summaryrefslogtreecommitdiffstats
path: root/sys/kern/kern_umtx.c
Commit message (Expand)AuthorAgeFilesLines
* Further system call comment cleanup:rwatson2007-03-051-1/+0
* Add a lwpid field into per-cpu structure, the lwpid represents currentdavidxu2006-12-201-0/+70
* Threading cleanup.. part 2 of several.julian2006-12-061-6/+0
* if a thread blocked on userland condition variable isdavidxu2006-12-041-5/+10
* Introduce userspace condition variable, since we have already POSIXdavidxu2006-12-031-4/+200
* Sweep kernel replacing suser(9) calls with priv(9) calls, assigningrwatson2006-11-061-2/+3
* Make KSE a kernel option, turned on by default in all GENERICjb2006-10-261-0/+6
* Optimize umtx_lock_pi() a bit by moving some heavy code out of the loop,davidxu2006-10-261-32/+27
* In order to eliminate a branch, convert opcode to unsigned integer.davidxu2006-10-251-2/+2
* Eliminate an unnecessary `if' statement.davidxu2006-10-251-1/+2
* o Add keyword volatile for user mutex owner field.davidxu2006-10-171-29/+23
* Implement 32bit umtx_lock and umtx_unlock system calls, these two systemdavidxu2006-10-061-0/+14
* Fix umtx command order error for freebsd 32bit.davidxu2006-09-221-1/+1
* Add umtx support for 32bit process on AMD64 machine.davidxu2006-09-221-82/+440
* Merge all code of do_lock_normal, do_lock_pi and do_lock_pp intodavidxu2006-09-051-120/+52
* Check if it is root user in do_unlock_pp.davidxu2006-09-031-2/+4
* Make sure we get new m_owner value if we can not unlock it indavidxu2006-09-021-8/+12
* Reorder some statments. Fix typo and remove stale comments.davidxu2006-08-301-20/+16
* Update comments about interrupted mutex locking.davidxu2006-08-281-12/+5
* This is initial version of POSIX priority mutex support, a new userlanddavidxu2006-08-281-187/+1710
* Add user priority loaning code to support priority propagation fordavidxu2006-08-251-0/+5
* Move flag TDF_UMTXQ into structure umtxq, this eliminates the requirementdavidxu2006-05-181-15/+12
* Use wakeup_one to avoid thundering herd.davidxu2006-05-091-6/+6
* Change msleep() and tsleep() to not alter the calling thread's priorityjhb2006-04-171-6/+4
* Axe unused code.davidxu2006-02-041-96/+0
* do umtx_wake at userland thread exit address, so that others userlanddavidxu2005-10-261-3/+3
* Allocate umtx_q from heap instead of stack, this avoidsdavidxu2005-03-051-34/+46
* Revert my previous errno hack, that is certainly an issue,davidxu2005-01-181-2/+1
* make umtx timeout relative so userland can select different clock type,davidxu2005-01-141-46/+51
* Comment out debugging printf which doesn't compile on amd64.phk2005-01-121-0/+2
* Let _umtx_op directly return error code rather than from errno becausedavidxu2005-01-121-12/+23
* Break out of loop earlier if it is not timeout.davidxu2005-01-081-1/+1
* /* -> /*- for copyright notices, minor format tweaks as necessaryimp2005-01-061-1/+1
* Return ETIMEDOUT when thread is timeouted since POSIX threaddavidxu2005-01-061-5/+7
* Make umtx_wait and umtx_wake more like linux futex does, it isdavidxu2004-12-301-41/+9
* Make _umtx_op() as more general interface, the final parameter needn't bedavidxu2004-12-251-4/+4
* 1. introduce umtx_owner to get an owner of a umtx.davidxu2004-12-251-3/+1
* Add umtxq_lock/unlock around umtx_signal, fix debug kernel compiling,davidxu2004-12-241-5/+9
* 1. Fix race condition between umtx lock and unlock, heavy testingdavidxu2004-12-241-133/+104
* 1. msleep returns EWOULDBLOCK not ETIMEDOUT, use EWOULDBLOCK instead.davidxu2004-12-181-8/+6
* 1. make umtx sharable between processes, the way is two or more processesdavidxu2004-12-181-170/+544
* Forgot to inline umtxq_unlock.davidxu2004-11-301-1/+1
* 1. use per-chain mutex instead of global mutex to reducedavidxu2004-11-301-115/+212
* writers must hold both sched_lock and the process lock; therefore, readersmtm2004-07-121-5/+3
* Change the thread ID (thr_id_t) used for 1:1 threading from being amarcel2004-07-021-3/+5
* Use the proc lock to sleep on a libthr umtx.mtm2004-03-271-2/+6
* Return EINVAL if the contested bit is not set on the umtx passed totjr2003-09-071-1/+2
* Initialize 'blocked' to NULL. I think this was a real problem, but Ipeter2003-07-231-0/+1
* Turn a KASSERT back into an EINVAL return value. So, next time someonemtm2003-07-191-2/+4
* Remove a lock held across casuptr() that snuck in last commit.mtm2003-07-181-2/+5
OpenPOWER on IntegriCloud