Slave device having independent error recovery

• US 7,526,676 B2
• Filed: 09/03/2004
• Issued: 04/28/2009
• Est. Priority Date: 09/03/2004
• Status: Expired due to Fees

• Current Assignee: Avago Technologies General IP PTE Limited (Broadcom, Inc.)
• Original Assignee: Avago Technologies General IP PTE Limited (Broadcom, Inc.)
• Inventors: Li, Whitney, Chou, Norman C.
• Primary Examiner(s): Baderman; Scott T
• Assistant Examiner(s): SCHELL, JOSEPH O
• Application Number:
  US10/933,750
• Publication Number:
  US 20060053331A1
• Time in Patent Office: 1,698 Days
• Field of Search: 714/4, 714/18, 714/43, 714/44, 714/748, 714/16, 714/17, 714/23
• US Class Current: 714/16
• CPC Class Codes: G06F 11/0745 in an input/output transactions management context input/output processing in general G06F13/00
• CPC Class Codes: G06F 11/0793 Remedial or corrective actions recovery from an exception in an instruction pipeline G06F9/3861; by retry G06F11/1402; for recovering from a failure of a protocol instance or entity H04L69/40
• CPC Class Codes: G06F 11/1471 involving logging of persistent data for recovery

Abstract

A slave device adapted to couple to a master processor and including an error handler and a communication controller. The error handler is configured to detect errors internal to the slave device and, in response to detecting at least one error and independent of the master processor, configured to select an error recovery operation based on the at least one detected error and to initiate and perform the selected error recovery operation. The communication controller is configured to communicate with the master processor according to a master/slave protocol, and configured to maintain the master/slave protocol during performance of the selected error recovery operation by the error handler.

Claims

1. A slave device adapted to couple to a master processor, the slave device comprising:

an error handler configured to detect errors internal to the slave device and, in response to detecting at least one error and independent of the master processor, configured to select an error recovery operation based on the at least one detected error and configured to initiate and perform the selected error recovery operation; and

a communication controller configured to communicate with the master processor according to a master/slave protocol, the master/slave protocol comprising completion of a transaction that is on-going with the master processor when the error handler initiates the selected error recovery operation, the communication controller further configured to provide a reset interrupt signal to the master processor to notify the master processor that the error recovery operation has been completed, and to further notify the master processor that at least a portion of data transferred between the slave device and the master processor may have been lost or corrupted.

2. The slave device of claim 1, wherein the error handler is configured to detect a plurality of types of error and to select the desired error recovery operation from a plurality of error recovery operations based on the type of error detected.

3. The slave device of claim 1, wherein the reset interrupt signal is operative to instruct the master processor to perform a check-point retry operation whereby the master processor initiates operations from a previously check-pointed state.

4. The slave device of claim 1, wherein the communication controller is configured to perform data transactions with the master processor according to the master/slave protocol, each transaction comprising a predetermined amount of data.

5. The slave device of claim 4, wherein the communication controller is configured to fabricate an amount of data necessary to complete a data transaction on-going from the slave device to the master processor after the error handler initiates the selected error recovery operation, wherein the amount of fabricated data comprises a difference between the predetermined amount of data and an amount of data received by the master processor from the slave device prior to the initiation of the selected error recovery operation.

6. A packet routing device configured as a slave to a master processor for routing data packets between the master processor and a network fabric, the packet routing device comprising:

an error handler configured to detect errors internal to the packet routing device and, in response to detecting at least one error and independent of the master processor, configured to select and initiate an error recovery operation based on the at least one detected error, the error handler further configured to assert an error recovery flag that remains asserted until the error recovery operation is completed; and

a communication controller configured to assert a transaction flag when a data transfer is taking place between the master processor and the packet routing device over a communication bus, the communication controller further configured to recognize the assertion of the error recovery flag yet complete the data transfer in order to avoid hanging the communication bus.

7. The packet routing device of claim 6, wherein the data transfer comprises a series of data transactions according to a communication protocol, each data transaction comprising a predetermined amount of data.

8. The packet routing device of claim 7, wherein the communication controller is configured to provide a fabricated amount of data necessary to complete a data transaction on-going from the packet routing device to the master processor after the error handler initiates the selected error recovery operation, wherein the amount of fabricated data comprises a difference between the predetermined amount of data and an amount of data received by the master processor from the packet routing device prior to the initiation of the selected error recovery operation.

