chan_sip: Address runaway when realtime peers subscribe to mailboxes

Users upgrading from asterisk 13.5 to a later version and who use
realtime with peers that have mailboxes were experiencing runaway
situations that manifested as a continuous stream of taskprocessor
congestion errors, memory leaks and an unresponsive chan_sip.

A related issue was that setting rtcachefriends=no NEVER worked in
asterisk 13 (since the move to stasis).  In 13.5 and earlier, when a
peer tried to register, all of the stasis threads would block and
chan_sip would again become unresponsive.  After 13.5, the runaway
would happen.

There were a number of causes...
* mwi_event_cb was (indirectly) calling build_peer even though calls to
  mwi_event_cb are often caused by build_peer.
* In an effort to prevent chan_sip from being unloaded while messages
  were still in flight, destroy_mailboxes was calling
  stasis_unsubscribe_and_join but in some cases waited forever for the
  final message.
* add_peer_mailboxes wasn't properly marking the existing mailboxes
  on a peer as "keep" so build_peer would always delete them all.
* add_peer_mwi_subs was unsubscribing existing mailbox subscriptions
  then just creating them again.

All of this was causing a flood of subscribes and unsubscribes on
multiple threads all for the same peer and mailbox.

Fixes...
* add_peer_mailboxes now marks mailboxes correctly and build_peer only
  deletes the ones that really are no longer needed by the peer.
* add_peer_mwi_subs now only adds subscriptions marked as "new" instead
  of unsubscribing and resubscribing everything.  It also adds the peer
  object's address to the mailbox instead of its name to the subscription
  userdata so mwi_event_cb doesn't have to call build_peer.

With these changes, with rtcachefriends=yes (the most common setting),
there are no leaks, locks, loops or crashes at shutdown.

rtcachefriends=no still causes leaks but at least it doesn't lock, loop
or crash.  Since making rtcachefriends=no work wasnt in scope for this
issue, further work will have to be deferred to a separate patch.

Side fixes...
 * The ast_lock_track structure had a member named "thread" which gdb
   doesn't like since it conflicts with it's "thread" command.  That
   member was renamed to "thread_id".

ASTERISK-25468 #close

Change-Id: I07519ef7f092629e1e844f855abd279d6475cdd0
This commit is contained in:
George Joseph
2016-09-20 08:42:15 -06:00
committed by Joshua Colp
parent a6b05e6371
commit 0056bcaebd
4 changed files with 67 additions and 39 deletions

