https://github.com/CanadaHonk/porffor

A from-scratch experimental AOT JS engine, written in JS

A from-scratch experimental AOT optimizing JS/TS -> Wasm/C engine/compiler/runtime in JS. Research project, not yet intended for serious use.

Porffor is a very unique JS engine, due many wildly different approaches. It is seriously limited, but what it can do, it does pretty well. Key differences:

• 100% AOT compiled (no JIT)
• No constant runtime/preluded code
• Least Wasm imports possible (only I/O)

Porffor is primarily built from scratch, the only thing that is not is the parser (using Acorn). Binaryen/etc is not used, we make final wasm binaries ourself. You could imagine it as compiling a language which is a sub (some things unsupported) and super (new/custom apis) set of javascript. Not based on any particular spec version.

Expect nothing to work! Only very limited JS is currently supported. See files in bench for examples.

npm install -g porffor@latest. It's that easy (hopefully) :)

porf. Just run it with no script file argument.

porf path/to/script.js

porf wasm path/to/script.js out.wasm. Currently it does not use an import standard like WASI, so it is mostly unusable on its own.

porf native path/to/script.js out(.exe). You can specify the compiler with --compiler=clang|gcc|zig (clang by default), and which optimization level to use with --cO=Ofast|O3|O2|O1|O0 (Ofast by default). Output binaries are also stripped by default.

porf c path/to/script.js (out.c). When not including an output file, it will be printed to stdout instead.

porf profile path/to/script.js

porf debug path/to/script.js

porf debug-wasm path/to/script.js

• --parser=acorn|@babel/parser|meriyah|hermes-parser (default: acorn) to set which parser to use
• --parse-types to enable parsing type annotations/typescript. if -parser is unset, changes default to @babel/parser. does not type check
• --opt-types to perform optimizations using type annotations as compiler hints. does not type check
• --valtype=i32|i64|f64 (default: f64) to set valtype
• -O0 to disable opt
• -O1 (default) to enable basic opt (simplify insts, treeshake wasm imports)
• -O2 to enable advanced opt (inlining). unstable!
• -O3 to enable advanceder opt (precompute const math). unstable!

• Limited async support
• No variables between scopes (except args and globals)
• No eval()/Function() etc (since it is AOT)

Asur is Porffor's own Wasm engine; it is an intentionally simple interpreter written in JS. It is very WIP. See its readme for more details.

Rhemyn is Porffor's own regex engine; it compiles literal regex to Wasm bytecode AOT (remind you of anything?). It is quite basic and WIP. See its readme for more details.

2c is Porffor's own Wasm -> C compiler, using generated Wasm bytecode and internal info to generate specific and efficient/fast C code. Little boilerplate/preluded code or required external files, just for CLI binaries (not like wasm2c very much).

See optimizations for opts implemented/supported.

These include some early (stage 1/0) and/or dead (last commit years ago) proposals but I think they are pretty neat, so.

• Math.clamp Proposal: Math.clamp (stage 0 - last commit april 2023)
• Math Extensions Proposal: Math.scale, Math.radians, Math.degrees, Math.RAD_PER_DEG, Math.DEG_PER_RAD (stage 1 - last commit september 2020)
• Math.signbit Proposal: Math.signbit (stage 1 - last commit february 2020)

• Number literals
• Declaring functions
• Calling functions
• return
• Basic declarations (let/const/var)
• Some basic integer operators (+-/*%)
• Some basic integer bitwise operators (&|)
• Equality operators (==, !=, etc)
• GT/LT operators (>, <, >=, etc)
• Some unary operators (!, +, -)
• Logical operators (&&, ||)
• Declaring multiple variables in one (let a, b = 0)
• Array destructuring (let [a, ...b] = foo)
• Global variables (var/none in top scope)
• Booleans
• if and if ... else
• Anonymous functions
• Setting functions using vars (const foo = function() { ... })
• Arrow functions
• undefined/null
• Update expressions (a++, ++b, c--, etc)
• for loops (for (let i = 0; i < N; i++), etc)
• Basic objects (no prototypes)
• console.log
• while loops
• break and continue
• Named export funcs
• IIFE support
• Assignment operators (+=, -=, >>=, &&=, etc)
• Conditional/ternary operator (cond ? a : b)
• Recursive functions
• Bare returns (return)
• throw (literals only, hack for new Error)
• Basic try { ... } catch { ... } (no error given)
• Calling functions with non-matching arguments (eg f(a, b); f(0); f(1, 2, 3);)
• typeof
• Runtime errors for undeclared variables (ReferenceError), not functions (TypeError)
• Array creation via [] (eg let arr = [ 1, 2, 3 ])
• Array member access via arr[ind] (eg arr[0])
• String literals ('hello world')
• String member (char) access via str[ind] (eg str[0])
• String concat (+) (eg 'a' + 'b')
• Truthy/falsy (eg !'' == true)
• String comparison (eg 'a' == 'a', 'a' != 'b')
• Nullish coalescing operator (??)
• for...of (arrays and strings)
• for...in
• Array member setting (arr[0] = 2, arr[0] += 2, etc)
• Array constructor (Array(5), new Array(1, 2, 3))
• Labelled statements (foo: while (...))
• do...while loops
• Optional parameters ((foo = 'bar') => { ... })
• Rest parameters ((...foo) => { ... })
• this
• Constructors (new Foo)
• Classes (class A {})
• Await (await promise)

