Link level retry scheme

• US 7,991,875 B2
• Filed: 01/06/2006
• Issued: 08/02/2011
• Est. Priority Date: 05/18/2001
• Status: Expired due to Fees

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: Khan, Andalib, Ram, Rajee S., Padwekar, Kiran, Chou, Ching-Tsun, Venkatesan, Srinand, Chittor, Suresh, Sen, Sujoy, Kumar, Akhilesh, Mannava, Phanindra K.
• Primary Examiner(s): Flynn; Nathan
• Assistant Examiner(s): Chang; Kai J
• Application Number:
  US11/326,982
• Publication Number:
  US 20070130353A1
• Time in Patent Office: 2,034 Days
• Field of Search: 709230-237, 709223-226, 714/746, 714/749, 714/748, 340/5.1, 370/229, 370/241, 370/254, 370/235, 370/236, 370412-416, 370392-394, 370470-474
• US Class Current: 709/224
• CPC Class Codes: H04L 1/1614 using bitmaps
• CPC Class Codes: H04L 1/1671 the supervisory signal being transmitted together with control information
• CPC Class Codes: H04L 1/1874 Buffer management
• CPC Class Codes: H04L 1/188 Time-out mechanisms
• CPC Class Codes: H04L 2001/125 Arrangements for preventing errors in the return channel
• CPC Class Codes: H04L 69/324 in the data link layer [OSI layer 2], e.g. HDLC
• CPC Class Codes: H04L 9/40 Network security protocols

Abstract

A link layer system is provided. The link layer system a first link layer control module and a retry queue for storing a transmitted data packet. The retry control module is coupled to the first link layer control module, which directs the retry queue to discard the transmitted data packet when an acknowledgment bit is received by the first link layer control module.

Claims

1. A link layer system, comprising:

a first link layer control module;

a retry queue for storing a transmitted data packet;

a retry control module coupled to said first link layer control module and said retry queue, wherein said retry control module directs said retry queue to discard said transmitted data packet when an acknowledgment bit piggybacked on an info flit is received by said first link layer control module and re-transmit data packets from said retry queue if an error is received, wherein a re-transmit starting point is established by counting info flits received without error using one or more pointers;

a first control flit module coupled to the first link layer control module;

a second control flit module coupled to a second link layer control module, the second control flit module sends a retry request to the first link layer control module and refuses to accept any more flits if the error is detected in the transmitted data packet; and

a processor implementing one or more of the first link layer control module, the retry queue, the retry control module, the first control flit module, the second control flit module, or the second link layer control module.

2. The link layer system as recited by claim 1, further comprising the second link layer control module for receiving the transmitted data packet and transmitting the acknowledgment bit to the first link layer control module if the error is not detected in the transmitted data packet.

3. The link layer system as recited by claim 1, wherein the first link layer control module receives the retry request and directs the retry control module to send the transmitted data packet from the retry queue.

4. The link layer system as recited by claim 3, wherein the retry queue includes a head pointer indicating a location of the transmitted data packet, wherein the transmitted data packet is the oldest transmitted data packet in the retry queue.

5. The link layer system as recited by claim 4, wherein the retry queue the head pointer is incremented and the transmitted data packet is dequeued from the retry queue when the acknowledgment bit for the transmitted data packet is received.

6. The link layer system as recited by claim 3, wherein the retry queue includes a tail pointer indicating a location of an unoccupied entry to store a next transmitted data packet.

7. The link layer system as recited by claim 6, wherein said tail pointer is incremented each time a new data packet is stored into the retry queue.

8. A method, comprising:

storing a transmitted data packet in a retry queue;

receiving an acknowledgment bit piggybacked on an info flit;

discarding the transmitted data packet from the retry queue when the acknowledgment bit piggybacked on the info flit is received; and

re-transmitting data packets from the retry queue if an error is received, wherein a re-transmit starting point is established by counting info flits received without error using one or more pointers;

sending a retry request if the error is detected in the transmitted data; and

refusing to accept any more flits if the error is detected.

9. The method as recited in claim 8, further comprising:

providing the acknowledgment bit piggybacked on the info flit if the error is not detected in the transmitted data packet.

10. The method as recited in claim 9, further comprising:

receiving the retry request; and

sending the transmitted data packet from the retry queue.

11. The method as recited in claim 10, further comprising:

indicating a location of the transmitted data packet in the retry queue with a head pointer, wherein the transmitted data packet is the oldest transmitted data packet in the retry queue.

12. The method as recited in claim 11, further comprising:

incrementing the head pointer and dequeuing the transmitted data packet from the retry queue when the acknowledgment bit for the transmitted data packet is received.

13. The method as recited in claim 10, further comprising:

indicating a location of an unoccupied entry to store a next transmitted data packet in the retry queue with a tail pointer.

14. The method as recited in claim 13, further comprising:

incrementing the tail pointer each time a new data packet is stored into the retry queue.

15. A set of instructions residing in a non-transitory storage medium, the set of instructions capable of being executed by a processor for recovering from transmission errors in data communication channels, comprising:

storing a transmitted data packet in a retry queue;

receiving an acknowledgment bit piggybacked on an info flit;

discarding the transmitted data packet from the retry queue when the acknowledgment bit piggybacked on the info flit is received; and

re-transmitting data packets from the retry queue if an error is received, wherein a re-transmit starting point is established by counting info flits received without error using one or more pointers;

sending a retry request if the error is detected in the transmitted data packet; and

refusing to accept any more flits if the error is detected.

16. The set of instructions as recited in claim 15, further comprising:

providing the acknowledgment bit piggybacked on the info flit if the error is not detected in the transmitted data packet.

17. The set of instructions as recited in claim 15, further comprising:

receiving the retry request; and

sending the transmitted data packet from the retry queue.

18. The set of instructions as recited in claim 17, further comprising:

indicating a location of the transmitted data packet in the retry queue with a head pointer, wherein the transmitted data packet is the oldest transmitted data packet in the retry queue.

19. The set of instructions as recited in claim 18, further comprising:

incrementing the head pointer and dequeuing the transmitted data packet from the retry queue when the acknowledgment bit for the transmitted data packet is received.

20. The set of instructions as recited in claim 17, further comprising:

indicating a location of an unoccupied entry to store a next transmitted data packet in the retry queue with a tail pointer.

21. The set of instructions as recited in claim 20, further comprising:

incrementing the tail pointer each time a new data packet is stored into the retry queue.