Method and apparatus for shared I/O in a load/store fabric

• US 7,046,668 B2
• Filed: 01/14/2004
• Issued: 05/16/2006
• Est. Priority Date: 01/21/2003
• Status: Active Grant

• Current Assignee: Mellanox Technologies Limited (NVIDIA Corporation)
• Original Assignee: NextIO Incorporated
• Inventors: Pagan, Annette, Pekkala, Richard E., Utley, Robert Haskell, Pettey, Christopher J., Khan, Asif
• Primary Examiner(s): NGUYEN, BRIAN D
• Application Number:
  US10/757,714
• Publication Number:
  US 20040179529A1
• Time in Patent Office: 853 Days
• Field of Search: 370/392, 370/389, 370/352, 370/401, 370/254, 370/360, 370/381, 370/395.7, 370/395.71, 370/395.72, 370/412, 370/429, 370/471, 709/203, 709/245, 709/249, 709/223, 710/2, 710/100, 710/300, 710/316
• US Class Current: 370/392
• CPC Class Codes: G06F 13/12 using hardware independent of the central processor, e.g. channel or peripheral processor

Abstract

An apparatus and method is provided for allowing I/O devices to be shared and/or partitioned among a plurality of processing complexes within the load/store fabric of each of the processing complexes without requiring modification to the operating system or driver software of the processing complexes. The apparatus and method includes a switch for selectively coupling each of the processing complexes to one or more shared I/O devices. The apparatus and method further includes placing information within packets transmitted between the switch and the I/O devices to identify which of the processing complexes the packets are associated with. The invention further includes an apparatus and method within the shared I/O devices to allow the shared I/O devices to service each of the processing complexes independently.

Claims

1. A method for transferring a packet from a shared input/output (I/O) endpoint to one of a plurality of OS Domains, within a load/store fabric of the plurality of OS Domains, comprising:

embedding an OS Domain number with the packet to associate the packet with one of the plurality of OS Domains;

transferring the packet with the embedded OS Domain number to a shared I/O switch;

examining the embedded OS Domain number to determine a port within the shared I/O switch associated with the one of the plurality of OS Domains; and

transferring the packet to the one of the plurality of OS Domains using the port

wherein the load/store fabric maps memory space for the shared I/O switch within memory space of the one of the plurality of OS Domains.

2. The method as recited in claim 1 wherein the shared input/output (I/O) endpoint comprises a network interface controller.

3. The method as recited in claim 1 wherein each of the plurality of OS Domains comprise:

a processing complex; and

memory, coupled to the processing complex for storing data utilized by the processing complex.

4. The method as recited in claim 3 wherein the processing complex comprises one or more processors.

5. The method as recited in claim 1 wherein the load/store fabric utilizes PCI Express.

6. The method as recited in claim 1 wherein said step of embedding comprises:

forming an OS Header field; and

including the OS Header field within the packet for transfer to the shared I/O switch.

7. The method as recited in claim 1 wherein said step of exammining comprises:

performing a table lookup to associate the OS Domain number with a PCI bus hierarchy for the packet; and

determining a port associated with the PCI bus hierarchy which is coupled to the OS Domain for the PCI bus hierarchy.

8. The method as recited in claim 1 wherein the port couples the shared I/O switch to the one of the plurality of OS Domains.