);
#endif
hlist: drop the node parameter from iterators I'm not sure why, but the hlist for each entry iterators were conceived list_for_each_entry(pos, head, member) The hlist ones were greedy and wanted an extra parameter: hlist_for_each_entry(tpos, pos, head, member) Why did they need an extra pos parameter? I'm not quite sure. Not only they don't really need it, it also prevents the iterator from looking exactly like the list iterator, which is unfortunate. Besides the semantic patch, there was some manual work required: - Fix up the actual hlist iterators in linux/list.h - Fix up the declaration of other iterators based on the hlist ones. - A very small amount of places were using the 'node' parameter, this was modified to use 'obj->member' instead. - Coccinelle didn't handle the hlist_for_each_entry_safe iterator properly, so those had to be fixed up manually. The semantic patch which is mostly the work of Peter Senna Tschudin is here: @@ iterator name hlist_for_each_entry, hlist_for_each_entry_continue, hlist_for_each_entry_from, hlist_for_each_entry_rcu, hlist_for_each_entry_rcu_bh, hlist_for_each_entry_continue_rcu_bh, for_each_busy_worker, ax25_uid_for_each, ax25_for_each, inet_bind_bucket_for_each, sctp_for_each_hentry, sk_for_each, sk_for_each_rcu, sk_for_each_from, sk_for_each_safe, sk_for_each_bound, hlist_for_each_entry_safe, hlist_for_each_entry_continue_rcu, nr_neigh_for_each, nr_neigh_for_each_safe, nr_node_for_each, nr_node_for_each_safe, for_each_gfn_indirect_valid_sp, for_each_gfn_sp, for_each_host; type T; expression a,c,d,e; identifier b; statement S; @@ -T b; <+... when != b ( hlist_for_each_entry(a, - b, c, d) S | hlist_for_each_entry_continue(a, - b, c) S | hlist_for_each_entry_from(a, - b, c) S | hlist_for_each_entry_rcu(a, - b, c, d) S | hlist_for_each_entry_rcu_bh(a, - b, c, d) S | hlist_for_each_entry_continue_rcu_bh(a, - b, c) S | for_each_busy_worker(a, c, - b, d) S | ax25_uid_for_each(a, - b, c) S | ax25_for_each(a, - b, c) S | inet_bind_bucket_for_each(a, - b, c) S | sctp_for_each_hentry(a, - b, c) S | sk_for_each(a, - b, c) S | sk_for_each_rcu(a, - b, c) S | sk_for_each_from -(a, b) +(a) S + sk_for_each_from(a) S | sk_for_each_safe(a, - b, c, d) S | sk_for_each_bound(a, - b, c) S | hlist_for_each_entry_safe(a, - b, c, d, e) S | hlist_for_each_entry_continue_rcu(a, - b, c) S | nr_neigh_for_each(a, - b, c) S | nr_neigh_for_each_safe(a, - b, c, d) S | nr_node_for_each(a, - b, c) S | nr_node_for_each_safe(a, - b, c, d) S | - for_each_gfn_sp(a, c, d, b) S + for_each_gfn_sp(a, c, d) S | - for_each_gfn_indirect_valid_sp(a, c, d, b) S + for_each_gfn_indirect_valid_sp(a, c, d) S | for_each_host(a, - b, c) S | for_each_host_safe(a, - b, c, d) S | for_each_mesh_entry(a, - b, c, d) S ) ...+> [akpm@linux-foundation.org: drop bogus change from net/ipv4/raw.c] [akpm@linux-foundation.org: drop bogus hunk from net/ipv6/raw.c] [akpm@linux-foundation.org: checkpatch fixes] [akpm@linux-foundation.org: fix warnings] [akpm@linux-foudnation.org: redo intrusive kvm changes] Tested-by: Peter Senna Tschudin <peter.senna@gmail.com> Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com> Cc: Wu Fengguang <fengguang.wu@intel.com> Cc: Marcelo Tosatti <mtosatti@redhat.com> Cc: Gleb Natapov <gleb@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-02-28 01:06:00 +00:00
hlist_for_each_entry_rcu(key, &md5sig->head, node) {
if (key->family != family)
continue;
if (!memcmp(&key->addr, addr, size))
return key;
}
return NULL;
}
EXPORT_SYMBOL(tcp_md5_do_lookup);
struct tcp_md5sig_key *tcp_v4_md5_lookup(const struct sock *sk,
const struct sock *addr_sk)
{
const union tcp_md5_addr *addr;
addr = (const union tcp_md5_addr *)&addr_sk->sk_daddr;
return tcp_md5_do_lookup(sk, addr, AF_INET);
}
EXPORT_SYMBOL(tcp_v4_md5_lookup);
/* This can be called on a newly created socket, from other files */
int tcp_md5_do_add(struct sock *sk, const union tcp_md5_addr *addr,
int family, const u8 *newkey, u8 newkeylen, gfp_t gfp)
{
/* Add Key to the list */
struct tcp_md5sig_key *key;
struct tcp_sock *tp = tcp_sk(sk);
struct tcp_md5sig_info *md5sig;
key = tcp_md5_do_lookup(sk, addr, family);
if (key) {
/* Pre-existing entry - just update that one. */
memcpy(key->key, newkey, newkeylen);
key->keylen = newkeylen;
return 0;
}
md5sig = rcu_dereference_protected(tp->md5sig_info,
lockdep_sock_is_held(sk));
if (!md5sig) {
md5sig = kmalloc(sizeof(*md5sig), gfp);
if (!md5sig)
return -ENOMEM;
sk_nocaps_add(sk, NETIF_F_GSO_MASK);
INIT_HLIST_HEAD(&md5sig->head);
rcu_assign_pointer(tp->md5sig_info, md5sig);
}
key = sock_kmalloc(sk, sizeof(*key), gfp);
if (!key)
return -ENOMEM;
if (!tcp_alloc_md5sig_pool()) {
sock_kfree_s(sk, key, sizeof(*key));
return -ENOMEM;
}
memcpy(key->key, newkey, newkeylen);
key->keylen = newkeylen;
key->family = family;
memcpy(&key->addr, addr,
(family == AF_INET6) ? sizeof(struct in6_addr) :
sizeof(struct in_addr));
hlist_add_head_rcu(&key->node, &md5sig->head);
return 0;
}
EXPORT_SYMBOL(tcp_md5_do_add);
int tcp_md5_do_del(struct sock *sk, const union tcp_md5_addr *addr, int family)
{
struct tcp_md5sig_key *key;
key = tcp_md5_do_lookup(sk, addr, family);
if (!key)
return -ENOENT;
hlist_del_rcu(&key->node);
atomic_sub(sizeof(*key), &sk->sk_omem_alloc);
kfree_rcu(key, rcu);
return 0;
}
EXPORT_SYMBOL(tcp_md5_do_del);
static void tcp_clear_md5_list(struct sock *sk)
{
struct tcp_sock *tp = tcp_sk(sk);
struct tcp_md5sig_key *key;
hlist: drop the node parameter from iterators I'm not sure why, but the hlist for each entry iterators were conceived list_for_each_entry(pos, head, member) The hlist ones were greedy and wanted an extra parameter: hlist_for_each_entry(tpos, pos, head, member) Why did they need an extra pos parameter? I'm not quite sure. Not only they don't really need it, it also prevents the iterator from looking exactly like the list iterator, which is unfortunate. Besides the semantic patch, there was some manual work required: - Fix up the actual hlist iterators in linux/list.h - Fix up the declaration of other iterators based on the hlist ones. - A very small amount of places were using the 'node' parameter, this was modified to use 'obj->member' instead. - Coccinelle didn't handle the hlist_for_each_entry_safe iterator properly, so those had to be fixed up manually. The semantic patch which is mostly the work of Peter Senna Tschudin is here: @@ iterator name hlist_for_each_entry, hlist_for_each_entry_continue, hlist_for_each_entry_from, hlist_for_each_entry_rcu, hlist_for_each_entry_rcu_bh, hlist_for_each_entry_continue_rcu_bh, for_each_busy_worker, ax25_uid_for_each, ax25_for_each, inet_bind_bucket_for_each, sctp_for_each_hentry, sk_for_each, sk_for_each_rcu, sk_for_each_from, sk_for_each_safe, sk_for_each_bound, hlist_for_each_entry_safe, hlist_for_each_entry_continue_rcu, nr_neigh_for_each, nr_neigh_for_each_safe, nr_node_for_each, nr_node_for_each_safe, for_each_gfn_indirect_valid_sp, for_each_gfn_sp, for_each_host; type T; expression a,c,d,e; identifier b; statement S; @@ -T b; <+... when != b ( hlist_for_each_entry(a, - b, c, d) S | hlist_for_each_entry_continue(a, - b, c) S | hlist_for_each_entry_from(a, - b, c) S | hlist_for_each_entry_rcu(a, - b, c, d) S | hlist_for_each_entry_rcu_bh(a, - b, c, d) S | hlist_for_each_entry_continue_rcu_bh(a, - b, c) S | for_each_busy_worker(a, c, - b, d) S | ax25_uid_for_each(a, - b, c) S | ax25_for_each(a, - b, c) S | inet_bind_bucket_for_each(a, - b, c) S | sctp_for_each_hentry(a, - b, c) S | sk_for_each(a, - b, c) S | sk_for_each_rcu(a, - b, c) S | sk_for_each_from -(a, b) +(a) S + sk_for_each_from(a) S | sk_for_each_safe(a, - b, c, d) S | sk_for_each_bound(a, - b, c) S | hlist_for_each_entry_safe(a, - b, c, d, e) S | hlist_for_each_entry_continue_rcu(a, - b, c) S | nr_neigh_for_each(a, - b, c) S | nr_neigh_for_each_safe(a, - b, c, d) S | nr_node_for_each(a, - b, c) S | nr_node_for_each_safe(a, - b, c, d) S | - for_each_gfn_sp(a, c, d, b) S + for_each_gfn_sp(a, c, d) S | - for_each_gfn_indirect_valid_sp(a, c, d, b) S + for_each_gfn_indirect_valid_sp(a, c, d) S | for_each_host(a, - b, c) S | for_each_host_safe(a, - b, c, d) S | for_each_mesh_entry(a, - b, c, d) S ) ...+> [akpm@linux-foundation.org: drop bogus change from net/ipv4/raw.c] [akpm@linux-foundation.org: drop bogus hunk from net/ipv6/raw.c] [akpm@linux-foundation.org: checkpatch fixes] [akpm@linux-foundation.org: fix warnings] [akpm@linux-foudnation.org: redo intrusive kvm changes] Tested-by: Peter Senna Tschudin <peter.senna@gmail.com> Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com> Cc: Wu Fengguang <fengguang.wu@intel.com> Cc: Marcelo Tosatti <mtosatti@redhat.com> Cc: Gleb Natapov <gleb@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-02-28 01:06:00 +00:00
struct hlist_node *n;
struct tcp_md5sig_info *md5sig;
md5sig = rcu_dereference_protected(tp->md5sig_info, 1);
hlist: drop the node parameter from iterators I'm not sure why, but the hlist for each entry iterators were conceived list_for_each_entry(pos, head, member) The hlist ones were greedy and wanted an extra parameter: hlist_for_each_entry(tpos, pos, head, member) Why did they need an extra pos parameter? I'm not quite sure. Not only they don't really need it, it also prevents the iterator from looking exactly like the list iterator, which is unfortunate. Besides the semantic patch, there was some manual work required: - Fix up the actual hlist iterators in linux/list.h - Fix up the declaration of other iterators based on the hlist ones. - A very small amount of places were using the 'node' parameter, this was modified to use 'obj->member' instead. - Coccinelle didn't handle the hlist_for_each_entry_safe iterator properly, so those had to be fixed up manually. The semantic patch which is mostly the work of Peter Senna Tschudin is here: @@ iterator name hlist_for_each_entry, hlist_for_each_entry_continue, hlist_for_each_entry_from, hlist_for_each_entry_rcu, hlist_for_each_entry_rcu_bh, hlist_for_each_entry_continue_rcu_bh, for_each_busy_worker, ax25_uid_for_each, ax25_for_each, inet_bind_bucket_for_each, sctp_for_each_hentry, sk_for_each, sk_for_each_rcu, sk_for_each_from, sk_for_each_safe, sk_for_each_bound, hlist_for_each_entry_safe, hlist_for_each_entry_continue_rcu, nr_neigh_for_each, nr_neigh_for_each_safe, nr_node_for_each, nr_node_for_each_safe, for_each_gfn_indirect_valid_sp, for_each_gfn_sp, for_each_host; type T; expression a,c,d,e; identifier b; statement S; @@ -T b; <+... when != b ( hlist_for_each_entry(a, - b, c, d) S | hlist_for_each_entry_continue(a, - b, c) S | hlist_for_each_entry_from(a, - b, c) S | hlist_for_each_entry_rcu(a, - b, c, d) S | hlist_for_each_entry_rcu_bh(a, - b, c, d) S | hlist_for_each_entry_continue_rcu_bh(a, - b, c) S | for_each_busy_worker(a, c, - b, d) S | ax25_uid_for_each(a, - b, c) S | ax25_for_each(a, - b, c) S | inet_bind_bucket_for_each(a, - b, c) S | sctp_for_each_hentry(a, - b, c) S | sk_for_each(a, - b, c) S | sk_for_each_rcu(a, - b, c) S | sk_for_each_from -(a, b) +(a) S + sk_for_each_from(a) S | sk_for_each_safe(a, - b, c, d) S | sk_for_each_bound(a, - b, c) S | hlist_for_each_entry_safe(a, - b, c, d, e) S | hlist_for_each_entry_continue_rcu(a, - b, c) S | nr_neigh_for_each(a, - b, c) S | nr_neigh_for_each_safe(a, - b, c, d) S | nr_node_for_each(a, - b, c) S | nr_node_for_each_safe(a, - b, c, d) S | - for_each_gfn_sp(a, c, d, b) S + for_each_gfn_sp(a, c, d) S | - for_each_gfn_indirect_valid_sp(a, c, d, b) S + for_each_gfn_indirect_valid_sp(a, c, d) S | for_each_host(a, - b, c) S | for_each_host_safe(a, - b, c, d) S | for_each_mesh_entry(a, - b, c, d) S ) ...+> [akpm@linux-foundation.org: drop bogus change from net/ipv4/raw.c] [akpm@linux-foundation.org: drop bogus hunk from net/ipv6/raw.c] [akpm@linux-foundation.org: checkpatch fixes] [akpm@linux-foundation.org: fix warnings] [akpm@linux-foudnation.org: redo intrusive kvm changes] Tested-by: Peter Senna Tschudin <peter.senna@gmail.com> Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com> Cc: Wu Fengguang <fengguang.wu@intel.com> Cc: Marcelo Tosatti <mtosatti@redhat.com> Cc: Gleb Natapov <gleb@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-02-28 01:06:00 +00:00
hlist_for_each_entry_safe(key, n, &md5sig->head, node) {
hlist_del_rcu(&key->node);
atomic_sub(sizeof(*key), &sk->sk_omem_alloc);
kfree_rcu(key, rcu);
}
}
static int tcp_v4_parse_md5_keys(struct sock *sk, char __user *optval,
int optlen)
{
struct tcp_md5sig cmd;
struct sockaddr_in *sin = (struct sockaddr_in *)&cmd.tcpm_addr;
if (optlen < sizeof(cmd))
return -EINVAL;
if (copy_from_user(&cmd, optval, sizeof(cmd)))
return -EFAULT;
if (sin->sin_family != AF_INET)
return -EINVAL;
if (!cmd.tcpm_keylen)
return tcp_md5_do_del(sk, (union tcp_md5_addr *)&sin->sin_addr.s_addr,
AF_INET);
if (cmd.tcpm_keylen > TCP_MD5SIG_MAXKEYLEN)
return -EINVAL;
return tcp_md5_do_add(sk, (union tcp_md5_addr *)&sin->sin_addr.s_addr,
AF_INET, cmd.tcpm_key, cmd.tcpm_keylen,
GFP_KERNEL);
}
static int tcp_v4_md5_hash_pseudoheader(struct tcp_md5sig_pool *hp,
__be32 daddr, __be32 saddr, int nbytes)
{
struct tcp4_pseudohdr *bp;
struct scatterlist sg;
bp = &hp->md5_blk.ip4;
/*
* 1. the TCP pseudo-header (in the order: source IP address,
* destination IP address, zero-padded protocol number, and
* segment length)
*/
bp->saddr = saddr;
bp->daddr = daddr;
bp->pad = 0;
bp->protocol = IPPROTO_TCP;
bp->len = cpu_to_be16(nbytes);
sg_init_one(&sg, bp, sizeof(*bp));
ahash_request_set_crypt(hp->md5_req, &sg, NULL, sizeof(*bp));
return crypto_ahash_update(hp->md5_req);
}
static int tcp_v4_md5_hash_hdr(char *md5_hash, const struct tcp_md5sig_key *key,
__be32 daddr, __be32 saddr, const struct tcphdr *th)
{
struct tcp_md5sig_pool *hp;
struct ahash_request *req;
hp = tcp_get_md5sig_pool();
if (!hp)
goto clear_hash_noput;
req = hp->md5_req;
if (crypto_ahash_init(req))
goto clear_hash;
if (tcp_v4_md5_hash_pseudoheader(hp, daddr, saddr, th->doff << 2))
goto clear_hash;
if (tcp_md5_hash_header(hp, th))
goto clear_hash;
if (tcp_md5_hash_key(hp, key))
goto clear_hash;
ahash_request_set_crypt(req, NULL, md5_hash, 0);
if (crypto_ahash_final(req))
goto clear_hash;
tcp_put_md5sig_pool();
return 0;
clear_hash:
tcp_put_md5sig_pool();
clear_hash_noput:
memset(md5_hash, 0, 16);
return 1;
}
int tcp_v4_md5_hash_skb(char *md5_hash, const struct tcp_md5sig_key *key,
const struct sock *sk,
const struct sk_buff *skb)
{
struct tcp_md5sig_pool *hp;
struct ahash_request *req;
const struct tcphdr *th = tcp_hdr(skb);
__be32 saddr, daddr;
if (sk) { /* valid for establish/request sockets */
saddr = sk->sk_rcv_saddr;
daddr = sk->sk_daddr;
} else {
const struct iphdr *iph = ip_hdr(skb);
saddr = iph->saddr;
daddr = iph->daddr;
}
hp = tcp_get_md5sig_pool();
if (!hp)
goto clear_hash_noput;
req = hp->md5_req;
if (crypto_ahash_init(req))
goto clear_hash;
if (tcp_v4_md5_hash_pseudoheader(hp, daddr, saddr, skb->len))
goto clear_hash;
if (tcp_md5_hash_header(hp, th))
goto clear_hash;
if (tcp_md5_hash_skb_data(hp, skb, th->doff << 2))
goto clear_hash;
if (tcp_md5_hash_key(hp, key))
goto clear_hash;
ahash_request_set_crypt(req, NULL, md5_hash, 0);
if (crypto_ahash_final(req))
goto clear_hash;
tcp_put_md5sig_pool();
return 0;
clear_hash:
tcp_put_md5sig_pool();
clear_hash_noput:
memset(md5_hash, 0, 16);
return 1;
}
EXPORT_SYMBOL(tcp_v4_md5_hash_skb);
#endif
/* Called with rcu_read_lock() */
static bool tcp_v4_inbound_md5_hash(const struct sock *sk,
const struct sk_buff *skb)
{
#ifdef CONFIG_TCP_MD5SIG
/*
* This gets called for each TCP segment that arrives
* so we want to be efficient.
