


A brief discussion of PHP source code 31: Basics of the heap layer in the PHP memory pool
This article mainly introduces about PHP source code 31: The foundation of the heap layer in the PHP memory pool, which has a certain reference value. Now I share it with you. Friends in need can refer to it
A brief discussion on PHP source code 31: The basics of the heap layer in the PHP memory pool
[Overview]
PHP’s memory manager is hierarchical . This manager has three layers: storage layer, heap layer and emalloc/efree layer. The storage layer is introduced in PHP source code reading notes 30: Storage layer in PHP memory pool. The storage layer actually applies for memory to the system through functions such as malloc() and mmap(), and releases the requested memory through the free() function. Memory. The storage layer usually applies for relatively large memory blocks. The large memory applied here does not mean the memory required by the storage layer structure. It is just that when the heap layer calls the allocation method of the storage layer, the memory it applies for in the segment format is relatively large. , the role of the storage layer is to make the memory allocation method transparent to the heap layer.
Above the storage layer is the heap layer we want to learn about today. The heap layer is a scheduling layer that interacts with the emalloc/efree layer above to split the large blocks of memory applied for through the storage layer and provide them on demand. There is a set of memory scheduling strategies in the heap layer, which is the core area of the entire PHP memory allocation management.
All the following sharing is based on the situation that ZEND_DEBUG has not been opened.
First look at the structures involved in the heap layer:
[Structure]
/* mm block type */typedef struct _zend_mm_block_info { size_t _size;/* block的大小*/ size_t _prev;/* 计算前一个块有用到*/} zend_mm_block_info; typedef struct _zend_mm_block { zend_mm_block_info info;} zend_mm_block; typedef struct _zend_mm_small_free_block {/* 双向链表 */ zend_mm_block_info info; struct _zend_mm_free_block *prev_free_block;/* 前一个块 */ struct _zend_mm_free_block *next_free_block;/* 后一个块 */} zend_mm_small_free_block;/* 小的空闲块*/ typedef struct _zend_mm_free_block {/* 双向链表 + 树结构 */ zend_mm_block_info info; struct _zend_mm_free_block *prev_free_block;/* 前一个块 */ struct _zend_mm_free_block *next_free_block;/* 后一个块 */ struct _zend_mm_free_block **parent;/* 父结点 */ struct _zend_mm_free_block *child[2];/* 两个子结点*/} zend_mm_free_block; struct _zend_mm_heap { int use_zend_alloc;/* 是否使用zend内存管理器 */ void *(*_malloc)(size_t);/* 内存分配函数*/ void (*_free)(void*);/* 内存释放函数*/ void *(*_realloc)(void*, size_t); size_t free_bitmap;/* 小块空闲内存标识 */ size_t large_free_bitmap; /* 大块空闲内存标识*/ size_t block_size;/* 一次内存分配的段大小,即ZEND_MM_SEG_SIZE指定的大小,默认为ZEND_MM_SEG_SIZE (256 * 1024)*/ size_t compact_size;/* 压缩操作边界值,为ZEND_MM_COMPACT指定大小,默认为 2 * 1024 * 1024*/ zend_mm_segment *segments_list;/* 段指针列表 */ zend_mm_storage *storage;/* 所调用的存储层 */ size_t real_size;/* 堆的真实大小 */ size_t real_peak;/* 堆真实大小的峰值 */ size_t limit;/* 堆的内存边界 */ size_t size;/* 堆大小 */ size_t peak;/* 堆大小的峰值*/ size_t reserve_size;/* 备用堆大小*/ void *reserve;/* 备用堆 */ int overflow;/* 内存溢出数*/ int internal;#if ZEND_MM_CACHE unsigned int cached;/* 已缓存大小 */ zend_mm_free_block *cache[ZEND_MM_NUM_BUCKETS];/* 缓存数组/ #endif zend_mm_free_block *free_buckets[ZEND_MM_NUM_BUCKETS*2];/* 小块空闲内存数组 */ zend_mm_free_block *large_free_buckets[ZEND_MM_NUM_BUCKETS];/* 大块空闲内存数组*/ zend_mm_free_block *rest_buckets[2];/* 剩余内存数组 */ };
For the memory operation function in the heap structure, if use_zend_alloc is no, malloc-type memory allocation is used, this At this time, all operations do not go through the memory management in the heap layer, and directly use functions such as malloc.
The default size of compact_size is 2 * 1024 * 1024 (2M). If the variable ZEND_MM_COMPACT is set, the size is specified. If the peak memory exceeds this value, the compact function of storage will be called, just this function The current implementation is empty and may be added in subsequent versions.
reserve_size is the size of the reserve heap, by default it is ZEND_MM_RESERVE_SIZE, its size is (8*1024)
