summaryrefslogtreecommitdiffstats
path: root/testsuites/smptests/smpscheduler02 (follow)
Commit message (Collapse)AuthorAgeFilesLines
* rtems: Require RTEMS_PRIORITY for MrsP semaphoresSebastian Huber2021-03-161-1/+2
| | | | | | | | | MrsP semaphores are a generalization of the priority ceiling semaphores for SMP configurations. Priority ceiling semaphores are required to use the priority task wait queue discipline. Require this discipline also for MrsP semaphores. Close #4347.
* Canonicalize config.h includeSebastian Huber2020-04-161-1/+1
| | | | | | | | Use the following variant which was already used by most source files: #ifdef HAVE_CONFIG_H #include "config.h" #endif
* config: Improve EDF SMP scheduler configurationSebastian Huber2019-12-191-1/+1
| | | | | | | | Use CONFIGURE_MAXIMUM_PROCESSORS to configure the EDF SMP scheduler context. This avoids hard to debug configuration errors resulting in memory corruptions. Close #3815.
* rtems: Simplify semaphore configurationSebastian Huber2019-12-111-1/+0
| | | | | | | | | | | | | | | The MrsP semaphore implementation predates the addition of self-contained synchronization objects. At this time, the potential memory reduction was justified considering the more complex configuration and additional use of the workspace. With the availability of self-contained synchronization options, e.g. POSIX mutexes, this is no longer justified. Memory constrained applications should use the self-contained synchronization objects. Remove the CONFIGURE_MAXIMUM_MRSP_SEMAPHORES configuration option. This has only an impact on applications which use SMP and a large number of scheduler instances. Update #3833.
* rtems: Add rtems_scheduler_get_processor_maximum()Sebastian Huber2019-04-091-2/+2
| | | | | | | | | | | Add rtems_scheduler_get_processor_maximum() as a replacement for rtems_get_processor_count(). The rtems_get_processor_count() is a bit orphaned. Adopt it by the Scheduler Manager. The count is also misleading, since the processor set may have gaps and the actual count of online processors may be less than the value returned by rtems_get_processor_count(). Update #3732.
* rtems: Add rtems_scheduler_get_processor()Sebastian Huber2019-04-091-7/+7
| | | | | | | | | | | Add rtems_scheduler_get_processor() as a replacement for rtems_get_current_processor(). The rtems_get_current_processor() is a bit orphaned. Adopt it by the Scheduler Manager. This is in line with the glibc sched_getcpu() function. Deprecate rtems_get_current_processor(). Update #3731.
* testsuite/smptests: Merged nested Makefile.am files into one Makefile.amChris Johns2018-04-101-19/+0
| | | | | | This change is part of the testsuite Makefile.am reorganization. Update #3382
* config: Use new scheduler configuration definesSebastian Huber2018-03-121-8/+8
| | | | Update #3325.
* tests: Use simple console driverSebastian Huber2017-11-061-1/+1
| | | | | Update #3170. Update #3199.
* smpscheduler02: Avoid sporadic test failuresSebastian Huber2017-11-031-0/+4
|
* tests: Remove TEST_INITSebastian Huber2017-10-281-2/+0
| | | | | | | | The TEST_EXTERN is a used only by the system.h style tests and they use CONFIGURE_INIT appropriately. Update #3170. Update #3199.
* testsuite: Use printk for all test output where possible.Chris Johns2017-10-231-0/+2
| | | | | | | | | | - Remove the printf support leaving the direct printk support configured with TESTS_USE_PRINTK and all other output goes via a buffered vsniprintf call to printk. - Control the test's single init for functions and global data with TEST_INIT and not CONFIGURE_INIT. They are now separate. Updates #3170.
* Include missing <limits.h>Sebastian Huber2017-08-251-0/+1
| | | | Update #2132.
* smptests/smpscheduler02: Remove invalid assertSebastian Huber2017-07-191-3/+1
| | | | Update #3059.
* rtems: Add rtems_scheduler_ident_by_processor_setSebastian Huber2017-07-111-0/+11
| | | | Update #3070.
* score: Fix set schedulerSebastian Huber2017-07-101-6/+10
| | | | | | | Ensure that the thread processor affinity fits the new scheduler instance. Update #3059.
* score: Fix default set affinitySebastian Huber2017-07-071-3/+4
| | | | | | | | The set of online processors must be a subset of the thread processor affinity for the schedulers without arbitrary processor affinity support to avoid problems in case of processor addition and removal. Update #3059.
* score: Move processor affinity to Thread_ControlSebastian Huber2017-07-071-5/+14
| | | | Update #3059.