* We have 3 drop cases:
* o No MD5 hash and one expected.
* o MD5 hash and we're not expecting one.
* o MD5 hash and its wrong.
*/
const __u8 *hash_location = NULL;
struct tcp_md5sig_key *hash_expected;
const struct iphdr *iph = ip_hdr(skb);
const struct tcphdr *th = tcp_hdr(skb);
int genhash;
unsigned char newhash[16];
hash_expected = tcp_md5_do_lookup(sk, (union tcp_md5_addr *)&iph->saddr,
AF_INET);
hash_location = tcp_parse_md5sig_option(th);
/* We've parsed the options - do we have a hash? */
if (!hash_expected && !hash_location)
return false;
if (hash_expected && !hash_location) {
NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPMD5NOTFOUND);
return true;
}
if (!hash_expected && hash_location) {
NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPMD5UNEXPECTED);
return true;
}
/* Okay, so this is hash_expected and hash_location -
* so we need to calculate the checksum.
*/
genhash = tcp_v4_md5_hash_skb(newhash,
hash_expected,
NULL, skb);
if (genhash || memcmp(hash_location, newhash, 16) != 0) {
net_info_ratelimited("MD5 Hash failed for (%pI4, %d)->(%pI4, %d)%s\n",
&iph->saddr, ntohs(th->source),
&iph->daddr, ntohs(th->dest),
genhash ? " tcp_v4_calc_md5_hash failed"
: "");
return true;
}
return false;
#endif
return false;
}
static void tcp_v4_init_req(struct request_sock *req,
const struct sock *sk_listener,
struct sk_buff *skb)
{
struct inet_request_sock *ireq = inet_rsk(req);
sk_rcv_saddr_set(req_to_sk(req), ip_hdr(skb)->daddr);
sk_daddr_set(req_to_sk(req), ip_hdr(skb)->saddr);
ireq->no_srccheck = inet_sk(sk_listener)->transparent;
ireq->opt = tcp_v4_save_options(skb);
}
static struct dst_entry *tcp_v4_route_req(const struct sock *sk,
struct flowi *fl,
const struct request_sock *req,
bool *strict)
{
struct dst_entry *dst = inet_csk_route_req(sk, &fl->u.ip4, req);
if (strict) {
if (fl->u.ip4.daddr == inet_rsk(req)->ir_rmt_addr)
*strict = true;
else
*strict = false;
}
return dst;
}
struct request_sock_ops tcp_request_sock_ops __read_mostly = {
.family = PF_INET,
.obj_size = sizeof(struct tcp_request_sock),
.rtx_syn_ack = tcp_rtx_synack,
.send_ack = tcp_v4_reqsk_send_ack,
.destructor = tcp_v4_reqsk_destructor,
.send_reset = tcp_v4_send_reset,
.syn_ack_timeout = tcp_syn_ack_timeout,
};
static const struct tcp_request_sock_ops tcp_request_sock_ipv4_ops = {
.mss_clamp = TCP_MSS_DEFAULT,
#ifdef CONFIG_TCP_MD5SIG
.req_md5_lookup = tcp_v4_md5_lookup,
.calc_md5_hash = tcp_v4_md5_hash_skb,
#endif
.init_req = tcp_v4_init_req,
#ifdef CONFIG_SYN_COOKIES
.cookie_init_seq = cookie_v4_init_sequence,
#endif
.route_req = tcp_v4_route_req,
.init_seq = tcp_v4_init_sequence,
.send_synack = tcp_v4_send_synack,
};
int tcp_v4_conn_request(struct sock *sk, struct sk_buff *skb)
{
/* Never answer to SYNs send to broadcast or multicast */
if (skb_rtable(skb)->rt_flags & (RTCF_BROADCAST | RTCF_MULTICAST))
goto drop;
return tcp_conn_request(&tcp_request_sock_ops,
&tcp_request_sock_ipv4_ops, sk, skb);
drop:
tcp_listendrop(sk);
return 0;
}
EXPORT_SYMBOL(tcp_v4_conn_request);
/*
* The three way handshake has completed - we got a valid synack -
* now create the new socket.
*/
struct sock *tcp_v4_syn_recv_sock(const struct sock *sk, struct sk_buff *skb,
struct request_sock *req,
struct dst_entry *dst,
struct request_sock *req_unhash,
bool *own_req)
{
struct inet_request_sock *ireq;
struct inet_sock *newinet;
struct tcp_sock *newtp;
struct sock *newsk;
#ifdef CONFIG_TCP_MD5SIG
struct tcp_md5sig_key *key;
#endif
struct ip_options_rcu *inet_opt;
if (sk_acceptq_is_full(sk))
goto exit_overflow;
newsk = tcp_create_openreq_child(sk, req, skb);
if (!newsk)
goto exit_nonewsk;
newsk->sk_gso_type = SKB_GSO_TCPV4;
inet_sk_rx_dst_set(newsk, skb);
newtp = tcp_sk(newsk);
newinet = inet_sk(newsk);
ireq = inet_rsk(req);
sk_daddr_set(newsk, ireq->ir_rmt_addr);
sk_rcv_saddr_set(newsk, ireq->ir_loc_addr);
newsk->sk_bound_dev_if = ireq->ir_iif;
newinet->inet_saddr = ireq->ir_loc_addr;
inet_opt = ireq->opt;
rcu_assign_pointer(newinet->inet_opt, inet_opt);
ireq->opt = NULL;
newinet->mc_index = inet_iif(skb);
newinet->mc_ttl = ip_hdr(skb)->ttl;
newinet->rcv_tos = ip_hdr(skb)->tos;
inet_csk(newsk)->icsk_ext_hdr_len = 0;
if (inet_opt)
inet_csk(newsk)->icsk_ext_hdr_len = inet_opt->opt.optlen;
newinet->inet_id = newtp->write_seq ^ jiffies;
if (!dst) {
dst = inet_csk_route_child_sock(sk, newsk, req);
if (!dst)
goto put_and_exit;
} else {
/* syncookie case : see end of cookie_v4_check() */
}
sk_setup_caps(newsk, dst);
net: tcp: add per route congestion control This work adds the possibility to define a per route/destination congestion control algorithm. Generally, this opens up the possibility for a machine with different links to enforce specific congestion control algorithms with optimal strategies for each of them based on their network characteristics, even transparently for a single application listening on all links. For our specific use case, this additionally facilitates deployment of DCTCP, for example, applications can easily serve internal traffic/dsts in DCTCP and external one with CUBIC. Other scenarios would also allow for utilizing e.g. long living, low priority background flows for certain destinations/routes while still being able for normal traffic to utilize the default congestion control algorithm. We also thought about a per netns setting (where different defaults are possible), but given its actually a link specific property, we argue that a per route/destination setting is the most natural and flexible. The administrator can utilize this through ip-route(8) by appending "congctl [lock] <name>", where <name> denotes the name of a congestion control algorithm and the optional lock parameter allows to enforce the given algorithm so that applications in user space would not be allowed to overwrite that algorithm for that destination. The dst metric lookups are being done when a dst entry is already available in order to avoid a costly lookup and still before the algorithms are being initialized, thus overhead is very low when the feature is not being used. While the client side would need to drop the current reference on the module, on server side this can actually even be avoided as we just got a flat-copied socket clone. Joint work with Florian Westphal. Suggested-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by: Florian Westphal <fw@strlen.de> Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-01-05 22:57:48 +00:00
tcp_ca_openreq_child(newsk, dst);
tcp_sync_mss(newsk, dst_mtu(dst));
newtp->advmss = dst_metric_advmss(dst);
if (tcp_sk(sk)->rx_opt.user_mss &&
tcp_sk(sk)->rx_opt.user_mss < newtp->advmss)
newtp->advmss = tcp_sk(sk)->rx_opt.user_mss;
tcp_initialize_rcv_mss(newsk);
#ifdef CONFIG_TCP_MD5SIG
/* Copy over the MD5 key from the original socket */
key = tcp_md5_do_lookup(sk, (union tcp_md5_addr *)&newinet->inet_daddr,
AF_INET);
if (key) {
/*
* We're using one, so create a matching key
* on the newsk structure. If we fail to get
* memory, then we end up not copying the key
* across. Shucks.
*/
tcp_md5_do_add(newsk, (union tcp_md5_addr *)&newinet->inet_daddr,
AF_INET, key->key, key->keylen, GFP_ATOMIC);
sk_nocaps_add(newsk, NETIF_F_GSO_MASK);
}
#endif
if (__inet_inherit_port(sk, newsk) < 0)
goto put_and_exit;
*own_req = inet_ehash_nolisten(newsk, req_to_sk(req_unhash));
if (*own_req)
tcp_move_syn(newtp, req);
return newsk;
exit_overflow:
NET_INC_STATS(sock_net(sk), LINUX_MIB_LISTENOVERFLOWS);
exit_nonewsk:
dst_release(dst);
exit:
tcp_listendrop(sk);
return NULL;
put_and_exit:
inet: Fix kmemleak in tcp_v4/6_syn_recv_sock and dccp_v4/6_request_recv_sock If in either of the above functions inet_csk_route_child_sock() or __inet_inherit_port() fails, the newsk will not be freed: unreferenced object 0xffff88022e8a92c0 (size 1592): comm "softirq", pid 0, jiffies 4294946244 (age 726.160s) hex dump (first 32 bytes): 0a 01 01 01 0a 01 01 02 00 00 00 00 a7 cc 16 00 ................ 02 00 03 01 00 00 00 00 00 00 00 00 00 00 00 00 ................ backtrace: [<ffffffff8153d190>] kmemleak_alloc+0x21/0x3e [<ffffffff810ab3e7>] kmem_cache_alloc+0xb5/0xc5 [<ffffffff8149b65b>] sk_prot_alloc.isra.53+0x2b/0xcd [<ffffffff8149b784>] sk_clone_lock+0x16/0x21e [<ffffffff814d711a>] inet_csk_clone_lock+0x10/0x7b [<ffffffff814ebbc3>] tcp_create_openreq_child+0x21/0x481 [<ffffffff814e8fa5>] tcp_v4_syn_recv_sock+0x3a/0x23b [<ffffffff814ec5ba>] tcp_check_req+0x29f/0x416 [<ffffffff814e8e10>] tcp_v4_do_rcv+0x161/0x2bc [<ffffffff814eb917>] tcp_v4_rcv+0x6c9/0x701 [<ffffffff814cea9f>] ip_local_deliver_finish+0x70/0xc4 [<ffffffff814cec20>] ip_local_deliver+0x4e/0x7f [<ffffffff814ce9f8>] ip_rcv_finish+0x1fc/0x233 [<ffffffff814cee68>] ip_rcv+0x217/0x267 [<ffffffff814a7bbe>] __netif_receive_skb+0x49e/0x553 [<ffffffff814a7cc3>] netif_receive_skb+0x50/0x82 This happens, because sk_clone_lock initializes sk_refcnt to 2, and thus a single sock_put() is not enough to free the memory. Additionally, things like xfrm, memcg, cookie_values,... may have been initialized. We have to free them properly. This is fixed by forcing a call to tcp_done(), ending up in inet_csk_destroy_sock, doing the final sock_put(). tcp_done() is necessary, because it ends up doing all the cleanup on xfrm, memcg, cookie_values, xfrm,... Before calling tcp_done, we have to set the socket to SOCK_DEAD, to force it entering inet_csk_destroy_sock. To avoid the warning in inet_csk_destroy_sock, inet_num has to be set to 0. As inet_csk_destroy_sock does a dec on orphan_count, we first have to increase it. Calling tcp_done() allows us to remove the calls to tcp_clear_xmit_timer() and tcp_cleanup_congestion_control(). A similar approach is taken for dccp by calling dccp_done(). This is in the kernel since 093d282321 (tproxy: fix hash locking issue when using port redirection in __inet_inherit_port()), thus since version >= 2.6.37. Signed-off-by: Christoph Paasch <christoph.paasch@uclouvain.be> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-12-14 04:07:58 +00:00
inet_csk_prepare_forced_close(newsk);
tcp_done(newsk);
goto exit;
}
EXPORT_SYMBOL(tcp_v4_syn_recv_sock);
static struct sock *tcp_v4_cookie_check(struct sock *sk, struct sk_buff *skb)
{
#ifdef CONFIG_SYN_COOKIES
const struct tcphdr *th = tcp_hdr(skb);
if (!th->syn)
sk = cookie_v4_check(sk, skb);
#endif
return sk;
}
/* The socket must have it's spinlock held when we get
* here, unless it is a TCP_LISTEN socket.
*
* We have a potential double-lock case here, so even when
* doing backlog processing we use the BH locking scheme.
* This is because we cannot sleep with the original spinlock
* held.
