linux/net/ipv4/af_inet.c

1767 lines
42 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.
*
* PF_INET protocol family socket handler.
*
* Authors: Ross Biro
* Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
* Florian La Roche, <flla@stud.uni-sb.de>
* Alan Cox, <A.Cox@swansea.ac.uk>
*
* Changes (see also sock.c)
*
* piggy,
* Karl Knutson : Socket protocol table
* A.N.Kuznetsov : Socket death error in accept().
* John Richardson : Fix non blocking error in connect()
* so sockets that fail to connect
* don't return -EINPROGRESS.
* Alan Cox : Asynchronous I/O support
* Alan Cox : Keep correct socket pointer on sock
* structures
* when accept() ed
* Alan Cox : Semantics of SO_LINGER aren't state
* moved to close when you look carefully.
* With this fixed and the accept bug fixed
* some RPC stuff seems happier.
* Niibe Yutaka : 4.4BSD style write async I/O
* Alan Cox,
* Tony Gale : Fixed reuse semantics.
* Alan Cox : bind() shouldn't abort existing but dead
* sockets. Stops FTP netin:.. I hope.
* Alan Cox : bind() works correctly for RAW sockets.
* Note that FreeBSD at least was broken
* in this respect so be careful with
* compatibility tests...
* Alan Cox : routing cache support
* Alan Cox : memzero the socket structure for
* compactness.
* Matt Day : nonblock connect error handler
* Alan Cox : Allow large numbers of pending sockets
* (eg for big web sites), but only if
* specifically application requested.
* Alan Cox : New buffering throughout IP. Used
* dumbly.
* Alan Cox : New buffering now used smartly.
* Alan Cox : BSD rather than common sense
* interpretation of listen.
* Germano Caronni : Assorted small races.
* Alan Cox : sendmsg/recvmsg basic support.
* Alan Cox : Only sendmsg/recvmsg now supported.
* Alan Cox : Locked down bind (see security list).
* Alan Cox : Loosened bind a little.
* Mike McLagan : ADD/DEL DLCI Ioctls
* Willy Konynenberg : Transparent proxying support.
* David S. Miller : New socket lookup architecture.
* Some other random speedups.
* Cyrus Durgin : Cleaned up file for kmod hacks.
* Andi Kleen : Fix inet_stream_connect TCP race.
*
* 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.
*/
#include <linux/err.h>
#include <linux/errno.h>
#include <linux/types.h>
#include <linux/socket.h>
#include <linux/in.h>
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/sched.h>
#include <linux/timer.h>
#include <linux/string.h>
#include <linux/sockios.h>
#include <linux/net.h>
#include <linux/capability.h>
#include <linux/fcntl.h>
#include <linux/mm.h>
#include <linux/interrupt.h>
#include <linux/stat.h>
#include <linux/init.h>
#include <linux/poll.h>
#include <linux/netfilter_ipv4.h>
#include <linux/random.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 <asm/uaccess.h>
#include <asm/system.h>
#include <linux/inet.h>
#include <linux/igmp.h>
#include <linux/inetdevice.h>
#include <linux/netdevice.h>
#include <net/checksum.h>
#include <net/ip.h>
#include <net/protocol.h>
#include <net/arp.h>
#include <net/route.h>
#include <net/ip_fib.h>
#include <net/inet_connection_sock.h>
#include <net/tcp.h>
#include <net/udp.h>
[NET]: Supporting UDP-Lite (RFC 3828) in Linux This is a revision of the previously submitted patch, which alters the way files are organized and compiled in the following manner: * UDP and UDP-Lite now use separate object files * source file dependencies resolved via header files net/ipv{4,6}/udp_impl.h * order of inclusion files in udp.c/udplite.c adapted accordingly [NET/IPv4]: Support for the UDP-Lite protocol (RFC 3828) This patch adds support for UDP-Lite to the IPv4 stack, provided as an extension to the existing UDPv4 code: * generic routines are all located in net/ipv4/udp.c * UDP-Lite specific routines are in net/ipv4/udplite.c * MIB/statistics support in /proc/net/snmp and /proc/net/udplite * shared API with extensions for partial checksum coverage [NET/IPv6]: Extension for UDP-Lite over IPv6 It extends the existing UDPv6 code base with support for UDP-Lite in the same manner as per UDPv4. In particular, * UDPv6 generic and shared code is in net/ipv6/udp.c * UDP-Litev6 specific extensions are in net/ipv6/udplite.c * MIB/statistics support in /proc/net/snmp6 and /proc/net/udplite6 * support for IPV6_ADDRFORM * aligned the coding style of protocol initialisation with af_inet6.c * made the error handling in udpv6_queue_rcv_skb consistent; to return `-1' on error on all error cases * consolidation of shared code [NET]: UDP-Lite Documentation and basic XFRM/Netfilter support The UDP-Lite patch further provides * API documentation for UDP-Lite * basic xfrm support * basic netfilter support for IPv4 and IPv6 (LOG target) Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-11-27 19:10:57 +00:00
#include <net/udplite.h>
#include <linux/skbuff.h>
#include <net/sock.h>
#include <net/raw.h>
#include <net/icmp.h>
#include <net/ipip.h>
#include <net/inet_common.h>
#include <net/xfrm.h>
#include <net/net_namespace.h>
#ifdef CONFIG_IP_MROUTE
#include <linux/mroute.h>
#endif
/* The inetsw table contains everything that inet_create needs to
* build a new socket.
*/
static struct list_head inetsw[SOCK_MAX];
static DEFINE_SPINLOCK(inetsw_lock);
struct ipv4_config ipv4_config;
EXPORT_SYMBOL(ipv4_config);
/* New destruction routine */
void inet_sock_destruct(struct sock *sk)
{
struct inet_sock *inet = inet_sk(sk);
__skb_queue_purge(&sk->sk_receive_queue);
__skb_queue_purge(&sk->sk_error_queue);
sk_mem_reclaim(sk);
if (sk->sk_type == SOCK_STREAM && sk->sk_state != TCP_CLOSE) {
pr_err("Attempt to release TCP socket in state %d %p\n",
sk->sk_state, sk);
return;
}
if (!sock_flag(sk, SOCK_DEAD)) {
pr_err("Attempt to release alive inet socket %p\n", sk);
return;
}
WARN_ON(atomic_read(&sk->sk_rmem_alloc));
WARN_ON(atomic_read(&sk->sk_wmem_alloc));
WARN_ON(sk->sk_wmem_queued);
WARN_ON(sk->sk_forward_alloc);
kfree(inet->opt);
dst_release(rcu_dereference_check(sk->sk_dst_cache, 1));
sk_refcnt_debug_dec(sk);
}
EXPORT_SYMBOL(inet_sock_destruct);
/*
* The routines beyond this point handle the behaviour of an AF_INET
* socket object. Mostly it punts to the subprotocols of IP to do
* the work.
*/
/*
* Automatically bind an unbound socket.
*/
static int inet_autobind(struct sock *sk)
{
struct inet_sock *inet;
/* We may need to bind the socket. */
lock_sock(sk);
inet = inet_sk(sk);
if (!inet->inet_num) {
if (sk->sk_prot->get_port(sk, 0)) {
release_sock(sk);
return -EAGAIN;
}
inet->inet_sport = htons(inet->inet_num);
}
release_sock(sk);
return 0;
}
/*
* Move a socket into listening state.
*/
int inet_listen(struct socket *sock, int backlog)
{
struct sock *sk = sock->sk;
unsigned char old_state;
int err;
lock_sock(sk);
err = -EINVAL;
if (sock->state != SS_UNCONNECTED || sock->type != SOCK_STREAM)
goto out;
old_state = sk->sk_state;
if (!((1 << old_state) & (TCPF_CLOSE | TCPF_LISTEN)))
goto out;
/* Really, if the socket is already in listen state
* we can only allow the backlog to be adjusted.
*/
if (old_state != TCP_LISTEN) {
err = inet_csk_listen_start(sk, backlog);
if (err)
goto out;
}
sk->sk_max_ack_backlog = backlog;
err = 0;
out:
release_sock(sk);
return err;
}
EXPORT_SYMBOL(inet_listen);
u32 inet_ehash_secret __read_mostly;
EXPORT_SYMBOL(inet_ehash_secret);
/*
* inet_ehash_secret must be set exactly once
*/
void build_ehash_secret(void)
{
u32 rnd;
do {
get_random_bytes(&rnd, sizeof(rnd));
} while (rnd == 0);
cmpxchg(&inet_ehash_secret, 0, rnd);
}
EXPORT_SYMBOL(build_ehash_secret);
static inline int inet_netns_ok(struct net *net, int protocol)
{
int hash;
const struct net_protocol *ipprot;
if (net_eq(net, &init_net))
return 1;
hash = protocol & (MAX_INET_PROTOS - 1);
ipprot = rcu_dereference(inet_protos[hash]);
if (ipprot == NULL)
/* raw IP is OK */
return 1;
return ipprot->netns_ok;
}
/*
* Create an inet socket.
*/
static int inet_create(struct net *net, struct socket *sock, int protocol,
int kern)
{
struct sock *sk;
struct inet_protosw *answer;
struct inet_sock *inet;
struct proto *answer_prot;
unsigned char answer_flags;
char answer_no_check;
int try_loading_module = 0;
int err;
if (unlikely(!inet_ehash_secret))
if (sock->type != SOCK_RAW && sock->type != SOCK_DGRAM)
build_ehash_secret();
sock->state = SS_UNCONNECTED;
/* Look for the requested type/protocol pair. */
lookup_protocol:
err = -ESOCKTNOSUPPORT;
rcu_read_lock();
list_for_each_entry_rcu(answer, &inetsw[sock->type], list) {
err = 0;
/* Check the non-wild match. */
if (protocol == answer->protocol) {
if (protocol != IPPROTO_IP)
break;
} else {
/* Check for the two wild cases. */
if (IPPROTO_IP == protocol) {
protocol = answer->protocol;
break;
}
if (IPPROTO_IP == answer->protocol)
break;
}
err = -EPROTONOSUPPORT;
}
if (unlikely(err)) {
if (try_loading_module < 2) {
rcu_read_unlock();
/*
* Be more specific, e.g. net-pf-2-proto-132-type-1
* (net-pf-PF_INET-proto-IPPROTO_SCTP-type-SOCK_STREAM)
*/
if (++try_loading_module == 1)
request_module("net-pf-%d-proto-%d-type-%d",
PF_INET, protocol, sock->type);
/*
* Fall back to generic, e.g. net-pf-2-proto-132
* (net-pf-PF_INET-proto-IPPROTO_SCTP)
*/
else
request_module("net-pf-%d-proto-%d",
PF_INET, protocol);
goto lookup_protocol;
} else
goto out_rcu_unlock;
}
err = -EPERM;
if (sock->type == SOCK_RAW && !kern && !capable(CAP_NET_RAW))
goto out_rcu_unlock;
err = -EAFNOSUPPORT;
if (!inet_netns_ok(net, protocol))
goto out_rcu_unlock;
sock->ops = answer->ops;
answer_prot = answer->prot;
answer_no_check = answer->no_check;
answer_flags = answer->flags;
rcu_read_unlock();
WARN_ON(answer_prot->slab == NULL);
err = -ENOBUFS;
sk = sk_alloc(net, PF_INET, GFP_KERNEL, answer_prot);
if (sk == NULL)
goto out;
err = 0;
sk->sk_no_check = answer_no_check;
if (INET_PROTOSW_REUSE & answer_flags)
sk->sk_reuse = 1;
inet = inet_sk(sk);
inet->is_icsk = (INET_PROTOSW_ICSK & answer_flags) != 0;
inet->nodefrag = 0;
if (SOCK_RAW == sock->type) {
inet->inet_num = protocol;
if (IPPROTO_RAW == protocol)
inet->hdrincl = 1;
}
if (ipv4_config.no_pmtu_disc)
inet->pmtudisc = IP_PMTUDISC_DONT;
else
inet->pmtudisc = IP_PMTUDISC_WANT;
inet->inet_id = 0;
sock_init_data(sock, sk);
sk->sk_destruct = inet_sock_destruct;
sk->sk_protocol = protocol;
sk->sk_backlog_rcv = sk->sk_prot->backlog_rcv;
inet->uc_ttl = -1;
inet->mc_loop = 1;
inet->mc_ttl = 1;
inet->mc_all = 1;
inet->mc_index = 0;
inet->mc_list = NULL;
sk_refcnt_debug_inc(sk);
if (inet->inet_num) {
/* It assumes that any protocol which allows
* the user to assign a number at socket
* creation time automatically
* shares.
