\section[ByteCodeGen]{Generate bytecode from Core}
\begin{code}
-module ByteCodeGen ( byteCodeGen ) where
+module ByteCodeGen ( UnlinkedBCO, UnlinkedBCOExpr, ItblEnv, ClosureEnv, HValue,
+ filterNameMap,
+ byteCodeGen, coreExprToBCOs,
+ linkIModules, linkIExpr
+ ) where
#include "HsVersions.h"
---import Id
---import Name
---import PrimOp
-
import Outputable
-import Name ( Name, getName )
-import Id ( Id, idType, isDataConId_maybe )
+import Name ( Name, getName, mkSysLocalName )
+import Id ( Id, idType, isDataConId_maybe, mkVanillaId,
+ isPrimOpId_maybe, idPrimRep )
import OrdList ( OrdList, consOL, snocOL, appOL, unitOL,
- nilOL, toOL, concatOL )
-import FiniteMap ( FiniteMap, addListToFM, listToFM,
- addToFM, lookupFM, fmToList )
+ nilOL, toOL, concatOL, fromOL )
+import FiniteMap ( FiniteMap, addListToFM, listToFM,
+ addToFM, lookupFM, fmToList, plusFM )
import CoreSyn
-import Literal ( Literal(..) )
+import PprCore ( pprCoreExpr )
+import Literal ( Literal(..), literalPrimRep )
import PrimRep ( PrimRep(..) )
+import PrimOp ( PrimOp(..) )
import CoreFVs ( freeVars )
import Type ( typePrimRep )
-import DataCon ( DataCon, dataConTag, fIRST_TAG )
-import Util ( zipEqual, zipWith4Equal )
+import DataCon ( dataConTag, fIRST_TAG, dataConTyCon, dataConWrapId )
+import TyCon ( TyCon, tyConFamilySize )
+import Class ( Class, classTyCon )
+import Util ( zipEqual, zipWith4Equal, naturalMergeSortLe, nOfThem )
+import Var ( isTyVar )
import VarSet ( VarSet, varSetElems )
---import FastTypes
+import PrimRep ( getPrimRepSize, isFollowableRep )
+import CmdLineOpts ( DynFlags, DynFlag(..) )
+import ErrUtils ( showPass, dumpIfSet_dyn )
+import Unique ( mkPseudoUnique3 )
+import FastString ( FastString(..) )
+import PprType ( pprType )
+import ByteCodeInstr ( BCInstr(..), ProtoBCO(..), nameOfProtoBCO )
+import ByteCodeItbls ( ItblEnv, mkITbls )
+import ByteCodeLink ( UnlinkedBCO, UnlinkedBCOExpr, assembleBCO,
+ ClosureEnv, HValue, linkSomeBCOs, filterNameMap )
+
+import List ( intersperse, sortBy )
+import Foreign ( Ptr(..), mallocBytes )
+import Addr ( Addr(..), addrToInt, writeCharOffAddr )
+import CTypes ( CInt )
+
+import PrelBase ( Int(..) )
+import PrelGHC ( ByteArray# )
+import IOExts ( unsafePerformIO )
+import PrelIOBase ( IO(..) )
+
\end{code}
-Entry point.
+%************************************************************************
+%* *
+\subsection{Functions visible from outside this module.}
+%* *
+%************************************************************************
\begin{code}
-byteCodeGen :: [CoreBind] -> [BCO Name]
-byteCodeGen binds
- = let flatBinds = concatMap getBind binds
- getBind (NonRec bndr rhs) = [(bndr, freeVars rhs)]
- getBind (Rec binds) = [(bndr, freeVars rhs) | (bndr,rhs) <- binds]
- in
- snd (initBc [] (mapBc schemeR flatBinds))
+
+byteCodeGen :: DynFlags
+ -> [CoreBind]
+ -> [TyCon] -> [Class]
+ -> IO ([UnlinkedBCO], ItblEnv)
+byteCodeGen dflags binds local_tycons local_classes
+ = do showPass dflags "ByteCodeGen"
+ let tycs = local_tycons ++ map classTyCon local_classes
+ itblenv <- mkITbls tycs
+
+ let flatBinds = concatMap getBind binds
+ getBind (NonRec bndr rhs) = [(bndr, freeVars rhs)]
+ getBind (Rec binds) = [(bndr, freeVars rhs) | (bndr,rhs) <- binds]
+ final_state = runBc (BcM_State [] 0)
+ (mapBc (schemeR True) flatBinds
+ `thenBc_` returnBc ())
+ (BcM_State proto_bcos final_ctr) = final_state
+
+ dumpIfSet_dyn dflags Opt_D_dump_BCOs
+ "Proto-bcos" (vcat (intersperse (char ' ') (map ppr proto_bcos)))
+
+ bcos <- mapM assembleBCO proto_bcos
+
+ return (bcos, itblenv)
+
+
+-- Returns: (the root BCO for this expression,
+-- a list of auxilary BCOs resulting from compiling closures)
+coreExprToBCOs :: DynFlags
+ -> CoreExpr
+ -> IO UnlinkedBCOExpr
+coreExprToBCOs dflags expr
+ = do showPass dflags "ByteCodeGen"
+
+ -- create a totally bogus name for the top-level BCO; this
+ -- should be harmless, since it's never used for anything
+ let invented_name = mkSysLocalName (mkPseudoUnique3 0) SLIT("Expr-Top-Level")
+ let invented_id = mkVanillaId invented_name (panic "invented_id's type")
+
+ let (BcM_State all_proto_bcos final_ctr)
+ = runBc (BcM_State [] 0)
