# A Gentle Introduction to Assembly Language Programming This textbook provides a gentle introduction to assembly language programming. What makes this introduction "gentle" is that it assumes the reader is already comfortable with C or C++ coding. We use this assumed knowledge to **bridge** backward towards the low level ISA (Instruction Set Architecture). We drive home a very sharp point:
Assembly language is nothing to be scared of!


Quick Link to Table of Contents


## Front Matter ### For Whom Is This Book Intended? As mentioned, if you are already familiar with C (or languages descended from C such as C++), this book begins with what you already know. Later chapters dive more deeply into the corners and recesses of the ARM V8 ISA and are suitable for those wishing to master the rich instruction set of the 64 bit ARM processors. ### Can This Book Be Used In Courses Covering Assembly Language? Yes, absolutely. ### Calling Convention Used In This Book Assembly language programming is quite closely dependent upon the underlying hardware architecture. The host operating environment plays an outsized role in determining how assembly language programs are constructed. A "calling convention" refers to how functions are called and how parameters are passed. In this book we will use the ARM LINUX conventions. This means: * You *may* need to run a ARM Linux VM on the Macintosh - even on ARM-based Macs. Why? Apple uses a different calling convention. Keep reading before you get upset. The convention used in this book should work on all ARM Linux machines while the Apple calling convention is specific to Apple Silicon-based machines. This necessity for a VM even when running on an Apple Silicon machine did not sit well with some, who made this criticism known. We assessed this to be a valid and constructive criticism and have responded. We now have a chapter devoted to bringing Linux and Apple code together to the degree possible. * The macros are a work in progress. [This link](./macros/) will lead to a current copy of them as well as documentation. Macros that make programming a bit easier are also included. * [This chapter](./more/apple_silicon/) provides some additional information about Apple Silicon assembly language programming. * You will need to run WSL (Windows Subsystem for Linux) on ARM-based Windows machines. These do exist! * You will need to run an ARM Linux VM on x86-based Windows machines. This is true even if you are on an ARM-based Windows machine as there are so many differences between a Unix-like environment and Windows. You'll notice that we make use of the C-runtime directly rather than make OS system calls. So, for instance, if we want to call `write()`, we call `write` from the assembly language. This version of the system call `write` is a wrapper function built into the C-runtime (CRT) which handles the lower level details of performing a system call. See the [here](./more/system_calls/README.md) on what actually happens inside these wrapper functions. The benefit of using the CRT wrappers is that there are details, explained in the chapter, that differ from system to system and architecture to architecture even for making the same system call. The CRT hides these differences. ### A Lot of Names As commendable as the ARM designs are, ARM's naming conventions for their Intellectual Properties are horrid. In this book, AARCH64 and ARM V8 are taken to be synonyms for the 64 bit ARM Instruction Set Architecture (ISA). It is very difficult to find documentation at the ARM site because they have *so many versions*, so many names for the same thing and so much documentation in general. It really can be maddening. Within the text we will provide germane links as appropriate. [Here]() is a link to "a" main instruction set page. ### What you need to work with assembly language on Linux Getting the tools for assembly language development is quite straight forward - perhaps you already have them. Using `apt` from the Linux terminal, say: ```text sudo apt update sudo apt install build-essential gdb ``` On the Macintosh type: `xcode-select --install` into a terminal and follow directions. Note that `gdb` is replaced by `lldb` with just enough differences to make you cry. Then you'll need your favorite editor. We currently use `vi` for quick edits and Visual Studio Code for any heavy lifting. ### How to build an assembly language We use `gcc`, the C "compiler". `g++` could also be used. On the Mac, `clang` can also be used. What sense does that make... using the "compiler" to "compile" assembly language? Well, to answer that one must understand that the word "compiler" refers to only one step in a build sequence. What we talk about as being the "compiler" is actually an umbrella that includes: * A preprocessor that acts on any `#` preprocessor command like `#include`. These commands are not part of C or C++. Rather they are commands to the preprocessor. Note that `gcc` will invoke the C preprocessor only if your assembly language file ends in `.S` - capital S. It may not be invoked if your file ends in a lower case s or any other file extension. * The *actual* compiler, whose job it is turn high level