*/
inet->inet_sport = htons(inet->inet_num);
/* Add to protocol hash chains. */
sk->sk_prot->hash(sk);
}
if (sk->sk_prot->init) {
err = sk->sk_prot->init(sk);
if (err)
sk_common_release(sk);
}
out:
return err;
out_rcu_unlock:
rcu_read_unlock();
goto out;
}
/*
* The peer socket should always be NULL (or else). When we call this
* function we are destroying the object and from then on nobody
* should refer to it.
*/
int inet_release(struct socket *sock)
{
struct sock *sk = sock->sk;
if (sk) {
long timeout;
sock_rps_reset_flow(sk);
rfs: Receive Flow Steering This patch implements receive flow steering (RFS). RFS steers received packets for layer 3 and 4 processing to the CPU where the application for the corresponding flow is running. RFS is an extension of Receive Packet Steering (RPS). The basic idea of RFS is that when an application calls recvmsg (or sendmsg) the application's running CPU is stored in a hash table that is indexed by the connection's rxhash which is stored in the socket structure. The rxhash is passed in skb's received on the connection from netif_receive_skb. For each received packet, the associated rxhash is used to look up the CPU in the hash table, if a valid CPU is set then the packet is steered to that CPU using the RPS mechanisms. The convolution of the simple approach is that it would potentially allow OOO packets. If threads are thrashing around CPUs or multiple threads are trying to read from the same sockets, a quickly changing CPU value in the hash table could cause rampant OOO packets-- we consider this a non-starter. To avoid OOO packets, this solution implements two types of hash tables: rps_sock_flow_table and rps_dev_flow_table. rps_sock_table is a global hash table. Each entry is just a CPU number and it is populated in recvmsg and sendmsg as described above. This table contains the "desired" CPUs for flows. rps_dev_flow_table is specific to each device queue. Each entry contains a CPU and a tail queue counter. The CPU is the "current" CPU for a matching flow. The tail queue counter holds the value of a tail queue counter for the associated CPU's backlog queue at the time of last enqueue for a flow matching the entry. Each backlog queue has a queue head counter which is incremented on dequeue, and so a queue tail counter is computed as queue head count + queue length. When a packet is enqueued on a backlog queue, the current value of the queue tail counter is saved in the hash entry of the rps_dev_flow_table. And now the trick: when selecting the CPU for RPS (get_rps_cpu) the rps_sock_flow table and the rps_dev_flow table for the RX queue are consulted. When the desired CPU for the flow (found in the rps_sock_flow table) does not match the current CPU (found in the rps_dev_flow table), the current CPU is changed to the desired CPU if one of the following is true: - The current CPU is unset (equal to RPS_NO_CPU) - Current CPU is offline - The current CPU's queue head counter >= queue tail counter in the rps_dev_flow table. This checks if the queue tail has advanced beyond the last packet that was enqueued using this table entry. This guarantees that all packets queued using this entry have been dequeued, thus preserving in order delivery. Making each queue have its own rps_dev_flow table has two advantages: 1) the tail queue counters will be written on each receive, so keeping the table local to interrupting CPU s good for locality. 2) this allows lockless access to the table-- the CPU number and queue tail counter need to be accessed together under mutual exclusion from netif_receive_skb, we assume that this is only called from device napi_poll which is non-reentrant. This patch implements RFS for TCP and connected UDP sockets. It should be usable for other flow oriented protocols. There are two configuration parameters for RFS. The "rps_flow_entries" kernel init parameter sets the number of entries in the rps_sock_flow_table, the per rxqueue sysfs entry "rps_flow_cnt" contains the number of entries in the rps_dev_flow table for the rxqueue. Both are rounded to power of two. The obvious benefit of RFS (over just RPS) is that it achieves CPU locality between the receive processing for a flow and the applications processing; this can result in increased performance (higher pps, lower latency). The benefits of RFS are dependent on cache hierarchy, application load, and other factors. On simple benchmarks, we don't necessarily see improvement and sometimes see degradation. However, for more complex benchmarks and for applications where cache pressure is much higher this technique seems to perform very well. Below are some benchmark results which show the potential benfit of this patch. The netperf test has 500 instances of netperf TCP_RR test with 1 byte req. and resp. The RPC test is an request/response test similar in structure to netperf RR test ith 100 threads on each host, but does more work in userspace that netperf. e1000e on 8 core Intel No RFS or RPS 104K tps at 30% CPU No RFS (best RPS config): 290K tps at 63% CPU RFS 303K tps at 61% CPU RPC test tps CPU% 50/90/99% usec latency Latency StdDev No RFS/RPS 103K 48% 757/900/3185 4472.35 RPS only: 174K 73% 415/993/2468 491.66 RFS 223K 73% 379/651/1382 315.61 Signed-off-by: Tom Herbert <therbert@google.com> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-16 23:01:27 +00:00
/* Applications forget to leave groups before exiting */
ip_mc_drop_socket(sk);
/* If linger is set, we don't return until the close
* is complete. Otherwise we return immediately. The
* actually closing is done the same either way.
*
* If the close is due to the process exiting, we never
* linger..
*/
timeout = 0;
if (sock_flag(sk, SOCK_LINGER) &&
!(current->flags & PF_EXITING))
timeout = sk->sk_lingertime;
sock->sk = NULL;
sk->sk_prot->close(sk, timeout);
}
return 0;
}
EXPORT_SYMBOL(inet_release);
/* It is off by default, see below. */
int sysctl_ip_nonlocal_bind __read_mostly;
EXPORT_SYMBOL(sysctl_ip_nonlocal_bind);
int inet_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
{
struct sockaddr_in *addr = (struct sockaddr_in *)uaddr;
struct sock *sk = sock->sk;
struct inet_sock *inet = inet_sk(sk);
unsigned short snum;
int chk_addr_ret;
int err;
/* If the socket has its own bind function then use it. (RAW) */
if (sk->sk_prot->bind) {
err = sk->sk_prot->bind(sk, uaddr, addr_len);
goto out;
}
err = -EINVAL;
if (addr_len < sizeof(struct sockaddr_in))
goto out;
chk_addr_ret = inet_addr_type(sock_net(sk), addr->sin_addr.s_addr);
/* Not specified by any standard per-se, however it breaks too
* many applications when removed. It is unfortunate since
* allowing applications to make a non-local bind solves
* several problems with systems using dynamic addressing.
* (ie. your servers still start up even if your ISDN link
* is temporarily down)
*/
err = -EADDRNOTAVAIL;
if (!sysctl_ip_nonlocal_bind &&
!(inet->freebind || inet->transparent) &&
addr->sin_addr.s_addr != htonl(INADDR_ANY) &&
chk_addr_ret != RTN_LOCAL &&
chk_addr_ret != RTN_MULTICAST &&
chk_addr_ret != RTN_BROADCAST)
goto out;
snum = ntohs(addr->sin_port);
err = -EACCES;
if (snum && snum < PROT_SOCK && !capable(CAP_NET_BIND_SERVICE))
goto out;
/* We keep a pair of addresses. rcv_saddr is the one
* used by hash lookups, and saddr is used for transmit.
*
* In the BSD API these are the same except where it
* would be illegal to use them (multicast/broadcast) in
* which case the sending device address is used.
*/
lock_sock(sk);
/* Check these errors (active socket, double bind). */
err = -EINVAL;
if (sk->sk_state != TCP_CLOSE || inet->inet_num)
goto out_release_sock;
inet->inet_rcv_saddr = inet->inet_saddr = addr->sin_addr.s_addr;
if (chk_addr_ret == RTN_MULTICAST || chk_addr_ret == RTN_BROADCAST)
inet->inet_saddr = 0; /* Use device */
/* Make sure we are allowed to bind here. */
if (sk->sk_prot->get_port(sk, snum)) {
inet->inet_saddr = inet->inet_rcv_saddr = 0;
err = -EADDRINUSE;
goto out_release_sock;
}
if (inet->inet_rcv_saddr)
sk->sk_userlocks |= SOCK_BINDADDR_LOCK;
if (snum)
sk->sk_userlocks |= SOCK_BINDPORT_LOCK;
inet->inet_sport = htons(inet->inet_num);
inet->inet_daddr = 0;
inet->inet_dport = 0;
sk_dst_reset(sk);
err = 0;
out_release_sock:
release_sock(sk);
out:
return err;
}
EXPORT_SYMBOL(inet_bind);
int inet_dgram_connect(struct socket *sock, struct sockaddr * uaddr,
int addr_len, int flags)
{
struct sock *sk = sock->sk;
if (addr_len < sizeof(uaddr->sa_family))
return -EINVAL;
if (uaddr->sa_family == AF_UNSPEC)
return sk->sk_prot->disconnect(sk, flags);
if (!inet_sk(sk)->inet_num && inet_autobind(sk))
return -EAGAIN;
return sk->sk_prot->connect(sk, (struct sockaddr *)uaddr, addr_len);
}
EXPORT_SYMBOL(inet_dgram_connect);
static long inet_wait_for_connect(struct sock *sk, long timeo)
{
DEFINE_WAIT(wait);
prepare_to_wait(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE);
/* Basic assumption: if someone sets sk->sk_err, he _must_
* change state of the socket from TCP_SYN_*.
* Connect() does not allow to get error notifications
* without closing the socket.
*/
while ((1 << sk->sk_state) & (TCPF_SYN_SENT | TCPF_SYN_RECV)) {
release_sock(sk);
timeo = schedule_timeout(timeo);
lock_sock(sk);
if (signal_pending(current) || !timeo)
break;
prepare_to_wait(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE);
}
finish_wait(sk_sleep(sk), &wait);
return timeo;
}
/*
* Connect to a remote host. There is regrettably still a little
* TCP 'magic' in here.
