navm 2.0.1

A barebones VM for use in scripting languages


To use this package, run the following command in your project's root directory:

Manual usage
Put the following dependency into your project's dependences section:


This package provides sub packages which can be used individually:

navm:stackvm - A basic stack based VM using NaVM

A barebones VM, intended to be used for scripting applications.

Created primarily for use in qscript.

Getting Started

NaVM comes with a demo stack based VM stackvm or svm. These instructions will build that.

See the documents:

  • docs/binarybytecode.md - Binary ByteCode format
  • syntax.md - Syntax description
  • tests/* - Some test bytecodes, to be run using the NaVM demo

Prerequisites

You need to have these present on your machine to build NaVM:

  1. dub
  2. dlang compiler (ldc usually generates best performing code)
  3. git

Building

git clone https://github.com/Nafees10/navm.git
cd navm
dub build :stackvm -b=release -c=demo # --compiler=ldc

This will compile the NaVM demo binary named svm.

You can now run NaVM bytecode using:

./svm tests/default [numberOfTimesToRun]

Replace tests/default with a path to a bytecode file.

Creating a VM

A VM using NaVM is created using simple functions. A very basic VM with only 2 instructions, printSum int int, and print int, can be created as:

import navm.navm;
void printSum(ptrdiff_t a, ptrdiff_t b){
	writeln(a + b);
}
void print(ptrdiff_t a){
	writeln(a);
}
// load bytecode. bytecodeSource is the bytecode in a string
ByteCode code;
try {
	code = parseByteCode!(printSum, print)(bytecodeSource);
} catch (Exception e){
	// probably an error in the bytecode
}
// execute the code
execute!(printSum, print)(code);

Starting at a label

import std.algorithm : countUntil, canFind;
ByteCode code;
// locate the index of the label
ptrdiff_t index = code.labelNames.countUntil("labelName");
if (index == -1){
	throw new Exception("labelName does not exist");
}
execute!(..)(code, index);

Jumps

An instruction can cause execution to jump to another place in the bytecode, by receiving references to the instruction counter _ic.

The instruction counter stores index of the next instruction to execute.

The most common use case for jumps would be jumping to some label. A jump to a label could be implemented as:

void jump(ref size_t _ic, size_t label){
	_ic = label;
}

Example usage:

start:
	printS "Hello\n"
loop:
	printS "Looping forever\n"
	jump @loop # @loop is replaced with the index of instruction after loop

Shared state

Since the instruction functions have to be functions and not delegates, there was one way to have shared data between instructions: global variables.

However that would get tricky with multiple execute calls, so an alternative is to use the _state parameter. An overload of execute can be passed a ref of a struct instance, which it will pass on to any instruction that needs it:

struct Stack {
	ptrdiff_t[512] arr;
	size_t top;
}
void push(ref Stack _state, size_t value){
	_state.arr[top ++] = value;
}
void pop(ref Stack _state){
	_state.top --;
}
import std.meta : AliasSeq;
alias InstrucionSet = AliasSeq!(push, pop);
// load
ByteCode code = parseByteCode!InstrucionSet(bytecodeSource);
Stack stack;
// execute
execute!(Stack, InstrucionSet)(code, stack);

License

NaVM is licensed under the MIT License - see LICENSE for details

Authors:
  • Nafees Hassan
Sub packages:
navm:stackvm
Dependencies:
utils
Versions:
2.0.2 2024-Apr-10
2.0.1 2024-Jan-28
2.0.0 2024-Jan-28
2.0.0-beta-1 2024-Jan-21
1.4.1 2023-Apr-14
Show all 31 versions
Download Stats:
  • 0 downloads today

  • 0 downloads this week

  • 0 downloads this month

  • 124 downloads total

Score:
1.5
Short URL:
navm.dub.pm