linux/net/rxrpc/transport.c
David Howells 4a3388c803 rxrpc: Use IDR to allocate client conn IDs on a machine-wide basis
Use the IDR facility to allocate client connection IDs on a machine-wide
basis so that each client connection has a unique identifier.  When the
connection ID space wraps, we advance the epoch by 1, thereby effectively
having a 62-bit ID space.  The IDR facility is then used to look up client
connections during incoming packet routing instead of using an rbtree
rooted on the transport.

This change allows for the removal of the transport in the future and also
means that client connections can be looked up directly in the data-ready
handler by connection ID.

The ID management code is placed in a new file, conn-client.c, to which all
the client connection-specific code will eventually move.

Note that the IDR tree gets very expensive on memory if the connection IDs
are widely scattered throughout the number space, so we shall need to
retire connections that have, say, an ID more than four times the maximum
number of client conns away from the current allocation point to try and
keep the IDs concentrated.  We will also need to retire connections from an
old epoch.

Also note that, for the moment, a pointer to the transport has to be passed
through into the ID allocation function so that we can take a BH lock to
prevent a locking issue against in-BH lookup of client connections.  This
will go away later when RCU is used for server connections also.

Signed-off-by: David Howells <dhowells@redhat.com>
2016-06-22 09:10:02 +01:00

268 lines
6.8 KiB
C

/* RxRPC point-to-point transport session management
*
* Copyright (C) 2007 Red Hat, Inc. All Rights Reserved.
* Written by David Howells (dhowells@redhat.com)
*
* 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.
*/
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
#include <linux/module.h>
#include <linux/net.h>
#include <linux/skbuff.h>
#include <linux/slab.h>
#include <net/sock.h>
#include <net/af_rxrpc.h>
#include "ar-internal.h"
/*
* Time after last use at which transport record is cleaned up.
*/
unsigned int rxrpc_transport_expiry = 3600 * 24;
static void rxrpc_transport_reaper(struct work_struct *work);
static LIST_HEAD(rxrpc_transports);
static DEFINE_RWLOCK(rxrpc_transport_lock);
static DECLARE_DELAYED_WORK(rxrpc_transport_reap, rxrpc_transport_reaper);
/*
* allocate a new transport session manager
*/
static struct rxrpc_transport *rxrpc_alloc_transport(struct rxrpc_local *local,
struct rxrpc_peer *peer,
gfp_t gfp)
{
struct rxrpc_transport *trans;
_enter("");
trans = kzalloc(sizeof(struct rxrpc_transport), gfp);
if (trans) {
trans->local = local;
trans->peer = peer;
INIT_LIST_HEAD(&trans->link);
trans->bundles = RB_ROOT;
trans->server_conns = RB_ROOT;
spin_lock_init(&trans->client_lock);
rwlock_init(&trans->conn_lock);
atomic_set(&trans->usage, 1);
trans->debug_id = atomic_inc_return(&rxrpc_debug_id);
}
_leave(" = %p", trans);
return trans;
}
/*
* obtain a transport session for the nominated endpoints
*/
struct rxrpc_transport *rxrpc_get_transport(struct rxrpc_local *local,
struct rxrpc_peer *peer,
gfp_t gfp)
{
struct rxrpc_transport *trans, *candidate;
const char *new = "old";
int usage;
_enter("{%pI4+%hu},{%pI4+%hu},",
&local->srx.transport.sin.sin_addr,
ntohs(local->srx.transport.sin.sin_port),
&peer->srx.transport.sin.sin_addr,
ntohs(peer->srx.transport.sin.sin_port));
/* search the transport list first */
read_lock_bh(&rxrpc_transport_lock);
list_for_each_entry(trans, &rxrpc_transports, link) {
if (trans->local == local && trans->peer == peer)
goto found_extant_transport;
}
read_unlock_bh(&rxrpc_transport_lock);
/* not yet present - create a candidate for a new record and then
* redo the search */
candidate = rxrpc_alloc_transport(local, peer, gfp);
if (!candidate) {
_leave(" = -ENOMEM");
return ERR_PTR(-ENOMEM);
}
write_lock_bh(&rxrpc_transport_lock);
list_for_each_entry(trans, &rxrpc_transports, link) {
if (trans->local == local && trans->peer == peer)
