Efficient sharing of memory between applications running under different operating systems on a shared hardware system

• US 20050246505A1
• Filed: 04/29/2004
• Published: 11/03/2005
• Est. Priority Date: 04/29/2004
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: McKenney, Paul E., Krieger, Orran Y., Ostrowski, Michal
• Granted Patent: US 7,478,204 B2
• Time in Patent Office: Days
• US Class Current: 711/151
• CPC Class Codes: G06F 12/1018 involving hashing techniques, e.g. inverted page tables
• CPC Class Codes: G06F 12/1036 for multiple virtual address spaces, e.g. segmentation G06F12/1045 takes precedence
• CPC Class Codes: G06F 2212/656 Address space sharing
• CPC Class Codes: G06F 9/52 Program synchronisation; Mutual exclusion, e.g. by means of semaphores

Abstract

A system, method and computer program product for efficient sharing of memory between first and second applications running under first and second operating systems on a shared hardware system. The hardware system runs a hypervisor that supports concurrent execution of the first and second operating systems, and further includes a region of shared memory managed on behalf of the first and second applications. Techniques are used to avoid preemption when the first application is accessing the shared memory region. In this way, the second application will not be unduly delayed when attempting to access the shared memory region due to delays stemming from the first application'"'"'s access of the shared memory region. This is especially advantageous when the second application and operating system are adapted for real-time processing. Additional benefits can be obtained by taking steps to minimize memory access faults.

Claims

1. A system for efficient sharing of memory between applications running under different operating systems on a shared hardware system, comprising:

a first operating system;

a first application adapted to execute under control of said first operating system;

a second real-time operating system;

a second real-time application adapted to execute under control of said second operating system;

a hypervisor supporting concurrent execution of said first and second operating systems on a system of shared hardware resources;

a region of shared memory managed on behalf of said first and second applications; and

means for avoiding preemption of said first application when it is accessing said shared memory region;

whereby said second application will not be unduly delayed when attempting to access said shared memory region due to delays stemming from said first application'"'"'s access of said shared memory region.

2. A system in accordance with claim 1 further including means for minimizing memory access faults when said first application is accessing said shared memory region.

3. A system in accordance with claim 1 wherein said means for avoiding preemption includes first means for indicating that said first application accessing said shared memory region is not to be preempted, and second means for indicating that preemption of said first application has been deferred.

4. A system in accordance with claim 3 wherein said first means comprises a do_not_preempt indicator mechanism that is controlled by said first application.

5. A system in accordance with claim 3 wherein said first means comprises a unique program counter range associated with a critical code section implemented by said first application when it accesses said shared memory region.

6. A system in accordance with claim 3 wherein said second means comprises a need_to_preempt indicator mechanism that is controlled by one or both of said first operating system and said hypervisor.

7. A system in accordance with claim 2 wherein said means for minimizing memory access faults comprises means for fixing said shared memory region in physical memory.

8. A system in accordance with claim 2 wherein said means for minimizing memory access faults comprises means for minimizing translation look-aside buffer faults, segment faults, inverted-page-table faults and cache misses.

9. A system in accordance with claim 2 wherein said means for minimizing memory access faults comprises means for placing data needed by said first application in a buffer area of said shared memory region while accessing said shared memory region.

10. A system in accordance with claim 2 wherein said means for minimizing memory access faults comprises one of a means for placing code and stack pages used by said application in a buffer area of said shared memory region, or an alternative means for pinning code and stack pages used by said application.

11. A method for efficient sharing of memory between applications running under different operating systems on a shared hardware system, comprising:

concurrently executing a first operating system, a first application adapted to execute under control of said first operating system, a second real-time operating system and a second real-time application adapted to execute under control of said second operating system, on a system of shared hardware resources with support from a hypervisor;

managing a region of shared memory on behalf of said first and second applications; and

avoiding preemption of said first application when it is accessing said shared memory region;

whereby said second application will not be unduly delayed when attempting to access said shared memory region due to delays stemming from said first application'"'"'s access of said shared memory region.

