linux/net/rds/tcp.c
Sowmini Varadhan 1edd6a14d2 RDS-TCP: Do not bloat sndbuf/rcvbuf in rds_tcp_tune
Using the value of RDS_TCP_DEFAULT_BUFSIZE (128K)
clobbers efficient use of TSO because it inflates the size_goal
that is computed in tcp_sendmsg/tcp_sendpage and skews packet
latency, and the default values for these parameters actually
results in significantly better performance.

In request-response tests using rds-stress with a packet size of
100K with 16 threads (test parameters -q 100000 -a 256 -t16 -d16)
between a single pair of IP addresses achieves a throughput of
6-8 Gbps. Without this patch, throughput maxes at 2-3 Gbps under
equivalent conditions on these platforms.

Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-10-05 03:34:53 -07:00

447 lines
12 KiB
C

/*
* Copyright (c) 2006 Oracle. All rights reserved.
*
* This software is available to you under a choice of one of two
* licenses. You may choose to be licensed under the terms of the GNU
* General Public License (GPL) Version 2, available from the file
* COPYING in the main directory of this source tree, or the
* OpenIB.org BSD license below:
*
* Redistribution and use in source and binary forms, with or
* without modification, are permitted provided that the following
* conditions are met:
*
* - Redistributions of source code must retain the above
* copyright notice, this list of conditions and the following
* disclaimer.
*
* - Redistributions in binary form must reproduce the above
* copyright notice, this list of conditions and the following
* disclaimer in the documentation and/or other materials
* provided with the distribution.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
* NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
* BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
* ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
* CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
* SOFTWARE.
*
*/
#include <linux/kernel.h>
#include <linux/slab.h>
#include <linux/in.h>
#include <linux/module.h>
#include <net/tcp.h>
#include <net/net_namespace.h>
#include <net/netns/generic.h>
#include <net/tcp.h>
#include "rds.h"
#include "tcp.h"
/* only for info exporting */
static DEFINE_SPINLOCK(rds_tcp_tc_list_lock);
static LIST_HEAD(rds_tcp_tc_list);
static unsigned int rds_tcp_tc_count;
/* Track rds_tcp_connection structs so they can be cleaned up */
static DEFINE_SPINLOCK(rds_tcp_conn_lock);
static LIST_HEAD(rds_tcp_conn_list);
static struct kmem_cache *rds_tcp_conn_slab;
#define RDS_TCP_DEFAULT_BUFSIZE (128 * 1024)
/* doing it this way avoids calling tcp_sk() */
void rds_tcp_nonagle(struct socket *sock)
{
mm_segment_t oldfs = get_fs();
int val = 1;
set_fs(KERNEL_DS);
sock->ops->setsockopt(sock, SOL_TCP, TCP_NODELAY, (char __user *)&val,
sizeof(val));
set_fs(oldfs);
}
/* All module specific customizations to the RDS-TCP socket should be done in
* rds_tcp_tune() and applied after socket creation. In general these
* customizations should be tunable via module_param()
*/
void rds_tcp_tune(struct socket *sock)
{
rds_tcp_nonagle(sock);
}
u32 rds_tcp_snd_nxt(struct rds_tcp_connection *tc)
{
return tcp_sk(tc->t_sock->sk)->snd_nxt;
}
u32 rds_tcp_snd_una(struct rds_tcp_connection *tc)
{
return tcp_sk(tc->t_sock->sk)->snd_una;
}
void rds_tcp_restore_callbacks(struct socket *sock,
struct rds_tcp_connection *tc)
{
rdsdebug("restoring sock %p callbacks from tc %p\n", sock, tc);
write_lock_bh(&sock->sk->sk_callback_lock);
/* done under the callback_lock to serialize with write_space */
spin_lock(&rds_tcp_tc_list_lock);
list_del_init(&tc->t_list_item);
rds_tcp_tc_count--;
spin_unlock(&rds_tcp_tc_list_lock);
tc->t_sock = NULL;
sock->sk->sk_write_space = tc->t_orig_write_space;
sock->sk->sk_data_ready = tc->t_orig_data_ready;
sock->sk->sk_state_change = tc->t_orig_state_change;
sock->sk->sk_user_data = NULL;
write_unlock_bh(&sock->sk->sk_callback_lock);
}
/*
* This is the only path that sets tc->t_sock. Send and receive trust that
* it is set. The RDS_CONN_CONNECTED bit protects those paths from being
* called while it isn't set.