*/
int tcp_v4_do_rcv(struct sock *sk, struct sk_buff *skb)
{
struct sock *rsk;
if (sk->sk_state == TCP_ESTABLISHED) { /* Fast path */
struct dst_entry *dst = sk->sk_rx_dst;
sock_rps_save_rxhash(sk, skb);
sk_mark_napi_id(sk, skb);
if (dst) {
if (inet_sk(sk)->rx_dst_ifindex != skb->skb_iif ||
!dst->ops->check(dst, 0)) {
dst_release(dst);
sk->sk_rx_dst = NULL;
}
}
tcp_rcv_established(sk, skb, tcp_hdr(skb), skb->len);
return 0;
}
if (tcp_checksum_complete(skb))
goto csum_err;
if (sk->sk_state == TCP_LISTEN) {
struct sock *nsk = tcp_v4_cookie_check(sk, skb);
if (!nsk)
goto discard;
if (nsk != sk) {
sock_rps_save_rxhash(nsk, skb);
sk_mark_napi_id(nsk, skb);
if (tcp_child_process(sk, nsk, skb)) {
rsk = nsk;
goto reset;
}
return 0;
}
} else
sock_rps_save_rxhash(sk, skb);
if (tcp_rcv_state_process(sk, skb)) {
rsk = sk;
goto reset;
}
return 0;
reset:
tcp_v4_send_reset(rsk, skb);
discard:
kfree_skb(skb);
/* Be careful here. If this function gets more complicated and
* gcc suffers from register pressure on the x86, sk (in %ebx)
* might be destroyed here. This current version compiles correctly,
* but you have been warned.
*/
return 0;
csum_err:
TCP_INC_STATS(sock_net(sk), TCP_MIB_CSUMERRORS);
TCP_INC_STATS(sock_net(sk), TCP_MIB_INERRS);
goto discard;
}
EXPORT_SYMBOL(tcp_v4_do_rcv);
void tcp_v4_early_demux(struct sk_buff *skb)
{
const struct iphdr *iph;
const struct tcphdr *th;
struct sock *sk;
if (skb->pkt_type != PACKET_HOST)
return;
if (!pskb_may_pull(skb, skb_transport_offset(skb) + sizeof(struct tcphdr)))
return;
iph = ip_hdr(skb);
th = tcp_hdr(skb);
if (th->doff < sizeof(struct tcphdr) / 4)
return;
sk = __inet_lookup_established(dev_net(skb->dev), &tcp_hashinfo,
iph->saddr, th->source,
iph->daddr, ntohs(th->dest),
skb->skb_iif);
if (sk) {
skb->sk = sk;
skb->destructor = sock_edemux;
if (sk_fullsock(sk)) {
struct dst_entry *dst = READ_ONCE(sk->sk_rx_dst);
if (dst)
dst = dst_check(dst, 0);
if (dst &&
inet_sk(sk)->rx_dst_ifindex == skb->skb_iif)
skb_dst_set_noref(skb, dst);
}
}
}
/* Packet is added to VJ-style prequeue for processing in process
* context, if a reader task is waiting. Apparently, this exciting
* idea (VJ's mail "Re: query about TCP header on tcp-ip" of 07 Sep 93)
* failed somewhere. Latency? Burstiness? Well, at least now we will
* see, why it failed. 8)8) --ANK
*
*/
bool tcp_prequeue(struct sock *sk, struct sk_buff *skb)
{
struct tcp_sock *tp = tcp_sk(sk);
if (sysctl_tcp_low_latency || !tp->ucopy.task)
return false;
if (skb->len <= tcp_hdrlen(skb) &&
skb_queue_len(&tp->ucopy.prequeue) == 0)
return false;
tcp: remove dst refcount false sharing for prequeue mode Alexander Duyck reported high false sharing on dst refcount in tcp stack when prequeue is used. prequeue is the mechanism used when a thread is blocked in recvmsg()/read() on a TCP socket, using a blocking model rather than select()/poll()/epoll() non blocking one. We already try to use RCU in input path as much as possible, but we were forced to take a refcount on the dst when skb escaped RCU protected region. When/if the user thread runs on different cpu, dst_release() will then touch dst refcount again. Commit 093162553c33 (tcp: force a dst refcount when prequeue packet) was an example of a race fix. It turns out the only remaining usage of skb->dst for a packet stored in a TCP socket prequeue is IP early demux. We can add a logic to detect when IP early demux is probably going to use skb->dst. Because we do an optimistic check rather than duplicate existing logic, we need to guard inet_sk_rx_dst_set() and inet6_sk_rx_dst_set() from using a NULL dst. Many thanks to Alexander for providing a nice bug report, git bisection, and reproducer. Tested using Alexander script on a 40Gb NIC, 8 RX queues. Hosts have 24 cores, 48 hyper threads. echo 0 >/proc/sys/net/ipv4/tcp_autocorking for i in `seq 0 47` do for j in `seq 0 2` do netperf -H $DEST -t TCP_STREAM -l 1000 \ -c -C -T $i,$i -P 0 -- \ -m 64 -s 64K -D & done done Before patch : ~6Mpps and ~95% cpu usage on receiver After patch : ~9Mpps and ~35% cpu usage on receiver. Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-09-08 15:06:07 +00:00
/* Before escaping RCU protected region, we need to take care of skb
* dst. Prequeue is only enabled for established sockets.
* For such sockets, we might need the skb dst only to set sk->sk_rx_dst
* Instead of doing full sk_rx_dst validity here, let's perform
* an optimistic check.
*/
if (likely(sk->sk_rx_dst))
skb_dst_drop(skb);
else
net: fix IP early demux races David Wilder reported crashes caused by dst reuse. <quote David> I am seeing a crash on a distro V4.2.3 kernel caused by a double release of a dst_entry. In ipv4_dst_destroy() the call to list_empty() finds a poisoned next pointer, indicating the dst_entry has already been removed from the list and freed. The crash occurs 18 to 24 hours into a run of a network stress exerciser. </quote> Thanks to his detailed report and analysis, we were able to understand the core issue. IP early demux can associate a dst to skb, after a lookup in TCP/UDP sockets. When socket cache is not properly set, we want to store into sk->sk_dst_cache the dst for future IP early demux lookups, by acquiring a stable refcount on the dst. Problem is this acquisition is simply using an atomic_inc(), which works well, unless the dst was queued for destruction from dst_release() noticing dst refcount went to zero, if DST_NOCACHE was set on dst. We need to make sure current refcount is not zero before incrementing it, or risk double free as David reported. This patch, being a stable candidate, adds two new helpers, and use them only from IP early demux problematic paths. It might be possible to merge in net-next skb_dst_force() and skb_dst_force_safe(), but I prefer having the smallest patch for stable kernels : Maybe some skb_dst_force() callers do not expect skb->dst can suddenly be cleared. Can probably be backported back to linux-3.6 kernels Reported-by: David J. Wilder <dwilder@us.ibm.com> Tested-by: David J. Wilder <dwilder@us.ibm.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-12-14 22:08:53 +00:00
skb_dst_force_safe(skb);
tcp: remove dst refcount false sharing for prequeue mode Alexander Duyck reported high false sharing on dst refcount in tcp stack when prequeue is used. prequeue is the mechanism used when a thread is blocked in recvmsg()/read() on a TCP socket, using a blocking model rather than select()/poll()/epoll() non blocking one. We already try to use RCU in input path as much as possible, but we were forced to take a refcount on the dst when skb escaped RCU protected region. When/if the user thread runs on different cpu, dst_release() will then touch dst refcount again. Commit 093162553c33 (tcp: force a dst refcount when prequeue packet) was an example of a race fix. It turns out the only remaining usage of skb->dst for a packet stored in a TCP socket prequeue is IP early demux. We can add a logic to detect when IP early demux is probably going to use skb->dst. Because we do an optimistic check rather than duplicate existing logic, we need to guard inet_sk_rx_dst_set() and inet6_sk_rx_dst_set() from using a NULL dst. Many thanks to Alexander for providing a nice bug report, git bisection, and reproducer. Tested using Alexander script on a 40Gb NIC, 8 RX queues. Hosts have 24 cores, 48 hyper threads. echo 0 >/proc/sys/net/ipv4/tcp_autocorking for i in `seq 0 47` do for j in `seq 0 2` do netperf -H $DEST -t TCP_STREAM -l 1000 \ -c -C -T $i,$i -P 0 -- \ -m 64 -s 64K -D & done done Before patch : ~6Mpps and ~95% cpu usage on receiver After patch : ~9Mpps and ~35% cpu usage on receiver. Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-09-08 15:06:07 +00:00
__skb_queue_tail(&tp->ucopy.prequeue, skb);
tp->ucopy.memory += skb->truesize;
tcp: give prequeue mode some care TCP prequeue goal is to defer processing of incoming packets to user space thread currently blocked in a recvmsg() system call. Intent is to spend less time processing these packets on behalf of softirq handler, as softirq handler is unfair to normal process scheduler decisions, as it might interrupt threads that do not even use networking. Current prequeue implementation has following issues : 1) It only checks size of the prequeue against sk_rcvbuf It was fine 15 years ago when sk_rcvbuf was in the 64KB vicinity. But we now have ~8MB values to cope with modern networking needs. We have to add sk_rmem_alloc in the equation, since out of order packets can definitely use up to sk_rcvbuf memory themselves. 2) Even with a fixed memory truesize check, prequeue can be filled by thousands of packets. When prequeue needs to be flushed, either from sofirq context (in tcp_prequeue() or timer code), or process context (in tcp_prequeue_process()), this adds a latency spike which is often not desirable. I added a fixed limit of 32 packets, as this translated to a max flush time of 60 us on my test hosts. Also note that all packets in prequeue are not accounted for tcp_mem, since they are not charged against sk_forward_alloc at this point. This is probably not a big deal. Note that this might increase LINUX_MIB_TCPPREQUEUEDROPPED counts, which is misnamed, as packets are not dropped at all, but rather pushed to the stack (where they can be either consumed or dropped) Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-27 17:12:25 +00:00
if (skb_queue_len(&tp->ucopy.prequeue) >= 32 ||
tp->ucopy.memory + atomic_read(&sk->sk_rmem_alloc) > sk->sk_rcvbuf) {
struct sk_buff *skb1;
BUG_ON(sock_owned_by_user(sk));
tcp: give prequeue mode some care TCP prequeue goal is to defer processing of incoming packets to user space thread currently blocked in a recvmsg() system call. Intent is to spend less time processing these packets on behalf of softirq handler, as softirq handler is unfair to normal process scheduler decisions, as it might interrupt threads that do not even use networking. Current prequeue implementation has following issues : 1) It only checks size of the prequeue against sk_rcvbuf It was fine 15 years ago when sk_rcvbuf was in the 64KB vicinity. But we now have ~8MB values to cope with modern networking needs. We have to add sk_rmem_alloc in the equation, since out of order packets can definitely use up to sk_rcvbuf memory themselves. 2) Even with a fixed memory truesize check, prequeue can be filled by thousands of packets. When prequeue needs to be flushed, either from sofirq context (in tcp_prequeue() or timer code), or process context (in tcp_prequeue_process()), this adds a latency spike which is often not desirable. I added a fixed limit of 32 packets, as this translated to a max flush time of 60 us on my test hosts. Also note that all packets in prequeue are not accounted for tcp_mem, since they are not charged against sk_forward_alloc at this point. This is probably not a big deal. Note that this might increase LINUX_MIB_TCPPREQUEUEDROPPED counts, which is misnamed, as packets are not dropped at all, but rather pushed to the stack (where they can be either consumed or dropped) Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-27 17:12:25 +00:00
__NET_ADD_STATS(sock_net(sk), LINUX_MIB_TCPPREQUEUEDROPPED,
skb_queue_len(&tp->ucopy.prequeue));
tcp: give prequeue mode some care TCP prequeue goal is to defer processing of incoming packets to user space thread currently blocked in a recvmsg() system call. Intent is to spend less time processing these packets on behalf of softirq handler, as softirq handler is unfair to normal process scheduler decisions, as it might interrupt threads that do not even use networking. Current prequeue implementation has following issues : 1) It only checks size of the prequeue against sk_rcvbuf It was fine 15 years ago when sk_rcvbuf was in the 64KB vicinity. But we now have ~8MB values to cope with modern networking needs. We have to add sk_rmem_alloc in the equation, since out of order packets can definitely use up to sk_rcvbuf memory themselves. 2) Even with a fixed memory truesize check, prequeue can be filled by thousands of packets. When prequeue needs to be flushed, either from sofirq context (in tcp_prequeue() or timer code), or process context (in tcp_prequeue_process()), this adds a latency spike which is often not desirable. I added a fixed limit of 32 packets, as this translated to a max flush time of 60 us on my test hosts. Also note that all packets in prequeue are not accounted for tcp_mem, since they are not charged against sk_forward_alloc at this point. This is probably not a big deal. Note that this might increase LINUX_MIB_TCPPREQUEUEDROPPED counts, which is misnamed, as packets are not dropped at all, but rather pushed to the stack (where they can be either consumed or dropped) Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-27 17:12:25 +00:00
while ((skb1 = __skb_dequeue(&tp->ucopy.prequeue)) != NULL)
sk_backlog_rcv(sk, skb1);
tp->ucopy.memory = 0;
} else if (skb_queue_len(&tp->ucopy.prequeue) == 1) {
wake_up_interruptible_sync_poll(sk_sleep(sk),
POLLIN | POLLRDNORM | POLLRDBAND);
if (!inet_csk_ack_scheduled(sk))
inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK,
(3 * tcp_rto_min(sk)) / 4,
TCP_RTO_MAX);
}
return true;
}
EXPORT_SYMBOL(tcp_prequeue);
/*
* From tcp_input.c
*/
int tcp_v4_rcv(struct sk_buff *skb)
{
struct net *net = dev_net(skb->dev);
const struct iphdr *iph;
const struct tcphdr *th;
bool refcounted;
struct sock *sk;
int ret;
if (skb->pkt_type != PACKET_HOST)
goto discard_it;
/* Count it even if it's bad */
__TCP_INC_STATS(net, TCP_MIB_INSEGS);
if (!pskb_may_pull(skb, sizeof(struct tcphdr)))
goto discard_it;
th = tcp_hdr(skb);
if (th->doff < sizeof(struct tcphdr) / 4)
goto bad_packet;
if (!pskb_may_pull(skb, th->doff * 4))
goto discard_it;
/* An explanation is required here, I think.
* Packet length and doff are validated by header prediction,
* provided case of th->doff==0 is eliminated.
* So, we defer the checks. */
if (skb_checksum_init(skb, IPPROTO_TCP, inet_compute_pseudo))
goto csum_error;
th = tcp_hdr(skb);
iph = ip_hdr(skb);
/* This is tricky : We move IPCB at its correct location into TCP_SKB_CB()
* barrier() makes sure compiler wont play fool^Waliasing games.
*/
memmove(&TCP_SKB_CB(skb)->header.h4, IPCB(skb),
sizeof(struct inet_skb_parm));
barrier();
TCP_SKB_CB(skb)->seq = ntohl(th->seq);
TCP_SKB_CB(skb)->end_seq = (TCP_SKB_CB(skb)->seq + th->syn + th->fin +
skb->len - th->doff * 4);
TCP_SKB_CB(skb)->ack_seq = ntohl(th->ack_seq);
TCP_SKB_CB(skb)->tcp_flags = tcp_flag_byte(th);
TCP_SKB_CB(skb)->tcp_tw_isn = 0;
TCP_SKB_CB(skb)->ip_dsfield = ipv4_get_dsfield(iph);
TCP_SKB_CB(skb)->sacked = 0;
lookup:
sk = __inet_lookup_skb(&tcp_hashinfo, skb, __tcp_hdrlen(th), th->source,
th->dest, &refcounted);
if (!sk)
goto no_tcp_socket;
process:
if (sk->sk_state == TCP_TIME_WAIT)
goto do_time_wait;
if (sk->sk_state == TCP_NEW_SYN_RECV) {
struct request_sock *req = inet_reqsk(sk);
struct sock *nsk;
sk = req->rsk_listener;
if (unlikely(tcp_v4_inbound_md5_hash(sk, skb))) {
reqsk_put(req);
goto discard_it;
}
if (unlikely(sk->sk_state != TCP_LISTEN)) {
inet_csk_reqsk_queue_drop_and_put(sk, req);
goto lookup;
}
/* We own a reference on the listener, increase it again
* as we might lose it too soon.
