2013-05-13 13:29:47 +02:00
|
|
|
/*
|
|
|
|
* urcu-mb.c
|
|
|
|
*
|
|
|
|
* Userspace RCU library with explicit memory barriers
|
|
|
|
*
|
|
|
|
* Copyright (c) 2009 Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
|
|
|
|
* Copyright (c) 2009 Paul E. McKenney, IBM Corporation.
|
|
|
|
* Copyright 2015 Red Hat, Inc.
|
|
|
|
*
|
|
|
|
* Ported to QEMU by Paolo Bonzini <pbonzini@redhat.com>
|
|
|
|
*
|
|
|
|
* This library is free software; you can redistribute it and/or
|
|
|
|
* modify it under the terms of the GNU Lesser General Public
|
|
|
|
* License as published by the Free Software Foundation; either
|
|
|
|
* version 2.1 of the License, or (at your option) any later version.
|
|
|
|
*
|
|
|
|
* This library is distributed in the hope that it will be useful,
|
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|
|
|
* Lesser General Public License for more details.
|
|
|
|
*
|
|
|
|
* You should have received a copy of the GNU Lesser General Public
|
|
|
|
* License along with this library; if not, write to the Free Software
|
|
|
|
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
|
|
|
|
*
|
|
|
|
* IBM's contributions to this file may be relicensed under LGPLv2 or later.
|
|
|
|
*/
|
|
|
|
|
2016-01-29 18:49:55 +01:00
|
|
|
#include "qemu/osdep.h"
|
2013-05-13 13:29:47 +02:00
|
|
|
#include "qemu/rcu.h"
|
|
|
|
#include "qemu/atomic.h"
|
2013-05-13 17:49:24 +02:00
|
|
|
#include "qemu/thread.h"
|
2015-02-11 17:15:18 +01:00
|
|
|
#include "qemu/main-loop.h"
|
2017-12-20 14:16:46 +01:00
|
|
|
#if defined(CONFIG_MALLOC_TRIM)
|
|
|
|
#include <malloc.h>
|
|
|
|
#endif
|
2013-05-13 13:29:47 +02:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Global grace period counter. Bit 0 is always one in rcu_gp_ctr.
|
|
|
|
* Bits 1 and above are defined in synchronize_rcu.
|
|
|
|
*/
|
|
|
|
#define RCU_GP_LOCKED (1UL << 0)
|
|
|
|
#define RCU_GP_CTR (1UL << 1)
|
|
|
|
|
|
|
|
unsigned long rcu_gp_ctr = RCU_GP_LOCKED;
|
|
|
|
|
|
|
|
QemuEvent rcu_gp_event;
|
2015-07-27 04:24:18 +02:00
|
|
|
static QemuMutex rcu_registry_lock;
|
|
|
|
static QemuMutex rcu_sync_lock;
|
2013-05-13 13:29:47 +02:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Check whether a quiescent state was crossed between the beginning of
|
|
|
|
* update_counter_and_wait and now.
|
|
|
|
*/
|
|
|
|
static inline int rcu_gp_ongoing(unsigned long *ctr)
|
|
|
|
{
|
|
|
|
unsigned long v;
|
|
|
|
|
|
|
|
v = atomic_read(ctr);
|
|
|
|
return v && (v != rcu_gp_ctr);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Written to only by each individual reader. Read by both the reader and the
|
|
|
|
* writers.
|
|
|
|
*/
|
|
|
|
__thread struct rcu_reader_data rcu_reader;
|
|
|
|
|
2015-07-27 04:24:18 +02:00
|
|
|
/* Protected by rcu_registry_lock. */
|
2013-05-13 13:29:47 +02:00
|
|
|
typedef QLIST_HEAD(, rcu_reader_data) ThreadList;
|
|
|
|
static ThreadList registry = QLIST_HEAD_INITIALIZER(registry);
|
|
|
|
|
|
|
|
/* Wait for previous parity/grace period to be empty of readers. */
|
|
|
|
static void wait_for_readers(void)
|
|
|
|
{
|
|
|
|
ThreadList qsreaders = QLIST_HEAD_INITIALIZER(qsreaders);
|
|
|
|
struct rcu_reader_data *index, *tmp;
|
|
|
|
|
|
|
|
for (;;) {
|
|
|
|
/* We want to be notified of changes made to rcu_gp_ongoing
|
|
|
|
* while we walk the list.
