METHOD AND SYSTEM FOR ALLOCATING MEMORY TO A PIPELINE

• US 20120066470A1
• Filed: 08/11/2011
• Published: 03/15/2012
• Est. Priority Date: 09/10/2010
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Okada, Masakuni, Yamafuji, Kazuto, Yonezawa, Takashi, Fukuda, Atsushi
• Granted Patent: US 8,775,767 B2
• Time in Patent Office: Days
• US Class Current: 711/171
• CPC Class Codes: G06F 12/023 Free address space management
• CPC Class Codes: G06F 15/17375 One dimensional, e.g. linear array, ring
• CPC Class Codes: G06F 9/3851 from multiple instruction streams, e.g. multistreaming
• CPC Class Codes: G06F 9/3885 using a plurality of independent parallel functional units

Abstract

A method, and associated system, for allocating memory to a first pipeline that includes a sequence of filters. Each filter is configured to execute a process specific to each filter, receive input data, and generate output data. The output data from each filter, except the last filter in the sequence, serves as the input data to the next filter in the sequence. An optimum memory capacity is allocated to the first pipeline if possible. Otherwise, a guaranteed memory bandwidth is allocated to the first pipeline if possible. Otherwise, extra memory currently allocated to a second pipeline is currently released if the second pipeline not currently performing processing or subsequently released when the second pipeline subsequently completes performing processing that is currently being performed, followed by allocating the extra memory to the first pipeline.

Claims

1. A method for allocating memory to a first pipeline, said method comprising:

ascertaining, by a processor of a computer system, whether an optimum memory capacity can be allocated to a first pipeline that comprises a sequence of filters, wherein each filter is configured to execute a process specific to each filter, receive input data to enable execution the process, and generate output data resulting from execution of the process, and wherein the output data generated by each filter except the last filter in the sequence is the input data to a next filter in the sequence;

if said ascertaining whether the optimum memory capacity can be allocated to the first pipeline ascertains that the optimum memory capacity can be allocated to the first pipeline, then said processor allocating the optimum memory capacity to the first pipeline, else said processor ascertaining whether a guaranteed memory bandwidth can be allocated to the first pipeline;

if said ascertaining whether the guaranteed memory bandwidth can be allocated to the first pipeline ascertains that the guaranteed memory bandwidth can be allocated to the first pipeline, then said processor allocating the guaranteed memory bandwidth to the first pipeline, else said processor currently releasing extra memory from a second pipeline if the second pipeline is not currently performing processing or subsequently releasing the extra memory from the second pipeline when the second pipeline subsequently completes performing processing that is currently being performed, followed by said processor allocating the extra memory to the first pipeline.

2. The method of claim 1, wherein said ascertaining whether the optimum memory capacity can be allocated to the first pipeline ascertains that the optimum memory capacity can be allocated to the first pipeline.

3. The method of claim 1, wherein said ascertaining whether the optimum memory capacity can be allocated to the first pipeline ascertains that the optimum memory capacity cannot be allocated to the first pipeline, and wherein said ascertaining whether the guaranteed memory bandwidth can be allocated to the first pipeline ascertains that the guaranteed memory bandwidth can be allocated to the first pipeline.

4. The method of claim 1, wherein said ascertaining whether the optimum memory capacity can be allocated to the first pipeline ascertains that the optimum memory capacity cannot be allocated to the first pipeline, wherein said ascertaining whether the guaranteed memory bandwidth can be allocated to the first pipeline ascertains that the guaranteed memory bandwidth cannot be allocated to the first pipeline, and wherein said ascertaining whether the extra memory can be currently released ascertains that the extra memory can be currently released.

5. The method of claim 1, wherein said ascertaining whether the optimum memory capacity can be allocated to the first pipeline ascertains that the optimum memory capacity cannot be allocated to the first pipeline, wherein said ascertaining whether the guaranteed memory bandwidth can be allocated to the first pipeline ascertains that the guaranteed memory bandwidth cannot be allocated to the first pipeline, and wherein said ascertaining whether the extra memory can be currently released ascertains that the extra memory cannot be currently released.

