IP Library Granted Patent US 7,660,306
Granted Patent B1
US 7,660,306 · App. 11/330,898 · Granted Feb 9, 2010

Virtualizing the operation of intelligent network interface circuitry

View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 7,660,306
App. No.
11/330,898
Granted
Feb 9, 2010
Kind
B1
Abstract

The present invention is in the field of virtualization of Network Interface Cards and Controllers (NIC) that connect to for example a host computer that can have multiple functions and OS's (referred to as guest functions and guest-OS's) running on it simultaneously and, in particular, relates to the virtualization of the operation of a single NIC such that it can be used simultaneously by multiple guests in such a manner that memory references due to the network traffic originating from and destined to a particular guest is kept separate from other network traffic, and that allows the Media Access Controller (MAC) within the NIC to operate such that it only accepts network packets that are destined to one of the guest-functions or guest-OS's, and the MAC is not required to operate in promiscuous mode where it accept all incoming packets in order to implement the virtualization features.

Claims (63)

1. A method of operating network interface circuitry, wherein the network interface circuitry is configured to couple a host computer to a network to facilitate communication over the network between the host computer and a peer, the method comprising:

by the network interface circuitry,

receiving data from the peer via the network; and

processing the received data, including:

where there is a layer-2 classification rule associated with characteristics of the received data, automatically applying the classification rule to the received data; and

demultiplexing the received layer-2 packets according to destination queues associated with the host computer,

wherein the method further comprises:

applying to the received data a layer-3 classification rule associated with characteristics of the received data, thereby filtering ingress packets that have layer-2 characteristics not consistent with the layer-3 characteristics; and

applying to the received data a VLAN classification rule associated with characteristics of the received data, thereby filtering ingress packets that have layer-2 characteristics and layer-3 characteristics that are not consistent with the VLAN characteristics.

2. A method of operating network interface circuitry, wherein the network interface circuitry is configured to couple a host computer, executing a plurality of guest functions, to a network to facilitate communication over the network between the host computer and a peer, the method comprising:

by the network interface circuitry,

receiving data packets from the peer via the network; and

processing the received data packets, including:

processing a classification rule associated with characteristics of the received data packets and automatically applying the classification rule to the received data packets to determine at least one set of at least one destination queue associated with the host computer; and

demultiplexing the received data packets according to the determined at least one set of at least one destination queue, each separate set of at least one destination queue associated with a separate guest function configured for execution on the host computer, wherein:

the host computer is executing a virtual operating system on top of which a plurality of operating system instances are executing, each operating system instance being an instance of a guest-OS having at least one of the guest-functions associated therewith;

at least one of the set of destination queues is associated with the virtual operating system; and

the classification rules are such that broadcast and/or multicast packets are demultiplexed according to the at least one of the set of destination queues that is associated with the virtual operating system.

3. The method of claim 2 , wherein:

the method further comprises performing protocol processing at the higher layer within the network interface circuitry.

4. The method of claim 2 , wherein:

demultiplexing the received data packets according to separate destination queues associated with the host computer includes directly placing at least a portion of a received packet into memory associated with the separate guest function with which the separate destination queue is associated.

5. The method of claim 2 , wherein:

for at least one of the guest-OS instances, that guest-OS instance is itself a guest-function associated with that guest-OS instance.

6. The method of claim 2 , wherein:

the classification rules are further such that the demultiplexing operation includes providing an indication of at least a particular one of the plurality of guest-functions.

7. The method of claim 2 , wherein:

the classification rules are such that broadcast and/or multicast packets are demultiplexed according to the sets of destination queues configured to receive the broadcast and/or multicast packets.

8. The method of claim 7 , wherein:

the classification rules are such that multicast packets, which cannot be demultiplexed to at least one set of destination queues configured to receive the multicast packets, are filtered.

9. The method of claim 2 , further comprising:

for at least one of the destination queues, controlling a rate at which the data is provided from that destination queue to the host.

10. A method of operating network interface circuitry, wherein the network interface circuitry is configured to couple a host computer, executing a plurality of guest functions, to a network to facilitate communication over the network between the host computer and a peer, the method comprising:

by the network interface circuitry,

receiving data packets from the peer via the network; and

processing the received data packets, including:

processing a classification rule associated with characteristics of the received data packets and automatically applying the classification rule to the received data packets to determine at least one set of at least one destination queue associated with the host computer; and

demultiplexing the received data packets according to the determined at least one set of at least one destination queue, each separate set of at least one destination queue associated with a separate guest function configured for execution on the host computer,

wherein:

the received data packets are network layer packets encapsulating packets at a higher layer than the network layer;

applying the classification rule to the received data packets includes applying a classification rule associated with characteristics of the received data at the network layer, thereby filtering packets that have higher layer characteristics not consistent with the network layer characteristics; and

processing the classification rule associated with characteristics of the received data packets and automatically applying the classification rule to the received data packets to determine at least one set of at least one destination queue associated with the host computer includes classifying the packets according to a network layer header and at least one header associated with the higher layer.