+ (schemeR True (invented_id, freeVars expr))
+ dumpIfSet_dyn dflags Opt_D_dump_BCOs
+ "Proto-bcos" (vcat (intersperse (char ' ') (map ppr all_proto_bcos)))
+
+ let root_proto_bco
+ = case filter ((== invented_name).nameOfProtoBCO) all_proto_bcos of
+ [root_bco] -> root_bco
+ auxiliary_proto_bcos
+ = filter ((/= invented_name).nameOfProtoBCO) all_proto_bcos
+
+ auxiliary_bcos <- mapM assembleBCO auxiliary_proto_bcos
+ root_bco <- assembleBCO root_proto_bco
+
+ return (root_bco, auxiliary_bcos)
+
+
+-- Linking stuff
+linkIModules :: ItblEnv -- incoming global itbl env; returned updated
+ -> ClosureEnv -- incoming global closure env; returned updated
+ -> [([UnlinkedBCO], ItblEnv)]
+ -> IO ([HValue], ItblEnv, ClosureEnv)
+linkIModules gie gce mods
+ = do let (bcoss, ies) = unzip mods
+ bcos = concat bcoss
+ final_gie = foldr plusFM gie ies
+ (final_gce, linked_bcos) <- linkSomeBCOs final_gie gce bcos
+ return (linked_bcos, final_gie, final_gce)
+
+
+linkIExpr :: ItblEnv -> ClosureEnv -> UnlinkedBCOExpr
+ -> IO HValue -- IO BCO# really
+linkIExpr ie ce (root_ul_bco, aux_ul_bcos)
+ = do (aux_ce, _) <- linkSomeBCOs ie ce aux_ul_bcos
+ (_, [root_bco]) <- linkSomeBCOs ie aux_ce [root_ul_bco]
+ return root_bco
\end{code}
-The real machinery.
+%************************************************************************
+%* *
+\subsection{Compilation schema for the bytecode generator.}
+%* *
+%************************************************************************
\begin{code}
-data BCInstr
- -- Messing with the stack
- = ARGCHECK Int
- | PUSH_L Int{-size-} Int{-offset-}
- | PUSH_G Name
- | PUSH_ALTS Name{-labels the alt BCO; derived from case binder-}
- | PUSH_I Integer
- | SLIDE Int{-this many-} Int{-down by this much-}
- -- To do with the heap
- | ALLOC Int
- | MKAP Int{-place ptr to heap this far down stack-} Int{-# words-}
- | UNPACK Int
- | PACK DataCon Int
- -- Casery (in French: caseage)
- | CASE_PTR [(Discr, BCInstrList)]
- | CASE_INT [(Discr, BCInstrList)]
- | CASE_FLOAT [(Discr, BCInstrList)]
- | CASE_DOUBLE [(Discr, BCInstrList)]
- -- To Infinity And Beyond
- | ENTER
-
type BCInstrList = OrdList BCInstr
-data BCO a = BCO a BCInstrList
-
-
type Sequel = Int -- back off to this depth before ENTER
-- Maps Ids to the offset from the stack _base_ so we don't have
-- to mess with it after each push/pop.
type BCEnv = FiniteMap Id Int -- To find vars on the stack
-lookupBCEnv :: BCEnv -> Id -> Int
-lookupBCEnv env nm
- = case lookupFM env nm of
- Nothing -> pprPanic "lookupBCEnv"
- (ppr nm $$ char ' ' $$ vcat (map ppr (fmToList env)))
- Just xx -> xx
-
-lookupBCEnv_maybe :: BCEnv -> Id -> Maybe Int
-lookupBCEnv_maybe = lookupFM
-
-
--- Describes case alts
-data Discr
- = DiscrI Int
- | DiscrF Rational
- | DiscrD Rational
- | DiscrP Int
- | NoDiscr
-
--- Hmm. This isn't really right (ie on Alpha, idSizeW Double -> 2)
--- There must be an Officially Approved way to do this somewhere.
-idSizeW :: Id -> Int
-idSizeW nm
- = let pr = typePrimRep (idType nm)
- in case pr of IntRep -> 2
- FloatRep -> 2
- DoubleRep -> 3
- PtrRep -> 1
- other -> pprPanic "ByteCodeGen.idSizeW" (ppr pr)
+ppBCEnv :: BCEnv -> SDoc
+ppBCEnv p
+ = text "begin-env"
+ $$ nest 4 (vcat (map pp_one (sortBy cmp_snd (fmToList p))))
+ $$ text "end-env"
+ where
+ pp_one (var, offset) = int offset <> colon <+> ppr var
+ cmp_snd x y = compare (snd x) (snd y)
+-- Create a BCO and do a spot of peephole optimisation on the insns
+-- at the same time.
+mkProtoBCO nm instrs_ordlist origin
+ = ProtoBCO nm (peep (fromOL instrs_ordlist)) origin
+ where
+ peep (PUSH_L off1 : PUSH_L off2 : PUSH_L off3 : rest)
+ = PUSH_LLL off1 (off2-1) (off3-2) : peep rest
+ peep (PUSH_L off1 : PUSH_L off2 : rest)
+ = PUSH_LL off1 (off2-1) : peep rest
+ peep (i:rest)
+ = i : peep rest
+ peep []
+ = []
-- Compile code for the right hand side of a let binding.
-- Park the resulting BCO in the monad. Also requires the
-- variable to which this value was bound, so as to give the
--- resulting BCO a name.
-schemeR :: (Id, AnnExpr Id VarSet) -> BcM ()
-schemeR (nm, rhs) = schemeR_wrk nm (collect [] rhs)
+-- resulting BCO a name. Bool indicates top-levelness.
+
+schemeR :: Bool -> (Id, AnnExpr Id VarSet) -> BcM ()
+schemeR is_top (nm, rhs)
+{-
+ | trace (showSDoc (
+ (char ' '
+ $$ (ppr.filter (not.isTyVar).varSetElems.fst) rhs
+ $$ pprCoreExpr (deAnnotate rhs)
+ $$ char ' '
+ ))) False
+ = undefined
+-}
+ | otherwise
+ = schemeR_wrk is_top rhs nm (collect [] rhs)
+
+
+collect xs (_, AnnNote note e)
+ = collect xs e
+collect xs (_, AnnLam x e)
+ = collect (if isTyVar x then xs else (x:xs)) e
+collect xs not_lambda
+ = (reverse xs, not_lambda)
-collect xs (_, AnnLam x e) = collect (x:xs) e
-collect xs not_lambda = (reverse xs, not_lambda)
+schemeR_wrk is_top original_body nm (args, body)
+ | Just dcon <- maybe_toplevel_null_con_rhs
+ = --trace ("nullary constructor! " ++ showSDocDebug (ppr nm)) (
+ emitBc (mkProtoBCO (getName nm) (toOL [PACK dcon 0, ENTER])
+ (Right original_body))
+ --)
-schemeR_wrk nm (args, body)
- = let fvs = fst body
- all_args = varSetElems fvs ++ args
- szsw_args = map idSizeW all_args
+ | otherwise
+ = let fvs = filter (not.isTyVar) (varSetElems (fst original_body))
+ all_args = reverse args ++ fvs
+ szsw_args = map taggedIdSizeW all_args
szw_args = sum szsw_args
- p_init = listToFM (zip all_args (scanl (+) 0 szsw_args))
- argcheck = if null args then nilOL else unitOL (ARGCHECK szw_args)
+ p_init = listToFM (zip all_args (mkStackOffsets 0 szsw_args))
+ argcheck = unitOL (ARGCHECK szw_args)
in
schemeE szw_args 0 p_init body `thenBc` \ body_code ->
- emitBc (BCO (getName nm) (appOL argcheck body_code))
+ emitBc (mkProtoBCO (getName nm) (appOL argcheck body_code)
+ (Right original_body))
+ where
+ maybe_toplevel_null_con_rhs
+ | is_top && null args
+ = case snd body of
+ AnnVar v_wrk
+ -> case isDataConId_maybe v_wrk of
+ Nothing -> Nothing
+ Just dc_wrk | nm == dataConWrapId dc_wrk
+ -> Just dc_wrk
+ | otherwise
+ -> Nothing
+ other -> Nothing
+ | otherwise
+ = Nothing
+
+-- Let szsw be the sizes in words of some items pushed onto the stack,
+-- which has initial depth d'. Return the values which the stack environment
+-- should map these items to.
+mkStackOffsets :: Int -> [Int] -> [Int]
+mkStackOffsets original_depth szsw
+ = map (subtract 1) (tail (scanl (+) original_depth szsw))
-- Compile code to apply the given expression to the remaining args
-- on the stack, returning a HNF.
schemeE :: Int -> Sequel -> BCEnv -> AnnExpr Id VarSet -> BcM BCInstrList
-- Delegate tail-calls to schemeT.
-schemeE d s p (fvs, AnnApp f a) = returnBc (schemeT d s 0 p (fvs, AnnApp f a))
-schemeE d s p (fvs, AnnVar v) = returnBc (schemeT d s 0 p (fvs, AnnVar v))
+schemeE d s p e@(fvs, AnnApp f a)
+ = returnBc (schemeT d s p (fvs, AnnApp f a))
+schemeE d s p e@(fvs, AnnVar v)
+ | isFollowableRep v_rep
+ = returnBc (schemeT d s p (fvs, AnnVar v))
+
+ | otherwise
+ = -- returning an unboxed value. Heave it on the stack, SLIDE, and RETURN.
+ let (push, szw) = pushAtom True d p (AnnVar v)
+ in returnBc (push -- value onto stack
+ `appOL` mkSLIDE szw (d-s) -- clear to sequel
+ `snocOL` RETURN v_rep) -- go
+ where
+ v_rep = typePrimRep (idType v)
+
+schemeE d s p (fvs, AnnLit literal)
+ = let (push, szw) = pushAtom True d p (AnnLit literal)
+ l_rep = literalPrimRep literal
+ in returnBc (push -- value onto stack
+ `appOL` mkSLIDE szw (d-s) -- clear to sequel
+ `snocOL` RETURN l_rep) -- go
schemeE d s p (fvs, AnnLet binds b)
= let (xs,rhss) = case binds of AnnNonRec x rhs -> ([x],[rhs])
AnnRec xs_n_rhss -> unzip xs_n_rhss
- in
- mapBc schemeR (zip xs rhss) `thenBc_`
- let n = length xs
- fvss = map (varSetElems.fst) rhss
- sizes = map (\rhs_fvs -> 1 + sum (map idSizeW rhs_fvs)) fvss
- p' = addListToFM p (zipE xs [d .. d+n-1])
+ n = length xs
+ fvss = map (filter (not.isTyVar).varSetElems.fst) rhss
+
+ -- Sizes of tagged free vars, + 1 for the fn
+ sizes = map (\rhs_fvs -> 1 + sum (map taggedIdSizeW rhs_fvs)) fvss
+
+ -- This p', d' defn is safe because all the items being pushed
+ -- are ptrs, so all have size 1. d' and p' reflect the stack
+ -- after the closures have been allocated in the heap (but not
+ -- filled in), and pointers to them parked on the stack.
