Method and system for implementing a multi-threaded API stream replay

• US 9,477,575 B2
• Filed: 06/12/2013
• Issued: 10/25/2016
• Est. Priority Date: 06/12/2013
• Status: Active Grant

• Current Assignee: NVIDIA Corporation
• Original Assignee: NVIDIA Corporation
• Inventors: Strauss, Michael C.
• Primary Examiner(s): Deng, Anna
• Assistant Examiner(s): Wu, Junchun
• Application Number:
  US13/916,041
• Publication Number:
  US 20140372990A1
• Time in Patent Office: 1,231 Days
• Field of Search: 717/133
• US Class Current: 1/1
• CPC Class Codes: G06F 11/3003 Monitoring arrangements specially adapted to the computing system or computing system component being monitored
• CPC Class Codes: G06F 11/3093 Configuration details thereof, e.g. installation, enabling, spatial arrangement of the probes
• CPC Class Codes: G06F 11/34 Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
• CPC Class Codes: G06F 11/3414 Workload generation, e.g. scripts, playback
• CPC Class Codes: G06F 11/3466 Performance evaluation by tracing or monitoring
• CPC Class Codes: G06F 11/362 Software debugging
• CPC Class Codes: G06F 11/3636 by tracing the execution of the program
• CPC Class Codes: G06F 2201/865 Monitoring of software

Abstract

A method for debugging and includes receiving a request for capturing a frame generated by a graphics application implementing application threads executing function calls. The function calls are associated with one or more thread specific resources used at the beginning of the capturing process. For each application thread, a corresponding state is determined for each thread specific resource utilized, and a corresponding capture stream is established. For each application thread, executed function calls are captured into the corresponding capture stream. A plurality of captured function calls is arranged in the order they were executed by the graphics application. For each capture stream, a corresponding replay thread is established. Application threads, capture streams, and replay threads exist in a one-to-one-to-one relationship. Captured function calls are replayed in order, wherein each captured function call is executed in a corresponding replay thread based on which application thread executed the captured function call.

Claims

1. A non-transitory computer-readable medium having computer-executable instructions for causing a computer system to perform a method for debugging, comprising:

receiving a request for capturing a frame generated by a graphics application implementing a plurality of application threads for executing a plurality of function calls generating said frame, wherein said plurality of function calls is associated with one or more thread specific resources used at an entry point of said capturing;

for each application thread, determining a corresponding state at said entry point for each corresponding thread specific resource utilized;

for each application thread, establishing a corresponding capture stream, wherein application threads and capture streams exist in a one-to-one relationship;

for each application thread, capturing executed function calls into said corresponding capture stream;

ordering a plurality of captured function calls captured across said plurality of application threads in the order they were executed by said graphics application;

for each capture stream, establishing a corresponding replay thread, wherein capture streams and replay threads exist in a one-to-one relationship; and

replaying a plurality of captured function calls in the order they were captured, wherein a captured function call is executed in a corresponding replay thread based on which application thread executed said captured function call.

2. The non-transitory computer-readable medium of claim 1, wherein said ordering in said method comprises:

associating a time-stamp for each captured function call.

3. The non-transitory computer-readable medium of claim 1, wherein said ordering in said method comprises:

determining a corresponding sequence number for each captured function call, wherein no two captured function calls have the same sequence number.

4. The non-transitory computer-readable medium of claim 1, wherein said replaying in said method comprises:

receiving from each replay thread information related to its next captured function call in a position for execution;

determining a smallest ordered captured function call; and

executing said smallest ordered captured function call.

5. The non-transitory computer-readable medium of claim 1, wherein said replaying in said method comprises:

determining a pool of non-executed captured function calls;

determining a smallest ordered captured function call; and

executing said smallest ordered captured function call.

6. The non-transitory computer-readable medium of claim 1, wherein said determining a corresponding state in said method further comprises:

for each thread, capturing a context utilized for executing a corresponding first function call.

7. The non-transitory computer-readable medium of claim 1, wherein said capturing executed function calls in said method further comprises:

capturing information related to a corresponding captured executed function call.

