Data Communications In A Parallel Active Messaging Interface Of A Parallel Computer

• US 20120210094A1
• Filed: 02/15/2011
• Published: 08/16/2012
• Est. Priority Date: 02/15/2011
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Blocksome, Michael A., Ratterman, Joseph D., Smith, Brian E.
• Granted Patent: US 8,825,983 B2
• Time in Patent Office: Days
• US Class Current: 711/206
• CPC Class Codes: G06F 12/1036 for multiple virtual address spaces, e.g. segmentation G06F12/1045 takes precedence
• CPC Class Codes: G06F 15/17337 Direct connection machines, e.g. completely connected computers, point to point communication networks coupling between buses G06F13/4004

Abstract

Eager send data communications in a parallel active messaging interface (PAMI) of a parallel computer, the PAMI composed of data communications endpoints that specify a client, a context, and a task, including receiving an eager send data communications instruction with transfer data disposed in a send buffer characterized by a read/write send buffer memory address in a read/write virtual address space of the origin endpoint; determining for the send buffer a read-only send buffer memory address in a read-only virtual address space, the read-only virtual address space shared by both the origin endpoint and the target endpoint, with all frames of physical memory mapped to pages of virtual memory in the read-only virtual address space; and communicating by the origin endpoint to the target endpoint an eager send message header that includes the read-only send buffer memory address.

Claims

1. A method of eager send data communications in a parallel active messaging interface (PAMI) of a parallel computer, the parallel computer comprising a plurality of compute nodes that execute a parallel application, the PAMI comprising data communications endpoints, each endpoint comprising a specification of data communications parameters for a thread of execution on a compute node, including specifications of a client, a context, and a task, the compute nodes and the endpoints coupled for data communications through the PAMI and through data communications resources, the method comprising:

receiving in an origin endpoint of the PAMI an eager send data communications instruction that specifies a transmission of transfer data from the origin endpoint to a target endpoint, the transfer data disposed in a send buffer, the send buffer characterized by a read/write send buffer memory address in a read/write virtual address space of the origin endpoint;

determining by the origin endpoint for the send buffer a read-only send buffer memory address in a read-only virtual address space, the read-only virtual address space shared by both the origin endpoint and the target endpoint, the read-only virtual address space non-overlapping with respect to the read/write virtual address space, the read-only virtual address space established by an operating system of the parallel computer with all frames of physical memory mapped to pages of virtual memory in the read-only virtual address space; and

communicating by the origin endpoint to the target endpoint an eager send message header that includes the read-only send buffer memory address.

2. The method of claim 1 wherein:

the read/write virtual address space and the read-only virtual address space are both characterized by a same virtual address space size;

all frames of physical memory mapped to pages of virtual memory further comprises frames mapped to pages with memory addresses that are identical except for the most significant bit in each address; and

determining for the send buffer a read-only send buffer memory address in a read-only virtual address space comprises setting a most significant bit in the read/write send buffer memory address.

3. The method of claim 1 wherein determining for the send buffer a read-only send buffer memory address in a read-only virtual address space comprises determining for the send buffer a read-only send buffer memory address in a read-only virtual address space through a PAMI query to the operating system.

4. The method of claim 1 wherein:

the operating system maintains records of associations among addresses in physical memory, in read/write virtual address spaces, and in the read-only virtual address space; and

determining for the send buffer a read-only send buffer memory address in a read-only virtual address space further comprises finding through the operating system a physical memory address associated with the read/write send buffer memory address and a read-only virtual memory address associated with the physical memory address.

5. The method of claim 1 wherein:

each endpoint of the PAMI operates with a separate read/write virtual address space backed by real memory; and

all endpoints of the PAMI share the read-only virtual address space, the read-only virtual address space being non-overlapping with respect to all of the separate read/write virtual address spaces of the endpoints.

6. The method of claim 1 further comprising copying by the target endpoint transfer data from the send buffer to a target receive buffer that is characterized by a read/write receive buffer memory address in a read/write virtual address space of the target endpoint.

7. The method of claim 1 further comprising using by the target endpoint the send buffer as a target receive buffer without copying by the target endpoint transfer data from the send buffer to a receive buffer.

8. The method of claim 1 wherein:

each client comprises a collection of data communications resources dedicated to the exclusive use of an application-level data processing entity;

each context comprises a subset of the collection of data processing resources of a client, context functions, and a work queue of data transfer instructions to be performed by use of the subset through the context functions operated by an assigned thread of execution; and

each task represents a process of execution of the parallel application.

9. The method of claim 1 wherein each context carries out, through post and advance operations, data communications for the application on data communications resources in the exclusive possession of that context.

10. The method of claim 1 wherein each context carries out data communications operations independently and in parallel with other contexts.

