Framework for domain-specific run-time environment acceleration using virtualization technology

• US 20070226736A1
• Filed: 03/21/2006
• Published: 09/27/2007
• Est. Priority Date: 03/21/2006
• Status: Active Grant

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: Johnson, Erik, Kunze, Aaron
• Granted Patent: US 8,042,109 B2
• Time in Patent Office: Days
• US Class Current: 718/1
• CPC Class Codes: G06F 9/45533 Hypervisors; Virtual machine monitors
• CPC Class Codes: G06F 9/4843 by program, e.g. task dispatcher, supervisor, operating system

Abstract

In some embodiments, the invention involves a framework for using virtualization technology to efficiently support a domain-specific run-time environment. In at least one embodiment, a framework is utilized to take advantage of virtualization technology (VT) to partition performance critical and non-performance critical tasks of the same domain-specific application. An embodiment of the invention utilizes a general-purpose operating system to execute non-performance critical aspects of a domain, and uses a privileged VT-root mode to execute performance critical aspects of the domain. Another embodiment uses one or more guest VMs to execute the performance critical aspects of the domain-specific run-time environment. Other embodiments are described and claimed.

Claims

1. A system comprising:

a platform having virtualization capabilities;

at least one processor coupled to the platform, the at least one processor to run a general-purpose operating system (GPOS) in a first virtual machine (VM) on the platform; and

a domain specific run-time environment (DSRTE) partitioned into at least two portions, wherein a first portion comprises non-performance critical processes to run under the GPOS in the first VM, and wherein at least one additional portion comprises at least one performance critical process to run on the platform outside of the first VM running the GPOS.

2. The system as recited in claim 1, further comprising:

a privileged root domain comprising a virtual machine monitor (VMM) to control each VM on the platform, wherein the at least one performance critical process is to run in a second virtual machine, where the second VM is non-privileged VM.

3. The system as recited in claim 2, wherein the first VM and the second VM communicate via at least one communication method selected from the group of communication methods consisting of mailboxes, shared memory, and network packets.

4. The system as recited in claim 2, further comprising:

at least one additional virtual machine to run one or more processes, wherein each of the one or more processes comprise one of a performance critical process and a non-performance critical process.

5. The system as recited in claim 1, further comprising:

a privileged root domain, wherein the at least one performance critical process is to run in the privileged root domain.

6. The system as recited in claim 5, wherein the privileged root domain is to be entered in response to a VM-EXIT event, and is to return control to a VM in response to a VM-ENTER event, and wherein initiating a performance critical task is to generate a VM-EXIT from the GPOS in the first VM.

7. The system as recited in claim 5, wherein the privileged root domain comprises a domain specific run-time environment customized to execute the at least one performance critical process efficiently, wherein the GPOS comprises a GPOS that is not customized to execute the at least one performance critical process.

8. A computer implemented method comprising:

partitioning a domain specific run-time environment (DSRTE) into at least two partitions, a first partition comprising non-performance critical tasks to run under a general-purpose operating system (GPOS) in a virtual machine (VM) and at least one additional partition comprising performance critical tasks, the DSRTE to reside on a platform having virtualization capabilities;

generating an Exit to a privileged root domain from the first partition, in response to an event for executing a performance critical task;

executing the performance critical task in the at least one additional partition; and

returning control to the first partition in response to an Enter event.

9. The method as recited in claim 8, wherein the at least one additional partition comprises two or more partitions, each of the two or more partitions to run a performance critical operation corresponding to a specific input/output (I/O) device.

10. The method as recited in claim 8, wherein the performance critical task comprises a network communication task.

11. The method as recited in claim 8, wherein the non-performance critical tasks include a user interface, and the performance critical tasks include network packet communication.

12. The method as recited in claim 8, wherein the platform comprises a set-top box environment, and the performance critical tasks include coding and decoding of audio-visual streams.

13. The method as recited in claim 8, wherein the at least one additional partition comprises a privileged root domain running on the platform.

14. The method as recited in claim 8, wherein the at least one additional partition comprises at least one additional corresponding non-privileged VM, wherein a privileged virtual machine monitor (VMM) controls the execution of each VM on the platform.

15. A computer implemented method comprising:

partitioning a domain specific run-time environment (DSRTE) into at least two partitions, a first partition comprising non-performance critical tasks to run under a general-purpose operating system (GPOS) in a virtual machine (VM) and a second partition comprising performance critical tasks, the DSRTE to reside on a platform having virtualization capabilities; and

executing the performance critical tasks in the second partition, wherein the first partition is used to expand a set of services provided to the DSRTE.

16. The method as recited in claim 15, wherein the second partition comprises a primary and privileged execution environment of the DSRTE.

17. The method as recited in claim 15, wherein the second partition comprises a second non-privileged virtual machine (VM).

18. The method as recited in claim 15, further comprising:

entering the first partition when the performance critical tasks need a service of the GPOS; and

exiting to the second partition to continue processing performance critical tasks.

19. A machine readable medium having instructions that when executed cause the machine to:

execute a domain specific run-time environment (DSRTE) having at least two partitions, wherein a first partition of the domain specific run-time environment (DSRTE) comprises non-performance critical tasks to run under a general-purpose operating system (GPOS) in a virtual machine (VM), and wherein at least one additional partition comprises performance critical tasks, and wherein the DSRTE is to reside on a platform having virtualization capabilities;

generate an Exit to a privileged root domain from the first partition, in response to an event for executing a performance critical task;

execute the performance critical tasks in the at least one additional partition; and

return control to the first partition in response to an Enter event.

20. The medium as recited in claim 19, wherein the at least one additional partition comprises two or more partitions, each of the two or more partitions to run a performance critical operation corresponding to a specific input/output (I/O) device.

21. The medium as recited in claim 19, wherein the performance critical task comprises a network communication task.

22. The medium as recited in claim 19, wherein the non-performance critical tasks include a user interface, and the performance critical tasks include network packet communication.

23. The medium as recited in claim 19, wherein the platform comprises a set-top box environment, and the performance critical tasks include coding and decoding of audio-visual streams.

24. The medium as recited in claim 19, wherein the at least one additional partition comprises a privileged root domain running on the platform.

25. The medium as recited in claim 19, wherein the at least one additional partition comprises at least one additional corresponding non-privileged VM, wherein a privileged virtual machine monitor (VMM) controls the execution of each VM on the platform.

26. A machine readable medium having instructions that when executed cause the machine to:

partition a domain specific run-time environment (DSRTE) into at least two partitions, a first partition comprising non-performance critical tasks to run under a general-purpose operating system (GPOS) in a virtual machine (VM) and a second partition comprising performance critical tasks, the DSRTE to reside on a platform having virtualization capabilities; and

execute the performance critical tasks in the second partition, wherein the first partition is used to expand a set of services provided to the DSRTE.

27. The medium as recited in claim 26, wherein the second partition comprises a primary and privileged execution environment of the DSRTE,

28. The medium as recited in claim 26, further comprising instructions to:

enter the first partition when the performance critical tasks need a service of the GPOS; and

exit to the second partition to continue processing performance critical tasks.