Hi Stefan,
On Tue, Jan 24, 2017 at 1:37 PM, Stefan Eissing
<[email protected]> wrote:
> Yann, thanks for the patch. I agree that the cleanups need to be killed in
> the right place. Not certain if it was wrong before, but that part is not
> easy to see for every combination.
>
> I did some rework and hope this makes it more readable. If you find the time
> to look at it, feedback welcome.
I still fear that if beam->pool gets destroyed while both
beam_send_cleanup() and beam_cleanup() are registered, the former is
called twice.
I'd change:
if (safe_send) {
if (beam->send_pool && beam->send_pool != beam->pool) {
apr_pool_cleanup_kill(beam->send_pool, beam, beam_send_cleanup);
}
status = beam_send_cleanup(beam);
}
with:
if (safe_send) {
if (beam->send_pool) {
if (beam->send_pool != beam->pool) {
apr_pool_cleanup_kill(beam->send_pool, beam, beam_send_cleanup);
}
status = beam_send_cleanup(beam);
}
}
since in the above case beam_send_cleanup is run first and sets send_pool=NULL.
Attached v3 with this only change w.r.t. v2.
Otherwise, looks good to me, thanks!
Index: modules/http2/h2_bucket_beam.c
===================================================================
--- modules/http2/h2_bucket_beam.c (revision 1780129)
+++ modules/http2/h2_bucket_beam.c (working copy)
@@ -438,18 +438,37 @@ static apr_status_t beam_recv_cleanup(void *data)
return APR_SUCCESS;
}
+static int pool_register(h2_bucket_beam *beam, apr_pool_t *pool,
+ apr_status_t (*cleanup)(void *))
+{
+ if (pool && pool != beam->pool) {
+ apr_pool_pre_cleanup_register(pool, beam, cleanup);
+ return 1;
+ }
+ return 0;
+}
+
+static int pool_kill(h2_bucket_beam *beam, apr_pool_t *pool,
+ apr_status_t (*cleanup)(void *)) {
+ if (pool && pool != beam->pool) {
+ apr_pool_cleanup_kill(pool, beam, cleanup);
+ return 1;
+ }
+ return 0;
+}
+
static void beam_set_recv_pool(h2_bucket_beam *beam, apr_pool_t *pool)
{
- /* if the beam owner is the sender, monitor receiver pool lifetime */
- if (beam->owner == H2_BEAM_OWNER_SEND && beam->recv_pool != pool) {
- if (beam->recv_pool) {
- apr_pool_cleanup_kill(beam->recv_pool, beam, beam_recv_cleanup);
- }
- beam->recv_pool = pool;
- if (beam->recv_pool) {
- apr_pool_pre_cleanup_register(beam->recv_pool, beam,
beam_recv_cleanup);
- }
+ if (beam->recv_pool == pool ||
+ (beam->recv_pool && pool
+ && apr_pool_is_ancestor(beam->recv_pool, pool))) {
+ /* when receiver same or sub-pool of existing, stick
+ * to the the pool we already have. */
+ return;
}
+ pool_kill(beam, beam->recv_pool, beam_recv_cleanup);
+ beam->recv_pool = pool;
+ pool_register(beam, beam->recv_pool, beam_recv_cleanup);
}
static apr_status_t beam_send_cleanup(void *data)
@@ -473,22 +492,16 @@ static apr_status_t beam_send_cleanup(void *data)
static void beam_set_send_pool(h2_bucket_beam *beam, apr_pool_t *pool)
{
- /* if the beam owner is the receiver, monitor sender pool lifetime */
- if (beam->owner == H2_BEAM_OWNER_RECV && beam->send_pool != pool) {
- if (beam->send_pool && pool
- && apr_pool_is_ancestor(beam->send_pool, pool)) {
- /* when sender uses sub-pools to transmit data, stick
- * to the lifetime of the pool we already have. */
- return;
- }
- if (beam->send_pool) {
- apr_pool_cleanup_kill(beam->send_pool, beam, beam_send_cleanup);
- }
- beam->send_pool = pool;
- if (beam->send_pool) {
- apr_pool_pre_cleanup_register(beam->send_pool, beam,
beam_send_cleanup);
- }
+ if (beam->send_pool == pool ||
+ (beam->send_pool && pool
+ && apr_pool_is_ancestor(beam->send_pool, pool))) {
+ /* when sender is same or sub-pool of existing, stick
+ * to the the pool we already have. */
+ return;
}
+ pool_kill(beam, beam->send_pool, beam_send_cleanup);
+ beam->send_pool = pool;
+ pool_register(beam, beam->send_pool, beam_send_cleanup);
}
static apr_status_t beam_cleanup(void *data)
@@ -495,44 +508,57 @@ static apr_status_t beam_cleanup(void *data)
{
h2_bucket_beam *beam = data;
apr_status_t status = APR_SUCCESS;
- /* owner of the beam is going away, depending on its role, cleanup
- * strategies differ. */
- beam_close(beam);
- switch (beam->owner) {
- case H2_BEAM_OWNER_SEND:
- status = beam_send_cleanup(beam);
- beam->recv_buffer = NULL;
+ int safe_send = !beam->m_enter || (beam->owner == H2_BEAM_OWNER_SEND);
+ int safe_recv = !beam->m_enter || (beam->owner == H2_BEAM_OWNER_RECV);
+
+ /*
+ * Owner of the beam is going away, depending on which side it owns,
+ * cleanup strategies will differ with multi-thread protection
+ * still in place (beam->m_enter).
+ *
+ * In general, receiver holds references to memory from sender.
+ * Clean up receiver first, if safe, then cleanup sender, if safe.
