Chipset support for managing hardware interrupts in a virtual machine system

• US 20060036791A1
• Filed: 10/13/2005
• Published: 02/16/2006
• Est. Priority Date: 09/30/2003
• Status: Active Grant

• Current Assignee: Alain Kagi, Andrew V. Anderson, Erik Cota-Robles, Gilbert Neiger, Richard Uhlig, Stalinselvaraj Jeyasingh, Steven M. Bennett
• Original Assignee: Alain Kagi, Andrew V. Anderson, Erik Cota-Robles, Gilbert Neiger, Richard Uhlig, Stalinselvaraj Jeyasingh, Steven M. Bennett
• Inventors: Anderson, Andrew V., Neiger, Gilbert, Bennett, Steven M., Uhlig, Richard, Jeyasingh, Stalinselvaraj, Kagi, Alain, Cota-Robles, Erik
• Granted Patent: US 7,302,511 B2
• Time in Patent Office: Days
• US Class Current: 710/260
• CPC Class Codes: G06F 13/24 using interrupt G06F13/32 takes precedence
• CPC Class Codes: G06F 9/45533 Hypervisors; Virtual machine monitors
• CPC Class Codes: G06F 9/4812 by interrupt, e.g. masked

Abstract

In one embodiment, an apparatus includes a set of multiplex blocks coupled with an interrupt controller and multiple interrupt request lines, and a virtual machine monitor block (VMM) coupled to the set of multiplex blocks. Each multiplex block corresponds to a distinct interrupt request line. Each multiplex block is to route the interrupt request signal received via the corresponding interrupt request line either to the interrupt controller or the VMM block depending on a current configuration value of this multiplex block.

Claims

1-12. -12. (canceled)

13. A method comprising:

receiving, at a multiplex block, an interrupt request signal from a device via an interrupt request line coupled with the multiplex block; and

determining, based on a current configuration value of the multiplex block, whether the interrupt request signal is to be sent to an interrupt controller or a virtual machine monitor (VMM) block coupled to the multiplex block.

14. The method of claim 13 wherein:

the current configuration value of the multiplex block requires that the interrupt request signal be sent to the interrupt controller if the device is managed by a currently-operating virtual machine (VM); and

the current configuration value of the multiplex block requires that the interrupt request signal be sent to the VMM block if the device is not managed by the currently-operating VM.

15. The method of claim 13 further comprising determining whether the interrupt request signal sent to the VMM block is masked based on data stored in a mask register of the VMM block.

16. The method of claim 13 further comprising updating a corresponding indicator in a status register of the VMM block upon receiving the interrupt request signal at the VMM block.

17. The method of claim 13 further comprising:

determining that the interrupt request signal received at the VMM block is not masked;

asserting an internal signal;

combining the internal signal with an external signal generated by an external signal source; and

transmitting the combined signal to a processor.

18. The method of claim 17 wherein:

the external signal is a non-maskable interrupt (NMI) signal; and

the designated external signal source is a NMI source.

19-21. -21. (canceled)

22. A method comprising:

identifying one or more interrupt request lines that are coupled to one or more devices managed by a virtual machine (VM);

configuring one or more multiplex blocks to route interrupt request signals on the one or more interrupt request lines to an interrupt controller; and

generating a request to transfer control to the VM.

23. The method of claim 22 further comprising:

configuring one or more multiplex blocks to route interrupt request signals on interrupt request lines that are not managed by the VM to a virtual machine monitor (VMM) block;

configuring a mask register in the VMM block to cause blocking of interrupt request signals routed to the VMM block.

24. The method of claim 22 further comprising:

restoring a state saved during a previous operation of the VM in the interrupt controller.

25-27. -27. (canceled)

28. A machine-readable medium containing instructions which, when executed by a processing system, cause the processing system to perform a method, the method comprising:

identifying one or more interrupt request lines that are coupled to one or more devices managed by a virtual machine (VM);

configuring one or more multiplex blocks to route interrupt request signals on the one or more interrupt request lines to an interrupt controller; and

generating a request to transfer control to the VM.

29. The machine-readable medium of claim 28 wherein the method further comprises:

configuring one or more multiplex blocks to route interrupt request signals on interrupt request lines that are not managed by the VM to a virtual machine monitor (VMM) block; and

configuring a mask register in the VMM block to cause masking/unmasking of interrupt request signals routed to the VMM block.

30. The machine-readable medium of claim 28 wherein the method further comprises:

restoring a state saved during a previous operation of the VM in the interrupt controller.