12. A method in accordance with claim 11 further including minimizing memory access faults when said first application is accessing said shared memory region.

13. A method in accordance with claim 11 wherein said step of avoiding preemption includes a first step of indicating that said first application accessing said shared memory region is not to be preempted, and a second step of indicating that preemption of said first application has been deferred.

14. A method in accordance with claim 13 wherein said first indicating step comprises setting a do_not_preempt indicator that is controlled by said first application.

15. A method in accordance with claim 13 wherein said first indicating step comprises using a unique program counter range associated with a critical code section implemented by said first application when it accesses said shared memory region.

16. A method in accordance with claim 13 wherein said second indicating step comprises setting a need_to_preempt indicator that is controlled by one or both of said first operating system and said hypervisor.

17. A method in accordance with claim 12 wherein said step of minimizing memory access faults comprises fixing said shared memory region in physical memory.

18. A method in accordance with claim 12 wherein said step of minimizing memory access faults comprises minimizing translation look-aside buffer faults, segment faults, inverted-page-table faults and cache misses.

19. A method in accordance with claim 12 wherein said step of minimizing memory access faults comprises placing data needed by said first application in a buffer area of said shared memory region while accessing said shared memory region.

20. A method in accordance with claim 12 wherein said step of minimizing memory access faults comprises one of a step of placing code and stack pages used by said application in a buffer area of said shared memory region, or an alternative step of pinning code and stack pages used by said application.

21. A computer program product for efficient sharing of memory between applications running under different operating systems on a shared hardware system, comprising:

one or more data storage media;

programming means recorded on said data storage media for programming a system of shared hardware resources to operate as by;

concurrently executing a first operating system, a first application adapted to execute under control of said first operating system, a second real-time operating system and a second real-time application adapted to execute under control of said second operating system, on a system of shared hardware resources with support from a hypervisor;

managing a region of shared memory on behalf of said first and second applications; and

avoiding preemption of said first application when it is accessing said shared memory region;

whereby said second application will not be unduly delayed when attempting to access said shared memory region due to delays stemming from said first application'"'"'s access of said shared memory region.

22. A computer program product in accordance with claim 21 wherein said programming means is adapted to program said system of shared hardware resources to operate as by minimizing memory access faults when said first application is accessing said shared memory region.

23. A computer program product in accordance with claim 21 wherein said programming means is adapted to program said system of shared hardware resources such that said step of avoiding preemption includes a first step of indicating that said first application accessing said shared memory region is not to be preempted, and a second step of indicating that preemption of said first application has been deferred.

24. A computer program product in accordance with claim 23 wherein said programming means is adapted to program said system of shared hardware resources such that said first indicating step comprises setting a do_not_preempt indicator that is controlled by said first application.

25. A computer program product in accordance with claim 23 wherein said programming means is adapted to program said system of shared hardware resources such that said first indicating step comprises using a unique program counter range associated with a critical code section implemented by said first application when it accesses said shared memory region.

26. A computer program product in accordance with claim 23 wherein said programming means is adapted to program said system of shared hardware resources such that said second indicating step comprises setting a need_to_preempt indicator that is controlled by one or both of said first operating system and said hypervisor.

27. A computer program product in accordance with claim 22 wherein said programming means is adapted to program said system of shared hardware resources such that said step of minimizing memory access faults comprises fixing said shared memory region in physical memory.

28. A computer program product in accordance with claim 22 wherein said programming means is adapted to program said system of shared hardware resources such that said step of minimizing memory access faults comprises minimizing translation look-aside buffer faults, segment faults, inverted-page-table faults and cache misses.

29. A computer program product in accordance with claim 22 wherein said programming means is adapted to program said system of shared hardware resources such that said step of minimizing memory access faults comprises placing data needed by said first application in a buffer area of said shared memory region while accessing said shared memory region.

