Computer system serialization control method involving unlocking global lock of one partition, after completion of machine check analysis regardless of state of other partition locks

• US 6,886,064 B2
• Filed: 03/28/2002
• Issued: 04/26/2005
• Est. Priority Date: 03/28/2002
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: McIntosh, Gordon D., Lee, Van Hoa, Dawkins, George John, Desai, Prakash Vinodrai
• Primary Examiner(s): Myers, Paul R.
• Assistant Examiner(s): PHAN, RAYMOND NGAN
• Application Number:
  US10/112,477
• Publication Number:
  US 20030208670A1
• Time in Patent Office: 1,125 Days
• Field of Search: 710200-220, 710104-110, 707 8- 10, 707200-204, 707100-103, 709201-203, 709208-216, 709223-228, 709/248
• US Class Current: 710/200
• CPC Class Codes: G06F 9/52 Program synchronisation; Mutual exclusion, e.g. by means of semaphores
• CPC Class Codes: Y10S 707/99938 Concurrency, e.g. lock management in shared database

Abstract

In a computer system having a logical-partitioned server, each partition of the server is provided with its own separate lock and access corridor, in addition to a global lock. When the locking of a partition lock is followed by the locking of the global lock, the system is serialized. The partition locks are controlled by system firmware on behalf of an OS isolating each partition; however, the global lock is controlled by the system firmware to be unlocked independent of the lock/unlock status of the partition locks. In this manner, the ability or inability of an OS that issued a machine check interrupt to unlock its partition lock after the machine check analysis is complete is irrelevant; once the machine check analysis is complete, the system firmware unlocks the global lock, giving other partitions access to shared system resources to run their own machine checks.

Claims

1. A method of controlling the serializing of a computer system having a logical-partitioned server operating under the control of a system controller, said logical-partitioned server having plural partitions, said method comprising the steps of:

providing an individual partition lock for each of said plural partitions;

providing a global lock associated with said system controller;

locking said global lock and the individual partition lock associated with a first partition when an operating system (OS) associated with said first partition requests serialization of said computer system to perform a first operation, thereby serializing said computer system on said first operation;

performing said first operation; and

unlocking said global lock by said system controller upon completion of said first operation, regardless of the locked/unlocked state of any of said partition locks.

2. A method as set forth in claim 1, further comprising the steps of:

locking the partition lock associated with a second of said plural partitions when an OS associated with said second partition requests serialization of said computer system to perform a second operation;

determining the locked/unlocked state of said global lock; and

if said global lock is determined to be unlocked, locking said global lock, thereby serializing said computer system on said second operation;

performing said second operation; and

unlocking said global lock upon completion of said second operation, regardless of the locked/unlocked state of any of said partition locks.

3. A method as set forth in claim 2, wherein said steps are repeated whenever a request for serialization is made by any OS=s associated with any of said plural partitions.

4. A method as set forth in claim 2, wherein said system controller comprises a hypervisor.

5. A method as set forth in claim 2, wherein said first operation comprises a machine-check analysis.

6. A method as set forth in claim 5, wherein said second operation comprises a machine check analysis.

7. A method of managing the handling of machine checks in a logical-partitioned server operating under the control of a system controller, said logical-partitioned server having plural partitions, said method comprising the steps of:

providing a discrete lockable access corridor between each said partition and said system controller, the locked/unlocked state of each discrete lockable access corridor being controlled by its associated partition;

providing a global lock associated with said system controller;

locking said global lock and the discrete lockable access corridor associated with a first of said plural partitions when an operating system (OS) associated with said first of said plural partitions issues a machine-check interrupt;

performing a first machine check analysis based on said machine check interrupt issued by said OS of said first partition; and

unlocking said global lock by said system controller upon completion of said first machine check analysis, regardless of the locked/unlocked state of any of said discrete lockable access corridors.

8. A method as set forth in claim 7, further comprising the steps of:

locking the discrete access corridor associated with a second of said plural partitions when an OS associated with said second partition issues a machine-check interrupt;

determining the locked/unlocked state of said global lock; and

if said global lock is determined to be unlocked, locking said global lock, thereby connecting said second partition and said system controller along the discrete access corridor associated with said second partition;

performing a second machine check analysis based on said machine check interrupt issued by said OS of said second partition; and

unlocking said global lock upon completion of said second machine check analysis, regardless of the locked/unlocked state of any of said discrete lockable access corridors.

