Dynamic reconfiguration of applications on a multi-processor embedded system

Patent metadata

Field	Value
Publication number	US-11023272-B2
Application number	US-201815976021-A
Country	US
Kind code	B2
Filing date	May 10, 2018
Priority date	May 17, 2013
Publication date	Jun 1, 2021
Grant date	Jun 1, 2021

Abstract

A multiprocessor system and method for swapping applications executing on the multiprocessor system are disclosed. The plurality of applications may include a first application and a plurality of other applications. The first application may be dynamically swapped with a second application. The swapping may be performed without stopping the plurality of other applications. The plurality of other applications may continue to execute during the swapping to perform a real-time operation and process real-time data. After the swapping, the plurality of other applications may continue to execute with the second application, and at least a subset of the plurality of other applications may communicate with the second application to perform the real time operation and process the real time data.

First claim

What is claimed is:

1. A method, comprising:
executing a master task on a given processor of a plurality of processors included in a multiprocessor system that also includes a plurality of data memory routers interspersed among the plurality of processors;
executing a plurality of applications on a subset of the plurality of processors excluding a particular processor, wherein the plurality of applications execute together and communicate, via a first set of routes established between a first subset of the plurality of data memory routers, with each other to perform a real time operation;
sending a command to the master task by a system controller that is external to the plurality of processors, wherein the command indicates to initiate swapping a first application with a second application;
in response to the master task receiving the command:
in response to receiving a command from the master task via a serial bus included in the multiprocessor system, executing, by the particular processor of the plurality of processors, a loading task to load, via an input/output port, the second application;
distributing instructions associated with the second application among different ones of the plurality of data memory routers via a second set of routes established between a second subset of the plurality of data memory routers;
swapping the first application with the second application;

wherein swapping the first application with the second application includes continuing execution of at least one of the plurality of applications; and
wherein upon completion of the swapping, the at least one of the plurality of applications continue to execute with the second application, and wherein the at least one of the plurality of applications communicates with the second application to perform the real time operation.

2. The method of claim 1 , wherein prior to swapping the first application with the second application, the first application executes on a first subset of the plurality of processors; and wherein swapping the first application with the second application includes:
stopping the first application on the first subset of the plurality of processors;
saving a state of the first application; and
loading the second application into a third subset of the plurality of data memory routers associated with the first subset of the plurality of processors.

3. The method of claim 2 , wherein swapping the first application with the second application further includes decoupling communication of the first application with one or more of the plurality of applications by altering communication settings for at least one of the plurality of data memory routers.

4. The method of claim 3 , wherein swapping the first application with the second application further includes, after loading the second application, coupling communication of the second application with the one or more of the plurality of applications.

5. The method of claim 1 , wherein swapping the first application with the second application includes loading the second application into two or more of the plurality of data memory routers, wherein loading the second application includes sending program instructions of the second application through the multiprocessor system along two or more swapping routes included in the second set of routes, and wherein each of the swapping routes is associated with one of the two or more of the plurality of data memory routers.

6. The method of claim 5 , wherein the two or more swapping routes share a first part in common, and wherein the first part includes a route from the input/output port of the multiprocessor system to an endpoint within the multiprocessor system.

7. A multiprocessor system, comprising:
a plurality of processors, wherein a given processor of the plurality of processors is configured to execute a master task;
a plurality of memories interspersed among the processors; and
a communication fabric interconnecting the plurality of processors and the plurality of memories, wherein the communication fabric includes a plurality of buffer memories interspersed among at least a subset of the plurality of processors and is configured to send a command to the master task from a system controller external to the plurality of processors, wherein the command indicates to initiate swapping a first application with a second application;
wherein the plurality of buffer memories is configured to communicate data between applications executing in the multiprocessor system via a first set of routes between a first subset of the plurality of buffer memories;
wherein, during execution of a plurality of applications, on a subset of the plurality of processors that excludes the given processor and a particular processor, to perform a real time operation, the plurality of buffer memories are configured to break and restore connections between active applications during swapping of the first application with the second application;
wherein swapping of the first application with the second application is initiated by the master task in response to receiving the command from the system controller; and
wherein the particular processor of the plurality of processor is configured, in response to receiving an indication to swap the first and second applications via a serial bus coupled to the plurality of processors, to:
execute, by the particular processor of the plurality of processors, a loading task to load, via an input/output port, the second application; and
distribute instructions associated with the second application among different ones of the plurality of memories via a second set of routes between a second subset of the plurality of buffer memories.

8. The multiprocessor system of claim 7 , wherein prior to swapping the first application with the second application, the first application communicates structured data elements via a first buffer memory, wherein each data element has a beginning boundary and an ending boundary, and wherein the first buffer memory is configured to:
receive a request to discontinue data communication after transmission of a first data element has been initiated; and
continue transmission of the first data element up to the ending boundary of the first data element.

9. The multiprocessor system of claim 7 , wherein the first application is on a downstream side of a first buffer memory of the plurality of buffer memories, wherein the first buffer memory is configurable to discontinue acceptance of data from a sending application during said swapping.

10. The multiprocessor system of claim 7 , further comprising a plurality of data memory routers coupled between the plurality of processors, wherein the data memory routers include the plurality of memories interspersed among the processors.

11. The multiprocessor system of claim 7 , wherein the communication fabric is further configured to load the second application into two or more of the plurality of memories, wherein to load the second application, the communication fabric is further configured to send program instructions of the second application via two or more swapping routes included in the second set of routes, and wherein each of the swapping routes is associated with one of the two or more of the plurality of memories.

12. The multiprocessor system of claim 11 , wherein the two or more swapping routes share a first part in common, and wherein the first part includes a route from the input/output port of the multiprocessor system to an endpoint within the multiprocessor system.

13. The multiprocessor system of claim 7 , wherein the first application is on a downstream side of a

Assignees

• Coherent Logix Inc

Inventors

• Kaku Wilbur William
• Purnell Michael Lyle
• Ellis Geoffrey Neil
• Beardslee John Mark
• Shang Zhong Qing
• Wang Teng-I
• Lim Stephen E

Classifications

• G06F9/4856

  resumption being on a different machine, e.g. task migration, virtual machine migration (G06F9/5088 takes precedence) · CPC title

• G06F12/00

  Accessing, addressing or allocating within memory systems or architectures (digital input from, or digital output to record carriers, e.g. to disk storage units, G06F3/06) · CPC title

• G06F9/5066

  Algorithms for mapping a plurality of inter-dependent sub-tasks onto a plurality of physical CPUs (mappping at compile time, see G06F8/451) · CPC title

• G06F9/461

  Saving or restoring of program or task context · CPC title

• G06F9/485 Primary

  Task life-cycle, e.g. stopping, restarting, resuming execution (G06F9/4881 takes precedence) · CPC title