Method and apparatus to monitor the run state of a multi-partitioned computer system

• US 6,711,700 B2
• Filed: 04/23/2001
• Issued: 03/23/2004
• Est. Priority Date: 04/23/2001
• Status: Active Grant

• Current Assignee: Globalfoundries US Incorporated (GlobalFoundries, Inc.)
• Original Assignee: International Business Machines Corporation
• Inventors: Armstrong, William Joseph, Wottreng, Andy, Forsman, Stephanie Maria, Nayar, Naresh, Armstrong, Troy David, Scheel, Jeffrey Jay
• Primary Examiner(s): BADERMAN, SCOTT T
• Application Number:
  US09/840,345
• Publication Number:
  US 20020156939A1
• Time in Patent Office: 1,065 Days
• Field of Search: 714/23, 714/55, 714/10, 712/13, 711/153, 711/173
• US Class Current: 714/23
• CPC Class Codes: G06F 11/0712 in a virtual computing platform, e.g. logically partitioned systems
• CPC Class Codes: G06F 11/0724 in a multiprocessor or a multi-core unit multiprocessors per se G06F15/80
• CPC Class Codes: G06F 11/0757 by exceeding a time limit, i.e. time-out, e.g. watchdogs
• CPC Class Codes: G06F 11/1441 Resetting or repowering

Abstract

An apparatus and method for monitoring the state of a computer system running multiple operating systems shared by a partition manager is provided. A dedicated service processor monitors the individual run state condition of a plurality of processors running a plurality of operating systems. The service processor executes a routine to poll a memory location in each processor in the system to determine if the processor has entered an error loop with interrupts disabled. If any one of the plurality of processors are in an error loop, the service processor executes a routine to send a non-maskable interrupt to the looped processor so that the partition manager may regain control of the processor.

Claims

1. A computer system including a plurality of processors, the system comprising:

a timestamp clock;

a memory containing a data structure comprising a respective memory location for each of the plurality of processors;

wherein the each of the plurality of processors, during normal operation, is configured to periodically write a timestamp value read from the timestamp clock to its respective timestamp memory location;

a service processor electrically coupled to the plurality of processors, wherein the service processor generates an electrical signal to reset a processor of the plurality of processors if a previously read timestamp value read by the service processor from the respective timestamp memory location is unchanged after a subsequent reading.

2. The computer system of claim 1, further comprising an incremental counter to provide an index of the plurality of processors.

3. The computer system of claim 1, wherein the timestamp clock runs in real time.

4. The computer system of claim 1, wherein the memory is integral to the service processor.

5. The computer system of claim 4, wherein the memory contains a partition manager and the plurality of processors are logically partitioned.

6. The computer system of claim 1, wherein the service processor is configured to read the respective timestamp memory location for each processor at a first rate less frequent then a second rate at which each processor, during normal operation, writes the timestamp value to its respective memory location.

7. The computer system of claim 1, wherein the plurality of processors are multi-threaded processors.

8. A method for resetting one of a plurality of processors when a service processor detects a looped condition in any one of the plurality of processors, comprising:

reading from a timestamp memory location at a first time;

wherein the timestamp memory location is configured to store a timestamp read by one of the plurality of processors from a timestamp clock;

waiting a predetermined period;

reading from the timestamp memory location at a second time;

comparing a first timestamp read at the first time to a second timestamp read at the second time; and

generating an electrical signal to re-set the one of the plurality of processors if the first timestamp is unchanged from the second timestamp.

9. The method of claim 8, wherein the predetermined period allows the service processor to read a timestamp memory location at a first time for each processor at a first rate less frequent then a second rate at which each processor, during normal operation, reads a second timestamp.

10. The method of claim 8, further comprising generating an index of the plurality of processors by incrementally counting the plurality of processors.

11. The method of claim 10, wherein claim 8 is performed for each of the plurality of processors specified by the index.

12. A computer system including a plurality of processors, operating systems and logical partitions;

the system comprising;

a timestamp clock;

a partition manager to allocate system resources among the plurality of logical partitions;

a memory integral to the partition manager containing a data structure comprising a respective memory location for each of the plurality of processors;

wherein the each of the plurality of processors, during normal operation, is configured to periodically write a timestamp value read from the timestamp clock to its respective timestamp memory location;

a service processor electrically coupled to the plurality of processors, wherein the service processor generates an electrical signal to reset a processor of the plurality of processors if a previously read timestamp value read by the service processor from the respective timestamp memory location is unchanged after a subsequent reading.

13. The computer system of claim 12, wherein the partition manager further comprises an incremental counter to provide an index of the plurality of processors.

14. The computer system of claim 12, wherein the service processor is configured to read the respective timestamp memory location for each processor at a first rate less frequent then a second rate at which each processor, during normal operation, writes the timestamp value to its respective memory location.

15. The computer system of claim 12, wherein the timestamp clock runs in real time.

16. The computer system of claim 12, wherein the service processor comprises integral memory.

17. The computer system of claim 16, wherein the previously read timestamp value is stored in memory in the service processor.

18. The computer system of claim 12, wherein the plurality of processors are multi-threaded processors.

19. In a computer system including a plurality of processors contained in a plurality of logical partitions, wherein at least one operating system is running in each logical partition, a method for re-setting one of a plurality of processors, comprising:

reading from a timestamp memory location at a first time;

wherein the timestamp memory location is configured to store a timestamp read by one of the plurality of processors from a timestamp clock;

waiting a predetermined period;

reading from the timestamp memory location at a second time;

comparing a first timestamp read at the first time to a second timestamp read at the second time; and

generating an electrical signal to re-set one of the plurality of processors if the first timestamp is unchanged from the second timestamp.

20. The method of claim 19, wherein the predetermined period allows the service processor to read a timestamp memory location at a first time for each processor at a first rate less frequent then a second rate at which each processor, during normal operation, reads a second timestamp.

21. The method of claim 20, further comprising generating an index of the plurality of processors by incrementally counting the plurality of processors.

22. The method of claim 21, wherein claim 19 is performed for each of the plurality of processors specified by the index.

23. The method of claim 19, further including assigning a first memory location for each of the plurality of processors, wherein the first read timestamp by each of the plurality of processors, is stored in the first assigned memory location.

24. The method of claim 23, further including assigning a second memory location for each of the plurality of processors, wherein the second read timestamp by each of the plurality of processors, is stored in the second assigned memory location.