30. A computer program product in accordance with claim 22 wherein said programming means is adapted to program said system of shared hardware resources such that said step of minimizing memory access faults comprises one of a step of placing code and stack pages used by said application in a buffer area of said shared memory region, or an alternative step of pinning code and stack pages used by said application.

31. An application computer program product for implementing a first application that runs under a first operating system while efficiently sharing memory with a second real-time application that runs under a second real-time operating system on a shared hardware system running a hypervisor to support concurrent execution of said operating systems, and in which a region of shared memory is managed for use by said applications, comprising:

one or more data storage media;

programming means recorded on said data storage media for programming said system of shared hardware resources to operate as by;

said first application providing a first indication to said first operating system or said hypervisor that said first application is not to be preempted when it is accessing said shared memory region; and

said first application responding to a second indication from said first operating system or said hypervisor that said first application needs to be preempted when it finishes accessing said shared memory region;

whereby said second application will not be unduly delayed when attempting to access said shared memory region due to delays stemming from said first application'"'"'s access of said shared memory region.

32. An operating system computer program product for implementing a first operating system that runs a first application which efficiently shares memory with a second real-time application that runs under a second real-time operating system on a shared hardware system running a hypervisor to support concurrent execution of said operating systems, and in which a region of shared memory is managed for use by said applications, comprising:

one or more data storage media;

programming means recorded on said data storage media for programming said system of shared hardware resources to operate as by;

said first operating system checking a first indication that said first application is not to be preempted when it is accessing said shared memory region; and

said first operating system providing a second indication to said first application that said first application needs to be preempted when it finishes accessing said shared memory region;

whereby said second application will not be unduly delayed when attempting to access said shared memory region due to delays stemming from said first application'"'"'s access of said shared memory region.

33. An operating system computer program product in accordance with claim 32 wherein said programming means is further adapted to program said system of shared hardware resources such that said first operating system determines whether said first application has permission to suppress preemption.

34. An operating system computer program product in accordance with claim 33 wherein said programming means is further adapted to program said system of shared hardware resources such that said first operating system provides an indication of whether said first application has permission to suppress preemption to said hypervisor.

36. A hypervisor computer program product in accordance with claim 34 wherein said programming means is further adapted to program said system of shared hardware resources such that said hypervisor receives an indication of whether said first application has permission to suppress preemption from said first operating system.

37. A hypervisor computer program product in accordance with claim 36 wherein said programming means is further adapted to program said system of shared hardware resources such that said hypervisor determines whether said first application has permission to suppress preemption.

35. A hypervisor computer program product for implementing a hypervisor that supports concurrent execution of a first application running under a first operating system that efficiently shares a common memory region with a second real-time application running under a second real-time operating system on a shared hardware system that runs said hypervisor, comprising:

one or more data storage media;

programming means recorded on said data storage media for programming said system of shared hardware resources to operate as by;

said hypervisor checking a first indication that said first application is not to be preempted when it is accessing said shared memory region; and

said hypervisor providing a second indication to said first application that said first application needs to be preempted when it finishes accessing said shared memory region;

whereby said second application will not be unduly delayed when attempting to access said shared memory region due to delays stemming from said first application'"'"'s access of said shared memory region.

38. A system for efficient sharing of memory between applications running under different operating systems on a shared hardware system, comprising:

a first operating system;

a first application adapted to execute under control of said first operating system;

a second operating system;

a second application adapted to execute under control of said second operating system;

a hypervisor supporting concurrent execution of said first and second operating systems on a system of shared hardware resources;

a region of shared memory managed on behalf of said first and second applications;

means for avoiding preemption of said first application when it is accessing said shared memory region; and

means for minimizing memory access faults when said first application is accessing said shared memory region;

whereby said second application will not be unduly delayed when attempting to access said shared memory region due to delays stemming from said first application'"'"'s access of said shared memory region.