What is the Kernel Object in OS

Hi all,

As a Windows software developer, I often need to create, open and operate a variety of kernel objects. System to create and operate several types of kernel objects, such as access to symbolic objects, events, objects, file objects, file mapping objects, I / O completion port objects, job objects, mailbox objects, mutex objects, pipe objects, process objects, beacon object, the thread object and wait for the timer objects. These objects are created by calling the function. Please suggest.

As the kernel object's data structure can only be the kernel access, so the application can not find these data structures in memory and directly to change their content. Microsoft provides this restriction is aimed at ensuring that the kernel object structure maintain a state of consensus. This restriction also enabled Microsoft can not destroy the case of any application of these structures to add, delete and modify data members. If we can not directly change the data structure, then our application, how to operate these kernel objects? Solution is, Windows provides a set of functions in order to use the definition of a very good way to operate on these structures. All the best.

These kernel objects can always be accessed through functions. When you call a function used to create the kernel object, the function returns a handle used to identify the object. The handle can be seen as an opaque value, you are in the process of any thread can use this value. This handle will be passed to every function of Windows, so that the system will know which kernel you want to operate the object. Later in this chapter also handles the details of these features. Best of luck.

To make the operating system become more robust, these handle values are closely related with the process. Therefore, if the handle value is passed to another process in a thread using some form of inter-process communication so that another process using the process handle value of your call by the will fail. Check and reply.

Kernel objects owned by the kernel, rather than owned by the process. In other words, if you call the process of creating a kernel object's function, and then you run the process to terminate, then the kernel object is not necessarily being taken away. In most cases, the object will be withdrawn, but if another process is using the process of creating the core of your object, then the kernel know that in another process to stop using the object will not rescind the previous object, it must be remembered that the , the kernel object's existence time can be compared with a long process that created the object. All the best.

The kernel to know how many processes are using a kernel object, because each object contains a usage count. Usage count of all the kernel object type commonly used data member. When an object just created, its usage count is set to 1. Then, when another process to access an existing kernel object, use the count is incremented 1. When the process termination running, the kernel will automatically determine if the process is still open to all kernel object's usage count. If the kernel object's usage count reduced to 0, the kernel on the withdrawal of the object. This would ensure that the process in the absence of reference to the object, the system does not retain any kernel objects. All the best. :cool: