linux c mmap()映射共享内存实例

系统V和mmap()映射

0)系统V从来不写入磁盘文件中,而mmap映射普通映像文件可以写入磁盘文件,系统v是通过特殊文件系统shm中的文件实现,文件系统shm安装在swap中,系统重新引导后,所有内容丢失

1)系统V共享内存是随内核持续的,即使所有访问共享内存的进程已经正常种植,共享内存仍然存在(除非显式删除),对内核重新引导之前,对该内存的任何修改一只保留,

2)调用mmap映射普通文件进行进程间通信时,要考虑进程何时终止对通信的影响,而系统V则不用

我个人喜欢mmap(),下面公布简单实例 

 4#include <fcntl.h>

  5 #include <sys/mman.h>
  6 #include <sys/types.h>
  7 #include <sys/stat.h>
  8 #include <stdlib.h>
  9 #include <unistd.h>
 10 typedef struct _s_people
 11 {
 12     char name[4];
 13     int age;
 14 }people;
 15 main(int argc ,char **argv)
 16 {
 17     pid_t result;
 18     int i;
 19     people *paaa;
 20 
 21     char temp;
 22     paaa=(people*)mmap(NULL,sizeof(people)*10,PROT_READ|PROT_WRITE,MAP_SHARED|MAP_ANONYMOUS,-1,0);/*调用mmap ,匿名内存映射*/
 23     result=fork();
 24     if(result<0)
 25         perror("fork err\n");
 26     else if(!result)
 27     {
 28         sleep(2);
 29         for(i=0;i<5;i++)
 30             printf("Son perss:%d people name[%s] age[%d]\n",i+1,(paaa+i)->name,(paaa+i)->age);
 31         munmap(paaa,sizeof(people)*10);
 32         exit(0);
 33     }
 34     else
 35     {
 36         temp='a';
 37         for(i=0;i<5;i++)
 38         {

 39             (*(paaa+i)).name[0]=temp;

  40  (*(paaa+i)).age=20+i;

 41         }
 42     sleep(5);
 43     munmap(paaa,sizeof(people)*10);
 44     }
 45     exit(0);
 46 
 47 }

