| Commit message (Expand) | Author | Age | Files | Lines |
|---|
| * | Revert to using the SCHED_LOCK() to protect time accounting. |   mpi | 2019-06-01 | 1 | -5/+1 |
| * | Use a per-process mutex to protect time accounting instead of SCHED_LOCK(). | mpi | 2019-05-31 | 1 | -1/+5 |
| * | Make sure that each ci has its spc_deferred queue initialized. |   visa | 2019-03-26 | 1 | -3/+2 |
| * | Introduce safe memory reclamation, a mechanism for reclaiming shared | visa | 2019-02-26 | 1 | -1/+6 |
| * | Add new KERN_CPUSTATS sysctl(2) so we can identify offline CPUs. |   cheloha | 2018-11-17 | 1 | -1/+7 |
| * | Revert KERN_CPTIME2 ENODEV changes in kernel and userspace. | cheloha | 2018-10-05 | 1 | -7/+1 |
| * | KERN_CPTIME2: set ENODEV if the CPU is offline. | cheloha | 2018-09-26 | 1 | -1/+7 |
| * | Add hw.ncpuonline to count the number of online CPUs. | cheloha | 2018-07-12 | 1 | -1/+21 |
| * | Release the kernel lock fully on thread exit. This prevents a locking | visa | 2018-07-07 | 1 | -2/+5 |
| * | Don't steal processes from other CPUs if we're not scheduling processes on |   kettenis | 2018-06-30 | 1 | -1/+5 |
| * | SMT (Simultanious Multi Threading) implementations typically share | kettenis | 2018-06-19 | 1 | -2/+52 |
| * | make sched_barrier use cond_wait/cond_signal. |   dlg | 2017-12-14 | 1 | -20/+16 |
| * | Raise the IPL of the sbar taskq to avoid lock order issues | visa | 2017-11-28 | 1 | -2/+2 |
| * | Split up fork1(): |   guenther | 2017-02-12 | 1 | -2/+2 |
| * | p_comm is the process's command and isn't per thread, so move it from | guenther | 2017-01-21 | 1 | -2/+3 |
| * | Allow pegged process on secondary CPUs to continue to be scheduled when | kettenis | 2016-06-03 | 1 | -2/+5 |
| * | Replace curcpu_is_idle() by cpu_is_idle() and use it instead of rolling | mpi | 2016-03-17 | 1 | -2/+2 |
| * | One "sbar" taskq is enough. | kettenis | 2015-12-23 | 1 | -8/+7 |
| * | Make the cost of moving a process to the primary cpu a bit higher. This is | kettenis | 2015-12-17 | 1 | -1/+10 |
| * | Make sched_barrier() use its own task queue to avoid deadlocks. | mpi | 2015-10-16 | 1 | -2/+13 |
| * | Short circuit if we're running on the CPU that we want to sync with. Fixes | kettenis | 2015-09-20 | 1 | -2/+5 |
| * | Introduce sched_barrier(9), an interface that acts as a scheduler barrier in | kettenis | 2015-09-13 | 1 | -1/+46 |
| * | Remove some includes include-what-you-use claims don't |   jsg | 2015-03-14 | 1 | -4/+1 |
| * | Keep under #ifdef MULTIPROCESSOR the code that deals with SPCF_SHOULDHALT | mpi | 2014-09-24 | 1 | -1/+5 |
| * | If we're stopping a secondary cpu, don't let sched_choosecpu() short-circuit | kettenis | 2014-07-26 | 1 | -1/+3 |
| * | Fix sched_stop_secondary_cpus() to properly drain CPUs |   matthew | 2014-07-13 | 1 | -2/+2 |
| * | Add PS_SYSTEM, the process-level mirror of the thread-level P_SYSTEM, | guenther | 2014-05-04 | 1 | -7/+2 |
| * | Eliminate the exit sig handling, which was only invokable via the | guenther | 2014-02-12 | 1 | -2/+2 |
| * | Prevent idle thread from being stolen on startup. |   haesbaert | 2013-06-06 | 1 | -2/+13 |
| * | Convert some internal APIs to use timespecs instead of timevals | guenther | 2013-06-03 | 1 | -5/+5 |
| * | sprinkle ifdef MP to disable cpu migration code when not needed. |   tedu | 2013-04-19 | 1 | -7/+17 |
| * | Make sure that we don't schedule processes on CPUs that we havetaken out of | kettenis | 2012-07-10 | 1 | -1/+5 |
| * | Make rusage totals, itimers, and profile settings per-process instead | guenther | 2012-03-23 | 1 | -2/+2 |
| * | Account for sched_noidle and document the scheduler variables. | haesbaert | 2012-03-10 | 1 | -11/+13 |
| * | Remove all MD diagnostics in cpu_switchto(), and move them to MI code if |   miod | 2011-10-12 | 1 | -1/+4 |
| * | Clean up after P_BIGLOCK removal. |   art | 2011-07-06 | 1 | -3/+3 |
| * | Delete a fallback definition for CPU_INFO_UNIT that's both unnecessary | guenther | 2010-05-28 | 1 | -8/+1 |
| * | Actively remove processes from the runqueues of a CPU when we stop it. | kettenis | 2010-05-25 | 1 | -5/+22 |
| * | Make sure we initialize sched_lock before we try to use it. | kettenis | 2010-05-14 | 1 | -4/+1 |
| * | Merge the only relevant (for now) parts of simplelock.h into lock.h |   deraadt | 2010-04-23 | 1 | -2/+1 |
| * | Implement functions to take away the secondary CPUs from the scheduler and | kettenis | 2010-04-06 | 1 | -2/+51 |
| * | Add code to stop scheduling processes on CPUs, effectively halting that CPU. | kettenis | 2010-01-09 | 1 | -2/+15 |
| * | Backout previous commit. There is a possible race which makes it possible | kettenis | 2009-11-29 | 1 | -13/+1 |
| * | Add a mechanism to stop the scheduler from scheduling processes on a | kettenis | 2009-11-25 | 1 | -1/+13 |
| * | Don't drop the big lock at the end of exit1(), but move it into the middle of | deraadt | 2009-10-05 | 1 | -4/+3 |
| * | When starting up idle, explicitly set p_cpu and the peg flag for the | art | 2009-04-22 | 1 | -1/+3 |
| * | Make pegging a proc work when there are idle cpus that are looking for | art | 2009-04-20 | 1 | -7/+12 |
| * | Some tweaks to the cpu affinity code. | art | 2009-04-14 | 1 | -56/+103 |
| * | sched_peg_curproc_to_cpu() - function to force a proc to stay on a cpu | art | 2009-04-03 | 1 | -1/+27 |
| * | Processor affinity for processes. | art | 2009-03-23 | 1 | -35/+344 |
