USING TRANSACTIONAL EXECUTION FOR RELIABILITY AND RECOVERY OF TRANSIENT FAILURES

• US 20150212906A1
• Filed: 01/24/2014
• Published: 07/30/2015
• Est. Priority Date: 01/24/2014
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Gschwind, Michael Karl, Salapura, Valentina
• Granted Patent: US 9,317,379 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 11/1407 Checkpointing the instruction stream
• CPC Class Codes: G06F 11/1438 Restarting or rejuvenating
• CPC Class Codes: G06F 11/1451 by selection of backup contents
• CPC Class Codes: G06F 11/1474 in transactions G06F16/20 takes precedence
• CPC Class Codes: G06F 13/24 using interrupt G06F13/32 takes precedence
• CPC Class Codes: G06F 2201/805 Real-time
• CPC Class Codes: G06F 2201/825 the problem or solution involving locking
• CPC Class Codes: G06F 2201/84 Using snapshots, i.e. a logical point-in-time copy of the data
• CPC Class Codes: G06F 3/0619 in relation to data integrity, e.g. data losses, bit errors
• CPC Class Codes: G06F 3/0629 Configuration or reconfiguration of storage systems
• CPC Class Codes: G06F 3/0673 Single storage device
• CPC Class Codes: G06F 9/466 Transaction processing

Abstract

Executing each portion of a stream of program instructions as a transaction for reliability, a computer system supporting transactional execution mode processing is provided. Included is determining that an instruction in a portion of the stream of program instructions begins a transaction; based on beginning the transaction, saving a snapshot of system state information and executing the portion of the stream of program instructions as a transaction until an end-mode test point in the stream of program instruction is reached. Based on reaching the end-mode test point, committing store data of the transaction to memory; and based on the stream of program instructions not being complete, automatically beginning a new transaction of a next portion of the stream of program instructions or based on aborting the transaction, re-executing the transaction based on the saved snapshot of the system state information.

Claims

1. A method for executing each portion of a stream of program instructions as a transaction for reliability, the method comprising:

determining, by a computer system configured to support transactional execution mode processing, that an instruction in a portion of the stream of program instructions is to begin a transaction in transactional execution mode;

based on beginning the transaction in transactional execution mode, saving, by the computer system, a snapshot of a system state information and executing, by the computer system, the portion of the stream of program instructions as a transaction until an end-mode test point in the stream of program instruction is reached;

based on reaching the end-mode test point in the stream of program instructions, committing store data of the transaction to memory; and

performing, by the computer system, any one of a) and b);

a) based on the stream of program instructions not being complete, automatically beginning a new transaction in transactional execution mode of a next portion of the stream of program instructions; and

b) based on aborting the transaction, re-executing the transaction based on the saved snapshot of the system state information.

2. The method according to claim 1, wherein the end-mode test point is indicated by any one of:

execution of an end-mode test point instruction, the end-mode test point instruction causing a transaction to end and a next transaction to begin; and

execution of an end-mode test point parameter instruction, the execution of the end-mode test point parameter instruction setting parameters used by the computer system to determine the end-mode test points in the stream of program instructions.

3. The method according to claim 1, further comprising:

encountering a transient failure; and

causing the abort based on the encountered transient failure.

4. The method according to claim 3, further comprising:

returning a specialized abort code with the abort for identifying the transient failure.

5. The method according to claim 3, further comprising:

creating a hardware interrupt, the interrupt handled by one of firmware or privileged software; and

providing failure mode information.

6. A computer program product for executing each portion of a stream of program instructions as a transaction for reliability, the computer program product comprising:

a computer readable storage medium readable by a processing circuit and storing instructions for execution by the processing circuit for performing a method comprising;

determining, by a computer system configured to support transactional execution mode processing, that an instruction in a portion of the stream of program instructions is to begin a transaction in transactional execution mode;

based on beginning the transaction in transactional execution mode, saving, by the computer system, a snapshot of a system state information and executing, by the computer system, the portion of the stream of program instructions as a transaction until an end-mode test point in the stream of program instruction is reached;

based on reaching the end-mode test point in the stream of program instructions, committing store data of the transaction to memory; and

performing, by the computer system, any one of a) and b);

a) based on the stream of program instructions not being complete, automatically beginning a new transaction in transactional execution mode of a next portion of the stream of program instructions; and

b) based on aborting the transaction, re-executing the transaction based on the saved snapshot of the system state information.

7. The computer program product according to claim 6, wherein the end-mode test point is indicated by any one of:

execution of an end-mode test point instruction, the end-mode test point instruction causing a transaction to end and a next transaction to begin; and

execution of an end-mode test point parameter instruction, the execution of the end-mode test point parameter instruction setting parameters used by the computer system to determine the end-mode test points in the stream of program instructions.

8. The computer program product according to claim 6, further comprising:

encountering a transient failure; and

causing the abort based on the encountered transient failure.

9. The computer program product according to claim 8, further comprising:

returning a specialized abort code with the abort for identifying the transient failure.

10. The computer program product according to claim 8, further comprising:

creating a hardware interrupt, the interrupt handled by one of firmware or privileged software; and

providing failure mode information.

11. The computer program product according to claim 8, further comprising:

passing control to a hardware event handler through event based branching.

12. A computer system for executing each portion of a stream of program instructions as a transaction for reliability, the computer system comprising:

a memory; and

a processor in communication with the memory, wherein the computer system is configured to perform a method, said method comprising;

determining, by a computer system configured to support transactional execution mode processing, that an instruction in a portion of the stream of program instructions is to begin a transaction in transactional execution mode;

based on beginning the transaction in transactional execution mode, saving, by the computer system, a snapshot of a system state information and executing, by the computer system, the portion of the stream of program instructions as a transaction until an end-mode test point in the stream of program instruction is reached;

based on reaching the end-mode test point in the stream of program instructions, committing store data of the transaction to memory; and

performing, by the computer system, any one of a) and b);

a) based on the stream of program instructions not being complete, automatically beginning a new transaction in transactional execution mode of a next portion of the stream of program instructions; and

b) based on aborting the transaction, re-executing the transaction based on the saved snapshot of the system state information.

13. The computer system according to claim 12, wherein the end-mode test point is indicated by any one of:

execution of an end-mode test point instruction, the end-mode test point instruction causing a transaction to end and a next transaction to begin; and

execution of an end-mode test point parameter instruction, the execution of the end-mode test point parameter instruction setting parameters used by the computer system to determine the end-mode test points in the stream of program instructions.

14. The computer system according to claim 12, further comprising:

encountering a transient failure; and

causing the abort based on the encountered transient failure.

15. The computer system according to claim 12, wherein a number of instructions after the end-mode test point are executed non-transactionally before automatically beginning the new transaction of the next portion of the stream of program instructions.

16. The computer system according to claim 14, further comprising:

returning a specialized abort code with the abort for identifying the transient failure.

17. The computer system according to claim 14, further comprising:

creating a hardware interrupt, the interrupt handled by one of firmware or privileged software; and

providing failure mode information.

18. The computer system according to claim 14, further comprising:

passing control to a hardware event handler through event based branching.

19. The computer system according to claim 12, wherein the re-executing the transaction is handled by the computer system.

20. The computer system according to claim 12, wherein re-executing the transaction occurs on an alternate processor thread.