Intel Architecture

When speaking about the Assembly language, people usually imagine a sort of unknown and dangerous beast, which obeys only the weirdest representatives of the programming community, or a gun that may only be used for shooting your own leg. Just as any prejudice, this one is rooted in ignorance and the primal fear of the unknown. The purpose of this book is not only to help you overcome this prejudice, but also to show how the Assembly language may become a powerful tool, a sharp lancet, that will help you perform certain tasks, even sophisticated ones, with elegance and relative simplicity, avoiding the unnecessary complications which are, sometimes, implied by high-level languages.

First of all, what is the Assembly language? To put it simply and precisely, we may safely define the Assembly language as symbolic or human readable machine code as each Assembly instruction translates into a single machine instruction (with a few exceptions). To be even more precise, there is no such thing as a single Assembly language as, instead, there are numerous Assembly languages--one per platform, where a platform is a programmable device. Almost any programmable device with a certain instruction set may have its own Assembly language, but this is not always so. Exceptions are devices such as, for example, NAND flash chips, which have their own command set, but have no means for fetching instructions from memory and executing them without implicitly being told to do so.

In order to be able to effectively use the Assembly language, one has to have a precise understanding of the underlying platform, as programming in the Assembly language means "talking" directly to the device. The deeper such understanding is, the more efficient is Assembly programming; however, we are not going to look at this in great detail, as this is beyond the scope of the book. One book would not be enough to cover each and every aspect of the specific architecture. Since we are going to concentrate on the Intel architecture during the course of this book, let's try to obtain at least a general understanding of Intel's x86/AMD64 architectures, and try to enrich it and make it deeper.

This chapter, in particular, covers processor registers and the functionality thereof and briefly describes memory organization (for example, segmentation and paging).

• General purpose registers: Despite the fact that some of them have special meanings under certain circumstances, these registers, as the name of the group states, may be used for any purpose.
• Floating point registers: These registers are used for floating point operations.
• Segment registers: These registers are hardly accessed by applications (the most common case is setting up structured exception handlers on Windows); however, it is important to cover them here so we may better understand the way the CPU percives RAM. The part of the chapter that discusses segment registers also addresses a few memory organization aspects, such as segmentation and paging.
• Control registers: This is a tiny group of registers of registers of high importance, as they control the behavior of the processor as well as enable or disable certain features.
• Debug registers: Although registers of this group are mainly used by debuggers, they add some interesting abilities to our code, for example the ability to set hardware breakpoints when tracing execution of a program.
• EFlags register: This is also known as the status register on some platforms. This one provides us with the information regarding the result of the latest arithmetic logic unit (ALU) operation performed, as well as some settings of the CPU itself.