Granted Patent B2

US 8,848,741 · App. 13/529,786 · Granted Sep 30, 2014

Abstract

Claims (54)

1. A method of offloading data intensive tasks from a processor, comprising:

at a processor:

preparing a plurality of TCP packets, each TCP packet comprising a TCP header and a data payload,

adding a prepend header to each of the TCP packets, the prepend header comprising a TCP segmentation flag and a maximum segment size value indicating a maximum segment size for TCP segmentation; and

transmitting each TCP packet to a configurable logic device (CLD); and

at the CLD:

receiving a TCP packet from the processor;

determining whether to generate a set of TCP segment packets based on the TCP segmentation flag for the TCP packet;

generating a set of TCP segment packets if the TCP segmentation flag indicates that TCP segment packets are to be generated, each TCP segment packet being no longer than the maximum segmentation size for the TCP packet and comprising:

a copy of the TCP header,

an incrementing segment sequence identifier, and

a portion of the data payload;

transmitting the set of TCP segment packets to an external network interface; and

repeating the receiving, determining, generating, and transmitting steps for a plurality of TCP packets received from the processor.

2. The method of claim 1 , wherein the CLD is a field programmable gate array.

3. A tangible, non-transitory computer-readable media comprising:

instructions that when executed by a processor are operable to:

prepare a plurality of TCP packets, each TCP packet comprising a TCP header and a data payload,

adding a prepend header to each of the TCP packets, the prepend header comprising a TCP segmentation flag and a maximum segment size value indicating a maximum segment size for TCP segmentation, and

transmit each TCP packet to a configurable logic device (CLD); and

a configuration file that when loaded by a configurable logic device (CLD) configures the CLD to receive a plurality of TCP packets from the processor and for each received TCP packet to:

determine whether to generate a set of TCP segment packets based on the TCP segmentation flag for the TCP packet;

generate a set of TCP segment packets if the TCP segmentation flag indicates that TCP segment packets are to be generated, each TCP segment packet being no longer than the maximum segmentation size for the TCP packet and comprising:

a copy of the TCP header,

an incrementing segment sequence identifier, and

a portion of the data payload; and

transmit the set of TCP segment packets to an external network interface.

4. The media of claim 3 , wherein the configuration file is targeted for a field programmable gate array.

5. A computing system comprising:

a processor programmed to:

prepare a plurality of TCP packets, each TCP packet comprising a TCP header and a data payload,

add a prepend header to each of the TCP packets, the prepend header comprising a TCP segmentation flag and a maximum segment size value indicating a maximum segment size for TCP segmentation; and

transmit each TCP packet to a configurable logic device (CLD); and

a configurable logic device (CLD) configured to receive a plurality of annotated TCP packets from the processor and for each received annotated TCP packet to:

determine whether to generate a set of TCP segment packets based on the TCP segmentation flag for the TCP packet;

generate a set of TCP segment packets if the TCP segmentation flag indicates that TCP segment packets are to be generated, each TCP segment packet being no longer than the maximum segmentation size for the TCP packet and comprising:

a copy of the TCP header,

an incrementing segment sequence identifier, and

a portion of the data payload; and

transmit the set of TCP segment packets to an external network interface.

6. The system of claim 5 , wherein the CLD is a field programmable gate array.

7. The method of claim 1 , further comprising queuing the TCP segment packets at the CLD prior to transmitting the TCP segment packets to the external network interface.

8. The method of claim 1 , wherein the transmitting step at the CLD comprises transmitting the TCP segment packets to a plurality of different external network interfaces.

9. The method of claim 1 , wherein the steps performed at the processor are performed at a plurality of processors.

10. The method of claim 1 , wherein the wherein the steps performed at the CLD are performed at a plurality of CLDs.

11. The method of claim 1 , further comprising stripping the prepend header from the TCP packet at the CLD.

12. The tangible, non-transitory computer-readable media of claim 3 , wherein the CLD is further configured to transmit the TCP segment packets to a plurality of different external network interfaces.

13. The tangible, non-transitory computer-readable media of claim 3 , wherein the CLD is further configured to queue the TCP segment packets prior to transmitting the TCP segment packets to the external network interface.

14. The tangible, non-transitory computer-readable media of claim 3 , wherein the CLD is further configured to strip the prepend header from the TCP packet.

15. The computing system of claim 5 , wherein the CLD is further configured to queue the TCP segment packets at the CLD prior to transmitting the TCP segment packets to the external network interface.

16. The computing system of claim 5 , wherein the CLD is further configured to transmit the TCP segment packets to a plurality of different external network interfaces.

17. The computing system of claim 5 , wherein the system comprises a plurality of the processors.

18. The computing system of claim 5 , wherein the system comprises a plurality of the CLDs.

19. The computing system of claim 5 , wherein the CLD is further configured to strip the prepend header from the TCP packet.

Assignments (6)