goto found_extant_second;
}
/* we can now add the new candidate to the list */
trans = candidate;
candidate = NULL;
usage = atomic_read(&trans->usage);
rxrpc_get_local(trans->local);
rxrpc_get_peer(trans->peer);
list_add_tail(&trans->link, &rxrpc_transports);
write_unlock_bh(&rxrpc_transport_lock);
new = "new";
success:
_net("TRANSPORT %s %d local %d -> peer %d",
new,
trans->debug_id,
trans->local->debug_id,
trans->peer->debug_id);
_leave(" = %p {u=%d}", trans, usage);
return trans;
/* we found the transport in the list immediately */
found_extant_transport:
usage = atomic_inc_return(&trans->usage);
read_unlock_bh(&rxrpc_transport_lock);
goto success;
/* we found the transport on the second time through the list */
found_extant_second:
usage = atomic_inc_return(&trans->usage);
write_unlock_bh(&rxrpc_transport_lock);
kfree(candidate);
goto success;
}
/*
* find the transport connecting two endpoints
*/
struct rxrpc_transport *rxrpc_find_transport(struct rxrpc_local *local,
struct rxrpc_peer *peer)
{
struct rxrpc_transport *trans;
_enter("{%pI4+%hu},{%pI4+%hu},",
&local->srx.transport.sin.sin_addr,
ntohs(local->srx.transport.sin.sin_port),
&peer->srx.transport.sin.sin_addr,
ntohs(peer->srx.transport.sin.sin_port));
/* search the transport list */
read_lock_bh(&rxrpc_transport_lock);
list_for_each_entry(trans, &rxrpc_transports, link) {
if (trans->local == local && trans->peer == peer)
goto found_extant_transport;
}
read_unlock_bh(&rxrpc_transport_lock);
_leave(" = NULL");
return NULL;
found_extant_transport:
atomic_inc(&trans->usage);
read_unlock_bh(&rxrpc_transport_lock);
_leave(" = %p", trans);
return trans;
}
/*
* release a transport session
*/
void rxrpc_put_transport(struct rxrpc_transport *trans)
{
_enter("%p{u=%d}", trans, atomic_read(&trans->usage));
ASSERTCMP(atomic_read(&trans->usage), >, 0);
trans->put_time = ktime_get_seconds();
if (unlikely(atomic_dec_and_test(&trans->usage))) {
_debug("zombie");
/* let the reaper determine the timeout to avoid a race with
* overextending the timeout if the reaper is running at the
* same time */
rxrpc_queue_delayed_work(&rxrpc_transport_reap, 0);
}
_leave("");
}
/*
* clean up a transport session
*/
static void rxrpc_cleanup_transport(struct rxrpc_transport *trans)
{
_net("DESTROY TRANS %d", trans->debug_id);
rxrpc_put_local(trans->local);
rxrpc_put_peer(trans->peer);
kfree(trans);
}
/*
* reap dead transports that have passed their expiry date
*/
static void rxrpc_transport_reaper(struct work_struct *work)
{
struct rxrpc_transport *trans, *_p;
unsigned long now, earliest, reap_time;
LIST_HEAD(graveyard);
_enter("");
now = ktime_get_seconds();
earliest = ULONG_MAX;
/* extract all the transports that have been dead too long */
write_lock_bh(&rxrpc_transport_lock);
list_for_each_entry_safe(trans, _p, &rxrpc_transports, link) {
_debug("reap TRANS %d { u=%d t=%ld }",
trans->debug_id, atomic_read(&trans->usage),
(long) now - (long) trans->put_time);
if (likely(atomic_read(&trans->usage) > 0))
continue;
reap_time = trans->put_time + rxrpc_transport_expiry;
if (reap_time <= now)
list_move_tail(&trans->link, &graveyard);
else if (reap_time < earliest)
earliest = reap_time;
}
write_unlock_bh(&rxrpc_transport_lock);
if (earliest != ULONG_MAX) {
_debug("reschedule reaper %ld", (long) earliest - now);
ASSERTCMP(earliest, >, now);
rxrpc_queue_delayed_work(&rxrpc_transport_reap,
(earliest - now) * HZ);
}
/* then destroy all those pulled out */
while (!list_empty(&graveyard)) {
trans = list_entry(graveyard.next, struct rxrpc_transport,
link);
list_del_init(&trans->link);
ASSERTCMP(atomic_read(&trans->usage), ==, 0);
rxrpc_cleanup_transport(trans);
}
_leave("");
}
/*
* preemptively destroy all the transport session records rather than waiting
* for them to time out
*/
void __exit rxrpc_destroy_all_transports(void)
{
_enter("");
rxrpc_transport_expiry = 0;
cancel_delayed_work(&rxrpc_transport_reap);
rxrpc_queue_delayed_work(&rxrpc_transport_reap, 0);
_leave("");
}