|
|
|
|
*/
|
|
|
|
qemu_event_reset(&rcu_gp_event);
|
|
|
|
|
|
|
|
/* Instead of using atomic_mb_set for index->waiting, and
|
|
|
|
* atomic_mb_read for index->ctr, memory barriers are placed
|
|
|
|
* manually since writes to different threads are independent.
|
2016-09-19 11:27:46 +02:00
|
|
|
* qemu_event_reset has acquire semantics, so no memory barrier
|
|
|
|
* is needed here.
|
2013-05-13 13:29:47 +02:00
|
|
|
*/
|
|
|
|
QLIST_FOREACH(index, ®istry, node) {
|
|
|
|
atomic_set(&index->waiting, true);
|
|
|
|
}
|
|
|
|
|
2018-02-16 09:23:31 +01:00
|
|
|
/* Here, order the stores to index->waiting before the loads of
|
2018-02-16 10:04:18 +01:00
|
|
|
* index->ctr. Pairs with smp_mb_placeholder() in rcu_read_unlock(),
|
2018-02-16 09:23:31 +01:00
|
|
|
* ensuring that the loads of index->ctr are sequentially consistent.
|
2016-09-19 11:27:46 +02:00
|
|
|
*/
|
2018-02-16 10:04:18 +01:00
|
|
|
smp_mb_global();
|
2013-05-13 13:29:47 +02:00
|
|
|
|
|
|
|
QLIST_FOREACH_SAFE(index, ®istry, node, tmp) {
|
|
|
|
if (!rcu_gp_ongoing(&index->ctr)) {
|
|
|
|
QLIST_REMOVE(index, node);
|
|
|
|
QLIST_INSERT_HEAD(&qsreaders, index, node);
|
|
|
|
|
|
|
|
/* No need for mb_set here, worst of all we
|
|
|
|
* get some extra futex wakeups.
|
|
|
|
*/
|
|
|
|
atomic_set(&index->waiting, false);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (QLIST_EMPTY(®istry)) {
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2015-07-27 04:24:18 +02:00
|
|
|
/* Wait for one thread to report a quiescent state and try again.
|
|
|
|
* Release rcu_registry_lock, so rcu_(un)register_thread() doesn't
|
|
|
|
* wait too much time.
|
|
|
|
*
|
|
|
|
* rcu_register_thread() may add nodes to ®istry; it will not
|
|
|
|
* wake up synchronize_rcu, but that is okay because at least another
|
|
|
|
* thread must exit its RCU read-side critical section before
|
|
|
|
* synchronize_rcu is done. The next iteration of the loop will
|
|
|
|
* move the new thread's rcu_reader from ®istry to &qsreaders,
|
|
|
|
* because rcu_gp_ongoing() will return false.
|
|
|
|
*
|
|
|
|
* rcu_unregister_thread() may remove nodes from &qsreaders instead
|
|
|
|
* of ®istry if it runs during qemu_event_wait. That's okay;
|
|
|
|
* the node then will not be added back to ®istry by QLIST_SWAP
|
|
|
|
* below. The invariant is that the node is part of one list when
|
|
|
|
* rcu_registry_lock is released.
|
2013-05-13 13:29:47 +02:00
|
|
|
*/
|
2015-07-27 04:24:18 +02:00
|
|
|
qemu_mutex_unlock(&rcu_registry_lock);
|
2013-05-13 13:29:47 +02:00
|
|
|
qemu_event_wait(&rcu_gp_event);
|
2015-07-27 04:24:18 +02:00
|
|
|
qemu_mutex_lock(&rcu_registry_lock);
|
2013-05-13 13:29:47 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/* put back the reader list in the registry */
|
|
|
|
QLIST_SWAP(®istry, &qsreaders, node);
|
|
|
|
}
|
|
|
|
|
|
|
|
void synchronize_rcu(void)
|
|
|
|
{
|
2015-07-27 04:24:18 +02:00
|
|
|
qemu_mutex_lock(&rcu_sync_lock);
|
2013-05-13 13:29:47 +02:00
|
|
|
|
2018-02-16 09:23:31 +01:00
|
|
|
/* Write RCU-protected pointers before reading p_rcu_reader->ctr.
|
2018-02-16 10:04:18 +01:00
|
|
|
* Pairs with smp_mb_placeholder() in rcu_read_lock().