*/
int inet_stream_connect(struct socket *sock, struct sockaddr *uaddr,
int addr_len, int flags)
{
struct sock *sk = sock->sk;
int err;
long timeo;
if (addr_len < sizeof(uaddr->sa_family))
return -EINVAL;
lock_sock(sk);
if (uaddr->sa_family == AF_UNSPEC) {
err = sk->sk_prot->disconnect(sk, flags);
sock->state = err ? SS_DISCONNECTING : SS_UNCONNECTED;
goto out;
}
switch (sock->state) {
default:
err = -EINVAL;
goto out;
case SS_CONNECTED:
err = -EISCONN;
goto out;
case SS_CONNECTING:
err = -EALREADY;
/* Fall out of switch with err, set for this state */
break;
case SS_UNCONNECTED:
err = -EISCONN;
if (sk->sk_state != TCP_CLOSE)
goto out;
err = sk->sk_prot->connect(sk, uaddr, addr_len);
if (err < 0)
goto out;
sock->state = SS_CONNECTING;
/* Just entered SS_CONNECTING state; the only
* difference is that return value in non-blocking
* case is EINPROGRESS, rather than EALREADY.
*/
err = -EINPROGRESS;
break;
}
timeo = sock_sndtimeo(sk, flags & O_NONBLOCK);
if ((1 << sk->sk_state) & (TCPF_SYN_SENT | TCPF_SYN_RECV)) {
/* Error code is set above */
if (!timeo || !inet_wait_for_connect(sk, timeo))
goto out;
err = sock_intr_errno(timeo);
if (signal_pending(current))
goto out;
}
/* Connection was closed by RST, timeout, ICMP error
* or another process disconnected us.
*/
if (sk->sk_state == TCP_CLOSE)
goto sock_error;
/* sk->sk_err may be not zero now, if RECVERR was ordered by user
* and error was received after socket entered established state.
* Hence, it is handled normally after connect() return successfully.
*/
sock->state = SS_CONNECTED;
err = 0;
out:
release_sock(sk);
return err;
sock_error:
err = sock_error(sk) ? : -ECONNABORTED;
sock->state = SS_UNCONNECTED;
if (sk->sk_prot->disconnect(sk, flags))
sock->state = SS_DISCONNECTING;
goto out;
}
EXPORT_SYMBOL(inet_stream_connect);
/*
* Accept a pending connection. The TCP layer now gives BSD semantics.
*/
int inet_accept(struct socket *sock, struct socket *newsock, int flags)
{
struct sock *sk1 = sock->sk;
int err = -EINVAL;
struct sock *sk2 = sk1->sk_prot->accept(sk1, flags, &err);
if (!sk2)
goto do_err;
lock_sock(sk2);
WARN_ON(!((1 << sk2->sk_state) &
(TCPF_ESTABLISHED | TCPF_CLOSE_WAIT | TCPF_CLOSE)));
sock_graft(sk2, newsock);
newsock->state = SS_CONNECTED;
err = 0;
release_sock(sk2);
do_err:
return err;
}
EXPORT_SYMBOL(inet_accept);
/*
* This does both peername and sockname.
*/
int inet_getname(struct socket *sock, struct sockaddr *uaddr,
int *uaddr_len, int peer)
{
struct sock *sk = sock->sk;
struct inet_sock *inet = inet_sk(sk);
DECLARE_SOCKADDR(struct sockaddr_in *, sin, uaddr);
sin->sin_family = AF_INET;
if (peer) {
if (!inet->inet_dport ||
(((1 << sk->sk_state) & (TCPF_CLOSE | TCPF_SYN_SENT)) &&
peer == 1))
return -ENOTCONN;
sin->sin_port = inet->inet_dport;
sin->sin_addr.s_addr = inet->inet_daddr;
} else {
__be32 addr = inet->inet_rcv_saddr;
if (!addr)
addr = inet->inet_saddr;
sin->sin_port = inet->inet_sport;
sin->sin_addr.s_addr = addr;
}
memset(sin->sin_zero, 0, sizeof(sin->sin_zero));
*uaddr_len = sizeof(*sin);
return 0;
}
EXPORT_SYMBOL(inet_getname);
int inet_sendmsg(struct kiocb *iocb, struct socket *sock, struct msghdr *msg,
size_t size)
{
struct sock *sk = sock->sk;
sock_rps_record_flow(sk);
rfs: Receive Flow Steering This patch implements receive flow steering (RFS). RFS steers received packets for layer 3 and 4 processing to the CPU where the application for the corresponding flow is running. RFS is an extension of Receive Packet Steering (RPS). The basic idea of RFS is that when an application calls recvmsg (or sendmsg) the application's running CPU is stored in a hash table that is indexed by the connection's rxhash which is stored in the socket structure. The rxhash is passed in skb's received on the connection from netif_receive_skb. For each received packet, the associated rxhash is used to look up the CPU in the hash table, if a valid CPU is set then the packet is steered to that CPU using the RPS mechanisms. The convolution of the simple approach is that it would potentially allow OOO packets. If threads are thrashing around CPUs or multiple threads are trying to read from the same sockets, a quickly changing CPU value in the hash table could cause rampant OOO packets-- we consider this a non-starter. To avoid OOO packets, this solution implements two types of hash tables: rps_sock_flow_table and rps_dev_flow_table. rps_sock_table is a global hash table. Each entry is just a CPU number and it is populated in recvmsg and sendmsg as described above. This table contains the "desired" CPUs for flows. rps_dev_flow_table is specific to each device queue. Each entry contains a CPU and a tail queue counter. The CPU is the "current" CPU for a matching flow. The tail queue counter holds the value of a tail queue counter for the associated CPU's backlog queue at the time of last enqueue for a flow matching the entry. Each backlog queue has a queue head counter which is incremented on dequeue, and so a queue tail counter is computed as queue head count + queue length. When a packet is enqueued on a backlog queue, the current value of the queue tail counter is saved in the hash entry of the rps_dev_flow_table. And now the trick: when selecting the CPU for RPS (get_rps_cpu) the rps_sock_flow table and the rps_dev_flow table for the RX queue are consulted. When the desired CPU for the flow (found in the rps_sock_flow table) does not match the current CPU (found in the rps_dev_flow table), the current CPU is changed to the desired CPU if one of the following is true: - The current CPU is unset (equal to RPS_NO_CPU) - Current CPU is offline - The current CPU's queue head counter >= queue tail counter in the rps_dev_flow table. This checks if the queue tail has advanced beyond the last packet that was enqueued using this table entry. This guarantees that all packets queued using this entry have been dequeued, thus preserving in order delivery. Making each queue have its own rps_dev_flow table has two advantages: 1) the tail queue counters will be written on each receive, so keeping the table local to interrupting CPU s good for locality. 2) this allows lockless access to the table-- the CPU number and queue tail counter need to be accessed together under mutual exclusion from netif_receive_skb, we assume that this is only called from device napi_poll which is non-reentrant. This patch implements RFS for TCP and connected UDP sockets. It should be usable for other flow oriented protocols. There are two configuration parameters for RFS. The "rps_flow_entries" kernel init parameter sets the number of entries in the rps_sock_flow_table, the per rxqueue sysfs entry "rps_flow_cnt" contains the number of entries in the rps_dev_flow table for the rxqueue. Both are rounded to power of two. The obvious benefit of RFS (over just RPS) is that it achieves CPU locality between the receive processing for a flow and the applications processing; this can result in increased performance (higher pps, lower latency). The benefits of RFS are dependent on cache hierarchy, application load, and other factors. On simple benchmarks, we don't necessarily see improvement and sometimes see degradation. However, for more complex benchmarks and for applications where cache pressure is much higher this technique seems to perform very well. Below are some benchmark results which show the potential benfit of this patch. The netperf test has 500 instances of netperf TCP_RR test with 1 byte req. and resp. The RPC test is an request/response test similar in structure to netperf RR test ith 100 threads on each host, but does more work in userspace that netperf. e1000e on 8 core Intel No RFS or RPS 104K tps at 30% CPU No RFS (best RPS config): 290K tps at 63% CPU RFS 303K tps at 61% CPU RPC test tps CPU% 50/90/99% usec latency Latency StdDev No RFS/RPS 103K 48% 757/900/3185 4472.35 RPS only: 174K 73% 415/993/2468 491.66 RFS 223K 73% 379/651/1382 315.61 Signed-off-by: Tom Herbert <therbert@google.com> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-16 23:01:27 +00:00
/* We may need to bind the socket. */
if (!inet_sk(sk)->inet_num && !sk->sk_prot->no_autobind &&
inet_autobind(sk))
return -EAGAIN;
return sk->sk_prot->sendmsg(iocb, sk, msg, size);
}
EXPORT_SYMBOL(inet_sendmsg);
ssize_t inet_sendpage(struct socket *sock, struct page *page, int offset,
size_t size, int flags)
{
struct sock *sk = sock->sk;
sock_rps_record_flow(sk);
rfs: Receive Flow Steering This patch implements receive flow steering (RFS). RFS steers received packets for layer 3 and 4 processing to the CPU where the application for the corresponding flow is running. RFS is an extension of Receive Packet Steering (RPS). The basic idea of RFS is that when an application calls recvmsg (or sendmsg) the application's running CPU is stored in a hash table that is indexed by the connection's rxhash which is stored in the socket structure. The rxhash is passed in skb's received on the connection from netif_receive_skb. For each received packet, the associated rxhash is used to look up the CPU in the hash table, if a valid CPU is set then the packet is steered to that CPU using the RPS mechanisms. The convolution of the simple approach is that it would potentially allow OOO packets. If threads are thrashing around CPUs or multiple threads are trying to read from the same sockets, a quickly changing CPU value in the hash table could cause rampant OOO packets-- we consider this a non-starter. To avoid OOO packets, this solution implements two types of hash tables: rps_sock_flow_table and rps_dev_flow_table. rps_sock_table is a global hash table. Each entry is just a CPU number and it is populated in recvmsg and sendmsg as described above. This table contains the "desired" CPUs for flows. rps_dev_flow_table is specific to each device queue. Each entry contains a CPU and a tail queue counter. The CPU is the "current" CPU for a matching flow. The tail queue counter holds the value of a tail queue counter for the associated CPU's backlog queue at the time of last enqueue for a flow matching the entry. Each backlog queue has a queue head counter which is incremented on dequeue, and so a queue tail counter is computed as queue head count + queue length. When a packet is enqueued on a backlog queue, the current value of the queue tail counter is saved in the hash entry of the rps_dev_flow_table. And now the trick: when selecting the CPU for RPS (get_rps_cpu) the rps_sock_flow table and the rps_dev_flow table for the RX queue are consulted. When the desired CPU for the flow (found in the rps_sock_flow table) does not match the current CPU (found in the rps_dev_flow table), the current CPU is changed to the desired CPU if one of the following is true: - The current CPU is unset (equal to RPS_NO_CPU) - Current CPU is offline - The current CPU's queue head counter >= queue tail counter in the rps_dev_flow table. This checks if the queue tail has advanced beyond the last packet that was enqueued using this table entry. This guarantees that all packets queued using this entry have been dequeued, thus preserving in order delivery. Making each queue have its own rps_dev_flow table has two advantages: 1) the tail queue counters will be written on each receive, so keeping the table local to interrupting CPU s good for locality. 2) this allows lockless access to the table-- the CPU number and queue tail counter need to be accessed together under mutual exclusion from netif_receive_skb, we assume that this is only called from device napi_poll which is non-reentrant. This patch implements RFS for TCP and connected UDP sockets. It should be usable for other flow oriented protocols. There are two configuration parameters for RFS. The "rps_flow_entries" kernel init parameter sets the number of entries in the rps_sock_flow_table, the per rxqueue sysfs entry "rps_flow_cnt" contains the number of entries in the rps_dev_flow table for the rxqueue. Both are rounded to power of two. The obvious benefit of RFS (over just RPS) is that it achieves CPU locality between the receive processing for a flow and the applications processing; this can result in increased performance (higher pps, lower latency). The benefits of RFS are dependent on cache hierarchy, application load, and other factors. On simple benchmarks, we don't necessarily see improvement and sometimes see degradation. However, for more complex benchmarks and for applications where cache pressure is much higher this technique seems to perform very well. Below are some benchmark results which show the potential benfit of this patch. The netperf test has 500 instances of netperf TCP_RR test with 1 byte req. and resp. The RPC test is an request/response test similar in structure to netperf RR test ith 100 threads on each host, but does more work in userspace that netperf. e1000e on 8 core Intel No RFS or RPS 104K tps at 30% CPU No RFS (best RPS config): 290K tps at 63% CPU RFS 303K tps at 61% CPU RPC test tps CPU% 50/90/99% usec latency Latency StdDev No RFS/RPS 103K 48% 757/900/3185 4472.35 RPS only: 174K 73% 415/993/2468 491.66 RFS 223K 73% 379/651/1382 315.61 Signed-off-by: Tom Herbert <therbert@google.com> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-16 23:01:27 +00:00