+ */
+
+ /* When modify send is not safe, this means we still have multi-thread
+ * protection and the owner is receiving the buckets. If the sending
+ * side has not gone away, this means we could have dangling buckets
+ * in our lists that never get destroyed. This should not happen. */
+ ap_assert(safe_send || !beam->send_pool);
+ if (!H2_BLIST_EMPTY(&beam->send_list)) {
+ ap_assert(beam->send_pool);
+ }
+
+ if (safe_recv) {
+ if (beam->recv_pool) {
+ apr_pool_cleanup_kill(beam->recv_pool, beam, beam_recv_cleanup);
beam->recv_pool = NULL;
- break;
- case H2_BEAM_OWNER_RECV:
- if (beam->recv_buffer) {
- apr_brigade_destroy(beam->recv_buffer);
- }
+ }
+ if (beam->recv_buffer) {
+ apr_brigade_destroy(beam->recv_buffer);
beam->recv_buffer = NULL;
- beam->recv_pool = NULL;
- if (!H2_BLIST_EMPTY(&beam->send_list)) {
- ap_assert(beam->send_pool);
- }
- if (beam->send_pool) {
- /* sender has not cleaned up, its pool still lives.
- * this is normal if the sender uses cleanup via a bucket
- * such as the BUCKET_EOR for requests. In that case, the
- * beam should have lost its mutex protection, meaning
- * it is no longer used multi-threaded and we can safely
- * purge all remaining sender buckets. */
+ }
+ }
+ else {
+ beam->recv_buffer = NULL;
+ beam->recv_pool = NULL;
+ }
+
+ if (safe_send) {
+ if (beam->send_pool) {
+ if (beam->send_pool != beam->pool) {
apr_pool_cleanup_kill(beam->send_pool, beam,
beam_send_cleanup);
- ap_assert(!beam->m_enter);
- beam_send_cleanup(beam);
}
- ap_assert(H2_BPROXY_LIST_EMPTY(&beam->proxies));
- ap_assert(H2_BLIST_EMPTY(&beam->send_list));
- ap_assert(H2_BLIST_EMPTY(&beam->hold_list));
- ap_assert(H2_BLIST_EMPTY(&beam->purge_list));
- break;
- default:
- ap_assert(NULL);
- break;
+ status = beam_send_cleanup(beam);
+ }
}
+
+ if (safe_recv) {
+ ap_assert(H2_BPROXY_LIST_EMPTY(&beam->proxies));
+ ap_assert(H2_BLIST_EMPTY(&beam->send_list));
+ ap_assert(H2_BLIST_EMPTY(&beam->hold_list));
+ ap_assert(H2_BLIST_EMPTY(&beam->purge_list));
+ }
return status;
}
Index: modules/proxy/mod_proxy.h
===================================================================
--- modules/proxy/mod_proxy.h (revision 1780129)
+++ modules/proxy/mod_proxy.h (working copy)
@@ -347,9 +347,9 @@ PROXY_WORKER_HC_FAIL )
#define PROXY_WORKER_MAX_SCHEME_SIZE 16
#define PROXY_WORKER_MAX_ROUTE_SIZE 64
#define PROXY_BALANCER_MAX_ROUTE_SIZE PROXY_WORKER_MAX_ROUTE_SIZE
-#define PROXY_WORKER_MAX_NAME_SIZE 96
+#define PROXY_WORKER_MAX_NAME_SIZE 544
#define PROXY_BALANCER_MAX_NAME_SIZE PROXY_WORKER_MAX_NAME_SIZE
-#define PROXY_WORKER_MAX_HOSTNAME_SIZE 64
+#define PROXY_WORKER_MAX_HOSTNAME_SIZE 512
#define PROXY_BALANCER_MAX_HOSTNAME_SIZE PROXY_WORKER_MAX_HOSTNAME_SIZE
#define PROXY_BALANCER_MAX_STICKY_SIZE 64
Index: modules/slotmem/mod_slotmem_shm.c
===================================================================
--- modules/slotmem/mod_slotmem_shm.c (revision 1780129)
+++ modules/slotmem/mod_slotmem_shm.c (working copy)
@@ -386,22 +386,26 @@ static apr_status_t slotmem_create(ap_slotmem_inst
if (apr_shm_size_get(shm) != size) {
apr_shm_detach(shm);
ap_log_error(APLOG_MARK, APLOG_ERR, 0, ap_server_conf,
APLOGNO(02599)
- "existing shared memory for %s could not be used
(failed size check)",
+ "existing shared memory for %s could not be reused
(failed size check)",
fname);
- return APR_EINVAL;
+ rv = APR_EINVAL;
}
- ptr = (char *)apr_shm_baseaddr_get(shm);
- memcpy(&desc, ptr, sizeof(desc));
- if (desc.size != item_size || desc.num != item_num) {
- apr_shm_detach(shm);
- ap_log_error(APLOG_MARK, APLOG_ERR, 0, ap_server_conf,
APLOGNO(02600)
- "existing shared memory for %s could not be used
(failed contents check)",
- fname);
- return APR_EINVAL;
+ else {
+ ptr = (char *)apr_shm_baseaddr_get(shm);
+ memcpy(&desc, ptr, sizeof(desc));
+ if (desc.size != item_size || desc.num != item_num) {
+ apr_shm_detach(shm);
+ ap_log_error(APLOG_MARK, APLOG_ERR, 0, ap_server_conf,
APLOGNO(02600)
+ "existing shared memory for %s could not be
reused (failed contents check)",
+ fname);
+ rv = APR_EINVAL;
+ }
+ else {
+ ptr += AP_SLOTMEM_OFFSET;
+ }
}
- ptr += AP_SLOTMEM_OFFSET;
}
- else {
+ if (rv != APR_SUCCESS) {
apr_size_t dsize = size - AP_SLOTMEM_OFFSET;
if (fbased) {
apr_shm_remove(fname, gpool);
Index: server/mpm/event/event.c
===================================================================
--- server/mpm/event/event.c (revision 1780129)
+++ server/mpm/event/event.c (working copy)
@@ -177,6 +177,7 @@ static int dying = 0;
static int workers_may_exit = 0;
static int start_thread_may_exit = 0;
static int listener_may_exit = 0;
+static int listener_is_wakeable = 0; /* Pollset supports
APR_POLLSET_WAKEABLE */
static int num_listensocks = 0;
static apr_int32_t conns_this_child; /* MaxConnectionsPerChild, only
access
in listener thread */
@@ -199,6 +200,19 @@ module AP_MODULE_DECLARE_DATA mpm_event_module;
struct event_srv_cfg_s;
typedef struct event_srv_cfg_s event_srv_cfg;
+static apr_pollfd_t *listener_pollfd;
+
+/*
+ * The pollset for sockets that are in any of the timeout queues. Currently
+ * we use the timeout_mutex to make sure that connections are added/removed
+ * atomically to/from both event_pollset and a timeout queue. Otherwise
+ * some confusion can happen under high load if timeout queues and pollset
+ * get out of sync.
