android内存管理分析（Android memory management analysis）.docVIP

android内存管理分析（Android memory management analysis） Many developers come from J2ME or J2EE, and the memory usage and understanding are not in place. The Android development network gives you some architectural guidance to prevent the emergence of bean curd waste engineering. As an intelligent platform based on Java, Android is necessary for developing high performance and quality software to understand the memory management mechanism of Android programs. Android Dalvik VM and SUN JVM on the basis of what no big difference is only the optimized bytecode, we want to know what time GC what time with recycle, in the end do not use finalization, because of the release of memory allocation of Java memory requires only new developers do not need to be displayed, but the probability of memory leaks caused by this the more high. 1., for regular developers, you need to understand four reference methods for Java, such as strong reference, soft reference, weak reference, and virtual reference. Some of the more complex programs are likely to have OutOfMemoryError like exceptions in the long run. 2. and not too much to expect GC, no object can be displayed is set to null, such as obj=null, here to remind everybody, Java GC is using a directed graph, determine whether an object is effective to see other objects can reach the object vertex, directed graph for phase list two fork tree overhead is as can be imagined. The 3.Android for each program to distribute memory through Runtime (totalMemory) freeMemory (two) some memory information acquisition method of VM system for heap memory access, through the Dalvik.VMRuntime class (getMinimumHeapSize) method to obtain the minimum heap memory, and display release soft references can call the class (gcSoftReferences) methods to obtain more memory to run. 4. for multi-threaded processing, if concurrent threads are many, and frequently created and released, you can solve the thread creation bottleneck through the thread pool of the concurrent