Adaptive Lock for Multicore Systems with Suspendable Threads
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for operating locks in multicore computing systems are inadequate for managing both suspendable and non-suspendable threads across multiple runtime environments, as they primarily focus on suspendable threads and do not account for the presence of non-suspendable threads.
Innovation Solution
A method for operating a lock in a computing system with multiple processing units, where threads can be either suspendable or non-suspendable, involves determining the suspendability of the holder thread and configuring the requester thread to either spin or wait, depending on the holder's status, and includes a wait queue for suspended threads to be resumed when the lock is released.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If existing lock techniques are used in multicore systems with multiple runtime environments, then suspendable threads can be managed, but non-suspendable threads cannot be properly handled
Solution Approach 1:
The lock operation mode is dynamically adjusted based on the holder thread's suspendability status. When the holder is non-suspendable, requester threads spin continuously. When the holder is suspendable, requester threads are added to a wait queue. This dynamic adaptation allows the same lock mechanism to correctly handle different thread types without compromising reliability.
Solution Approach 2:
The system changes the operational parameters of the lock based on thread characteristics. The key parameter change is the state of the requester thread: spinning state for non-suspendable holders versus suspended state in wait queue for suspendable holders. This parameter change enables the lock to be versatile across different thread types while maintaining correct operation.
2Speed
If requester threads continuously spin to acquire a lock held by a non-suspendable thread, then lock acquisition responsiveness is improved, but system resource consumption increases
Solution Approach 1:
Different resource consumption strategies are applied locally based on the holder thread's characteristics. For non-suspendable holders, requester threads use spinning (higher local resource consumption) to achieve fast acquisition. For suspendable holders, requester threads use wait queue suspension (lower local resource consumption). This local differentiation optimizes the overall system by matching resource usage to actual needs.
3Use of energy by moving object
If requester threads are added to a wait queue when the holder is suspendable, then system resource efficiency is improved, but lock acquisition time may increase
Solution Approach 1:
The potential harm of increased lock acquisition time is converted into a benefit by efficiently utilizing system resources. When threads are suspended in the wait queue, the processing unit is freed to execute other tasks, improving overall system throughput. The time delay is transformed from a pure loss into an opportunity for better resource utilization, where other threads can make progress while waiting threads are suspended.
Data Source
AI summary
A method for operating a lock in a computing system having plural processing units and running under multiple runtime environments is provided. When a requester thread attempts to acquire the lock while the lock is held by a holder thread, determine whether the holder thread is suspendable or non-suspendable. If the holder thread is non-suspendable, put the requester thread in a spin state regardless of whether the requester thread is suspendable or non-suspendable; otherwise determines whether the requester thread is suspendable or non-suspendable unless the requester thread quits acquiring the lock. If the requester thread is non-suspendable, arrange the requester thread to attempt acquiring the lock again; otherwise add the requester thread to a wait queue as an additional suspended thread. Suspended threads stored in the wait queue are allowable to be resumed later for lock acquisition. The method is applicable for the computing system with a multicore processor.