+ * XXX: It should be possible to make the lock unnecessary in many or even all
+ * XXX: cases.
+ */
+static apr_pollset_t *event_pollset;
+
struct event_conn_state_t {
/** APR_RING of expiration timeouts */
APR_RING_ENTRY(event_conn_state_t) timeout_list;
@@ -228,9 +242,10 @@ APR_RING_HEAD(timeout_head_t, event_conn_state_t);
struct timeout_queue {
struct timeout_head_t head;
- int count, *total;
apr_interval_time_t timeout;
- struct timeout_queue *next;
+ apr_uint32_t count; /* for this queue */
+ apr_uint32_t *total; /* for all chained/related queues */
+ struct timeout_queue *next; /* chaining */
};
/*
* Several timeout queues that use different timeouts, so that we always can
@@ -244,52 +259,65 @@ static struct timeout_queue *write_completion_q,
*keepalive_q,
*linger_q,
*short_linger_q;
+static volatile apr_time_t queues_next_expiry;
-static apr_pollfd_t *listener_pollfd;
+/* Prevent extra poll/wakeup calls for timeouts close in the future (queues
+ * have the granularity of a second anyway).
+ * XXX: Wouldn't 0.5s (instead of 0.1s) be "enough"?
+ */
+#define TIMEOUT_FUDGE_FACTOR apr_time_from_msec(100)
/*
* Macros for accessing struct timeout_queue.
* For TO_QUEUE_APPEND and TO_QUEUE_REMOVE, timeout_mutex must be held.
*/
-#define TO_QUEUE_APPEND(q, el) \
- do { \
- APR_RING_INSERT_TAIL(&(q)->head, el, event_conn_state_t, \
- timeout_list); \
- ++*(q)->total; \
- ++(q)->count; \
- } while (0)
+static void TO_QUEUE_APPEND(struct timeout_queue *q, event_conn_state_t *el)
+{
+ apr_time_t q_expiry;
+ apr_time_t next_expiry;
-#define TO_QUEUE_REMOVE(q, el) \
- do { \
- APR_RING_REMOVE(el, timeout_list); \
- --*(q)->total; \
- --(q)->count; \
- } while (0)
+ APR_RING_INSERT_TAIL(&q->head, el, event_conn_state_t, timeout_list);
+ apr_atomic_inc32(q->total);
+ ++q->count;
-#define TO_QUEUE_INIT(q, p, t, v) \
- do { \
- struct timeout_queue *b = (v); \
- (q) = apr_palloc((p), sizeof *(q)); \
- APR_RING_INIT(&(q)->head, event_conn_state_t, timeout_list); \
- (q)->total = (b) ? (b)->total : apr_pcalloc((p), sizeof *(q)->total); \
- (q)->count = 0; \
- (q)->timeout = (t); \
- (q)->next = NULL; \
- } while (0)
+ /* Cheaply update the overall queues' next expiry according to the
+ * first entry of this queue (oldest), if necessary.
+ */
+ el = APR_RING_FIRST(&q->head);
+ q_expiry = el->queue_timestamp + q->timeout;
+ next_expiry = queues_next_expiry;
+ if (!next_expiry || next_expiry > q_expiry + TIMEOUT_FUDGE_FACTOR) {
+ queues_next_expiry = q_expiry;
+ /* Unblock the poll()ing listener for it to update its timeout. */
+ if (listener_is_wakeable) {
+ apr_pollset_wakeup(event_pollset);
+ }
+ }
+}
-#define TO_QUEUE_ELEM_INIT(el) APR_RING_ELEM_INIT(el, timeout_list)
+static void TO_QUEUE_REMOVE(struct timeout_queue *q, event_conn_state_t *el)
+{
+ APR_RING_REMOVE(el, timeout_list);
+ apr_atomic_dec32(q->total);
+ --q->count;
+}
-/*
- * The pollset for sockets that are in any of the timeout queues. Currently
- * we use the timeout_mutex to make sure that connections are added/removed
- * atomically to/from both event_pollset and a timeout queue. Otherwise
- * some confusion can happen under high load if timeout queues and pollset
- * get out of sync.
- * XXX: It should be possible to make the lock unnecessary in many or even all
- * XXX: cases.