*/
void rds_tcp_set_callbacks(struct socket *sock, struct rds_connection *conn)
{
struct rds_tcp_connection *tc = conn->c_transport_data;
rdsdebug("setting sock %p callbacks to tc %p\n", sock, tc);
write_lock_bh(&sock->sk->sk_callback_lock);
/* done under the callback_lock to serialize with write_space */
spin_lock(&rds_tcp_tc_list_lock);
list_add_tail(&tc->t_list_item, &rds_tcp_tc_list);
rds_tcp_tc_count++;
spin_unlock(&rds_tcp_tc_list_lock);
/* accepted sockets need our listen data ready undone */
if (sock->sk->sk_data_ready == rds_tcp_listen_data_ready)
sock->sk->sk_data_ready = sock->sk->sk_user_data;
tc->t_sock = sock;
tc->conn = conn;
tc->t_orig_data_ready = sock->sk->sk_data_ready;
tc->t_orig_write_space = sock->sk->sk_write_space;
tc->t_orig_state_change = sock->sk->sk_state_change;
sock->sk->sk_user_data = conn;
sock->sk->sk_data_ready = rds_tcp_data_ready;
sock->sk->sk_write_space = rds_tcp_write_space;
sock->sk->sk_state_change = rds_tcp_state_change;
write_unlock_bh(&sock->sk->sk_callback_lock);
}
static void rds_tcp_tc_info(struct socket *sock, unsigned int len,
struct rds_info_iterator *iter,
struct rds_info_lengths *lens)
{
struct rds_info_tcp_socket tsinfo;
struct rds_tcp_connection *tc;
unsigned long flags;
struct sockaddr_in sin;
int sinlen;
spin_lock_irqsave(&rds_tcp_tc_list_lock, flags);
if (len / sizeof(tsinfo) < rds_tcp_tc_count)
goto out;
list_for_each_entry(tc, &rds_tcp_tc_list, t_list_item) {
sock->ops->getname(sock, (struct sockaddr *)&sin, &sinlen, 0);
tsinfo.local_addr = sin.sin_addr.s_addr;
tsinfo.local_port = sin.sin_port;
sock->ops->getname(sock, (struct sockaddr *)&sin, &sinlen, 1);
tsinfo.peer_addr = sin.sin_addr.s_addr;
tsinfo.peer_port = sin.sin_port;
tsinfo.hdr_rem = tc->t_tinc_hdr_rem;
tsinfo.data_rem = tc->t_tinc_data_rem;
tsinfo.last_sent_nxt = tc->t_last_sent_nxt;
tsinfo.last_expected_una = tc->t_last_expected_una;
tsinfo.last_seen_una = tc->t_last_seen_una;
rds_info_copy(iter, &tsinfo, sizeof(tsinfo));
}
out:
lens->nr = rds_tcp_tc_count;
lens->each = sizeof(tsinfo);
spin_unlock_irqrestore(&rds_tcp_tc_list_lock, flags);
}
static int rds_tcp_laddr_check(struct net *net, __be32 addr)
{
if (inet_addr_type(net, addr) == RTN_LOCAL)
return 0;
return -EADDRNOTAVAIL;
}
static int rds_tcp_conn_alloc(struct rds_connection *conn, gfp_t gfp)
{
struct rds_tcp_connection *tc;
tc = kmem_cache_alloc(rds_tcp_conn_slab, gfp);
if (!tc)
return -ENOMEM;
tc->t_sock = NULL;
tc->t_tinc = NULL;
tc->t_tinc_hdr_rem = sizeof(struct rds_header);
tc->t_tinc_data_rem = 0;
conn->c_transport_data = tc;
spin_lock_irq(&rds_tcp_conn_lock);
list_add_tail(&tc->t_tcp_node, &rds_tcp_conn_list);
spin_unlock_irq(&rds_tcp_conn_lock);
rdsdebug("alloced tc %p\n", conn->c_transport_data);
return 0;
}
static void rds_tcp_conn_free(void *arg)
