4 Commits
Author SHA1 Message Date
msyds 80164acb96 doc
build / build (push) Failing after 1m25s
2026-07-22 23:48:40 -06:00
msyds 1c7322614c fart 2026-07-22 23:48:02 -06:00
msyds 81a136fcf2 inspect-wasm 2026-07-22 12:22:41 -06:00
msyds be1d7566f4 idk ^w^
build / build (push) Failing after 1m10s
2026-07-20 13:09:59 -06:00
10 changed files with 472 additions and 12 deletions
+53
View File
@@ -0,0 +1,53 @@
#+title: Gyehoek Scheme
#+begin_center
(this document is written in present tense as if the project is complete, but Gyehoek is a work-in-progress.)
#+end_center
Gyehoek is an R⁷RS-compliant Scheme compiler targeting WebAssembly 3.0, relying principally on the recently standardised garbage collector and tail call proposals. the Gyehoek compiler is implemented in Haskell, and the Gyehoek runtime is a Rust program providing primitive routines and WebAssembly execution via the Wasmtime library.
primitives are implemented as native Rust functions made available to the guest by Wasmtime. in the future, it would be ideal to provide the primitives as a WASI interface to help decouple ourselves from a specific Wasm runtime, but it is not a priority.
Gyehoek allows separate compilation, ~eval~, first-class continuations, and so on.
* pipeline
a Scheme program's journey through Gyehoek is as follows:
1. read (source code → Scheme data)
2. parse (Scheme data → AST)
3. expand(?) (AST → AST)
4. contify (AST → CPS)
5. close (CPS → CPS)
6. lower (CPS → Wasm)
** read
in the read phase, Gyehoek's reader serialises textual source code into a sequence of tokens, which are then parsed into S-expressions. this phase is completely agnostic towards any interpretation of the data — it's just data, not code (yet). this distinction between reading and parsing is made so that the reader can easily be shared amongst many parsers, allowing convenient definition of human-readable representations for all sorts of compiler internals. Gyehoek's intermediate languages and WebAssembly text format are of particular interest.
the reader may be configured to extend R⁷RS's syntax with a special "antiquotation" notation, used internally in the compiler to elegantly interpolate and splice S-expression literals via Haskell's quasiquotation.
#+begin_src haskell
let meta = 123 :: Int
in [sx|(a b c #{meta} d)|] -- ⇒ (a b c 123 d)
let metas = ["c","d"] :: List Text
in [sx|(a b ##{metas} e f)|] -- ⇒ (a b "c" "d" e f)
#+end_src
Gyehoek's lexer and parser are generated by Alex and Happy, respectively.
unless otherwise noted, the term "parse" will be used in reference to the phase taking S-expressions to ASTs, while "read" refers to the combined Alex/Happy process. if the tokenisation process (Alex) must be distinguished from the "parse" process (Happy), the former is called "lexical analysis" and the latter "syntactic analysis."
** parse
- use invertible-grammar library
** expand
** contify
- procedures are distinguished from continuations, and procedure applications are distinguished from continuation jumps.
- all continuations and lambda will be named i think. the exception is continuations for primitive calls.
** close
** lower
+2 -1
View File
@@ -66,7 +66,7 @@ library
, base ^>=4.21.2.0
, binary
, containers
, cradle
, typed-process
, effectful
, effectful-core
, effectful-plugin
@@ -89,6 +89,7 @@ library
, text-short
, unordered-containers
, vector
, bytestring
hs-source-dirs: src
default-language: GHC2024
+5 -4
View File
@@ -28,6 +28,7 @@ import qualified Gyehoek.Sexp
import Data.Text qualified as T
import Data.Foldable (fold)
import Gyehoek.Sexp (encodeOrShow, toSexp)
import Debug.Pretty.Simple
data Env = MkEnv
@@ -182,10 +183,10 @@ lower' g e@(ExpApply f xs ktail) = do
(ref.cast (ref $closure))
(struct.get $closure $code)
(return_call_ref $cont-type)
;; (@gyehoek todo
;; (f' ##{f'})
;; (ktail #{l}))
|]
(@gyehoek todo
(f' ##{f'})
(ktail #{l}))
|]
lower' g e@(ExpContinue k xs) = do
let nargs = length xs
+92 -4
View File
@@ -2,6 +2,8 @@
{-# LANGUAGE TemplateHaskell #-}
{-# LANGUAGE PatternSynonyms #-}
{-# LANGUAGE ViewPatterns #-}
{-# LANGUAGE TypeFamilies #-}
{-# LANGUAGE FunctionalDependencies #-}
module Gyehoek.CPS.Syntax
( Val(..)