- */
-static apr_pollset_t *event_pollset;
+static struct timeout_queue *TO_QUEUE_MAKE(apr_pool_t *p, apr_time_t t,
+ struct timeout_queue *ref)
+{
+ struct timeout_queue *q;
+
+ q = apr_pcalloc(p, sizeof *q);
+ APR_RING_INIT(&q->head, event_conn_state_t, timeout_list);
+ q->total = (ref) ? ref->total : apr_pcalloc(p, sizeof *q->total);
+ q->timeout = t;
+ return q;
+}
+
+#define TO_QUEUE_ELEM_INIT(el) \
+ APR_RING_ELEM_INIT((el), timeout_list)
+
/* The structure used to pass unique initialization info to each thread */
typedef struct
{
@@ -474,6 +502,11 @@ static void wakeup_listener(void)
return;
}
+ /* Unblock the listener if it's poll()ing */
+ if (listener_is_wakeable) {
+ apr_pollset_wakeup(event_pollset);
+ }
+
/* unblock the listener if it's waiting for a worker */
ap_queue_info_term(worker_queue_info);
@@ -647,7 +680,11 @@ static apr_status_t decrement_connection_count(voi
default:
break;
}
- apr_atomic_dec32(&connection_count);
+ /* Unblock the listener if it's waiting for connection_count = 0 */
+ if (!apr_atomic_dec32(&connection_count)
+ && listener_is_wakeable && listener_may_exit) {
+ apr_pollset_wakeup(event_pollset);
+ }
return APR_SUCCESS;
}
@@ -810,6 +847,7 @@ static void notify_resume(event_conn_state_t *cs,
static int start_lingering_close_common(event_conn_state_t *cs, int in_worker)
{
+ int done = 0;
apr_status_t rv;
struct timeout_queue *q;
apr_socket_t *csd = cs->pfd.desc.s;
@@ -842,25 +880,24 @@ static int start_lingering_close_common(event_conn
else {
cs->c->sbh = NULL;
}
- apr_thread_mutex_lock(timeout_mutex);
- TO_QUEUE_APPEND(q, cs);
cs->pfd.reqevents = (
cs->pub.sense == CONN_SENSE_WANT_WRITE ? APR_POLLOUT :
APR_POLLIN) | APR_POLLHUP | APR_POLLERR;
cs->pub.sense = CONN_SENSE_DEFAULT;
+ apr_thread_mutex_lock(timeout_mutex);
rv = apr_pollset_add(event_pollset, &cs->pfd);
+ if (rv == APR_SUCCESS || APR_STATUS_IS_EEXIST(rv)) {
+ TO_QUEUE_APPEND(q, cs);
+ done = 1;
+ }
apr_thread_mutex_unlock(timeout_mutex);
- if (rv != APR_SUCCESS && !APR_STATUS_IS_EEXIST(rv)) {
+ if (!done) {
ap_log_error(APLOG_MARK, APLOG_ERR, rv, ap_server_conf, APLOGNO(03092)
"start_lingering_close: apr_pollset_add failure");
- apr_thread_mutex_lock(timeout_mutex);
- TO_QUEUE_REMOVE(q, cs);
- apr_thread_mutex_unlock(timeout_mutex);
apr_socket_close(cs->pfd.desc.s);
ap_push_pool(worker_queue_info, cs->p);
- return 0;
}
- return 1;
+ return done;
}
/*
@@ -1124,16 +1161,27 @@ read_request:
* Set a write timeout for this connection, and let the
* event thread poll for writeability.
*/
+ int done = 0;
cs->queue_timestamp = apr_time_now();
notify_suspend(cs);
- apr_thread_mutex_lock(timeout_mutex);
- TO_QUEUE_APPEND(cs->sc->wc_q, cs);
cs->pfd.reqevents = (
cs->pub.sense == CONN_SENSE_WANT_READ ? APR_POLLIN :
APR_POLLOUT) | APR_POLLHUP | APR_POLLERR;
cs->pub.sense = CONN_SENSE_DEFAULT;
+ apr_thread_mutex_lock(timeout_mutex);
rc = apr_pollset_add(event_pollset, &cs->pfd);
+ if (rc == APR_SUCCESS || APR_STATUS_IS_EEXIST(rc)) {
+ TO_QUEUE_APPEND(cs->sc->wc_q, cs);
+ done = 1;
+ }
apr_thread_mutex_unlock(timeout_mutex);
+ if (!done) {
+ ap_log_error(APLOG_MARK, APLOG_ERR, rc, ap_server_conf,
APLOGNO(03465)
+ "process_socket: apr_pollset_add failure for "
+ "write completion");
+ apr_socket_close(cs->pfd.desc.s);
+ ap_push_pool(worker_queue_info, cs->p);
+ }
return;
}
else if (c->keepalive != AP_CONN_KEEPALIVE || c->aborted ||
@@ -1161,20 +1209,26 @@ read_request:
* timeout today. With a normal client, the socket will be readable in
* a few milliseconds anyway.
*/
+ int done = 0;
cs->queue_timestamp = apr_time_now();
notify_suspend(cs);
- apr_thread_mutex_lock(timeout_mutex);
- TO_QUEUE_APPEND(cs->sc->ka_q, cs);
/* Add work to pollset. */
cs->pfd.reqevents = APR_POLLIN;
+ apr_thread_mutex_lock(timeout_mutex);
rc = apr_pollset_add(event_pollset, &cs->pfd);
+ if (rc == APR_SUCCESS || APR_STATUS_IS_EEXIST(rc)) {
+ TO_QUEUE_APPEND(cs->sc->ka_q, cs);
+ done = 1;
+ }
apr_thread_mutex_unlock(timeout_mutex);
-
- if (rc != APR_SUCCESS) {
+ if (!done) {
ap_log_error(APLOG_MARK, APLOG_ERR, rc, ap_server_conf,
APLOGNO(03093)
- "process_socket: apr_pollset_add failure");
- AP_DEBUG_ASSERT(rc == APR_SUCCESS);
+ "process_socket: apr_pollset_add failure for "
+ "keep alive");
+ apr_socket_close(cs->pfd.desc.s);
+ ap_push_pool(worker_queue_info, cs->p);
+ return;
}
}
else if (cs->pub.state == CONN_STATE_SUSPENDED) {
@@ -1345,7 +1399,14 @@ static void get_worker(int *have_idle_worker_p, in
static APR_RING_HEAD(timer_free_ring_t, timer_event_t) timer_free_ring;
static apr_skiplist *timer_skiplist;
+static volatile apr_time_t timers_next_expiry;
+/* Same goal as for TIMEOUT_FUDGE_FACTOR (avoid extra poll calls), but applied
+ * to timers. Since their timeouts are custom (user defined), we can't be too
+ * approximative here (hence using 0.01s).
