Systems and methods of reporting multiple threads involved in a potential data race

• US 7,620,852 B2
• Filed: 03/02/2005
• Issued: 11/17/2009
• Est. Priority Date: 03/02/2005
• Status: Expired due to Fees

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Rodeheffer, Thomas L., Yu, Yuan
• Primary Examiner(s): Beausoliel; Robert
• Assistant Examiner(s): Riad; Amine
• Application Number:
  US11/070,693
• Publication Number:
  US 20060200823A1
• Time in Patent Office: 1,721 Days
• Field of Search: 714/38, 714/48, 714/42, 718/104
• US Class Current: 714/42
• CPC Class Codes: G06F 11/0715 in a system implementing multitasking multitasking per se G06F9/46
• CPC Class Codes: G06F 11/073 in a memory management context, e.g. virtual memory or cache management memory management G06F12/00; testing of static memory units G11C29/00
• CPC Class Codes: G06F 11/0751 Error or fault detection not based on redundancy power supply failures G06F1/30; network fault management H04L41/06
• CPC Class Codes: G06F 11/0778 Dumping, i.e. gathering error/state information after a fault for later diagnosis
• CPC Class Codes: G06F 11/3632 of specific synchronisation aspects
• CPC Class Codes: G06F 11/3636 by tracing the execution of the program
• CPC Class Codes: G06F 2209/523 Mode
• CPC Class Codes: G06F 9/524 Deadlock detection or avoidance

Abstract

Criteria are defined for determining when additional stack dumps would be saved and presented (e.g., to the developer). This provides more information for the developer to use for debugging, and yet does not overburden the developer with irrelevant information. Suspicious activity worthy of an additional stack dumps may be defined as a thread accessing an already suspicious location that is different than the original thread, a write access being made and no write access having yet been made to the target location, or if the lockset being held by the thread accessing the location is omitted from the intersection of all locksets involved in stack dumps for the location.

Claims

1. A method of reporting a warning involving a prospective data race among a plurality of threads of a multithreaded computer program, comprising:

monitoring lock acquisition and memory accesses to a memory location and computing a memory access pattern for each memory access;

determining whether a memory access to the location by a thread results in a suspicious memory access pattern for the location and, if so, dumping the stack of the thread

detecting a subsequent memory access to the location;

determining if at least one predetermined criterion regarding the subsequent memory access is met; and

dumping the stack of the thread performing the subsequent memory access if the at least one predetermined criterion is met.

2. The method of claim 1, wherein the at least one predetermined criterion comprises whether the thread making the subsequent access has not yet been involved in a stack dump for the location.

3. The method of claim 1, wherein the subsequent memory access comprises a write access for a location, and wherein the at least one predetermined criterion comprises no write access has been involved in a stack dump for the location.

4. The method of claim 1, wherein the at least one predetermined criterion comprises a lock set associated with the subsequent memory access omits a lock from the intersection of all lock sets involved in stack dumps for the location.

5. The method of claim 1, further comprising issuing a warning if a suspicious memory access pattern is detected.

6. The method of claim 1, further comprising setting the state of the location to a warning state after detecting the suspicious access pattern.

7. The method of claim 6, further comprising storing an identity of the thread whose memory access resulted in the suspicious access pattern.

8. The method of claim 6, further comprising setting the state of the location to a second warning state if the subsequent memory access is by a different thread.

9. The method of claim 8, further comprising storing an identity of the different thread.

10. The method of claim 8, further comprising reporting a warning pursuant to the second warning state.

11. The method of claim 10, wherein the warning comprises information about the subsequent memory access.

12. A computer-readable storage medium with computer-executable instructions stored thereon, the instructions, when executed by a computer, causing the computer to perform the steps of:

monitoring lock acquisition and memory accesses to a memory location and computing a memory access pattern for each memory access;

determining whether a memory access to the location by a thread results in a suspicious memory access pattern for the location and, if so, dumping the stack of the thread;

detecting a subsequent memory access to the location;

determining if at least one predetermined criterion regarding the subsequent memory access is met; and

dumping the stack of the thread performing the subsequent memory access if the at least one predetermined criterion is met.

13. The computer-readable storage medium of claim 12, wherein the suspicious memory access pattern is related to a prospective data race.

14. The computer-readable storage medium of claim 12, wherein the at least one predetermined criterion comprises whether the thread making the subsequent access has not yet been involved in a stack dump for the location.

15. The computer-readable storage medium of claim 12, wherein the subsequent memory access comprises a write access for a location, and wherein the at least one predetermined criterion comprises no write access has been involved in a stack dump for the location.

16. The computer-readable storage medium of claim 12, wherein the at least one predetermined criterion comprises a lock set associated with the subsequent memory access omits a lock from the intersection of all lock sets involved in stack dumps for the location.

17. The computer-readable storage medium of claim 12, wherein the instructions further cause the computer to set the state of the location to a warning state after detecting the suspicious access pattern.

18. The computer-readable storage medium of claim 17, wherein the instructions further cause the computer to store an identity of the thread whose memory access resulted in the suspicious access pattern.

19. The computer-readable storage medium of claim 17, wherein the instructions further cause the computer to set the state of the location to a second warning state if the subsequent memory access is by a different thread.

20. The computer-readable storage medium of claim 19, wherein the instructions further cause the computer to store an identity of the different thread.