• NaN and Infinity
• isNaN() and isFinite()
• Most of Number (MAX_VALUE, MIN_VALUE, MAX_SAFE_INTEGER, MIN_SAFE_INTEGER, POSITIVE_INFINITY, NEGATIVE_INFINITY, EPSILON, NaN, isNaN, isFinite, isInteger, isSafeInteger)
• Most Math funcs (sqrt, abs, floor, sign, round, trunc, clz32, fround, random, exp, log, log2, log10, pow, expm1, log1p, sqrt, cbrt, hypot, sin, cos, tan, sinh, cosh, tanh, asinh, acosh, atanh, asin, acos, atan, atan2)
• Basic globalThis support
• Basic Boolean and Number
• Basic eval for literals
• Math.random() using self-made xorshift128+ PRNG
• Some of performance (now(), timeOrigin)
• Most of Array.prototype (at, push, pop, shift, fill, slice, indexOf, lastIndexOf, includes, with, reverse, toReversed, forEach, filter, map, find, findLast, findIndex, findLastIndex, every, some, reduce, reduceRight, join, toString)
• Most of Array (of, isArray)
• Most of String.prototype (at, charAt, charCodeAt, toUpperCase, toLowerCase, startsWith, endsWith, indexOf, lastIndexOf, includes, padStart, padEnd, substring, substr, slice, trimStart, trimEnd, trim, toString, big, blink, bold, fixed, italics, small, strike, sub, sup, trimLeft, trimRight, trim)
• Some of crypto (randomUUID)
• escape
• btoa
• Most of Number.prototype (toString, toFixed, toExponential)
• parseInt
• Spec-compliant Date
• WIP typed arrays (Uint8Array, Int32Array, etc)
• Synchronous Promise

• Supports i32, i64, and f64 for valtypes
• Intrinsic functions (see below)
• Inlining wasm via asm`...` "macro"

Porffor uses a unique versioning system, here's an example: 0.18.2+2aa3f0589. Let's break it down:

1. 0 - major, always 0 as Porffor is not ready yet
2. 18 - minor, total Test262 pass percentage (floored to nearest int)
3. 2 - micro, build number for that minor (incremented each publish/git push)
4. 2aa3f0589 - commit hash

For the features it supports most of the time, Porffor is blazingly fast compared to most interpreters and common engines running without JIT. For those with JIT, it is usually slower by default, but can catch up with compiler arguments and typed input, even more so when compiling to native binaries.

Mostly for reducing size. I do not really care about compiler perf/time as long as it is reasonable. We do not use/rely on external opt tools (wasm-opt, etc), instead doing optimization inside the compiler itself creating even smaller code sizes than wasm-opt itself can produce as we have more internal information.

• Inlining functions (WIP, limited)
• Inline const math ops
• Tail calls (behind flag --tail-call)

• local.set, local.get -> local.tee
• i32.const 0, i32.eq -> i32.eqz
• i64.extend_i32_s, i32.wrap_i64 -> ``
• f64.convert_i32_u, i32.trunc_sat_f64_s -> ``
• return, end -> end
• Change const, convert to const of converted valtype (eg f64.const, i32.trunc_sat_f64_s -> i32.const`)
• Remove some redundant sets/gets
• Remove unneeded single just used vars
• Remove unneeded blocks (no brs inside)
• Remove unused imports
• Use data segments for initing arrays/strings
• (Likely more not documented yet, todo)

• Type cache/index (no repeated types)
• No main func if empty (and other exports)
• No tags if unused/optimized out

Porffor can run Test262 via some hacks/transforms which remove unsupported features whilst still doing the same asserts (eg simpler error messages using literals only). It currently passes >14% (see latest commit desc for latest and details). Use node test262 to test, it will also show a difference of overall results between the last commit and current results.