+ p' = addListToFM p (zipE xs (mkStackOffsets d (nOfThem n 1)))
d' = d + n
+
infos = zipE4 fvss sizes xs [n, n-1 .. 1]
zipE = zipEqual "schemeE"
zipE4 = zipWith4Equal "schemeE" (\a b c d -> (a,b,c,d))
-- ToDo: don't build thunks for things with no free variables
- buildThunk (fvs, size, id, off)
- = case unzip (map (pushAtom d' p . AnnVar) (reverse fvs)) of
- (push_codes, pushed_szsw)
- -> ASSERT(sum pushed_szsw == size - 1)
- (toOL push_codes `snocOL` PUSH_G (getName id)
- `appOL` unitOL (MKAP off size))
-
- thunkCode = concatOL (map buildThunk infos)
+ buildThunk dd ([], size, id, off)
+ = PUSH_G (Left (getName id))
+ `consOL` unitOL (MKAP (off+size-1) size)
+ buildThunk dd ((fv:fvs), size, id, off)
+ = case pushAtom True dd p' (AnnVar fv) of
+ (push_code, pushed_szw)
+ -> push_code `appOL`
+ buildThunk (dd+pushed_szw) (fvs, size, id, off)
+
+ thunkCode = concatOL (map (buildThunk d') infos)
allocCode = toOL (map ALLOC sizes)
in
- schemeE d' s p' b `thenBc` \ bodyCode ->
- mapBc schemeR (zip xs rhss) `thenBc` \_ ->
+ schemeE d' s p' b `thenBc` \ bodyCode ->
+ mapBc (schemeR False) (zip xs rhss) `thenBc_`
returnBc (allocCode `appOL` thunkCode `appOL` bodyCode)
-- Env and depth in which to compile the alts, not including
-- any vars bound by the alts themselves
- d' = d + ret_frame_sizeW + idSizeW bndr
- p' = addToFM p bndr d'
-
- (case_instr, isAlgCase)
- = case typePrimRep (idType bndr) of
- IntRep -> (CASE_INT, False)
- FloatRep -> (CASE_FLOAT, False)
- DoubleRep -> (CASE_DOUBLE, False)
- PtrRep -> (CASE_PTR, True)
- other -> pprPanic "ByteCodeGen.schemeE" (ppr other)
-
- -- make the code for an alt
- codeAlt (discr, binds, rhs)
+ d' = d + ret_frame_sizeW + taggedIdSizeW bndr
+ p' = addToFM p bndr (d' - 1)
+
+ scrut_primrep = typePrimRep (idType bndr)
+ isAlgCase
+ = case scrut_primrep of
+ CharRep -> False ; AddrRep -> False ; WordRep -> False
+ IntRep -> False ; FloatRep -> False ; DoubleRep -> False
+ PtrRep -> True
+ other -> pprPanic "ByteCodeGen.schemeE" (ppr other)
+
+ -- given an alt, return a discr and code for it.
+ codeAlt alt@(discr, binds_f, rhs)
| isAlgCase
- = let binds_szsw = map idSizeW binds
- binds_szw = sum binds_szsw
- p'' = addListToFM p' (zip binds (scanl (+) d' binds_szsw))
- d'' = d' + binds_szw
- in schemeE d'' s p'' rhs `thenBc` \ rhs_code ->
- returnBc (UNPACK binds_szw `consOL` rhs_code)
+ = let (unpack_code, d_after_unpack, p_after_unpack)
+ = mkUnpackCode binds_f d' p'
+ in schemeE d_after_unpack s p_after_unpack rhs
+ `thenBc` \ rhs_code ->
+ returnBc (my_discr alt, unpack_code `appOL` rhs_code)
| otherwise
- = ASSERT(null binds) schemeE d' s p' rhs
+ = ASSERT(null binds_f)
+ schemeE d' s p' rhs `thenBc` \ rhs_code ->
+ returnBc (my_discr alt, rhs_code)
- discr (DEFAULT, binds, rhs) = NoDiscr
- discr (DataAlt dc, binds, rhs) = DiscrP (dataConTag dc - fIRST_TAG)
- discr (LitAlt l, binds, rhs)
+ my_discr (DEFAULT, binds, rhs) = NoDiscr
+ my_discr (DataAlt dc, binds, rhs) = DiscrP (dataConTag dc - fIRST_TAG)
+ my_discr (LitAlt l, binds, rhs)
= case l of MachInt i -> DiscrI (fromInteger i)
- MachFloat r -> DiscrF r
- MachDouble r -> DiscrD r
+ MachFloat r -> DiscrF (fromRational r)
+ MachDouble r -> DiscrD (fromRational r)
+ MachChar i -> DiscrI i
+ _ -> pprPanic "schemeE(AnnCase).my_discr" (ppr l)
+
+ maybe_ncons
+ | not isAlgCase = Nothing
+ | otherwise
+ = case [dc | (DataAlt dc, _, _) <- alts] of
+ [] -> Nothing
+ (dc:_) -> Just (tyConFamilySize (dataConTyCon dc))
- discrs = map discr alts
in
- mapBc codeAlt alts `thenBc` \ alt_codes ->
+ mapBc codeAlt alts `thenBc` \ alt_stuff ->
+ mkMultiBranch maybe_ncons alt_stuff `thenBc` \ alt_final ->
let
- alt_code = case_instr (zip discrs alt_codes)
+ alt_final_ac = ARGCHECK (taggedIdSizeW bndr) `consOL` alt_final
alt_bco_name = getName bndr
- alt_bco = BCO alt_bco_name (unitOL alt_code)
+ alt_bco = mkProtoBCO alt_bco_name alt_final_ac (Left alts)
in
schemeE (d + ret_frame_sizeW)
(d + ret_frame_sizeW) p scrut `thenBc` \ scrut_code ->
emitBc alt_bco `thenBc_`
- returnBc (PUSH_ALTS alt_bco_name `consOL` scrut_code)
+ returnBc (PUSH_AS alt_bco_name scrut_primrep `consOL` scrut_code)
+
+
+schemeE d s p (fvs, AnnNote note body)
+ = schemeE d s p body
+
+schemeE d s p other
+ = pprPanic "ByteCodeGen.schemeE: unhandled case"
+ (pprCoreExpr (deAnnotate other))
+
+
+-- Compile code to do a tail call. Three cases:
+--
+-- 1. A nullary constructor. Push its closure on the stack
+-- and SLIDE and RETURN.
