Using writeable page tables for memory address translation in a hypervisor environment

• US 7,814,287 B2
• Filed: 07/17/2006
• Issued: 10/12/2010
• Est. Priority Date: 07/15/2005
• Status: Active Grant

• Current Assignee: Citrix Systems, Inc. (Cloud Software Group)
• Original Assignee: XenSource, Inc. (Cloud Software Group)
• Inventors: Pratt, Ian
• Primary Examiner(s): Chace; Christian P
• Assistant Examiner(s): Namazi; Mehdi
• Application Number:
  US11/487,945
• Publication Number:
  US 20070016755A1
• Time in Patent Office: 1,548 Days
• Field of Search: 711/207, 711/163, 711/203, 711/205, 711/6, 718/1, 719/319
• US Class Current: 711/163
• CPC Class Codes: G06F 12/1009 using page tables, e.g. page table structures
• CPC Class Codes: G06F 12/1027 using associative or pseudo-associative address translation means, e.g. translation look-aside buffer [TLB]
• CPC Class Codes: G06F 12/1036 for multiple virtual address spaces, e.g. segmentation G06F12/1045 takes precedence
• CPC Class Codes: G06F 2212/151 Emulated environment, e.g. virtual machine
• CPC Class Codes: G06F 2212/682 Multiprocessor TLB consistency

Abstract

A method and system for using writeable page tables to increase performance of memory address translation in computing environments utilizing a hypervisor. Guest operating systems are given temporary write-access to a page table page after the system confirms that such page is not part of the current address space (i.e., confirming that the page is part of a different page table from the one that is currently in use, such as a different user-space process). Alternatively, if the page is part of the currently running page table, the system invalidates the appropriate entry in the root page directory, thus “unlinking” it, and ensuring that the appropriate region of virtual address space is flushed from the translation lookaside buffer (TLB) in the current CPU and others that may be using it. After giving the OS write-access, the page is added to a validation queue. Validation of all 1024 entries and “re-hooking” of the page occurs whenever a page fault is taken due to the page being unhooked, or before context switching to another page table.

Claims

1. A method of managing memory in a processing system capable of simultaneously executing, within at least one virtual machine, multiple instantiations of one or more guest operating systems, without duplication of page tables provided by the processing system, said method comprising:

requesting, by a guest operating system executing within a virtual machine, access to write to a page in a page table provided by the processing system;

determining, by a hypervisor executing on the processing system, whether the requested page table page to which the guest operating system is attempting to write is part of a page table page currently in use by the processing system;

upon determining that said page table page is part of a page table page currently in use by said processing system, unlinking, by the hypervisor, said page table page from a page directory to which said page table page belongs; and

granting said guest operating system temporary write access to said page table page.

2. The method of claim 1, further comprising:

validating entries in said page table page to confirm that such entries refer to a page table page attributed to said guest operating system, and that such entries do not act to create a writeable mapping to any other page table page.

3. The method of claim 2, further comprising:

upon said guest operating system attempting to write to said page table page, generating a copy of the contents of such page table page;

said validating step further comprising comparing said copy to said page table page after said guest operating system has modified said page table page, and validating only those entries in said page table page that are different from said copy.

4. The method of claim 2, wherein said validating step is initiated in response to an attempt by said guest operating system to access a virtual memory address that requires said page table page.

5. The method of claim 2, wherein said validating step is initiated in response to said guest operating system attempting to context switch from a first process to a second process.

6. The method of claim 2, further comprising:

before said validating step, terminating said temporary write access granted to said guest operating system so as to allow said guest operating system read-only access to said page table page; and

after said validating step, linking said page table page back to said page directory to which said page table page belongs.

7. The method of claim 6, further comprising after said linking step, causing said guest operating system to reissue said instruction.

8. The method of claim 1, wherein said unlinking step further comprises invalidating an entry in said page directory corresponding to said page table page.

9. The method of claim 8, said invalidating step further comprising converting bits in the entry in said page directory corresponding to said page table page to 0.

10. The method of claim 1, said determining step further comprising:

detecting an offset of said page table page in said page directory;

determining a page table page entry in a page directory currently in use by hardware on which said processing system is operating at a location in said page directory designated by said offset; and

determining whether bits in said page table page entry refer to said page table page.

11. A memory having computer readable program code means embodied therein for causing simultaneous execution, within at least one virtual machine, of multiple instantiations of one or more guest operating systems, without duplication of page tables provided by a processing system, the memory comprising:

means for determining, by a hypervisor, whether a page table page to which a guest operating system, executing within a virtual machine, is attempting to write is part of a page table page currently in use by said processing system;

means for unlinking, by the hypervisor, said page table page from a page directory to which said page table page belongs, when a determination is made that said page table page is part of a page table page currently in use by said processing system; and

means for granting, by the hypervisor, said guest operating system temporary write access to said page table page.

12. A processing system for simultaneous execution, within at least one virtual machine, of multiple instantiations of one or more guest operating systems, without duplication of page tables provided by the processing system, comprising:

means for determining whether a page table page to which a guest operating system, executing within a virtual machine, is attempting to write is part of a page table page currently in use by said processing system;

means for unlinking, by the hypervisor, said page table page from a page directory to which said page table page belongs, when a determination is made that said page table page is part of a page table page currently in use by said processing system; and

means for granting said guest operating system temporary write access to said page table page.

13. The system of claim 12 further comprising means for validating entries in said page table page to confirm that such entries refer to a page table page attributed to said guest operating system and that such entries do not act to create a writeable mapping to any other page table page.

14. The system of claim 13 further comprising:

means for generating a copy of the contents of such page table page, upon attempting, by said guest operating system, to write to said page table page; and

means for comparing said copy to said page table page after said guest operating system has modified said page table page and validating only those entries in said page table page that are different from said copy.

15. The system of claim 13, wherein said means for validating entries further comprises means for validating entries in response to an attempt by said guest operating system to access a virtual memory address that requires said page table page.

16. The system of claim 13, wherein said means for validating entries further comprising means for validating entries initiated in response to said guest operating system attempting to context switch from a first process to a second process.

17. The system of claim 13, further comprising:

means for terminating, before validating the entries in said page table page, said temporary write access granted to said guest operating system so as to allow said guest operating system read-only access to said page table page; and

means for linking, after validating the entries in said page table page, said page table page back to said page directory to which said page table page belongs.

18. The system of claim 17, further comprising means for causing, after linking, said guest operating system to reissue said instruction.

19. The system of claim 12 further comprising means for invalidating an entry in said page directory corresponding to said page table page.

20. The system of claim 19 further comprising means for converting bits in the entry in said page directory corresponding to said page table page to 0.

21. The system of claim 12, further comprising:

means for detecting an offset of said page table page in said directory;

means for determining a page table page entry in a page directory currently in use by hardware on which said processing system in operating at a location in said page directory designated by said offset; and

means for determining whether bits in said page table page entry refer to said page table page.