diff options
author | Thomas Pfaff <tpfaff@gmx.net> | 2003-12-01 22:10:57 +0000 |
---|---|---|
committer | Thomas Pfaff <tpfaff@gmx.net> | 2003-12-01 22:10:57 +0000 |
commit | 94d2416049fd158471da45380df829ffce203671 (patch) | |
tree | 21c164059860c586b4a8a72c8d0ecf42f6d40695 /winsup/cygwin/thread.cc | |
parent | bd16a3a8a8a1b0f2971e6bac6c896320f02208c2 (diff) | |
download | cygnal-94d2416049fd158471da45380df829ffce203671.tar.gz cygnal-94d2416049fd158471da45380df829ffce203671.tar.bz2 cygnal-94d2416049fd158471da45380df829ffce203671.zip |
* thread.cc (pthread_rwlock::add_reader): Remove mx parameter for
List_insert call.
(pthread::prepare): Ensure race safeness when adding function
pointers to atfork lists by using List_insert.
* thread.h (List_insert): Use InterlockedCompareExchangePointer to
ensure race safeness without using a mutex.
(List_remove): Use InterlockedCompareExchangePointer to
ensure race safeness with List_insert.
(List::insert): Remove mx parameter for List_insert call.
Diffstat (limited to 'winsup/cygwin/thread.cc')
-rw-r--r-- | winsup/cygwin/thread.cc | 24 |
1 files changed, 4 insertions, 20 deletions
diff --git a/winsup/cygwin/thread.cc b/winsup/cygwin/thread.cc index db12e5cb7..cb99be031 100644 --- a/winsup/cygwin/thread.cc +++ b/winsup/cygwin/thread.cc @@ -1259,7 +1259,7 @@ pthread_rwlock::unlock () void pthread_rwlock::add_reader (struct RWLOCK_READER *rd) { - List_insert (readers_mx, readers, rd); + List_insert (readers, rd); } void @@ -1998,22 +1998,6 @@ pthread::cancel (pthread_t thread) return thread->cancel (); } -/* Races in pthread_atfork: - We are race safe in that any additions to the lists are made via - InterlockedExchangePointer. - However, if the user application doesn't perform syncronisation of some sort - It's not guaranteed that a near simultaneous call to pthread_atfork and fork - will result in the new atfork handlers being calls. - More rigorous internal syncronisation isn't needed as the user program isn't - guaranteeing their own state. - - as far as multiple calls to pthread_atfork, the worst case is simultaneous calls - will result in an indeterminate order for parent and child calls (what gets inserted - first isn't guaranteed.) - - There is one potential race... Does the result of InterlockedExchangePointer - get committed to the return location _before_ any context switches can occur? - If yes, we're safe, if no, we're not. */ void pthread::atforkprepare (void) { @@ -2090,7 +2074,7 @@ pthread::atfork (void (*prepare)(void), void (*parent)(void), void (*child)(void if (prepcb) { prepcb->cb = prepare; - prepcb->next = (callback *) InterlockedExchangePointer ((LONG *) &MT_INTERFACE->pthread_prepare, (long int) prepcb); + List_insert (MT_INTERFACE->pthread_prepare, prepcb); } if (parentcb) { @@ -2099,7 +2083,7 @@ pthread::atfork (void (*prepare)(void), void (*parent)(void), void (*child)(void while (*t) t = &(*t)->next; /* t = pointer to last next in the list */ - parentcb->next = (callback *) InterlockedExchangePointer ((LONG *) t, (long int) parentcb); + List_insert (*t, parentcb); } if (childcb) { @@ -2108,7 +2092,7 @@ pthread::atfork (void (*prepare)(void), void (*parent)(void), void (*child)(void while (*t) t = &(*t)->next; /* t = pointer to last next in the list */ - childcb->next = (callback *) InterlockedExchangePointer ((LONG *) t, (long int) childcb); + List_insert (*t, childcb); } return 0; } |