+ */
+#define EVENT_FUDGE_FACTOR apr_time_from_msec(10)
+
/* The following compare function is used by apr_skiplist_insert() to keep the
* elements (timers) sorted and provide O(log n) complexity (this is also true
* for apr_skiplist_{find,remove}(), but those are not used in MPM event where
@@ -1391,9 +1452,25 @@ static apr_status_t event_register_timed_callback(
/* XXXXX: optimize */
te->when = t + apr_time_now();
- /* Okay, add sorted by when.. */
- apr_skiplist_insert(timer_skiplist, te);
+ {
+ apr_time_t next_expiry;
+ /* Okay, add sorted by when.. */
+ apr_skiplist_insert(timer_skiplist, te);
+
+ /* Cheaply update the overall timers' next expiry according to
+ * this event, if necessary.
+ */
+ next_expiry = timers_next_expiry;
+ if (!next_expiry || next_expiry > te->when + EVENT_FUDGE_FACTOR) {
+ timers_next_expiry = te->when;
+ /* Unblock the poll()ing listener for it to update its timeout. */
+ if (listener_is_wakeable) {
+ apr_pollset_wakeup(event_pollset);
+ }
+ }
+ }
+
apr_thread_mutex_unlock(g_timer_skiplist_mtx);
return APR_SUCCESS;
@@ -1427,15 +1504,14 @@ static void process_lingering_close(event_conn_sta
apr_thread_mutex_lock(timeout_mutex);
rv = apr_pollset_remove(event_pollset, pfd);
- AP_DEBUG_ASSERT(rv == APR_SUCCESS);
+ TO_QUEUE_REMOVE(q, cs);
+ apr_thread_mutex_unlock(timeout_mutex);
+ AP_DEBUG_ASSERT(rv == APR_SUCCESS || APR_STATUS_IS_NOTFOUND(rv));
+ TO_QUEUE_ELEM_INIT(cs);
rv = apr_socket_close(csd);
AP_DEBUG_ASSERT(rv == APR_SUCCESS);
- TO_QUEUE_REMOVE(q, cs);
- apr_thread_mutex_unlock(timeout_mutex);
- TO_QUEUE_ELEM_INIT(cs);
-
ap_push_pool(worker_queue_info, cs->p);
if (dying)
ap_queue_interrupt_one(worker_queue);
@@ -1449,13 +1525,13 @@ static void process_timeout_queue(struct timeout_q
apr_time_t timeout_time,
int (*func)(event_conn_state_t *))
{
- int total = 0, count;
+ apr_uint32_t total = 0, count;
event_conn_state_t *first, *cs, *last;
struct timeout_head_t trash;
struct timeout_queue *qp;
apr_status_t rv;
- if (!*q->total) {
+ if (!apr_atomic_read32(q->total)) {
return;
}
@@ -1464,20 +1540,32 @@ static void process_timeout_queue(struct timeout_q
count = 0;
cs = first = last = APR_RING_FIRST(&qp->head);
while (cs != APR_RING_SENTINEL(&qp->head, event_conn_state_t,
- timeout_list)
- /* Trash the entry if:
- * - no timeout_time was given (asked for all), or
- * - it expired (according to the queue timeout), or
- * - the system clock skewed in the past: no entry should be
- * registered above the given timeout_time (~now) + the queue
- * timeout, we won't keep any here (eg. for centuries).
- * Stop otherwise, no following entry will match thanks to the
- * single timeout per queue (entries are added to the end!).
- * This allows maintenance in O(1).
- */
- && (!timeout_time
- || cs->queue_timestamp + qp->timeout < timeout_time
- || cs->queue_timestamp > timeout_time + qp->timeout)) {
+ timeout_list)) {
+ /* Trash the entry if:
+ * - no timeout_time was given (asked for all), or
+ * - it expired (according to the queue timeout), or
+ * - the system clock skewed in the past: no entry should be
+ * registered above the given timeout_time (~now) + the queue
+ * timeout, we won't keep any here (eg. for centuries).
+ *
+ * Otherwise stop, no following entry will match thanks to the
+ * single timeout per queue (entries are added to the end!).
+ * This allows maintenance in O(1).
+ */
+ if (timeout_time
+ && cs->queue_timestamp + qp->timeout > timeout_time
+ && cs->queue_timestamp < timeout_time + qp->timeout) {
+ /* Since this is the next expiring of this queue, update the
+ * overall queues' next expiry if it's later than this one.
+ */
+ apr_time_t q_expiry = cs->queue_timestamp + qp->timeout;
+ apr_time_t next_expiry = queues_next_expiry;
+ if (!next_expiry || next_expiry > q_expiry) {
+ queues_next_expiry = q_expiry;
+ }
+ break;
+ }
+
last = cs;
rv = apr_pollset_remove(event_pollset, &cs->pfd);
if (rv != APR_SUCCESS && !APR_STATUS_IS_NOTFOUND(rv)) {
@@ -1493,6 +1581,8 @@ static void process_timeout_queue(struct timeout_q
APR_RING_UNSPLICE(first, last, timeout_list);
APR_RING_SPLICE_TAIL(&trash, first, last, event_conn_state_t,
timeout_list);
+ AP_DEBUG_ASSERT(apr_atomic_read32(q->total) >= count);
+ apr_atomic_sub32(q->total, count);
qp->count -= count;
total += count;
}
@@ -1499,8 +1589,6 @@ static void process_timeout_queue(struct timeout_q
if (!total)
return;
- AP_DEBUG_ASSERT(*q->total >= total);
- *q->total -= total;
apr_thread_mutex_unlock(timeout_mutex);
first = APR_RING_FIRST(&trash);
do {
@@ -1512,13 +1600,28 @@ static void process_timeout_queue(struct timeout_q
apr_thread_mutex_lock(timeout_mutex);
}
+static void process_keepalive_queue(apr_time_t timeout_time)
+{
+ /* If all workers are busy, we kill older keep-alive connections so
+ * that they may connect to another process.
