summaryrefslogtreecommitdiffstats
path: root/lib/libpthread
Commit message (Collapse)AuthorAgeFilesLines
* The second call to _thread_kern_sig_defer() in sem_post() should be a calljasone2000-08-231-1/+1
| | | | to _thread_kern_sig_undefer().
* Fix an off-by-one error in the recursive mutex handling that made italfred2000-08-131-3/+3
| | | | | | | prematurely release recursive mutexes. Test case provided by: Bradley T. Hughes <bhughes@trolltech.com> Reviewed by: deischen
* Add wrapper for kevent() syscalljlemon2000-08-074-3/+5
| | | | Noted as missing by: nicolas.leonard@animaths.com
* Make sem_post() safe to call from within a signal handler, as required byjasone2000-08-012-0/+10
| | | | POSIX/SUSv2.
* Call _thread_init() from pthread_once() if it has not already been called.dfr2000-07-211-0/+2
| | | | | | | This fixes a segfault in some C++ programs which use exceptions before main() has been called (i.e. from global constructors). Reviewed by: deischen
* pthread_once --> pthread_once_t.jasone2000-07-191-1/+1
|
* Change my email address in the copyright notices for the sake of consistencyjasone2000-07-1823-23/+23
| | | | (jasone@canonware.com --> jasone@freebsd.org).
* Deal correctly with statically initialized condition variables injasone2000-07-171-18/+17
| | | | | | | | | pthread_cond_signal(), pthread_cond_broadcast(), and pthread_cond_timedwait(). Do not dump core in pthread_cond_timedwait() (due to a NULL pointer dereference) if attempting to wait on an uninitialized condition variable. PR: bin/18099
* Reshuffle the SEE ALSO section.jasone2000-07-173-7/+7
| | | | Prompted by: sheldonh
* Remove DEBUG_FLAGS=-g3, which never should have been committed.jasone2000-07-051-2/+0
|
* Fix typo in SEE ALSO section.jasone2000-06-281-1/+1
|
* If multiple threads are blocked in sigwait() for the same signal that doesjasone2000-06-273-4/+37
| | | | | | | | | | | | | | | not have a user-supplied signal handler, when a signal is delivered, one thread will receive the signal, and then the code reverts to having no signal handler for the signal. This can leave the other sigwait()ing threads stranded permanently if the signal is later ignored, or can result in process termination when the process should have delivered the signal to one of the threads in sigwait(). To fix this problem, maintain a count of sigwait()ers for each signal that has no default signal handler. Use the count to correctly install/uninstall dummy signal handlers. Reviewed by: deischen
* pthread_mutex_lock(), pthread_cond_trywait(), and pthread_cond_wait() arejasone2000-06-146-36/+125
| | | | | | | | | | not allowed to return EINTR, but use of pthread_suspend_np() could cause EINTR to be returned. To fix this, restructure pthread_suspend_np() so that it does not interrupt a thread that is waiting on a mutex or condition, and keep enough state around that pthread_resume_np() can fix things up afterwards. Reviewed by: deischen
* Fix a misspelling: undefied -> undefinedchris2000-05-271-1/+1
|
* Back out the previous change to the queue(3) interface.jake2000-05-261-20/+20
| | | | | | It was not discussed and should probably not happen. Requested by: msmith and others
* Change the way that the queue(3) structures are declared; don't assume thatjake2000-05-231-20/+20
| | | | | | | | the type argument to *_HEAD and *_ENTRY is a struct. Suggested by: phk Reviewed by: phk Approved by: mdodd
* Fix a memory leak. pthread_set_name_np() allocates space for a name, butjasone2000-05-161-2/+6
| | | | | | was not deallocating space for the previous name, if any. PR: misc/18504
* Fix a memory leak. pthread_set_name_np() allocates space for a name, butjasone2000-05-161-0/+4
| | | | | | _thread_gc() was not deallocating it. PR: misc/18504
* add MLINKS: pthread_testcancel(3) -> pthread_setcanceltype(3)phantom2000-05-121-0/+1
|
* "Fixed" missing include in synopsis. POSIX.1-1996 only specifiesbde2000-05-112-0/+2
| | | | | including <signal.h>, but that must be a bug in POSIX.1, because it also specifies that the relevant prototype is [only] in <pthread.h>.
