summaryrefslogtreecommitdiffstats
path: root/lib/libthr
Commit message (Expand)AuthorAgeFilesLines
* add code to support stack unwinding when thread exits. note that onlydavidxu2010-09-156-1/+184
* Move back IN_GCLIST flag into field tlflags, since thread list and gc listdavidxu2010-09-151-6/+6
* Don't compare thread pointers again.davidxu2010-09-131-3/+3
* Fix copy&paste problem.davidxu2010-09-131-2/+2
* Update symbol.davidxu2010-09-131-1/+1
* PS_DEAD state needs not be checked because _thr_find_thread() has alreadydavidxu2010-09-131-4/+0
* Convert thread list lock from mutex to rwlock.davidxu2010-09-1315-129/+153
* Merge from tbemd, with a small amount of rework:imp2010-09-139-16/+11
* Because POSIX does not allow EINTR to be returned from sigwait(),davidxu2010-09-101-6/+8
* To avoid possible race condition, SIGCANCEL is always sent except thedavidxu2010-09-081-1/+1
* Fix off-by-one error in function _thr_sigact_unload, also disable thedavidxu2010-09-061-3/+5
* Remove incorrect comments, also make sure signal isdavidxu2010-09-011-5/+4
* In function __pthread_cxa_finalize(), also make code for removingdavidxu2010-09-011-1/+8
* pthread_atfork should acquire writer lock and protect the codedavidxu2010-09-011-1/+3
* Change atfork lock from mutex to rwlock, also make mutexes used by malloc()davidxu2010-09-016-28/+60
* Add signal handler wrapper, the reason to add it becauses there aredavidxu2010-09-0114-273/+591
* Unregister thread specific data destructor when a corresponding dsodavidxu2010-08-273-0/+23
* clear lock to zero state if it is destroyed.davidxu2010-08-271-0/+3
* eliminate unused code.davidxu2010-08-262-17/+0
* Decrease rdlock count only when thread unlocked a reader lock.davidxu2010-08-261-1/+1
* Unify 32-bit and 64-bit PowerPC libthr support. This reduces codenwhitehorn2010-08-246-147/+16
* Remove unused source.kib2010-08-241-57/+0
* The __hidden definition is provided by sys/cdefs.h.kib2010-08-241-4/+0
* Add wrapper for setcontext() and swapcontext(), the wrappersdavidxu2010-08-243-14/+55
* On shared object unload, in __cxa_finalize, call and clear all installedkib2010-08-233-0/+26
* Reduce redundant code.davidxu2010-08-201-76/+21
* In current implementation, thread cancellation is done in signal handler,davidxu2010-08-206-99/+284
* According to specification, function fcntl() is a cancellation point onlydavidxu2010-08-201-4/+11
* Tweak code a bit to be POSIX compatible, when a cancellation requestdavidxu2010-08-171-0/+2
* Use _SIG_VALID instead of expanded form of the macro.kib2010-07-121-1/+1
* Fix SVN mismerge. We somehow ended up with the 32-bit powerpc versionnwhitehorn2010-07-112-4/+4
* Powerpc64 thread libraries support.nwhitehorn2010-07-103-0/+143
* Coalesce one more broken line.deischen2010-05-241-2/+1
* Coalesce a couple of broken lines since they can fit within 80deischen2010-05-241-4/+2
* mdoc: order prologue macros consistently by Dd/Dt/Osuqs2010-04-141-1/+1
* Merge r195129 from project/mips to head by hand:imp2010-01-091-2/+6
* remove file thr_sem_new.c.davidxu2010-01-051-1/+0
* Remove extra new semaphore stubs, because libc already has them, anddavidxu2010-01-052-110/+0
* Use umtx to implement process sharable semaphore, to make this work,davidxu2010-01-056-232/+172
* Work-around a race condition on ia64 while unlocking a contested lock.marcel2009-12-141-0/+3
* Revert r199830 for now. Too many ports dlopen() libraries linked withkib2009-11-281-1/+1
* Libthr cannot be dynamically loaded into the running process.kib2009-11-261-1/+1
* Current pselect(3) is implemented in usermode and thus vulnerable tokib2009-10-271-1/+3
* Implement _umtx_op_err() for ia64.marcel2009-10-243-1/+38
* Make openat(2) a cancellation point.jilles2009-10-113-0/+34
* don't report error if key was deleted.davidxu2009-09-251-1/+11
* rwlock implemented from libthr need to fall through the 'hard path' andattilio2009-09-231-2/+5
* In the current code, rdlock_count is not correctly handled for some cases.attilio2009-07-062-1/+4
* These are some cosmetic changes to improve the clarity of libthr's fork imple...green2009-05-111-9/+9
* Now that the kernel defines CACHE_LINE_SIZE in machine/param.h, userwatson2009-04-191-2/+0
OpenPOWER on IntegriCloud