• compiler: contains the compiler itself

  • 2c.js: porffor's custom wasm-to-c engine
  • allocators.js: static and dynamic allocators to power various language features
  • assemble.js: assembles wasm ops and metadata into a wasm module/file
  • builtins.js: all manually written built-ins of the engine (spec, custom. vars, funcs)
  • builtins_object.js: all the various built-in objects (think String, globalThis, etc.)
  • builtins_precompiled.js: dynamically generated builtins from the builtins/ folder
  • codegen.js: code (wasm) generation, ast -> wasm. The bulk of the effort
  • cyclone.js: wasm partial constant evaluator (it is fast and dangerous hence "cyclone")
  • decompile.js: basic wasm decompiler for debug info
  • diagram.js: produces Mermaid graphs
  • embedding.js: utils for embedding consts
  • encoding.js: utils for encoding things as bytes as wasm expects
  • expression.js: mapping most operators to an opcode (advanced are as built-ins eg f64_%)
  • havoc.js: wasm rewrite library (it wreaks havoc upon wasm bytecode hence "havoc")
  • index.js: doing all the compiler steps, takes code in, wasm out
  • opt.js: self-made wasm bytecode optimizer
  • parse.js: parser simply wrapping acorn
  • pgo.js: a profile guided optimizer
  • precompile.js: the tool to generate builtins_precompied.js
  • prefs.js: a utility to read command line arguments
  • prototype.js: some builtin prototype functions
  • types.js: definitions for each of the builtin types
  • wasmSpec.js: "enums"/info from wasm spec
  • wrap.js: wrapper for compiler which instantiates and produces nice exports

• runner: contains utils for running JS with the compiler

  • index.js: the main file, you probably want to use this
  • info.js: runs with extra info printed
  • repl.js: basic repl (uses node:repl)

• rhemyn: contains Rhemyn - our regex engine (used by Porffor)

  • compile.js: compiles regex ast into wasm bytecode
  • parse.js: own regex parser

• test: contains many test files for majority of supported features

• test262: test262 runner and utils

Currently, Porffor is seriously limited in features and functionality, however it has some key benefits:

• Safety. As Porffor is written in JS, a memory-safe language*, and compiles JS to Wasm, a fully sandboxed environment*, it is quite safe. (* These rely on the underlying implementations being secure. You could also run Wasm, or even Porffor itself, with an interpreter instead of a JIT for bonus security points too.)
• Compiling JS to native binaries. This is still very early!
• Inline Wasm for when you want to beat the compiler in performance, or just want fine grained functionality.
• Potential for SIMD operations and other lower level concepts.
• More in future probably?

No particular order and no guarentees, just what could happen soon™

• Asur
  • Support memory
  • Support exceptions
• Exceptions
  • Rethrowing inside catch
• Optimizations
  • Rewrite local indexes per func for smallest local header and remove unused idxs
  • Smarter inline selection (snapshots?)
  • Memory alignment
• Runtime
  • WASI target
  • Run precompiled Wasm file if given
• Cool proposals
  • Optional Chaining Assignment
  • Modulus and Additional Integer Math
  • Array Equality
  • Declarations in Conditionals
  • Seeded Pseudo-Random Numbers
  • do expressions
  • String Trim Characters
• Posts
  • Inlining investigation
  • JS -> Native
  • Precompiled TS built-ins
  • Asur
  • escape() optimization
  • PGO
• Self hosted testing?

There is a vscode extension in vscode-ext which tweaks JS syntax highlighting to be nicer with porffor features (eg highlighting wasm inside of inline asm).

Porffor intentionally does not use Wasm proposals which are not commonly implemented yet (eg GC) so it can be used in as many places as possible.

• Multi-value (required)
• Non-trapping float-to-int conversions (required)
• Bulk memory operations (optional, can get away without sometimes)
• Exception handling (optional, only for errors)
• Tail calls (opt-in, off by default)

purple in Welsh is porffor. Why purple?

• No other JS engine is purple colored
• Purple is pretty cool
• Purple apparently represents "ambition", which is one word to describe this project

Yes!

No. they are not alike at all internally and have very different goals/ideals:

• Porffor is made as a generic JS engine, not for Wasm stuff specifically
• Porffor primarily consumes JS
• Porffor is written in pure JS and compiles itself, not using Binaryen/etc
• (Also I didn't know it existed when I started this, lol)