, Kappa(..)
@@ -18,10 +20,17 @@ module Gyehoek.CPS.Syntax
, _ExpPrim
, _ExpLetRec
, _ExpApply
, binders
, body
, op
, args
, cont
, cps
, pattern AbsLambda'
, pattern AbsKappa'
, Abs(..)
, free
, free'
)
where
@@ -41,6 +50,11 @@ import Language.Sexp.Located (Sexp)
import qualified Data.InvertibleGrammar.Base as IG
import qualified Gyehoek.Scheme.Syntax as Gyehoek
import Data.InvertibleGrammar.Base (type (:-)((:-)))
import Data.HashSet (HashSet)
import qualified Data.HashSet as HS
import Data.Hashable (Hashable)
import Data.Monoid (Endo)
import Data.Containers.ListUtils (nubOrd)
-- Data types
@@ -49,10 +63,10 @@ data Val
| ValLit Lit
deriving (Show, Generic, Data, Eq)
data Kappa = MkKappa (List Name) Exp
data Kappa = MkKappa { binders :: List Name, body :: Exp }
deriving (Show, Generic, Data, Eq)
data Lambda = MkLambda (List Name) Name Exp
data Lambda = MkLambda { binders :: List Name, ktail :: Name, body :: Exp }
deriving (Show, Generic, Data, Eq)
data Abs
@@ -65,7 +79,7 @@ pattern AbsLambda' xs e ktail = AbsLambda (MkLambda xs e ktail)
data Exp
= ExpPrim (Prim Val) Kappa
| ExpLetRec (NonEmpty (Name, Abs)) Exp
| ExpLetRec { binders :: NonEmpty (Name, Abs), body :: Exp }
| ExpContinue Name (List Val)
| ExpIf Val Exp Exp
| ExpApply
@@ -90,14 +104,30 @@ data Program = MkProgram
deriving (Show, Generic, Data)
makePrisms ''Kappa
-- makeLenses ''Kappa
makePrisms ''Exp
-- makeLenses ''Exp
-- makeFieldsNoPrefix ''Exp
-- makeFieldsNoPrefix ''Kappa
-- makeLensesWith abbreviatedFields ''Exp
-- makeLensesFor [("binders", "_binders"), ("body", "_body")] ''Exp
makeFieldsId ''Exp
makeFieldsId ''Kappa
makeFieldsId ''Lambda
instance HasBinders Abs (List Name) where
binders k (AbsKappa kap) = AbsKappa <$> binders k kap
binders k (AbsLambda lam) = AbsLambda <$> binders k lam
instance HasBody Abs Exp where
body k (AbsKappa kap) = AbsKappa <$> body k kap
body k (AbsLambda lam) = AbsLambda <$> body k lam
-- SexpIso instances
instance S.SexpIso Val where
sexpIso = match
-- $ With (. label)
$ With (\var -> var . S.sexpIso)
$ With (\lit -> lit . S.sexpIso)
$ End
@@ -192,3 +222,61 @@ instance CPS Abs where toCPS = Gyehoek.Sexp.fromSexp
cps :: QuasiQuoter
cps = Gyehoek.Sexp.makeSx' [| toCPS |]
deleteFrom :: (Foldable f, Hashable a) => f a -> HashSet a -> HashSet a
deleteFrom = flip $ foldr HS.delete
insertFrom :: (Foldable f, Hashable a) => f a -> HashSet a -> HashSet a
insertFrom = flip $ foldr HS.insert
toHashSetOf :: Hashable a => Getting (Endo (HashSet a)) s a -> s -> HashSet a
toHashSetOf l = foldrOf l HS.insert mempty
free :: Exp -> HashSet Name
free = go where
gokap (MkKappa xs m) = go m & deleteFrom xs
golam (MkLambda xs k m) = go m & deleteFrom xs & sans k
goabs = \case
AbsKappa kap -> gokap kap
AbsLambda lam -> golam lam
go = \case
ExpPrim p k ->
p & toHashSetOf (folded . #ValVar)
& HS.union (gokap k)
ExpLetRec bs m ->
foldMapOf (each . _2) goabs bs <> go m
& deleteFrom (bs ^.. each . _1)
ExpContinue k xs -> HS.fromList $ k : xs ^.. each . #ValVar
ExpIf c t f -> toHashSetOf #ValVar c <> go t <> go f
ExpApply f xs k -> toHashSetOf (each . #ValVar) (f:xs) <> HS.singleton k
-- | Free variables given in the order of their appearance.