|
2018-02-16 09:23:31 +01:00
|
|
|
*/
|
2018-02-16 10:04:18 +01:00
|
|
|
smp_mb_global();
|
2018-02-16 09:23:31 +01:00
|
|
|
|
|
|
|
qemu_mutex_lock(&rcu_registry_lock);
|
2013-05-13 13:29:47 +02:00
|
|
|
if (!QLIST_EMPTY(®istry)) {
|
|
|
|
/* In either case, the atomic_mb_set below blocks stores that free
|
|
|
|
* old RCU-protected pointers.
|
|
|
|
*/
|
|
|
|
if (sizeof(rcu_gp_ctr) < 8) {
|
|
|
|
/* For architectures with 32-bit longs, a two-subphases algorithm
|
|
|
|
* ensures we do not encounter overflow bugs.
|
|
|
|
*
|
|
|
|
* Switch parity: 0 -> 1, 1 -> 0.
|
|
|
|
*/
|
|
|
|
atomic_mb_set(&rcu_gp_ctr, rcu_gp_ctr ^ RCU_GP_CTR);
|
|
|
|
wait_for_readers();
|
|
|
|
atomic_mb_set(&rcu_gp_ctr, rcu_gp_ctr ^ RCU_GP_CTR);
|
|
|
|
} else {
|
|
|
|
/* Increment current grace period. */
|
|
|
|
atomic_mb_set(&rcu_gp_ctr, rcu_gp_ctr + RCU_GP_CTR);
|
|
|
|
}
|
|
|
|
|
|
|
|
wait_for_readers();
|
|
|
|
}
|
|
|
|
|
2015-07-27 04:24:18 +02:00
|
|
|
qemu_mutex_unlock(&rcu_registry_lock);
|
|
|
|
qemu_mutex_unlock(&rcu_sync_lock);
|
2013-05-13 13:29:47 +02:00
|
|
|
}
|
|
|
|
|
2013-05-13 17:49:24 +02:00
|
|
|
|
|
|
|
#define RCU_CALL_MIN_SIZE 30
|
|
|
|
|
|
|
|
/* Multi-producer, single-consumer queue based on urcu/static/wfqueue.h
|
|
|
|
* from liburcu. Note that head is only used by the consumer.
|
|
|
|
*/
|
|
|
|
static struct rcu_head dummy;
|
|
|
|
static struct rcu_head *head = &dummy, **tail = &dummy.next;
|
|
|
|
static int rcu_call_count;
|
|
|
|
static QemuEvent rcu_call_ready_event;
|
|
|
|
|
|
|
|
static void enqueue(struct rcu_head *node)
|
|
|
|
{
|
|
|
|
struct rcu_head **old_tail;
|
|
|
|
|
|
|
|
node->next = NULL;
|
|
|
|
old_tail = atomic_xchg(&tail, &node->next);
|
|
|
|
atomic_mb_set(old_tail, node);
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct rcu_head *try_dequeue(void)
|
|
|
|
{
|
|
|
|
struct rcu_head *node, *next;
|
|
|
|
|
|
|
|
retry:
|
|
|
|
/* Test for an empty list, which we do not expect. Note that for
|
|
|
|
* the consumer head and tail are always consistent. The head
|
|
|
|
* is consistent because only the consumer reads/writes it.
|
|
|
|
* The tail, because it is the first step in the enqueuing.
|
|
|
|
* It is only the next pointers that might be inconsistent.
|
|
|
|
*/
|
|
|
|
if (head == &dummy && atomic_mb_read(&tail) == &dummy.next) {
|
|
|
|
abort();
|
|
|
|
}
|
|
|
|
|
|
|
|
/* If the head node has NULL in its next pointer, the value is
|
|
|
|
* wrong and we need to wait until its enqueuer finishes the update.
|
|
|
|
*/
|
|
|
|
node = head;
|
|
|
|
next = atomic_mb_read(&head->next);
|
|
|
|
if (!next) {
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Since we are the sole consumer, and we excluded the empty case
|
|
|
|
* above, the queue will always have at least two nodes: the
|
|
|
|
* dummy node, and the one being removed. So we do not need to update
|
|
|
|
* the tail pointer.