11. A parallel computer that implements eager send data communications in a parallel active messaging interface (PAMI) of a parallel computer, the parallel computer comprising a plurality of compute nodes that execute a parallel application, the PAMI comprising data communications endpoints, each endpoint comprising a specification of data communications parameters for a thread of execution on a compute node, including specifications of a client, a context, and a task, the compute nodes and the endpoints coupled for data communications through the PAMI and through data communications resources, the compute nodes comprising computer processors operatively coupled to computer memory having disposed within it computer program instructions that, when executed by the computer processors, cause the parallel computer to function by:

receiving in an origin endpoint of the PAMI an eager send data communications instruction that specifies a transmission of transfer data from the origin endpoint to a target endpoint, the transfer data disposed in a send buffer, the send buffer characterized by a read/write send buffer memory address in a read/write virtual address space of the origin endpoint;

determining by the origin endpoint for the send buffer a read-only send buffer memory address in a read-only virtual address space, the read-only virtual address space shared by both the origin endpoint and the target endpoint, the read-only virtual address space non-overlapping with respect to the read/write virtual address space, the read-only virtual address space established by an operating system of the parallel computer with all frames of physical memory mapped to pages of virtual memory in the read-only virtual address space; and

communicating by the origin endpoint to the target endpoint an eager send message header that includes the read-only send buffer memory address.

12. The parallel computer of claim 11 wherein determining for the send buffer a read-only send buffer memory address in a read-only virtual address space comprises determining for the send buffer a read-only send buffer memory address in a read-only virtual address space through a PAMI query to the operating system.

13. The parallel computer of claim 11 wherein:

the operating system maintains records of associations among addresses in physical memory, in read/write virtual address spaces, and in the read-only virtual address space; and

determining for the send buffer a read-only send buffer memory address in a read-only virtual address space further comprises finding through the operating system a physical memory address associated with the read/write send buffer memory address and a read-only virtual memory address associated with the physical memory address.

14. The parallel computer of claim 11 wherein:

each endpoint of the PAMI operates with a separate read/write virtual address space backed by real memory; and

all endpoints of the PAMI share the read-only virtual address space, the read-only virtual address space being non-overlapping with respect to all of the separate read/write virtual address spaces of the endpoints.

15. The parallel computer of claim 11 wherein:

each client comprises a collection of data communications resources dedicated to the exclusive use of an application-level data processing entity;

each context comprises a subset of the collection of data processing resources of a client, context functions, and a work queue of data transfer instructions to be performed by use of the subset through the context functions operated by an assigned thread of execution; and

each task represents a process of execution of the parallel application.

16. A computer program product for eager send data communications in a parallel active messaging interface (PAMI'"'"') of a parallel computer, the parallel computer comprising a plurality of compute nodes that execute a parallel application, the PAMI comprising data communications endpoints, each endpoint comprising a specification of data communications parameters for a thread of execution on a compute node, including specifications of a client, a context, and a task, the compute nodes and the endpoints coupled for data communications through the PAMI and through data communications resources, the computer program product comprising computer program instructions that, when installed and executed, cause the parallel computer to function by:

receiving in an origin endpoint of the PAMI an eager send data communications instruction that specifies a transmission of transfer data from the origin endpoint to a target endpoint, the transfer data disposed in a send buffer, the send buffer characterized by a read/write send buffer memory address in a read/write virtual address space of the origin endpoint;

determining by the origin endpoint for the send buffer a read-only send buffer memory address in a read-only virtual address space, the read-only virtual address space shared by both the origin endpoint and the target endpoint, the read-only virtual address space non-overlapping with respect to the read/write virtual address space, the read-only virtual address space established by an operating system of the parallel computer with all frames of physical memory mapped to pages of virtual memory in the read-only virtual address space; and

communicating by the origin endpoint to the target endpoint an eager send message header that includes the read-only send buffer memory address.

17. The computer program product of claim 16 wherein:

the read/write virtual address space and the read-only virtual address space are both characterized by a same virtual address space size;

all frames of physical memory mapped to pages of virtual memory further comprises frames mapped to pages with memory addresses that are identical except for the most significant bit in each address; and

determining for the send buffer a read-only send buffer memory address in a read-only virtual address space comprises setting a most significant bit in the read/write send buffer memory address.

18. The computer program product of claim 16 wherein:

the operating system maintains records of associations among addresses in physical memory, in read/write virtual address spaces, and in the read-only virtual address space; and

determining for the send buffer a read-only send buffer memory address in a read-only virtual address space further comprises finding through the operating system a physical memory address associated with the read/write send buffer memory address and a read-only virtual memory address associated with the physical memory address.

19. The computer program product of claim 16 wherein:

each endpoint of the PAMI operates with a separate read/write virtual address space backed by real memory; and

all endpoints of the PAMI share the read-only virtual address space, the read-only virtual address space being non-overlapping with respect to all of the separate read/write virtual address spaces of the endpoints.

20. The computer program product of claim 16 wherein:

each client comprises a collection of data communications resources dedicated to the exclusive use of an application-level data processing entity;

each context comprises a subset of the collection of data processing resources of a client, context functions, and a work queue of data transfer instructions to be performed by use of the subset through the context functions operated by an assigned thread of execution; and

each task represents a process of execution of the parallel application.