languages such as C and C++ into assembly language. * The assembler, which turns assembly language into machine code which is not quite ready for execution. * And finally, the linker, which combines potentially many intermediate machine code files (called object files), potentially many library files (statically linked .dlls on Windows and .a files on Linux). The linker is the last step in this chain. [Here](https://youtu.be/Iv3psS4n9j8) is a video explaining this process. We use gcc and g++ directly because, being umbrellas, they automate the above steps with other benefits such as automatically linking in the C runtime. Suppose you've implemented `main()` in a C file (main.c) and want to make use of an assembly language file you have written (asm.S). It can be done in several ways. #### All at once ```text gcc main.c asm.S ``` That's all you need for a minimal build. The resulting program will be written to `a.out`. All the intermediate files generated will be removed. #### Modularly ```text gcc -c main.c gcc -c asm.S gcc main.o asm.o ``` Used in this way, `.o` files are left on disk. Using the previous method, the `.o` files are removed without you seeing them. ### If there are no C or C++ modules used Suppose `main()` is implemented in assembly language and `main.s` is self-contained, then simply: ```text gcc main.S ``` Often, you will want to enable the debugger `gdb` or `lldb`. Do this: ```text gcc -g main.S ``` #### The C Pre-Processor If you want `gcc` to run your code through the C pre-processor (for handing `#include` for example), name your assembly language source code files with a capital S. So, on Linux: `gcc main.s` Will not go through the C pre-processor but `gcc main.S` will. #### Programs called by the "Compiler" To drive home the point that the "compiler" is an umbrella, using gcc to "compile" a program causes the following to be called on Ubuntu running on ARM: ```text /usr/bin/cpp /usr/lib/gcc/aarch64-linux-gnu/11/cc1 /usr/bin/as /usr/lib/gcc/aarch64-linux-gnu/11/collect2 which is... /usr/bin/ld ``` `cpp` is the C preprocessor - it is a general tool can is used by other languages as well (C++, for example). `cc1` is the actual compiler. `as` is the assembler. `ld` is the linker. You can see why we default to using the umbrella command in this book. ## Table of Contents ### Section 1 - Bridging from C / C++ to Assembly Language We start by providing what we're calling "bridging" from C and C++ to assembly language. We use the knowledge you already have to learn new knowledge - how cool is that! | Chapter | Markdown | PDF | | ------- | -------- | --- | | 1 | [Hello World](./section_1/hello_world/README.md) | [Link](./section_1/hello_world/README.pdf) | | 2 | [If Statements](./section_1/if/README.md) | [Link](./section_1/if/README.pdf) | | 3 | Loops | | | 3a | [While Loops](./section_1/while/README.md) | [Link](./section_1/while/README.pdf) | | 3b | [For Loops](./section_1/for/README.md) | [Link](./section_1/for/README.pdf) | | 3c | [Implementing Continue](./section_1/for/README.md#implementing-a-continue) | [Link](./section_1/for/README.pdf) | | 3d | [Implementing Break](./section_1/for/README.md#implementing-a-break) | [Link](./section_1/for/README.pdf) | | 4 | Interludes | | | 4a | [Registers](./section_1/regs/README.md) | [Link](./section_1/regs/README.pdf) | | 4b | [Load and Store](./section_1/regs/ldr.md) | [Link](./section_1/regs/ldr.pdf) | | 4c | [More About `ldr`](./section_1/regs/ldr2.md) | [Link](./section_1/regs/ldr2.pdf) | | 4d | [Register Sizes](./section_1/regs/widths.md) | [Link](./section_1/regs/widths.pdf) | | 4e | [Hexadecimal](./section_1/hex.md) | [Link](./section_1/hex.pdf) | | 5 | [`switch`](./section_1/jump_tables/README.md) | [Link](./section_1/jump_tables/README.pdf) | | 6 | Functions | | | 6a | [Calling and Returning](./section_1/funcs/README.md) | [Link](./section_1/funcs/README.pdf) | | 6b | [Passing Parameters](./section_1/funcs/README2.md) | [Link](./section_1/funcs/README2.pdf) | | 6c | [Example of calling some common C runtime functions](./section_1/funcs/README3.md) | [Link](./section_1/funcs/README3.pdf) | | 7 | [FizzBuzz - a Complete Program](./section_1/fizzbuzz/README.md) | [Link](./section_1/fizzbuzz/README.pdf) | | 8 | Structs | | | 8a | [Alignment](./section_1/structs/alignment.md) | [Link](./section_1/structs/alignment.pdf) | | 8b | [Defining](./section_1/structs/defining.md) | [Link](./section_1/structs/defining.pdf) | | 8c | [Using](./section_1/structs/using.md) | [Link](./section_1/structs/using.pdf) | | 9 | [`const`](./section_1/const/README.md) | [Link](./section_1/const/README.pdf) | | 10 | [Casting](./section_1/casting/README.md) | [Link](./section_1/casting/README.pdf) | ### Section 2 - Floating Point Floating point operations use their own instructions and their own set of registers. Therefore, floating point operations are covered in their own section: | Chapter | Markdown | PDF | | ------- | -------- | --- | | 0 | [Chapter Overview](./section_2/float/README.md) | [Link](./section_2/float/README.pdf) | | 1 | [What Are Floating Point Numbers?](./section_2/float/what.md) | [Link](./section_2/float/what.pdf) | | 2 | [Registers](./section_2/float/working.md) | [Link](./section_2/float/working.pdf) | | 3 | [Truncation and Rounding](./section_2/float/rounding.md) | [Link](./section_2/float/rounding.pdf) | | 4 | [Literals](./section_2/float/literals.md) | [Link](./section_2/float/literals.pdf) | | 5 | [`fmov`](./section_2/float/fmov.md) | [Link](./section_2/float/fmov.pdf) | | 6 | [Half Precision Floats](./section_2/float/half.md) | [Link](./section_2/float/half.pdf) | | 7 | [NEON SIMD Not Yet Written](./not_written_yet.md) | [Link](./not_written_yet.pdf) | ### Section 3 - Bit Manipulation What would a book about assembly language be without bit bashing? | Chapter | Markdown | PDF | | ------- | -------- | --- | | 1 | Bit Fields | | | 1a | [Without Bit Fields](./section_3/bitfields/README.md) | [Link](./section_3/bitfields/README.pdf) | | 1b | [With Bit Fields](./section_3/bitfields/with.md) | [Link](./section_3/bitfields/with.pdf) | | 1c | [Review of Newly Described Instructions](./section_3/bitfields/review.md) | [Link](./section_3/bitfields/review.pdf) | | 2 | [Endianness](./section_3/endian/README.md) | [Link](./section_3/endian/README.pdf) | ### Section 4 - More Stuff In this section, we present miscellaneous material. | Chapter | Markdown | PDF | | ------- | -------- | --- | | 1 | [Apple Silicon](./more/apple_silicon/README.md) | [Link](./more/apple_silicon/README.pdf) | | 2 | [Apple / Linux Convergence](./macros) | [Link](./macros/README.pdf) | | 3 | [Variadic Functions](./more/varargs/README.md) | [Link](./more/varargs/README.pdf) | | 4 | [Under the hood: System Calls](./more/system_calls/README.md) | [Link](./more/system_calls/README.pdf) | | 5 | [Determining string literal lengths for C functions](./more/strlen_for_c/README.md) | [Link](./more/strlen_for_c/README.pdf) | ## Macro Suite As indicated immediately above, the macro suite [can be found here](./macros/). ## Projects [Here](./projects/README.md) are some project specifications to offer a challenge to your growing mastery. Here are very brief descriptions presented in alphabetical order. Perhaps before you tackle these, check out the fully described [FIZZBUZZ](./section_1/fizzbuzz/README.md) program first. * Then try [this](./projects/first_project/README.md) as your very first project. With some blank lines and comments it weighs in at 35 lines. * The [DIRENT](./projects/DIRENT/README.md) project demonstrates how a complex `struct` can be used in assembly language. * The [PI](./projects/PI/README.md) project demonstrates floating point instructions. The program will "throw darts at a target," calculating an approximation of PI by tracking how many darts "hit the target" versus the total number of darts "thrown". * The [SINE](./projects/SINE/README.md) project stresses floating point math and functions. * The [SNOW](./projects/snow/README.md) project uses 1970's era tech to animate a simple particle system. This project demonstrates a reasonable design process of breaking down complex problems into simpler parts. * The [WALKIES](./projects/walkies/README.md) presents a cute little animation demonstrating looping with some pointer dereferencing. ## About The Author Perry Kivolowitz's career in the Computer Sciences spans just under five decades. He launched more than 5 companies, mostly relating to hardware, image processing and visual effects (for motion pictures and television). Perry received Emmy recognition for his work on the The Gathering, the pilot episode of Babylon 5. Later he received an Emmy Award for Engineering along with his colleagues at [SilhouetteFX, LLC](https://en.wikipedia.org/wiki/SilhouetteFX). SilhouetteFX is used in almost every significant motion picture for rotoscoping, paint, tracking, 2D to 3D reconstruction, compositing and more. In 1996 Perry received an [Academy Award for Scientific and Technical Achievement](https://en.wikipedia.org/wiki/Academy_Award_for_Technical_Achievement) for his invention of Shape Driven Warping and Morphing. This is the technique responsible for many of the famous effects in Forrest Gump, Titanic and Stargate. Twenty twenty three marks Perry's 19th year teaching Computer Science at the college level, ten years at the UW Madison and now 8+ at Carthage College. Assembly language is a passion for Perry having worked in the following ISAs (in chronological order): * Univac 1100 * Digital Equipment Corporation PDP-11 * Digital Equipment Corporation VAX-11 * Motorola 68000 * ARM beginning with AARCH64 This work is dedicated to my wife Sara and sons Ian and Evan. ### Gratuitous Plugs Perry has created a library of about 200 programming projects suitable for CS 1, CS 2, Data Structures, Networking, Operating Systems and Computer Organization classes. If a publisher of CS text books (or other CS related content) would be interested in purchasing the library, please reach out. Also, check out [Get Off My L@wn](https://www.amazon.com/Get-Off-My-Zombie-Novel-ebook/dp/B00DQ26J8G), a Zombie novel for coders. You read that right... elite programmer Doug Handsman retires to his wife Ruth Ann's native northern Wisconsin. And then, well, the apocalypse happens. Bummer. Rated 4.3 out of 5 with more than 70 reviews, it's a fun read and costs next to nothing.