11. A method of operating network interface circuitry, wherein the network interface circuitry is configured to couple a host computer to a network to facilitate communication over the network between the host computer and a peer, the method comprising:

by the network interface circuitry,

receiving data from the peer via the network; and

processing the received data, including:

where there is a layer-2 classification rule associated with characteristics of the received data, automatically applying the classification rule to the received data; and

demultiplexing the received layer-2 packets according to destination queues associated with the host computer,

wherein the method further comprises:

applying to the received data a layer-3 classification rule associated with characteristics of the received data, thereby filtering ingress packets that have layer-2 characteristics not consistent with the layer-3 characteristics; and

applying to the received data a layer-4 classification rule associated with characteristics of the received data, thereby filtering ingress packets that have layer-2 and layer-3 characteristics not consistent with the layer-4 characteristics.

12. The method of claim 11 , wherein:

separate sets of at least one destination queue are associated with separate one of a plurality of guest operating systems executing on the host computer.

13. The method of claim 11 , wherein:

the host computer includes a plurality of processors;

the classification rule includes an indication of a particular one of the plurality of processors; and

demultiplexing the received data packet according to the separate destination queues associated with the host computer includes providing at least a portion of the received data to the indicated particular one of the plurality of processors.

14. The method of claim 11 , wherein:

the layer-4 protocol is TCP and the received data includes a TCP SYN flag.

15. The method of claim 14 , wherein:

the layer-4 classification rule is associated with at least the local TCP port number in the received data.

16. The method of claim 11 , further comprising:

applying to the received data a VLAN classification rule associated with characteristics of the received data, thereby filtering ingress packets that have layer-2 characteristics, layer-3 characteristics and layer-4 characteristics that are not consistent with the VLAN characteristics.

Assignments (9)
RELEASE OF SECURITY INTEREST Recorded Dec 26, 2025
From: CSNK WORKING CAPITAL FINANCE CORP.
To: CHELSIO COMMUNICATIONS, INC.
Reel/Frame 074070/0784 →
RELEASE OF SECURITY INTEREST Recorded Dec 18, 2025
From: WESTERN ALLIANCE BANK, AN ARIZONA CORPORATION
To: CHELSIO COMMUNICATIONS, INC.
Reel/Frame 073264/0525 →
CORRECTIVE ASSIGNMENT TO CORRECT THE THE FIRST PAGE OF THE INTELLECTUAL PROPERTY SECURITY AGREEMENT HAS AN INCORRECT DATE THAT NEEDS TO BE CORRECTED PREVIOUSLY RECORDED ON REEL 050050 FRAME 0396. ASSIGNOR(S) HEREBY CONFIRMS THE SECURITY INTEREST. Recorded Aug 15, 2019
From: CHELSIO COMMUNICATIONS, INC.
To: WESTERN ALLIANCE BANK, AN ARIZONA CORPORATION
Reel/Frame 050112/0234 →
SECURITY INTEREST Recorded Aug 14, 2019
From: CHELSIO COMMUNICATIONS, INC.
To: WESTERN ALLIANCE BANK, AN ARIZONA CORPORATION
Reel/Frame 050050/0396 →
RELEASE OF SECURITY INTEREST Recorded Apr 25, 2017
From: NOVIRIAN CAPITAL
To: CHELSIO COMMUNICATIONS, INC.
Reel/Frame 042142/0232 →
SECURITY INTEREST Recorded Jul 29, 2016
From: CHELSIO COMMUNICATIONS, INC.
To: NOVIRIAN CAPITAL
Reel/Frame 039296/0550 →
RELEASE OF SECURITY INTEREST Recorded Jul 15, 2016
From: EAST WEST BANK
To: CHELSIO COMMUNICATIONS, INC.
Reel/Frame 039360/0596 →
RELEASE OF SECURITY INTEREST Recorded Oct 21, 2014
From: EAST WEST BANK
To: CHELSIO COMMUNICATIONS, INC.
Reel/Frame 033989/0930 →
SECURITY INTEREST Recorded Oct 21, 2014
From: CHELSIO COMMUNICATIONS, INC.
To: SILICON VALLEY BANK
Reel/Frame 033990/0451 →