Defining and Calling Procedures

In assembly, a procedure (often called a function or subroutine) is a block of code designed to perform a specific task. They help organize your program and avoid repeating code.

Think of them like functions in high-level languages like C++ or Python. They allow you to break down complex problems into smaller, manageable parts, making your code modular and easier to read.

To execute a procedure, we use the CALL instruction. When CALL is executed, two important things happen:

• The address of the instruction immediately after CALL is pushed onto the stack. This is known as the return address.
• The CPU then jumps to the starting address of the procedure you specified.

This mechanism ensures that the program knows exactly where to resume execution once the procedure has completed its work.

Once a procedure has finished its assigned task, it needs to return control to the code that called it. This is achieved using the RET instruction.

When RET is executed:

• The CPU pops the return address from the top of the stack.
• The CPU then jumps to this popped address, effectively resuming execution at the instruction immediately following the original CALL.

Together, CALL and RET form the fundamental pair for managing program flow between different procedures.

Let's look at a simple assembly program that demonstrates a basic procedure call and return. We'll define a procedure named print_hello and call it from our program's entry point, _start.

This example uses Linux system calls for output and program exit.

When you run the previous code, it will print "Hello from proc!" to your console. Here's a step-by-step breakdown of what happened:

• The _start routine executed call print_hello.
• The address of the instruction mov eax, 1 (which is right after call print_hello) was pushed onto the stack.
• The CPU jumped to the print_hello procedure.
• print_hello executed its instructions to print the message.
• ret popped the saved return address from the stack and jumped back to the _start routine.
• _start then executed the system call to exit the program.

When a procedure is called, it often needs its own private workspace on the stack to manage its data. This dedicated region on the stack is called a stack frame.

A stack frame typically holds several key pieces of information for a procedure:

• The return address (pushed by the CALL instruction).
• Saved register values (e.g., the caller's base pointer).
• Local variables specific to that procedure.
• Arguments passed to the procedure (we'll cover this in the next lesson!).

The base pointer register (`EBP` in 32-bit, `RBP` in 64-bit) is a crucial tool for managing stack frames. It provides a stable reference point within the current stack frame, making it easy to access local variables and arguments.

A common setup sequence at the very beginning of a procedure is:

• push ebp: This saves the caller's current `EBP` value onto the stack, so it can be restored later.
• mov ebp, esp: This sets `EBP` to the current value of the stack pointer (`ESP`), establishing the base of the new stack frame.

After setting up `EBP`, a procedure can allocate space for its own local variables on the stack. This is typically done by simply decrementing the stack pointer (`ESP`).

Here, `N` represents the total number of bytes required for all local variables. For example, sub esp, 4 allocates enough space for one 32-bit integer.

These local variables can then be accessed efficiently relative to `EBP` (e.g., [ebp-4], [ebp-8], etc.).

Before a procedure returns, its stack frame must be properly dismantled to restore the stack to its original state. This involves deallocating local variables and restoring the caller's base pointer.

The LEAVE instruction is a convenient way to perform these two actions in one step:

• mov esp, ebp: This deallocates any local variables by moving `ESP` back to where `EBP` points (the base of the frame).
• pop ebp: This restores the caller's `EBP` value, which was saved at the beginning of the procedure.

After LEAVE, the stack is correctly positioned for the RET instruction to pop the return address.

This example demonstrates a complete procedure that sets up a proper stack frame, allocates space for a hypothetical local variable, and then correctly tears down the frame before returning.

Notice how `push ebp`, `mov ebp, esp`, `sub esp, 4`, `leave`, and `ret` work together.

Consider the following x86 assembly snippet:

  call my_function
  mov eax, 1
my_function:
  ret

What specific address is pushed onto the stack by the call my_function instruction?

We've covered the essential concepts of defining and calling procedures in x86 assembly. Here are the key takeaways from this lesson:

• The CALL instruction pushes the return address onto the stack and transfers control to a procedure.
• The RET instruction pops the return address from the stack and transfers control back to the caller.
• Stack frames, managed primarily with the EBP/RBP register, provide a dedicated and organized workspace on the stack for a procedure's local variables and saved registers.
• A typical stack frame setup involves push ebp, mov ebp, esp, and allocating local variables with sub esp, N.
• Tearing down the stack frame is done using the LEAVE instruction (or manually with mov esp, ebp and pop ebp) before RET.

Next, we'll build on this by learning how to pass arguments to procedures and retrieve return values.