6. The method of claim 1, wherein said ascertaining whether the optimum memory capacity can be allocated to the first pipeline ascertains that the optimum memory capacity cannot be allocated to the first pipeline, wherein each filter is characterized by a guaranteed ratio of memory usage expressed as a ratio of a numerator to a denominator such that each ratio is specific to each filter, and wherein the method further comprises:

prior to said ascertaining whether the guaranteed memory bandwidth can be allocated to the first pipeline, computing the guaranteed memory bandwidth of the first pipeline as N1+D, wherein N1 is the numerator of the guaranteed ratio of the first filter in the sequence of filters and D is a sum of the denominators of the guaranteed ratio of all filters in the sequence of filters.

7. The method of claim 1, wherein the method further comprises:

prior to said ascertaining whether the optimum memory capacity can be allocated to the first pipeline, ascertaining the optimum memory capacity of the first pipeline by performing the smoothing procedure.

8. The method of claim 1, wherein the method further comprises:

prior to said ascertaining whether the optimum memory capacity can be allocated to the first pipeline, ascertaining the optimum memory capacity of the first pipeline by performing the addition procedure.

9. A computer program product, comprising a computer readable physically tangible storage device having program code stored therein, said program code configured to be executed by a processor of a computer to implement a method for allocating memory to a first pipeline, said method comprising:

ascertaining whether an optimum memory capacity can be allocated to a first pipeline that comprises a sequence of filters, wherein each filter is configured to execute a process specific to each filter, receive input data to enable execution the process, and generate output data resulting from execution of the process, and wherein the output data generated by each filter except the last filter in the sequence is the input data to a next filter in the sequence;

if said ascertaining whether the optimum memory capacity can be allocated to the first pipeline ascertains that the optimum memory capacity can be allocated to the first pipeline, then allocating the optimum memory capacity to the first pipeline, else ascertaining whether a guaranteed memory bandwidth can be allocated to the first pipeline;

if said ascertaining whether the guaranteed memory bandwidth can be allocated to the first pipeline ascertains that the guaranteed memory bandwidth can be allocated to the first pipeline, then allocating the guaranteed memory bandwidth to the first pipeline, else currently releasing extra memory from a second pipeline if the second pipeline is not currently performing processing or subsequently releasing the extra memory from the second pipeline when the second pipeline subsequently completes performing processing that is currently being performed, followed by allocating the extra memory to the first pipeline.

10. The computer program product of claim 9, wherein said ascertaining whether the optimum memory capacity can be allocated to the first pipeline ascertains that the optimum memory capacity can be allocated to the first pipeline.

11. The computer program product of claim 9, wherein said ascertaining whether the optimum memory capacity can be allocated to the first pipeline ascertains that the optimum memory capacity cannot be allocated to the first pipeline, and wherein said ascertaining whether the guaranteed memory bandwidth can be allocated to the first pipeline ascertains that the guaranteed memory bandwidth can be allocated to the first pipeline.

12. The computer program product of claim 9, wherein said ascertaining whether the optimum memory capacity can be allocated to the first pipeline ascertains that the optimum memory capacity cannot be allocated to the first pipeline, wherein said ascertaining whether the guaranteed memory bandwidth can be allocated to the first pipeline ascertains that the guaranteed memory bandwidth cannot be allocated to the first pipeline, and wherein said ascertaining whether the extra memory can be currently released ascertains that the extra memory can be currently released.

13. The computer program product of claim 9, wherein said ascertaining whether the optimum memory capacity can be allocated to the first pipeline ascertains that the optimum memory capacity cannot be allocated to the first pipeline, wherein said ascertaining whether the guaranteed memory bandwidth can be allocated to the first pipeline ascertains that the guaranteed memory bandwidth cannot be allocated to the first pipeline, and wherein said ascertaining whether the extra memory can be currently released ascertains that the extra memory cannot be currently released.

