System and method for handling frames in multiple stack environments

• US 7,426,579 B2
• Filed: 11/21/2002
• Issued: 09/16/2008
• Est. Priority Date: 09/17/2002
• Status: Active Grant

• Current Assignee: Avago Technologies International Sales Pte Limited (Broadcom, Inc.)
• Original Assignee: Broadcom Corporation (Broadcom, Inc.)
• Inventors: McDaniel, Scott S.
• Primary Examiner(s): Follansbee; John
• Assistant Examiner(s): PHILLIPS, HASSAN A
• Application Number:
  US10/302,474
• Publication Number:
  US 20040054814A1
• Time in Patent Office: 2,126 Days
• Field of Search: 709/250
• US Class Current: 709/250
• CPC Class Codes: H04L 69/12 Protocol engines
• CPC Class Codes: H04L 69/32 Architecture of open systems interconnection [OSI] 7-layer type protocol stacks, e.g. the interfaces between the data link level and the physical level
• CPC Class Codes: H04L 69/329 in the application layer [OSI layer 7]

Abstract

Systems and methods that handle frames in multiple stack environments are provided. In one embodiment, a system may include, for example, a non-offload protocol stack and an offload protocol stack, the offload protocol stack being coupled to the non-offload protocol stack. The non-offload protocol stack may include, for example, one or more partially processed frames. The one or more partially processed frames may be sent to the offload protocol stack to complete processing in the offload protocol stack.

Claims

1. A system for processing incoming frames in a multiple stack environment, the system comprising:

a non-offload protocol stack adapted to process at least a first connection protocol layer for incoming frames received from a network interface card; and

an offload protocol stack coupled to the non-offload protocol stack and adapted to process at least the first connection protocol layer for the incoming frames, at least a portion of the offload protocol stack implemented in hardware,

wherein, prior to offloading, the first connection protocol layer of the non-offload protocol stack comprises one or more partially processed incoming frames, and

wherein, during offloading, the one or more partially processed incoming frames are sent from the first connection protocol layer of the non-offload protocol stack to the first connection protocol layer of the offload protocol stack, via a first catch-up interface dedicated exclusively to the first connection protocol layer, to complete processing in the offload protocol stack.

2. The system according to claim 1,

wherein the one or more partially processed incoming frames comprise a set of incoming frames that has completed processing in accordance with a second connection protocol layer, but not processing in accordance with the first connection protocol layer, in the non-offload protocol stack, and

wherein the set of incoming frames is sent to the first connection protocol layer of the offload protocol stack for processing in accordance with the first connection protocol layer.

3. The system according to claim 2,

wherein the first connection protocol layer corresponds to L3 processing, and

wherein the second connection protocol layer corresponds to L2 processing.

4. The system according to claim 3, wherein the first catch-up interface dedicated exclusively to the first connection protocol layer communicates the sent partially processed incoming frames from the first connection protocol layer of the non-offload protocol stack to the first connection protocol layer of the offload protocol stack without passing through other protocol layers of the non-offload and offload protocol stacks.

5. The system according to claim 2, wherein:

the non-offload protocol stack is adapted to process a second connection protocol layer for incoming frames;

the offload protocol stack is adapted to process the second connection protocol layer for incoming frames; and

prior to offloading, the second connection protocol layer of the non-offload protocol stack comprises one or more partially processed incoming frames that, during offloading, are sent from the second connection protocol layer of the non-offload protocol stack to the second connection protocol layer of the offload protocol stack via a second catch-up interface dedicated exclusively to the second connection protocol layer.

6. The system according to claim 5,

wherein the first connection protocol layer corresponds to L4 processing, and

wherein the second connection protocol layer corresponds to L3 processing.

7. The system according to claim 2, wherein the first catch-up interface passes through a host interface.

8. The system according to claim 1,

wherein the non-offload protocol stack comprises a layer that provides a first level of processing and a layer that provides a second level of processing, and

wherein the layer that provides the first level of processing of the non-offload stack sends frames in order to the layer that provides the second level of processing of the non-offload stack.

9. The system according to claim 8,

wherein the first level of processing comprises L4 processing;

and

wherein the second level of processing comprises L5 processing.

10. The system according to claim 8, wherein the layer that provides the first level of processing of the non-offload stack postpones delivery of a particular frame that is out of order.

11. The system according to claim 1,

wherein the offload protocol stack comprises a layer that provides a first level of processing and a layer that provides a second level of processing, and

wherein the layer that provides the first level of processing of the offload stack sends frames in order to the layer that provides the second level of processing to the non-offload stack.

12. The system according to claim 1,

wherein the offload protocol stack comprises a layer that provides a first level of processing,

wherein the non-offload protocol stack comprises a layer that provides a higher level of processing than the first level, and

wherein the layer that provides the first level of processing of the offload stack sends frames in order to the layer that provides the higher level of processing of the non-offload stack.

13. The system according to claim 1,

wherein the offload protocol stack comprises a layer that provides a first level of processing,

wherein the non-offload protocol stack comprises a layer that provides a higher level of processing than the first level, and

wherein the layer that provides the first level of processing of the offload stack sends out-of-order frames to the layer that provides the higher level of processing of the non-offload stack.

14. The system according to claim 1, wherein frames are processed or received out of order in the non-offload stack or the offload stack until an application layer of the respective stack.

15. A system for processing incoming frames in a multiple stack environment, comprising:

a host that runs a software portion of a non-offload protocol stack, the software portion adapted to process at least a first connection protocol layer for incoming frames received from a network interface card; and

an offload engine coupled to the host and comprising a hardware portion of an offload protocol stack, the hardware portion adapted to process at least the first connection protocol layer for the incoming frames;

wherein, prior to offloading, the first connection protocol layer of the non-offload protocol stack comprises one or more partially processed incoming frames for processing in accordance with the first connection protocol layer, and

wherein, during offloading, the one or more partially processed incoming frames are sent from the first connection protocol layer of the host to the first connection protocol layer of the offload engine, via a first catch-up interface dedicated exclusively to the first connection protocol layer, for processing in accordance with the first connection protocol layer.

16. The system according to claim 15, wherein the one or more partially processed incoming frames are sent from the software portion of the non-offload protocol stack to the hardware portion of the offload protocol stack.

17. The system according to claim 15, wherein at least one partially processed incoming frame in a first level, software processing block of the non-offload protocol stack is sent to a first level, hardware processing block of the offload protocol stack.

18. The system according to claim 17,

wherein the first level, software processing block comprises a software L3 processing block, and

wherein the first level, hardware processing block comprises a hardware L3 processing block.

19. The system according to claim 17,

wherein the first level, software processing block comprises a software L4 processing block, and

wherein the first level, hardware processing block comprises a hardware L4 processing block.

20. The system according to claim 15, wherein frames are sent in order from a first level, hardware processing block in the offload engine to a second level, software processing block in the host.

21. The system according to claim 20,

wherein the first level, hardware processing block comprises a hardware L4 processing block, and

wherein the second level, software processing block comprises a software L5 processing block.

22. A system for handling frames in a multiple stack environment, comprising:

a first layered processing system comprising a first connection protocol layer; and

a second layered processing system coupled to the first layered processing system, the second layered processing system comprising a first connection protocol layer corresponding to the first connection protocol layer of the first layered processing system, at least a portion of the first connection protocol layer of the second layered processing system implemented in hardware,

wherein incoming frames of a particular connection received from a network interface card awaiting processing in the first connection protocol layer of the first layered processing system are moved to the first connection protocol layer of the second layered processing system via a first catch-up interface dedicated exclusively to the first connection protocol layer.

23. The system according to claim 22, wherein the incoming frames of the particular connection in the first layered processing system are moved to the second layered processing system after the particular connection has been moved from the first layered processing system to the second layered processing system.

24. The system according to claim 22, wherein the incoming frames of the particular connection in the first layered processing system are partially processed by the first connection protocol layer of the first layered processing system prior to being moved to the first connection protocol layer of the second layered processing system.

25. The system according to claim 22, wherein the incoming frames of the particular connection are moved to the second layered processing system to continue layer processing.

26. The system according to claim 22, wherein the incoming frames of the particular connection are trapped in the first layered processing system prior to being moved to the second layered processing system.

27. The system according to claim 22, wherein the particular connection is moved during a life of the particular connection.

28. The system according to claim 22,

wherein the first layered processing system comprises a first offload layered processing system, and

wherein the second layered processing system comprises a second offload layered processing system.

29. The system according to claim 22,

wherein the first layered processing system comprises a first non-offload layered processing system, and

wherein the second layered processing system comprises a second non-offload layered processing system.

30. The system according to claim 22,

wherein the first layered processing system comprises a non-offload layered processing system, and

wherein the second layered processing system comprises an offload layered processing system.

31. The system according to claim 22, wherein the first layered processing system handles a first set of connections and the second layered processing system handles a second set of connections.

32. The system according to claim 31, wherein a decision to move a connection of the first set of connections from the first layered processing system to the second layered processing system is made on a per connection basis.

33. A method for processing incoming frames in a multiple stack environment comprising a non-offload protocol stack adapted to process incoming frames in accordance with a first connection protocol layer and an offload protocol stack adapted to process incoming frames in accordance with the first connection protocol layer, at least a portion of the offload protocol stack implemented in hardware, the method comprising:

completing an offload of a connection from the first connection protocol layer of the non-offload protocol stack to the first connection protocol layer of the offload protocol stack; and

sending a partially processed incoming frame, which was received from a network interface card and partially processed by the first connection protocol layer of the non-offload protocol stack, from the first connection protocol layer of the non-offload protocol stack to the first connection protocol layer of the offload protocol stack, via a first catch-up interface dedicated exclusively to the first connection protocol layer, to complete processing of the partially processed incoming frame by the offload protocol stack in accordance with the first connection protocol layer.

34. The method according to claim 33,

wherein the partially processed incoming frame completed a first level of processing, but not a second level of processing, in the non-offload protocol stack, and

wherein the sending comprises sending the partially processed incoming frame to a second level, hardware processing block of the offload protocol stack.

35. The method according to claim 34,

wherein the first level of processing is L2 processing,

wherein the second level of processing is L3 processing, and

wherein the second level, hardware processing block comprises a hardware L3 processing block.

36. The method according to claim 34,

wherein the first level of processing is L3 processing,

wherein the second level of processing is L4 processing, and

wherein the second level, hardware processing block comprises a hardware L4 processing block.

37. The method according to claim 33, wherein a second level, software processing block of the offload protocol stack receives frames in order from a first level, hardware processing block of the offload protocol stack.

38. The method according to claim 37,

wherein the second level, software processing block comprises a software L5 processing block, and

wherein the first level, hardware processing block comprises a hardware L4 processing block.

39. The method according to claim 37, wherein the second level, software processing block of the offload protocol stack comprises a second level, software processing block of the non-offload protocol stack.

40. The method according to claim 33, where the non-offload protocol stack is adapted to process incoming frames in accordance with a second connection protocol layer, and the offload protocol stack is adapted to process incoming frames in accordance with the second connection protocol layer, the method further comprising:

completing an offload of the connection from the second connection protocol layer of the non-offload protocol stack to the second connection protocol layer of the offload protocol stack; and

sending a second partially processed incoming frame, which was partially processed by the second connection protocol layer of the non-offload protocol stack, from the second connection protocol layer of the non-offload protocol stack to the second connection protocol layer of the offload protocol stack, via a second catch-up interface dedicated exclusively to the second connection protocol layer, to complete processing of the second partially processed incoming frame by the offload protocol stack in accordance with the second connection protocol layer.

41. The method according to claim 33, wherein sending the partially processed incoming frame via the first catch-up interface dedicated exclusively to the first connection protocol layer comprises sending the partially processing incoming frame from the first connection protocol layer of the non-offload protocol stack to the first connection protocol layer of the offload protocol stack without passing through other protocol layers of the non-offload and offload protocol stacks.

42. The method according to claim 41, wherein the first connection protocol layer corresponds to L3 layer processing.