Framework for executing multiple threads and sharing resources in a multithreaded computer programming environment

• US 7,827,559 B1
• Filed: 04/24/2006
• Issued: 11/02/2010
• Est. Priority Date: 04/24/2006
• Status: Active Grant

• Current Assignee: Real-Time Innovations, Inc.
• Original Assignee: Real-Time Innovations, Inc.
• Inventors: Thiebaut, Stefaan Sonck, Sin, Elaine Yee Ting, Pardo-Castellote, Gerardo, Rhee, Stephen Jisoo, Joshi, Rajive
• Primary Examiner(s): An; Meng-Ai
• Assistant Examiner(s): KUMABE, BLAKE K
• Application Number:
  US11/410,563
• Time in Patent Office: 1,653 Days
• Field of Search: 718/102, 718/104
• US Class Current: 718/104
• CPC Class Codes: G06F 9/526 Mutual exclusion algorithms

Abstract

Techniques for execution of multiple threads in a multithreaded computing programming environment are disclosed. The techniques are especially well suited for environments that use multilayered programming architecture where a higher layer can build on the functions provided by a lower layer where the delay time is an important consideration. In one aspect, the conceptual notion of a “Worker” effectively serves to represent the thread-specific execution context for a thread of execution (“thread”) in a multithreaded computing environment. Another aspect, provides the notion of an Exclusion Area (EA) as logical lock that serves to protect shared resources in a multithreaded environment. The combination of the worker and EA are used to provide a powerful framework that, among other things, allows minimizing of the delay time.

Claims

1. A method of protecting a shared resource in a multithreaded computing environments where multiple threads of execution (threads) can access said shared resource, said method comprising:

associating a first thread of execution (first thread) with a first thread-specific execution context (first Worker) that can effectively provide the execution context for executing operational computer code for said first thread when said first thread effectively activates said operational code for execution;

associating a first shared resource with a first Exclusion Area (first EA) provided as logical lock to effectively protect said first shared resource from multiple threads that may try to access said shared resource at the same time;

determining whether said first Worker associated with said first thread is trying to visit (or enter) said first Exclusion Area (first EA), thereby determining whether said first thread is attempting to access said first shared resource effectively protected by said first Exclusion Area (first EA);

determining whether said first Exclusion Area (first EA) is being visited by another Worker associated with another thread when said determining determines that said first Worker is trying to visit said first Exclusion Area (first EA); and

not allowing said first Worker to visit said first Exclusion Area (first EA) when said determining determines that said first Exclusion Area (first EA) is being visited by said another Worker, thereby effectively preventing said first thread from accessing said first shared resource when said first shared resource is being accessed by said another thread.

8. A method as recited in claim 1, wherein said method further comprises:

mapping a plurality of EAs to a single EA bounded to a single physical lock, thereby protecting a plurality of shared resources associated with plurality of EA'"'"'s with said single physical lock.

9. A method as recited in claim 8, wherein said plurality of EA are associated with shared resources in a Data Reader'"'"'s or Data Writers incoming or outgoing data paths across a plurality of software layers, thereby allowing said plurality of Data Readers or Data Writers to deliver data concurrently.

10. A method as recited in claim 9, wherein said method comprises:

determining whether to use a single physical lock per each Data Reader or Data Writer or per group of Data Readers or Data Writers.

11. A method as recited in claim 10, wherein said determining of whether to use a single physical lock per each Data Reader or Data Writer or per group of Data Readers or Data Writers is made based on an application program code via a configuration policy.

13. A computer system operable to perform the method recited in claim 1.

2. A method as recited in 1, wherein said method further comprises:

allowing said first Worker to visit said first Exclusion Area (first EA) when said determining determines that said first Exclusion Area (first EA) is not being visited by said another Worker, thereby effectively allowing said first thread to access said first shared resource when said first shared resource is not being accessed by said another thread.

3. A method as recited in claim 2, wherein said method further comprises:

allowing said Worker to re-visit said first Exclusion Area (EA) that it has previously entered.

4. A method as recited in claim 3, wherein the Worker is considered to be visiting if it has entered an EA more times then it has left the EA.

5. A method as recited in claim 2, wherein said method further comprises:

assigning a first number to said first Exclusion Area (first EA) that effectively protects said first shared resource;

assigning a second number to a second Exclusion Area (second EA) that effectively protects a second shared resource;

determining whether said first Worker is trying to visit said second Exclusion Area (second EA) when said first Worker is visiting said first Exclusion Area (first EA);

comparing said first and second numbers when said determining determines that said first Worker is trying to visit said second Exclusion Area (second EA); and

determining, based on said comparing, whether to allow said first Worker to visit said second Exclusion Area (second EA).

6. A method as recited in claim 5, wherein said method further comprises:

allowing said first Worker to visit said second Exclusion Area (second EA), thereby allowing said first thread to access both said first and second shared resources.

7. A method as recited in claim 6, wherein said first and second Exclusion Areas (EAs) are mapped to a single physical lock, thereby allowing said first and second shared resources to be effectively protected by said single physical lock.

12. A non-transitory computer readable medium storing computer program code executed by a processor for protecting a shared resource in a multithreaded computing environments where multiple threads of execution (threads) can access said shared resource, said computer readable medium comprising:

computer program code for associating a first thread of execution (first thread) with a first thread-specific execution context (first Worker) that can effectively provide the execution context for executing operational computer code for said first thread when said first thread effectively activates said operational code for execution;

computer program code for associating a first shared resource with a first Exclusion Area (first EA) provided as logical lock to effectively protect said first shared resource from multiple threads that may try to access said shared resource at the same time;

computer program code for determining whether said first Worker associated with said first thread is trying to visit (or enter) said first Exclusion Area (first EA), thereby determining whether said first thread is attempting to access said first shared resource effectively protected by said first Exclusion Area (first EA);

computer program code for determining whether said first Exclusion Area (first EA) is being visited by another Worker associated with another thread when said determining determines that said first Worker is trying to visit said first Exclusion Area (first EA); and

computer program code for not allowing said first Worker to visit said first Exclusion Area (first EA) when said determining determines that said first Exclusion Area (first EA) is being visited by said another Worker, thereby effectively preventing said first thread form accessing said first shared resource when said first shared resource is being accessed by said another thread.