+--
+-- 2. Application of a non-nullary constructor, by defn saturated.
+-- Split the args into ptrs and non-ptrs, and push the nonptrs,
+-- then the ptrs, and then do PACK and RETURN.
+--
+-- 3. Otherwise, it must be a function call. Push the args
+-- right to left, SLIDE and ENTER.
+schemeT :: Int -- Stack depth
+ -> Sequel -- Sequel depth
+ -> BCEnv -- stack env
+ -> AnnExpr Id VarSet
+ -> BCInstrList
--- Compile code to do a tail call. Doesn't need to be monadic.
-schemeT :: Int -> Sequel -> Int -> BCEnv -> AnnExpr Id VarSet -> BCInstrList
+schemeT d s p app
+-- | trace ("schemeT: env in = \n" ++ showSDocDebug (ppBCEnv p)) False
+-- = panic "schemeT ?!?!"
-schemeT d s narg_words p (_, AnnApp f a)
- = let (push, arg_words) = pushAtom d p (snd a)
- in push
- `consOL` schemeT (d+arg_words) s (narg_words+arg_words) p f
+ -- Handle case 1
+ | is_con_call && null args_r_to_l
+ = (PUSH_G (Left (getName con)) `consOL` mkSLIDE 1 (d-s))
+ `snocOL` ENTER
-schemeT d s narg_words p (_, AnnVar f)
- | Just con <- isDataConId_maybe f
- = PACK con narg_words `consOL` SLIDE 1 (d-s-1) `consOL` unitOL ENTER
+ -- Cases 2 and 3
| otherwise
- = let (push, arg_words) = pushAtom d p (AnnVar f)
- in push
- `consOL` SLIDE (narg_words+arg_words) (d - s - narg_words)
- `consOL` unitOL ENTER
+ = code
+
+ where
+ -- Extract the args (R->L) and fn
+ (args_r_to_l_raw, fn) = chomp app
+ chomp expr
+ = case snd expr of
+ AnnVar v -> ([], v)
+ AnnApp f a -> case chomp f of (az, f) -> (snd a:az, f)
+ AnnNote n e -> chomp e
+ other -> pprPanic "schemeT"
+ (ppr (deAnnotate (panic "schemeT.chomp", other)))
+
+ args_r_to_l = filter (not.isTypeAtom) args_r_to_l_raw
+ isTypeAtom (AnnType _) = True
+ isTypeAtom _ = False
+
+ -- decide if this is a constructor call, and rearrange
+ -- args appropriately.
+ maybe_dcon = isDataConId_maybe fn
+ is_con_call = case maybe_dcon of Nothing -> False; Just _ -> True
+ (Just con) = maybe_dcon
+
+ args_final_r_to_l
+ | not is_con_call
+ = args_r_to_l
+ | otherwise
+ = filter (not.isPtr) args_r_to_l ++ filter isPtr args_r_to_l
+ where isPtr = isFollowableRep . atomRep
+
+ -- make code to push the args and then do the SLIDE-ENTER thing
+ code = do_pushery d args_final_r_to_l
+
+ tag_when_push = not is_con_call
+ narg_words = sum (map (get_arg_szw . atomRep) args_r_to_l)
+ get_arg_szw = if tag_when_push then taggedSizeW else untaggedSizeW
+
+ do_pushery d (arg:args)
+ = let (push, arg_words) = pushAtom tag_when_push d p arg
+ in push `appOL` do_pushery (d+arg_words) args
+ do_pushery d []
+ = case maybe_dcon of
+ Just con -> PACK con narg_words `consOL` (
+ mkSLIDE 1 (d - narg_words - s) `snocOL` ENTER)
+ Nothing
+ -> let (push, arg_words) = pushAtom True d p (AnnVar fn)
+ in push
+ `appOL` mkSLIDE (narg_words+arg_words)
+ (d - s - narg_words)
+ `snocOL` ENTER
+
+mkSLIDE n d
+ = if d == 0 then nilOL else unitOL (SLIDE n d)
+
+atomRep (AnnVar v) = typePrimRep (idType v)
+atomRep (AnnLit l) = literalPrimRep l
+atomRep (AnnNote n b) = atomRep (snd b)
+atomRep (AnnApp f (_, AnnType _)) = atomRep (snd f)
+atomRep (AnnLam x e) | isTyVar x = atomRep (snd e)
+atomRep other = pprPanic "atomRep" (ppr (deAnnotate (undefined,other)))
+
+
+-- Make code to unpack the top-of-stack constructor onto the stack,
+-- adding tags for the unboxed bits. Takes the PrimReps of the
+-- constructor's arguments. off_h and off_s are travelling offsets
+-- along the constructor and the stack.
+--
+-- Supposing a constructor in the heap has layout
+--
+-- Itbl p_1 ... p_i np_1 ... np_j
+--
+-- then we add to the stack, shown growing down, the following:
+--
+-- (previous stack)
+-- p_i
+-- ...