*/
sock_hold(sk);
refcounted = true;
nsk = tcp_check_req(sk, skb, req, false);
if (!nsk) {
reqsk_put(req);
goto discard_and_relse;
}
if (nsk == sk) {
reqsk_put(req);
} else if (tcp_child_process(sk, nsk, skb)) {
tcp_v4_send_reset(nsk, skb);
goto discard_and_relse;
} else {
sock_put(sk);
return 0;
}
}
if (unlikely(iph->ttl < inet_sk(sk)->min_ttl)) {
__NET_INC_STATS(net, LINUX_MIB_TCPMINTTLDROP);
goto discard_and_relse;
}
if (!xfrm4_policy_check(sk, XFRM_POLICY_IN, skb))
goto discard_and_relse;
if (tcp_v4_inbound_md5_hash(sk, skb))
goto discard_and_relse;
nf_reset(skb);
if (sk_filter(sk, skb))
goto discard_and_relse;
skb->dev = NULL;
if (sk->sk_state == TCP_LISTEN) {
ret = tcp_v4_do_rcv(sk, skb);
goto put_and_return;
}
sk_incoming_cpu_update(sk);
bh_lock_sock_nested(sk);
tcp: Add RFC4898 tcpEStatsPerfDataSegsOut/In Per RFC4898, they count segments sent/received containing a positive length data segment (that includes retransmission segments carrying data). Unlike tcpi_segs_out/in, tcpi_data_segs_out/in excludes segments carrying no data (e.g. pure ack). The patch also updates the segs_in in tcp_fastopen_add_skb() so that segs_in >= data_segs_in property is kept. Together with retransmission data, tcpi_data_segs_out gives a better signal on the rxmit rate. v6: Rebase on the latest net-next v5: Eric pointed out that checking skb->len is still needed in tcp_fastopen_add_skb() because skb can carry a FIN without data. Hence, instead of open coding segs_in and data_segs_in, tcp_segs_in() helper is used. Comment is added to the fastopen case to explain why segs_in has to be reset and tcp_segs_in() has to be called before __skb_pull(). v4: Add comment to the changes in tcp_fastopen_add_skb() and also add remark on this case in the commit message. v3: Add const modifier to the skb parameter in tcp_segs_in() v2: Rework based on recent fix by Eric: commit a9d99ce28ed3 ("tcp: fix tcpi_segs_in after connection establishment") Signed-off-by: Martin KaFai Lau <kafai@fb.com> Cc: Chris Rapier <rapier@psc.edu> Cc: Eric Dumazet <edumazet@google.com> Cc: Marcelo Ricardo Leitner <mleitner@redhat.com> Cc: Neal Cardwell <ncardwell@google.com> Cc: Yuchung Cheng <ycheng@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-03-14 17:52:15 +00:00
tcp_segs_in(tcp_sk(sk), skb);
ret = 0;
if (!sock_owned_by_user(sk)) {
if (!tcp_prequeue(sk, skb))
ret = tcp_v4_do_rcv(sk, skb);
} else if (unlikely(sk_add_backlog(sk, skb,
sk->sk_rcvbuf + sk->sk_sndbuf))) {
bh_unlock_sock(sk);
__NET_INC_STATS(net, LINUX_MIB_TCPBACKLOGDROP);
goto discard_and_relse;
}
bh_unlock_sock(sk);
put_and_return:
if (refcounted)
sock_put(sk);
return ret;
no_tcp_socket:
if (!xfrm4_policy_check(NULL, XFRM_POLICY_IN, skb))
goto discard_it;
if (tcp_checksum_complete(skb)) {
csum_error:
__TCP_INC_STATS(net, TCP_MIB_CSUMERRORS);
bad_packet:
__TCP_INC_STATS(net, TCP_MIB_INERRS);
} else {
tcp_v4_send_reset(NULL, skb);
}
discard_it:
/* Discard frame. */
kfree_skb(skb);
return 0;
discard_and_relse:
sk_drops_add(sk, skb);
if (refcounted)
sock_put(sk);
goto discard_it;
do_time_wait:
if (!xfrm4_policy_check(NULL, XFRM_POLICY_IN, skb)) {
inet_twsk_put(inet_twsk(sk));
goto discard_it;
}
if (tcp_checksum_complete(skb)) {
inet_twsk_put(inet_twsk(sk));
goto csum_error;
}
switch (tcp_timewait_state_process(inet_twsk(sk), skb, th)) {
case TCP_TW_SYN: {
struct sock *sk2 = inet_lookup_listener(dev_net(skb->dev),
&tcp_hashinfo, skb,
__tcp_hdrlen(th),
iph->saddr, th->source,
iph->daddr, th->dest,
inet_iif(skb));
if (sk2) {
inet_twsk_deschedule_put(inet_twsk(sk));
sk = sk2;
refcounted = false;
goto process;
}
/* Fall through to ACK */
}
case TCP_TW_ACK:
tcp_v4_timewait_ack(sk, skb);
break;
case TCP_TW_RST:
tcp_v4_send_reset(sk, skb);
inet_twsk_deschedule_put(inet_twsk(sk));
goto discard_it;
case TCP_TW_SUCCESS:;
}
goto discard_it;
}
static struct timewait_sock_ops tcp_timewait_sock_ops = {
.twsk_obj_size = sizeof(struct tcp_timewait_sock),
.twsk_unique = tcp_twsk_unique,
.twsk_destructor= tcp_twsk_destructor,
};
net: tcp: ipv6_mapped needs sk_rx_dst_set method commit 5d299f3d3c8a2fb (net: ipv6: fix TCP early demux) added a regression for ipv6_mapped case. [ 67.422369] SELinux: initialized (dev autofs, type autofs), uses genfs_contexts [ 67.449678] SELinux: initialized (dev autofs, type autofs), uses genfs_contexts [ 92.631060] BUG: unable to handle kernel NULL pointer dereference at (null) [ 92.631435] IP: [< (null)>] (null) [ 92.631645] PGD 0 [ 92.631846] Oops: 0010 [#1] SMP [ 92.632095] Modules linked in: autofs4 sunrpc ipv6 dm_mirror dm_region_hash dm_log dm_multipath dm_mod video sbs sbshc battery ac lp parport sg snd_hda_intel snd_hda_codec snd_seq_oss snd_seq_midi_event snd_seq snd_seq_device pcspkr snd_pcm_oss snd_mixer_oss snd_pcm snd_timer serio_raw button floppy snd i2c_i801 i2c_core soundcore snd_page_alloc shpchp ide_cd_mod cdrom microcode ehci_hcd ohci_hcd uhci_hcd [ 92.634294] CPU 0 [ 92.634294] Pid: 4469, comm: sendmail Not tainted 3.6.0-rc1 #3 [ 92.634294] RIP: 0010:[<0000000000000000>] [< (null)>] (null) [ 92.634294] RSP: 0018:ffff880245fc7cb0 EFLAGS: 00010282 [ 92.634294] RAX: ffffffffa01985f0 RBX: ffff88024827ad00 RCX: 0000000000000000 [ 92.634294] RDX: 0000000000000218 RSI: ffff880254735380 RDI: ffff88024827ad00 [ 92.634294] RBP: ffff880245fc7cc8 R08: 0000000000000001 R09: 0000000000000000 [ 92.634294] R10: 0000000000000000 R11: ffff880245fc7bf8 R12: ffff880254735380 [ 92.634294] R13: ffff880254735380 R14: 0000000000000000 R15: 7fffffffffff0218 [ 92.634294] FS: 00007f4516ccd6f0(0000) GS:ffff880256600000(0000) knlGS:0000000000000000 [ 92.634294] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b [ 92.634294] CR2: 0000000000000000 CR3: 0000000245ed1000 CR4: 00000000000007f0 [ 92.634294] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 92.634294] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 [ 92.634294] Process sendmail (pid: 4469, threadinfo ffff880245fc6000, task ffff880254b8cac0) [ 92.634294] Stack: [ 92.634294] ffffffff813837a7 ffff88024827ad00 ffff880254b6b0e8 ffff880245fc7d68 [ 92.634294] ffffffff81385083 00000000001d2680 ffff8802547353a8 ffff880245fc7d18 [ 92.634294] ffffffff8105903a ffff88024827ad60 0000000000000002 00000000000000ff [ 92.634294] Call Trace: [ 92.634294] [<ffffffff813837a7>] ? tcp_finish_connect+0x2c/0xfa [ 92.634294] [<ffffffff81385083>] tcp_rcv_state_process+0x2b6/0x9c6 [ 92.634294] [<ffffffff8105903a>] ? sched_clock_cpu+0xc3/0xd1 [ 92.634294] [<ffffffff81059073>] ? local_clock+0x2b/0x3c [ 92.634294] [<ffffffff8138caf3>] tcp_v4_do_rcv+0x63a/0x670 [ 92.634294] [<ffffffff8133278e>] release_sock+0x128/0x1bd [ 92.634294] [<ffffffff8139f060>] __inet_stream_connect+0x1b1/0x352 [ 92.634294] [<ffffffff813325f5>] ? lock_sock_nested+0x74/0x7f [ 92.634294] [<ffffffff8104b333>] ? wake_up_bit+0x25/0x25 [ 92.634294] [<ffffffff813325f5>] ? lock_sock_nested+0x74/0x7f [ 92.634294] [<ffffffff8139f223>] ? inet_stream_connect+0x22/0x4b [ 92.634294] [<ffffffff8139f234>] inet_stream_connect+0x33/0x4b [ 92.634294] [<ffffffff8132e8cf>] sys_connect+0x78/0x9e [ 92.634294] [<ffffffff813fd407>] ? sysret_check+0x1b/0x56 [ 92.634294] [<ffffffff81088503>] ? __audit_syscall_entry+0x195/0x1c8 [ 92.634294] [<ffffffff811cc26e>] ? trace_hardirqs_on_thunk+0x3a/0x3f [ 92.634294] [<ffffffff813fd3e2>] system_call_fastpath+0x16/0x1b [ 92.634294] Code: Bad RIP value. [ 92.634294] RIP [< (null)>] (null) [ 92.634294] RSP <ffff880245fc7cb0> [ 92.634294] CR2: 0000000000000000 [ 92.648982] ---[ end trace 24e2bed94314c8d9 ]--- [ 92.649146] Kernel panic - not syncing: Fatal exception in interrupt Fix this using inet_sk_rx_dst_set(), and export this function in case IPv6 is modular. Reported-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-08-09 14:11:00 +00:00
void inet_sk_rx_dst_set(struct sock *sk, const struct sk_buff *skb)
{
struct dst_entry *dst = skb_dst(skb);
net: fix IP early demux races David Wilder reported crashes caused by dst reuse. <quote David> I am seeing a crash on a distro V4.2.3 kernel caused by a double release of a dst_entry. In ipv4_dst_destroy() the call to list_empty() finds a poisoned next pointer, indicating the dst_entry has already been removed from the list and freed. The crash occurs 18 to 24 hours into a run of a network stress exerciser. </quote> Thanks to his detailed report and analysis, we were able to understand the core issue. IP early demux can associate a dst to skb, after a lookup in TCP/UDP sockets. When socket cache is not properly set, we want to store into sk->sk_dst_cache the dst for future IP early demux lookups, by acquiring a stable refcount on the dst. Problem is this acquisition is simply using an atomic_inc(), which works well, unless the dst was queued for destruction from dst_release() noticing dst refcount went to zero, if DST_NOCACHE was set on dst. We need to make sure current refcount is not zero before incrementing it, or risk double free as David reported. This patch, being a stable candidate, adds two new helpers, and use them only from IP early demux problematic paths. It might be possible to merge in net-next skb_dst_force() and skb_dst_force_safe(), but I prefer having the smallest patch for stable kernels : Maybe some skb_dst_force() callers do not expect skb->dst can suddenly be cleared. Can probably be backported back to linux-3.6 kernels Reported-by: David J. Wilder <dwilder@us.ibm.com> Tested-by: David J. Wilder <dwilder@us.ibm.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-12-14 22:08:53 +00:00
if (dst && dst_hold_safe(dst)) {
tcp: remove dst refcount false sharing for prequeue mode Alexander Duyck reported high false sharing on dst refcount in tcp stack when prequeue is used. prequeue is the mechanism used when a thread is blocked in recvmsg()/read() on a TCP socket, using a blocking model rather than select()/poll()/epoll() non blocking one. We already try to use RCU in input path as much as possible, but we were forced to take a refcount on the dst when skb escaped RCU protected region. When/if the user thread runs on different cpu, dst_release() will then touch dst refcount again. Commit 093162553c33 (tcp: force a dst refcount when prequeue packet) was an example of a race fix. It turns out the only remaining usage of skb->dst for a packet stored in a TCP socket prequeue is IP early demux. We can add a logic to detect when IP early demux is probably going to use skb->dst. Because we do an optimistic check rather than duplicate existing logic, we need to guard inet_sk_rx_dst_set() and inet6_sk_rx_dst_set() from using a NULL dst. Many thanks to Alexander for providing a nice bug report, git bisection, and reproducer. Tested using Alexander script on a 40Gb NIC, 8 RX queues. Hosts have 24 cores, 48 hyper threads. echo 0 >/proc/sys/net/ipv4/tcp_autocorking for i in `seq 0 47` do for j in `seq 0 2` do netperf -H $DEST -t TCP_STREAM -l 1000 \ -c -C -T $i,$i -P 0 -- \ -m 64 -s 64K -D & done done Before patch : ~6Mpps and ~95% cpu usage on receiver After patch : ~9Mpps and ~35% cpu usage on receiver. Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-09-08 15:06:07 +00:00
sk->sk_rx_dst = dst;
inet_sk(sk)->rx_dst_ifindex = skb->skb_iif;
}
}
net: tcp: ipv6_mapped needs sk_rx_dst_set method commit 5d299f3d3c8a2fb (net: ipv6: fix TCP early demux) added a regression for ipv6_mapped case. [ 67.422369] SELinux: initialized (dev autofs, type autofs), uses genfs_contexts [ 67.449678] SELinux: initialized (dev autofs, type autofs), uses genfs_contexts [ 92.631060] BUG: unable to handle kernel NULL pointer dereference at (null) [ 92.631435] IP: [< (null)>] (null) [ 92.631645] PGD 0 [ 92.631846] Oops: 0010 [#1] SMP [ 92.632095] Modules linked in: autofs4 sunrpc ipv6 dm_mirror dm_region_hash dm_log dm_multipath dm_mod video sbs sbshc battery ac lp parport sg snd_hda_intel snd_hda_codec snd_seq_oss snd_seq_midi_event snd_seq snd_seq_device pcspkr snd_pcm_oss snd_mixer_oss snd_pcm snd_timer serio_raw button floppy snd i2c_i801 i2c_core soundcore snd_page_alloc shpchp ide_cd_mod cdrom microcode ehci_hcd ohci_hcd uhci_hcd [ 92.634294] CPU 0 [ 92.634294] Pid: 4469, comm: sendmail Not tainted 3.6.0-rc1 #3 [ 92.634294] RIP: 0010:[<0000000000000000>] [< (null)>] (null) [ 92.634294] RSP: 0018:ffff880245fc7cb0 EFLAGS: 00010282 [ 92.634294] RAX: ffffffffa01985f0 RBX: ffff88024827ad00 RCX: 0000000000000000 [ 92.634294] RDX: 0000000000000218 RSI: ffff880254735380 RDI: ffff88024827ad00 [ 92.634294] RBP: ffff880245fc7cc8 R08: 0000000000000001 R09: 0000000000000000 [ 92.634294] R10: 0000000000000000 R11: ffff880245fc7bf8 R12: ffff880254735380 [ 92.634294] R13: ffff880254735380 R14: 0000000000000000 R15: 7fffffffffff0218 [ 92.634294] FS: 00007f4516ccd6f0(0000) GS:ffff880256600000(0000) knlGS:0000000000000000 [ 92.634294] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b [ 92.634294] CR2: 0000000000000000 CR3: 0000000245ed1000 CR4: 00000000000007f0 [ 92.634294] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 92.634294] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 [ 92.634294] Process sendmail (pid: 4469, threadinfo ffff880245fc6000, task ffff880254b8cac0) [ 92.634294] Stack: [ 92.634294] ffffffff813837a7 ffff88024827ad00 ffff880254b6b0e8 ffff880245fc7d68 [ 92.634294] ffffffff81385083 00000000001d2680 ffff8802547353a8 ffff880245fc7d18 [ 92.634294] ffffffff8105903a ffff88024827ad60 0000000000000002 00000000000000ff [ 92.634294] Call Trace: [ 92.634294] [<ffffffff813837a7>] ? tcp_finish_connect+0x2c/0xfa [ 92.634294] [<ffffffff81385083>] tcp_rcv_state_process+0x2b6/0x9c6 [ 92.634294] [<ffffffff8105903a>] ? sched_clock_cpu+0xc3/0xd1 [ 92.634294] [<ffffffff81059073>] ? local_clock+0x2b/0x3c [ 92.634294] [<ffffffff8138caf3>] tcp_v4_do_rcv+0x63a/0x670 [ 92.634294] [<ffffffff8133278e>] release_sock+0x128/0x1bd [ 92.634294] [<ffffffff8139f060>] __inet_stream_connect+0x1b1/0x352 [ 92.634294] [<ffffffff813325f5>] ? lock_sock_nested+0x74/0x7f [ 92.634294] [<ffffffff8104b333>] ? wake_up_bit+0x25/0x25 [ 92.634294] [<ffffffff813325f5>] ? lock_sock_nested+0x74/0x7f [ 92.634294] [<ffffffff8139f223>] ? inet_stream_connect+0x22/0x4b [ 92.634294] [<ffffffff8139f234>] inet_stream_connect+0x33/0x4b [ 92.634294] [<ffffffff8132e8cf>] sys_connect+0x78/0x9e [ 92.634294] [<ffffffff813fd407>] ? sysret_check+0x1b/0x56 [ 92.634294] [<ffffffff81088503>] ? __audit_syscall_entry+0x195/0x1c8 [ 92.634294] [<ffffffff811cc26e>] ? trace_hardirqs_on_thunk+0x3a/0x3f [ 92.634294] [<ffffffff813fd3e2>] system_call_fastpath+0x16/0x1b [ 92.634294] Code: Bad RIP value. [ 92.634294] RIP [< (null)>] (null) [ 92.634294] RSP <ffff880245fc7cb0> [ 92.634294] CR2: 0000000000000000 [ 92.648982] ---[ end trace 24e2bed94314c8d9 ]--- [ 92.649146] Kernel panic - not syncing: Fatal exception in interrupt Fix this using inet_sk_rx_dst_set(), and export this function in case IPv6 is modular. Reported-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-08-09 14:11:00 +00:00