*reserve is the reserve heap, its size is reserve_size, it is used to report errors when memory overflows .
[About USE_ZEND_ALLOC]
The environment variable USE_ZEND_ALLOC can be used to allow selection of malloc or emalloc memory allocation at runtime. Using malloc-type memory allocation will allow external debuggers to observe memory usage, while emalloc allocation will use the Zend memory manager abstraction, requiring internal debugging.
[zend_startup() -> start_memory_manager() -> alloc_globals_ctor()]
static void alloc_globals_ctor(zend_alloc_globals *alloc_globals TSRMLS_DC){ char *tmp; alloc_globals->mm_heap = zend_mm_startup(); tmp = getenv("USE_ZEND_ALLOC"); if (tmp) { alloc_globals->mm_heap->use_zend_alloc = zend_atoi(tmp, 0); if (!alloc_globals->mm_heap->use_zend_alloc) {/* 如果不使用zend的内存管理器,同直接使用malloc函数*/ alloc_globals->mm_heap->_malloc = malloc; alloc_globals->mm_heap->_free = free; alloc_globals->mm_heap->_realloc = realloc; } }}
[Initialization]
[zend_mm_startup()]
Initialize the allocation plan of the storage layer , initialize the segment size, compression boundary value, and call zend_mm_startup_ex() to initialize the heap layer.
[zend_mm_startup() -> zend_mm_startup_ex()]
[Memory Alignment]
Memory alignment is used in PHP’s memory allocation. The memory alignment calculation obviously has two goals: one is to reduce The number of CPU memory accesses; the second is to keep the storage space efficient enough.
# define ZEND_MM_ALIGNMENT 8 #define ZEND_MM_ALIGNMENT_MASK ~(ZEND_MM_ALIGNMENT-1) #define ZEND_MM_ALIGNED_SIZE(size)(((size) + ZEND_MM_ALIGNMENT - 1) & ZEND_MM_ALIGNMENT_MASK) #define ZEND_MM_ALIGNED_HEADER_SIZEZEND_MM_ALIGNED_SIZE(sizeof(zend_mm_block)) #define ZEND_MM_ALIGNED_FREE_HEADER_SIZEZEND_MM_ALIGNED_SIZE(sizeof(zend_mm_small_free_block))
PHP uses memory alignment in the memory of allocated blocks. If the lower three digits of the required memory size are not 0 (not divisible by 8), add 7 to the lower digits. And ~7 performs an AND operation, that is, for memory sizes that are not an integer multiple of 8, the memory size is completed to be divisible by 8.
On win32 machines, the numerical sizes corresponding to some macros are:
ZEND_MM_MIN_SIZE=8
ZEND_MM_MAX_SMALL_SIZE=272
ZEND_MM_ALIGNED_HEADER_SIZE=8
ZEND_MM_ALIGNED_FREE_HEADER_SIZE=16
ZEND_MM_MIN_ALLOC_BLOCK_SIZE=8
ZEND_MM_ALIGNED_MIN_HEADER_SIZE =16
ZEND_MM_ALIGNED_SEGMENT_SIZE=8
If you want to allocate a block with a size of 9 bytes, the actual allocated size is ZEND_MM_ALIGNED_SIZE(9 8)=24
[of the block Positioning】
The two right digits of the allocated memory are used to mark the type of memory.
The definition of its size is #define ZEND_MM_TYPE_MASK ZEND_MM_LONG_CONST(0×3)
The code shown below is the positioning of the block
#define ZEND_MM_NEXT_BLOCK(b)ZEND_MM_BLOCK_AT(b, ZEND_MM_BLOCK_SIZE(b)) #define ZEND_MM_PREV_BLOCK(b)ZEND_MM_BLOCK_AT(b, -(int)((b)->info._prev & ~ZEND_MM_TYPE_MASK)) #define ZEND_MM_BLOCK_AT(blk, offset)((zend_mm_block *) (((char *) (blk))+(offset))) #define ZEND_MM_BLOCK_SIZE(b)((b)->info._size & ~ZEND_MM_TYPE_MASK)#define ZEND_MM_TYPE_MASKZEND_MM_LONG_CONST(0x3)
The next element of the current block is the current block The head position plus the length of the entire block (minus the length of the type).
The previous element of the current block is the head position of the current block minus the length of the previous block (removing the length of the type).
Regarding the length of the previous block, it is set to the result of the OR operation of the size of the current block and the block type during the initialization of the block.
The above is the entire content of this article. I hope it will be helpful to everyone's study. For more related content, please pay attention to the PHP Chinese website!
Related recommendations:
A brief discussion of PHP source code 30: The storage layer in the PHP memory pool
A brief discussion of PHP Source code twenty-nine: About the inheritance of interfaces
## A brief discussion of PHP source code twenty-eight: About the class structure and inheritance
The above is the detailed content of A brief discussion of PHP source code 31: Basics of the heap layer in the PHP memory pool. For more information, please follow other related articles on the PHP Chinese website!

