- Verbesserter Scheduler der nicht nach Round-Robin die verfügbaren Tasks aktiviert, sondern über Priorisierungs möglichkeiten und verdrängung verfügt:
`For an full example (from a previous micro-kernel of mine), I used:
Policy 0: Intended for low latency tasks (e.g. device driver IRQ handlers). Uses a "highest priority task that can run does run" scheduling algorithm, and tasks within this policy have a priority (0 to 255). These tasks will preempt all tasks in lower policies and all tasks with lower priority in the same policy.
Policy 1: Intended for tasks with some latency requirements (e.g. user interfaces, GUIs, etc). Uses a "highest priority task that can run does run" scheduling algorithm, and tasks within this policy have a priority (0 to 255). These tasks will preempt all tasks in lower policies and all tasks with lower priority in the same policy.
Policy 2: Intended for normal tasks. Uses a "variable frequency, fixed time slice" scheduling algorithm. Tasks within this policy have a priority (0 to 255) that determines how many (fixed length) time slices the task gets. Tasks will preempt all tasks in lower policies but will not preempt tasks in the same policy.
Policy 3: Intended for background tasks. Uses a "variable time slice" scheduling algorithm, and tasks have a priority (0 to 255) that determines the length of the time slice they get. These tasks never preempt anything.`
- Anlage von Processen/Tasks entsprechend erweitern, dass die Stufen entsprechend initial gesetzt werden
- Noch keine Dynamische Anpassung von Prioritäten, wenn ein Task schon läuft!?!
`For an full example (from a previous micro-kernel of mine), I used:
Policy 0: Intended for low latency tasks (e.g. device driver IRQ handlers). Uses a "highest priority task that can run does run" scheduling algorithm, and tasks within this policy have a priority (0 to 255). These tasks will preempt all tasks in lower policies and all tasks with lower priority in the same policy.
Policy 1: Intended for tasks with some latency requirements (e.g. user interfaces, GUIs, etc). Uses a "highest priority task that can run does run" scheduling algorithm, and tasks within this policy have a priority (0 to 255). These tasks will preempt all tasks in lower policies and all tasks with lower priority in the same policy.
Policy 2: Intended for normal tasks. Uses a "variable frequency, fixed time slice" scheduling algorithm. Tasks within this policy have a priority (0 to 255) that determines how many (fixed length) time slices the task gets. Tasks will preempt all tasks in lower policies but will not preempt tasks in the same policy.
Policy 3: Intended for background tasks. Uses a "variable time slice" scheduling algorithm, and tasks have a priority (0 to 255) that determines the length of the time slice they get. These tasks never preempt anything.`