EXPORT_SYMBOL(inet_sk_rx_dst_set);
const struct inet_connection_sock_af_ops ipv4_specific = {
.queue_xmit = ip_queue_xmit,
.send_check = tcp_v4_send_check,
.rebuild_header = inet_sk_rebuild_header,
.sk_rx_dst_set = inet_sk_rx_dst_set,
.conn_request = tcp_v4_conn_request,
.syn_recv_sock = tcp_v4_syn_recv_sock,
.net_header_len = sizeof(struct iphdr),
.setsockopt = ip_setsockopt,
.getsockopt = ip_getsockopt,
.addr2sockaddr = inet_csk_addr2sockaddr,
.sockaddr_len = sizeof(struct sockaddr_in),
[SOCK] proto: Add hashinfo member to struct proto This way we can remove TCP and DCCP specific versions of sk->sk_prot->get_port: both v4 and v6 use inet_csk_get_port sk->sk_prot->hash: inet_hash is directly used, only v6 need a specific version to deal with mapped sockets sk->sk_prot->unhash: both v4 and v6 use inet_hash directly struct inet_connection_sock_af_ops also gets a new member, bind_conflict, so that inet_csk_get_port can find the per family routine. Now only the lookup routines receive as a parameter a struct inet_hashtable. With this we further reuse code, reducing the difference among INET transport protocols. Eventually work has to be done on UDP and SCTP to make them share this infrastructure and get as a bonus inet_diag interfaces so that iproute can be used with these protocols. net-2.6/net/ipv4/inet_hashtables.c: struct proto | +8 struct inet_connection_sock_af_ops | +8 2 structs changed __inet_hash_nolisten | +18 __inet_hash | -210 inet_put_port | +8 inet_bind_bucket_create | +1 __inet_hash_connect | -8 5 functions changed, 27 bytes added, 218 bytes removed, diff: -191 net-2.6/net/core/sock.c: proto_seq_show | +3 1 function changed, 3 bytes added, diff: +3 net-2.6/net/ipv4/inet_connection_sock.c: inet_csk_get_port | +15 1 function changed, 15 bytes added, diff: +15 net-2.6/net/ipv4/tcp.c: tcp_set_state | -7 1 function changed, 7 bytes removed, diff: -7 net-2.6/net/ipv4/tcp_ipv4.c: tcp_v4_get_port | -31 tcp_v4_hash | -48 tcp_v4_destroy_sock | -7 tcp_v4_syn_recv_sock | -2 tcp_unhash | -179 5 functions changed, 267 bytes removed, diff: -267 net-2.6/net/ipv6/inet6_hashtables.c: __inet6_hash | +8 1 function changed, 8 bytes added, diff: +8 net-2.6/net/ipv4/inet_hashtables.c: inet_unhash | +190 inet_hash | +242 2 functions changed, 432 bytes added, diff: +432 vmlinux: 16 functions changed, 485 bytes added, 492 bytes removed, diff: -7 /home/acme/git/net-2.6/net/ipv6/tcp_ipv6.c: tcp_v6_get_port | -31 tcp_v6_hash | -7 tcp_v6_syn_recv_sock | -9 3 functions changed, 47 bytes removed, diff: -47 /home/acme/git/net-2.6/net/dccp/proto.c: dccp_destroy_sock | -7 dccp_unhash | -179 dccp_hash | -49 dccp_set_state | -7 dccp_done | +1 5 functions changed, 1 bytes added, 242 bytes removed, diff: -241 /home/acme/git/net-2.6/net/dccp/ipv4.c: dccp_v4_get_port | -31 dccp_v4_request_recv_sock | -2 2 functions changed, 33 bytes removed, diff: -33 /home/acme/git/net-2.6/net/dccp/ipv6.c: dccp_v6_get_port | -31 dccp_v6_hash | -7 dccp_v6_request_recv_sock | +5 3 functions changed, 5 bytes added, 38 bytes removed, diff: -33 Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-02-03 12:06:04 +00:00
.bind_conflict = inet_csk_bind_conflict,
#ifdef CONFIG_COMPAT
.compat_setsockopt = compat_ip_setsockopt,
.compat_getsockopt = compat_ip_getsockopt,
#endif
.mtu_reduced = tcp_v4_mtu_reduced,
};
EXPORT_SYMBOL(ipv4_specific);
#ifdef CONFIG_TCP_MD5SIG
static const struct tcp_sock_af_ops tcp_sock_ipv4_specific = {
.md5_lookup = tcp_v4_md5_lookup,
.calc_md5_hash = tcp_v4_md5_hash_skb,
.md5_parse = tcp_v4_parse_md5_keys,
};
#endif
/* NOTE: A lot of things set to zero explicitly by call to
* sk_alloc() so need not be done here.
*/
static int tcp_v4_init_sock(struct sock *sk)
{
struct inet_connection_sock *icsk = inet_csk(sk);
tcp_init_sock(sk);
icsk->icsk_af_ops = &ipv4_specific;
#ifdef CONFIG_TCP_MD5SIG
tcp_sk(sk)->af_specific = &tcp_sock_ipv4_specific;
#endif
return 0;
}
void tcp_v4_destroy_sock(struct sock *sk)
{
struct tcp_sock *tp = tcp_sk(sk);
tcp_clear_xmit_timers(sk);
tcp_cleanup_congestion_control(sk);
/* Cleanup up the write buffer. */
tcp_write_queue_purge(sk);
/* Cleans up our, hopefully empty, out_of_order_queue. */
__skb_queue_purge(&tp->out_of_order_queue);
#ifdef CONFIG_TCP_MD5SIG
/* Clean up the MD5 key list, if any */
if (tp->md5sig_info) {
tcp_clear_md5_list(sk);
kfree_rcu(tp->md5sig_info, rcu);
tp->md5sig_info = NULL;
}
#endif
/* Clean prequeue, it must be empty really */
__skb_queue_purge(&tp->ucopy.prequeue);
/* Clean up a referenced TCP bind bucket. */
if (inet_csk(sk)->icsk_bind_hash)
[SOCK] proto: Add hashinfo member to struct proto This way we can remove TCP and DCCP specific versions of sk->sk_prot->get_port: both v4 and v6 use inet_csk_get_port sk->sk_prot->hash: inet_hash is directly used, only v6 need a specific version to deal with mapped sockets sk->sk_prot->unhash: both v4 and v6 use inet_hash directly struct inet_connection_sock_af_ops also gets a new member, bind_conflict, so that inet_csk_get_port can find the per family routine. Now only the lookup routines receive as a parameter a struct inet_hashtable. With this we further reuse code, reducing the difference among INET transport protocols. Eventually work has to be done on UDP and SCTP to make them share this infrastructure and get as a bonus inet_diag interfaces so that iproute can be used with these protocols. net-2.6/net/ipv4/inet_hashtables.c: struct proto | +8 struct inet_connection_sock_af_ops | +8 2 structs changed __inet_hash_nolisten | +18 __inet_hash | -210 inet_put_port | +8 inet_bind_bucket_create | +1 __inet_hash_connect | -8 5 functions changed, 27 bytes added, 218 bytes removed, diff: -191 net-2.6/net/core/sock.c: proto_seq_show | +3 1 function changed, 3 bytes added, diff: +3 net-2.6/net/ipv4/inet_connection_sock.c: inet_csk_get_port | +15 1 function changed, 15 bytes added, diff: +15 net-2.6/net/ipv4/tcp.c: tcp_set_state | -7 1 function changed, 7 bytes removed, diff: -7 net-2.6/net/ipv4/tcp_ipv4.c: tcp_v4_get_port | -31 tcp_v4_hash | -48 tcp_v4_destroy_sock | -7 tcp_v4_syn_recv_sock | -2 tcp_unhash | -179 5 functions changed, 267 bytes removed, diff: -267 net-2.6/net/ipv6/inet6_hashtables.c: __inet6_hash | +8 1 function changed, 8 bytes added, diff: +8 net-2.6/net/ipv4/inet_hashtables.c: inet_unhash | +190 inet_hash | +242 2 functions changed, 432 bytes added, diff: +432 vmlinux: 16 functions changed, 485 bytes added, 492 bytes removed, diff: -7 /home/acme/git/net-2.6/net/ipv6/tcp_ipv6.c: tcp_v6_get_port | -31 tcp_v6_hash | -7 tcp_v6_syn_recv_sock | -9 3 functions changed, 47 bytes removed, diff: -47 /home/acme/git/net-2.6/net/dccp/proto.c: dccp_destroy_sock | -7 dccp_unhash | -179 dccp_hash | -49 dccp_set_state | -7 dccp_done | +1 5 functions changed, 1 bytes added, 242 bytes removed, diff: -241 /home/acme/git/net-2.6/net/dccp/ipv4.c: dccp_v4_get_port | -31 dccp_v4_request_recv_sock | -2 2 functions changed, 33 bytes removed, diff: -33 /home/acme/git/net-2.6/net/dccp/ipv6.c: dccp_v6_get_port | -31 dccp_v6_hash | -7 dccp_v6_request_recv_sock | +5 3 functions changed, 5 bytes added, 38 bytes removed, diff: -33 Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-02-03 12:06:04 +00:00
inet_put_port(sk);
BUG_ON(tp->fastopen_rsk);
TCPCT part 1d: define TCP cookie option, extend existing struct's Data structures are carefully composed to require minimal additions. For example, the struct tcp_options_received cookie_plus variable fits between existing 16-bit and 8-bit variables, requiring no additional space (taking alignment into consideration). There are no additions to tcp_request_sock, and only 1 pointer in tcp_sock. This is a significantly revised implementation of an earlier (year-old) patch that no longer applies cleanly, with permission of the original author (Adam Langley): http://thread.gmane.org/gmane.linux.network/102586 The principle difference is using a TCP option to carry the cookie nonce, instead of a user configured offset in the data. This is more flexible and less subject to user configuration error. Such a cookie option has been suggested for many years, and is also useful without SYN data, allowing several related concepts to use the same extension option. "Re: SYN floods (was: does history repeat itself?)", September 9, 1996. http://www.merit.net/mail.archives/nanog/1996-09/msg00235.html "Re: what a new TCP header might look like", May 12, 1998. ftp://ftp.isi.edu/end2end/end2end-interest-1998.mail These functions will also be used in subsequent patches that implement additional features. Requires: TCPCT part 1a: add request_values parameter for sending SYNACK TCPCT part 1b: generate Responder Cookie secret TCPCT part 1c: sysctl_tcp_cookie_size, socket option TCP_COOKIE_TRANSACTIONS Signed-off-by: William.Allen.Simpson@gmail.com Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-02 18:17:05 +00:00
/* If socket is aborted during connect operation */
tcp_free_fastopen_req(tp);
tcp_saved_syn_free(tp);
sk_sockets_allocated_dec(sk);
if (mem_cgroup_sockets_enabled && sk->sk_memcg)
sock_release_memcg(sk);
}
EXPORT_SYMBOL(tcp_v4_destroy_sock);
#ifdef CONFIG_PROC_FS
/* Proc filesystem TCP sock list dumping. */
/*
* Get next listener socket follow cur. If cur is NULL, get first socket
* starting from bucket given in st->bucket; when st->bucket is zero the
* very first socket in the hash table is returned.
*/
static void *listening_get_next(struct seq_file *seq, void *cur)
{
struct tcp_iter_state *st = seq->private;
struct net *net = seq_file_net(seq);
struct inet_listen_hashbucket *ilb;
struct inet_connection_sock *icsk;
struct sock *sk = cur;
if (!sk) {
get_head:
ilb = &tcp_hashinfo.listening_hash[st->bucket];
spin_lock_bh(&ilb->lock);
sk = sk_head(&ilb->head);
st->offset = 0;
goto get_sk;
}
ilb = &tcp_hashinfo.listening_hash[st->bucket];
++st->num;
++st->offset;
sk = sk_next(sk);
get_sk:
sk_for_each_from(sk) {
if (!net_eq(sock_net(sk), net))
continue;
if (sk->sk_family == st->family)
return sk;
icsk = inet_csk(sk);
}
spin_unlock_bh(&ilb->lock);
st->offset = 0;
if (++st->bucket < INET_LHTABLE_SIZE)
goto get_head;
return NULL;
}
static void *listening_get_idx(struct seq_file *seq, loff_t *pos)
{
struct tcp_iter_state *st = seq->private;
void *rc;
st->bucket = 0;
st->offset = 0;
rc = listening_get_next(seq, NULL);
while (rc && *pos) {
rc = listening_get_next(seq, rc);
--*pos;
}
return rc;
}
tcp/dccp: remove twchain TCP listener refactoring, part 3 : Our goal is to hash SYN_RECV sockets into main ehash for fast lookup, and parallel SYN processing. Current inet_ehash_bucket contains two chains, one for ESTABLISH (and friend states) sockets, another for TIME_WAIT sockets only. As the hash table is sized to get at most one socket per bucket, it makes little sense to have separate twchain, as it makes the lookup slightly more complicated, and doubles hash table memory usage. If we make sure all socket types have the lookup keys at the same offsets, we can use a generic and faster lookup. It turns out TIME_WAIT and ESTABLISHED sockets already have common lookup fields for IPv4. [ INET_TW_MATCH() is no longer needed ] I'll provide a follow-up to factorize IPv6 lookup as well, to remove INET6_TW_MATCH() This way, SYN_RECV pseudo sockets will be supported the same. A new sock_gen_put() helper is added, doing either a sock_put() or inet_twsk_put() [ and will support SYN_RECV later ]. Note this helper should only be called in real slow path, when rcu lookup found a socket that was moved to another identity (freed/reused immediately), but could eventually be used in other contexts, like sock_edemux() Before patch : dmesg | grep "TCP established" TCP established hash table entries: 524288 (order: 11, 8388608 bytes) After patch : TCP established hash table entries: 524288 (order: 10, 4194304 bytes) Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-10-03 07:22:02 +00:00
static inline bool empty_bucket(const struct tcp_iter_state *st)
{
tcp/dccp: remove twchain TCP listener refactoring, part 3 : Our goal is to hash SYN_RECV sockets into main ehash for fast lookup, and parallel SYN processing. Current inet_ehash_bucket contains two chains, one for ESTABLISH (and friend states) sockets, another for TIME_WAIT sockets only. As the hash table is sized to get at most one socket per bucket, it makes little sense to have separate twchain, as it makes the lookup slightly more complicated, and doubles hash table memory usage. If we make sure all socket types have the lookup keys at the same offsets, we can use a generic and faster lookup. It turns out TIME_WAIT and ESTABLISHED sockets already have common lookup fields for IPv4. [ INET_TW_MATCH() is no longer needed ] I'll provide a follow-up to factorize IPv6 lookup as well, to remove INET6_TW_MATCH() This way, SYN_RECV pseudo sockets will be supported the same. A new sock_gen_put() helper is added, doing either a sock_put() or inet_twsk_put() [ and will support SYN_RECV later ]. Note this helper should only be called in real slow path, when rcu lookup found a socket that was moved to another identity (freed/reused immediately), but could eventually be used in other contexts, like sock_edemux() Before patch : dmesg | grep "TCP established" TCP established hash table entries: 524288 (order: 11, 8388608 bytes) After patch : TCP established hash table entries: 524288 (order: 10, 4194304 bytes) Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-10-03 07:22:02 +00:00
return hlist_nulls_empty(&tcp_hashinfo.ehash[st->bucket].chain);
}
/*
* Get first established socket starting from bucket given in st->bucket.