free' :: Exp -> List Name
free' = nubOrd . goFree HS.empty where
goFreeKap bound (MkKappa xs m) = goFree (bound & insertFrom xs) m
goFreeLam bound (MkLambda xs k m) = goFree (bound & insertFrom (k:xs)) m
goFreeAbs bound = \case
AbsKappa kap -> goFreeKap bound kap
AbsLambda lam -> goFreeLam bound lam
goFree :: HashSet Name -> Exp -> List Name
goFree bound = \case
ExpPrim p k ->
p & toListOf (folded . #ValVar . filtered (`notElem` bound))
& (<> goFreeKap bound k)
ExpLetRec bs m ->
foldMapOf (each . _2) (goFreeAbs bound') bs <> goFree bound' m
where bound' = bound & insertFrom (bs ^.. each . _1)
ExpContinue k xs -> filter (`notElem` bound) (k : xs ^.. each . #ValVar)
ExpIf c t f ->
(c ^.. #ValVar . filtered (`notElem` bound))
<> goFree bound t <> goFree bound f
ExpApply f xs k ->
(f:xs) ^.. (each . #ValVar . filtered (`notElem` bound))
<> (k ^.. filtered (`notElem` bound))
freeLambda :: Lambda -> List Name
freeLambda (MkLambda {binders,ktail,body}) = _
+37 -2
View File
@@ -16,11 +16,18 @@ import qualified Gyehoek.Sexp as Sexp
import qualified Gyehoek.Scheme.Syntax as Scm
import qualified Data.Text.Encoding as T
import System.IO (Handle)
import System.IO qualified as IO
import Gyehoek.CPS.Convert
import Gyehoek.CPS.Lower
import Gyehoek.CPS.Syntax qualified as Cps
import Control.Monad
import Text.Pretty.Simple (pShowNoColor)
import System.Process.Typed
import Data.Text.Encoding (encodeUtf8)
import System.Environment.Blank (getEnvDefault)
import GHC.Conc (atomically)
import qualified Data.Text.IO as TIO
import qualified Data.ByteString.Lazy as BS
main :: IO ()
@@ -59,6 +66,31 @@ readScm f =
Sexp.parseSexps @Scm.CommandOrDef (fileName f) <$> hGetContents h
>>= either error (pure . Scm.MkProgram)
inspectWasm :: IOE :> es => Text -> Eff es ()
inspectWasm wat = do
pager_cmd <- liftIO $ getEnvDefault "PAGER" "less"
let wasmtools_cfg
= proc "wasm-tools" ["print", "-pf", "--print-operand-stack"
,"--color", "always", "-"]
-- & setStdin (byteStringInput . view lazy . encodeUtf8 $ wat)
-- & setStdout byteStringOutput
& setStdin createPipe
& setStdout createPipe
& setStderr inherit
let pager_cfg = proc pager_cmd []
& setStdin createPipe
& setStdout inherit
& setStderr inherit
liftIO $ withProcessWait_ wasmtools_cfg \wasmtools -> do
TIO.hPutStrLn (getStdin wasmtools) wat
IO.hFlush (getStdin wasmtools)
IO.hClose (getStdin wasmtools)
withProcessWait_ pager_cfg \pager -> do
t <- BS.hGetContents (getStdout wasmtools)
BS.hPut (getStdin pager) t
IO.hFlush (getStdin pager)
IO.hClose (getStdin pager)
driver
:: (GenSym :> es, FileSystem :> es, IOE :> es)
=> Options -> Eff es ()
@@ -70,8 +102,11 @@ driver opts = do
when opts.dumpCPS do
hPutStrLn FS.stdout $ Sexp.encodePretty cps ^?! _Right