/* We may need to bind the socket. */
if (!inet_sk(sk)->inet_num && !sk->sk_prot->no_autobind &&
inet_autobind(sk))
return -EAGAIN;
if (sk->sk_prot->sendpage)
return sk->sk_prot->sendpage(sk, page, offset, size, flags);
return sock_no_sendpage(sock, page, offset, size, flags);
}
EXPORT_SYMBOL(inet_sendpage);
rfs: Receive Flow Steering This patch implements receive flow steering (RFS). RFS steers received packets for layer 3 and 4 processing to the CPU where the application for the corresponding flow is running. RFS is an extension of Receive Packet Steering (RPS). The basic idea of RFS is that when an application calls recvmsg (or sendmsg) the application's running CPU is stored in a hash table that is indexed by the connection's rxhash which is stored in the socket structure. The rxhash is passed in skb's received on the connection from netif_receive_skb. For each received packet, the associated rxhash is used to look up the CPU in the hash table, if a valid CPU is set then the packet is steered to that CPU using the RPS mechanisms. The convolution of the simple approach is that it would potentially allow OOO packets. If threads are thrashing around CPUs or multiple threads are trying to read from the same sockets, a quickly changing CPU value in the hash table could cause rampant OOO packets-- we consider this a non-starter. To avoid OOO packets, this solution implements two types of hash tables: rps_sock_flow_table and rps_dev_flow_table. rps_sock_table is a global hash table. Each entry is just a CPU number and it is populated in recvmsg and sendmsg as described above. This table contains the "desired" CPUs for flows. rps_dev_flow_table is specific to each device queue. Each entry contains a CPU and a tail queue counter. The CPU is the "current" CPU for a matching flow. The tail queue counter holds the value of a tail queue counter for the associated CPU's backlog queue at the time of last enqueue for a flow matching the entry. Each backlog queue has a queue head counter which is incremented on dequeue, and so a queue tail counter is computed as queue head count + queue length. When a packet is enqueued on a backlog queue, the current value of the queue tail counter is saved in the hash entry of the rps_dev_flow_table. And now the trick: when selecting the CPU for RPS (get_rps_cpu) the rps_sock_flow table and the rps_dev_flow table for the RX queue are consulted. When the desired CPU for the flow (found in the rps_sock_flow table) does not match the current CPU (found in the rps_dev_flow table), the current CPU is changed to the desired CPU if one of the following is true: - The current CPU is unset (equal to RPS_NO_CPU) - Current CPU is offline - The current CPU's queue head counter >= queue tail counter in the rps_dev_flow table. This checks if the queue tail has advanced beyond the last packet that was enqueued using this table entry. This guarantees that all packets queued using this entry have been dequeued, thus preserving in order delivery. Making each queue have its own rps_dev_flow table has two advantages: 1) the tail queue counters will be written on each receive, so keeping the table local to interrupting CPU s good for locality. 2) this allows lockless access to the table-- the CPU number and queue tail counter need to be accessed together under mutual exclusion from netif_receive_skb, we assume that this is only called from device napi_poll which is non-reentrant. This patch implements RFS for TCP and connected UDP sockets. It should be usable for other flow oriented protocols. There are two configuration parameters for RFS. The "rps_flow_entries" kernel init parameter sets the number of entries in the rps_sock_flow_table, the per rxqueue sysfs entry "rps_flow_cnt" contains the number of entries in the rps_dev_flow table for the rxqueue. Both are rounded to power of two. The obvious benefit of RFS (over just RPS) is that it achieves CPU locality between the receive processing for a flow and the applications processing; this can result in increased performance (higher pps, lower latency). The benefits of RFS are dependent on cache hierarchy, application load, and other factors. On simple benchmarks, we don't necessarily see improvement and sometimes see degradation. However, for more complex benchmarks and for applications where cache pressure is much higher this technique seems to perform very well. Below are some benchmark results which show the potential benfit of this patch. The netperf test has 500 instances of netperf TCP_RR test with 1 byte req. and resp. The RPC test is an request/response test similar in structure to netperf RR test ith 100 threads on each host, but does more work in userspace that netperf. e1000e on 8 core Intel No RFS or RPS 104K tps at 30% CPU No RFS (best RPS config): 290K tps at 63% CPU RFS 303K tps at 61% CPU RPC test tps CPU% 50/90/99% usec latency Latency StdDev No RFS/RPS 103K 48% 757/900/3185 4472.35 RPS only: 174K 73% 415/993/2468 491.66 RFS 223K 73% 379/651/1382 315.61 Signed-off-by: Tom Herbert <therbert@google.com> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-16 23:01:27 +00:00
int inet_recvmsg(struct kiocb *iocb, struct socket *sock, struct msghdr *msg,
size_t size, int flags)
{
struct sock *sk = sock->sk;
int addr_len = 0;
int err;
sock_rps_record_flow(sk);
rfs: Receive Flow Steering This patch implements receive flow steering (RFS). RFS steers received packets for layer 3 and 4 processing to the CPU where the application for the corresponding flow is running. RFS is an extension of Receive Packet Steering (RPS). The basic idea of RFS is that when an application calls recvmsg (or sendmsg) the application's running CPU is stored in a hash table that is indexed by the connection's rxhash which is stored in the socket structure. The rxhash is passed in skb's received on the connection from netif_receive_skb. For each received packet, the associated rxhash is used to look up the CPU in the hash table, if a valid CPU is set then the packet is steered to that CPU using the RPS mechanisms. The convolution of the simple approach is that it would potentially allow OOO packets. If threads are thrashing around CPUs or multiple threads are trying to read from the same sockets, a quickly changing CPU value in the hash table could cause rampant OOO packets-- we consider this a non-starter. To avoid OOO packets, this solution implements two types of hash tables: rps_sock_flow_table and rps_dev_flow_table. rps_sock_table is a global hash table. Each entry is just a CPU number and it is populated in recvmsg and sendmsg as described above. This table contains the "desired" CPUs for flows. rps_dev_flow_table is specific to each device queue. Each entry contains a CPU and a tail queue counter. The CPU is the "current" CPU for a matching flow. The tail queue counter holds the value of a tail queue counter for the associated CPU's backlog queue at the time of last enqueue for a flow matching the entry. Each backlog queue has a queue head counter which is incremented on dequeue, and so a queue tail counter is computed as queue head count + queue length. When a packet is enqueued on a backlog queue, the current value of the queue tail counter is saved in the hash entry of the rps_dev_flow_table. And now the trick: when selecting the CPU for RPS (get_rps_cpu) the rps_sock_flow table and the rps_dev_flow table for the RX queue are consulted. When the desired CPU for the flow (found in the rps_sock_flow table) does not match the current CPU (found in the rps_dev_flow table), the current CPU is changed to the desired CPU if one of the following is true: - The current CPU is unset (equal to RPS_NO_CPU) - Current CPU is offline - The current CPU's queue head counter >= queue tail counter in the rps_dev_flow table. This checks if the queue tail has advanced beyond the last packet that was enqueued using this table entry. This guarantees that all packets queued using this entry have been dequeued, thus preserving in order delivery. Making each queue have its own rps_dev_flow table has two advantages: 1) the tail queue counters will be written on each receive, so keeping the table local to interrupting CPU s good for locality. 2) this allows lockless access to the table-- the CPU number and queue tail counter need to be accessed together under mutual exclusion from netif_receive_skb, we assume that this is only called from device napi_poll which is non-reentrant. This patch implements RFS for TCP and connected UDP sockets. It should be usable for other flow oriented protocols. There are two configuration parameters for RFS. The "rps_flow_entries" kernel init parameter sets the number of entries in the rps_sock_flow_table, the per rxqueue sysfs entry "rps_flow_cnt" contains the number of entries in the rps_dev_flow table for the rxqueue. Both are rounded to power of two. The obvious benefit of RFS (over just RPS) is that it achieves CPU locality between the receive processing for a flow and the applications processing; this can result in increased performance (higher pps, lower latency). The benefits of RFS are dependent on cache hierarchy, application load, and other factors. On simple benchmarks, we don't necessarily see improvement and sometimes see degradation. However, for more complex benchmarks and for applications where cache pressure is much higher this technique seems to perform very well. Below are some benchmark results which show the potential benfit of this patch. The netperf test has 500 instances of netperf TCP_RR test with 1 byte req. and resp. The RPC test is an request/response test similar in structure to netperf RR test ith 100 threads on each host, but does more work in userspace that netperf. e1000e on 8 core Intel No RFS or RPS 104K tps at 30% CPU No RFS (best RPS config): 290K tps at 63% CPU RFS 303K tps at 61% CPU RPC test tps CPU% 50/90/99% usec latency Latency StdDev No RFS/RPS 103K 48% 757/900/3185 4472.35 RPS only: 174K 73% 415/993/2468 491.66 RFS 223K 73% 379/651/1382 315.61 Signed-off-by: Tom Herbert <therbert@google.com> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-16 23:01:27 +00:00
err = sk->sk_prot->recvmsg(iocb, sk, msg, size, flags & MSG_DONTWAIT,
flags & ~MSG_DONTWAIT, &addr_len);
if (err >= 0)
msg->msg_namelen = addr_len;
return err;
}
EXPORT_SYMBOL(inet_recvmsg);
int inet_shutdown(struct socket *sock, int how)
{
struct sock *sk = sock->sk;
int err = 0;
/* This should really check to make sure
* the socket is a TCP socket. (WHY AC...)