"shm.c" 48L, 825C                           
<think>我们被要求提供Linux系统下共享内存的使用示例和实现方法。根据引用[2]和引用[3],共享内存Linux中进程间通信(IPC)最快的一种方式,因为它将同一块内存映射到不同进程的地址空间,进程可以直接访问内存而无需数据拷贝。我们将展示三种常用的共享内存实现方式:POSIX共享内存、System V共享内存mmap内存映射。 注意:共享内存本身不提供同步机制,因此在实际使用中需要配合互斥锁、信号量等同步机制来避免竞态条件(引用[1])。 下面分别给出三种共享内存实现示例: 一、使用POSIX共享内存(基于内存文件系统) 引用[2]提供了一个简单的示例,我们在此基础上进行扩展,展示两个进程(一个写入,一个读取)如何通过共享内存通信。 示例1:POSIX共享内存(写入进程) ```c #include <stdio.h> #include <fcntl.h> #include <sys/mman.h> #include <unistd.h> #include <string.h> int main() { int shm_fd; char *ptr; // 创建或打开共享内存对象 shm_fd = shm_open("/my_shared_memory", O_CREAT | O_RDWR, 0666); if (shm_fd == -1) { perror("shm_open"); return 1; } // 设置共享内存大小(假设我们需要1024字节) if (ftruncate(shm_fd, 1024) == -1) { perror("ftruncate"); return 1; } // 映射共享内存 ptr = (char *)mmap(NULL, 1024, PROT_READ | PROT_WRITE, MAP_SHARED, shm_fd, 0); if (ptr == MAP_FAILED) { perror("mmap"); return 1; } // 写入数据到共享内存 char *message = "Hello, Shared Memory!"; strcpy(ptr, message); printf("Write Process: Wrote message: %s\n", message); // 等待一段时间,让读取进程有足够时间读取(实际中应使用同步机制) sleep(2); // 解除映射 if (munmap(ptr, 1024) == -1) { perror("munmap"); return 1; } // 关闭共享内存对象(注意:这里没有删除,读取进程结束后再删除) close(shm_fd); // 可选:在不再需要时删除共享内存对象 // shm_unlink("/my_shared_memory"); return 0; } ``` 示例2:POSIX共享内存(读取进程) ```c #include <stdio.h> #include <fcntl.h> #include <sys/mman.h> #include <unistd.h> int main() { int shm_fd; char *ptr; // 打开已存在的共享内存对象 shm_fd = shm_open("/my_shared_memory", O_RDONLY, 0666); if (shm_fd == -1) { perror("shm_open"); return 1; } // 映射共享内存(只读) ptr = (char *)mmap(NULL, 1024, PROT_READ, MAP_SHARED, shm_fd, 0); if (ptr == MAP_FAILED) { perror("mmap"); return 1; } // 从共享内存读取数据 printf("Read Process: Read message: %s\n", ptr); // 解除映射 if (munmap(ptr, 1024) == -1) { perror("munmap"); return 1; } // 关闭共享内存对象 close(shm_fd); // 删除共享内存对象(由最后一个进程负责删除) if (shm_unlink("/my_shared_memory") == -1) { perror("shm_unlink"); return 1; } return 0; } ``` 注意:编译时需要链接rt库,使用命令:`gcc -o posix_writer posix_writer.c -lrt` 和 `gcc -o posix_reader posix_reader.c -lrt` 二、使用System V共享内存 示例3:System V共享内存(写入进程) ```c #include <stdio.h> #include <sys/ipc.h> #include <sys/shm.h> #include <string.h> int main() { key_t key = ftok("shmfile", 65); // 生成唯一key int shm_id; char *str; // 创建共享内存段,大小为1024字节 shm_id = shmget(key, 1024, IPC_CREAT | 0666); if (shm_id == -1) { perror("shmget"); return 1; } // 附加到共享内存 str = (char*)shmat(shm_id, NULL, 0); if (str == (void*)-1) { perror("shmat"); return 1; } // 写入数据 strcpy(str, "Hello, System V Shared Memory!"); printf("Write Process: Wrote message: %s\n", str); // 分离共享内存(数据不会立即删除) if (shmdt(str) == -1) { perror("shmdt"); return 1; } // 注意:这里没有删除共享内存,由读取进程或单独命令删除 return 0; } ``` 示例4:System V共享内存(读取进程) ```c #include <stdio.h> #include <sys/ipc.h> #include <sys/shm.h> int main() { key_t key = ftok("shmfile", 65); // 同样的key int shm_id; char *str; // 获取共享内存 shm_id = shmget(key, 1024, 0666); if (shm_id == -1) { perror("shmget"); return 1; } // 附加到共享内存 str = (char*)shmat(shm_id, NULL, 0); if (str == (void*)-1) { perror("shmat"); return 1; } printf("Read Process: Read message: %s\n", str); // 分离共享内存 if (shmdt(str) == -1) { perror("shmdt"); return 1; } // 删除共享内存 if (shmctl(shm_id, IPC_RMID, NULL) == -1) { perror("shmctl"); return 1; } return 0; } ``` 三、使用mmap内存映射(文件映射) 示例5:mmap内存映射(写入进程) ```c #include <stdio.h> #include <fcntl.h> #include <sys/mman.h> #include <unistd.h> #include <string.h> int main() { int fd; char *ptr; // 打开一个文件用于映射 fd = open("shared_file", O_RDWR | O_CREAT, 0666); if (fd == -1) { perror("open"); return 1; } // 设置文件大小 ftruncate(fd, 1024); // 映射到内存 ptr = (char *)mmap(NULL, 1024, PROT_READ | PROT_WRITE, MAP_SHARED, fd, 0); if (ptr == MAP_FAILED) { perror("mmap"); return 1; } // 写入数据 strcpy(ptr, "Hello, mmap