{
struct rds_tcp_connection *tc = arg;
unsigned long flags;
rdsdebug("freeing tc %p\n", tc);
spin_lock_irqsave(&rds_tcp_conn_lock, flags);
list_del(&tc->t_tcp_node);
spin_unlock_irqrestore(&rds_tcp_conn_lock, flags);
kmem_cache_free(rds_tcp_conn_slab, tc);
}
static void rds_tcp_destroy_conns(void)
{
struct rds_tcp_connection *tc, *_tc;
LIST_HEAD(tmp_list);
/* avoid calling conn_destroy with irqs off */
spin_lock_irq(&rds_tcp_conn_lock);
list_splice(&rds_tcp_conn_list, &tmp_list);
INIT_LIST_HEAD(&rds_tcp_conn_list);
spin_unlock_irq(&rds_tcp_conn_lock);
list_for_each_entry_safe(tc, _tc, &tmp_list, t_tcp_node) {
if (tc->conn->c_passive)
rds_conn_destroy(tc->conn->c_passive);
rds_conn_destroy(tc->conn);
}
}
static void rds_tcp_exit(void);
struct rds_transport rds_tcp_transport = {
.laddr_check = rds_tcp_laddr_check,
.xmit_prepare = rds_tcp_xmit_prepare,
.xmit_complete = rds_tcp_xmit_complete,
.xmit = rds_tcp_xmit,
.recv = rds_tcp_recv,
.conn_alloc = rds_tcp_conn_alloc,
.conn_free = rds_tcp_conn_free,
.conn_connect = rds_tcp_conn_connect,
.conn_shutdown = rds_tcp_conn_shutdown,
.inc_copy_to_user = rds_tcp_inc_copy_to_user,
.inc_free = rds_tcp_inc_free,
.stats_info_copy = rds_tcp_stats_info_copy,
.exit = rds_tcp_exit,
.t_owner = THIS_MODULE,
.t_name = "tcp",
.t_type = RDS_TRANS_TCP,
.t_prefer_loopback = 1,
};
static int rds_tcp_netid;
/* per-network namespace private data for this module */
struct rds_tcp_net {
struct socket *rds_tcp_listen_sock;
struct work_struct rds_tcp_accept_w;
};
static void rds_tcp_accept_worker(struct work_struct *work)
{
struct rds_tcp_net *rtn = container_of(work,
struct rds_tcp_net,
rds_tcp_accept_w);
while (rds_tcp_accept_one(rtn->rds_tcp_listen_sock) == 0)
cond_resched();
}
void rds_tcp_accept_work(struct sock *sk)
{
struct net *net = sock_net(sk);
struct rds_tcp_net *rtn = net_generic(net, rds_tcp_netid);
queue_work(rds_wq, &rtn->rds_tcp_accept_w);
}
static __net_init int rds_tcp_init_net(struct net *net)
{
struct rds_tcp_net *rtn = net_generic(net, rds_tcp_netid);
rtn->rds_tcp_listen_sock = rds_tcp_listen_init(net);
if (!rtn->rds_tcp_listen_sock) {
pr_warn("could not set up listen sock\n");
return -EAFNOSUPPORT;
}
INIT_WORK(&rtn->rds_tcp_accept_w, rds_tcp_accept_worker);
return 0;
}
static void __net_exit rds_tcp_exit_net(struct net *net)
{
struct rds_tcp_net *rtn = net_generic(net, rds_tcp_netid);
/* If rds_tcp_exit_net() is called as a result of netns deletion,
* the rds_tcp_kill_sock() device notifier would already have cleaned
* up the listen socket, thus there is no work to do in this function.
*
* If rds_tcp_exit_net() is called as a result of module unload,
* i.e., due to rds_tcp_exit() -> unregister_pernet_subsys(), then
* we do need to clean up the listen socket here.
