Shulou(Shulou.com)06/02 Report--

What is the principle of CAS and the optimization of JDK8? aiming at this problem, this article introduces the corresponding analysis and solution in detail, hoping to help more partners who want to solve this problem to find a more simple and feasible method.

What is CAS?

CAS-CompareAndSet is the basis for the implementation of JDK atomic variable classes AtomicInteger, AtomicLong, AtomicInteger, AtomicBoolean, AtomicReference, etc. For example, for a shared variable int, even a simple self-increment operation is not atomic, and multithreading increases at the same time, which may cause the value of the variable to be smaller than expected. However, you can use the incrementAndGet () method of AtomicInteger to manipulate variables so that the result is the same as expected. Unlike traditional locking, the getAndDecrement () method does not lock the code. The code looks like this:

Public final int incrementAndGet () {for (;;) {int current = get (); int next = current + 1; if (compareAndSet (current, next)) return next;}}

The underlying layer is implemented through sun.misc.Unsafe 's native method compareAndSwapInt, which is atomic.

Comparison with synchronized

The difference between optimistic lock and pessimistic lock

Performance comparison

Synchronized is blocking, CAS update is non-blocking, but it will retry without thread context switching overhead. For most simple operations, whether in the case of low concurrency or high concurrency, the performance of this optimistic non-blocking method is much higher than that of pessimistic blocking mode.

Application scenario

Used to implement optimistic non-blocking algorithm to ensure that the shared variables used in the current thread method are not changed by other threads. CAS is widely used in non-blocking containers.

It is used to implement the pessimistic blocking algorithm, which is used in the principle implementation of explicit locks. For example, in the reentrant count, the lock () method is set to 1 through the CAS method, and the call to unlock is set to decrement 1. If multiple threads call the Lock method at the same time, it will inevitably lead to unsuccessful atomic modification, ensuring the lock mechanism and exclusiveness.

Possible problems

ABA problem, the ordinary CAS operation is not atomic, because it is possible that another thread changed the value but changed it back, then the optimistic locking method cannot guarantee atomicity. If the business needs to avoid this situation, you can use AtomicStampedReference's compareAndSet (V expectedReference, V newReference, int expectedStamp, int newStamp) method to update the atom only when the value and timestamp are equal, and each update modifies the current time into the atomic variable.

Optimization of JDK8

JAVA8 adds LongAdder and DoubleAdder to further optimize the atomic variables, mainly using the mechanism of segmented CAS. If you do not use LongAdder and use AtomicLong, it will produce continuous spin in the case of high concurrency, resulting in low efficiency. He divides a number into several numbers, and the parameter of CompareAndSet method is only one of these numbers, which reduces the probability of spin and improves the efficiency.

The answer to the question about the principle of CAS and the optimization of JDK8 is shared here. I hope the above content can be of some help to everyone. If you still have a lot of doubts to be solved, you can follow the industry information channel to learn more about it.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.