wat <- lowerProgram cps
withFile opts.output FS.WriteMode \h ->
hPutStrLn h wat
if not opts.inspectWasm then
withFile opts.output FS.WriteMode \h ->
hPutStrLn h wat
else
inspectWasm wat
parse_e2e :: FilePath -> IO Scm.Program
parse_e2e = runEff . runFileSystem . readScm
+3
View File
@@ -19,6 +19,7 @@ data Options = MkOptions
-- , dumpQBE :: Maybe FilePath
dumpCPS :: Bool
, dumpParsed :: Bool
, inspectWasm :: Bool
, output :: FilePath
, sourceFile :: FilePath
}
@@ -49,10 +50,12 @@ parseOutput = strOption
parseDumpCPS = switch (long "dump-cps")
parseDumpParsed = switch (long "dump-parsed")
parseInspectWasm = switch $ long "inspect-wasm" <> short 'p'
parser :: Parser Options
parser = MkOptions
<$> parseDumpCPS
<*> parseDumpParsed
<*> parseInspectWasm
<*> parseOutput
<*> argument str (metavar "FILE")
+1 -1
View File
@@ -52,7 +52,7 @@ import Language.Haskell.TH.Quote (QuasiQuoter)
newtype Name = MkName { inner :: Text }
deriving newtype (Show, Eq, IsString, Gen, Hashable)
deriving newtype (Show, Eq, Ord, IsString, Gen, Hashable)
deriving stock (Generic, Data)
getName :: Name -> Text
+1
View File
@@ -0,0 +1 @@
(values 1 2)
+265
View File
@@ -0,0 +1,265 @@
(module
(import
"gyehoek"
"write"
(func $gh-write (param (ref eq))))
(import
"gyehoek"
"truthy?"
(func $gh-truthy? (param (ref eq)) (result i32)))
(type $heap-object (sub (struct (field $hash (mut i32)))))
(type $cont-type (func (param i32)))
(type $cont-stack-type (array (mut (ref null $cont-type))))
(type
$closure
(sub
$heap-object
(struct
(field $hash (mut i32))
(field $code (ref $cont-type)))))
(global $cont-stack-top (mut i32) (i32.const 0))
(global
$cont-stack
(ref $cont-stack-type)
(array.new_default $cont-stack-type (i32.const 128)))
(type $arg-array-type (array (mut (ref null eq))))
(global
$arg-array
(ref $arg-array-type)
(array.new_default $arg-array-type (i32.const 32)))
(global $result (mut (ref null eq)) (ref.null eq))
(func
$halt
(param i32)
(@gyehoek "pop argument")
(global.get $arg-array)
(i32.const 0)
(array.get $arg-array-type)
ref.as_non_null
(global.set $result))
(func
(param i32)
(@gyehoek :origin "(κ (x5) (continue λ-tail1 x5))")
(local (ref eq) (ref eq) (ref eq) (ref eq) (ref eq))
(@gyehoek "pop argument")
(global.get $arg-array)
(i32.const 0)
(array.get $arg-array-type)
ref.as_non_null
(local.set 1)
(@gyehoek :origin "(continue λ-tail1 x5)")
(@gyehoek "push args")
(@gyehoek "push argument")
(global.get $arg-array)
(i32.const 0)
(local.get 4)
(array.set $arg-array-type)
(@gyehoek "nargs")
(i32.const 1)
(@gyehoek "pop cont stack")
(global.get $cont-stack-top)
(i32.const 1)
i32.sub
(global.set $cont-stack-top)
(global.get $cont-stack)
(global.get $cont-stack-top)
(array.get $cont-stack-type)
ref.as_non_null