8. The non-transitory computer-readable medium of claim 1, wherein said replaying a plurality of captured function calls in said method further comprises:

detecting an end-of-frame marker;

restoring for each thread its corresponding state for each of said corresponding thread specific resource utilized; and

determining a pool of non-executed captured function calls;

determining a smallest ordered captured function call; and

executing said smallest ordered captured function call.

9. A system for debugging, comprising:

a multi-threaded processor for executing a plurality of function calls generated by a graphics application in association with a frame;

a graphics application programming interface (API) configured to execute said plurality of function calls to generate said frame suitable for display, wherein said plurality of function calls is associated with one or more thread specific resources used at an entry point of capturing said frame;

an interception layer configured to intercept said plurality of function calls between said graphics application and said graphics API in response to a request for capturing said frame, wherein said interception layer is configured to capture for each thread, a corresponding state for each corresponding thread specific resource utilized at said entry point;

a plurality of capture streams each configured to capture executed function calls from a corresponding application thread, wherein application threads and capture streams exist in a one-to-one relationship; and

a plurality of replay threads each configured to replay corresponding captured function calls originally executed on a corresponding application thread, wherein capture streams and replay threads exist in a one-to-one relationship; and

an arbiter configured for ordering a plurality of captured function calls captured across said plurality of application threads in the order they were executed by said graphics application, and wherein said arbiter replays a plurality of captured function calls in the order they were captured.

10. The system of claim 9, wherein said arbiter is configured for associating a time-stamp for each captured function call.

11. The system of claim 9, wherein said arbiter is configured for determining a corresponding sequence number for each captured function call, wherein no two captured function calls have the same sequence number.

12. The system of claim 9, wherein said arbiter is configured for receiving from each replay thread information related to its next captured function call in a position for execution, and for determining a smallest ordered captured function call, and for executing said smallest ordered captured function call.

13. The system of claim 9, wherein said arbiter is configured for determining a pool of non-executed captured function calls, for determining a smallest ordered captured function call, and for executing said smallest ordered captured function call.

14. The system of claim 9, wherein said arbiter is configured for restoring for each thread its corresponding state for each of said corresponding thread specific resource utilized when detecting an end-of-frame marker, and determining a pool of non-executed captured function calls, and determining a smallest ordered captured function call, and executing said smallest ordered captured function call.

15. The system of claim 9, wherein said interception layer is configured for capturing information related to a corresponding captured executed function call.

16. A computer system comprising:

a processor; and

memory coupled to said processor and having stored therein instructions that, if executed by said computer system, cause said computer system to execute a method of debugging comprising;

receiving a request for capturing a frame generated by a graphics application implementing a plurality of application threads for executing a plurality of function calls generating said frame, wherein said plurality of function calls is associated with one or more thread specific resources used at an entry point of said capturing;

for each thread, determining a corresponding state at said entry point for each corresponding thread specific resource utilized;

for each application thread, establishing a corresponding capture stream, wherein application threads and capture streams exist in a one-to-one relationship;

for each thread, capturing executed function calls into said corresponding capture stream;

ordering a plurality of captured function calls captured across said plurality of application threads in the order they were executed by said graphics application;

for each capture stream establishing a corresponding replay thread, wherein capture streams and replay threads exist in a one-to-one relationship; and

replaying a plurality of captured function calls in the order they were captured, wherein a captured function call is executed in a corresponding replay thread based on which application thread executed said captured function call.

17. The computer system of claim 16, wherein said ordering in said method comprises:

determining a corresponding sequence number for each captured function call, wherein no two captured function calls have the same sequence number.

18. The computer system of claim 16, wherein said replaying in said method comprises:

determining a pool of non-executed captured function calls;

determining a smallest ordered captured function call; and

executing said smallest ordered captured function call.

19. The computer system of claim 16, wherein said replaying a plurality of captured function calls in said method further comprises:

detecting an end-of-frame marker;

restoring for each thread its corresponding state for each of said corresponding thread specific resource utilized; and

determining a pool of non-executed captured function calls;

determining a smallest ordered captured function call; and

executing said smallest ordered captured function call.

20. The computer system of claim 16, wherein said determining a corresponding state for each corresponding thread specific resource utilized defines a context in a corresponding application thread.