*/
how++; /* maps 0->1 has the advantage of making bit 1 rcvs and
1->2 bit 2 snds.
2->3 */
if ((how & ~SHUTDOWN_MASK) || !how) /* MAXINT->0 */
return -EINVAL;
lock_sock(sk);
if (sock->state == SS_CONNECTING) {
if ((1 << sk->sk_state) &
(TCPF_SYN_SENT | TCPF_SYN_RECV | TCPF_CLOSE))
sock->state = SS_DISCONNECTING;
else
sock->state = SS_CONNECTED;
}
switch (sk->sk_state) {
case TCP_CLOSE:
err = -ENOTCONN;
/* Hack to wake up other listeners, who can poll for
POLLHUP, even on eg. unconnected UDP sockets -- RR */
default:
sk->sk_shutdown |= how;
if (sk->sk_prot->shutdown)
sk->sk_prot->shutdown(sk, how);
break;
/* Remaining two branches are temporary solution for missing
* close() in multithreaded environment. It is _not_ a good idea,
* but we have no choice until close() is repaired at VFS level.
*/
case TCP_LISTEN:
if (!(how & RCV_SHUTDOWN))
break;
/* Fall through */
case TCP_SYN_SENT:
err = sk->sk_prot->disconnect(sk, O_NONBLOCK);
sock->state = err ? SS_DISCONNECTING : SS_UNCONNECTED;
break;
}
/* Wake up anyone sleeping in poll. */
sk->sk_state_change(sk);
release_sock(sk);
return err;
}
EXPORT_SYMBOL(inet_shutdown);
/*
* ioctl() calls you can issue on an INET socket. Most of these are
* device configuration and stuff and very rarely used. Some ioctls
* pass on to the socket itself.
*
* NOTE: I like the idea of a module for the config stuff. ie ifconfig
* loads the devconfigure module does its configuring and unloads it.
* There's a good 20K of config code hanging around the kernel.
*/
int inet_ioctl(struct socket *sock, unsigned int cmd, unsigned long arg)
{
struct sock *sk = sock->sk;
int err = 0;
struct net *net = sock_net(sk);
switch (cmd) {
case SIOCGSTAMP:
err = sock_get_timestamp(sk, (struct timeval __user *)arg);
break;
case SIOCGSTAMPNS:
err = sock_get_timestampns(sk, (struct timespec __user *)arg);
break;
case SIOCADDRT:
case SIOCDELRT:
case SIOCRTMSG:
err = ip_rt_ioctl(net, cmd, (void __user *)arg);
break;
case SIOCDARP:
case SIOCGARP:
case SIOCSARP:
err = arp_ioctl(net, cmd, (void __user *)arg);
break;
case SIOCGIFADDR:
case SIOCSIFADDR:
case SIOCGIFBRDADDR:
case SIOCSIFBRDADDR:
case SIOCGIFNETMASK:
case SIOCSIFNETMASK:
case SIOCGIFDSTADDR:
case SIOCSIFDSTADDR:
case SIOCSIFPFLAGS:
case SIOCGIFPFLAGS:
case SIOCSIFFLAGS:
err = devinet_ioctl(net, cmd, (void __user *)arg);
break;
default:
if (sk->sk_prot->ioctl)
err = sk->sk_prot->ioctl(sk, cmd, arg);
else
err = -ENOIOCTLCMD;
break;
}
return err;
}
EXPORT_SYMBOL(inet_ioctl);
#ifdef CONFIG_COMPAT
int inet_compat_ioctl(struct socket *sock, unsigned int cmd, unsigned long arg)
{
struct sock *sk = sock->sk;
int err = -ENOIOCTLCMD;
if (sk->sk_prot->compat_ioctl)
err = sk->sk_prot->compat_ioctl(sk, cmd, arg);
return err;
}
#endif
const struct proto_ops inet_stream_ops = {
.family = PF_INET,
.owner = THIS_MODULE,
.release = inet_release,
.bind = inet_bind,
.connect = inet_stream_connect,
.socketpair = sock_no_socketpair,
.accept = inet_accept,
.getname = inet_getname,
.poll = tcp_poll,
.ioctl = inet_ioctl,
.listen = inet_listen,
.shutdown = inet_shutdown,
.setsockopt = sock_common_setsockopt,
.getsockopt = sock_common_getsockopt,
.sendmsg = inet_sendmsg,
rfs: Receive Flow Steering This patch implements receive flow steering (RFS). RFS steers received packets for layer 3 and 4 processing to the CPU where the application for the corresponding flow is running. RFS is an extension of Receive Packet Steering (RPS). The basic idea of RFS is that when an application calls recvmsg (or sendmsg) the application's running CPU is stored in a hash table that is indexed by the connection's rxhash which is stored in the socket structure. The rxhash is passed in skb's received on the connection from netif_receive_skb. For each received packet, the associated rxhash is used to look up the CPU in the hash table, if a valid CPU is set then the packet is steered to that CPU using the RPS mechanisms. The convolution of the simple approach is that it would potentially allow OOO packets. If threads are thrashing around CPUs or multiple threads are trying to read from the same sockets, a quickly changing CPU value in the hash table could cause rampant OOO packets-- we consider this a non-starter. To avoid OOO packets, this solution implements two types of hash tables: rps_sock_flow_table and rps_dev_flow_table. rps_sock_table is a global hash table. Each entry is just a CPU number and it is populated in recvmsg and sendmsg as described above. This table contains the "desired" CPUs for flows. rps_dev_flow_table is specific to each device queue. Each entry contains a CPU and a tail queue counter. The CPU is the "current" CPU for a matching flow. The tail queue counter holds the value of a tail queue counter for the associated CPU's backlog queue at the time of last enqueue for a flow matching the entry. Each backlog queue has a queue head counter which is incremented on dequeue, and so a queue tail counter is computed as queue head count + queue length. When a packet is enqueued on a backlog queue, the current value of the queue tail counter is saved in the hash entry of the rps_dev_flow_table. And now the trick: when selecting the CPU for RPS (get_rps_cpu) the rps_sock_flow table and the rps_dev_flow table for the RX queue are consulted. When the desired CPU for the flow (found in the rps_sock_flow table) does not match the current CPU (found in the rps_dev_flow table), the current CPU is changed to the desired CPU if one of the following is true: - The current CPU is unset (equal to RPS_NO_CPU) - Current CPU is offline - The current CPU's queue head counter >= queue tail counter in the rps_dev_flow table. This checks if the queue tail has advanced beyond the last packet that was enqueued using this table entry. This guarantees that all packets queued using this entry have been dequeued, thus preserving in order delivery. Making each queue have its own rps_dev_flow table has two advantages: 1) the tail queue counters will be written on each receive, so keeping the table local to interrupting CPU s good for locality. 2) this allows lockless access to the table-- the CPU number and queue tail counter need to be accessed together under mutual exclusion from netif_receive_skb, we assume that this is only called from device napi_poll which is non-reentrant. This patch implements RFS for TCP and connected UDP sockets. It should be usable for other flow oriented protocols. There are two configuration parameters for RFS. The "rps_flow_entries" kernel init parameter sets the number of entries in the rps_sock_flow_table, the per rxqueue sysfs entry "rps_flow_cnt" contains the number of entries in the rps_dev_flow table for the rxqueue. Both are rounded to power of two. The obvious benefit of RFS (over just RPS) is that it achieves CPU locality between the receive processing for a flow and the applications processing; this can result in increased performance (higher pps, lower latency). The benefits of RFS are dependent on cache hierarchy, application load, and other factors. On simple benchmarks, we don't necessarily see improvement and sometimes see degradation. However, for more complex benchmarks and for applications where cache pressure is much higher this technique seems to perform very well. Below are some benchmark results which show the potential benfit of this patch. The netperf test has 500 instances of netperf TCP_RR test with 1 byte req. and resp. The RPC test is an request/response test similar in structure to netperf RR test ith 100 threads on each host, but does more work in userspace that netperf. e1000e on 8 core Intel No RFS or RPS 104K tps at 30% CPU No RFS (best RPS config): 290K tps at 63% CPU RFS 303K tps at 61% CPU RPC test tps CPU% 50/90/99% usec latency Latency StdDev No RFS/RPS 103K 48% 757/900/3185 4472.35 RPS only: 174K 73% 415/993/2468 491.66 RFS 223K 73% 379/651/1382 315.61 Signed-off-by: Tom Herbert <therbert@google.com> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-16 23:01:27 +00:00
.recvmsg = inet_recvmsg,
.mmap = sock_no_mmap,
.sendpage = inet_sendpage,
.splice_read = tcp_splice_read,
#ifdef CONFIG_COMPAT
.compat_setsockopt = compat_sock_common_setsockopt,
.compat_getsockopt = compat_sock_common_getsockopt,
.compat_ioctl = inet_compat_ioctl,
#endif
};
EXPORT_SYMBOL(inet_stream_ops);
const struct proto_ops inet_dgram_ops = {
.family = PF_INET,
.owner = THIS_MODULE,
.release = inet_release,
.bind = inet_bind,
.connect = inet_dgram_connect,
.socketpair = sock_no_socketpair,
.accept = sock_no_accept,
.getname = inet_getname,
.poll = udp_poll,
.ioctl = inet_ioctl,
.listen = sock_no_listen,
.shutdown = inet_shutdown,
.setsockopt = sock_common_setsockopt,
.getsockopt = sock_common_getsockopt,
.sendmsg = inet_sendmsg,
rfs: Receive Flow Steering This patch implements receive flow steering (RFS). RFS steers received packets for layer 3 and 4 processing to the CPU where the application for the corresponding flow is running. RFS is an extension of Receive Packet Steering (RPS). The basic idea of RFS is that when an application calls recvmsg (or sendmsg) the application's running CPU is stored in a hash table that is indexed by the connection's rxhash which is stored in the socket structure. The rxhash is passed in skb's received on the connection from netif_receive_skb. For each received packet, the associated rxhash is used to look up the CPU in the hash table, if a valid CPU is set then the packet is steered to that CPU using the RPS mechanisms. The convolution of the simple approach is that it would potentially allow OOO packets. If threads are thrashing around CPUs or multiple threads are trying to read from the same sockets, a quickly changing CPU value in the hash table could cause rampant OOO packets-- we consider this a non-starter. To avoid OOO packets, this solution implements two types of hash tables: rps_sock_flow_table and rps_dev_flow_table. rps_sock_table is a global hash table. Each entry is just a CPU number and it is populated in recvmsg and sendmsg as described above. This table contains the "desired" CPUs for flows. rps_dev_flow_table is specific to each device queue. Each entry contains a CPU and a tail queue counter. The CPU is the "current" CPU for a matching flow. The tail queue counter holds the value of a tail queue counter for the associated CPU's backlog queue at the time of last enqueue for a flow matching the entry. Each backlog queue has a queue head counter which is incremented on dequeue, and so a queue tail counter is computed as queue head count + queue length. When a packet is enqueued on a backlog queue, the current value of the queue tail counter is saved in the hash entry of the rps_dev_flow_table. And now the trick: when selecting the CPU for RPS (get_rps_cpu) the rps_sock_flow table and the rps_dev_flow table for the RX queue are consulted. When the desired CPU for the flow (found in the rps_sock_flow table) does not match the current CPU (found in the rps_dev_flow table), the current CPU is changed to the desired CPU if one of the following is true: - The current CPU is unset (equal to RPS_NO_CPU) - Current CPU is offline - The current CPU's queue head counter >= queue tail counter in the rps_dev_flow table. This checks if the queue tail has advanced beyond the last packet that was enqueued using this table entry. This guarantees that all packets queued using this entry have been dequeued, thus preserving in order delivery. Making each queue have its own rps_dev_flow table has two advantages: 1) the tail queue counters will be written on each receive, so keeping the table local to interrupting CPU s good for locality. 