Selective migration of resources or remapping of virtual processors to provide access to resources

1. A computer system, comprising:

one or more I/O devices;

physical memory; and

a plurality of physical nodes, wherein an operating system is run collectively across the plurality of physical nodes;

wherein each physical node includes one or more physical processors, and wherein each physical processor has one or more cores, and wherein each core has one or more hyperthreads;

wherein each core or hyperthread instantiates one or more virtual processors, wherein a virtual processor comprises a computing engine visible to the operating system run collectively across the plurality of physical nodes;

wherein each core or hyperthread is configured to invoke a hyper-kernel on its hosting physical processor when the core or hyperthread cannot access a resource needed by the core or hyperthread, the needed resource comprising at least one of: (1) a portion of the physical memory needed by the core or hyperthread, and (2) an I/O resource needed by the core or hyperthread; and

wherein the hyper-kernel remaps the virtual processor to another core or hyperthread that can access the needed resource.

2. The computer system of claim 1 wherein remapping the virtual processor comprises transferring a continuation to a remote node, wherein the continuation comprises a representation of a state of the virtual processor.

3. The computer system of claim 2 wherein the hyper-kernel provides the remote node with information usable to build the continuation.

4. The computer system of claim 2 wherein when the remote node is in possession of the needed resource, the continuation is placed in a hierarchical data structure usable to schedule execution of the continuation.

5. The computer system of claim 2 wherein when the remote node needs additional resources to handle the continuation, the continuation is placed in an event table local to the remote node.

6. A method, comprising:

receiving, at a hyper-kernel on a hosting physical processor, an indication that a core or hyperthread which instantiates one or more virtual processors cannot access a resource needed by the core or hyperthread, the needed resource comprising at least one of: (1) a portion of a physical memory needed by the core or hyperthread, and (2) an I/O resource needed by the core or hyperthread; and

remapping, by the hyper-kernel, the virtual processor to another core or hyperthread that can access the needed resource.

7. The method of claim 6 wherein remapping the virtual processor comprises transferring a continuation to a remote node, wherein the continuation comprises a representation of a state of the virtual processor.

8. The method of claim 7 wherein the hyper-kernel provides the remote node with information usable to build the continuation.

9. The method of claim 7 wherein when the remote node is in possession of the needed resource, the continuation is placed in a hierarchical data structure usable to schedule execution of the continuation.

10. The method of claim 7 wherein when the remote node needs additional resources to handle the continuation, the continuation is placed in an event table local to the remote node.

11. A computer system, comprising:

one or more I/O devices;

physical memory; and

a plurality of physical nodes, wherein an operating system is run collectively across the plurality of physical nodes;

wherein each physical node includes one or more physical processors, and wherein each physical processor has one or more cores, and wherein each core has one or more hyperthreads;

wherein each core or hyperthread instantiates one or more virtual processors, wherein a virtual processor comprises a computing engine visible to the operating system run collectively across the plurality of physical nodes;

wherein each core or hyperthread is configured to invoke a hyper-kernel on its hosting physical processor when the core or hyperthread cannot access a resource needed by the core or hyperthread, the needed resource comprising at least one of: (1) a portion of the physical memory needed by the core or hyperthread, and (2) an I/O resource needed by the core or hyperthread; and

wherein the hyper-kernel moves the needed resource to a location accessible by the hosting physical processor.

12. The computer system of claim 11 wherein the hyper-kernel requests the needed resource at least in part by sending a message to a remote node.

13. The computer system of claim 12 wherein the hyper-kernel determines the remote node based at least in part on a resource map.

14. The computer system of claim 11 wherein the hyper-kernel generates a continuation, and wherein the continuation comprises a representation of the virtual processor.

15. The computer system of claim 14 wherein the hyper-kernel places the continuation in an event table.

16. A method, comprising:

receiving, at a hyper-kernel on a hosting physical processor, an indication that a core or hyperthread which instantiates one or more virtual processors cannot access a resource needed by the core or hyperthread, the needed resource comprising at least one of: (1) a portion of a physical memory needed by the core or hyperthread, and (2) an I/O resource needed by the core or hyperthread; and

moving, by the hyper-kernel, the needed resource to a location accessible by the hosting physical processor.

17. The method of claim 16 wherein the hyper-kernel requests the needed resource at least in part by sending a message to a remote node.

18. The method of claim 17 wherein the hyper-kernel determines the remote node based at least in part on a resource map.

19. The method of claim 16 wherein the hyper-kernel generates a continuation, and wherein the continuation comprises a representation of the virtual processor.

20. The method of claim 19 wherein the hyper-kernel places the continuation in an event table.