14. The computer program product of claim 9, wherein said ascertaining whether the optimum memory capacity can be allocated to the first pipeline ascertains that the optimum memory capacity cannot be allocated to the first pipeline, wherein each filter is characterized by a guaranteed ratio of memory usage expressed as a ratio of a numerator to a denominator such that each ratio is specific to each filter, and wherein the method further comprises:

prior to said ascertaining whether the guaranteed memory bandwidth can be allocated to the first pipeline, computing the guaranteed memory bandwidth of the first pipeline as N1+D, wherein N1 is the numerator of the guaranteed ratio of the first filter in the sequence of filters and D is a sum of the denominators of the guaranteed ratio of all filters in the sequence of filters.

15. A computer system comprising a processor and a computer readable memory unit coupled to the processor, said memory unit containing program code configured to be executed by the processor to implement a method for allocating memory to a first pipeline, said method comprising:

ascertaining whether an optimum memory capacity can be allocated to a first pipeline that comprises a sequence of filters, wherein each filter is configured to execute a process specific to each filter, receive input data to enable execution the process, and generate output data resulting from execution of the process, and wherein the output data generated by each filter except the last filter in the sequence is the input data to a next filter in the sequence;

if said ascertaining whether the optimum memory capacity can be allocated to the first pipeline ascertains that the optimum memory capacity can be allocated to the first pipeline, then allocating the optimum memory capacity to the first pipeline, else ascertaining whether a guaranteed memory bandwidth can be allocated to the first pipeline;

if said ascertaining whether the guaranteed memory bandwidth can be allocated to the first pipeline ascertains that the guaranteed memory bandwidth can be allocated to the first pipeline, then allocating the guaranteed memory bandwidth to the first pipeline, else currently releasing extra memory from a second pipeline if the second pipeline is not currently performing processing or subsequently releasing the extra memory from the second pipeline when the second pipeline subsequently completes performing processing that is currently being performed, followed by allocating the extra memory to the first pipeline.

16. The computer system of claim 15, wherein said ascertaining whether the optimum memory capacity can be allocated to the first pipeline ascertains that the optimum memory capacity can be allocated to the first pipeline.

17. The computer system of claim 15, wherein said ascertaining whether the optimum memory capacity can be allocated to the first pipeline ascertains that the optimum memory capacity cannot be allocated to the first pipeline, and wherein said ascertaining whether the guaranteed memory bandwidth can be allocated to the first pipeline ascertains that the guaranteed memory bandwidth can be allocated to the first pipeline.

18. The computer system of claim 15, wherein said ascertaining whether the optimum memory capacity can be allocated to the first pipeline ascertains that the optimum memory capacity cannot be allocated to the first pipeline, wherein said ascertaining whether the guaranteed memory bandwidth can be allocated to the first pipeline ascertains that the guaranteed memory bandwidth cannot be allocated to the first pipeline, and wherein said ascertaining whether the extra memory can be currently released ascertains that the extra memory can be currently released.

19. The computer system of claim 15, wherein said ascertaining whether the optimum memory capacity can be allocated to the first pipeline ascertains that the optimum memory capacity cannot be allocated to the first pipeline, wherein said ascertaining whether the guaranteed memory bandwidth can be allocated to the first pipeline ascertains that the guaranteed memory bandwidth cannot be allocated to the first pipeline, and wherein said ascertaining whether the extra memory can be currently released ascertains that the extra memory cannot be currently released.

20. The computer system of claim 15, wherein said ascertaining whether the optimum memory capacity can be allocated to the first pipeline ascertains that the optimum memory capacity cannot be allocated to the first pipeline, wherein each filter is characterized by a guaranteed ratio of memory usage expressed as a ratio of a numerator to a denominator such that each ratio is specific to each filter, and wherein the method further comprises:

prior to said ascertaining whether the guaranteed memory bandwidth can be allocated to the first pipeline, computing the guaranteed memory bandwidth of the first pipeline as N1+D, wherein N1 is the numerator of the guaranteed ratio of the first filter in the sequence of filters and D is a sum of the denominators of the guaranteed ratio of all filters in the sequence of filters.