2) this allows lockless access to the table-- the CPU number and queue tail counter need to be accessed together under mutual exclusion from netif_receive_skb, we assume that this is only called from device napi_poll which is non-reentrant. This patch implements RFS for TCP and connected UDP sockets. It should be usable for other flow oriented protocols. There are two configuration parameters for RFS. The "rps_flow_entries" kernel init parameter sets the number of entries in the rps_sock_flow_table, the per rxqueue sysfs entry "rps_flow_cnt" contains the number of entries in the rps_dev_flow table for the rxqueue. Both are rounded to power of two. The obvious benefit of RFS (over just RPS) is that it achieves CPU locality between the receive processing for a flow and the applications processing; this can result in increased performance (higher pps, lower latency). The benefits of RFS are dependent on cache hierarchy, application load, and other factors. On simple benchmarks, we don't necessarily see improvement and sometimes see degradation. However, for more complex benchmarks and for applications where cache pressure is much higher this technique seems to perform very well. Below are some benchmark results which show the potential benfit of this patch. The netperf test has 500 instances of netperf TCP_RR test with 1 byte req. and resp. The RPC test is an request/response test similar in structure to netperf RR test ith 100 threads on each host, but does more work in userspace that netperf. e1000e on 8 core Intel No RFS or RPS 104K tps at 30% CPU No RFS (best RPS config): 290K tps at 63% CPU RFS 303K tps at 61% CPU RPC test tps CPU% 50/90/99% usec latency Latency StdDev No RFS/RPS 103K 48% 757/900/3185 4472.35 RPS only: 174K 73% 415/993/2468 491.66 RFS 223K 73% 379/651/1382 315.61 Signed-off-by: Tom Herbert <therbert@google.com> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-16 23:01:27 +00:00
.recvmsg = inet_recvmsg,
.mmap = sock_no_mmap,
.sendpage = inet_sendpage,
#ifdef CONFIG_COMPAT
.compat_setsockopt = compat_sock_common_setsockopt,
.compat_getsockopt = compat_sock_common_getsockopt,
.compat_ioctl = inet_compat_ioctl,
#endif
};
EXPORT_SYMBOL(inet_dgram_ops);
/*
* For SOCK_RAW sockets; should be the same as inet_dgram_ops but without
* udp_poll
*/
static const struct proto_ops inet_sockraw_ops = {
.family = PF_INET,
.owner = THIS_MODULE,
.release = inet_release,
.bind = inet_bind,
.connect = inet_dgram_connect,
.socketpair = sock_no_socketpair,
.accept = sock_no_accept,
.getname = inet_getname,
.poll = datagram_poll,
.ioctl = inet_ioctl,
.listen = sock_no_listen,
.shutdown = inet_shutdown,
.setsockopt = sock_common_setsockopt,
.getsockopt = sock_common_getsockopt,
.sendmsg = inet_sendmsg,
rfs: Receive Flow Steering This patch implements receive flow steering (RFS). RFS steers received packets for layer 3 and 4 processing to the CPU where the application for the corresponding flow is running. RFS is an extension of Receive Packet Steering (RPS). The basic idea of RFS is that when an application calls recvmsg (or sendmsg) the application's running CPU is stored in a hash table that is indexed by the connection's rxhash which is stored in the socket structure. The rxhash is passed in skb's received on the connection from netif_receive_skb. For each received packet, the associated rxhash is used to look up the CPU in the hash table, if a valid CPU is set then the packet is steered to that CPU using the RPS mechanisms. The convolution of the simple approach is that it would potentially allow OOO packets. If threads are thrashing around CPUs or multiple threads are trying to read from the same sockets, a quickly changing CPU value in the hash table could cause rampant OOO packets-- we consider this a non-starter. To avoid OOO packets, this solution implements two types of hash tables: rps_sock_flow_table and rps_dev_flow_table. rps_sock_table is a global hash table. Each entry is just a CPU number and it is populated in recvmsg and sendmsg as described above. This table contains the "desired" CPUs for flows. rps_dev_flow_table is specific to each device queue. Each entry contains a CPU and a tail queue counter. The CPU is the "current" CPU for a matching flow. The tail queue counter holds the value of a tail queue counter for the associated CPU's backlog queue at the time of last enqueue for a flow matching the entry. Each backlog queue has a queue head counter which is incremented on dequeue, and so a queue tail counter is computed as queue head count + queue length. When a packet is enqueued on a backlog queue, the current value of the queue tail counter is saved in the hash entry of the rps_dev_flow_table. And now the trick: when selecting the CPU for RPS (get_rps_cpu) the rps_sock_flow table and the rps_dev_flow table for the RX queue are consulted. When the desired CPU for the flow (found in the rps_sock_flow table) does not match the current CPU (found in the rps_dev_flow table), the current CPU is changed to the desired CPU if one of the following is true: - The current CPU is unset (equal to RPS_NO_CPU) - Current CPU is offline - The current CPU's queue head counter >= queue tail counter in the rps_dev_flow table. This checks if the queue tail has advanced beyond the last packet that was enqueued using this table entry. This guarantees that all packets queued using this entry have been dequeued, thus preserving in order delivery. Making each queue have its own rps_dev_flow table has two advantages: 1) the tail queue counters will be written on each receive, so keeping the table local to interrupting CPU s good for locality. 2) this allows lockless access to the table-- the CPU number and queue tail counter need to be accessed together under mutual exclusion from netif_receive_skb, we assume that this is only called from device napi_poll which is non-reentrant. This patch implements RFS for TCP and connected UDP sockets. It should be usable for other flow oriented protocols. There are two configuration parameters for RFS. The "rps_flow_entries" kernel init parameter sets the number of entries in the rps_sock_flow_table, the per rxqueue sysfs entry "rps_flow_cnt" contains the number of entries in the rps_dev_flow table for the rxqueue. Both are rounded to power of two. The obvious benefit of RFS (over just RPS) is that it achieves CPU locality between the receive processing for a flow and the applications processing; this can result in increased performance (higher pps, lower latency). The benefits of RFS are dependent on cache hierarchy, application load, and other factors. On simple benchmarks, we don't necessarily see improvement and sometimes see degradation. However, for more complex benchmarks and for applications where cache pressure is much higher this technique seems to perform very well. Below are some benchmark results which show the potential benfit of this patch. The netperf test has 500 instances of netperf TCP_RR test with 1 byte req. and resp. The RPC test is an request/response test similar in structure to netperf RR test ith 100 threads on each host, but does more work in userspace that netperf. e1000e on 8 core Intel No RFS or RPS 104K tps at 30% CPU No RFS (best RPS config): 290K tps at 63% CPU RFS 303K tps at 61% CPU RPC test tps CPU% 50/90/99% usec latency Latency StdDev No RFS/RPS 103K 48% 757/900/3185 4472.35 RPS only: 174K 73% 415/993/2468 491.66 RFS 223K 73% 379/651/1382 315.61 Signed-off-by: Tom Herbert <therbert@google.com> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-16 23:01:27 +00:00
.recvmsg = inet_recvmsg,
.mmap = sock_no_mmap,
.sendpage = inet_sendpage,
#ifdef CONFIG_COMPAT
.compat_setsockopt = compat_sock_common_setsockopt,
.compat_getsockopt = compat_sock_common_getsockopt,
.compat_ioctl = inet_compat_ioctl,
#endif
};
static const struct net_proto_family inet_family_ops = {
.family = PF_INET,
.create = inet_create,
.owner = THIS_MODULE,
};
/* Upon startup we insert all the elements in inetsw_array[] into
* the linked list inetsw.
*/
static struct inet_protosw inetsw_array[] =
{
{
.type = SOCK_STREAM,
.protocol = IPPROTO_TCP,
.prot = &tcp_prot,
.ops = &inet_stream_ops,
.no_check = 0,
.flags = INET_PROTOSW_PERMANENT |
INET_PROTOSW_ICSK,
},
{
.type = SOCK_DGRAM,
.protocol = IPPROTO_UDP,
.prot = &udp_prot,
.ops = &inet_dgram_ops,
.no_check = UDP_CSUM_DEFAULT,
.flags = INET_PROTOSW_PERMANENT,
},
{
.type = SOCK_RAW,
.protocol = IPPROTO_IP, /* wild card */
.prot = &raw_prot,
.ops = &inet_sockraw_ops,
.no_check = UDP_CSUM_DEFAULT,
.flags = INET_PROTOSW_REUSE,
}
};
#define INETSW_ARRAY_LEN ARRAY_SIZE(inetsw_array)
void inet_register_protosw(struct inet_protosw *p)
{
struct list_head *lh;
struct inet_protosw *answer;
int protocol = p->protocol;
struct list_head *last_perm;
spin_lock_bh(&inetsw_lock);
if (p->type >= SOCK_MAX)
goto out_illegal;
/* If we are trying to override a permanent protocol, bail. */
answer = NULL;
last_perm = &inetsw[p->type];
list_for_each(lh, &inetsw[p->type]) {
answer = list_entry(lh, struct inet_protosw, list);
/* Check only the non-wild match. */
if (INET_PROTOSW_PERMANENT & answer->flags) {
if (protocol == answer->protocol)
break;
last_perm = lh;
}
answer = NULL;
}
if (answer)
goto out_permanent;
/* Add the new entry after the last permanent entry if any, so that
* the new entry does not override a permanent entry when matched with
* a wild-card protocol. But it is allowed to override any existing
* non-permanent entry. This means that when we remove this entry, the
* system automatically returns to the old behavior.
*/
list_add_rcu(&p->list, last_perm);
out:
spin_unlock_bh(&inetsw_lock);
return;
out_permanent:
printk(KERN_ERR "Attempt to override permanent protocol %d.\n",
protocol);
goto out;
out_illegal:
printk(KERN_ERR
"Ignoring attempt to register invalid socket type %d.\n",
p->type);
goto out;
}
EXPORT_SYMBOL(inet_register_protosw);
void inet_unregister_protosw(struct inet_protosw *p)
{
if (INET_PROTOSW_PERMANENT & p->flags) {
printk(KERN_ERR
"Attempt to unregister permanent protocol %d.\n",
p->protocol);
} else {
spin_lock_bh(&inetsw_lock);
list_del_rcu(&p->list);
spin_unlock_bh(&inetsw_lock);
synchronize_net();
}
}
EXPORT_SYMBOL(inet_unregister_protosw);
/*
* Shall we try to damage output packets if routing dev changes?