(return_call_ref $cont-type))
(elem declare funcref (ref.func 3))
(func
(param i32)
(@gyehoek
:origin
"(κ (x3) (letrec ((r4 (κ (x5) (continue λ-tail1 x5)))) (f x3 r4)))")
(local (ref eq) (ref eq) (ref eq) (ref eq) (ref eq))
(@gyehoek "pop argument")
(global.get $arg-array)
(i32.const 0)
(array.get $arg-array-type)
ref.as_non_null
(local.set 1)
(@gyehoek :origin "(f x3 r4)")
(@gyehoek "push cont" :idx 3)
(array.set
$cont-stack-type
(global.get $cont-stack)
(global.get $cont-stack-top)
(ref.func 3))
(global.set
$cont-stack-top
(i32.add (global.get $cont-stack-top) (i32.const 1)))
(@gyehoek :origin "(f x3 r4)")
(@gyehoek "load args")
(@gyehoek "push argument")
(global.get $arg-array)
(i32.const 0)
(local.get 3)
(array.set $arg-array-type)
(i32.const 1)
(local.get 1)
(ref.cast (ref $closure))
(struct.get $closure $code)
(return_call_ref $cont-type))
(elem declare funcref (ref.func 4))
(func
(param i32)
(@gyehoek
:origin
"(λ (f x λ-tail1) (letrec ((r2 (κ (x3) (letrec ((r4 (κ (x5) (continue λ-tail1 x5)))) (f x3 r4))))) (f x r2)))")
(local (ref eq) (ref eq) (ref eq) (ref eq) (ref eq))
(@gyehoek "pop argument")
(global.get $arg-array)
(i32.const 0)
(array.get $arg-array-type)
ref.as_non_null
(local.set 1)
(@gyehoek "pop argument")
(global.get $arg-array)
(i32.const 1)
(array.get $arg-array-type)
ref.as_non_null
(local.set 2)
(@gyehoek :origin "(f x r2)")
(@gyehoek "push cont" :idx 4)
(array.set
$cont-stack-type
(global.get $cont-stack)
(global.get $cont-stack-top)
(ref.func 4))
(global.set
$cont-stack-top
(i32.add (global.get $cont-stack-top) (i32.const 1)))
(@gyehoek :origin "(f x r2)")
(@gyehoek "load args")
(@gyehoek "push argument")
(global.get $arg-array)
(i32.const 0)
(local.get 2)
(array.set $arg-array-type)
(i32.const 1)
(local.get 1)
(ref.cast (ref $closure))
(struct.get $closure $code)
(return_call_ref $cont-type))
(elem declare funcref (ref.func 5))
(func
(param i32)
(@gyehoek
:origin
"(λ (x λ-tail7) (prim (+ x 4) (κ (r8) (continue λ-tail7 r8))))")
(local (ref eq) (ref eq) (ref eq) (ref eq) (ref eq))
(@gyehoek "pop argument")
(global.get $arg-array)
(i32.const 0)
(array.get $arg-array-type)
ref.as_non_null
(local.set 1)
(@gyehoek
:origin
"(prim (+ x 4) (κ (r8) (continue λ-tail7 r8)))")
(local.get 1)
(i31.get_s (ref.cast (ref i31)))
(i32.const 1)
i32.shr_u
(i32.const 4)
(@gyehoek "construct small fixnum")
(i32.const 1)
i32.shl
ref.i31
(i31.get_s (ref.cast (ref i31)))
(i32.const 1)
i32.shr_u
i32.add
(@gyehoek "construct small fixnum")
(i32.const 1)
i32.shl
ref.i31
(local.set 2)
(@gyehoek :origin "(continue λ-tail7 r8)")
(@gyehoek "push args")
(@gyehoek "push argument")
(global.get $arg-array)
(i32.const 0)
(local.get 2)
(array.set $arg-array-type)
