How to install Randal Hyde’s xi386 HLA compiler on 64bit Linux.

This guide boils down to 3 basic steps: Get the source code, put the compiled bianries in your system path somehow, and add some aliases in your shell’s rc so you can easily run it with specific settings from your terminal.

0. Get the code
   • Go to the author’s homepage.
   • Download the taballs titled HLA for Linux and HLA Stdlib Source Code for Linux. You can save these wherever, like ~/Desktop.
   • Open a terminal and navigate to the download location, for me this was ~/Desktop, but it might be somewhere like ~/Downloads for you.

cd ~/path/to/download.tar.gz

1. Now you need to extract the tarballs. you can run the commands:

tar -xvf linux.hla.tar.gz
tar -xvf linux.hlalib.tar.gz

• This will extract both archives’ contents into a folder in the current directory: ./usr/hla/* (ie. ~/Desktop/usr/hla).
• Optional: Here, the HLA author recommends you set permissions on the HLA files to ‘All users can read and execute’. Unless you have multiple users on your linux install this step isn’t really necessary, but for completion’s sake you can use this command to accomplish this:

chmod -R 555 ./usr/hla

• NOTE: Make sure to type the command correctly!! otherwise you might end up changing permissions on files that don’t need to be changed; don’t miss the period in the path! It means ‘current directory’.

2. The next step is to get the hla binaries into your system path so you can simply type ‘hla file’ to compile your programs:

• Type the following command to copy the HLA compiler source code and binaries to your Unix System Resources (/usr/) directory:

cp -r ./usr/hla /usr/

• Optional, but recommended: The author recommends this step if you prefer a ‘Unix-like’ environment; it copies the executables to /usr/local/bin:

• Change directory to where we copied the binaries and source files in previous step:

cd /usr/hla

• We’ll now place copies of the compiled hla executables to the folder that normally stores user-compiled programs. This folder should be in your system path, so your shell should find it with little effort.

cp hla delete hlacmp hlaparse /usr/local/bin

3. If you didn’t perform the previous step, or if you’re on an atypical setup, or if you simply want the HLA compiler in a specific directory, then you can add the location of the executables to your system path. The process for this is out of the scope of this how-to, I’ll assume if you go this route then you know what you’re doing and can figure out how to append a specific directory to your system path.

4. Now you should set some environment variables:
   • open the run commands file in your home directory. For most people, this is ~/.bashrc. You can figure out which shell you’re using with the command echo $SHELL if you have a different shell, you can consult it’s documentation to figure out how to export environment variables if the following doesn’t work.
   • at the bottom of .bashrc (or whatever shell you’re configuring) add the lines:

hlalib=/usr/hla/hlalib
export hlalib
hlainc=/usr/hla/include
export hlainc

• Optional: add a temp directory, but make sure /tmp exists:

hlatemp=/tmp
export hlainc

• Save the file and ‘activate’ it with:

source .bashrc

• tip: if that doesn’t work, try restarting your session (log out or restart the machine).

5. HLA should be properly installed now, type the following into your terminal to confirm: hla -?
   • if a help message gets printed out, congrats! you did it! However, if you’re on a 64 bit OS there is still one more important step before you can compile programs.

6. The author doesn’t mention this, but it’s crucial for most people running on modern PCs. The HLA compiler library source code was compiled for 32bit architecture, if you are on a 32bit machine you shouldn’t need this step. If you’re on a 64bit machine (most computers nowadays) running a raw hla file.hla command will probably print out a bunch of linker errors. They look somehting like ld: system architecture input source file.o is for i386.
   • Explanation from th bottom of this page (optional reading):

The thing is; on x64 systems ld tries to produce x64 output at its default. Since at its current state hla outputs x86 compatible code and provides x86 libraries, ld simply prints those irritating lines.
The solution resides in a very cleverly figured switch of hla; the “-l”. What this switch does is simply to pass its argument to ld. But, if you try to pass it in a “hla -l -melf_i386 -v helloWorld” format hla will prompt an error since it doesn’t have a “-m” switch. As it is stated in hla’s help; this switch’s argument needs to be in the “-lxxxxx” format where “xxxxx” corresponds to the linker parameter. Similarly a command as “hla -lm elf_i386 -v helloWorld” is false, because the “elf_i386” part of it is recognized as an input file by the hla. Again, “hla -l-melf_i386 -v helloWorld” also doesn’t function as intended because this time ld will be angry since it doesn’t recognize the “option ‘–melf_i386’”. The reason to it is that the argument of this switch is passed to ld with a minus sign prefixed to it.
And about this “-melf_i386”. “-m” switch of ld configures the “emulation mode”. To be honest I don’t know what this means accurately but it does work!
The correct command is;
<user>@<user>-GA-770TA-UD3:/usr/local/samples/hla$ hla -lmelf_i386 -v helloWorld

• In short, you want to run hla with the -lmelf_i386 flag, you can make life easier on yourself by adding this as an alias to your shell config file (.bashrc):
• On some systems you may ned to use the -a32 flag to “let the compiler know you’re compiling for 32bit” -instructor.

alias hla='hla -lmelf_i386 -a32'

• save and run source .bashrc
• You might need to restart your session after editing this file. source can give mixed results (it didn’t work for me).

7. Verify the environment has been set up by compiling a simple hello world program:

cd ~/Desktop
mkdir hw
cd hw
touch hw.hla

• Place this code in the file:

program HelloWorld;
#include( "stdlib.hhf" )

begin HelloWorld;

    stdout.put( "Hello, World of Assembly Language", nl );

end HelloWorld;

• run: hla -v hw
• this should print out a success message that looks like this:

HLA (High Level Assembler)
Use '-license' to see licensing information.
Version 2.16 build 4409 (prototype)
ELF output
OBJ output using HLA Back Engine
-test active

HLA Lib Path:     /usr/hla/hlalib/hlalib.a
HLA include path: /usr/hla/include
HLA temp path:    /tmp
Files:
1: hw.hla

Compiling 'hw.hla' to 'hw.o'
using command line:
[hlaparse -LINUX -level=high  -v -test "hw.hla"]

----------------------
HLA (High Level Assembler) Parser
use '-license' to view license information
Version 2.16 build 4409 (prototype)
-test active
File: hw.hla
Output Path: ""
hlainc Path: "/usr/hla/include"
hlaauxinc Path: ""
Compiler generating code for Linux OS
Back-end assembler: HLABE
Language Level: high

Assembling "hw.hla" to "hw.o"
HLAPARSE assembly complete, 48347 lines,   0.057 seconds,  846708 lines/second
------------
HLA Back Engine Object code formatter

HLABE compiling 'hw.hla' to 'hw.o'
Optimization passes: 3+2
----------------------
Linking via [ld   -melf_i386   -o "hw"   "hw.o" "/usr/hla/hlalib/hlalib.a"]

========================================
     HLA Compilation Complete
========================================

8. If this runs with no problems, then congrats!! you’ve successfully installed the High Level Assembly compiler for use on a 64bit linux system. Leave questions in the comments and I’ll do my best to help out. I’m considering adding a section for windows installs too if needed.

• This guide is subject to edits and fixes! Feel free to suggest any in the comments below. cheers!

• For more information you can read the Author’s installation guide here.