*/
int sysctl_ip_dynaddr __read_mostly;
static int inet_sk_reselect_saddr(struct sock *sk)
{
struct inet_sock *inet = inet_sk(sk);
__be32 old_saddr = inet->inet_saddr;
__be32 daddr = inet->inet_daddr;
struct rtable *rt;
__be32 new_saddr;
if (inet->opt && inet->opt->srr)
daddr = inet->opt->faddr;
/* Query new route. */
rt = ip_route_connect(daddr, 0, RT_CONN_FLAGS(sk),
sk->sk_bound_dev_if, sk->sk_protocol,
inet->inet_sport, inet->inet_dport, sk, false);
if (IS_ERR(rt))
return PTR_ERR(rt);
sk_setup_caps(sk, &rt->dst);
new_saddr = rt->rt_src;
if (new_saddr == old_saddr)
return 0;
if (sysctl_ip_dynaddr > 1) {
printk(KERN_INFO "%s(): shifting inet->saddr from %pI4 to %pI4\n",
__func__, &old_saddr, &new_saddr);
}
inet->inet_saddr = inet->inet_rcv_saddr = new_saddr;
/*
* XXX The only one ugly spot where we need to
* XXX really change the sockets identity after
* XXX it has entered the hashes. -DaveM
*
* Besides that, it does not check for connection
* uniqueness. Wait for troubles.
*/
__sk_prot_rehash(sk);
return 0;
}
int inet_sk_rebuild_header(struct sock *sk)
{
struct inet_sock *inet = inet_sk(sk);
struct rtable *rt = (struct rtable *)__sk_dst_check(sk, 0);
__be32 daddr;
int err;
/* Route is OK, nothing to do. */
if (rt)
return 0;
/* Reroute. */
daddr = inet->inet_daddr;
if (inet->opt && inet->opt->srr)
daddr = inet->opt->faddr;
rt = ip_route_output_ports(sock_net(sk), sk, daddr, inet->inet_saddr,
inet->inet_dport, inet->inet_sport,
sk->sk_protocol, RT_CONN_FLAGS(sk),
sk->sk_bound_dev_if);
if (!IS_ERR(rt)) {
err = 0;
sk_setup_caps(sk, &rt->dst);
} else {
err = PTR_ERR(rt);
/* Routing failed... */
sk->sk_route_caps = 0;
/*
* Other protocols have to map its equivalent state to TCP_SYN_SENT.
* DCCP maps its DCCP_REQUESTING state to TCP_SYN_SENT. -acme
*/
if (!sysctl_ip_dynaddr ||
sk->sk_state != TCP_SYN_SENT ||
(sk->sk_userlocks & SOCK_BINDADDR_LOCK) ||
(err = inet_sk_reselect_saddr(sk)) != 0)
sk->sk_err_soft = -err;
}
return err;
}
EXPORT_SYMBOL(inet_sk_rebuild_header);
static int inet_gso_send_check(struct sk_buff *skb)
{
struct iphdr *iph;
const struct net_protocol *ops;
int proto;
int ihl;
int err = -EINVAL;
if (unlikely(!pskb_may_pull(skb, sizeof(*iph))))
goto out;
iph = ip_hdr(skb);
ihl = iph->ihl * 4;
if (ihl < sizeof(*iph))
goto out;
if (unlikely(!pskb_may_pull(skb, ihl)))
goto out;
__skb_pull(skb, ihl);
skb_reset_transport_header(skb);
iph = ip_hdr(skb);
proto = iph->protocol & (MAX_INET_PROTOS - 1);
err = -EPROTONOSUPPORT;
rcu_read_lock();
ops = rcu_dereference(inet_protos[proto]);
if (likely(ops && ops->gso_send_check))
err = ops->gso_send_check(skb);
rcu_read_unlock();
out:
return err;
}
static struct sk_buff *inet_gso_segment(struct sk_buff *skb, u32 features)
{
struct sk_buff *segs = ERR_PTR(-EINVAL);
struct iphdr *iph;
const struct net_protocol *ops;
int proto;
int ihl;
int id;
unsigned int offset = 0;
if (!(features & NETIF_F_V4_CSUM))
features &= ~NETIF_F_SG;
if (unlikely(skb_shinfo(skb)->gso_type &
~(SKB_GSO_TCPV4 |
SKB_GSO_UDP |
SKB_GSO_DODGY |
SKB_GSO_TCP_ECN |
0)))
goto out;
if (unlikely(!pskb_may_pull(skb, sizeof(*iph))))
goto out;
iph = ip_hdr(skb);
ihl = iph->ihl * 4;
if (ihl < sizeof(*iph))
goto out;
if (unlikely(!pskb_may_pull(skb, ihl)))
goto out;
__skb_pull(skb, ihl);
skb_reset_transport_header(skb);
iph = ip_hdr(skb);
id = ntohs(iph->id);
proto = iph->protocol & (MAX_INET_PROTOS - 1);
segs = ERR_PTR(-EPROTONOSUPPORT);
rcu_read_lock();
ops = rcu_dereference(inet_protos[proto]);
if (likely(ops && ops->gso_segment))
segs = ops->gso_segment(skb, features);
rcu_read_unlock();
if (!segs || IS_ERR(segs))
goto out;
skb = segs;
do {
iph = ip_hdr(skb);
if (proto == IPPROTO_UDP) {
iph->id = htons(id);
iph->frag_off = htons(offset >> 3);
if (skb->next != NULL)
iph->frag_off |= htons(IP_MF);
offset += (skb->len - skb->mac_len - iph->ihl * 4);
} else
iph->id = htons(id++);
iph->tot_len = htons(skb->len - skb->mac_len);
iph->check = 0;
iph->check = ip_fast_csum(skb_network_header(skb), iph->ihl);
} while ((skb = skb->next));
out:
return segs;
}
static struct sk_buff **inet_gro_receive(struct sk_buff **head,
struct sk_buff *skb)
{
const struct net_protocol *ops;
struct sk_buff **pp = NULL;
struct sk_buff *p;
struct iphdr *iph;
unsigned int hlen;
unsigned int off;
unsigned int id;
int flush = 1;
int proto;
off = skb_gro_offset(skb);
hlen = off + sizeof(*iph);
iph = skb_gro_header_fast(skb, off);
if (skb_gro_header_hard(skb, hlen)) {
iph = skb_gro_header_slow(skb, hlen, off);
if (unlikely(!iph))
goto out;
}
proto = iph->protocol & (MAX_INET_PROTOS - 1);
rcu_read_lock();
ops = rcu_dereference(inet_protos[proto]);
if (!ops || !ops->gro_receive)
goto out_unlock;
if (*(u8 *)iph != 0x45)
goto out_unlock;
if (unlikely(ip_fast_csum((u8 *)iph, iph->ihl)))
goto out_unlock;
id = ntohl(*(__be32 *)&iph->id);
flush = (u16)((ntohl(*(__be32 *)iph) ^ skb_gro_len(skb)) | (id ^ IP_DF));
id >>= 16;
for (p = *head; p; p = p->next) {
struct iphdr *iph2;
if (!NAPI_GRO_CB(p)->same_flow)
continue;
iph2 = ip_hdr(p);
if ((iph->protocol ^ iph2->protocol) |
(iph->tos ^ iph2->tos) |
((__force u32)iph->saddr ^ (__force u32)iph2->saddr) |
((__force u32)iph->daddr ^ (__force u32)iph2->daddr)) {
NAPI_GRO_CB(p)->same_flow = 0;
continue;
}
/* All fields must match except length and checksum. */
NAPI_GRO_CB(p)->flush |=
(iph->ttl ^ iph2->ttl) |
((u16)(ntohs(iph2->id) + NAPI_GRO_CB(p)->count) ^ id);
NAPI_GRO_CB(p)->flush |= flush;
}
NAPI_GRO_CB(skb)->flush |= flush;
skb_gro_pull(skb, sizeof(*iph));
skb_set_transport_header(skb, skb_gro_offset(skb));
pp = ops->gro_receive(head, skb);
out_unlock:
rcu_read_unlock();
out:
NAPI_GRO_CB(skb)->flush |= flush;
return pp;
}
static int inet_gro_complete(struct sk_buff *skb)
{
const struct net_protocol *ops;
struct iphdr *iph = ip_hdr(skb);
int proto = iph->protocol & (MAX_INET_PROTOS - 1);
int err = -ENOSYS;
__be16 newlen = htons(skb->len - skb_network_offset(skb));
csum_replace2(&iph->check, iph->tot_len, newlen);
iph->tot_len = newlen;
rcu_read_lock();
ops = rcu_dereference(inet_protos[proto]);
if (WARN_ON(!ops || !ops->gro_complete))
goto out_unlock;
err = ops->gro_complete(skb);
out_unlock:
rcu_read_unlock();
return err;
}
int inet_ctl_sock_create(struct sock **sk, unsigned short family,
unsigned short type, unsigned char protocol,
struct net *net)
{
struct socket *sock;
int rc = sock_create_kern(family, type, protocol, &sock);
if (rc == 0) {
*sk = sock->sk;
(*sk)->sk_allocation = GFP_ATOMIC;
/*
* Unhash it so that IP input processing does not even see it,
* we do not wish this socket to see incoming packets.
