Method and apparatus for thread scheduling
A method is provided for use in computing system, comprising: instantiating a plurality of scheduler queues; instantiating a plurality of type-1 threads; assigning each of the plurality of type-1 threads to a respective scheduler queue; allocating a plurality of budgets to the plurality of scheduling of queues, each of the plurality of budgets being allocated to a different one of the plurality of scheduling queues; selecting one of the plurality of scheduling queues; retrieving a type-1 thread from the selected scheduling queue when one of a first condition or a second condition is satisfied, the first condition being satisfied when the respective budget of the queue is not exhausted, and the second condition being satisfied when another one of the plurality of scheduling queues are empty; and executing the retrieved type-1 thread.
1. A method for use in a computing system, comprising:
instantiating a plurality of scheduling queues;
instantiating a plurality of type-1 threads;
assigning each of the plurality of type-1 threads to a respective scheduling queue;
allocating a plurality of budgets to the plurality of scheduling of queues, each of the plurality of budgets being allocated to a different one of the plurality of scheduling queues, the plurality of budgets being allocated for a duration of a scheduling cycle;
selecting one of the plurality of scheduling queues;
retrieving a type-1 thread from the selected scheduling queue when one of a first condition or a second condition is satisfied, the first condition being satisfied when the respective budget of the queue is not exhausted, and the second condition being satisfied when another one of the plurality of scheduling queues is empty;
executing the retrieved type-1 thread; and
executing one or more type-2 threads after the scheduling cycle is completed and before a next scheduling cycle has begun.
2. The method of claim 1 , wherein the second condition is determined to be satisfied further upon confirming that the budget of the selected scheduling queue is exhausted.
3. The method of claim 1 , wherein the second condition is determined to be satisfied further upon confirming that all scheduling queues in the plurality of scheduling queues, which have a higher priority than the selected scheduling queue, are empty.
4. The method of claim 1 , wherein the respective budget of any of the scheduling queues specifies a maximum amount of time for which threads in that scheduling queue can be executed.
5. The method of claim 1 , wherein assigning each of the plurality of threads to a respective scheduling queue includes:
retrieving a queue identifier from the thread;
selecting one of the plurality of scheduling queues based on the retrieved identifier; and
adding the thread to the selected queue.
6. The method of claim 1 , wherein, wherein the selected scheduling queue is non-empty, and the selected scheduling queue has a highest priority among all non-empty scheduling queues in the plurality of scheduling queues.
7. A system comprising:
a memory; and
at least one processor operatively coupled to the memory, the at least one processor being configured to perform the operations of:
instantiating a plurality of scheduling queues;
instantiating a plurality of type-1 threads;
assigning each of the plurality of type-1 threads to a respective scheduling queue;
allocating a plurality of budgets to the plurality of scheduling of queues, each of the plurality of budgets being allocated to a different one of the plurality of scheduling queues, the plurality of budgets being allocated for a duration of a scheduling cycle;
selecting one of the plurality of scheduling queues;
retrieving a type-1 thread from the selected scheduling queue when one of a first condition or a second condition is satisfied, the first condition being satisfied when the respective budget of the queue is not exhausted, and the second condition being satisfied when another one of the plurality of scheduling queues is empty; and executing the retrieved type-1 thread; and
executing one or more type-2 threads after the scheduling cycle is completed and before a next scheduling cycle has begun.
8. The system of claim 7 , wherein the second condition is satisfied when the budget of the selected scheduling queue is exhausted.
9. The system of claim 7 , wherein the second condition is satisfied when all scheduling queues in the plurality of scheduling queues, which have a higher priority than the selected scheduling queue, are empty.
10. The system of claim 7 , wherein the respective budget of any of the scheduling queues specifies a maximum amount of time for which threads in that scheduling queue can be executed.
11. The system of claim 7 , wherein assigning each of the plurality of threads to a respective scheduling queue includes:
retrieving a queue identifier from the thread;
selecting one of the plurality of scheduling queues based on the retrieved identifier; and
adding the thread to the selected queue.
12. The system of claim 7 , wherein the selected scheduling queue is non-empty, and the selected scheduling queue has a highest priority among all non-empty scheduling queues in the plurality of scheduling queues.
13. A non-transitory computer-readable medium storing processor-executable instructions, which when executed by at least one processor cause the at least one processor to perform the operations of:
instantiating a plurality of scheduling queues;
instantiating a plurality of type-1 threads;
assigning each of the plurality of type-1 threads to a respective scheduling queue;
allocating a plurality of budgets to the plurality of scheduling of queues, each of the plurality of budgets being allocated to a different one of the plurality of scheduling queues, the plurality of budgets being allocated for a duration of a scheduling cycle;
selecting one of the plurality of scheduling queues;
retrieving a type-1 thread from the selected scheduling queue when one of a first condition or a second condition is satisfied, the first condition being satisfied when the respective budget of the queue is not exhausted, and the second condition being satisfied when another one of the plurality of scheduling queues is empty;
executing the retrieved type-1 thread; and
executing one or more type-2 threads after the scheduling cycle is completed and before a next scheduling cycle has begun.
14. The non-transitory computer-readable medium of claim 13 , wherein the second condition is satisfied when the budget of the selected scheduling queue is exhausted.
15. The non-transitory computer-readable medium of claim 13 , wherein the second condition is satisfied when all scheduling queues in the plurality of scheduling queues, which have a higher priority than the selected scheduling queue, are empty.
16. The non-transitory computer-readable medium of claim 13 , wherein the respective budget of any of the scheduling queues specifies a maximum amount of time for which threads in that scheduling queue can be executed.
17. The non-transitory computer-readable medium of claim 13 , wherein assigning each of the plurality of threads to a respective scheduling queue includes:
retrieving a queue identifier from the thread;
selecting one of the plurality of scheduling queues based on the retrieved identifier; and
adding the thread to the selected queue.