System and method for TCP offload

• US 7,346,701 B2
• Filed: 08/29/2003
• Issued: 03/18/2008
• Est. Priority Date: 08/30/2002
• Status: Active Grant

• Current Assignee: Avago Technologies International Sales Pte Limited (Broadcom, Inc.)
• Original Assignee: Broadcom Corporation (Broadcom, Inc.)
• Inventors: Lindsay, Steve, Elzur, Uri, McDaniel, Scott S., Fan, Frankie
• Primary Examiner(s): Avellino; Joseph E.
• Application Number:
  US10/652,267
• Publication Number:
  US 20040042483A1
• Time in Patent Office: 1,663 Days
• Field of Search: 709/232, 709/236, 370/463, 370/419, 370/392
• US Class Current: 709/232
• CPC Class Codes: H04L 47/193 at the transport layer, e.g. TCP related
• CPC Class Codes: H04L 47/2441 relying on flow classification, e.g. using integrated services [IntServ]
• CPC Class Codes: H04L 47/34 ensuring sequence integrity, e.g. using sequence numbers
• CPC Class Codes: H04L 49/90 Buffering arrangements
• CPC Class Codes: H04L 49/9063 Intermediate storage in different physical parts of a node or terminal
• CPC Class Codes: H04L 49/9073 Early interruption upon arrival of a fraction of a packet
• CPC Class Codes: H04L 49/9094 Arrangements for simultaneous transmit and receive, e.g. simultaneous reading/writing from/to the storage element
• CPC Class Codes: H04L 69/10 Streamlined, light-weight or high-speed protocols, e.g. express transfer protocol [XTP] or byte stream
• CPC Class Codes: H04L 69/12 Protocol engines
• CPC Class Codes: H04L 69/16 Implementation or adaptation of Internet protocol [IP], of transmission control protocol [TCP] or of user datagram protocol [UDP]
• CPC Class Codes: H04L 69/161 Implementation details of TCP/IP or UDP/IP stack architecture; Specification of modified or new header fields
• CPC Class Codes: H04L 69/162 involving adaptations of sockets based mechanisms secure socket layer H04L63/168
• CPC Class Codes: H04L 69/163 In-band adaptation of TCP data exchange; In-band control procedures
• CPC Class Codes: H04L 69/166 IP fragmentation; TCP segmentation
• CPC Class Codes: H04L 9/40 Network security protocols

Abstract

Aspects of the invention may comprise receiving an incoming TCP packet at a TEEC and processing at least a portion of the incoming packet once by the TEEC without having to do any reassembly and/or retransmission by the TEEC. At least a portion of the incoming TCP packet may be buffered in at least one internal elastic buffer of the TEEC. The internal elastic buffer may comprise a receive internal elastic buffer and/or a transmit internal elastic buffer. Accordingly, at least a portion of the incoming TCP packet may be buffered in the receive internal elastic buffer. At least a portion of the processed incoming packet may be placed in a portion of a host memory for processing by a host processor or CPU. Furthermore, at least a portion of the processed incoming TCP packet may be DMA transferred to a portion of the host memory.

Claims

1. A system for offloading TCP processing, the system comprising:

a host;

a network interface card (NIC) coupled to said host, said NIC comprising,

a TCP enabled Ethernet controller (TEEC), said TEEC comprising,

at least one internal elastic buffer, wherein said TEEC processes an incoming TOP packet once and temporarily buffers at least a portion of said incoming TOP packet in said internal elastic buffer, said processing occurring without reassembly.

2. The system according to claim 1, wherein said at least one internal elastic buffer comprises at least one of a receive internal elastic buffer and a transmit internal elastic buffer.

3. The system according to claim 2, wherein said at least a portion of said incoming TCP packet is temporarily buffered in said receive internal elastic buffer.

4. The system according to claim 2, wherein at least a portion of a TCP packet to be transmitted is temporarily buffered in said transmit internal elastic buffer.

5. The system according to claim 1, wherein said TEEC places at least a portion of said incoming TCP packet data into at least a portion of a host memory.

6. The system according to claim 1, wherein said NIC utilizes only said at least one internal elastic buffer to temporarily buffer said at least a portion of said incoming TCP packet.

7. The system according to claim 1, wherein out-of-order TCP packets are not at least one of stored, re-ordered and re-assembled in a TEEC buffer.