+ */
+ if (!timeout_time) {
+ ap_log_error(APLOG_MARK, APLOG_TRACE1, 0, ap_server_conf,
+ "All workers are busy or dying, will close %u "
+ "keep-alive connections",
+ apr_atomic_read32(keepalive_q->total));
+ }
+ process_timeout_queue(keepalive_q, timeout_time,
+ start_lingering_close_nonblocking);
+}
+
static void * APR_THREAD_FUNC listener_thread(apr_thread_t * thd, void *dummy)
{
- timer_event_t *ep;
timer_event_t *te;
apr_status_t rc;
proc_info *ti = dummy;
int process_slot = ti->pslot;
+ struct process_score *ps = ap_get_scoreboard_process(process_slot);
apr_pool_t *tpool = apr_thread_pool_get(thd);
void *csd = NULL;
apr_pool_t *ptrans; /* Pool for per-transaction stuff */
@@ -1534,14 +1637,6 @@ static void * APR_THREAD_FUNC listener_thread(apr_
last_log = apr_time_now();
free(ti);
- /* the following times out events that are really close in the future
- * to prevent extra poll calls
- *
- * current value is .1 second
- */
-#define TIMEOUT_FUDGE_FACTOR 100000
-#define EVENT_FUDGE_FACTOR 10000
-
rc = init_pollset(tpool);
if (rc != APR_SUCCESS) {
ap_log_error(APLOG_MARK, APLOG_ERR, rc, ap_server_conf,
@@ -1559,6 +1654,7 @@ static void * APR_THREAD_FUNC listener_thread(apr_
for (;;) {
int workers_were_busy = 0;
+
if (listener_may_exit) {
close_listeners(process_slot, &closed);
if (terminate_mode == ST_UNGRACEFUL
@@ -1572,7 +1668,7 @@ static void * APR_THREAD_FUNC listener_thread(apr_
now = apr_time_now();
if (APLOGtrace6(ap_server_conf)) {
/* trace log status every second */
- if (now - last_log > apr_time_from_msec(1000)) {
+ if (now - last_log > apr_time_from_sec(1)) {
last_log = now;
apr_thread_mutex_lock(timeout_mutex);
ap_log_error(APLOG_MARK, APLOG_TRACE6, 0, ap_server_conf,
@@ -1580,8 +1676,8 @@ static void * APR_THREAD_FUNC listener_thread(apr_
"keep-alive: %d lingering: %d suspended: %u)",
apr_atomic_read32(&connection_count),
apr_atomic_read32(&clogged_count),
- *write_completion_q->total,
- *keepalive_q->total,
+ apr_atomic_read32(write_completion_q->total),
+ apr_atomic_read32(keepalive_q->total),
apr_atomic_read32(&lingering_count),
apr_atomic_read32(&suspended_count));
if (dying) {
@@ -1594,32 +1690,71 @@ static void * APR_THREAD_FUNC listener_thread(apr_
}
}
- apr_thread_mutex_lock(g_timer_skiplist_mtx);
- te = apr_skiplist_peek(timer_skiplist);
- if (te) {
- if (te->when > now) {
- timeout_interval = te->when - now;
+ /* Start with an infinite poll() timeout and update it according to
+ * the next expiring timer or queue entry. If there are none, either
+ * the listener is wakeable and it can poll() indefinitely until a wake
+ * up occurs, otherwise periodic checks (maintenance, shutdown, ...)
+ * must be performed.
+ */
+ timeout_interval = -1;
+
+ /* Push expired timers to a worker, the first remaining one determines
+ * the maximum time to poll() below, if any.
+ */
+ timeout_time = timers_next_expiry;
+ if (timeout_time && timeout_time < now + EVENT_FUDGE_FACTOR) {
+ apr_thread_mutex_lock(g_timer_skiplist_mtx);
+ while ((te = apr_skiplist_peek(timer_skiplist))) {
+ if (te->when > now + EVENT_FUDGE_FACTOR) {
+ timers_next_expiry = te->when;
+ timeout_interval = te->when - now;
+ break;
+ }
+ apr_skiplist_pop(timer_skiplist, NULL);
+ push_timer2worker(te);
}
- else {
- timeout_interval = 1;
+ if (!te) {
+ timers_next_expiry = 0;
}
+ apr_thread_mutex_unlock(g_timer_skiplist_mtx);
}
- else {
- timeout_interval = apr_time_from_msec(100);
+
+ /* Same for queues, use their next expiry, if any. */
+ timeout_time = queues_next_expiry;
+ if (timeout_time
+ && (timeout_interval < 0
+ || timeout_time <= now
+ || timeout_interval > timeout_time - now)) {
+ timeout_interval = timeout_time > now ? timeout_time - now : 1;
}
- apr_thread_mutex_unlock(g_timer_skiplist_mtx);
+ /* When non-wakeable, don't wait more than 100 ms, in any case. */
+#define NON_WAKEABLE_POLL_TIMEOUT apr_time_from_msec(100)
+ if (!listener_is_wakeable
+ && (timeout_interval < 0
+ || timeout_interval > NON_WAKEABLE_POLL_TIMEOUT)) {
+ timeout_interval = NON_WAKEABLE_POLL_TIMEOUT;
+ }
+
rc = apr_pollset_poll(event_pollset, timeout_interval, &num, &out_pfd);
if (rc != APR_SUCCESS) {
if (APR_STATUS_IS_EINTR(rc)) {
- continue;
+ /* Woken up, if we are exiting we must fall through to kill
+ * kept-alive connections, otherwise we only need to update
+ * timeouts (logic is above, so restart the loop).