* If st->bucket is zero, the very first socket in the hash is returned.
*/
static void *established_get_first(struct seq_file *seq)
{
struct tcp_iter_state *st = seq->private;
struct net *net = seq_file_net(seq);
void *rc = NULL;
st->offset = 0;
for (; st->bucket <= tcp_hashinfo.ehash_mask; ++st->bucket) {
struct sock *sk;
struct hlist_nulls_node *node;
spinlock_t *lock = inet_ehash_lockp(&tcp_hashinfo, st->bucket);
/* Lockless fast path for the common case of empty buckets */
if (empty_bucket(st))
continue;
spin_lock_bh(lock);
sk_nulls_for_each(sk, node, &tcp_hashinfo.ehash[st->bucket].chain) {
if (sk->sk_family != st->family ||
!net_eq(sock_net(sk), net)) {
continue;
}
rc = sk;
goto out;
}
spin_unlock_bh(lock);
}
out:
return rc;
}
static void *established_get_next(struct seq_file *seq, void *cur)
{
struct sock *sk = cur;
struct hlist_nulls_node *node;
struct tcp_iter_state *st = seq->private;
struct net *net = seq_file_net(seq);
++st->num;
++st->offset;
tcp/dccp: remove twchain TCP listener refactoring, part 3 : Our goal is to hash SYN_RECV sockets into main ehash for fast lookup, and parallel SYN processing. Current inet_ehash_bucket contains two chains, one for ESTABLISH (and friend states) sockets, another for TIME_WAIT sockets only. As the hash table is sized to get at most one socket per bucket, it makes little sense to have separate twchain, as it makes the lookup slightly more complicated, and doubles hash table memory usage. If we make sure all socket types have the lookup keys at the same offsets, we can use a generic and faster lookup. It turns out TIME_WAIT and ESTABLISHED sockets already have common lookup fields for IPv4. [ INET_TW_MATCH() is no longer needed ] I'll provide a follow-up to factorize IPv6 lookup as well, to remove INET6_TW_MATCH() This way, SYN_RECV pseudo sockets will be supported the same. A new sock_gen_put() helper is added, doing either a sock_put() or inet_twsk_put() [ and will support SYN_RECV later ]. Note this helper should only be called in real slow path, when rcu lookup found a socket that was moved to another identity (freed/reused immediately), but could eventually be used in other contexts, like sock_edemux() Before patch : dmesg | grep "TCP established" TCP established hash table entries: 524288 (order: 11, 8388608 bytes) After patch : TCP established hash table entries: 524288 (order: 10, 4194304 bytes) Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-10-03 07:22:02 +00:00
sk = sk_nulls_next(sk);
sk_nulls_for_each_from(sk, node) {
if (sk->sk_family == st->family && net_eq(sock_net(sk), net))
tcp/dccp: remove twchain TCP listener refactoring, part 3 : Our goal is to hash SYN_RECV sockets into main ehash for fast lookup, and parallel SYN processing. Current inet_ehash_bucket contains two chains, one for ESTABLISH (and friend states) sockets, another for TIME_WAIT sockets only. As the hash table is sized to get at most one socket per bucket, it makes little sense to have separate twchain, as it makes the lookup slightly more complicated, and doubles hash table memory usage. If we make sure all socket types have the lookup keys at the same offsets, we can use a generic and faster lookup. It turns out TIME_WAIT and ESTABLISHED sockets already have common lookup fields for IPv4. [ INET_TW_MATCH() is no longer needed ] I'll provide a follow-up to factorize IPv6 lookup as well, to remove INET6_TW_MATCH() This way, SYN_RECV pseudo sockets will be supported the same. A new sock_gen_put() helper is added, doing either a sock_put() or inet_twsk_put() [ and will support SYN_RECV later ]. Note this helper should only be called in real slow path, when rcu lookup found a socket that was moved to another identity (freed/reused immediately), but could eventually be used in other contexts, like sock_edemux() Before patch : dmesg | grep "TCP established" TCP established hash table entries: 524288 (order: 11, 8388608 bytes) After patch : TCP established hash table entries: 524288 (order: 10, 4194304 bytes) Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-10-03 07:22:02 +00:00
return sk;
}
tcp/dccp: remove twchain TCP listener refactoring, part 3 : Our goal is to hash SYN_RECV sockets into main ehash for fast lookup, and parallel SYN processing. Current inet_ehash_bucket contains two chains, one for ESTABLISH (and friend states) sockets, another for TIME_WAIT sockets only. As the hash table is sized to get at most one socket per bucket, it makes little sense to have separate twchain, as it makes the lookup slightly more complicated, and doubles hash table memory usage. If we make sure all socket types have the lookup keys at the same offsets, we can use a generic and faster lookup. It turns out TIME_WAIT and ESTABLISHED sockets already have common lookup fields for IPv4. [ INET_TW_MATCH() is no longer needed ] I'll provide a follow-up to factorize IPv6 lookup as well, to remove INET6_TW_MATCH() This way, SYN_RECV pseudo sockets will be supported the same. A new sock_gen_put() helper is added, doing either a sock_put() or inet_twsk_put() [ and will support SYN_RECV later ]. Note this helper should only be called in real slow path, when rcu lookup found a socket that was moved to another identity (freed/reused immediately), but could eventually be used in other contexts, like sock_edemux() Before patch : dmesg | grep "TCP established" TCP established hash table entries: 524288 (order: 11, 8388608 bytes) After patch : TCP established hash table entries: 524288 (order: 10, 4194304 bytes) Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-10-03 07:22:02 +00:00
spin_unlock_bh(inet_ehash_lockp(&tcp_hashinfo, st->bucket));
++st->bucket;
return established_get_first(seq);
}
static void *established_get_idx(struct seq_file *seq, loff_t pos)
{
struct tcp_iter_state *st = seq->private;
void *rc;
st->bucket = 0;
rc = established_get_first(seq);
while (rc && pos) {
rc = established_get_next(seq, rc);
--pos;
}
return rc;
}
static void *tcp_get_idx(struct seq_file *seq, loff_t pos)
{
void *rc;
struct tcp_iter_state *st = seq->private;
st->state = TCP_SEQ_STATE_LISTENING;
rc = listening_get_idx(seq, &pos);
if (!rc) {
st->state = TCP_SEQ_STATE_ESTABLISHED;
rc = established_get_idx(seq, pos);
}
return rc;
}
static void *tcp_seek_last_pos(struct seq_file *seq)
{
struct tcp_iter_state *st = seq->private;
int offset = st->offset;
int orig_num = st->num;
void *rc = NULL;
switch (st->state) {
case TCP_SEQ_STATE_LISTENING:
if (st->bucket >= INET_LHTABLE_SIZE)
break;
st->state = TCP_SEQ_STATE_LISTENING;
rc = listening_get_next(seq, NULL);
while (offset-- && rc)
rc = listening_get_next(seq, rc);
if (rc)
break;
st->bucket = 0;
tcp/dccp: remove twchain TCP listener refactoring, part 3 : Our goal is to hash SYN_RECV sockets into main ehash for fast lookup, and parallel SYN processing. Current inet_ehash_bucket contains two chains, one for ESTABLISH (and friend states) sockets, another for TIME_WAIT sockets only. As the hash table is sized to get at most one socket per bucket, it makes little sense to have separate twchain, as it makes the lookup slightly more complicated, and doubles hash table memory usage. If we make sure all socket types have the lookup keys at the same offsets, we can use a generic and faster lookup. It turns out TIME_WAIT and ESTABLISHED sockets already have common lookup fields for IPv4. [ INET_TW_MATCH() is no longer needed ] I'll provide a follow-up to factorize IPv6 lookup as well, to remove INET6_TW_MATCH() This way, SYN_RECV pseudo sockets will be supported the same. A new sock_gen_put() helper is added, doing either a sock_put() or inet_twsk_put() [ and will support SYN_RECV later ]. Note this helper should only be called in real slow path, when rcu lookup found a socket that was moved to another identity (freed/reused immediately), but could eventually be used in other contexts, like sock_edemux() Before patch : dmesg | grep "TCP established" TCP established hash table entries: 524288 (order: 11, 8388608 bytes) After patch : TCP established hash table entries: 524288 (order: 10, 4194304 bytes) Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-10-03 07:22:02 +00:00
st->state = TCP_SEQ_STATE_ESTABLISHED;
/* Fallthrough */
case TCP_SEQ_STATE_ESTABLISHED:
if (st->bucket > tcp_hashinfo.ehash_mask)
break;
rc = established_get_first(seq);
while (offset-- && rc)
rc = established_get_next(seq, rc);
}
st->num = orig_num;
return rc;
}
static void *tcp_seq_start(struct seq_file *seq, loff_t *pos)
{
struct tcp_iter_state *st = seq->private;
void *rc;
if (*pos && *pos == st->last_pos) {
rc = tcp_seek_last_pos(seq);
if (rc)
goto out;
}
st->state = TCP_SEQ_STATE_LISTENING;
st->num = 0;
st->bucket = 0;
st->offset = 0;
rc = *pos ? tcp_get_idx(seq, *pos - 1) : SEQ_START_TOKEN;
out:
st->last_pos = *pos;
return rc;
}
static void *tcp_seq_next(struct seq_file *seq, void *v, loff_t *pos)
{
struct tcp_iter_state *st = seq->private;
void *rc = NULL;
if (v == SEQ_START_TOKEN) {
rc = tcp_get_idx(seq, 0);
goto out;
}
switch (st->state) {
case TCP_SEQ_STATE_LISTENING:
rc = listening_get_next(seq, v);
if (!rc) {
st->state = TCP_SEQ_STATE_ESTABLISHED;
st->bucket = 0;
st->offset = 0;
rc = established_get_first(seq);
}
break;
case TCP_SEQ_STATE_ESTABLISHED:
rc = established_get_next(seq, v);
break;
}
out:
++*pos;
st->last_pos = *pos;
return rc;
}
static void tcp_seq_stop(struct seq_file *seq, void *v)
{
struct tcp_iter_state *st = seq->private;
switch (st->state) {
case TCP_SEQ_STATE_LISTENING:
if (v != SEQ_START_TOKEN)
spin_unlock_bh(&tcp_hashinfo.listening_hash[st->bucket].lock);
break;
case TCP_SEQ_STATE_ESTABLISHED:
if (v)
spin_unlock_bh(inet_ehash_lockp(&tcp_hashinfo, st->bucket));
break;
}
}
int tcp_seq_open(struct inode *inode, struct file *file)
{
struct tcp_seq_afinfo *afinfo = PDE_DATA(inode);
struct tcp_iter_state *s;
int err;
err = seq_open_net(inode, file, &afinfo->seq_ops,
sizeof(struct tcp_iter_state));
if (err < 0)
return err;
s = ((struct seq_file *)file->private_data)->private;
s->family = afinfo->family;
s->last_pos = 0;
return 0;
}
EXPORT_SYMBOL(tcp_seq_open);
int tcp_proc_register(struct net *net, struct tcp_seq_afinfo *afinfo)
{
int rc = 0;
struct proc_dir_entry *p;
afinfo->seq_ops.start = tcp_seq_start;
afinfo->seq_ops.next = tcp_seq_next;
afinfo->seq_ops.stop = tcp_seq_stop;
p = proc_create_data(afinfo->name, S_IRUGO, net->proc_net,
afinfo->seq_fops, afinfo);
if (!p)
rc = -ENOMEM;
return rc;
}
EXPORT_SYMBOL(tcp_proc_register);
void tcp_proc_unregister(struct net *net, struct tcp_seq_afinfo *afinfo)
{
remove_proc_entry(afinfo->name, net->proc_net);
}
EXPORT_SYMBOL(tcp_proc_unregister);
static void get_openreq4(const struct request_sock *req,
struct seq_file *f, int i)
{
const struct inet_request_sock *ireq = inet_rsk(req);
long delta = req->rsk_timer.expires - jiffies;
seq_printf(f, "%4d: %08X:%04X %08X:%04X"
" %02X %08X:%08X %02X:%08lX %08X %5u %8d %u %d %pK",
i,
ireq->ir_loc_addr,
ireq->ir_num,
ireq->ir_rmt_addr,
ntohs(ireq->ir_rmt_port),
TCP_SYN_RECV,
0, 0, /* could print option size, but that is af dependent. */
1, /* timers active (only the expire timer) */
jiffies_delta_to_clock_t(delta),
tcp: better retrans tracking for defer-accept For passive TCP connections using TCP_DEFER_ACCEPT facility, we incorrectly increment req->retrans each time timeout triggers while no SYNACK is sent. SYNACK are not sent for TCP_DEFER_ACCEPT that were established (for which we received the ACK from client). Only the last SYNACK is sent so that we can receive again an ACK from client, to move the req into accept queue. We plan to change this later to avoid the useless retransmit (and potential problem as this SYNACK could be lost) TCP_INFO later gives wrong information to user, claiming imaginary retransmits. Decouple req->retrans field into two independent fields : num_retrans : number of retransmit num_timeout : number of timeouts num_timeout is the counter that is incremented at each timeout, regardless of actual SYNACK being sent or not, and used to compute the exponential timeout. Introduce inet_rtx_syn_ack() helper to increment num_retrans only if ->rtx_syn_ack() succeeded. Use inet_rtx_syn_ack() from tcp_check_req() to increment num_retrans when we re-send a SYNACK in answer to a (retransmitted) SYN. Prior to this patch, we were not counting these retransmits. Change tcp_v[46]_rtx_synack() to increment TCP_MIB_RETRANSSEGS only if a synack packet was successfully queued. Reported-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Julian Anastasov <ja@ssi.bg> Cc: Vijay Subramanian <subramanian.vijay@gmail.com> Cc: Elliott Hughes <enh@google.com> Cc: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-10-27 23:16:46 +00:00
req->num_timeout,
from_kuid_munged(seq_user_ns(f),
sock_i_uid(req->rsk_listener)),
0, /* non standard timer */
0, /* open_requests have no inode */
0,
req);
}
static void get_tcp4_sock(struct sock *sk, struct seq_file *f, int i)
{
int timer_active;
unsigned long timer_expires;
const struct tcp_sock *tp = tcp_sk(sk);
const struct inet_connection_sock *icsk = inet_csk(sk);
const struct inet_sock *inet = inet_sk(sk);
const struct fastopen_queue *fastopenq = &icsk->icsk_accept_queue.fastopenq;
__be32 dest = inet->inet_daddr;
__be32 src = inet->inet_rcv_saddr;
__u16 destp = ntohs(inet->inet_dport);
__u16 srcp = ntohs(inet->inet_sport);
int rx_queue;
int state;
tcp: Tail loss probe (TLP) This patch series implement the Tail loss probe (TLP) algorithm described in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The first patch implements the basic algorithm. TLP's goal is to reduce tail latency of short transactions. It achieves this by converting retransmission timeouts (RTOs) occuring due to tail losses (losses at end of transactions) into fast recovery. TLP transmits one packet in two round-trips when a connection is in Open state and isn't receiving any ACKs. The transmitted packet, aka loss probe, can be either new or a retransmission. When there is tail loss, the ACK from a loss probe triggers FACK/early-retransmit based fast recovery, thus avoiding a costly RTO. In the absence of loss, there is no change in the connection state. PTO stands for probe timeout. It is a timer event indicating that an ACK is overdue and triggers a loss probe packet. The PTO value is set to max(2*SRTT, 10ms) and is adjusted to account for delayed ACK timer when there is only one oustanding packet. TLP Algorithm On transmission of new data in Open state: -> packets_out > 1: schedule PTO in max(2*SRTT, 10ms). -> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms) -> PTO = min(PTO, RTO) Conditions for scheduling PTO: -> Connection is in Open state. -> Connection is either cwnd limited or no new data to send. -> Number of probes per tail loss episode is limited to one. -> Connection is SACK enabled. When PTO fires: new_segment_exists: -> transmit new segment. -> packets_out++. cwnd remains same. no_new_packet: -> retransmit the last segment. Its ACK triggers FACK or early retransmit based recovery. ACK path: -> rearm RTO at start of ACK processing. -> reschedule PTO if need be. In addition, the patch includes a small variation to the Early Retransmit (ER) algorithm, such that ER and TLP together can in principle recover any N-degree of tail loss through fast recovery. TLP is controlled by the same sysctl as ER, tcp_early_retrans sysctl. tcp_early_retrans==0; disables TLP and ER. ==1; enables RFC5827 ER. ==2; delayed ER. ==3; TLP and delayed ER. [DEFAULT] ==4; TLP only. The TLP patch series have been extensively tested on Google Web servers. It is most effective for short Web trasactions, where it reduced RTOs by 15% and improved HTTP response time (average by 6%, 99th percentile by 10%). The transmitted probes account for <0.5% of the overall transmissions. Signed-off-by: Nandita Dukkipati <nanditad@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Acked-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 10:00:43 +00:00
if (icsk->icsk_pending == ICSK_TIME_RETRANS ||
icsk->icsk_pending == ICSK_TIME_EARLY_RETRANS ||
icsk->icsk_pending == ICSK_TIME_LOSS_PROBE) {
timer_active = 1;
timer_expires = icsk->icsk_timeout;
} else if (icsk->icsk_pending == ICSK_TIME_PROBE0) {
timer_active = 4;
timer_expires = icsk->icsk_timeout;
} else if (timer_pending(&sk->sk_timer)) {
timer_active = 2;
timer_expires = sk->sk_timer.expires;
} else {
timer_active = 0;
timer_expires = jiffies;
}
state = sk_state_load(sk);
if (state == TCP_LISTEN)
rx_queue = sk->sk_ack_backlog;
else
/* Because we don't lock the socket,
* we might find a transient negative value.