+-- p_1
+-- np_j
+-- tag_for(np_j)
+-- ..
+-- np_1
+-- tag_for(np_1)
+--
+-- so that in the common case (ptrs only) a single UNPACK instr can
+-- copy all the payload of the constr onto the stack with no further ado.
+
+mkUnpackCode :: [Id] -- constr args
+ -> Int -- depth before unpack
+ -> BCEnv -- env before unpack
+ -> (BCInstrList, Int, BCEnv)
+mkUnpackCode vars d p
+ = --trace ("mkUnpackCode: " ++ showSDocDebug (ppr vars)
+ -- ++ " --> " ++ show d' ++ "\n" ++ showSDocDebug (ppBCEnv p')
+ -- ++ "\n") (
+ (code_p `appOL` code_np, d', p')
+ --)
+ where
+ -- vars with reps
+ vreps = [(var, typePrimRep (idType var)) | var <- vars]
+
+ -- ptrs and nonptrs, forward
+ vreps_p = filter (isFollowableRep.snd) vreps
+ vreps_np = filter (not.isFollowableRep.snd) vreps
+
+ -- the order in which we will augment the environment
+ vreps_env = reverse vreps_p ++ reverse vreps_np
+
+ -- new env and depth
+ vreps_env_tszsw = map (taggedSizeW.snd) vreps_env
+ p' = addListToFM p (zip (map fst vreps_env)
+ (mkStackOffsets d vreps_env_tszsw))
+ d' = d + sum vreps_env_tszsw
+
+ -- code to unpack the ptrs
+ ptrs_szw = sum (map (untaggedSizeW.snd) vreps_p)
+ code_p | null vreps_p = nilOL
+ | otherwise = unitOL (UNPACK ptrs_szw)
+
+ -- code to unpack the nonptrs
+ vreps_env_uszw = sum (map (untaggedSizeW.snd) vreps_env)
+ code_np = do_nptrs vreps_env_uszw ptrs_szw (reverse (map snd vreps_np))
+ do_nptrs off_h off_s [] = nilOL
+ do_nptrs off_h off_s (npr:nprs)
+ = case npr of
+ IntRep -> approved ; FloatRep -> approved
+ DoubleRep -> approved ; AddrRep -> approved
+ CharRep -> approved
+ _ -> pprPanic "ByteCodeGen.mkUnpackCode" (ppr npr)
+ where
+ approved = UPK_TAG usizeW (off_h-usizeW) off_s `consOL` theRest
+ theRest = do_nptrs (off_h-usizeW) (off_s + tsizeW) nprs
+ usizeW = untaggedSizeW npr
+ tsizeW = taggedSizeW npr
-- Push an atom onto the stack, returning suitable code & number of
--- stack words used.
-pushAtom d p (AnnVar v)
- = case lookupBCEnv_maybe p v of
- Just offset -> (PUSH_L sz offset, sz)
- Nothing -> ASSERT(sz == 1) (PUSH_G nm, 1)
- where
- nm = getName v
- sz = idSizeW v
+-- stack words used. Pushes it either tagged or untagged, since
+-- pushAtom is used to set up the stack prior to copying into the
+-- heap for both APs (requiring tags) and constructors (which don't).
+--
+-- NB this means NO GC between pushing atoms for a constructor and
+-- copying them into the heap. It probably also means that
+-- tail calls MUST be of the form atom{atom ... atom} since if the
+-- expression head was allowed to be arbitrary, there could be GC
+-- in between pushing the arg atoms and completing the head.
+-- (not sure; perhaps the allocate/doYouWantToGC interface means this
+-- isn't a problem; but only if arbitrary graph construction for the
+-- head doesn't leave this BCO, since GC might happen at the start of
+-- each BCO (we consult doYouWantToGC there).
+--
+-- Blargh. JRS 001206
+--
+-- NB (further) that the env p must map each variable to the highest-
+-- numbered stack slot for it. For example, if the stack has depth 4
+-- and we tagged-ly push (v :: Int#) on it, the value will be in stack[4],
+-- the tag in stack[5], the stack will have depth 6, and p must map v to
+-- 5 and not to 4. Stack locations are numbered from zero, so a depth
+-- 6 stack has valid words 0 .. 5.
+
+pushAtom :: Bool -> Int -> BCEnv -> AnnExpr' Id VarSet -> (BCInstrList, Int)
+pushAtom tagged d p (AnnVar v)
+
+ | idPrimRep v == VoidRep
+ = ASSERT(tagged)
+ (unitOL (PUSH_TAG 0), 1)
+
+ | Just primop <- isPrimOpId_maybe v
+ = case primop of
+ CCallOp _ -> panic "pushAtom: byte code generator can't handle CCalls"
+ other -> (unitOL (PUSH_G (Right primop)), 1)
+
+ | otherwise
+ = let str = "\npushAtom " ++ showSDocDebug (ppr v)
+ ++ " :: " ++ showSDocDebug (pprType (idType v))
+ ++ ", depth = " ++ show d
+ ++ ", tagged = " ++ show tagged ++ ", env =\n" ++
+ showSDocDebug (ppBCEnv p)
+ ++ " --> words: " ++ show (snd result) ++ "\n" ++
+ showSDoc (nest 4 (vcat (map ppr (fromOL (fst result)))))
+ ++ "\nendPushAtom " ++ showSDocDebug (ppr v)
+ where
+ cmp_snd x y = compare (snd x) (snd y)
+ str' = if str == str then str else str
+
+ result
+ = case lookupBCEnv_maybe p v of
+ Just d_v -> (toOL (nOfThem nwords (PUSH_L (d-d_v+sz_t-2))), nwords)
+ Nothing -> ASSERT(sz_t == 1) (unitOL (PUSH_G (Left nm)), nwords)
+
+ nm = case isDataConId_maybe v of
+ Just c -> getName c
+ Nothing -> getName v
+
+ sz_t = taggedIdSizeW v
+ sz_u = untaggedIdSizeW v
+ nwords = if tagged then sz_t else sz_u
+ in
+ --trace str'
+ result
+
+pushAtom True d p (AnnLit lit)
+ = let (ubx_code, ubx_size) = pushAtom False d p (AnnLit lit)
+ in (ubx_code `snocOL` PUSH_TAG ubx_size, 1 + ubx_size)
-pushAtom d p (AnnLit lit)
+pushAtom False d p (AnnLit lit)
= case lit of
- MachInt i -> (PUSH_I i, 2)
+ MachWord w -> code WordRep
+ MachInt i -> code IntRep
+ MachFloat r -> code FloatRep
+ MachDouble r -> code DoubleRep
+ MachChar c -> code CharRep
+ MachStr s -> pushStr s
+ where
+ code rep
+ = let size_host_words = untaggedSizeW rep
+ in (unitOL (PUSH_UBX lit size_host_words), size_host_words)
+
+ pushStr s
+ = let mallocvilleAddr
+ = case s of
+ CharStr s i -> A# s
+
+ FastString _ l ba ->
+ -- sigh, a string in the heap is no good to us.
