Method and apparatus for aggregating input data streams

• US 8,493,988 B2
• Filed: 09/13/2010
• Issued: 07/23/2013
• Est. Priority Date: 03/26/2004
• Status: Active Grant

• Current Assignee: Avago Technologies International Sales Pte Limited (Broadcom, Inc.)
• Original Assignee: Foundry Networks LLC (Broadcom, Inc.)
• Inventors: Wong, Yuen Fai, Lin, Yu-Mei, Grenier, Richard A.
• Primary Examiner(s): PATEL, CHANDRAHAS B
• Application Number:
  US12/880,518
• Publication Number:
  US 20110110237A1
• Time in Patent Office: 1,044 Days
• Field of Search: 370/422
• US Class Current: 370/422
• CPC Class Codes: H04L 47/2433 Allocation of priorities to traffic types
• CPC Class Codes: H04L 47/6275 based on priority
• CPC Class Codes: H04L 49/30 Peripheral units, e.g. input or output ports
• CPC Class Codes: H04L 49/352 Gigabit ethernet switching [GBPS]

Abstract

A method and apparatus aggregate a plurality of input data streams from first processors into one data stream for a second processor, the circuit and the first and second processors being provided on an electronic circuit substrate. The aggregation circuit includes (a) a plurality of ingress data ports, each ingress data port adapted to receive an input data stream from a corresponding first processor, each input data stream formed of ingress data packets, each ingress data packet including priority factors coded therein, (b) an aggregation module coupled to the ingress data ports, adapted to analyze and combine the plurality of input data steams into one aggregated data stream in response to the priority factors, (c) a memory coupled to the aggregation module, adapted to store analyzed data packets, and (d) an output data port coupled to the aggregation module, adapted to output the aggregated data stream to the second processor.

Claims

1. A method for aggregating data packets received from a first processor for a second processor, said method comprising:

receiving an input data stream from the first processor using a data link having a first bandwidth, the input data stream comprising ingress data packets, each ingress data packet comprising at least one priority factor coded therein;

analyzing and classifying each of the ingress data packets into one of predetermined priority classes based on the at least one priority factor;

storing an analyzed data packet in a memory;

generating a packet descriptor for the analyzed ingress data packet, the packet descriptor containing a reference to a memory location of its analyzed data packet stored in the memory;

placing the packet descriptor in a priority queue corresponding to the priority class of the data packet;

arbitrating and selecting a packet descriptor from among the priority queues using selection logic implementing a queue scheme;

reading a data packet corresponding to the selected packet descriptor from the memory; and

sending the data packets read from the memory to the second processor using a second data link as an aggregated data stream, wherein the first bandwidth is greater than a second bandwidth of the second data link.

2. The method of claim 1, wherein said analyzing comprises:

decoding a header of each ingress data packet to extract the at least one priority factor.

3. The method of claim 1, further comprising:

buffering the analyzed data packet in a write buffer before storing in the memory.

4. The method of claim 3, further comprising:

asserting a flow control signal if an amount of data stored in the write buffer exceeds a threshold.

5. The method of claim 1, further comprising:

buffering the data packet read from the memory in a read buffer.

6. The method of claim 1, further comprising:

encoding the data packets into an interface format before sending to the second processor.

7. The method of claim 1, further comprising:

asserting a flow control signal if a length of a corresponding priority queue exceeds a threshold.

8. The method of claim 1, wherein said memory is an external buffer memory.

9. The method of claim 1, wherein the first processor and the second processor transmits and receives a data stream through a logical interface providing logical interconnection between a Media Access Control sublayer (MAC) and a Physical layer (PHY).

10. The method of claim 1, wherein said arbitrating and selecting, said reading, and said sending are performed as a single data channel.

11. The method of claim 1, wherein said analyzing and classifying comprises:

protocol-filtering to determine if the ingress data packet is a certain protocol packet.

12. The method of claim 11, wherein the at least one priority factor comprises:

protocol filter priority;

per-port priority; and

virtual LAN priority.

13. The method of claim 1, wherein the first processor is a Layer-2 switching processor.

14. The method of claim 1, wherein the second processor is a data packet processor.