From 6e3ac5c605d5062279178b3ea0b853d0e9cf7cc9 Mon Sep 17 00:00:00 2001 From: rpj Date: Thu, 10 Mar 2011 13:40:16 +0000 Subject: Replace global Critical Sections with MCS Queue locks --- ptw32_cond_check_need_init.c | 22 +++------------------- 1 file changed, 3 insertions(+), 19 deletions(-) (limited to 'ptw32_cond_check_need_init.c') diff --git a/ptw32_cond_check_need_init.c b/ptw32_cond_check_need_init.c index 31359ad..ec3e8bb 100644 --- a/ptw32_cond_check_need_init.c +++ b/ptw32_cond_check_need_init.c @@ -43,29 +43,13 @@ INLINE int ptw32_cond_check_need_init (pthread_cond_t * cond) { int result = 0; + ptw32_mcs_local_node_t node; /* * The following guarded test is specifically for statically * initialised condition variables (via PTHREAD_OBJECT_INITIALIZER). - * - * Note that by not providing this synchronisation we risk - * introducing race conditions into applications which are - * correctly written. - * - * Approach - * -------- - * We know that static condition variables will not be PROCESS_SHARED - * so we can serialise access to internal state using - * Win32 Critical Sections rather than Win32 Mutexes. - * - * If using a single global lock slows applications down too much, - * multiple global locks could be created and hashed on some random - * value associated with each mutex, the pointer perhaps. At a guess, - * a good value for the optimal number of global locks might be - * the number of processors + 1. - * */ - EnterCriticalSection (&ptw32_cond_test_init_lock); + ptw32_mcs_lock_acquire(&ptw32_cond_test_init_lock, &node); /* * We got here possibly under race @@ -88,7 +72,7 @@ ptw32_cond_check_need_init (pthread_cond_t * cond) result = EINVAL; } - LeaveCriticalSection (&ptw32_cond_test_init_lock); + ptw32_mcs_lock_release(&node); return result; } -- cgit v1.2.3