The Process Stack

Notes

• This is from Chapter 9 of Jorgensen.
• Draw the memory of a process on the board.
• A stack is the best solution we have found so far for function management.
  • When we call a function, we place local variables on the stack.
  • Along with other record keeping information
  • When we return from a function, we remove this from the stack.
• The stack grows down from higher memory to lower memory
  • So the data at the bottom of the stack have a higher address then the data at the top of the stack.
  • This is upside down, but live with it.
• Intel assembly supports the stack with
  • push operand
  • pop operand
    • Register or memory or address only, no immediate.
    • Quadwords (8 bytes) only
    • So push will move the top of the stack by -8 bytes
    • And pop will move the top of the stack by 8 bytes.
• The rsp and rbp registers are associated with the stack.
  • rsp holds the address of the top of the stack.
  • push operand

    •  rsp-= 8
       memory[rsp] = operand
  • We need to be exceedingly careful when using the stack.
    • The linux calling convention expects the stack to be 16 byte aligned.
    • Ie rsp % 16 = 0
    • So we need to push a multiple of two if we are going to call a function
• Let's use the stack to reverse a string. (reverse.asm).