+ */
+ if (!listener_may_exit) {
+ continue;
+ }
+ timeout_time = 0;
}
- if (!APR_STATUS_IS_TIMEUP(rc)) {
+ else if (!APR_STATUS_IS_TIMEUP(rc)) {
ap_log_error(APLOG_MARK, APLOG_CRIT, rc, ap_server_conf,
"apr_pollset_poll failed. Attempting to "
"shutdown process gracefully");
signal_threads(ST_GRACEFUL);
}
+ num = 0;
}
if (listener_may_exit) {
@@ -1629,21 +1764,6 @@ static void * APR_THREAD_FUNC listener_thread(apr_
break;
}
- now = apr_time_now();
- apr_thread_mutex_lock(g_timer_skiplist_mtx);
- ep = apr_skiplist_peek(timer_skiplist);
- while (ep) {
- if (ep->when < now + EVENT_FUDGE_FACTOR) {
- apr_skiplist_pop(timer_skiplist, NULL);
- push_timer2worker(ep);
- }
- else {
- break;
- }
- ep = apr_skiplist_peek(timer_skiplist);
- }
- apr_thread_mutex_unlock(g_timer_skiplist_mtx);
-
while (num) {
pt = (listener_poll_type *) out_pfd->client_data;
if (pt->type == PT_CSD) {
@@ -1666,7 +1786,6 @@ static void * APR_THREAD_FUNC listener_thread(apr_
TO_QUEUE_REMOVE(remove_from_q, cs);
rc = apr_pollset_remove(event_pollset, &cs->pfd);
apr_thread_mutex_unlock(timeout_mutex);
-
/*
* Some of the pollset backends, like KQueue or Epoll
* automagically remove the FD if the socket is closed,
@@ -1808,52 +1927,55 @@ static void * APR_THREAD_FUNC listener_thread(apr_
/* XXX possible optimization: stash the current time for use as
* r->request_time for new requests
*/
- now = apr_time_now();
- /* We only do this once per 0.1s (TIMEOUT_FUDGE_FACTOR), or on a clock
- * skew (if the system time is set back in the meantime, timeout_time
- * will exceed now + TIMEOUT_FUDGE_FACTOR, can't happen otherwise).
+ /* We process the timeout queues here only when their overall next
+ * expiry (read once above) is over. This happens accurately since
+ * adding to the queues (in workers) can only decrease this expiry,
+ * while latest ones are only taken into account here (in listener)
+ * during queues' processing, with the lock held. This works both
+ * with and without wake-ability.
*/
- if (now > timeout_time || now + TIMEOUT_FUDGE_FACTOR < timeout_time ) {
- struct process_score *ps;
+ if (timeout_time && timeout_time < (now = apr_time_now())) {
timeout_time = now + TIMEOUT_FUDGE_FACTOR;
/* handle timed out sockets */
apr_thread_mutex_lock(timeout_mutex);
+ /* Processing all the queues below will recompute this. */
+ queues_next_expiry = 0;
+
/* Step 1: keepalive timeouts */
- /* If all workers are busy, we kill older keep-alive connections
so that they
- * may connect to another process.
- */
- if ((workers_were_busy || dying) && *keepalive_q->total) {
- if (!dying)
- ap_log_error(APLOG_MARK, APLOG_TRACE1, 0, ap_server_conf,
- "All workers are busy, will close %d
keep-alive "
- "connections",
- *keepalive_q->total);
- process_timeout_queue(keepalive_q, 0,
- start_lingering_close_nonblocking);
+ if (workers_were_busy || dying) {
+ process_keepalive_queue(0); /* kill'em all \m/ */
}
else {
- process_timeout_queue(keepalive_q, timeout_time,
- start_lingering_close_nonblocking);
+ process_keepalive_queue(timeout_time);
}
/* Step 2: write completion timeouts */
process_timeout_queue(write_completion_q, timeout_time,
start_lingering_close_nonblocking);
/* Step 3: (normal) lingering close completion timeouts */
- process_timeout_queue(linger_q, timeout_time,
stop_lingering_close);
+ process_timeout_queue(linger_q, timeout_time,
+ stop_lingering_close);
/* Step 4: (short) lingering close completion timeouts */
- process_timeout_queue(short_linger_q, timeout_time,
stop_lingering_close);
+ process_timeout_queue(short_linger_q, timeout_time,
+ stop_lingering_close);
- ps = ap_get_scoreboard_process(process_slot);
- ps->write_completion = *write_completion_q->total;
- ps->keep_alive = *keepalive_q->total;
apr_thread_mutex_unlock(timeout_mutex);
+ ps->keep_alive = apr_atomic_read32(keepalive_q->total);
+ ps->write_completion =
apr_atomic_read32(write_completion_q->total);
ps->connections = apr_atomic_read32(&connection_count);
ps->suspended = apr_atomic_read32(&suspended_count);
ps->lingering_close = apr_atomic_read32(&lingering_count);
}
+ else if ((workers_were_busy || dying)
+ && apr_atomic_read32(keepalive_q->total)) {
+ apr_thread_mutex_lock(timeout_mutex);
+ process_keepalive_queue(0); /* kill'em all \m/ */
+ apr_thread_mutex_unlock(timeout_mutex);
+ ps->keep_alive = apr_atomic_read32(keepalive_q->total);
+ }
+
if (listeners_disabled && !workers_were_busy
&& (int)apr_atomic_read32(&connection_count)
- (int)apr_atomic_read32(&lingering_count)
@@ -2064,6 +2186,8 @@ static void *APR_THREAD_FUNC start_threads(apr_thr
int prev_threads_created;
int max_recycled_pools = -1;
int good_methods[] = {APR_POLLSET_KQUEUE, APR_POLLSET_PORT,
APR_POLLSET_EPOLL};
+ /* XXX don't we need more to handle K-A or lingering close? */
+ const apr_uint32_t pollset_size = threads_per_child * 2;
/* We must create the fd queues before we start up the listener
* and worker threads. */
@@ -2103,24 +2227,24 @@ static void *APR_THREAD_FUNC start_threads(apr_thr
/* Create the main pollset */
for (i = 0; i < sizeof(good_methods) / sizeof(good_methods[0]); i++) {
- rv = apr_pollset_create_ex(&event_pollset,
- threads_per_child*2, /* XXX don't we need more, to
handle
- * connections in K-A or
lingering
- * close?