9. A method as set forth in claim 8, wherein said steps are repeated whenever a request for serialization is made by any OS=s associated with any of said plural partitions.

10. A method as set forth in claim 8, wherein said system controller comprises a hypervisor.

11. A system of managing the handling of machine checks in a logical-partitioned server operating under the control of a system controller, said logical-partitioned server having plural partitions, said system comprising:

means for providing a discrete lockable access corridor between each of said plural partitions and said system controller, the locked/unlocked state of each discrete lockable access corridor being controlled by its associated partition;

means for providing a global lock associated with said system controller;

means for locking said global lock and the discrete lockable access corridor associated with a first of said plural partitions when an operating system (OS) associated with said first of said plural partitions issues a machine-check interrupt;

means for performing a first machine check analysis based on said machine check interrupt issued by said OS of said first partition; and

means for unlocking said global lock by said system controller upon completion of said first machine check analysis, regardless of the locked/unlocked state of any of said discrete lockable access corridors.

12. A system as set forth in claim 11, further comprising:

means for locking the discrete access corridor associated with a second of said plural partitions when an OS associated with said second partition issues a machine-check interrupt;

means for determining the locked/unlocked state of said global lock; and

means for locking said global lock if said global lock is determined to be unlocked, thereby connecting said second partition and said system controller along the discrete access corridor associated with said second partition;

means for performing a second machine check analysis based on said machine check interrupt issued by said OS of said second partition; and

means for unlocking said global lock upon completion of said second machine check analysis, regardless of the locked/unlocked state of any of said discrete lockable access corridors.

13. A system as set forth in claim 12, wherein said system controller comprises a hypervisor.

14. A computer program product recorded on a computer readable medium that manages the handling of machine checks in a logical-partitioned server operating under the control of a system controller, said logical-partitioned server having plural partitions, said computer program product comprising:

computer-readable means for providing a discrete lockable access corridor between each said partition and said system controller, the locked/unlocked state of each discrete lockable access corridor being controlled by its associated partition;

computer-readable means for providing a global lock associated with said system controller;

computer-readable means for locking said global lock and the discrete lockable access corridor associated with a first of said plural partitions when an operating system (OS) associated with said first of said plural partitions issues a machine-check interrupt;

computer-readable means for performing a first machine check analysis based on said machine check interrupt issued by said OS of said first partition; and

computer-readable means for unlocking said global lock by said system controller upon completion of said first machine check analysis, regardless of the locked/unlocked state of any of said discrete lockable access corridors.

15. A computer program product as set forth in claim 14, further comprising:

computer-readable means for locking the discrete access corridor associated with a second of said plural partitions when an OS associated with said second partition issues a machine-check interrupt;

computer-readable means for determining the locked/unlocked state of said global lock; and

computer-readable means for locking said global lock if said global lock is determined to be unlocked, thereby connecting said second partition and said system controller along the discrete access corridor associated with said second partition;

computer-readable means for performing a second machine check analysis based on said machine check interrupt issued by said OS of said second partition; and

computer-readable means for unlocking said global lock upon completion of said second machine check analysis, regardless of the locked/unlocked state of any of said discrete lockable access corridors.

16. A computer program product as set forth in claim 15, wherein said system controller comprises a hypervisor.

17. An apparatus for managing the handling of machine checks in a logical-partitioned server operating under the control of a system controller, said logical-partitioned server having plural partitions, said apparatus comprising:

a discrete lockable access corridor between each of said plural partitions and said system controller, the locked/unlocked state of each discrete lockable access corridor being controlled by its associated partition;

a global lock associated with said system controller;

an instruction processor executing instructions which lock said global lock and the discrete lockable access corridor associated with a first of said plural partitions when an operating system (OS) associated with said first of said plural partitions issues a machine-check interrupt;

a hardware processor performing a first machine check analysis based on said machine check interrupt issued by said OS of said first partition; and

an instruction processor configured with instructions for said system controller to unlock said global lock upon completion of said first machine check analysis, regardless of the locked/unlocked state of any of said discrete lockable access corridors.