*/
rx_queue = max_t(int, tp->rcv_nxt - tp->copied_seq, 0);
seq_printf(f, "%4d: %08X:%04X %08X:%04X %02X %08X:%08X %02X:%08lX "
"%08X %5u %8d %lu %d %pK %lu %lu %u %u %d",
i, src, srcp, dest, destp, state,
tp->write_seq - tp->snd_una,
rx_queue,
timer_active,
jiffies_delta_to_clock_t(timer_expires - jiffies),
icsk->icsk_retransmits,
from_kuid_munged(seq_user_ns(f), sock_i_uid(sk)),
icsk->icsk_probes_out,
sock_i_ino(sk),
atomic_read(&sk->sk_refcnt), sk,
jiffies_to_clock_t(icsk->icsk_rto),
jiffies_to_clock_t(icsk->icsk_ack.ato),
(icsk->icsk_ack.quick << 1) | icsk->icsk_ack.pingpong,
tp->snd_cwnd,
state == TCP_LISTEN ?
fastopenq->max_qlen :
(tcp_in_initial_slowstart(tp) ? -1 : tp->snd_ssthresh));
}
static void get_timewait4_sock(const struct inet_timewait_sock *tw,
struct seq_file *f, int i)
{
tcp/dccp: get rid of central timewait timer Using a timer wheel for timewait sockets was nice ~15 years ago when memory was expensive and machines had a single processor. This does not scale, code is ugly and source of huge latencies (Typically 30 ms have been seen, cpus spinning on death_lock spinlock.) We can afford to use an extra 64 bytes per timewait sock and spread timewait load to all cpus to have better behavior. Tested: On following test, /proc/sys/net/ipv4/tcp_tw_recycle is set to 1 on the target (lpaa24) Before patch : lpaa23:~# ./super_netperf 200 -H lpaa24 -t TCP_CC -l 60 -- -p0,0 419594 lpaa23:~# ./super_netperf 200 -H lpaa24 -t TCP_CC -l 60 -- -p0,0 437171 While test is running, we can observe 25 or even 33 ms latencies. lpaa24:~# ping -c 1000 -i 0.02 -qn lpaa23 ... 1000 packets transmitted, 1000 received, 0% packet loss, time 20601ms rtt min/avg/max/mdev = 0.020/0.217/25.771/1.535 ms, pipe 2 lpaa24:~# ping -c 1000 -i 0.02 -qn lpaa23 ... 1000 packets transmitted, 1000 received, 0% packet loss, time 20702ms rtt min/avg/max/mdev = 0.019/0.183/33.761/1.441 ms, pipe 2 After patch : About 90% increase of throughput : lpaa23:~# ./super_netperf 200 -H lpaa24 -t TCP_CC -l 60 -- -p0,0 810442 lpaa23:~# ./super_netperf 200 -H lpaa24 -t TCP_CC -l 60 -- -p0,0 800992 And latencies are kept to minimal values during this load, even if network utilization is 90% higher : lpaa24:~# ping -c 1000 -i 0.02 -qn lpaa23 ... 1000 packets transmitted, 1000 received, 0% packet loss, time 19991ms rtt min/avg/max/mdev = 0.023/0.064/0.360/0.042 ms Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-04-13 01:51:09 +00:00
long delta = tw->tw_timer.expires - jiffies;
__be32 dest, src;
__u16 destp, srcp;
dest = tw->tw_daddr;
src = tw->tw_rcv_saddr;
destp = ntohs(tw->tw_dport);
srcp = ntohs(tw->tw_sport);
seq_printf(f, "%4d: %08X:%04X %08X:%04X"
" %02X %08X:%08X %02X:%08lX %08X %5d %8d %d %d %pK",
i, src, srcp, dest, destp, tw->tw_substate, 0, 0,
3, jiffies_delta_to_clock_t(delta), 0, 0, 0, 0,
atomic_read(&tw->tw_refcnt), tw);
}
#define TMPSZ 150
static int tcp4_seq_show(struct seq_file *seq, void *v)
{
struct tcp_iter_state *st;
tcp/dccp: remove twchain TCP listener refactoring, part 3 : Our goal is to hash SYN_RECV sockets into main ehash for fast lookup, and parallel SYN processing. Current inet_ehash_bucket contains two chains, one for ESTABLISH (and friend states) sockets, another for TIME_WAIT sockets only. As the hash table is sized to get at most one socket per bucket, it makes little sense to have separate twchain, as it makes the lookup slightly more complicated, and doubles hash table memory usage. If we make sure all socket types have the lookup keys at the same offsets, we can use a generic and faster lookup. It turns out TIME_WAIT and ESTABLISHED sockets already have common lookup fields for IPv4. [ INET_TW_MATCH() is no longer needed ] I'll provide a follow-up to factorize IPv6 lookup as well, to remove INET6_TW_MATCH() This way, SYN_RECV pseudo sockets will be supported the same. A new sock_gen_put() helper is added, doing either a sock_put() or inet_twsk_put() [ and will support SYN_RECV later ]. Note this helper should only be called in real slow path, when rcu lookup found a socket that was moved to another identity (freed/reused immediately), but could eventually be used in other contexts, like sock_edemux() Before patch : dmesg | grep "TCP established" TCP established hash table entries: 524288 (order: 11, 8388608 bytes) After patch : TCP established hash table entries: 524288 (order: 10, 4194304 bytes) Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-10-03 07:22:02 +00:00
struct sock *sk = v;
seq_setwidth(seq, TMPSZ - 1);
if (v == SEQ_START_TOKEN) {
seq_puts(seq, " sl local_address rem_address st tx_queue "
"rx_queue tr tm->when retrnsmt uid timeout "
"inode");
goto out;
}
st = seq->private;
if (sk->sk_state == TCP_TIME_WAIT)
get_timewait4_sock(v, seq, st->num);
else if (sk->sk_state == TCP_NEW_SYN_RECV)
get_openreq4(v, seq, st->num);
else
get_tcp4_sock(v, seq, st->num);
out:
seq_pad(seq, '\n');
return 0;
}
static const struct file_operations tcp_afinfo_seq_fops = {
.owner = THIS_MODULE,
.open = tcp_seq_open,
.read = seq_read,
.llseek = seq_lseek,
.release = seq_release_net
};
static struct tcp_seq_afinfo tcp4_seq_afinfo = {
.name = "tcp",
.family = AF_INET,
.seq_fops = &tcp_afinfo_seq_fops,
.seq_ops = {
.show = tcp4_seq_show,
},
};
static int __net_init tcp4_proc_init_net(struct net *net)
{
return tcp_proc_register(net, &tcp4_seq_afinfo);
}
static void __net_exit tcp4_proc_exit_net(struct net *net)
{
tcp_proc_unregister(net, &tcp4_seq_afinfo);
}
static struct pernet_operations tcp4_net_ops = {
.init = tcp4_proc_init_net,
.exit = tcp4_proc_exit_net,
};
int __init tcp4_proc_init(void)
{
return register_pernet_subsys(&tcp4_net_ops);
}
void tcp4_proc_exit(void)
{
unregister_pernet_subsys(&tcp4_net_ops);
}
#endif /* CONFIG_PROC_FS */
struct proto tcp_prot = {
.name = "TCP",
.owner = THIS_MODULE,
.close = tcp_close,
.connect = tcp_v4_connect,
.disconnect = tcp_disconnect,
.accept = inet_csk_accept,
.ioctl = tcp_ioctl,
.init = tcp_v4_init_sock,
.destroy = tcp_v4_destroy_sock,
.shutdown = tcp_shutdown,
.setsockopt = tcp_setsockopt,
.getsockopt = tcp_getsockopt,
.recvmsg = tcp_recvmsg,
.sendmsg = tcp_sendmsg,
.sendpage = tcp_sendpage,
.backlog_rcv = tcp_v4_do_rcv,
tcp: TCP Small Queues This introduce TSQ (TCP Small Queues) TSQ goal is to reduce number of TCP packets in xmit queues (qdisc & device queues), to reduce RTT and cwnd bias, part of the bufferbloat problem. sk->sk_wmem_alloc not allowed to grow above a given limit, allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a given time. TSO packets are sized/capped to half the limit, so that we have two TSO packets in flight, allowing better bandwidth use. As a side effect, setting the limit to 40000 automatically reduces the standard gso max limit (65536) to 40000/2 : It can help to reduce latencies of high prio packets, having smaller TSO packets. This means we divert sock_wfree() to a tcp_wfree() handler, to queue/send following frames when skb_orphan() [2] is called for the already queued skbs. Results on my dev machines (tg3/ixgbe nics) are really impressive, using standard pfifo_fast, and with or without TSO/GSO. Without reduction of nominal bandwidth, we have reduction of buffering per bulk sender : < 1ms on Gbit (instead of 50ms with TSO) < 8ms on 100Mbit (instead of 132 ms) I no longer have 4 MBytes backlogged in qdisc by a single netperf session, and both side socket autotuning no longer use 4 Mbytes. As skb destructor cannot restart xmit itself ( as qdisc lock might be taken at this point ), we delegate the work to a tasklet. We use one tasklest per cpu for performance reasons. If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag. This flag is tested in a new protocol method called from release_sock(), to eventually send new segments. [1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable [2] skb_orphan() is usually called at TX completion time, but some drivers call it in their start_xmit() handler. These drivers should at least use BQL, or else a single TCP session can still fill the whole NIC TX ring, since TSQ will have no effect. Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Dave Taht <dave.taht@bufferbloat.net> Cc: Tom Herbert <therbert@google.com> Cc: Matt Mathis <mattmathis@google.com> Cc: Yuchung Cheng <ycheng@google.com> Cc: Nandita Dukkipati <nanditad@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 05:50:31 +00:00
.release_cb = tcp_release_cb,
[SOCK] proto: Add hashinfo member to struct proto This way we can remove TCP and DCCP specific versions of sk->sk_prot->get_port: both v4 and v6 use inet_csk_get_port sk->sk_prot->hash: inet_hash is directly used, only v6 need a specific version to deal with mapped sockets sk->sk_prot->unhash: both v4 and v6 use inet_hash directly struct inet_connection_sock_af_ops also gets a new member, bind_conflict, so that inet_csk_get_port can find the per family routine. Now only the lookup routines receive as a parameter a struct inet_hashtable. With this we further reuse code, reducing the difference among INET transport protocols. Eventually work has to be done on UDP and SCTP to make them share this infrastructure and get as a bonus inet_diag interfaces so that iproute can be used with these protocols. net-2.6/net/ipv4/inet_hashtables.c: struct proto | +8 struct inet_connection_sock_af_ops | +8 2 structs changed __inet_hash_nolisten | +18 __inet_hash | -210 inet_put_port | +8 inet_bind_bucket_create | +1 __inet_hash_connect | -8 5 functions changed, 27 bytes added, 218 bytes removed, diff: -191 net-2.6/net/core/sock.c: proto_seq_show | +3 1 function changed, 3 bytes added, diff: +3 net-2.6/net/ipv4/inet_connection_sock.c: inet_csk_get_port | +15 1 function changed, 15 bytes added, diff: +15 net-2.6/net/ipv4/tcp.c: tcp_set_state | -7 1 function changed, 7 bytes removed, diff: -7 net-2.6/net/ipv4/tcp_ipv4.c: tcp_v4_get_port | -31 tcp_v4_hash | -48 tcp_v4_destroy_sock | -7 tcp_v4_syn_recv_sock | -2 tcp_unhash | -179 5 functions changed, 267 bytes removed, diff: -267 net-2.6/net/ipv6/inet6_hashtables.c: __inet6_hash | +8 1 function changed, 8 bytes added, diff: +8 net-2.6/net/ipv4/inet_hashtables.c: inet_unhash | +190 inet_hash | +242 2 functions changed, 432 bytes added, diff: +432 vmlinux: 16 functions changed, 485 bytes added, 492 bytes removed, diff: -7 /home/acme/git/net-2.6/net/ipv6/tcp_ipv6.c: tcp_v6_get_port | -31 tcp_v6_hash | -7 tcp_v6_syn_recv_sock | -9 3 functions changed, 47 bytes removed, diff: -47 /home/acme/git/net-2.6/net/dccp/proto.c: dccp_destroy_sock | -7 dccp_unhash | -179 dccp_hash | -49 dccp_set_state | -7 dccp_done | +1 5 functions changed, 1 bytes added, 242 bytes removed, diff: -241 /home/acme/git/net-2.6/net/dccp/ipv4.c: dccp_v4_get_port | -31 dccp_v4_request_recv_sock | -2 2 functions changed, 33 bytes removed, diff: -33 /home/acme/git/net-2.6/net/dccp/ipv6.c: dccp_v6_get_port | -31 dccp_v6_hash | -7 dccp_v6_request_recv_sock | +5 3 functions changed, 5 bytes added, 38 bytes removed, diff: -33 Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-02-03 12:06:04 +00:00
.hash = inet_hash,
.unhash = inet_unhash,
.get_port = inet_csk_get_port,
.enter_memory_pressure = tcp_enter_memory_pressure,
tcp: TCP_NOTSENT_LOWAT socket option Idea of this patch is to add optional limitation of number of unsent bytes in TCP sockets, to reduce usage of kernel memory. TCP receiver might announce a big window, and TCP sender autotuning might allow a large amount of bytes in write queue, but this has little performance impact if a large part of this buffering is wasted : Write queue needs to be large only to deal with large BDP, not necessarily to cope with scheduling delays (incoming ACKS make room for the application to queue more bytes) For most workloads, using a value of 128 KB or less is OK to give applications enough time to react to POLLOUT events in time (or being awaken in a blocking sendmsg()) This patch adds two ways to set the limit : 1) Per socket option TCP_NOTSENT_LOWAT 2) A sysctl (/proc/sys/net/ipv4/tcp_notsent_lowat) for sockets not using TCP_NOTSENT_LOWAT socket option (or setting a zero value) Default value being UINT_MAX (0xFFFFFFFF), meaning this has no effect. This changes poll()/select()/epoll() to report POLLOUT only if number of unsent bytes is below tp->nosent_lowat Note this might increase number of sendmsg()/sendfile() calls when using non blocking sockets, and increase number of context switches for blocking sockets. Note this is not related to SO_SNDLOWAT (as SO_SNDLOWAT is defined as : Specify the minimum number of bytes in the buffer until the socket layer will pass the data to the protocol) Tested: netperf sessions, and watching /proc/net/protocols "memory" column for TCP With 200 concurrent netperf -t TCP_STREAM sessions, amount of kernel memory used by TCP buffers shrinks by ~55 % (20567 pages instead of 45458) lpq83:~# echo -1 >/proc/sys/net/ipv4/tcp_notsent_lowat lpq83:~# (super_netperf 200 -t TCP_STREAM -H remote -l 90 &); sleep 60 ; grep TCP /proc/net/protocols TCPv6 1880 2 45458 no 208 yes ipv6 y y y y y y y y y y y y y n y y y y y TCP 1696 508 45458 no 208 yes kernel y y y y y y y y y y y y y n y y y y y lpq83:~# echo 131072 >/proc/sys/net/ipv4/tcp_notsent_lowat lpq83:~# (super_netperf 200 -t TCP_STREAM -H remote -l 90 &); sleep 60 ; grep TCP /proc/net/protocols TCPv6 1880 2 20567 no 208 yes ipv6 y y y y y y y y y y y y y n y y y y y TCP 1696 508 20567 no 208 yes kernel y y y y y y y y y y y y y n y y y y y Using 128KB has no bad effect on the throughput or cpu usage of a single flow, although there is an increase of context switches. A bonus is that we hold socket lock for a shorter amount of time and should improve latencies of ACK processing. lpq83:~# echo -1 >/proc/sys/net/ipv4/tcp_notsent_lowat lpq83:~# perf stat -e context-switches ./netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3 OMNI Send TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 7.7.7.84 () port 0 AF_INET : +/-2.500% @ 99% conf. Local Remote Local Elapsed Throughput Throughput Local Local Remote Remote Local Remote Service Send Socket Recv Socket Send Time Units CPU CPU CPU CPU Service Service Demand Size Size Size (sec) Util Util Util Util Demand Demand Units Final Final % Method % Method 1651584 6291456 16384 20.00 17447.90 10^6bits/s 3.13 S -1.00 U 0.353 -1.000 usec/KB Performance counter stats for './netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3': 412,514 context-switches 200.034645535 seconds time elapsed lpq83:~# echo 131072 >/proc/sys/net/ipv4/tcp_notsent_lowat lpq83:~# perf stat -e context-switches ./netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3 OMNI Send TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 7.7.7.84 () port 0 AF_INET : +/-2.500% @ 99% conf. Local Remote Local Elapsed Throughput Throughput Local Local Remote Remote Local Remote Service Send Socket Recv Socket Send Time Units CPU CPU CPU CPU Service Service Demand Size Size Size (sec) Util Util Util Util Demand Demand Units Final Final % Method % Method 1593240 6291456 16384 20.00 17321.16 10^6bits/s 3.35 S -1.00 U 0.381 -1.000 usec/KB Performance counter stats for './netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3': 2,675,818 context-switches 200.029651391 seconds time elapsed Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Neal Cardwell <ncardwell@google.com> Cc: Yuchung Cheng <ycheng@google.com> Acked-By: Yuchung Cheng <ycheng@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-07-23 03:27:07 +00:00