* Rename CONFIGURE_SMP_MAXIMUM_PROCESSORSSebastian Huber2017-02-141-1/+1
| | | | | | | Rename CONFIGURE_SMP_MAXIMUM_PROCESSORS to CONFIGURE_MAXIMUM_PROCESSORS since the SMP part is superfluous. Update #2894.
* Remove CONFIGURE_SMP_APPLICATIONSebastian Huber2017-02-021-2/+0
| | | | | | Enable the SMP support if CONFIGURE_SMP_MAXIMUM_PROCESSORS > 1. Update #2893.
* smptests/smpscheduler02: Fix task initializationSebastian Huber2016-11-231-2/+2
| | | | Set scheduler before the task start.
* rtems: Add scheduler processor add/removeSebastian Huber2016-11-101-0/+130
| | | | Update #2797.
* score: Simplify _Scheduler_Get_by_id()Sebastian Huber2016-11-091-4/+7
| | | | | | | | | | Avoid dead code in non-SMP configurations. Return scheduler identifier independent of the current processor count of the scheduler via rtems_scheduler_ident(), since this value may change during run-time. Check the processor count in _Scheduler_Set() under scheduler lock protection. Update #2797.
* score: Relax _Scheduler_Set() restrictionsSebastian Huber2016-11-031-18/+42
| | | | | | No longer unconditionally prevent scheduler changes if the thread owns resources. Prevent a scheduler change only in case other threads wait for the resource.
* rtems: Fix rtems_task_set_scheduler() APISebastian Huber2016-07-011-3/+3
| | | | | | | | | | | Task priorities are only valid within a scheduler instance. The rtems_task_set_scheduler() directive moves a task from one scheduler instance to another using the current priority of the thread. However, the current task priority of the source scheduler instance is undefined in the target scheduler instance. Add a third parameter to specify the priority. Close #2749.
* score: Introduce map priority scheduler operationSebastian Huber2016-06-221-2/+33
| | | | | | | | | | | Introduce map/unmap priority scheduler operations to map thread priority values from/to the user domain to/from the scheduler domain. Use the map priority operation to validate the thread priority. The EDF schedulers use this new operation to distinguish between normal priorities and priorities obtain through a job release. Update #2173. Update #2556.
* posix: sched_get_priority_max()Sebastian Huber2016-06-141-4/+4
| | | | | | Enable for all configurations since it pulls in no additional dependencies. Return value of the scheduler instance of the executing thread.
* posix: sched_get_priority_min()Sebastian Huber2016-06-141-1/+6
| | | | | Enable for all configurations since it pulls in no additional dependencies.
* smptests/smpscheduler02: Reduce required CPU countSebastian Huber2015-06-031-41/+50
|
* tests: Add documentationSebastian Huber2014-09-011-0/+12
|
* score: _Scheduler_Set_affinity()Sebastian Huber2014-06-021-3/+3
| | | | | Do not change the scheduler with this function. Documentation. Coding style.
* score: Implement forced thread migrationSebastian Huber2014-05-071-2/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | The current implementation of task migration in RTEMS has some implications with respect to the interrupt latency. It is crucial to preserve the system invariant that a task can execute on at most one processor in the system at a time. This is accomplished with a boolean indicator in the task context. The processor architecture specific low-level task context switch code will mark that a task context is no longer executing and waits that the heir context stopped execution before it restores the heir context and resumes execution of the heir task. So there is one point in time in which a processor is without a task. This is essential to avoid cyclic dependencies in case multiple tasks migrate at once. Otherwise some supervising entity is necessary to prevent life-locks. Such a global supervisor would lead to scalability problems so this approach is not used. Currently the thread dispatch is performed with interrupts disabled. So in case the heir task is currently executing on another processor then this prolongs the time of disabled interrupts since one processor has to wait for another processor to make progress. It is difficult to avoid this issue with the interrupt latency since interrupts normally store the context of the interrupted task on its stack. In case a task is marked as not executing we must not use its task stack to store such an interrupt context. We cannot use the heir stack before it stopped execution on another processor. So if we enable interrupts during this transition we have to provide an alternative task independent stack for this time frame. This issue needs further investigation.
* score: SMP_FATAL_SCHEDULER_WITHOUT_PROCESSORSSebastian Huber2014-05-051-7/+51
| | | | | Avoid the SMP_FATAL_SCHEDULER_WITHOUT_PROCESSORS fatal error and make it a run-time error in rtems_scheduler_ident() and _Scheduler_Get_by_id().
* score: Add clustered/partitioned schedulingSebastian Huber2014-04-154-0/+255
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.
generated by cgit v1.2.3 (git 2.25.1) at 2024-05-05 07:28:43 +0000