第6章インフラストラクチャーの抽象化層：オペレーティングシステム

By the end of this section, you will be able to:

• Explain various key concepts and components of operating systems

• Discuss the various designs of operating systems

An OS manages computer resources (hardware) and provides services for computer programs (software). The OS works as an interface between the computer user and the system. The OS manages the memory, the files, the hardware, and the software; it also handles the inputs and the outputs such as the keyboard and printer (Figure 6.4).

In this module, we study OS components such as process management and threads, memory and address space management, and device drivers and I/O devices. In addition, we cover OS structures such as monolithic OS design, layered OS design, hardware abstraction layer (HAL), and microkernels.

OS is a complex system that is typically created using the divide and conquer mechanism. That is, the system is divided into small pieces, each of which defines a part of the system. The OS component’s structure is static, and the OS and the hardware are tightly coupled together. The application programming interface (API) is a set of rules and tools that allows different software applications to communicate with each other. Applications make requests to the OS through API. The user can, using a keyboard and/or mouse, interact with the OS through the OS interface. The OS interface could be a graphical user interface (GUI), which allows users to interact with electronic devices through graphical icons and visual indicators (e.g., Windows), or a command line (e.g., DOS). For example, in Figure 6.5, the OS interface would be the window into which the Chrome browser opens.

With a single click, end users seamlessly launch applications, enjoying the simplicity of the process. Have you ever wondered what manages the application once it is launched? The direct answer is the OS. An OS executes many kinds of activities starting from users’ programs, background jobs or scripts, to system programs. There are many system programs such as print managers, name servers, and file servers. Each one of these programs or activities is encapsulated within a process. As you learned in the preceding section, a process is a running program that has an execution context plus the runtime instance of the program itself. Examples of an execution context are program counters (PCs), registers, VM maps, OS resources, and examples of runtime instance are code and data.

Here are some other process-related concepts that will be introduced here and developed as the chapter progresses:

• A register is a high-speed memory storing unit.

• A process can be in running state (when it is executed by the CPU), ready state (when it is waiting in the CPU), or blocked state (when it is waiting for an event to occur).

• The OS’s process module manages the processes via creation, destruction, and scheduling. One way an OS controls access to a resource is through a data type called a semaphore.

• Managing the sharing of a system’s resources to avoid interference and errors is called process synchronization.

Given the complexity of an OS’s tasks (i.e., the management of a computer system’s resources and the scheduling of tasks to make an application do what the user expects it to do) and the speed at which these tasks need to happen, it’s amazing how often technology doesn’t fail. But, of course, it does at times.

Let’s suppose you ran into an OS-related failure this morning—namely, your laptop did not fully boot. What are some steps you would take to diagnose the problem on your own before seeking help or a repair?

As we have learned, a process is an active program. A thread is a smaller or lightweight portion of a process that is managed independently. Table 6.2 shows a comparison between the process and the thread.

A program is passive; it is just bytes on a disk that encode instructions to be run. A process is an instance of a program being executed—or processed—by a processor. Thus, processing involves a program, a process, and a processor. The processor can be a real processor or a virtual processor (i.e., CPU core assigned to a virtual machine). At any time, there may be many processes running copies of the same program (e.g., an editor); each process is separate and usually independent. An OS is responsible for managing these processes.

Different OSs approach process management in different ways. For example, the Windows operating system’s approach consists of adding an operating system call to create a process and other system calls for process operation. The approach used in the UNIX OS is different from this as it consists of splitting the process into two steps using fork and exec functionalities. The fork functionality is used to set up privileges by creating a complete copy of the process, and the exec functionality brings the executable file into memory to start the execution.