+ -- We need a static C pointer, since the type of
+ -- a string literal is Addr#. So, copy the string
+ -- into C land and introduce a memory leak
+ -- at the same time.
+ let n = I# l
+ -- CAREFUL! Chars are 32 bits in ghc 4.09+
+ in unsafePerformIO (
+ do (Ptr a#) <- mallocBytes (n+1)
+ strncpy (Ptr a#) ba (fromIntegral n)
+ writeCharOffAddr (A# a#) n '\0'
+ return (A# a#)
+ )
+ _ -> panic "StgInterp.lit2expr: unhandled string constant type"
+
+ addrLit
+ = MachInt (toInteger (addrToInt mallocvilleAddr))
+ in
+ -- Get the addr on the stack, untaggedly
+ (unitOL (PUSH_UBX addrLit 1), 1)
+
+
+
+
+
+pushAtom tagged d p (AnnApp f (_, AnnType _))
+ = pushAtom tagged d p (snd f)
+
+pushAtom tagged d p (AnnNote note e)
+ = pushAtom tagged d p (snd e)
+
+pushAtom tagged d p (AnnLam x e)
+ | isTyVar x
+ = pushAtom tagged d p (snd e)
+
+pushAtom tagged d p other
+ = pprPanic "ByteCodeGen.pushAtom"
+ (pprCoreExpr (deAnnotate (undefined, other)))
+
+foreign import "strncpy" strncpy :: Ptr a -> ByteArray# -> CInt -> IO ()
+
+
+-- Given a bunch of alts code and their discrs, do the donkey work
+-- of making a multiway branch using a switch tree.
+-- What a load of hassle!
+mkMultiBranch :: Maybe Int -- # datacons in tycon, if alg alt
+ -- a hint; generates better code
+ -- Nothing is always safe
+ -> [(Discr, BCInstrList)]
+ -> BcM BCInstrList
+mkMultiBranch maybe_ncons raw_ways
+ = let d_way = filter (isNoDiscr.fst) raw_ways
+ notd_ways = naturalMergeSortLe
+ (\w1 w2 -> leAlt (fst w1) (fst w2))
+ (filter (not.isNoDiscr.fst) raw_ways)
+
+ mkTree :: [(Discr, BCInstrList)] -> Discr -> Discr -> BcM BCInstrList
+ mkTree [] range_lo range_hi = returnBc the_default
+
+ mkTree [val] range_lo range_hi
+ | range_lo `eqAlt` range_hi
+ = returnBc (snd val)
+ | otherwise
+ = getLabelBc `thenBc` \ label_neq ->
+ returnBc (mkTestEQ (fst val) label_neq
+ `consOL` (snd val
+ `appOL` unitOL (LABEL label_neq)
+ `appOL` the_default))
+
+ mkTree vals range_lo range_hi
+ = let n = length vals `div` 2
+ vals_lo = take n vals
+ vals_hi = drop n vals
+ v_mid = fst (head vals_hi)
+ in
+ getLabelBc `thenBc` \ label_geq ->
+ mkTree vals_lo range_lo (dec v_mid) `thenBc` \ code_lo ->
+ mkTree vals_hi v_mid range_hi `thenBc` \ code_hi ->
+ returnBc (mkTestLT v_mid label_geq
+ `consOL` (code_lo
+ `appOL` unitOL (LABEL label_geq)
+ `appOL` code_hi))
+
+ the_default
+ = case d_way of [] -> unitOL CASEFAIL
+ [(_, def)] -> def
+
+ -- None of these will be needed if there are no non-default alts
+ (mkTestLT, mkTestEQ, init_lo, init_hi)
+ | null notd_ways
+ = panic "mkMultiBranch: awesome foursome"
+ | otherwise
+ = case fst (head notd_ways) of {
+ DiscrI _ -> ( \(DiscrI i) fail_label -> TESTLT_I i fail_label,
+ \(DiscrI i) fail_label -> TESTEQ_I i fail_label,
+ DiscrI minBound,
+ DiscrI maxBound );
+ DiscrF _ -> ( \(DiscrF f) fail_label -> TESTLT_F f fail_label,
+ \(DiscrF f) fail_label -> TESTEQ_F f fail_label,
+ DiscrF minF,
+ DiscrF maxF );
+ DiscrD _ -> ( \(DiscrD d) fail_label -> TESTLT_D d fail_label,
+ \(DiscrD d) fail_label -> TESTEQ_D d fail_label,
+ DiscrD minD,
+ DiscrD maxD );
+ DiscrP _ -> ( \(DiscrP i) fail_label -> TESTLT_P i fail_label,
+ \(DiscrP i) fail_label -> TESTEQ_P i fail_label,
+ DiscrP algMinBound,
+ DiscrP algMaxBound )
+ }
+
+ (algMinBound, algMaxBound)
+ = case maybe_ncons of
+ Just n -> (0, n - 1)
+ Nothing -> (minBound, maxBound)
+
+ (DiscrI i1) `eqAlt` (DiscrI i2) = i1 == i2
+ (DiscrF f1) `eqAlt` (DiscrF f2) = f1 == f2
+ (DiscrD d1) `eqAlt` (DiscrD d2) = d1 == d2
