601.229 (S20): Assembly language exercise

Task

Your task in this exercise is to write an x86-64 assembly language program which reads 10 integer values from the user, stores them in an array, finds the maximum value, and then prints the maximum value.

Note that it’s not sufficient to simply keep track of the maximum value as values are read: the program should first store the 10 input values in an array, then find the maximum from the values in the array.

Here is an example session showing assembling and running the program (user input in bold):

$ make arrayMax
gcc -c -g -no-pie -o arrayMax.o arrayMax.S
gcc -no-pie -o arrayMax arrayMax.o
$ ./arrayMax
Enter 10 integer values: 3 2 61 35 74 73 70 7 94 53
Max is 94

If you finish this task and are looking for a more challenging task, you can try the second assembly language exercise.

Solution

Here is a solution: asmExerciseSoln.zip

Getting started

Download the following zipfile and unzip it: asmExercise.zip

Make your changes to arrayMax.S. You can assemble the program using the command make arrayMax. Run it using the command ./arrayMax.

An example assembly language program hello.S is provided. You can assemble it using the command make hello and run it using the command ./hello.

Tips and suggestions

Allocating storage. The easiest way to allocate storage for the array is to make it a global variable in the .bss segment. For example:

	.section .bss

	.align 8
dataValues: .space (10 * 8)

would reserve space for 10 8-byte (64-bit) values. Storage allocated in the .bss segment is guaranteed to be filled with zeroes.

If you want a challenge, allocate the arrays on the stack. The frame pointer register (%rbp) can help you keep track of stack-allocated storage: see Lecture 8.

Use callee-saved registers for variables. You can use callee-saved registers as variables in your computation. Callee-saved registers have the significant advantage (over caller-saved registers) of being preserved accross procedure calls. Make sure that you push their original values onto the stack before modifying them, and restore their original values from the stack when they are no longer needed. For example, let’s say that your entire computation is implemented in main, and you intend to use %r12 and %r13 as variables. You could put the following code at the beginning of main:

pushq %r12
pushq %r13

Then, put the following code near the end of main (just before the ret instruction at the very end):

popq %r13
popq %r12

Note that values must be popped from the the stack in the reverse of the order in which they were pushed. (It’s a stack!)

Don’t forget that the stack pointer (%rsp) must be an exact multiple of 16 at the point of any call instruction. Each push of a 64 bit value will decrease %rsp by 8. Depending on how many pushq/popq instructions you have, you may need to adjust the stack pointer using subq $8, %rsp and addq $8, %rsp to ensure correct stack alignment.

Accessing array elements. One challenge in this exercise is accessing array elements. Assuming you use 64-bit integers, each array element will occupy 8 bytes of storage. The indexed/scaled addressing mode is very convenient for directly accessing an array element based on its displacement from the array’s base address. Let’s say that %r12 contains the base address of the array (i.e., it points to the first element of the array), and that %r13 contains the index of an element. You can store the address of the chosen element in %rsi with the instruction

leaq (%r12,%r13,8), %rsi

You can load the value of the chosen element into %rsi with the instruction

movq (%r12,%r13,8), %rsi

Note that when specifying the address of a global variable or array, prefix it with $ (because you’re referring to the constant address of the variable, not referring to the data stored in the variable.) For example, to load the base address of the array called dataValues into the %r12 register, you would use the following instruction:

movq $dataValues, %r12

Use gdb. Use gdb to trace through the execution of your program. The Resources page has links to some useful information about using gdb to debug assembly language. Lecture 8 also has some useful gdb tips on the last two pages.