View File

@@ -218,7 +218,7 @@ lt_cleanup:
lt->lineno[0] = lineno;
lt->func[0] = func;
lt->reentrancy = 0;
lt->thread[0] = 0;
lt->thread_id[0] = 0;
#ifdef HAVE_BKTR
memset(&lt->backtrace[0], 0, sizeof(lt->backtrace[0]));
#endif
@@ -322,7 +322,7 @@ int __ast_pthread_mutex_lock(const char *filename, int lineno, const char *func,
lt->file[lt->reentrancy] = filename;
lt->lineno[lt->reentrancy] = lineno;
lt->func[lt->reentrancy] = func;
lt->thread[lt->reentrancy] = pthread_self();
lt->thread_id[lt->reentrancy] = pthread_self();
lt->reentrancy++;
} else {
__ast_mutex_logger("%s line %d (%s): '%s' really deep reentrancy!\n",
@@ -402,7 +402,7 @@ int __ast_pthread_mutex_trylock(const char *filename, int lineno, const char *fu
lt->file[lt->reentrancy] = filename;
lt->lineno[lt->reentrancy] = lineno;
lt->func[lt->reentrancy] = func;
lt->thread[lt->reentrancy] = pthread_self();
lt->thread_id[lt->reentrancy] = pthread_self();
lt->reentrancy++;
} else {
__ast_mutex_logger("%s line %d (%s): '%s' really deep reentrancy!\n",
@@ -445,7 +445,7 @@ int __ast_pthread_mutex_unlock(const char *filename, int lineno, const char *fun
if (lt) {
ast_reentrancy_lock(lt);
if (lt->reentrancy && (lt->thread[ROFFSET] != pthread_self())) {
if (lt->reentrancy && (lt->thread_id[ROFFSET] != pthread_self())) {
__ast_mutex_logger("%s line %d (%s): attempted unlock mutex '%s' without owning it!\n",
filename, lineno, func, mutex_name);
__ast_mutex_logger("%s line %d (%s): '%s' was locked here.\n",
@@ -466,7 +466,7 @@ int __ast_pthread_mutex_unlock(const char *filename, int lineno, const char *fun
lt->file[lt->reentrancy] = NULL;
lt->lineno[lt->reentrancy] = 0;
lt->func[lt->reentrancy] = NULL;
lt->thread[lt->reentrancy] = 0;
lt->thread_id[lt->reentrancy] = 0;
}
#ifdef HAVE_BKTR
@@ -536,7 +536,7 @@ static void restore_lock_tracking(struct ast_lock_track *lt, struct ast_lock_tra
memcpy(lt->lineno, lt_saved->lineno, sizeof(lt->lineno));
lt->reentrancy = lt_saved->reentrancy;
memcpy(lt->func, lt_saved->func, sizeof(lt->func));
memcpy(lt->thread, lt_saved->thread, sizeof(lt->thread));
memcpy(lt->thread_id, lt_saved->thread_id, sizeof(lt->thread_id));
#ifdef HAVE_BKTR
memcpy(lt->backtrace, lt_saved->backtrace, sizeof(lt->backtrace));
#endif
@@ -571,7 +571,7 @@ int __ast_cond_wait(const char *filename, int lineno, const char *func,
if (lt) {
ast_reentrancy_lock(lt);
if (lt->reentrancy && (lt->thread[ROFFSET] != pthread_self())) {
if (lt->reentrancy && (lt->thread_id[ROFFSET] != pthread_self())) {
__ast_mutex_logger("%s line %d (%s): attempted wait using mutex '%s' without owning it!\n",
filename, lineno, func, mutex_name);
__ast_mutex_logger("%s line %d (%s): '%s' was locked here.\n",
@@ -639,7 +639,7 @@ int __ast_cond_timedwait(const char *filename, int lineno, const char *func,
if (lt) {
ast_reentrancy_lock(lt);
if (lt->reentrancy && (lt->thread[ROFFSET] != pthread_self())) {
if (lt->reentrancy && (lt->thread_id[ROFFSET] != pthread_self())) {
__ast_mutex_logger("%s line %d (%s): attempted wait using mutex '%s' without owning it!\n",
filename, lineno, func, mutex_name);
__ast_mutex_logger("%s line %d (%s): '%s' was locked here.\n",
@@ -747,7 +747,7 @@ lt_cleanup:
lt->lineno[0] = lineno;
lt->func[0] = func;
lt->reentrancy = 0;
lt->thread[0] = 0;
lt->thread_id[0] = 0;
#ifdef HAVE_BKTR
memset(&lt->backtrace[0], 0, sizeof(lt->backtrace[0]));
#endif
@@ -790,13 +790,13 @@ int __ast_rwlock_unlock(const char *filename, int line, const char *func, ast_rw
int i;
pthread_t self = pthread_self();
for (i = lt->reentrancy - 1; i >= 0; --i) {
if (lt->thread[i] == self) {
if (lt->thread_id[i] == self) {
lock_found = 1;
if (i != lt->reentrancy - 1) {
lt->file[i] = lt->file[lt->reentrancy - 1];
lt->lineno[i] = lt->lineno[lt->reentrancy - 1];
lt->func[i] = lt->func[lt->reentrancy - 1];
lt->thread[i] = lt->thread[lt->reentrancy - 1];
lt->thread_id[i] = lt->thread_id[lt->reentrancy - 1];
}
#ifdef HAVE_BKTR
bt = &lt->backtrace[i];
@@ -804,7 +804,7 @@ int __ast_rwlock_unlock(const char *filename, int line, const char *func, ast_rw
lt->file[lt->reentrancy - 1] = NULL;
lt->lineno[lt->reentrancy - 1] = 0;
lt->func[lt->reentrancy - 1] = NULL;
lt->thread[lt->reentrancy - 1] = AST_PTHREADT_NULL;
lt->thread_id[lt->reentrancy - 1] = AST_PTHREADT_NULL;
break;
}
}
@@ -918,7 +918,7 @@ int __ast_rwlock_rdlock(const char *filename, int line, const char *func, ast_rw
lt->file[lt->reentrancy] = filename;
lt->lineno[lt->reentrancy] = line;
lt->func[lt->reentrancy] = func;
lt->thread[lt->reentrancy] = pthread_self();
lt->thread_id[lt->reentrancy] = pthread_self();
lt->reentrancy++;
}
ast_reentrancy_unlock(lt);
@@ -1027,7 +1027,7 @@ int __ast_rwlock_wrlock(const char *filename, int line, const char *func, ast_rw
lt->file[lt->reentrancy] = filename;
lt->lineno[lt->reentrancy] = line;
lt->func[lt->reentrancy] = func;
lt->thread[lt->reentrancy] = pthread_self();
lt->thread_id[lt->reentrancy] = pthread_self();
lt->reentrancy++;
}
ast_reentrancy_unlock(lt);
@@ -1120,7 +1120,7 @@ int __ast_rwlock_timedrdlock(const char *filename, int line, const char *func, a
lt->file[lt->reentrancy] = filename;
lt->lineno[lt->reentrancy] = line;
lt->func[lt->reentrancy] = func;
lt->thread[lt->reentrancy] = pthread_self();
lt->thread_id[lt->reentrancy] = pthread_self();
lt->reentrancy++;
}
ast_reentrancy_unlock(lt);
@@ -1213,7 +1213,7 @@ int __ast_rwlock_timedwrlock(const char *filename, int line, const char *func, a
lt->file[lt->reentrancy] = filename;
lt->lineno[lt->reentrancy] = line;
lt->func[lt->reentrancy] = func;
lt->thread[lt->reentrancy] = pthread_self();
lt->thread_id[lt->reentrancy] = pthread_self();
lt->reentrancy++;
}
ast_reentrancy_unlock(lt);
@@ -1288,7 +1288,7 @@ int __ast_rwlock_tryrdlock(const char *filename, int line, const char *func, ast
lt->file[lt->reentrancy] = filename;
lt->lineno[lt->reentrancy] = line;
lt->func[lt->reentrancy] = func;
lt->thread[lt->reentrancy] = pthread_self();
lt->thread_id[lt->reentrancy] = pthread_self();
lt->reentrancy++;
}
ast_reentrancy_unlock(lt);
@@ -1347,7 +1347,7 @@ int __ast_rwlock_trywrlock(const char *filename, int line, const char *func, ast
lt->file[lt->reentrancy] = filename;
lt->lineno[lt->reentrancy] = line;
lt->func[lt->reentrancy] = func;
lt->thread[lt->reentrancy] = pthread_self();
lt->thread_id[lt->reentrancy] = pthread_self();
lt->reentrancy++;
}
ast_reentrancy_unlock(lt);