Multiprocess programming (V): semaphores

Reference link ： Semaphore

The difference of communication mode between processes

1、 First of all, the simplest is the pipe , Pipelines are divided into anonymous pipelines and famous pipelines
Anonymous pipes can only be used for interprocess communication with kinship , One way transmission ; If you want two-way communication , We need to set up two pipelines ; Hide Name pipeline is a special kind of file , There are buffers in the kernel , The transmission data is unformatted byte stream data , Size is limited .
And famous pipeline , By creating files in the file system , It solves the problem that anonymous pipes can only be used for the communication of processes with kinship problem , Thus, unrelated processes can also communicate .

2、 Message queue , It is actually a message linked list in the kernel , It overcomes the problem that the pipeline communication data is an unformatted byte stream . But the communication speed of message queue is not timely , Because every time data is written and read, it must be copied in user mode and kernel mode

3、 Shared memory , By allocating a shared memory space directly in memory , Solved the problem of data copy , Multiple processes can Through the address To directly access this shared memory space , However, data disorder may occur when multiple processes compete at the same time
Shared memory is the most efficient way

4、 Semaphore , It is a means of process synchronization , It can be used to protect shared resources , Ensure that only one process is accessing shared resources at any one time .
Usually used with shared memory , That is, after obtaining the lock , Then operate on the shared memory （ Write / read ）