Scheduling of Execution Units

• US 20080168447A1
• Filed: 01/09/2007
• Published: 07/10/2008
• Est. Priority Date: 01/09/2007
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Lewis, Russell Lee
• Granted Patent: US 8,024,739 B2
• Time in Patent Office: Days
• US Class Current: 718/102
• CPC Class Codes: G06F 9/461 Saving or restoring of program or task context
• CPC Class Codes: G06F 9/485 Task life-cycle, e.g. stopping, restarting, resuming execution G06F9/4881 takes precedence

Abstract

Provided are a method, system, and article of manufacture, wherein an execution unit requests additional execution time from a kernel to prevent the execution unit from being swapped out during an execution of a critical section. The kernel determines whether the execution unit has previously avoided yielding execution to other execution units within a predetermined period of time after being allocated the additional execution time in response to at least one previous request for the additional execution time. The kernel allocates the additional execution time to the execution unit, in response to determining that the execution unit has not previously avoided yielding execution to the other execution units within the predetermined period of time after being allocated the additional execution time in response to the at least one previous request for the additional execution time.

Claims

1. A method, comprising:

requesting, by an execution unit, additional execution time from a kernel to prevent the execution unit from being swapped out during an execution of a critical section;

determining, by the kernel, whether the execution unit has previously avoided yielding execution to other execution units within a predetermined period of time after being allocated the additional execution time in response to at least one previous request for the additional execution time; and

allocating, by the kernel, the additional execution time to the execution unit, in response to determining that the execution unit has not previously avoided yielding execution to the other execution units within the predetermined period of time after being allocated the additional execution time in response to the at least one previous request for the additional execution time.

2. The method of claim 1, further comprising:

maintaining a first indicator in the kernel, wherein the first indicator indicates whether the execution unit has previously avoided yielding execution to the other execution units within the predetermined period of time after being allocated the additional execution time in response to the at least one previous request for the additional execution time;

maintaining at least one second indicator that indicates whether the additional time has been granted by the kernel to the execution unit; and

maintaining a third indicator that indicates whether the additional time has been requested by the execution unit from the kernel.

3. The method of claim 2, the method further comprising:

updating, by the execution unit, the third indicator to indicate that the additional time is longer needed by the execution unit, in response to the execution unit exiting from the critical section; and

swapping in at least one other execution unit, by the kernel, in response to the updating by the execution unit of third indicator to indicate that the additional time is longer needed by the execution unit.

4. The method of claim 1, wherein the execution unit is a first thread, the method further comprising:

determining, by the kernel, that a second thread that is to be swapped in to replace the first thread is part of the same process as the first thread; and

allocating, by the kernel, the additional execution time to the first thread in response to determining that the second thread that is to be swapped in to replace the first thread is part of the same process as the first thread whether or not the process has previously avoided yielding execution for other processes within the predetermined period of time.

5. The method of claim 1, wherein the execution unit is one of a process or a thread, wherein the kernel is part of an operating system, and wherein the execution unit is not part of the kernel, and wherein the requesting, the determining, and the allocating are performed via system calls.

6. A system, comprising:

memory; and

processor coupled to the memory, wherein the processor executes operations, the operations comprising;

(i) requesting, by an execution unit, additional execution time from a kernel to prevent the execution unit from being swapped out during an execution of a critical section;

(ii) determining, by the kernel, whether the execution unit has previously avoided yielding execution to other execution units within a predetermined period of time after being allocated the additional execution time in response to at least one previous request for the additional execution time; and

(iii) allocating, by the kernel, the additional execution time to the execution unit, in response to determining that the execution unit has not previously avoided yielding execution to the other execution units within the predetermined period of time after being allocated the additional execution time in response to the at least one previous request for the additional execution time.

7. The system of claim 6, the operations further comprising:

maintaining a first indicator in the kernel, wherein the first indicator indicates whether the execution unit has previously avoided yielding execution to the other execution units within the predetermined period of time after being allocated the additional execution time in response to the at least one previous request for the additional execution time;

maintaining at least one second indicator that indicates whether the additional time has been granted by the kernel to the execution unit; and

maintaining a third indicator that indicates whether the additional time has been requested by the execution unit from the kernel.

8. The system of claim 7, the operations further comprising:

updating, by the execution unit, the third indicator to indicate that the additional time is longer needed by the execution unit, in response to the execution unit exiting from the critical section; and

swapping in at least one other execution unit, by the kernel, in response to the updating by the execution unit of third indicator to indicate that the additional time is longer needed by the execution unit.

9. The system of claim 6, wherein the execution unit is a first thread, the operations further comprising:

determining, by the kernel, that a second thread that is to be swapped in to replace the first thread is part of the same process as the first thread; and

allocating, by the kernel, the additional execution time to the first thread in response to determining that the second thread that is to be swapped in to replace the first thread is part of the same process as the first thread whether or not the process has previously avoided yielding execution for other processes within the predetermined period of time.

10. The system of claim 6, wherein the execution unit is one of a process or a thread, wherein the kernel is part of an operating system, and wherein the execution unit is not part of the kernel, and wherein the requesting, the determining, and the allocating are performed via system calls.

11. An article of manufacture, wherein the article of manufacture is capable of causing operations on a computer, the operations comprising:

requesting, by an execution unit, additional execution time from a kernel to prevent the execution unit from being swapped out during an execution of a critical section;

determining, by the kernel, whether the execution unit has previously avoided yielding execution to other execution units within a predetermined period of time after being allocated the additional execution time in response to at least one previous request for the additional execution time; and

allocating, by the kernel, the additional execution time to the execution unit, in response to determining that the execution unit has not previously avoided yielding execution to the other execution units within the predetermined period of time after being allocated the additional execution time in response to the at least one previous request for the additional execution time.

12. The article of manufacture of claim 11, the operations further comprising:

maintaining a first indicator in the kernel, wherein the first indicator indicates whether the execution unit has previously avoided yielding execution to the other execution units within the predetermined period of time after being allocated the additional execution time in response to the at least one previous request for the additional execution time;

maintaining at least one second indicator that indicates whether the additional time has been granted by the kernel to the execution unit; and

maintaining a third indicator that indicates whether the additional time has been requested by the execution unit from the kernel.

13. The article of manufacture of claim 12, the operations further comprising:

updating, by the execution unit, the third indicator to indicate that the additional time is longer needed by the execution unit, in response to the execution unit exiting from the critical section; and

swapping in at least one other execution unit, by the kernel, in response to the updating by the execution unit of third indicator to indicate that the additional time is longer needed by the execution unit.

14. The article of manufacture of claim 11, wherein the execution unit is a first thread, the operations further comprising:

determining, by the kernel, that a second thread that is to be swapped in to replace the first thread is part of the same process as the first thread; and

allocating, by the kernel, the additional execution time to the first thread in response to determining that the second thread that is to be swapped in to replace the first thread is part of the same process as the first thread whether or not the process has previously avoided yielding execution for other processes within the predetermined period of time.

15. The article of manufacture of claim 11, wherein the execution unit is one of a process or a thread, wherein the kernel is part of an operating system, and wherein the execution unit is not part of the kernel, and wherein the requesting, the determining, and the allocating are performed via system calls.

16. A method for deploying computing infrastructure, comprising integrating computer-readable code into a computer, wherein the code in combination with the computer is capable of performing:

requesting, by an execution unit, additional execution time from a kernel to prevent the execution unit from being swapped out during an execution of a critical section;

determining, by the kernel, whether the execution unit has previously avoided yielding execution to other execution units within a predetermined period of time after being allocated the additional execution time in response to at least one previous request for the additional execution time; and

allocating, by the kernel, the additional execution time to the execution unit, in response to determining that the execution unit has not previously avoided yielding execution to the other execution units within the predetermined period of time after being allocated the additional execution time in response to the at least one previous request for the additional execution time.

17. The method for deploying computing infrastructure of claim 16, wherein the code in combination with the computer is further capable of performing:

maintaining a first indicator in the kernel, wherein the first indicator indicates whether the execution unit has previously avoided yielding execution to the other execution units within the predetermined period of time after being allocated the additional execution time in response to the at least one previous request for the additional execution time;

maintaining at least one second indicator that indicates whether the additional time has been granted by the kernel to the execution unit; and

maintaining a third indicator that indicates whether the additional time has been requested by the execution unit from the kernel.

18. The method for deploying computing infrastructure of claim 17, wherein the code in combination with the computer is further capable of performing:

updating, by the execution unit, the third indicator to indicate that the additional time is longer needed by the execution unit, in response to the execution unit exiting from the critical section; and

swapping in at least one other execution unit, by the kernel, in response to the updating by the execution unit of third indicator to indicate that the additional time is longer needed by the execution unit.

19. The method for deploying computing infrastructure of claim 16, wherein the execution unit is a first thread, wherein the code in combination with the computer is further capable of performing:

determining, by the kernel, that a second thread that is to be swapped in to replace the first thread is part of the same process as the first thread; and

allocating, by the kernel, the additional execution time to the first thread in response to determining that the second thread that is to be swapped in to replace the first thread is part of the same process as the first thread whether or not the process has previously avoided yielding execution for other processes within the predetermined period of time.

20. The method for deploying computing infrastructure of claim 16, wherein the execution unit is one of a process or a thread, wherein the kernel is part of an operating system, and wherein the execution unit is not part of the kernel, and wherein the requesting, the determining, and the allocating are performed via system calls.