Shulou(Shulou.com)06/03 Report--

This article mainly shows you "there are several GC algorithms in JVM". The content is simple and clear. I hope it can help you solve your doubts. Let me lead you to study and learn this article "there are several GC algorithms in JVM".

Introduction

In the process of running the program, it will produce a lot of memory garbage (some memory objects that do not have a reference point to belong to memory garbage, because these objects can no longer be accessed, the program can not use them, and they have died for the program). In order to ensure the performance of the program, the java virtual machine continues to carry out automatic garbage collection (GC) while the program is running. There are four main GC algorithms for JVM:

Reference counting algorithm (Reference counting) algorithm idea:

When each object is created, a counter is bound to the object. Whenever there is a reference to the object, the counter is incremented by one; whenever a reference to it is deleted, the counter is minus one. In this way, when no reference points to the object, the object dies and the counter is 0, and the object should be garbage collected.

Core ideas:

An additional counter RC is stored for each object, and the value of RC is used to determine whether the object is dead or not, so as to determine whether the GC operation is performed.

Advantages:

simple

The computational cost is dispersed

The "ghost time" is short (ghost time refers to the time between the death and recovery of the object, being in a ghostly state)

Disadvantages:

Incomplete (objects that are easy to miss circular references)

Weak concurrency support

Take up extra memory space

Examples are shown in the figure:

Initial state:

After changing the reference:

Tag-cleanup algorithm (Mark-Sweep) algorithm idea:

Stores a tag bit for each object to record the state of the object (alive or dead). It is divided into two stages, one is the marking phase, in which the tag bits are updated for each object to check whether the object is dead; the second stage is the cleanup phase, which clears the dead objects and performs GC operations.

Advantages

The biggest advantage is that, compared with the reference counting method, the reference of each living object in the tag-clear algorithm only needs to find one, and it can be judged to be alive by finding one.

In addition, this algorithm is more comprehensive than reference counting and does not cost much in pointer operation. More importantly, this algorithm does not move the position of the object (the latter two algorithms involve the problem of moving position).

Shortcoming

For a long ghost time, it takes a lot of time to judge that the object is dead, so the time from the object's death to the time it is reclaimed is too long.

Every living object is traversed in the marking phase; all objects are scanned in the cleanup phase, so the algorithm is complex.

The failure to move objects can lead to situations where a lot of debris space is not available.

The example is shown in the picture.

In this picture, the gray object in the circle is the dead object, marked as dead, waiting to be cleared.

The idea of marking-finishing algorithm

The mark-finishing method is an improved version of the mark-removal method. Similarly, in the marking phase, the algorithm marks all objects as alive and dead; the difference is that in the second stage, the algorithm does not directly clean up the dead objects, but organizes all the living objects. put it in another space, and then clear all the remaining objects. In this way, the purpose of marking-finishing is achieved.

Advantages

This algorithm does not generate a large amount of fragment space like the mark-clear algorithm.

Shortcoming

If there are too many surviving objects, more replication operations will be performed in the finishing phase, resulting in a reduction in the efficiency of the algorithm.

Examples

As shown in the figure:

The above is the marking phase, and the following is the status after finishing. As you can see, this algorithm does not generate a lot of fragmented memory space.

The idea of replication algorithm

The algorithm divides the memory into two parts on average, and then uses only one part of the memory at a time. When this part of the memory is full, all the living objects in the memory are copied to another memory, and then the previous memory is emptied, using only this part of the memory. Loop.

Note:

The difference between this algorithm and the tag-collation algorithm is that instead of replicating in the same area, all living objects are copied to another region.

Advantages

Easy to implement

No memory fragmentation

Shortcoming

Each time you run, half of the memory is empty, resulting in only half the available memory space.

These are all the contents of the article "there are several GC algorithms in JVM". Thank you for reading! I believe we all have a certain understanding, hope to share the content to help you, if you want to learn more knowledge, welcome to follow the industry information channel!

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.