Internal memory distribution, cleaning and releasing method, and internal memory management apparatus

Abstract

The present invention discloses a memory distributing, clearing and releasing method and also discloses a memory management device which comprises a memory distributing module used for applying and distributing memory blocks for the memory from a vacant queue with most proper size, a fragment recovering module which is used for recovering internal chips which are produced when the memory is distributed and plugs the internal chips into different vacant queue and a memory combining module which is used for combining the memory blocks when a maximum vacant memory block can not satisfy the memory application. The present invention distributes the memory application according to need and can reduce the internal chips. The released memory blocks are processed for prolonging combination, and memory vibration caused by frequent cleavage and combination can be avoided, thereby reducing the cost for the frequent cleavage and combination of a system and improving the performance of the system.

Description

technical field [0001] The invention relates to memory management technology, in particular to a method for allocating, clearing and releasing memory and a device for memory management. Background technique [0002] With the widespread use of linux systems, the management of dynamic memory in the system has become more and more important. Because if the dynamic memory is short or improperly managed, it will cause the entire system to respond slowly, or even crash the entire system. [0003] In order to ensure that the Linux memory management mechanism can run efficiently, many technologies are used in modern operating systems, among which the buddy algorithm is more commonly used. The algorithm was first proposed by Donald E. Knuth in 1968, and it is a fast classic algorithm for dynamic memory management. [0004] In this algorithm there are multiple free queues with a block length of 2 k The free blocks of pages are all in the same queue. When you want to allocate a mem...

AI Technical Summary

Problems solved by technology

Although these algorithms provide more suitable memory block sizes, it seems to achieve the purpose of reducing internal fragmentation, but after careful analysis, it is not difficult to find that when more memory block sizes are provided, the system needs to increase the maintenance of new size memory blocks. Table entries, thus increasing the space overhead; at the same time, due to the finer granularity of the memory block size, the number of times to split and recycle partners is increased, which increases the time overhead; and experiments have proved that the total amount of internal fragments in the system is always maintained at 25%-40 % This constant interval does not effectively reduce

Images