libc malloc与Linux内核伙伴分配器 [英] libc malloc vs linux kernel buddy allocator

问题描述

malloc是否担心Linux内核中的内部碎片?例如，当我要分配5个页面时，将malloc舍入为2:5-> 8的幂，以避免内核内部碎片，因为linux内核使用伙伴系统作为页面分配器.

Does malloc wory about internal fragmentation in linux kernel? For example when I want to allocate 5 pages, will malloc round up size to make it power of 2: 5->8 to avoid internal fragmentation in kernel, because linux kernel uses buddy system as page allocator.

推荐答案

至少对于glibc，它实际上并不关心内核中的碎片.除了很小或非常大的分配外，它大多是最佳匹配"分配器.这是glibc"

At least for glibc, it doesn't really care about fragmentation in the kernel. It is mostly a "best-fit" allocator except for very small or very large allocations. Here is an extract from the comments near the top of glibc's "malloc.c":

• 为什么使用此malloc?

• Why use this malloc?

这不是最快，最节省空间，最可移植或有史以来最可调的malloc.但是它是最快的同时也是最节省空间，最便于携带和可调的设备之一.这些因素之间的一致平衡可带来良好的通用性malloc密集型程序的分配器.该算法的主要特性是:

This is not the fastest, most space-conserving, most portable, or most tunable malloc ever written. However it is among the fastest while also being among the most space-conserving, portable and tunable. Consistent balance across these factors results in a good general-purpose allocator for malloc-intensive programs. The main properties of the algorithms are:

• 对于大型(> = 512字节)请求，它是一个最适合的分配器，通常通过FIFO(即最近最少使用)确定联系.
• 对于较小的请求(默认情况下为< = 64字节)，这是一个缓存分配器，用于维护快速回收的块池.
• 介于两者之间，并且对于大型和小型请求，它确实它可以尝试同时实现两个目标的最佳方法.
• 对于非常大的请求(默认情况下== 128KB)，它依赖于系统内存映射工具(如果支持).

有关较长的但有点过时的高级描述，请参阅 http://gee.cs.oswego.edu/dl/html/malloc.html

For a longer but slightly out of date high-level description, see http://gee.cs.oswego.edu/dl/html/malloc.html

如手册页 查看全文