.stream_memory_free = tcp_stream_memory_free,
.sockets_allocated = &tcp_sockets_allocated,
.orphan_count = &tcp_orphan_count,
.memory_allocated = &tcp_memory_allocated,
.memory_pressure = &tcp_memory_pressure,
.sysctl_mem = sysctl_tcp_mem,
.sysctl_wmem = sysctl_tcp_wmem,
.sysctl_rmem = sysctl_tcp_rmem,
.max_header = MAX_TCP_HEADER,
.obj_size = sizeof(struct tcp_sock),
.slab_flags = SLAB_DESTROY_BY_RCU,
.twsk_prot = &tcp_timewait_sock_ops,
.rsk_prot = &tcp_request_sock_ops,
.h.hashinfo = &tcp_hashinfo,
.no_autobind = true,
#ifdef CONFIG_COMPAT
.compat_setsockopt = compat_tcp_setsockopt,
.compat_getsockopt = compat_tcp_getsockopt,
#endif
.diag_destroy = tcp_abort,
};
EXPORT_SYMBOL(tcp_prot);
static void __net_exit tcp_sk_exit(struct net *net)
{
int cpu;
for_each_possible_cpu(cpu)
inet_ctl_sock_destroy(*per_cpu_ptr(net->ipv4.tcp_sk, cpu));
free_percpu(net->ipv4.tcp_sk);
}
static int __net_init tcp_sk_init(struct net *net)
{
int res, cpu;
net->ipv4.tcp_sk = alloc_percpu(struct sock *);
if (!net->ipv4.tcp_sk)
return -ENOMEM;
for_each_possible_cpu(cpu) {
struct sock *sk;
res = inet_ctl_sock_create(&sk, PF_INET, SOCK_RAW,
IPPROTO_TCP, net);
if (res)
goto fail;
sock_set_flag(sk, SOCK_USE_WRITE_QUEUE);
*per_cpu_ptr(net->ipv4.tcp_sk, cpu) = sk;
}
tcp: add rfc3168, section 6.1.1.1. fallback This work as a follow-up of commit f7b3bec6f516 ("net: allow setting ecn via routing table") and adds RFC3168 section 6.1.1.1. fallback for outgoing ECN connections. In other words, this work adds a retry with a non-ECN setup SYN packet, as suggested from the RFC on the first timeout: [...] A host that receives no reply to an ECN-setup SYN within the normal SYN retransmission timeout interval MAY resend the SYN and any subsequent SYN retransmissions with CWR and ECE cleared. [...] Schematic client-side view when assuming the server is in tcp_ecn=2 mode, that is, Linux default since 2009 via commit 255cac91c3c9 ("tcp: extend ECN sysctl to allow server-side only ECN"): 1) Normal ECN-capable path: SYN ECE CWR -----> <----- SYN ACK ECE ACK -----> 2) Path with broken middlebox, when client has fallback: SYN ECE CWR ----X crappy middlebox drops packet (timeout, rtx) SYN -----> <----- SYN ACK ACK -----> In case we would not have the fallback implemented, the middlebox drop point would basically end up as: SYN ECE CWR ----X crappy middlebox drops packet (timeout, rtx) SYN ECE CWR ----X crappy middlebox drops packet (timeout, rtx) SYN ECE CWR ----X crappy middlebox drops packet (timeout, rtx) In any case, it's rather a smaller percentage of sites where there would occur such additional setup latency: it was found in end of 2014 that ~56% of IPv4 and 65% of IPv6 servers of Alexa 1 million list would negotiate ECN (aka tcp_ecn=2 default), 0.42% of these webservers will fail to connect when trying to negotiate with ECN (tcp_ecn=1) due to timeouts, which the fallback would mitigate with a slight latency trade-off. Recent related paper on this topic: Brian Trammell, Mirja Kühlewind, Damiano Boppart, Iain Learmonth, Gorry Fairhurst, and Richard Scheffenegger: "Enabling Internet-Wide Deployment of Explicit Congestion Notification." Proc. PAM 2015, New York. http://ecn.ethz.ch/ecn-pam15.pdf Thus, when net.ipv4.tcp_ecn=1 is being set, the patch will perform RFC3168, section 6.1.1.1. fallback on timeout. For users explicitly not wanting this which can be in DC use case, we add a net.ipv4.tcp_ecn_fallback knob that allows for disabling the fallback. tp->ecn_flags are not being cleared in tcp_ecn_clear_syn() on output, but rather we let tcp_ecn_rcv_synack() take that over on input path in case a SYN ACK ECE was delayed. Thus a spurious SYN retransmission will not prevent ECN being negotiated eventually in that case. Reference: https://www.ietf.org/proceedings/92/slides/slides-92-iccrg-1.pdf Reference: https://www.ietf.org/proceedings/89/slides/slides-89-tsvarea-1.pdf Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: Florian Westphal <fw@strlen.de> Signed-off-by: Mirja Kühlewind <mirja.kuehlewind@tik.ee.ethz.ch> Signed-off-by: Brian Trammell <trammell@tik.ee.ethz.ch> Cc: Eric Dumazet <edumazet@google.com> Cc: Dave That <dave.taht@gmail.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-05-19 19:04:22 +00:00
net->ipv4.sysctl_tcp_ecn = 2;
tcp: add rfc3168, section 6.1.1.1. fallback This work as a follow-up of commit f7b3bec6f516 ("net: allow setting ecn via routing table") and adds RFC3168 section 6.1.1.1. fallback for outgoing ECN connections. In other words, this work adds a retry with a non-ECN setup SYN packet, as suggested from the RFC on the first timeout: [...] A host that receives no reply to an ECN-setup SYN within the normal SYN retransmission timeout interval MAY resend the SYN and any subsequent SYN retransmissions with CWR and ECE cleared. [...] Schematic client-side view when assuming the server is in tcp_ecn=2 mode, that is, Linux default since 2009 via commit 255cac91c3c9 ("tcp: extend ECN sysctl to allow server-side only ECN"): 1) Normal ECN-capable path: SYN ECE CWR -----> <----- SYN ACK ECE ACK -----> 2) Path with broken middlebox, when client has fallback: SYN ECE CWR ----X crappy middlebox drops packet (timeout, rtx) SYN -----> <----- SYN ACK ACK -----> In case we would not have the fallback implemented, the middlebox drop point would basically end up as: SYN ECE CWR ----X crappy middlebox drops packet (timeout, rtx) SYN ECE CWR ----X crappy middlebox drops packet (timeout, rtx) SYN ECE CWR ----X crappy middlebox drops packet (timeout, rtx) In any case, it's rather a smaller percentage of sites where there would occur such additional setup latency: it was found in end of 2014 that ~56% of IPv4 and 65% of IPv6 servers of Alexa 1 million list would negotiate ECN (aka tcp_ecn=2 default), 0.42% of these webservers will fail to connect when trying to negotiate with ECN (tcp_ecn=1) due to timeouts, which the fallback would mitigate with a slight latency trade-off. Recent related paper on this topic: Brian Trammell, Mirja Kühlewind, Damiano Boppart, Iain Learmonth, Gorry Fairhurst, and Richard Scheffenegger: "Enabling Internet-Wide Deployment of Explicit Congestion Notification." Proc. PAM 2015, New York. http://ecn.ethz.ch/ecn-pam15.pdf Thus, when net.ipv4.tcp_ecn=1 is being set, the patch will perform RFC3168, section 6.1.1.1. fallback on timeout. For users explicitly not wanting this which can be in DC use case, we add a net.ipv4.tcp_ecn_fallback knob that allows for disabling the fallback. tp->ecn_flags are not being cleared in tcp_ecn_clear_syn() on output, but rather we let tcp_ecn_rcv_synack() take that over on input path in case a SYN ACK ECE was delayed. Thus a spurious SYN retransmission will not prevent ECN being negotiated eventually in that case. Reference: https://www.ietf.org/proceedings/92/slides/slides-92-iccrg-1.pdf Reference: https://www.ietf.org/proceedings/89/slides/slides-89-tsvarea-1.pdf Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: Florian Westphal <fw@strlen.de> Signed-off-by: Mirja Kühlewind <mirja.kuehlewind@tik.ee.ethz.ch> Signed-off-by: Brian Trammell <trammell@tik.ee.ethz.ch> Cc: Eric Dumazet <edumazet@google.com> Cc: Dave That <dave.taht@gmail.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-05-19 19:04:22 +00:00
net->ipv4.sysctl_tcp_ecn_fallback = 1;
net->ipv4.sysctl_tcp_base_mss = TCP_BASE_MSS;
net->ipv4.sysctl_tcp_probe_threshold = TCP_PROBE_THRESHOLD;
net->ipv4.sysctl_tcp_probe_interval = TCP_PROBE_INTERVAL;
net->ipv4.sysctl_tcp_keepalive_time = TCP_KEEPALIVE_TIME;
net->ipv4.sysctl_tcp_keepalive_probes = TCP_KEEPALIVE_PROBES;
net->ipv4.sysctl_tcp_keepalive_intvl = TCP_KEEPALIVE_INTVL;
net->ipv4.sysctl_tcp_syn_retries = TCP_SYN_RETRIES;
net->ipv4.sysctl_tcp_synack_retries = TCP_SYNACK_RETRIES;
net->ipv4.sysctl_tcp_syncookies = 1;
net->ipv4.sysctl_tcp_reordering = TCP_FASTRETRANS_THRESH;
net->ipv4.sysctl_tcp_retries1 = TCP_RETR1;
net->ipv4.sysctl_tcp_retries2 = TCP_RETR2;
net->ipv4.sysctl_tcp_orphan_retries = 0;
net->ipv4.sysctl_tcp_fin_timeout = TCP_FIN_TIMEOUT;
net->ipv4.sysctl_tcp_notsent_lowat = UINT_MAX;
tcp: add rfc3168, section 6.1.1.1. fallback This work as a follow-up of commit f7b3bec6f516 ("net: allow setting ecn via routing table") and adds RFC3168 section 6.1.1.1. fallback for outgoing ECN connections. In other words, this work adds a retry with a non-ECN setup SYN packet, as suggested from the RFC on the first timeout: [...] A host that receives no reply to an ECN-setup SYN within the normal SYN retransmission timeout interval MAY resend the SYN and any subsequent SYN retransmissions with CWR and ECE cleared. [...] Schematic client-side view when assuming the server is in tcp_ecn=2 mode, that is, Linux default since 2009 via commit 255cac91c3c9 ("tcp: extend ECN sysctl to allow server-side only ECN"): 1) Normal ECN-capable path: SYN ECE CWR -----> <----- SYN ACK ECE ACK -----> 2) Path with broken middlebox, when client has fallback: SYN ECE CWR ----X crappy middlebox drops packet (timeout, rtx) SYN -----> <----- SYN ACK ACK -----> In case we would not have the fallback implemented, the middlebox drop point would basically end up as: SYN ECE CWR ----X crappy middlebox drops packet (timeout, rtx) SYN ECE CWR ----X crappy middlebox drops packet (timeout, rtx) SYN ECE CWR ----X crappy middlebox drops packet (timeout, rtx) In any case, it's rather a smaller percentage of sites where there would occur such additional setup latency: it was found in end of 2014 that ~56% of IPv4 and 65% of IPv6 servers of Alexa 1 million list would negotiate ECN (aka tcp_ecn=2 default), 0.42% of these webservers will fail to connect when trying to negotiate with ECN (tcp_ecn=1) due to timeouts, which the fallback would mitigate with a slight latency trade-off. Recent related paper on this topic: Brian Trammell, Mirja Kühlewind, Damiano Boppart, Iain Learmonth, Gorry Fairhurst, and Richard Scheffenegger: "Enabling Internet-Wide Deployment of Explicit Congestion Notification." Proc. PAM 2015, New York. http://ecn.ethz.ch/ecn-pam15.pdf Thus, when net.ipv4.tcp_ecn=1 is being set, the patch will perform RFC3168, section 6.1.1.1. fallback on timeout. For users explicitly not wanting this which can be in DC use case, we add a net.ipv4.tcp_ecn_fallback knob that allows for disabling the fallback. tp->ecn_flags are not being cleared in tcp_ecn_clear_syn() on output, but rather we let tcp_ecn_rcv_synack() take that over on input path in case a SYN ACK ECE was delayed. Thus a spurious SYN retransmission will not prevent ECN being negotiated eventually in that case. Reference: https://www.ietf.org/proceedings/92/slides/slides-92-iccrg-1.pdf Reference: https://www.ietf.org/proceedings/89/slides/slides-89-tsvarea-1.pdf Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: Florian Westphal <fw@strlen.de> Signed-off-by: Mirja Kühlewind <mirja.kuehlewind@tik.ee.ethz.ch> Signed-off-by: Brian Trammell <trammell@tik.ee.ethz.ch> Cc: Eric Dumazet <edumazet@google.com> Cc: Dave That <dave.taht@gmail.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-05-19 19:04:22 +00:00
return 0;
fail:
tcp_sk_exit(net);
return res;
}
static void __net_exit tcp_sk_exit_batch(struct list_head *net_exit_list)
{
inet_twsk_purge(&tcp_hashinfo, &tcp_death_row, AF_INET);
}
static struct pernet_operations __net_initdata tcp_sk_ops = {
.init = tcp_sk_init,
.exit = tcp_sk_exit,
.exit_batch = tcp_sk_exit_batch,
};
void __init tcp_v4_init(void)
{
inet_hashinfo_init(&tcp_hashinfo);
if (register_pernet_subsys(&tcp_sk_ops))
panic("Failed to create the TCP control socket.\n");
}