Optimization of paging cache protection in virtual environment

1. A system for virtualizing a processor and protecting paging cache, comprising:

a virtualization system running on a computer system and controlling memory paging through hardware support for maintaining real paging structures;

a Virtual Machine (VM) running guest code and having at least one set of guest paging structures that are mapped to guest physical pages in guest address space using linear addresses; and

a Virtual Machine Monitor (VMM) paging cache comprising mapped paging tables corresponding to pages of a guest physical memory,

wherein the VMM paging cache is validated by recognizing Translation Lookaside Buffer (TLB) invalidation events initiated by the guest OS.

2. The system of claim 1 , wherein at least a part of the guest paging tables is protected by a write protection bit.

3. The system of claim 2 , wherein the guest paging structures and the real paging structures are organized into a hierarchy of layers.

4. The system of claim 3 , further comprising an additional layer that utilizes a CR3 value for caching of paging translations for different values of CR3.

5. The system of claim 2 , wherein the protection bit is an RW (read/write)-bit.

6. The system of claim 1 , wherein at least some TLB invalidation events comprise any of:

the TLB is invalidated by performing an INVLPG instruction for page linear address; and

the TLB is reset by writing into a control register CR3; and

the TLB is reset by modifying a PGE bit (Page Global Enable) in a control register CR4.

7. The system of claim 6 , wherein the paging tables comprise active paging tables having a current CR3 register pointing to them and non-active paging tables that do not have a current CR3 register pointing to them.

8. The system of claim 7 , wherein the non-active paging tables are validated and protected only upon their activation by loading a value into the CR3 register.

9. The system of claim 1 , further comprising means for tracking Access (A) and Dirty (D) bits.

10. The system of claim 9 , wherein the A and D bits are synchronized in guest paging tables entries.

11. A method for optimization of paging cache protection, the method being performed on a computer having a processor and a memory, the method comprising:

initiating a virtualization system running on a computer system and controlling memory paging through hardware support for maintaining real paging structures;

launching a Virtual Machine (VM) running guest code and having at least one set of guest paging structures that correspond to guest physical pages in guest virtualized linear address space,

wherein at least some of the guest paging structures are mapped to the real paging structures; and

generating a Virtual Machine Monitor (VMM) paging cache comprising mapped paging tables corresponding to pages of a guest physical memory,

protecting the paging tables by using a write protection bit;

validating the VMM paging cache by tracking TLB resets and performing INVLPG instructions;

clearing the VMM paging cache by writing into a current control register CR3; and

tracking Access (A) and Dirty (D) bits and synchronizing the A and D bits in a guest physical memory,

wherein the VMM paging cache is validated by recognizing Translation Lookaside Buffer (TLB) invalidation events initiated by the guest OS.

12. The method of claim 11 , wherein at least some TLB invalidation events comprise any of:

the TLB is invalidated by performing an INVLPG instruction for page linear address; and

the TLB is reset by writing into a control register CR3; and

the TLB is reset by modifying a PGE bit (Page Global Enable) in a control register CR4.

13. The method of claim 11 , wherein the write protection bit is an RW (read/write)-bit.

14. The method of claim 11 , further comprising, protecting global pages by modifying a CR4.GE bit.

15. The method of claim 11 , wherein, if mapping entries become unsynchronized, protecting the paging root tables by setting the RW-bit and applying the INVLPG instructions to page directory and page table entries.

16. A method for optimization of paging cache protection, the method being performed on a computer having a processor and a memory, the method comprising:

initiating a virtualization system running on a computer system and controlling memory paging through hardware support for maintaining real paging structures;

launching a Virtual Machine (VM) running guest code and having at least one set of guest paging structures that correspond to guest physical pages in guest virtualized linear address space,

wherein at least some of the guest paging structures are mapped to the real paging structures using double-mapping;

generating a Virtual Machine Monitor (VMM) paging cache comprising mapped paging tables corresponding to pages of a guest physical memory;

protecting the paging tables by setting an Access bit (A) for each page as read-only;

executing re-mapping of the pages upon page access by resetting a Dirty bit (D);

resetting the Access (A) bit to read-write; and

synchronizing the A and D bits in a guest physical memory.

17. The method of claim 16 , wherein the Access (A) bit is set as read-only regardless of a page access status.