|
|
|
|
*/
|
|
|
|
head = next;
|
|
|
|
|
|
|
|
/* If we dequeued the dummy node, add it back at the end and retry. */
|
|
|
|
if (node == &dummy) {
|
|
|
|
enqueue(node);
|
|
|
|
goto retry;
|
|
|
|
}
|
|
|
|
|
|
|
|
return node;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void *call_rcu_thread(void *opaque)
|
|
|
|
{
|
|
|
|
struct rcu_head *node;
|
|
|
|
|
2015-07-09 08:55:38 +02:00
|
|
|
rcu_register_thread();
|
|
|
|
|
2013-05-13 17:49:24 +02:00
|
|
|
for (;;) {
|
|
|
|
int tries = 0;
|
|
|
|
int n = atomic_read(&rcu_call_count);
|
|
|
|
|
|
|
|
/* Heuristically wait for a decent number of callbacks to pile up.
|
|
|
|
* Fetch rcu_call_count now, we only must process elements that were
|
|
|
|
* added before synchronize_rcu() starts.
|
|
|
|
*/
|
2015-02-11 15:51:54 +01:00
|
|
|
while (n == 0 || (n < RCU_CALL_MIN_SIZE && ++tries <= 5)) {
|
|
|
|
g_usleep(10000);
|
|
|
|
if (n == 0) {
|
|
|
|
qemu_event_reset(&rcu_call_ready_event);
|
2013-05-13 17:49:24 +02:00
|
|
|
n = atomic_read(&rcu_call_count);
|
2015-02-11 15:51:54 +01:00
|
|
|
if (n == 0) {
|
2017-12-20 14:16:46 +01:00
|
|
|
#if defined(CONFIG_MALLOC_TRIM)
|
|
|
|
malloc_trim(4 * 1024 * 1024);
|
|
|
|
#endif
|
2015-02-11 15:51:54 +01:00
|
|
|
qemu_event_wait(&rcu_call_ready_event);
|
|
|
|
}
|
2013-05-13 17:49:24 +02:00
|
|
|
}
|
2015-02-11 15:51:54 +01:00
|
|
|
n = atomic_read(&rcu_call_count);
|
2013-05-13 17:49:24 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
atomic_sub(&rcu_call_count, n);
|
|
|
|
synchronize_rcu();
|
2015-02-11 17:15:18 +01:00
|
|
|
qemu_mutex_lock_iothread();
|
2013-05-13 17:49:24 +02:00
|
|
|
while (n > 0) {
|
|
|
|
node = try_dequeue();
|
|
|
|
while (!node) {
|
2015-02-11 17:15:18 +01:00
|
|
|
qemu_mutex_unlock_iothread();
|
2013-05-13 17:49:24 +02:00
|
|
|
qemu_event_reset(&rcu_call_ready_event);
|
|
|
|
node = try_dequeue();
|
|
|
|
if (!node) {
|
|
|
|
qemu_event_wait(&rcu_call_ready_event);
|
|
|
|
node = try_dequeue();
|
|
|
|
}
|
2015-02-11 17:15:18 +01:00
|
|
|
qemu_mutex_lock_iothread();
|
2013-05-13 17:49:24 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
n--;
|
|
|
|
node->func(node);
|
|
|
|
}
|
2015-02-11 17:15:18 +01:00
|
|
|
qemu_mutex_unlock_iothread();
|
2013-05-13 17:49:24 +02:00
|
|
|
}
|
|
|
|
abort();
|
|
|
|
}
|
|
|
|
|
|
|
|
void call_rcu1(struct rcu_head *node, void (*func)(struct rcu_head *node))
|
|
|
|
{
|
|
|
|
node->func = func;
|
|
|
|
enqueue(node);
|
|
|
|
atomic_inc(&rcu_call_count);
|
|
|
|
qemu_event_set(&rcu_call_ready_event);
|
|
|
|
}
|
|
|
|
|
2013-05-13 13:29:47 +02:00
|
|
|
void rcu_register_thread(void)
|
|
|
|
{
|
|
|
|
assert(rcu_reader.ctr == 0);
|
2015-07-27 04:24:18 +02:00
|
|
|
qemu_mutex_lock(&rcu_registry_lock);
|
2013-05-13 13:29:47 +02:00
|
|
|
QLIST_INSERT_HEAD(®istry, &rcu_reader, node);
|
2015-07-27 04:24:18 +02:00
|
|
|
qemu_mutex_unlock(&rcu_registry_lock);
|
2013-05-13 13:29:47 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
void rcu_unregister_thread(void)
|
|
|
|
{
|
2015-07-27 04:24:18 +02:00
|
|
|
qemu_mutex_lock(&rcu_registry_lock);
|
2013-05-13 13:29:47 +02:00
|
|
|
QLIST_REMOVE(&rcu_reader, node);
|
2015-07-27 04:24:18 +02:00
|
|
|
qemu_mutex_unlock(&rcu_registry_lock);
|
2013-05-13 13:29:47 +02:00
|
|
|
}
|
|
|
|
|
2015-03-05 16:53:48 +01:00
|
|
|
static void rcu_init_complete(void)
|
2013-05-13 13:29:47 +02:00
|
|
|
{
|
2013-05-13 17:49:24 +02:00
|
|
|
QemuThread thread;
|
|
|
|
|
2015-07-27 04:24:18 +02:00
|
|
|
qemu_mutex_init(&rcu_registry_lock);
|
|
|
|
qemu_mutex_init(&rcu_sync_lock);
|
2013-05-13 13:29:47 +02:00
|
|
|
qemu_event_init(&rcu_gp_event, true);
|
2013-05-13 17:49:24 +02:00
|
|
|
|
|
|
|
qemu_event_init(&rcu_call_ready_event, false);
|
2015-03-05 16:53:48 +01:00
|
|
|
|
|
|
|
/* The caller is assumed to have iothread lock, so the call_rcu thread
|
|
|
|
* must have been quiescent even after forking, just recreate it.
