Micro-virtualization architecture for threat-aware microvisor deployment in a node of a network environment

• US 9,292,686 B2
• Filed: 03/28/2014
• Issued: 03/22/2016
• Est. Priority Date: 01/16/2014
• Status: Active Grant

• Current Assignee: Fireeye Security Holdings US LLC
• Original Assignee: FireEye, Inc.
• Inventors: Ismael, Osman Abdoul, Aziz, Ashar
• Primary Examiner(s): Gee, Jason K.
• Assistant Examiner(s): Zhu, Zhimei
• Application Number:
  US14/229,626
• Publication Number:
  US 20150199532A1
• Time in Patent Office: 725 Days
• US Class Current: 1/1
• CPC Class Codes: G06F 2009/45587 Isolation or security of virtual machine instances
• CPC Class Codes: G06F 21/53 by executing in a restricted environment, e.g. sandbox or secure virtual machine
• CPC Class Codes: G06F 21/552 involving long-term monitoring or reporting
• CPC Class Codes: G06F 21/629 to features or functions of an application
• CPC Class Codes: G06F 9/45533 Hypervisors; Virtual machine monitors
• CPC Class Codes: G06F 9/45558 Hypervisor-specific management and integration aspects
• CPC Class Codes: G06F 9/5027 the resource being a machine, e.g. CPUs, Servers, Terminals

Abstract

A micro-virtualization architecture deploys a threat-aware microvisor as a module of a virtualization system configured to facilitate real-time security analysis, including exploit detection and threat intelligence, of operating system processes executing in a memory of a node in a network environment. The micro-virtualization architecture organizes the memory as a user space and kernel space, wherein the microvisor executes in the kernel space of the architecture, while the operating system processes, an operating system kernel, a virtual machine monitor (VMM) and its spawned virtual machines (VMs) execute in the user space. Notably, the microvisor executes at the highest privilege level of a central processing unit of the node to virtualize access to kernel resources. The operating system kernel executes under control of the microvisor at a privilege level lower than a highest privilege level of the microvisor. The VMM and its spawned VMs execute at the highest privilege level of the microvisor.

Claims

1. A system comprising:

a central processing unit (CPU) adapted to execute a process, a single instance of an operating system kernel, a virtual machine monitor (VMM) and a microvisor;

a memory configured to store the process, the operating system kernel, the VMM and the microvisor as a micro-virtualization architecture that organizes the memory as a user space and a kernel space, wherein the process, the operating system kernel and the VMM execute in the user space of the architecture, and wherein the microvisor executes in the kernel space of the architecture;

the microvisor disposed beneath the operating system kernel and configured to communicate with the VMM over a privileged interface, the microvisor further configured to execute at a highest privilege level of the CPU to control access permissions to a plurality of kernel resources accessible by the process; and

the VMM configured to execute at a highest privilege level of the microvisor to perform a one-to-one mapping between the plurality of kernel resources and the operating system kernel and to expose the mapped kernel resources to the operating system kernel, the operating system kernel configured to execute at a privilege level lower than the highest privilege level of the microvisor, the VMM configured to instantiate a virtual machine containing the operating system kernel, the VMM further configured to instantiate a micro-virtual machine restricted to containing the process, wherein access to the mapped kernel resources is controlled by the VMM among the virtual machine and the micro-virtual machine.

2. The system of claim 1 wherein the operating system kernel comprises an operating system (OS) specific VMM extension adapted to communicate with the VMM, the OS specific VMM extension configured to provide an interface to the VMM that allows introspection of contents of internal structures of the operating system kernel.

3. The system of claim 2 wherein introspection of the contents of the internal structures comprises examination of data structures of the operating system kernel without duplicating the structures.

4. The system of claim 2 wherein the OS specific VMM extension communicates with the VMM over an application programming interface.

5. The system of claim 2 wherein the microvisor comprises one or more protection domains, each protection domain including one or more execution contexts, each execution context tightly linked to a scheduling context and configured to interact with capabilities having contents that specify access control permissions to the kernel resources by the process.

6. The system of claim 5 wherein the scheduling context includes information defining a priority for execution of the execution context on the CPU and wherein the microvisor further comprises a global scheduler configured to cooperate with the scheduling context to schedule the execution context for execution on the CPU.

7. The system of claim 6 wherein the defined priority is implemented by one or more queues and wherein the global scheduler examines the queues to dispatch the scheduling context to the CPU for execution of its linked execution context.

8. The system of claim 5 wherein the microvisor further comprises an exception handler configured to service a trap generated by the CPU in response to a capability violation of an access control permission to a kernel resource.

9. The system of claim 8 wherein the trap comprises one of a page fault and a general protection fault.

10. The system of claim 8 wherein the exception handler is further configured to report the capability violation to the VMM.