+ (DiscrP i1) `eqAlt` (DiscrP i2) = i1 == i2
+ NoDiscr `eqAlt` NoDiscr = True
+ _ `eqAlt` _ = False
+
+ (DiscrI i1) `leAlt` (DiscrI i2) = i1 <= i2
+ (DiscrF f1) `leAlt` (DiscrF f2) = f1 <= f2
+ (DiscrD d1) `leAlt` (DiscrD d2) = d1 <= d2
+ (DiscrP i1) `leAlt` (DiscrP i2) = i1 <= i2
+ NoDiscr `leAlt` NoDiscr = True
+ _ `leAlt` _ = False
+
+ isNoDiscr NoDiscr = True
+ isNoDiscr _ = False
+
+ dec (DiscrI i) = DiscrI (i-1)
+ dec (DiscrP i) = DiscrP (i-1)
+ dec other = other -- not really right, but if you
+ -- do cases on floating values, you'll get what you deserve
+
+ -- same snotty comment applies to the following
+ minF, maxF :: Float
+ minD, maxD :: Double
+ minF = -1.0e37
+ maxF = 1.0e37
+ minD = -1.0e308
+ maxD = 1.0e308
+ in
+ mkTree notd_ways init_lo init_hi
+
\end{code}
-The bytecode generator's monad.
+%************************************************************************
+%* *
+\subsection{Supporting junk for the compilation schemes}
+%* *
+%************************************************************************
\begin{code}
-type BcM_State = [BCO Name] -- accumulates completed BCOs
-type BcM result = BcM_State -> (result, BcM_State)
+-- Describes case alts
+data Discr
+ = DiscrI Int
+ | DiscrF Float
+ | DiscrD Double
+ | DiscrP Int
+ | NoDiscr
-mkBcM_State :: [BCO Name] -> BcM_State
-mkBcM_State = id
+instance Outputable Discr where
+ ppr (DiscrI i) = int i
+ ppr (DiscrF f) = text (show f)
+ ppr (DiscrD d) = text (show d)
+ ppr (DiscrP i) = int i
+ ppr NoDiscr = text "DEF"
-initBc :: BcM_State -> BcM a -> (a, BcM_State)
-initBc init_st m = case m init_st of { (r,st) -> (r,st) }
+
+-- Find things in the BCEnv (the what's-on-the-stack-env)
+-- See comment preceding pushAtom for precise meaning of env contents
+--lookupBCEnv :: BCEnv -> Id -> Int
+--lookupBCEnv env nm
+-- = case lookupFM env nm of
+-- Nothing -> pprPanic "lookupBCEnv"
+-- (ppr nm $$ char ' ' $$ vcat (map ppr (fmToList env)))
+-- Just xx -> xx
+
+lookupBCEnv_maybe :: BCEnv -> Id -> Maybe Int
+lookupBCEnv_maybe = lookupFM
+
+
+-- When I push one of these on the stack, how much does Sp move by?
+taggedSizeW :: PrimRep -> Int
+taggedSizeW pr
+ | isFollowableRep pr = 1
+ | otherwise = 1{-the tag-} + getPrimRepSize pr
+
+
+-- The plain size of something, without tag.
+untaggedSizeW :: PrimRep -> Int
+untaggedSizeW pr
+ | isFollowableRep pr = 1
+ | otherwise = getPrimRepSize pr
+
+
+taggedIdSizeW, untaggedIdSizeW :: Id -> Int
+taggedIdSizeW = taggedSizeW . typePrimRep . idType
+untaggedIdSizeW = untaggedSizeW . typePrimRep . idType
+
+\end{code}
+
+%************************************************************************
+%* *
+\subsection{The bytecode generator's monad}
+%* *
+%************************************************************************
+
+\begin{code}
+data BcM_State
+ = BcM_State { bcos :: [ProtoBCO Name], -- accumulates completed BCOs
+ nextlabel :: Int } -- for generating local labels
+
+type BcM result = BcM_State -> (result, BcM_State)
+
+runBc :: BcM_State -> BcM () -> BcM_State
+runBc init_st m = case m init_st of { (r,st) -> st }
thenBc :: BcM a -> (a -> BcM b) -> BcM b
thenBc expr cont st
mapBc f xs `thenBc` \ rs ->
returnBc (r:rs)
-emitBc :: BCO Name -> BcM ()
-emitBc bco bcos
- = ((), bcos)
+emitBc :: ProtoBCO Name -> BcM ()
+emitBc bco st
+ = ((), st{bcos = bco : bcos st})
+
+getLabelBc :: BcM Int
+getLabelBc st
+ = (nextlabel st, st{nextlabel = 1 + nextlabel st})
+
\end{code}
projects / ghc-hetmet.git / blobdiff