9. The packet routing device of claim 6, wherein the communication controller is configured to provide a reset signal to the master processor after completion of the error recovery operation to notify the master processor of the error recovery operation and to further notify the master processor that at least a portion of data transferred between the master processor and the packet routing device may have been lost or corrupted.

10. The packet routing device of claim 6, wherein the communication controller further comprises:

a blocking mechanism, wherein the communication controller sets the blocking mechanism to prevent potentially corrupted data packets from being transferred from the packet routing device to the network fabric when the selected error recovery operation is initiated while the communication controller is transferring data between the packet routing device and the master processor.

11. The packet routing device of claim 10, wherein the blocking mechanism is configured to be reset by the master processor.

12. The packet routing device of claim 10, wherein the blocking mechanism comprises a semaphore.

13. A memory device configured as a slave to a master processor, the memory device comprising:

an error handler configured to detect errors internal to the memory device and, in response to detecting at least one error and independent of the master processor, configured to select and initiate an error recovery operation based on the at least one detected error, the error handler further configured to assert an error recovery flag upon initiation of the error recovery operation;

a storage medium configured to store data; and

a memory controller configured to assert a transaction flag when one of a write or a read operation is initiated upon the memory device for carrying out a data transfer between the master processor and the memory device over a communication bus, the communication controller further configured to recognize the assertion of the error recovery flag yet complete the data transfer in order to avoid hanging the communication bus.

14. The memory device of claim 13, wherein each data transfer comprises a predetermined amount of data in accordance with a communication protocol.

15. The memory device of claim 13, wherein the memory device further comprises:

a blocking mechanism, wherein the memory controller sets the blocking mechanism to prevent potentially corrupted data from being written to the storage medium when the error recovery flag is asserted.

16. The memory device of claim 13, wherein the storage medium comprises a write-once type storage medium, and a semaphore is used to prevent writing of potentially corrupt data into the write-once type storage medium.

17. The memory device of claim 13, wherein the memory controller is configured to provide a reset signal to the master processor after completion of the error recovery operation to notify the master processor of the error recovery operation and to further notify the master processor that at least a portion of data transferred between the master processor and the memory device may have been lost or corrupted.

18. A network system comprising:

a master processor; and

a slave device coupled to the master processor, the slave device comprising;

an error handler configured to detect errors internal to the slave device and, in response to detecting at least one error and independent of the master processor, configured to select an error recovery operation based on the at least one detect error and configured to initiate and perform the selected error recovery operation;

a communication controller configured to communicate with the master processor according to a master/slave protocol, and configured to complete an ongoing data transfer between the slave device and the master processor in spite of being aware of the error recovery operation;

and the communication controller further configured to provide a reset interrupt signal to the master processor to notify the master processor that the error recovery operation has been completed.

19. The network system of claim 18, wherein the communication controller is configured to perform data transactions with the master processor according to the master/slave protocol, each transaction comprising a predetermined amount of data, and wherein the communication controller is configured to complete the ongoing data transfer by providing fabricated data.

20. The network system of claim 18, wherein the communication controller, upon completion of the error recovery operation, provides to the master processor a reset signal for informing the master processor of the error recovery operation.

21. A method of operating a slave device adapted to couple to a master processor, the method comprising:

communicating with the master processor according to a master/slave protocol;

detecting at least one error internal to the slave device;

selecting an error recovery operation based on the at least one detected internal error;

initiating and performing the selected error recovery operation independent of the master processor;

completing at least one of a) a pending data transaction or b) an ongoing data transfer between the slave device and the master processor over a communication bus, regardless of a concurrent occurrence of the error recovery operation, in order to avoid disrupting the communication bus;

and providing a reset interrupt signal to the master processor to notify the master processor that the error recovery operation has been completed.

22. The method of claim 21, wherein communicating with the master processor further comprises:

performing data transactions with the master processor according to the master/slave protocol, each transaction comprising a predetermined amount of data.

23. The method of claim 22, wherein completing the ongoing data transfer comprises:

fabricating an amount of data comprising a difference between the predetermined amount of data and an amount of data transferred prior to the initiation of the selected error recovery operation.

24. The method of claim 21, further comprising:

using a semaphore to block transfer of data from the slave device to an external network.