What is stack and the difference between stacks

This article was first published on the official account 【 Programmer Huazai 】

------------------

First of all, let's explain , Heaps and stacks discussed here , It refers to “ Heap area ” and “ The stack area ”, It is not the heap and stack mentioned in the data structure .

To understand the stack in program memory , You need to know C Memory model of language .

C The memory model of language is divided into 5 Districts ： Code section , Global area , The constant area , Stack area and stack area .

1. Code section

Store binary code area , stay ELF called .text paragraph .

2. Global area

Area for storing global and static variables .

The initialized global and static variables are in an area of the global , stay ELF Named as .data paragraph . 

Uninitialized global variables and static variables are in the adjacent area , stay ELF In Chinese, it means .bss paragraph .

The memory of the global area is allocated by the system , And at the end of the program , Released by the system .

3. The constant area

The area where constants are stored . No modification allowed .

The memory of this area is at the end of the program , Released by the system .

4. Heap area （Heap）

An area for storing variables created by programmers , It is a discontinuous area , In general use new,alloc,calloc,malloc Keyword to create variables , These variables need to be called by the programmer delete,release,free Keywords can be released . If there is no call , It will cause a memory leak .

5. The stack area （Stack）

Store the parameters of the function , Region of local variables , Automatically allocated and released by the compiler . Variables are usually released outside the scope . Formally, it is similar to stack in data structure . Because in CPU Allocate memory in the instruction set of , So the operation is fast and efficient . But the space is very limited , for example iOS The stack size of is 2M.

Heap and stack concepts

Pile up It is a large discontinuous area , Generally, the linked list in the data structure is used to organize and manage , In the process of distribution , It is easy to produce memory fragments .

Stack It's a small continuous area , Generally, the queue is used according to the first in and last out （First In Last Out）（FILO） Sort in a different way , And only at one end （ To the top of the stack （top）） Insert and delete data .

The difference between heap and stack

Heap and stack , There are two different ways for the operating system to manage the memory space of the process , There are mainly the following differences ：

（1） Different ways of distribution ：

Stack ： Automatically assigned and released by the operating system , No need for us to operate it manually . The allocation method is similar to the stack in the data structure .

Pile up ： The application and release work is completed by the programmer , If programmers don't release , Memory leaks occur . The allocation method is similar to the linked list in the data structure .

（2） The size of the space is different ：

Stack ： Stack space is much smaller than heap space . such as iOS Stack as 2M,64 Bit Linux Default 10MB.

Pile up ： Theoretically , The heap size that the process can apply for is the size of virtual memory .

（3） Different storage contents ：

Stack ： Store function return address 、 Parameters 、 Local variables and registers .

Pile up ： The specific storage content is filled in by the programmer .

（4） Different caching methods ：

Stack ： Stored in L1 cache , Put it into storage space when called , Release immediately after the call .

Pile up ： Stored in L2 cache , Reference counting is generally used （iOS） Conduct management .

（5） The data structure is different ：

Pile up ： Generally, the tree in the data structure is used to manage .

Stack ： Generally, the queue of first in and last out is used to manage .

（6） Distribution efficiency is different ：

The stack is automatically allocated by the operating system , stay CPU The instruction set of completes the allocation and is supported by the hardware layer （ Special registers are responsible for storage , Special instructions are responsible for pressing the stack ）.

Heap is made up of system kernel API Interface to complete application and management , The implementation mechanism is complex , And frequent memory applications are prone to memory fragmentation . In this way, the allocation efficiency of stack is much higher than that of heap .

These are the six differences of stack .

iOS App Memory layout

So let's see iOS App Memory layout , Shown below .

Just like the above figure ,

1. In the whole memory space , The top is the high address , At the bottom is the low address ;

2. The code segment and data segment are in the lowest address bit , That is, the lowest end ;

3. On top of it is a heap , It's used to store variables ; about ObjC Come on , Namely new,alloc,calloc,malloc Keyword modifies the variable , The address expansion mode of the heap is from low to high , That is, when it is not enough, apply for a space address up

4. Up is the stack , It mainly stores the parameters of the function 、 Data such as local variables , Its expansion mode is from high to low .

 

That's all iOS App Memory layout of , Understand and master iOS App Memory layout of , Especially stack allocation and expansion trend , It can help us analyze the positioning problem well .