diff options
author | Sebastian Huber <sebastian.huber@embedded-brains.de> | 2014-04-09 15:07:54 +0200 |
---|---|---|
committer | Sebastian Huber <sebastian.huber@embedded-brains.de> | 2014-04-15 10:41:44 +0200 |
commit | c5831a3f9af11228dbdaabaf01f69d37e55684ef (patch) | |
tree | e9ddedd942f3f31d239820dfc4dbcef4cde0b09a /cpukit/score/src/threadstart.c | |
parent | rtems: Add task get/set scheduler (diff) | |
download | rtems-c5831a3f9af11228dbdaabaf01f69d37e55684ef.tar.bz2 |
score: Add clustered/partitioned scheduling
Clustered/partitioned scheduling helps to control the worst-case
latencies in the system. The goal is to reduce the amount of shared
state in the system and thus prevention of lock contention. Modern
multi-processor systems tend to have several layers of data and
instruction caches. With clustered/partitioned scheduling it is
possible to honour the cache topology of a system and thus avoid
expensive cache synchronization traffic.
We have clustered scheduling in case the set of processors of a system
is partitioned into non-empty pairwise-disjoint subsets. These subsets
are called clusters. Clusters with a cardinality of one are partitions.
Each cluster is owned by exactly one scheduler instance.
Diffstat (limited to 'cpukit/score/src/threadstart.c')
-rw-r--r-- | cpukit/score/src/threadstart.c | 8 |
1 files changed, 6 insertions, 2 deletions
diff --git a/cpukit/score/src/threadstart.c b/cpukit/score/src/threadstart.c index b65a2c3380..dda9495a35 100644 --- a/cpukit/score/src/threadstart.c +++ b/cpukit/score/src/threadstart.c @@ -46,8 +46,12 @@ bool _Thread_Start( if ( cpu == NULL ) { _Thread_Ready( the_thread ); } else { - the_thread->current_state = STATES_READY; - _Scheduler_Start_idle( _Scheduler_Get( the_thread ), the_thread, cpu ); + const Scheduler_Control *scheduler = _Scheduler_Get_by_CPU( cpu ); + + if ( scheduler != NULL ) { + the_thread->current_state = STATES_READY; + _Scheduler_Start_idle( scheduler, the_thread, cpu ); + } } _User_extensions_Thread_start( the_thread ); |