* Fixed misspelling of a struct tag in a function parameter type.bde2000-05-111-1/+1
|
* Fixed missing consts for function parameters, so that the code matchesbde2000-05-113-3/+6
| | | | | the man page and POSIX.1. Fixed nearby misformatting. Fixed a missing prototype.
* Fix miscellaneous mdoc macro argument limit infringements.sheldonh2000-05-091-1/+1
| | | | | PR: 18465 Reported by: Kazu TAKAMUNE <takamune@avrl.mei.co.jp>
* mdoc(7) cleanup:phantom2000-05-047-63/+79
| | | | | | | | | | | . use real function names as `.Nm' macro argument in NAME section. It allows them to appear in apropos(1) or whatis(1) output. . replace empty lines with `.Pp' macro. . replace hardcoded standard names with their `.St' macro equivalents. . sort cross references in SEE ALSO section
* Add missing .El macro.jasone2000-05-031-0/+1
|
* Add missing man pages. Fix various compliance bugs, mostly having to do withjasone2000-05-0213-25/+875
| | | | | | error return values. Implement pthread_mutexattr_gettype(). PR: docs/16537, docs/17538
* Add a wrapper for the sendfile() system call.jasone2000-04-272-1/+4
| | | | PR: bin/17366
* Add test to detect propagation of cancellation points within libc_r.jasone2000-04-262-1/+75
|
* Automated regression test harness for libc_r. Existing tests are integrated,jasone2000-04-2415-3/+3356
| | | | | a new test for POSIX semaphores was added, and examples of harness usage are included.
* Introduce .Lb macro to libc_r manpages.phantom2000-04-2241-0/+82
|
* Explicitly include sys/cdefs.h to get the definition of __strong_reference(),jasone2000-03-181-0/+1
| | | | rather than getting lucky due to header dependencies.
* Fix pthread_suspend_np/pthread_resume_np. For the record, suspending adeischen2000-03-159-31/+125
| | | | | | | | | | | | | | thread waiting on an event (I/O, condvar, etc) will, when resumed using pthread_resume_np, return with EINTR. For example, suspending and resuming a thread blocked on read() will not requeue the thread for the read, but will return -1 with errno = EINTR. If the suspended thread is in a critical region, the thread is suspended as soon as it leaves the critical region. Fix a bogon in pthread_kill() where a signal was being delivered twice to threads waiting in sigwait(). Reported by (suspend/resume bug): jdp Reviewed by: jasone
* Remove single-space hard sentence breaks. These degrade the qualitysheldonh2000-03-0213-55/+110
| | | | | of the typeset output, tend to make diffs harder to read and provide bad examples for new-comers to mdoc.
* Fix the installworld breakage from rev 1.9.ru2000-02-181-5/+4
| | | | | | | Submitted by: reg Reviewed by: ru Approved by: jasone (rev 1.9) Pointy hat to: unfurl
* fixed missing trailing \unfurl2000-02-181-1/+1
| | | | Submitted by: Thimble Smith <tim@mysql.com>
* Adds missing symlinks for pthread related manpages.unfurl2000-02-181-0/+3
| | | | | | PR: 16537 Submitted by: AnarCat <beaupran@iro.umontreal.ca> Approved by: jasone
* For errors, return -1 and set errno to indicate the error type, rather thanjasone2000-02-161-16/+33
| | | | | | | | | returning the error directly. For sem_post(), make sure that the correct thread is woken up. This has unfortunate performance implications, but is necessary for POSIX compliance. Approved by: jkh
* Add man pages for the sem_*() functions.jasone2000-02-167-4/+516
| | | | Approved by: jkh
* Use __strong_reference() instead of __weak_reference() to assure that thejasone2000-01-2911-11/+11
| | | | weak symbols of the same name are not used.
