summaryrefslogtreecommitdiff
path: root/lib/libpthread/uthread
AgeCommit message (Collapse)Author
2006-04-02malloc(x * y) -> calloc(x, y) from adobriyan AT gmail.com, with tweaksDamien Miller
suggested by kjell@; ok otto@ pat@ millert@ jaredy@
2006-03-06Add _thread_atexit_lock and _thread_atexit_unlock to the list ofOtto Moerbeek
weak functions; ok millert@.
2006-02-16cap seconds to 100 million per man page and kernel nanosleep impl.Kurt Miller
prevents userland from causing an overflow of tv_sec. okay marc@
2006-01-06Adapt things to use __type_t instead of _BSD_TYPE_T_Todd C. Miller
Add new sys/_types.h header Include machine/_types.h or sys/_types.h where applicable
2005-12-21Userland programs should include <errno.h> not <sys/errno.h>Todd C. Miller
OK deraadt@
2005-10-31- Prefix MUTEX_TYPE_MAX with PTHREAD_ to avoid namespace pollution.Brad Smith
- Remove the macros MUTEX_TYPE_FAST and MUTEX_TYPE_COUNTING_FAST. From stefanf FreeBSD ok fgsch@
2005-10-30Add pthread_atfork(3)Brad Smith
From FreeBSD 'looks ok' fgsch@ miod@ man page reviewed by jmc@
2005-10-30Don't use TAILQ_NEXT() on an element that has been removed. Similar toKenneth R Westerback
otto@'s diff for uvm_aobj.c. Identical to a diff canacar@ developed independantly. ok brad@ 'looks correct' fgsch@
2005-10-29Don't use TAILQ_NEXT() on an element that has been removed. Similar toKenneth R Westerback
otto@'s diff for uvm_aobj.c. ok deraadt@ pedro@
2005-10-14Use queue macros instead of directly accessing fields. ok marc@ miod@Otto Moerbeek
patrick@
2005-01-28Grab the current time before calculating thread wake-up time.Marco S Hyman
Solves a problem where select/poll calls would return early without indicating any error. Tested by otto@ and kurt@ -- thanks.
2005-01-23Add debug support.Mark Kettenis
ok marc@
2004-12-12Avoid using void pointers in additive expressionsBrad Smith
From FreeBSD ok marc@
2004-11-30* make sure we leave the thread's cancellation point before we returnPatrick Latifi
* plug a memory leak and FD_UNLOCK properly in writev ok marc
2004-09-18Allow thread in FD{R,W}_WAIT states to be suspended.Marco S Hyman
Patch from Kurt Miller <truk at optonline.net> Commit reminder from millert@. OK brad@
2004-06-07major bump to libc and libpthread to break the dependency of aMarco S Hyman
particular implementation of libpthread for libc. libc no longer needs pthread.h to compile. OK millert@, brad@, tedu@
2004-05-01pthead_supsend_all_np and pthread_resume_all_np from freebsd viaMarco S Hyman
Kurt Miller <truk at optonline.net>. Kurt says: They're needed for the java hotspot compiler in the 1.3+ jdk OK brad@
2004-04-13refresh curthread after a thread switch so it points to the realMarco S Hyman
current thread. ok deraadt@
2004-04-10ansiBrad Smith
2004-04-10styleBrad Smith
2004-04-09When poll(2)'ing for readability or writability of a file descriptorBrad Smith
on behalf of a thread, we should check the POLLERR, POLLHUP, and POLLNVAL flags as well to wake up the thread in these cases. From: FreeBSD's libc_r ok marc@
2004-04-06Apply patch from guenther at mho.com (pr3730) that fixes a doubleMarco S Hyman
free in the threads library
2004-02-22add pthread_attr_[get|set]stackBrad Smith
From: FreeBSD's libc_r ok marc@
2004-02-16fixes function parameters to match the man page as wellBrad Smith
as adding a more aggressive arguement check. From: FreeBSD's libc_r. ok marc@
2004-02-16Add missing const for function parameters to match POSIX.Brad Smith
From: FreeBSD's libc_r ok marc@
2004-02-01Add a simple work-around for deadlocking on recursiveBrad Smith
readlocks on a rwlock while there are writers waiting. From: FreeBSD's libc_r ok marc@
2004-01-26Change read_stackseg_np to work with any thread, not just the currentMarco S Hyman
thread. Requested by truk at optonline dot net and OK-ed by tedu @.