*/
(*sk)->sk_prot->unhash(*sk);
sk_change_net(*sk, net);
}
return rc;
}
EXPORT_SYMBOL_GPL(inet_ctl_sock_create);
unsigned long snmp_fold_field(void __percpu *mib[], int offt)
{
unsigned long res = 0;
int i;
for_each_possible_cpu(i) {
res += *(((unsigned long *) per_cpu_ptr(mib[0], i)) + offt);
res += *(((unsigned long *) per_cpu_ptr(mib[1], i)) + offt);
}
return res;
}
EXPORT_SYMBOL_GPL(snmp_fold_field);
#if BITS_PER_LONG==32
u64 snmp_fold_field64(void __percpu *mib[], int offt, size_t syncp_offset)
{
u64 res = 0;
int cpu;
for_each_possible_cpu(cpu) {
void *bhptr, *userptr;
struct u64_stats_sync *syncp;
u64 v_bh, v_user;
unsigned int start;
/* first mib used by softirq context, we must use _bh() accessors */
bhptr = per_cpu_ptr(SNMP_STAT_BHPTR(mib), cpu);
syncp = (struct u64_stats_sync *)(bhptr + syncp_offset);
do {
start = u64_stats_fetch_begin_bh(syncp);
v_bh = *(((u64 *) bhptr) + offt);
} while (u64_stats_fetch_retry_bh(syncp, start));
/* second mib used in USER context */
userptr = per_cpu_ptr(SNMP_STAT_USRPTR(mib), cpu);
syncp = (struct u64_stats_sync *)(userptr + syncp_offset);
do {
start = u64_stats_fetch_begin(syncp);
v_user = *(((u64 *) userptr) + offt);
} while (u64_stats_fetch_retry(syncp, start));
res += v_bh + v_user;
}
return res;
}
EXPORT_SYMBOL_GPL(snmp_fold_field64);
#endif
int snmp_mib_init(void __percpu *ptr[2], size_t mibsize, size_t align)
{
BUG_ON(ptr == NULL);
ptr[0] = __alloc_percpu(mibsize, align);
if (!ptr[0])
goto err0;
ptr[1] = __alloc_percpu(mibsize, align);
if (!ptr[1])
goto err1;
return 0;
err1:
free_percpu(ptr[0]);
ptr[0] = NULL;
err0:
return -ENOMEM;
}
EXPORT_SYMBOL_GPL(snmp_mib_init);
void snmp_mib_free(void __percpu *ptr[2])
{
BUG_ON(ptr == NULL);
free_percpu(ptr[0]);
free_percpu(ptr[1]);
ptr[0] = ptr[1] = NULL;
}
EXPORT_SYMBOL_GPL(snmp_mib_free);
#ifdef CONFIG_IP_MULTICAST
static const struct net_protocol igmp_protocol = {
.handler = igmp_rcv,
.netns_ok = 1,
};
#endif
static const struct net_protocol tcp_protocol = {
.handler = tcp_v4_rcv,
.err_handler = tcp_v4_err,
.gso_send_check = tcp_v4_gso_send_check,
.gso_segment = tcp_tso_segment,
.gro_receive = tcp4_gro_receive,
.gro_complete = tcp4_gro_complete,
.no_policy = 1,
.netns_ok = 1,
};
static const struct net_protocol udp_protocol = {
.handler = udp_rcv,
.err_handler = udp_err,
.gso_send_check = udp4_ufo_send_check,
.gso_segment = udp4_ufo_fragment,
.no_policy = 1,
.netns_ok = 1,
};
static const struct net_protocol icmp_protocol = {
.handler = icmp_rcv,
.no_policy = 1,
.netns_ok = 1,
};
static __net_init int ipv4_mib_init_net(struct net *net)
{
if (snmp_mib_init((void __percpu **)net->mib.tcp_statistics,
sizeof(struct tcp_mib),
__alignof__(struct tcp_mib)) < 0)
goto err_tcp_mib;
if (snmp_mib_init((void __percpu **)net->mib.ip_statistics,
sizeof(struct ipstats_mib),
__alignof__(struct ipstats_mib)) < 0)
goto err_ip_mib;
if (snmp_mib_init((void __percpu **)net->mib.net_statistics,
sizeof(struct linux_mib),
__alignof__(struct linux_mib)) < 0)
goto err_net_mib;
if (snmp_mib_init((void __percpu **)net->mib.udp_statistics,
sizeof(struct udp_mib),
__alignof__(struct udp_mib)) < 0)
goto err_udp_mib;
if (snmp_mib_init((void __percpu **)net->mib.udplite_statistics,
sizeof(struct udp_mib),
__alignof__(struct udp_mib)) < 0)
goto err_udplite_mib;
if (snmp_mib_init((void __percpu **)net->mib.icmp_statistics,
sizeof(struct icmp_mib),
__alignof__(struct icmp_mib)) < 0)
goto err_icmp_mib;
if (snmp_mib_init((void __percpu **)net->mib.icmpmsg_statistics,
sizeof(struct icmpmsg_mib),
__alignof__(struct icmpmsg_mib)) < 0)
goto err_icmpmsg_mib;
tcp_mib_init(net);
return 0;
err_icmpmsg_mib:
snmp_mib_free((void __percpu **)net->mib.icmp_statistics);
err_icmp_mib:
snmp_mib_free((void __percpu **)net->mib.udplite_statistics);
err_udplite_mib:
snmp_mib_free((void __percpu **)net->mib.udp_statistics);
err_udp_mib:
snmp_mib_free((void __percpu **)net->mib.net_statistics);
err_net_mib:
snmp_mib_free((void __percpu **)net->mib.ip_statistics);
err_ip_mib:
snmp_mib_free((void __percpu **)net->mib.tcp_statistics);
err_tcp_mib:
return -ENOMEM;
}
static __net_exit void ipv4_mib_exit_net(struct net *net)
{
snmp_mib_free((void __percpu **)net->mib.icmpmsg_statistics);
snmp_mib_free((void __percpu **)net->mib.icmp_statistics);
snmp_mib_free((void __percpu **)net->mib.udplite_statistics);
snmp_mib_free((void __percpu **)net->mib.udp_statistics);
snmp_mib_free((void __percpu **)net->mib.net_statistics);
snmp_mib_free((void __percpu **)net->mib.ip_statistics);
snmp_mib_free((void __percpu **)net->mib.tcp_statistics);
}
static __net_initdata struct pernet_operations ipv4_mib_ops = {
.init = ipv4_mib_init_net,
.exit = ipv4_mib_exit_net,
};
static int __init init_ipv4_mibs(void)
{
return register_pernet_subsys(&ipv4_mib_ops);
}
static int ipv4_proc_init(void);
/*
* IP protocol layer initialiser
*/
static struct packet_type ip_packet_type __read_mostly = {
.type = cpu_to_be16(ETH_P_IP),
.func = ip_rcv,
.gso_send_check = inet_gso_send_check,
.gso_segment = inet_gso_segment,
.gro_receive = inet_gro_receive,
.gro_complete = inet_gro_complete,
};
static int __init inet_init(void)
{
struct sk_buff *dummy_skb;
struct inet_protosw *q;
struct list_head *r;
int rc = -EINVAL;
BUILD_BUG_ON(sizeof(struct inet_skb_parm) > sizeof(dummy_skb->cb));
sysctl_local_reserved_ports = kzalloc(65536 / 8, GFP_KERNEL);
if (!sysctl_local_reserved_ports)
goto out;
rc = proto_register(&tcp_prot, 1);
if (rc)
goto out_free_reserved_ports;
rc = proto_register(&udp_prot, 1);
if (rc)
goto out_unregister_tcp_proto;
rc = proto_register(&raw_prot, 1);
if (rc)
goto out_unregister_udp_proto;
/*
* Tell SOCKET that we are alive...
*/
(void)sock_register(&inet_family_ops);
#ifdef CONFIG_SYSCTL
ip_static_sysctl_init();
#endif
/*
* Add all the base protocols.
*/
if (inet_add_protocol(&icmp_protocol, IPPROTO_ICMP) < 0)
printk(KERN_CRIT "inet_init: Cannot add ICMP protocol\n");
if (inet_add_protocol(&udp_protocol, IPPROTO_UDP) < 0)
printk(KERN_CRIT "inet_init: Cannot add UDP protocol\n");
if (inet_add_protocol(&tcp_protocol, IPPROTO_TCP) < 0)
printk(KERN_CRIT "inet_init: Cannot add TCP protocol\n");
#ifdef CONFIG_IP_MULTICAST
if (inet_add_protocol(&igmp_protocol, IPPROTO_IGMP) < 0)
printk(KERN_CRIT "inet_init: Cannot add IGMP protocol\n");
#endif
/* Register the socket-side information for inet_create. */
for (r = &inetsw[0]; r < &inetsw[SOCK_MAX]; ++r)
INIT_LIST_HEAD(r);
for (q = inetsw_array; q < &inetsw_array[INETSW_ARRAY_LEN]; ++q)
inet_register_protosw(q);
/*
* Set the ARP module up
*/
arp_init();
/*
* Set the IP module up
*/
ip_init();
tcp_v4_init();
/* Setup TCP slab cache for open requests. */
tcp_init();
/* Setup UDP memory threshold */
udp_init();
[NET]: Supporting UDP-Lite (RFC 3828) in Linux This is a revision of the previously submitted patch, which alters the way files are organized and compiled in the following manner: * UDP and UDP-Lite now use separate object files * source file dependencies resolved via header files net/ipv{4,6}/udp_impl.h * order of inclusion files in udp.c/udplite.c adapted accordingly [NET/IPv4]: Support for the UDP-Lite protocol (RFC 3828) This patch adds support for UDP-Lite to the IPv4 stack, provided as an extension to the existing UDPv4 code: * generic routines are all located in net/ipv4/udp.c * UDP-Lite specific routines are in net/ipv4/udplite.c * MIB/statistics support in /proc/net/snmp and /proc/net/udplite * shared API with extensions for partial checksum coverage [NET/IPv6]: Extension for UDP-Lite over IPv6 It extends the existing UDPv6 code base with support for UDP-Lite in the same manner as per UDPv4. In particular, * UDPv6 generic and shared code is in net/ipv6/udp.c * UDP-Litev6 specific extensions are in net/ipv6/udplite.c * MIB/statistics support in /proc/net/snmp6 and /proc/net/udplite6 * support for IPV6_ADDRFORM * aligned the coding style of protocol initialisation with af_inet6.c * made the error handling in udpv6_queue_rcv_skb consistent; to return `-1' on error on all error cases * consolidation of shared code [NET]: UDP-Lite Documentation and basic XFRM/Netfilter support The UDP-Lite patch further provides * API documentation for UDP-Lite * basic xfrm support * basic netfilter support for IPv4 and IPv6 (LOG target) Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-11-27 19:10:57 +00:00
/* Add UDP-Lite (RFC 3828) */
udplite4_register();
/*
* Set the ICMP layer up
*/
if (icmp_init() < 0)
panic("Failed to create the ICMP control socket.\n");
/*
* Initialise the multicast router
*/
#if defined(CONFIG_IP_MROUTE)
if (ip_mr_init())
printk(KERN_CRIT "inet_init: Cannot init ipv4 mroute\n");
#endif
/*
* Initialise per-cpu ipv4 mibs
*/
if (init_ipv4_mibs())
printk(KERN_CRIT "inet_init: Cannot init ipv4 mibs\n");
ipv4_proc_init();
ipfrag_init();
dev_add_pack(&ip_packet_type);
rc = 0;
out:
return rc;
out_unregister_udp_proto:
proto_unregister(&udp_prot);
out_unregister_tcp_proto:
proto_unregister(&tcp_prot);
out_free_reserved_ports:
kfree(sysctl_local_reserved_ports);
goto out;
}
fs_initcall(inet_init);
/* ------------------------------------------------------------------------ */
#ifdef CONFIG_PROC_FS
static int __init ipv4_proc_init(void)
{
int rc = 0;
if (raw_proc_init())
goto out_raw;
if (tcp4_proc_init())
goto out_tcp;
if (udp4_proc_init())
goto out_udp;
if (ip_misc_proc_init())
goto out_misc;
out:
return rc;
out_misc:
udp4_proc_exit();
out_udp:
tcp4_proc_exit();
out_tcp:
raw_proc_exit();
out_raw:
rc = -ENOMEM;
goto out;
}
#else /* CONFIG_PROC_FS */
static int __init ipv4_proc_init(void)
{
return 0;
}
#endif /* CONFIG_PROC_FS */
MODULE_ALIAS_NETPROTO(PF_INET);