Error detection and fault isolation for lockstep processor systems

• US 6,065,135 A
• Filed: 06/04/1999
• Issued: 05/16/2000
• Est. Priority Date: 06/07/1996
• Status: Expired due to Term

• Current Assignee: BAE Systems Information and Electronic Systems Integration Incorporated (BAE Systems Plc)
• Original Assignee: Lockheed Martin Corporation (Martin Marietta Corporation)
• Inventors: Langston, Dale G., Marshall, Joseph R.
• Primary Examiner(s): Beausoliel, Jr., Robert W.
• Assistant Examiner(s): ELISCA, PIERRE E
• Application Number:
  US09/325,641
• Time in Patent Office: 347 Days
• Field of Search: 714/12, 714/11, 714/48, 714/49, 714/52, 714/14, 370/228, 370/221, 370/217, 370/218, 370/220
• US Class Current: 714/11
• CPC Class Codes: G06F 11/10 Adding special bits or symbols to the coded information, e.g. parity check, casting out 9's or 11's
• CPC Class Codes: G06F 11/1044 with specific ECC/EDC distribution
• CPC Class Codes: G06F 11/1641 where the comparison is not performed by the redundant processing components
• CPC Class Codes: G06F 11/165 with continued operation after detection of the error
• CPC Class Codes: G06F 11/1654 where the output of only one of the redundant processing components can drive the attached hardware, e.g. memory or I/O

Abstract

A lockstep processor system adds error detection, isolation, and recovery logic to one or more lockstep processor system functions; namely, control outputs, processor inputs, I/O busses, memory address busses, and memory data busses.

Claims

1. A lockstep processor system including a master processor and a slave processor executing identical tasks independently in lockstep with one another, comprising in combination:

a receiver comprising bit compare logic wherein data on a bus of said master processor are compared with data on a corresponding bus of said slave processor and a compare error signal is generated when corresponding bits on the corresponding busses miscompare;

error detection and isolation logic wherein a source of said compare error is isolated as originating with said master processor or with said slave processor, said error detection and isolation logic including master code generation logic wherein an error detection code is generated in said data on said bus of said master processor, and slave code generation logic wherein an error detection code is generated in said data on said bus of said slave processor;

master error signal generation logic responsive to said error detection code in said data on said bus of said master processor, wherein a master processor error signal is generated, and slave error signal generation logic responsive to said error detection code in said data on said bus of said slave processor, wherein a slave processor error signal is generated; and

disable signal generation logic responsive to said compare error signal, said master processor error signal, and said slave processor error signal wherein a lockstep disable signal and a processor disable signal are generated.

2. A lockstep processor system as in claim 1 wherein said error detection code includes parity bits.

3. A lockstep processor system as in claim 1 wherein said bus of said master processor is a control output bus and said bus of said slave processor is a control output bus.

4. A lockstep processor system as in claim 1 wherein said bus of said master processor is a control output bus and said bus of said slave processor is a control input/output bus.

5. A lockstep processor system as in claim 1 wherein said bus of said master processor is a control output bus and said bus of said slave processor is a control memory address bus.

6. A lockstep processor system as in claim 1 wherein said bus of said master processor is a control output bus and said bus of said slave processor is a control memory data bus.

7. A method for operating a master processor and a slave processor to execute identical tasks independently in lockstep with one another, comprising:

comparing data on an output control bus of the master processor with data on a corresponding output control bus of the slave processor;

generating a compare error signal when corresponding bits on the corresponding busses miscompare;

isolating a source of said compare error as originating with the master processor or with the slave processor,

wherein the step of generating comprises generating an error detection code in said data on the output control bus of the master processor, generating an error detection code in said data on the output bus of the slave processor, generating a master processor error signal in response to said error detection code in said data on the output control bus of the master processor, and generating a slave processor error signal in response to said error detection code in said data on the output control bus of the slave processor; and

generating a lockstep disable signal and a processor disable signal in response to said compare error signal, said master processor error signal, and said slave processor error signal.