* Use _fcntl() (not fcntl()) inside of fcntl().jasone2000-01-281-3/+3
| | | | Reported by: green
* Simplify sytem call renaming. Instead of _foo() <-- _libc_foo <-- foo(),jasone2000-01-2733-81/+571
| | | | | | | | | | | | | | | | | just use _foo() <-- foo(). In the case of a libpthread that doesn't do call conversion (such as linuxthreads and our upcoming libpthread), this is adequate. In the case of libc_r, we still need three names, which are now _thread_sys_foo() <-- _foo() <-- foo(). Convert all internal libc usage of: aio_suspend(), close(), fsync(), msync(), nanosleep(), open(), fcntl(), read(), and write() to _foo() instead of foo(). Remove all internal libc usage of: creat(), pause(), sleep(), system(), tcdrain(), wait(), and waitpid(). Make thread cancellation fully POSIX-compliant. Suggested by: deischen
* Fix millisecond to nanosecond conversion.jasone2000-01-221-1/+1
| | | | PR: misc/16245
* Minor *jmp() cleanups.jasone2000-01-201-2/+3
|
* Add sem_*() functions. Named semaphores and process-shared semaphoresjasone2000-01-203-0/+237
| | | | | | | are not supported by this implementation, and the error return values from sem_init(), sem_open(), sem_close(), and sem_unlink() reflect this. Approved by: jkh
* Do signal deferral for pthread_kill() as it was done in the old days.jasone2000-01-203-30/+6
| | | | Submitted by: deischen
* Implement continuations to correctly handle [sig|_]longjmp() inside of ajasone2000-01-1912-61/+374
| | | | | | | | | | | | | | | | | | | | | | signal handler. Explicitly check for jumps to anywhere other than the current stack, since such jumps are undefined according to POSIX. While we're at it, convert thread cancellation to use continuations, since it's cleaner than the original cancellation code. Avoid delivering a signal to a thread twice. This was a pre-existing bug, but was likely unexposed until these other changes were made. Defer signals generated by pthread_kill() so that they can be delivered on the appropriate stack. deischen claims that this is unnecessary, which is likely true, but without this change, pthread_kill() can cause undefined priority queue states and/or PANICs in [sig|_]longjmp(), so I'm leaving this in for now. To compile this code out and exercise the bug, define the _NO_UNDISPATCH cpp macro. Defining _PTHREADS_INVARIANTS as well will cause earlier crashes. PR: kern/14685 Collaboration with: deischen
* Properly initialize the last active time of the initial thread. This fixesdeischen2000-01-181-0/+6
| | | | | | | the case that a CPU hungry main thread is prevented from being preempted due to a negative calculation of its time slice. Reported by: Alexander Litvin <archer@lucky.net>
* Track libc's three-tier symbol naming. libc_r must currently implementjasone2000-01-1226-32/+75
| | | | | the _libc_*() entry points and add *() weak aliases. This will all change for the better when libc_r becomes libpthread.
* Make sched_param parameter a const to comply with POSIX and SUSv2 specs.deischen2000-01-101-1/+1
| | | | | | | This doesn't need to be applied to stable, because somehow -stable seems to have gotten it right. Reviewed by: jasone
* Don't explicitly mmap() red zones at the bottom of thread stacks (exceptjasone1999-12-293-16/+26
| | | | | | | | | | the initial thread). Instead, just leave an unmapped gap between thread stacks and make sure that the thread stacks won't grow into these gaps, simply by limiting the size of the stacks with the 'len' argument to mmap(). This (if I understand correctly) reduces VM overhead considerably. Reviewed by: deischen
OpenPOWER on IntegriCloud