*/
if (rtn->rds_tcp_listen_sock) {
rds_tcp_listen_stop(rtn->rds_tcp_listen_sock);
rtn->rds_tcp_listen_sock = NULL;
flush_work(&rtn->rds_tcp_accept_w);
}
}
static struct pernet_operations rds_tcp_net_ops = {
.init = rds_tcp_init_net,
.exit = rds_tcp_exit_net,
.id = &rds_tcp_netid,
.size = sizeof(struct rds_tcp_net),
};
static void rds_tcp_kill_sock(struct net *net)
{
struct rds_tcp_connection *tc, *_tc;
struct sock *sk;
LIST_HEAD(tmp_list);
struct rds_tcp_net *rtn = net_generic(net, rds_tcp_netid);
rds_tcp_listen_stop(rtn->rds_tcp_listen_sock);
rtn->rds_tcp_listen_sock = NULL;
flush_work(&rtn->rds_tcp_accept_w);
spin_lock_irq(&rds_tcp_conn_lock);
list_for_each_entry_safe(tc, _tc, &rds_tcp_conn_list, t_tcp_node) {
struct net *c_net = read_pnet(&tc->conn->c_net);
if (net != c_net || !tc->t_sock)
continue;
list_move_tail(&tc->t_tcp_node, &tmp_list);
}
spin_unlock_irq(&rds_tcp_conn_lock);
list_for_each_entry_safe(tc, _tc, &tmp_list, t_tcp_node) {
sk = tc->t_sock->sk;
sk->sk_prot->disconnect(sk, 0);
tcp_done(sk);
if (tc->conn->c_passive)
rds_conn_destroy(tc->conn->c_passive);
rds_conn_destroy(tc->conn);
}
}
static int rds_tcp_dev_event(struct notifier_block *this,
unsigned long event, void *ptr)
{
struct net_device *dev = netdev_notifier_info_to_dev(ptr);
/* rds-tcp registers as a pernet subys, so the ->exit will only
* get invoked after network acitivity has quiesced. We need to
* clean up all sockets to quiesce network activity, and use
* the unregistration of the per-net loopback device as a trigger
* to start that cleanup.
*/
if (event == NETDEV_UNREGISTER_FINAL &&
dev->ifindex == LOOPBACK_IFINDEX)
rds_tcp_kill_sock(dev_net(dev));
return NOTIFY_DONE;
}
static struct notifier_block rds_tcp_dev_notifier = {
.notifier_call = rds_tcp_dev_event,
.priority = -10, /* must be called after other network notifiers */
};
static void rds_tcp_exit(void)
{
rds_info_deregister_func(RDS_INFO_TCP_SOCKETS, rds_tcp_tc_info);
unregister_pernet_subsys(&rds_tcp_net_ops);
if (unregister_netdevice_notifier(&rds_tcp_dev_notifier))
pr_warn("could not unregister rds_tcp_dev_notifier\n");
rds_tcp_destroy_conns();
rds_trans_unregister(&rds_tcp_transport);
rds_tcp_recv_exit();
kmem_cache_destroy(rds_tcp_conn_slab);
}
module_exit(rds_tcp_exit);
static int rds_tcp_init(void)
{
int ret;
rds_tcp_conn_slab = kmem_cache_create("rds_tcp_connection",
sizeof(struct rds_tcp_connection),
0, 0, NULL);
if (!rds_tcp_conn_slab) {
ret = -ENOMEM;
goto out;
}
ret = register_netdevice_notifier(&rds_tcp_dev_notifier);
if (ret) {
pr_warn("could not register rds_tcp_dev_notifier\n");
goto out;
}
ret = register_pernet_subsys(&rds_tcp_net_ops);
if (ret)
goto out_slab;
ret = rds_tcp_recv_init();
if (ret)
goto out_slab;
ret = rds_trans_register(&rds_tcp_transport);
if (ret)
goto out_recv;
rds_info_register_func(RDS_INFO_TCP_SOCKETS, rds_tcp_tc_info);
goto out;
out_recv:
rds_tcp_recv_exit();
out_slab:
unregister_pernet_subsys(&rds_tcp_net_ops);
kmem_cache_destroy(rds_tcp_conn_slab);
out:
return ret;
}
module_init(rds_tcp_init);
MODULE_AUTHOR("Oracle Corporation <rds-devel@oss.oracle.com>");
MODULE_DESCRIPTION("RDS: TCP transport");
MODULE_LICENSE("Dual BSD/GPL");