NASM Coding: Basics

The Linux kernel (= the main component of the Linux OS) keeps a list of all the files currently open by every running program. This list is called the file table, and its indexes are file descriptors (

fd

s), each uniquely representing a file.

• Most file system calls accept an

  fd

  as an argument (since they need to know which file(s) we want to change.)
• Each file table entry contains a pointer to the file's metadata (inode), including the location of the file's content on disk.
• Legit

  fd

  s are non-negative integers (0, 1, 2, ..., n), so

  fd == -1

  indicates an error. Each program has an upper limit of how many files can be opened at the same time (by default: n = 1000), but we can change this limit if wanted.
• Conventionally,

  fd == 0

  represents standard input (=

  stdin

  = keyboard),

  fd == 1

  is standard output (=

  stdout

  = display), and

  fd == 2

  represents standard error output (=

  stderr

  = also printed to display). All 3 are open by default.

• fd

  s can also represent special files (devices, disks, sockets, etc.) since Linux treats everything as a file.

In the scope of our study of NASM coding, the only files that we will interact with are standard input (=

stdin

= keyboard) and standard output (=

stdout

= screen), even though NASM has the tools for reading from and writing to other files (e.g., text files stored on the computer.)