linux/net/rds/tcp.c

570 lines
16 KiB
C
Raw Normal View History

/*
* 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 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 <linux/in.h>
#include <linux/module.h>
#include <net/tcp.h>
#include <net/net_namespace.h>
#include <net/netns/generic.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;
static int rds_tcp_skbuf_handler(struct ctl_table *ctl, int write,
void __user *buffer, size_t *lenp,
loff_t *fpos);
int rds_tcp_min_sndbuf = SOCK_MIN_SNDBUF;
int rds_tcp_min_rcvbuf = SOCK_MIN_RCVBUF;
static struct ctl_table rds_tcp_sysctl_table[] = {
#define RDS_TCP_SNDBUF 0
{
.procname = "rds_tcp_sndbuf",
/* data is per-net pointer */
.maxlen = sizeof(int),
.mode = 0644,
.proc_handler = rds_tcp_skbuf_handler,
.extra1 = &rds_tcp_min_sndbuf,
},
#define RDS_TCP_RCVBUF 1
{
.procname = "rds_tcp_rcvbuf",
/* data is per-net pointer */
.maxlen = sizeof(int),
.mode = 0644,
.proc_handler = rds_tcp_skbuf_handler,
.extra1 = &rds_tcp_min_rcvbuf,
},
{ }
};
/* 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);
}
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;
struct ctl_table_header *rds_tcp_sysctl;
struct ctl_table *ctl_table;
int sndbuf_size;
int rcvbuf_size;
};
/* All module specific customizations to the RDS-TCP socket should be done in
* rds_tcp_tune() and applied after socket creation.
*/
void rds_tcp_tune(struct socket *sock)
{
struct sock *sk = sock->sk;
struct net *net = sock_net(sk);
struct rds_tcp_net *rtn = net_generic(net, rds_tcp_netid);
rds_tcp_nonagle(sock);
lock_sock(sk);
if (rtn->sndbuf_size > 0) {
sk->sk_sndbuf = rtn->sndbuf_size;
sk->sk_userlocks |= SOCK_SNDBUF_LOCK;
}
if (rtn->rcvbuf_size > 0) {
sk->sk_sndbuf = rtn->rcvbuf_size;
sk->sk_userlocks |= SOCK_RCVBUF_LOCK;
}
release_sock(sk);
}
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);
struct ctl_table *tbl;
int err = 0;
memset(rtn, 0, sizeof(*rtn));
/* {snd, rcv}buf_size default to 0, which implies we let the
* stack pick the value, and permit auto-tuning of buffer size.
*/
if (net == &init_net) {
tbl = rds_tcp_sysctl_table;
} else {
tbl = kmemdup(rds_tcp_sysctl_table,
sizeof(rds_tcp_sysctl_table), GFP_KERNEL);
if (!tbl) {
pr_warn("could not set allocate syctl table\n");
return -ENOMEM;
}
rtn->ctl_table = tbl;
}
tbl[RDS_TCP_SNDBUF].data = &rtn->sndbuf_size;
tbl[RDS_TCP_RCVBUF].data = &rtn->rcvbuf_size;
rtn->rds_tcp_sysctl = register_net_sysctl(net, "net/rds/tcp", tbl);
if (!rtn->rds_tcp_sysctl) {
pr_warn("could not register sysctl\n");
err = -ENOMEM;
goto fail;
}
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");
unregister_net_sysctl_table(rtn->rds_tcp_sysctl);
rtn->rds_tcp_sysctl = NULL;
err = -EAFNOSUPPORT;
goto fail;
}
INIT_WORK(&rtn->rds_tcp_accept_w, rds_tcp_accept_worker);
return 0;
fail:
if (net != &init_net)
kfree(tbl);
return err;
}
static void __net_exit rds_tcp_exit_net(struct net *net)
{
struct rds_tcp_net *rtn = net_generic(net, rds_tcp_netid);
if (rtn->rds_tcp_sysctl)
unregister_net_sysctl_table(rtn->rds_tcp_sysctl);
if (net != &init_net && rtn->ctl_table)
kfree(rtn->ctl_table);
/* 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 */
};
/* when sysctl is used to modify some kernel socket parameters,this
* function resets the RDS connections in that netns so that we can
* restart with new parameters. The assumption is that such reset
* events are few and far-between.
*/
static void rds_tcp_sysctl_reset(struct net *net)
{
struct rds_tcp_connection *tc, *_tc;
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;
rds_conn_drop(tc->conn); /* reconnect with new parameters */
}
spin_unlock_irq(&rds_tcp_conn_lock);
}
static int rds_tcp_skbuf_handler(struct ctl_table *ctl, int write,
void __user *buffer, size_t *lenp,
loff_t *fpos)
{
struct net *net = current->nsproxy->net_ns;
int err;
err = proc_dointvec_minmax(ctl, write, buffer, lenp, fpos);
if (err < 0) {
pr_warn("Invalid input. Must be >= %d\n",
*(int *)(ctl->extra1));
return err;
}
if (write)
rds_tcp_sysctl_reset(net);
return 0;
}
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");