svn commit: r197445 - head/lib/libthr/thread

Attilio Rao attilio at FreeBSD.org
Wed Sep 23 21:38:58 UTC 2009


Author: attilio
Date: Wed Sep 23 21:38:57 2009
New Revision: 197445
URL: http://svn.freebsd.org/changeset/base/197445

Log:
  rwlock implemented from libthr need to fall through the 'hard path' and
  query umtx also if the shared waiters bit is set on a shared lock.
  The writer starvation avoidance technique, infact, can lead to shared
  waiters on a shared lock which can bring to a missed wakeup and thus
  to a deadlock if the right bit is not checked (a notable case is the
  writers counterpart to be handled through expired timeouts).
  
  Fix that by checking for the shared waiters bit also when unlocking the
  shared locks.
  
  That bug was causing a reported MySQL deadlock.
  Many thanks go to Nick Esborn and his employer DesertNet which provided
  time and machines to identify and fix this issue.
  
  PR:		thread/135673
  Reported by:	Nick Esborn <nick at desert dot net>
  Tested by:	Nick Esborn <nick at desert dot net>
  Reviewed by:	jeff

Modified:
  head/lib/libthr/thread/thr_umtx.h

Modified: head/lib/libthr/thread/thr_umtx.h
==============================================================================
--- head/lib/libthr/thread/thr_umtx.h	Wed Sep 23 20:49:14 2009	(r197444)
+++ head/lib/libthr/thread/thr_umtx.h	Wed Sep 23 21:38:57 2009	(r197445)
@@ -171,8 +171,11 @@ _thr_rwlock_unlock(struct urwlock *rwloc
 		for (;;) {
 			if (__predict_false(URWLOCK_READER_COUNT(state) == 0))
 				return (EPERM);
-			if (!((state & URWLOCK_WRITE_WAITERS) && URWLOCK_READER_COUNT(state) == 1)) {
-				if (atomic_cmpset_rel_32(&rwlock->rw_state, state, state-1))
+			if (!((state & (URWLOCK_WRITE_WAITERS |
+			    URWLOCK_READ_WAITERS)) &&
+			    URWLOCK_READER_COUNT(state) == 1)) {
+				if (atomic_cmpset_rel_32(&rwlock->rw_state,
+				    state, state-1))
 					return (0);
 				state = rwlock->rw_state;
 			} else {


More information about the svn-src-all mailing list