Rich Felker 6016457011 process-shared barrier support, based on discussion with bdonlan ... this implementation is rather heavy-weight, but it's the first solution i've found that's actually correct. all waiters actually wait twice at the barrier so that they can synchronize exit, and they hold a "vm lock" that prevents changes to virtual memory mappings (and blocks pthread_barrier_destroy) until all waiters are finished inspecting the barrier. thus, it is safe for any thread to destroy and/or unmap the barrier's memory as soon as pthread_barrier_wait returns, without further synchronization.		2011-09-27 13:50:29 -04:00
..
madvise.c	global cleanup to use the new syscall interface	2011-03-20 00:16:43 -04:00
mlock.c	global cleanup to use the new syscall interface	2011-03-20 00:16:43 -04:00
mlockall.c	global cleanup to use the new syscall interface	2011-03-20 00:16:43 -04:00
mmap.c	process-shared barrier support, based on discussion with bdonlan	2011-09-27 13:50:29 -04:00
mprotect.c	work around linux bug in mprotect	2011-06-29 00:42:42 -04:00
mremap.c	global cleanup to use the new syscall interface	2011-03-20 00:16:43 -04:00
msync.c	global cleanup to use the new syscall interface	2011-03-20 00:16:43 -04:00
munlock.c	global cleanup to use the new syscall interface	2011-03-20 00:16:43 -04:00
munlockall.c	global cleanup to use the new syscall interface	2011-03-20 00:16:43 -04:00
munmap.c	process-shared barrier support, based on discussion with bdonlan	2011-09-27 13:50:29 -04:00
posix_madvise.c	fix missing include in posix_madvise.c (compile error)	2011-04-20 15:36:15 -04:00
shm_open.c	implement POSIX shared memory	2011-03-03 00:30:31 -05:00
shm_unlink.c	implement POSIX shared memory	2011-03-03 00:30:31 -05:00