11. The system of claim 10 wherein the VMM comprises instrumentation logic that is invoked in response to the capability violation to determine whether the process is suspicious.

12. The system of claim 11 wherein, in response to determining that that process is suspicious, the VMM is configured to instantiate the micro-virtual machine for further analysis of the process.

13. A method comprising:

storing a process, a single instance of an operating system kernel, a virtual machine monitor (VMM) and a microvisor in a memory coupled to a central processing unit (CPU) of a node in a computer network;

organizing the memory as a user space and a kernel space of a micro-virtualization architecture;

executing the microvisor in the kernel space of the architecture and at a highest privilege level of the CPU to control access permissions to a plurality of kernel resources accessible by the process;

executing the VMM in the user space of the architecture and at a highest privilege level of the microvisor to perform a one-to-one mapping between the plurality of kernel resources and the operating system kernel and to expose the mapped kernel resources to the operating system kernel;

executing the operating system kernel in the user space of the architecture and at a privilege level lower than the highest privilege level of the microvisor;

instantiating a virtual machine containing the operating system kernel;

instantiating a micro-virtual machine restricted to containing the process; and

controlling access to the mapped kernel resources among the virtual machine and the micro-virtual machine.

14. The method of claim 13 further comprising:

organizing the microvisor to include a first protection domain bound to the VMM, the first protection domain having one or more execution contexts, each execution context tightly linked to a scheduling context and configured to interact with capabilities having contents that specify access control permissions to the kernel resources by the process.

15. The method of claim 14 further comprising:

switching privilege levels in response to an interception point, wherein the interception point includes one of a memory access request, a function call and a system call.

16. The method of claim 15 wherein instantiating the virtual machine further comprises:

instantiating first instrumentation logic of the virtual machine directed to determination of an exploit in the process.

17. The method of claim 16 wherein executing the VMM further comprises:

executing second instrumentation logic of the VMM to analyze the interception point originated by the process to invoke services of the operating system kernel.

18. The method of claim 17 wherein executing the VMM further comprises:

in response to the interception point, assuming control over the operating system kernel at the VMM to enable monitoring of activity of the process.

19. The method of claim 18 wherein executing the VMM further comprises:

enabling communication between the operating system kernel and microvisor over a privileged interface embodied as a set of hyper-calls.

20. A system comprising:

a central processing unit (CPU) adapted to execute a plurality of user mode processes, a single instance of an operating system kernel, a type 0 virtual machine monitor (VMM 0) and a microvisor;

a memory configured to store the user mode processes, the operating system kernel, the VMM 0 and the microvisor as a micro-virtualization architecture that organizes the memory as a user space and a kernel space, wherein the user mode processes, the operating system kernel and the VMM 0 execute in the user space of the architecture, and wherein the microvisor executes in the kernel space of the architecture;

the microvisor disposed beneath the operating system kernel and configured to communicate with the VMM 0 over a privileged interface, the microvisor further configured to execute at a highest privilege level of the CPU to control access permissions to a plurality of kernel resources accessible by the user mode processes;

the VMM 0 including instrumentation logic configured to analyze a system call issued by a first process to invoke services of the operating system kernel that include accesses to the kernel resources, the VMM 0 configured to execute at a highest privilege level of the to perform a one-to-one mapping between the plurality of kernel resources and the operating system kernel and to expose the mapped kernel resources to the operating system kernel, the VMM 0 configured to instantiate a virtual machine containing the operating system kernel, the VMM 0 further configured to instantiate a plurality of micro-virtual machines, each micro-virtual machine restricted to containing one of the processes, wherein access to the mapped kernel resources is controlled among the virtual machine and the micro-virtual machines by the VMM 0; and

the operating system kernel including an operating system specific VMM 0 extension adapted to communicate with the VMM 0, the operating system kernel configured to execute at a privilege level lower than the highest privilege level of the microvisor.

21. A non-transitory computer readable media containing instructions for execution on a processor for a method comprising:

storing a plurality of processes, a single instance of an operating system kernel, a virtual machine monitor (VMM) and a microvisor in a memory coupled to a central processing unit (CPU) of a node in a computer network;

organizing the memory as a user space and a kernel space of a micro-virtualization architecture;

executing the microvisor in the kernel space of the architecture and at a highest privilege level of the CPU to control access permissions to a plurality of kernel resources accessible by the processes;

executing the VMM in the user space of the architecture and at a highest privilege level of the microvisor to perform a one-to-one mapping between the plurality of kernel resources and the operating system kernel and to expose the mapped kernel resources to the operating system kernel;

executing the operating system kernel in the user space of the architecture and at a privilege level lower than the highest privilege level of the microvisor;

instantiating a virtual machine containing the operating system kernel;

instantiating a plurality of micro-virtual machines, each micro-virtual machine restricted to containing one of the processes; and

controlling access to the kernel resources among the virtual machine and the micro-virtual machine.