Sharing virtual memory-based multi-version data between the heterogenous processors of a computer platform

• US 8,868,848 B2
• Filed: 12/21/2009
• Issued: 10/21/2014
• Est. Priority Date: 12/21/2009
• Status: Active Grant

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: Gao, Ying, Chen, Hu, Yan, Shoumeng, Zhou, Xiaocheng, Luo, Sai, Saha, Bratin
• Primary Examiner(s): Savla, Arpan P.
• Application Number:
  US12/643,803
• Publication Number:
  US 20110153957A1
• Time in Patent Office: 1,765 Days
• Field of Search: 711/147, 711/154, 711/162, 707/695, 712/32, 712/34
• US Class Current: 711/147
• CPC Class Codes: G06F 12/109 for multiple virtual address spaces, e.g. segmentation G06F12/1036 takes precedence
• CPC Class Codes: G06F 15/167 using a common memory, e.g. mailbox
• CPC Class Codes: G06F 2212/656 Address space sharing
• CPC Class Codes: G06F 2212/657 Virtual address space management
• CPC Class Codes: G06F 9/466 Transaction processing
• CPC Class Codes: G06F 9/544 Buffers; Shared memory; Pipes
• CPC Class Codes: G06T 1/20 Processor architectures; Processor configuration, e.g. pipelining

Abstract

A computer system may comprise a computer platform and input-output devices. The computer platform may include a plurality of heterogeneous processors comprising a central processing unit (CPU) and a graphics processing unit (GPU) and a shared virtual memory supported by a physical private memory space of at least one heterogeneous processor or a physical shared memory shared by the heterogeneous processor. The CPU (producer) may create shared multi-version data and store such shared multi-version data in the physical private memory space or the physical shared memory. The GPU (consumer) may acquire or access the shared multi-version data.

Claims

1. A method to be performed in a computer platform having a plurality of heterogeneous processors including at least a first processor and a second processor, the method comprising:

creating multi-version data using the first processor, wherein the multi-version data includes an initial version and at least one additional version,

identifying, for each additional version that is created and released, difference content between a current version and a previous version of the multi-version data using the first processor,

storing the multi-version data including the initial version and the difference content for each additional version released in a shared virtual memory, wherein the shared virtual memory is accessible to at least a subset of the heterogeneous processors including the first processor and the second processor, and

acquiring a particular version of the multi-version data from the shared virtual memory for use by the second processor, wherein the acquiring includes retrieving the initial version and the difference content for each additional version up to the particular version and merging the difference content for each successive additional version into the initial version.

2. The method of claim 1, wherein the storing the multi-version data includes storing the difference content for each additional version released in a separate memory location within the shared virtual memory.

3. The method of claim 2, further comprises generating a new version number for the difference content and storing the new version number with the difference content in the separate memory location within the shared virtual memory.

4. The method of claim 3, further comprising tracking the difference content and associated version numbers stored in the shared virtual memory.

5. The method of claim 1, further comprising releasing an additional version of the multi-version data in write protection mode.

6. The method of claim 5, further comprising creating a back-up copy of the additional version of the multi-version data released in write protection mode, wherein the back-up copy is created in a handler and can be modified.

7. The method of claim 6, further comprising marking an auxiliary storage to indicate that the back-up copy is modified.

8. The method of claim 1, further comprising acquiring an updated version of the multi-version data from the shared virtual memory for use by the second processor, wherein the acquiring includes retrieving the difference content for each additional version after the particular version and merging the difference content for each successive additional version into the particular version.

9. A computer platform comprising:

a plurality of heterogeneous processors that includes a first processor acting as a producer and a second processor acting as a consumer and a shared virtual memory,

wherein the first processor is to

create multi-version data, wherein the multi-version data includes an initial version and at least one additional version,

identify, for each additional version that is created and released, difference content between a current version and a previous version of the multi-version data, and

store the multi-version data including the initial version and the difference content for each additional version released in the shared virtual memory, wherein the shared virtual memory is accessible to at least a subset of the heterogeneous processors including the first processor and the second processor, and

wherein the second processor is to acquire a particular version of the multi-version data from the shared virtual memory, wherein the acquiring includes retrieving the initial version and the difference content for each additional version up to the particular version and merging the difference content for each successive additional version into the initial version.

10. The computer platform of claim 9, wherein the first processor is to store the difference content for each additional version released in a separate memory location within the shared virtual memory.

11. The computer platform of claim 10, wherein the first processor is to further generate a new version number for the difference content and to store the new version number with the difference content in the separate memory location within the shared virtual memory.

12. The computer platform of claim 11, wherein the first processor is to track the difference content and associated version numbers stored in the shared virtual memory.

13. The computer platform of claim 9, wherein the first processor is to release an additional version of the multi-version data in write protection mode.

14. The computer platform of claim 13, wherein the first processor is to create a back-up copy of the additional version of the multi-version data released in write protection mode, wherein the back-up copy is created in a handler and can be modified.

15. The computer platform of claim 14, wherein the first processor is to mark an auxiliary storage to indicate that the back-up copy is modified.

16. The computer platform of claim 9, wherein the second processor is to acquire an updated version of the multi-version data from the shared virtual memory, wherein the acquiring includes retrieving the difference content for each additional version after the particular version and merging the difference content for each successive additional version into the particular version.