2004-01-19Fix -Wsign-compare warnings pointed out by kevlo@, ok marc@Todd C. Miller
2004-01-16Include correct includes to unbreak compilation on macppc. ok miod@Otto Moerbeek
2004-01-15threaded version of closefromMarco S Hyman
pthread_stackseg_np added. Minor bump for these changes occurred a day or so ago and will not be bumped again
2004-01-03Fixes from FreeBSD's libc_rBrad Smith
rev 1.21 Make the libc_r version of select() set the readable or writable file descriptor bit if poll() returns POLLERR, POLLHUP, or POLLNVAL. Othewise, it's possible for select() to return successfully but with no bits set. rev 1.19 Return correct number of total bits set in all fd_set's. Change case of POLLNVAL as an error. Remove POLLHUP and POLLERR from one case, their place is most likely amongst read events. ok marc@
2004-01-01more cancellation points.Brad Smith
ok marc@
2003-12-31there is no pause() wrapper (thread_sys_pause) so remove the prototypeBrad Smith
2003-12-31undefer sigs in all paths through the join code.Marco S Hyman
Based upon patch/report from Marc Olzheim <marcolz at stack.nl>
2003-12-23Add stub implementations of pthread_[gs]etconcurrencyBrad Smith
From: FreeBSD' libc_r ok marc@
2003-12-23change the variable name used for status checking,Brad Smith
LIBC_R_DEBUG -> LIBPTHREAD_DEBUG ok marc@
2003-12-23Based upon a freebsd change:Marco S Hyman
If an application closes one of its stdio descriptors (0..2), an excessive close() on one of these descriptors would cause a memory for this descriptor to be allocated in the internal descriptor table. When this descriptor gets used again, e.g. through the call to open() or socket(), the descriptor would be erroneously left in the blocking mode, and the whole application would get stuck on a blocking operation, e.g., in accept(2). but changed to not eat fds when a file that the thread kernel doesn't know about is closed.
2003-12-23Make accept(), connect(), recvfrom(), recvmsg(), sendmsg(),Brad Smith
and sendto() cancellation points, as required by POSIX.1-2001. From: FreeBSD' libc_r ok marc@
2003-12-22Fix from FreeBSD' libc_rBrad Smith
rev 1.21 Fix bogus return values from libc_r's writev() routine in situations where a partial-write is followed by an error. ok marc@
2003-12-22Fixes from FreeBSD' libc_rBrad Smith
rev 1.21 Fix bogus return values from libc_r's write() routine in situations where partial-write is followed by an error. rev 1.22 If __sys_write() returns 0, allow that to exit the loop in libc_r's wrapped version of write(). rev 1.23 Return to the caller if write() returns 0. ok marc@
2003-12-10Add an nfds_t type as per POSIX and also add pollfd_t like Solaris has.Todd C. Miller
2003-11-10missing proto for kqueue(); marc@ okMichael Shalayeff
2003-11-03replace magic number for min number of poll_data entriesMarco S Hyman
to allocate with a define
2003-10-22Add wrapper for kqueue() to keep track of the allocated fd and allow it toBrad Smith
be closed. This fixes a file descriptor leak when closing a kqueue() fd. from FreeBSD ok marc@
2003-10-22When about to do an execve(), don't bother resetting the O_NONBLOCK flagBrad Smith
on any file descriptors that have the close-on-exec flag set. from FreeBSD ok marc@
2003-10-19Fix optimized select handling buglet. Patch from tholo@Marco S Hyman
2003-10-19sync FreeBSD RCS idBrad Smith
2003-10-16Optimize conversion of fd_set -> pollfd. OK and test by marc@Todd C. Miller
2003-09-22Off-by-ones, from aaron@Miod Vallat
2003-08-06Remove some double semicolons (hmm, do two semis equal a maxi?).Todd C. Miller
I've skipped the GNU stuff for now. From Patrick Latifi.