Hot AI Tools

Undresser.AI Undress
AI-powered app for creating realistic nude photos

AI Clothes Remover
Online AI tool for removing clothes from photos.

Undress AI Tool
Undress images for free

Clothoff.io
AI clothes remover

Video Face Swap
Swap faces in any video effortlessly with our completely free AI face swap tool!

Hot Article

Hot Tools

Notepad++7.3.1
Easy-to-use and free code editor

SublimeText3 Chinese version
Chinese version, very easy to use

Zend Studio 13.0.1
Powerful PHP integrated development environment

Dreamweaver CS6
Visual web development tools

SublimeText3 Mac version
God-level code editing software (SublimeText3)

Hot Topics

Alipay PHP...

JWT is an open standard based on JSON, used to securely transmit information between parties, mainly for identity authentication and information exchange. 1. JWT consists of three parts: Header, Payload and Signature. 2. The working principle of JWT includes three steps: generating JWT, verifying JWT and parsing Payload. 3. When using JWT for authentication in PHP, JWT can be generated and verified, and user role and permission information can be included in advanced usage. 4. Common errors include signature verification failure, token expiration, and payload oversized. Debugging skills include using debugging tools and logging. 5. Performance optimization and best practices include using appropriate signature algorithms, setting validity periods reasonably,

Session hijacking can be achieved through the following steps: 1. Obtain the session ID, 2. Use the session ID, 3. Keep the session active. The methods to prevent session hijacking in PHP include: 1. Use the session_regenerate_id() function to regenerate the session ID, 2. Store session data through the database, 3. Ensure that all session data is transmitted through HTTPS.

The application of SOLID principle in PHP development includes: 1. Single responsibility principle (SRP): Each class is responsible for only one function. 2. Open and close principle (OCP): Changes are achieved through extension rather than modification. 3. Lisch's Substitution Principle (LSP): Subclasses can replace base classes without affecting program accuracy. 4. Interface isolation principle (ISP): Use fine-grained interfaces to avoid dependencies and unused methods. 5. Dependency inversion principle (DIP): High and low-level modules rely on abstraction and are implemented through dependency injection.

How to debug CLI mode in PHPStorm? When developing with PHPStorm, sometimes we need to debug PHP in command line interface (CLI) mode...

How to automatically set the permissions of unixsocket after the system restarts. Every time the system restarts, we need to execute the following command to modify the permissions of unixsocket: sudo...

Static binding (static::) implements late static binding (LSB) in PHP, allowing calling classes to be referenced in static contexts rather than defining classes. 1) The parsing process is performed at runtime, 2) Look up the call class in the inheritance relationship, 3) It may bring performance overhead.

Sending JSON data using PHP's cURL library In PHP development, it is often necessary to interact with external APIs. One of the common ways is to use cURL library to send POST�...