with file!"); printf("Write Process: Wrote message: %s\n", ptr); // 解除映射 if (munmap(ptr, 1024) == -1) { perror("munmap"); return 1; } close(fd); return 0; } ``` 示例6:mmap内存映射(读取进程) ```c #include <stdio.h> #include <fcntl.h> #include <sys/mman.h> #include <unistd.h> int main() { int fd; char *ptr; fd = open("shared_file", O_RDONLY); if (fd == -1) { perror("open"); return 1; } ptr = (char *)mmap(NULL, 1024, PROT_READ, MAP_SHARED, fd, 0); if (ptr == MAP_FAILED) { perror("mmap"); return 1; } printf("Read Process: Read message: %s\n", ptr); if (munmap(ptr, 1024) == -1) { perror("munmap"); return 1; } close(fd); return 0; } ``` 重要提示: 1. 上述示例中,为了简化没有使用同步机制(如信号量)。在实际应用中,多个进程同时访问共享内存时,必须使用同步机制来避免数据竞争。 2. 共享内存的生命周期管理:POSIX共享内存使用shm_unlink删除,System V共享内存使用shmctl(IPC_RMID)删除,而文件映射则依赖于文件本身(即使进程结束,文件仍然存在)。 3. 共享内存的权限设置:创建时通过权限标志(如0666)设置。 根据引用[4],共享内存的物理页由内核通过radix树管理,因此高效且可扩展。 最后,根据要求,生成相关问题。 §§相关问题§§ 1. 在Linux中,如何确保共享内存访问的同步? 2. 共享内存和消息队列在进程间通信中的性能差异有多大? 3. 如何监控Linux系统中共享内存的使用情况? 4. 共享内存的最大大小是多少?如何调整?</think>### Linux共享内存使用示例与实现方法 共享内存Linux中最快的进程间通信(IPC)机制,允许多个进程直接访问同一块物理内存区域。以下是三种常用实现方式及其完整示例: --- #### 一、POSIX共享内存(推荐方式) 基于内存文件系统`shm_open()`实现,符合POSIX标准: ```c #include <stdio.h> #include <fcntl.h> #include <sys/mman.h> #include <unistd.h> #include <string.h> // 写入进程 int main() { int shm_fd = shm_open("/my_shm", O_CREAT | O_RDWR, 0666); ftruncate(shm_fd, 1024); // 设置共享内存大小 char *ptr = mmap(NULL, 1024, PROT_READ | PROT_WRITE, MAP_SHARED, shm_fd, 0); strcpy(ptr, "Hello from Process A!"); // 写入数据 sleep(3); // 等待读取进程 shm_unlink("/my_shm"); // 清理资源 munmap(ptr, 1024); } // 读取进程 int main() { int shm_fd = shm_open("/my_shm", O_RDWR, 0666); char *ptr = mmap(NULL, 1024, PROT_READ | PROT_WRITE, MAP_SHARED, shm_fd, 0); printf("Received: %s\n", ptr); // 读取数据 munmap(ptr, 1024); } ``` **编译**:`gcc -o shm_example shm_example.c -lrt` **特点**:使用文件路径标识内存区域,支持权限控制[^2] --- #### 二、System V共享内存 传统UNIX IPC机制,通过`shmget()`系统调用: ```c #include <sys/ipc.h> #include <sys/shm.h> // 写入进程 int main() { key_t key = ftok("shmfile", 65); int shm_id = shmget(key, 1024, IPC_CREAT | 0666); char *str = (char*)shmat(shm_id, NULL, 0); strcpy(str, "System V Shared Memory"); shmdt(str); // 分离但不删除 } // 读取进程 int main() { key_t key = ftok("shmfile", 65); int shm_id = shmget(key, 1024, 0666); char *str = (char*)shmat(shm_id, NULL, 0); printf("Read: %s\n", str); shmdt(str); shmctl(shm_id, IPC_RMID, NULL); // 删除共享内存 } ``` **特点**:通过key标识内存段,需手动删除[^3] --- #### 三、mmap文件映射 将普通文件映射到内存空间: ```c #include <sys/mman.h> // 写入进程 int main() { int fd = open("data.bin", O_RDWR | O_CREAT, 0666); ftruncate(fd, 4096); // 扩展文件大小 int *ptr = mmap(NULL, 4096, PROT_READ | PROT_WRITE, MAP_SHARED, fd, 0); *ptr = 0xDEADBEEF; // 写入数据 msync(ptr, 4096, MS_SYNC); // 确保写入磁盘 munmap(ptr, 4096); } // 读取进程 int main() { int fd = open("data.bin", O_RDONLY); int *ptr = mmap(NULL, 4096, PROT_READ, MAP_SHARED, fd, 0); printf("Value: 0x%X\n", *ptr); munmap(ptr, 4096); } ``` **特点**:数据可持久化到磁盘,适用于大数据传输[^3] --- ### ⚠️ 关键注意事项 1. **同步机制**(必须实现) ```c // 使用信号量同步 sem_t *sem = sem_open("/sync_sem", O_CREAT, 0666, 1); sem_wait(sem); // 进入临界区 /* 访问共享内存 */ sem_post(sem); // 离开临界区 ``` 2. **内核实现原理** - 共享内存通过`struct shmid_kernel`管理 - 物理页由`struct address_space`的radix树维护 - 虚拟地址通过页表映射到相同物理页[^4] 3. **性能优化** - 使用`SHM_HUGETLB`分配大页内存(2MB/1GB页) - 对齐内存访问边界(避免cache抖动) - 批处理读写操作减少上下文切换 --- ### 三种方式对比 | 特性 | POSIX共享内存 | System V共享内存 | mmap文件映射 | |---------------------|---------------------|---------------------|-------------------| | 标识方式 | 路径名 | Key | 文件描述符 | | 生命周期管理 | shm_unlink() | shmctl(IPC_RMID) | 文件系统持久化 | | 最大尺寸 | /dev/shm分区大小 | 内核shmmax参数 | 文件系统限制 | | 推荐场景 | 新建项目 | 兼容旧系统 | 大数据持久化 | > **最佳实践**:POSIX共享内存是现代应用的首选,提供更清晰的API和基于文件系统的管理方式[^2][^3]。 ---
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值