|
|
|
|
*/
|
2013-05-13 17:49:24 +02:00
|
|
|
qemu_thread_create(&thread, "call_rcu", call_rcu_thread,
|
|
|
|
NULL, QEMU_THREAD_DETACHED);
|
|
|
|
|
2013-05-13 13:29:47 +02:00
|
|
|
rcu_register_thread();
|
|
|
|
}
|
2015-03-05 16:53:48 +01:00
|
|
|
|
2016-01-27 08:49:21 +01:00
|
|
|
static int atfork_depth = 1;
|
|
|
|
|
|
|
|
void rcu_enable_atfork(void)
|
|
|
|
{
|
|
|
|
atfork_depth++;
|
|
|
|
}
|
|
|
|
|
|
|
|
void rcu_disable_atfork(void)
|
|
|
|
{
|
|
|
|
atfork_depth--;
|
|
|
|
}
|
|
|
|
|
2015-03-05 16:53:48 +01:00
|
|
|
#ifdef CONFIG_POSIX
|
|
|
|
static void rcu_init_lock(void)
|
|
|
|
{
|
2016-01-27 08:49:21 +01:00
|
|
|
if (atfork_depth < 1) {
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2015-07-27 04:24:18 +02:00
|
|
|
qemu_mutex_lock(&rcu_sync_lock);
|
|
|
|
qemu_mutex_lock(&rcu_registry_lock);
|
2015-03-05 16:53:48 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
static void rcu_init_unlock(void)
|
|
|
|
{
|
2016-01-27 08:49:21 +01:00
|
|
|
if (atfork_depth < 1) {
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2015-07-27 04:24:18 +02:00
|
|
|
qemu_mutex_unlock(&rcu_registry_lock);
|
|
|
|
qemu_mutex_unlock(&rcu_sync_lock);
|
2015-03-05 16:53:48 +01:00
|
|
|
}
|
|
|
|
|
2016-03-25 14:00:51 +01:00
|
|
|
static void rcu_init_child(void)
|
2015-03-05 16:53:48 +01:00
|
|
|
{
|
2016-03-25 14:00:51 +01:00
|
|
|
if (atfork_depth < 1) {
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2015-03-05 16:53:48 +01:00
|
|
|
memset(®istry, 0, sizeof(registry));
|
|
|
|
rcu_init_complete();
|
|
|
|
}
|
2016-03-25 14:00:51 +01:00
|
|
|
#endif
|
2015-03-05 16:53:48 +01:00
|
|
|
|
|
|
|
static void __attribute__((__constructor__)) rcu_init(void)
|
|
|
|
{
|
2018-02-16 10:04:18 +01:00
|
|
|
smp_mb_global_init();
|
2015-03-05 16:53:48 +01:00
|
|
|
#ifdef CONFIG_POSIX
|
2016-03-25 14:00:51 +01:00
|
|
|
pthread_atfork(rcu_init_lock, rcu_init_unlock, rcu_init_child);
|
2015-03-05 16:53:48 +01:00
|
|
|
#endif
|
|
|
|
rcu_init_complete();
|
|
|
|
}
|