(@gyehoek "nargs")
(i32.const 1)
(@gyehoek "pop cont stack")
(global.get $cont-stack-top)
(i32.const 1)
i32.sub
(global.set $cont-stack-top)
(global.get $cont-stack)
(global.get $cont-stack-top)
(array.get $cont-stack-type)
ref.as_non_null
(return_call_ref $cont-type))
(elem declare funcref (ref.func 6))
(func
(param i32)
(@gyehoek :origin "(κ (x10) (continue halt x10))")
(local (ref eq) (ref eq) (ref eq) (ref eq) (ref eq))
(@gyehoek "pop argument")
(global.get $arg-array)
(i32.const 0)
(array.get $arg-array-type)
ref.as_non_null
(local.set 1)
(@gyehoek "push argument")
(global.get $arg-array)
(i32.const 0)
(local.get 3)
(array.set $arg-array-type)
(return_call $halt (i32.const 1)))
(elem declare funcref (ref.func 7))
(func
$scm-entry
(param i32)
(@gyehoek
:origin
"(letrec ((λ-body0 (λ (f x λ-tail1) (letrec ((r2 (κ (x3) (letrec ((r4 (κ (x5) (continue λ-tail1 x5)))) (f x3 r4))))) (f x r2))))) (letrec ((λ-body6 (λ (x λ-tail7) (prim (+ x 4) (κ (r8) (continue λ-tail7 r8)))))) (letrec ((r9 (κ (x10) (continue halt x10)))) (λ-body0 λ-body6 9 r9))))")
(local (ref eq) (ref eq) (ref eq) (ref eq) (ref eq))
(i32.const 0)
(ref.func 5)
(struct.new $closure)
(local.set 1)
(i32.const 0)
(ref.func 6)
(struct.new $closure)
(local.set 2)
(@gyehoek :origin "(λ-body0 λ-body6 9 r9)")
(@gyehoek "push cont" :idx 7)
(array.set
$cont-stack-type
(global.get $cont-stack)
(global.get $cont-stack-top)
(ref.func 7))
(global.set
$cont-stack-top
(i32.add (global.get $cont-stack-top) (i32.const 1)))
(@gyehoek :origin "(λ-body0 λ-body6 9 r9)")
(@gyehoek "load args")
(@gyehoek "push argument")
(global.get $arg-array)
(i32.const 0)
(local.get 2)
(array.set $arg-array-type)
(@gyehoek "push argument")
(global.get $arg-array)
(i32.const 1)
(i32.const 9)
(@gyehoek "construct small fixnum")
(i32.const 1)
i32.shl
ref.i31
(array.set $arg-array-type)
(i32.const 1)
(local.get 1)
(ref.cast (ref $closure))
(struct.get $closure $code)
(return_call_ref $cont-type))
(func
(export "main")
(call $scm-entry (i32.const 0))
(call $gh-write (ref.as_non_null (global.get $result)))))
+13
View File
@@ -1,3 +1,4 @@
{-# LANGUAGE OverloadedLists #-}
module Gyehoek.Test.CPS.Syntax (root) where
import Test.Tasty (TestTree, testGroup)
@@ -13,8 +14,20 @@ import Gyehoek.Test.Sexp (equivto)
root :: IO TestTree
root = pure . testGroup "cps syntax" $
[ qqTree
, freeTree
]
freeTree :: TestTree
freeTree = testCase "free" do
Sut.free [cps|
(letrec ((x (lambda (r k1) (continue k1 y)))
(y (lambda (r k2) (continue k2 x))))
(continue x y k3))|] @=? ["k3"]
Sut.free' [cps|
(letrec ((x (lambda (r k1) (continue k1 y)))
(y (lambda (r k2) (continue k2 x))))
(continue x y k3))|] @=? ["k3"]
qqTree :: TestTree
qqTree = testGroup "parser"
[ testCase "lambda" do