8. The system according to claim 1, wherein said NIC does not require a dedicated memory for re-ordering out-of-sequence TCP packets.

9. The system according to claim 1, wherein said NIC does not require a dedicated memory for assembling and re-ordering IP packets fragmented at the IP layer.

10. The system according to claim 1, wherein said TEEC places at least data from said incoming TCP packet into-a highest hierarchy of buffer available in a host memory by performing a single copy operation.

11. The system according to claim 1, wherein said TEEC DMA transfers at least a portion of said processed incoming TCP packet to at least a portion of a host memory.

12. The system according to claim 1, wherein said NIC does not require a TCP off load engine (TOE) dedicated memory for at least one of packet retransmission and packet reassembly.

13. The system according to claim 1, wherein said TEEC places at least a portion of said processed incoming TCP packets into host buffers in a host memory for reassembly.

14. The system according to claim 1, wherein said TEEC comprises a single chip, having integrated therein, said at least one internal elastic buffer.

15. The system according to claim 1, wherein said TEEC comprises a single chip, having integrated therein, said at least one internal elastic buffer, and no internal buffers and interfaces to external buffers, that are utilized for at least one of packet retransmission, packet reassembly and packet re-ordering.

16. A method for offloading TCP processing, the method comprising:

receiving an incoming TCP packet at a TCP enabled Ethernet controller (TEEC);

processing at least a portion of said incoming packet once by said TEEC without reassembly; and

temporarily buffering said at least a portion of said incoming TCP packet in at least one internal elastic buffer of said TEEC.

17. The method according to claim 16, wherein said at least one internal elastic buffer comprises at least one of a receive internal elastic buffer and a transmit internal elastic buffer.

18. The method according to claim 17, further comprising temporarily buffering said at least a portion of said incoming TCP packet in said receive internal elastic buffer.

19. The method according to claim 16, further comprising placing at least a portion of said processed at least a portion of said incoming packet in at least a portion of a host memory.

20. The method according to claim 16, wherein said placing further comprises placing at least a portion of said processed incoming TCP packet in a highest hierarchy of buffer available in a host memory by performing a single copy operation.

21. The method according to claim 16, further comprising DMA transferring at least a portion of said processed incoming TCP packet in at least a portion of a host memory.

22. The method according to claim 16, wherein packets temporarily buffered in said in at least one internal elastic buffer are not buffered for at least one of reassembly and retransmission.

23. The method according to claim 16, further comprising placing at least a portion of said processed incoming TCP packet in host buffers in a host memory for processing.

24. The method according to claim 16, wherein said TEEC comprises a single chip, having integrated therein, said at least one internal elastic buffer.

25. A machine-readable storage, having stored thereon, a computer program having at least one code section for providing TCP offload, the at least one code section being executable by a machine for causing the machine to perform steps comprising:

receiving an incoming TCP packet at a TCP enabled Ethernet controller (TEEC);

processing at least a portion of said incoming packet once by said TEEC without reassembly; and

temporarily buffering said at least a portion of said incoming TCP packet in at least one internal elastic buffer of said TEEC.

26. The machine-readable storage according to claim 25, wherein said at least one internal elastic buffer comprises at least one of a receive internal elastic buffer and a transmit internal elastic buffer.

27. The machine-readable storage according to claim 26, code for temporarily buffering said at least a portion of said incoming TCP packet in said receive internal elastic buffer.

28. The machine-readable storage according to claim 25, further comprising code for placing at least a portion of said processed at least a portion of said incoming packet in at least a portion of a host memory.

29. The machine-readable storage according to claim 25, further comprising code for placing at least a portion of said processed incoming TCP packet in a highest hierarchy of buffer available in a host memory by performing a single copy operation.

30. The machine-readable storage according to claim 25, further comprising code for DMA transferring at least a portion of said processed incoming TCP packet in at least a portion of a host memory.

31. The machine-readable storage according to claim 25, wherein packets temporarily buffered in said in at least one internal elastic buffer are not buffered for at least one of reassembly and retransmission.

32. The machine-readable storage according to claim 25, further comprising code for placing at least a portion of said processed incoming TCP packet in host buffers in a host memory for processing.

33. The machine-readable storage according to claim 25, wherein said TEEC comprises a single chip, having integrated therein, said at least one internal elastic buffer.