Anatomy Of An Assembly Program

Section

Sections define how the memory layout at runtime would be prepared. Common sections include:

1. .text, for code or instructions.
2. .data, for initialized data.
3. .bss, for uninitialized data, where bss stands for “Block Started by Symbol”.
   • It refers to a label (symbol) that marks the start of a block of uninitialized data in memory.
   • It helps in reducing the size of the object files by leaving a note for the system to allocate x bytes at runtime for this block and zero-initialize them. As allocating zeros at compile time makes no sense.
4. .rodata, for read-only data.

Registers (As Operands)

If you’ve attempted assembly before, you might have seen old lectures using ax, bx, cx, dx like registers, or eax, ebx, ecx, edx and the most recent ones might be using rax, rsi, rdi etc….

We know that x86 architecture emerged from 8086 processor, which was a 16-bit processor. The two lettered registers we see belongs to the 16-bit architecture.

• These 16-bit registers also have smaller ones. They are sized 8-bits each.
• They are called high and low. For example, ax has al and ah.

Intel extended them to 32-bit. This increased the register width and all the registers from 16-bit architecture got prefixed by an e. So, ax become eax, bx become ebx and so on. e stands for extended.

AMD extended it further to 64-bit. The register width increased again and we get new registers prefixed with r, while retaining the existing ones. So, eax become rax, ebx become rbx and so on. Along with this, we have got 8 new general purpose registers from r8-r15.

The newer systems are also backward compatible. This means that x86_32 still supports x86 registers, x86_64 still supports x86 and x86_32.

• When we use rax, we are using the complete 64-bit register.
• When we use eax, we are using the lower 32-bits of the rax register.
• When we use ax, we are using the lower 16-bits of the rax register.
• When we use ah, we are using the 8-bits after al, 8-15.
• When we use al, we are using the lowest 8-bits of the rax register, 0-7.
• a visual diagram

A complete list of general purpose registers, link.

Data Addressing Modes

1. Immediate Mode.
   • The simplest method.
   • Here, the data to access is embedded in the instruction itself.
   • Example: mov eax, 5 ; Move the value 5 into EAX register
2. Register Addressing Mode.
   • The instruction contains a register to access, rather than a memory location.
3. Direct Addressing Mode.
   • The instruction contains the reference to the memory address to access.
   • Example: mov eax, [some_address] ; Move data from memory at some_address into EAX
4. Indexed Addressing Mode.
   • The instruction contains a memory address to access, and also specifies an index register to offset that address.
   • Example: mov eax, [ebx + 4] ; Move data from the address in EBX + 4 into EAX
5. Indirect Addressing Mode.
   • The instruction contains a register that contains a pointer to where the data should be accessed.
   • Example: mov eax, [ebx] ; Move data from the address in EBX into EAX
6. Base Pointer Addressing Mode.
   • This is similar to indirect addressing, but you also include a number called the offset to add to the register’s value before using it for lookup.

Always leave an empty line at end of the program. This gracefully marks the end of assembly code. Otherwise, you’ll get a warning by the assembler.

System Calls

A system call is the controlled gateway between a user-space program and the kernel. It lets your code request services that require higher privileges — like writing to the screen, reading a file, or exiting the program.

User Mode vs Kernel Mode

The CPU operates in two modes:

• User Mode: Restricted environment in which our code runs.
• Kernel Mode: Full-access mode where the operating system runs.

Our program cannot perform privileged operations directly. Instead, it uses syscalls to request the kernel to perform them on its behalf.

Linux supports hundreds of syscalls. Here are a few common ones:

Variables v/s Labels

A variable is a container to store a value. A label is a named memory location. Both are different.

A label can point to a group of instructions, a constant value, a procedure, anything. But a variable only stores some value. It can store the result of a computation, but not the instruction itself.

In simple terms, every variable is a label, but every label need not to be a variable.