- */
- pchild, APR_POLLSET_THREADSAFE |
APR_POLLSET_NOCOPY | APR_POLLSET_NODEFAULT,
- good_methods[i]);
+ apr_uint32_t flags = APR_POLLSET_THREADSAFE | APR_POLLSET_NOCOPY |
+ APR_POLLSET_NODEFAULT | APR_POLLSET_WAKEABLE;
+ rv = apr_pollset_create_ex(&event_pollset, pollset_size, pchild, flags,
+ good_methods[i]);
if (rv == APR_SUCCESS) {
+ listener_is_wakeable = 1;
break;
}
+ flags &= ~APR_POLLSET_WAKEABLE;
+ rv = apr_pollset_create_ex(&event_pollset, pollset_size, pchild, flags,
+ good_methods[i]);
+ if (rv == APR_SUCCESS) {
+ break;
+ }
}
if (rv != APR_SUCCESS) {
- rv = apr_pollset_create(&event_pollset,
- threads_per_child*2, /* XXX don't we need more,
to handle
- * connections in K-A or
lingering
- * close?
- */
- pchild, APR_POLLSET_THREADSAFE |
APR_POLLSET_NOCOPY);
+ rv = apr_pollset_create(&event_pollset, pollset_size, pchild,
+ APR_POLLSET_THREADSAFE | APR_POLLSET_NOCOPY);
}
if (rv != APR_SUCCESS) {
ap_log_error(APLOG_MARK, APLOG_ERR, rv, ap_server_conf, APLOGNO(03103)
@@ -2129,7 +2253,9 @@ static void *APR_THREAD_FUNC start_threads(apr_thr
}
ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, ap_server_conf, APLOGNO(02471)
- "start_threads: Using %s",
apr_pollset_method_name(event_pollset));
+ "start_threads: Using %s (%swakeable)",
+ apr_pollset_method_name(event_pollset),
+ listener_is_wakeable ? "" : "not ");
worker_sockets = apr_pcalloc(pchild, threads_per_child
* sizeof(apr_socket_t *));
@@ -3262,10 +3388,10 @@ static int event_post_config(apr_pool_t *pconf, ap
wc.hash = apr_hash_make(ptemp);
ka.hash = apr_hash_make(ptemp);
- TO_QUEUE_INIT(linger_q, pconf,
- apr_time_from_sec(MAX_SECS_TO_LINGER), NULL);
- TO_QUEUE_INIT(short_linger_q, pconf,
- apr_time_from_sec(SECONDS_TO_LINGER), NULL);
+ linger_q = TO_QUEUE_MAKE(pconf, apr_time_from_sec(MAX_SECS_TO_LINGER),
+ NULL);
+ short_linger_q = TO_QUEUE_MAKE(pconf, apr_time_from_sec(SECONDS_TO_LINGER),
+ NULL);
for (; s; s = s->next) {
event_srv_cfg *sc = apr_pcalloc(pconf, sizeof *sc);
@@ -3273,11 +3399,11 @@ static int event_post_config(apr_pool_t *pconf, ap
ap_set_module_config(s->module_config, &mpm_event_module, sc);
if (!wc.tail) {
/* The main server uses the global queues */
- TO_QUEUE_INIT(wc.q, pconf, s->timeout, NULL);
+ wc.q = TO_QUEUE_MAKE(pconf, s->timeout, NULL);
apr_hash_set(wc.hash, &s->timeout, sizeof s->timeout, wc.q);
wc.tail = write_completion_q = wc.q;
- TO_QUEUE_INIT(ka.q, pconf, s->keep_alive_timeout, NULL);
+ ka.q = TO_QUEUE_MAKE(pconf, s->keep_alive_timeout, NULL);
apr_hash_set(ka.hash, &s->keep_alive_timeout,
sizeof s->keep_alive_timeout, ka.q);
ka.tail = keepalive_q = ka.q;
@@ -3287,7 +3413,7 @@ static int event_post_config(apr_pool_t *pconf, ap
* or their own queue(s) if there isn't */
wc.q = apr_hash_get(wc.hash, &s->timeout, sizeof s->timeout);
if (!wc.q) {
- TO_QUEUE_INIT(wc.q, pconf, s->timeout, wc.tail);
+ wc.q = TO_QUEUE_MAKE(pconf, s->timeout, wc.tail);
apr_hash_set(wc.hash, &s->timeout, sizeof s->timeout, wc.q);
wc.tail = wc.tail->next = wc.q;
}
@@ -3295,7 +3421,7 @@ static int event_post_config(apr_pool_t *pconf, ap
ka.q = apr_hash_get(ka.hash, &s->keep_alive_timeout,
sizeof s->keep_alive_timeout);
if (!ka.q) {
- TO_QUEUE_INIT(ka.q, pconf, s->keep_alive_timeout, ka.tail);
+ ka.q = TO_QUEUE_MAKE(pconf, s->keep_alive_timeout, ka.tail);
apr_hash_set(ka.hash, &s->keep_alive_timeout,
sizeof s->keep_alive_timeout, ka.q);
ka.tail = ka.tail->next = ka.q;
Index: server/mpm/event/fdqueue.c
===================================================================
--- server/mpm/event/fdqueue.c (revision 1780129)
+++ server/mpm/event/fdqueue.c (working copy)
@@ -269,9 +269,8 @@ void ap_pop_pool(apr_pool_t ** recycled_pool, fd_q
if (first_pool == NULL) {
break;
}
- if (apr_atomic_casptr
- ((void*) &(queue_info->recycled_pools),
- first_pool->next, first_pool) == first_pool) {
+ if (apr_atomic_casptr((void*) &(queue_info->recycled_pools),
+ first_pool->next, first_pool) == first_pool) {
*recycled_pool = first_pool->pool;
if (queue_info->max_recycled_pools >= 0)
apr_atomic_dec32(&queue_info->recycled_pools_count);
