Detailed introduction to python multi-thread applications

When using python multithreading, you need to pay attention to a lot of things. In fact, you only need to master these technical segments to ensure the integrity of this application. Next, let's take a look at how to perform the operation.

Today, I read about how to securely call multiple threads in python in multithreading applications in the last day. At the beginning, I thought it was a big fight, the Global Interpreter Lock and Thread State of the python language were a little dizzy. Later, I went through various articles and help documents for mutual reference, we found that for versions 2.4/2.5, we provided PyGILState_Ensure and PyGILState_Release.

1. Define an encapsulation class to ensure pair use of PyGILState_Ensure and PyGILState_Release, which can be nested.

 
 

  
  
#include <python.h> 
  
  
class PyThreadStateLock  
  
  
{  
  
  
public:  
  
  
PyThreadStateLock(void)  
  
  
{  
  
  
state = PyGILState_Ensure( );  
  
  
}  
  
  
~PyThreadStateLock(void)  
  
  
{  
  
  
PyGILState_Release( state );  
  
  
}  
  
  
private:  
  
  
PyGILState_STATE state;  
  
  
};  
 
 

2. In the main thread, this process

// Initialization
Py_Initialize ();
// Initialization thread support
PyEval_InitThreads ();
// Run the command before starting the child thread. To release the global lock obtained by PyEval_InitThreads, the Child thread may not be able to obtain the global lock.
PyEval_ReleaseThread (PyThreadState_Get ());

// Other processes, such as promoter threads
......

// Ensure that all sub-thread calls are completed
PyGILState_Ensure ();
Py_Finalize ();
// You cannot call any python API later

The above is a detailed introduction to python multithreading. Hope to help you.