Art of multi-processor programming Reading Notes (2)-mutex

Mutex is a multi-processor. Program The most common collaboration method in design. We put a resource into the critical section: A single point of time can only be executed by one thread. This feature is called mutex. The standard method for implementing mutex is to use an ilock object with the following interfaces. Public   Interface Ilock
{
Void Setlock ();
Void Unlock ();
}

Public class counter
{
Private int _ value;
Private ilock _ lock;

Public int getandincrement ()
{
_ Lock. setlock ();
Try
{
Int temp = _ value;
_ Value = temp + 1;
Return _ value;
}
Finally
{
_ Lock. Unlock ();
}
}
}

If a thread executes the setlock () method call, the thread obtains a lock (or lock). If the unlock () method call is executed, the thread is called releasing the lock (or unlocking ). In the implementation of the shared counter above, the ilock domain is used to ensure the mutex feature of the object. The thread must call setlock () and unlock () in the specified way (). If a thread meets the following conditions, it is well-structured:
1. A critical section is associated only with a unique ilock object.
2. The setlock () method of the object is called when the thread is about to enter the critical section.
3. Call the unlock () method when the thread leaves the critical section.