8. A method as in claim 7, wherein the step of generating comprises generating error detection codes including parity bits.

9. A method as in claim 7, wherein said bus of said master processor is a control output bus and said bus of said slave processor is a control output bus.

10. A method as in claim 7, wherein said bus of said master processor is a control output bus and said bus of said slave processor is a control input/output bus.

11. A method as in claim 7, wherein said bus of said master processor is a control output bus and said bus of said slave processor is a control memory address bus.

12. A method as in claim 7, wherein said bus of said master processor is a control output bus and said bus of said slave processor is a control memory data bus.

13. A lockstep processor system including a master processor and a slave processor executing identical tasks independently in lockstep with one another, comprising in combination:

means for comparing data on a bus of said master processor with data on a corresponding bus of said slave processor and for generating a compare error signal when corresponding bits on the corresponding busses miscompare;

means for isolating a source of said compare error as originating with said master processor or with said slave processor;

said means for isolating including;

means for generating an error detection code in said data on said bus of said master processor and means for generating an error detection code in said data on said bus of said slave processor;

means responsive to said error detection code in said data on said bus of said master processor for generating a master processor error signal and means responsive to said error detection code in said data on said bus of said slave processor for generating a slave processor error signal; and

means responsive to said compare error signal, said master processor error signal, and said slave processor error signal for generating a lockstep disable signal and a processor disable signal,

wherein said error detection code includes parity bits.

14. A lockstep processor system including a master processor and a slave processor executing identical tasks independently in lockstep with one another, comprising in combination:

means for comparing data on a bus of said master processor with data on a corresponding bus of said slave processor and for generating a compare error signal when corresponding bits on the corresponding busses miscompare;

means for isolating a source of said compare error as originating with said master processor or with said slave processor;

said means for isolating including;

means for generating an error detection code in said data on said bus of said master processor and means for generating an error detection code in said data on said bus of said slave processor;

means responsive to said error detection code in said data on said bus of said master processor for generating a master processor error signal and means responsive to said error detection code in said data on said bus of said slave processor for generating a slave processor error signal; and

means responsive to said compare error signal, said master processor error signal, and said slave processor error signal for generating a lockstep disable signal and a processor disable signal,

wherein said bus of said master processor is a control output bus and said bus of said slave processor is a control input/output bus.

15. A lockstep processor system including a master processor and a slave processor executing identical tasks independently in lockstep with one another, comprising in combination:

means for comparing data on a bus of said master processor with data on a corresponding bus of said slave processor and for generating a compare error signal when corresponding bits on the corresponding busses miscompare;

means for isolating a source of said compare error as originating with said master processor or with said slave processor;

said means for isolating including;

means for generating an error detection code in said data on said bus of said master processor and means for generating an error detection code in said data on said bus of said slave processor;

means responsive to said error detection code in said data on said bus of said master processor for generating a master processor error signal and means responsive to said error detection code in said data on said bus of said slave processor for generating a slave processor error signal; and

means responsive to said compare error signal, said master processor error signal, and said slave processor error signal for generating a lockstep disable signal and a processor disable signal,

wherein said bus of said master processor is a control output bus and said bus of said slave processor is a control memory address bus.

16. A lockstep processor system including a master processor and a slave processor executing identical tasks independently in lockstep with one another, comprising in combination:

means for comparing data on a bus of said master processor with data on a corresponding bus of said slave processor and for generating a compare error signal when corresponding bits on the corresponding busses miscompare;

means for isolating a source of said compare error as originating with said master processor or with said slave processor;

said means for isolating including;

means for generating an error detection code in said data on said bus of said master processor and means for generating an error detection code in said data on said bus of said slave processor;

means responsive to said error detection code in said data on said bus of said master processor for generating a master processor error signal and means responsive to said error detection code in said data on said bus of said slave processor for generating a slave processor error signal; and

means responsive to said compare error signal, said master processor error signal, and said slave processor error signal for generating a lockstep disable signal and a processor disable signal,

wherein said bus of said master processor is a control output bus and said bus of said slave processor is a control memory data bus.