X-Git-Url: http://git.megacz.com/?a=blobdiff_plain;f=ghc%2Fcompiler%2FbasicTypes%2FIdInfo.lhs;h=0db72f1bd6c4cc5a4290d214b382a31badd0f94f;hb=9adbdb312507dcc7d5777e36376535918549103b;hp=6eebe45348f4859f85e97daf0f624924a8fdb3e3;hpb=f9120c200bcf613b58d742802172fb4c08171f0d;p=ghc-hetmet.git diff --git a/ghc/compiler/basicTypes/IdInfo.lhs b/ghc/compiler/basicTypes/IdInfo.lhs index 6eebe45..0db72f1 100644 --- a/ghc/compiler/basicTypes/IdInfo.lhs +++ b/ghc/compiler/basicTypes/IdInfo.lhs @@ -1,5 +1,5 @@ % -% (c) The GRASP/AQUA Project, Glasgow University, 1993-1996 +% (c) The GRASP/AQUA Project, Glasgow University, 1993-1998 % \section[IdInfo]{@IdInfos@: Non-essential information about @Ids@} @@ -7,317 +7,258 @@ Haskell. [WDP 94/11]) \begin{code} -#include "HsVersions.h" - module IdInfo ( - IdInfo, -- abstract - noIdInfo, - boringIdInfo, - ppIdInfo, - applySubstToIdInfo, apply_to_IdInfo, -- not for general use, please - - OptIdInfo(..), -- class; for convenience only - -- all the *Infos herein are instances of it - - -- component "id infos"; also abstract: - SrcLoc, - getSrcLocIdInfo, - - ArityInfo, - mkArityInfo, unknownArity, arityMaybe, - - DemandInfo, - mkDemandInfo, - willBeDemanded, - - MatchEnv, -- the SpecEnv - StrictnessInfo(..), -- non-abstract - Demand(..), -- non-abstract - - wwLazy, wwStrict, wwUnpack, wwPrim, wwEnum, - indicatesWorker, nonAbsentArgs, - mkStrictnessInfo, mkBottomStrictnessInfo, - getWrapperArgTypeCategories, - getWorkerId, - workerExists, - bottomIsGuaranteed, - - mkUnfolding, - noInfo_UF, getInfo_UF, addInfo_UF, -- to avoid instance virus - - UpdateInfo, - mkUpdateInfo, - UpdateSpec(..), - updateInfoMaybe, - - DeforestInfo(..), - - ArgUsageInfo, - ArgUsage(..), - ArgUsageType(..), - mkArgUsageInfo, - getArgUsage, - - FBTypeInfo, - FBType(..), - FBConsum(..), - FBProd(..), - mkFBTypeInfo, - getFBType + IdInfo, -- Abstract - ) where + vanillaIdInfo, mkIdInfo, seqIdInfo, megaSeqIdInfo, -import Ubiq - -import IdLoop -- IdInfo is a dependency-loop ranch, and - -- we break those loops by using IdLoop and - -- *not* importing much of anything else, - -- except from the very general "utils". - -import CmdLineOpts ( opt_OmitInterfacePragmas ) -import Maybes ( firstJust ) -import MatchEnv ( nullMEnv, isEmptyMEnv, mEnvToList ) -import Outputable ( ifPprInterface, Outputable(..){-instances-} ) -import PprStyle ( PprStyle(..) ) -import Pretty -import SrcLoc ( mkUnknownSrcLoc ) -import Type ( eqSimpleTy ) -import Util ( mapAccumL, panic, assertPanic, pprPanic ) - -applySubstToTy = panic "IdInfo.applySubstToTy" -splitTypeWithDictsAsArgs = panic "IdInfo.splitTypeWithDictsAsArgs" -showTypeCategory = panic "IdInfo.showTypeCategory" -mkFormSummary = panic "IdInfo.mkFormSummary" -occurAnalyseGlobalExpr = panic "IdInfo.occurAnalyseGlobalExpr" -isWrapperFor = panic "IdInfo.isWrapperFor" -pprCoreUnfolding = panic "IdInfo.pprCoreUnfolding" -\end{code} + -- Zapping + zapFragileInfo, zapLamInfo, zapSpecPragInfo, shortableIdInfo, copyIdInfo, -An @IdInfo@ gives {\em optional} information about an @Id@. If -present it never lies, but it may not be present, in which case there -is always a conservative assumption which can be made. + -- Flavour + IdFlavour(..), flavourInfo, + setNoDiscardInfo, + ppFlavourInfo, -Two @Id@s may have different info even though they have the same -@Unique@ (and are hence the same @Id@); for example, one might lack -the properties attached to the other. + -- Arity + ArityInfo(..), + exactArity, atLeastArity, unknownArity, hasArity, + arityInfo, setArityInfo, ppArityInfo, arityLowerBound, -The @IdInfo@ gives information about the value, or definition, of the -@Id@. It does {\em not} contain information about the @Id@'s usage -(except for @DemandInfo@? ToDo). + -- Strictness; imported from Demand + StrictnessInfo(..), + mkStrictnessInfo, noStrictnessInfo, + ppStrictnessInfo,isBottomingStrictness, -\begin{code} -data IdInfo - = IdInfo - ArityInfo -- Its arity + strictnessInfo, setStrictnessInfo, - DemandInfo -- Whether or not it is definitely - -- demanded + -- Worker + WorkerInfo(..), workerExists, wrapperArity, workerId, + workerInfo, setWorkerInfo, ppWorkerInfo, - (MatchEnv [Type] CoreExpr) - -- Specialisations of this function which exist - -- This corresponds to a SpecEnv which we do - -- not import directly to avoid loop + -- Unfolding + unfoldingInfo, setUnfoldingInfo, - StrictnessInfo -- Strictness properties, notably - -- how to conjure up "worker" functions + -- DemandInfo + demandInfo, setDemandInfo, - UnfoldingDetails -- Its unfolding; for locally-defined - -- things, this can *only* be NoUnfoldingDetails + -- Inline prags + InlinePragInfo(..), + inlinePragInfo, setInlinePragInfo, pprInlinePragInfo, + isNeverInlinePrag, neverInlinePrag, - UpdateInfo -- Which args should be updated + -- Occurrence info + OccInfo(..), isFragileOcc, isDeadOcc, isLoopBreaker, + InsideLam, OneBranch, insideLam, notInsideLam, oneBranch, notOneBranch, + occInfo, setOccInfo, - DeforestInfo -- Whether its definition should be - -- unfolded during deforestation + -- Specialisation + specInfo, setSpecInfo, - ArgUsageInfo -- how this Id uses its arguments + -- CAF info + CafInfo(..), cafInfo, setCafInfo, ppCafInfo, - FBTypeInfo -- the Foldr/Build W/W property of this function. + -- Constructed Product Result Info + CprInfo(..), cprInfo, setCprInfo, ppCprInfo, noCprInfo, - SrcLoc -- Source location of definition + -- Lambda-bound variable info + LBVarInfo(..), lbvarInfo, setLBVarInfo, noLBVarInfo + ) where - -- ToDo: SrcLoc is in FullNames too (could rm?) but it - -- is needed here too for things like ConstMethodIds and the - -- like, which don't have full-names of their own Mind you, - -- perhaps the Name for a constant method could give the - -- class/type involved? -\end{code} +#include "HsVersions.h" -\begin{code} -noIdInfo = IdInfo noInfo noInfo noInfo noInfo noInfo_UF - noInfo noInfo noInfo noInfo mkUnknownSrcLoc - --- "boring" means: nothing to put in interface -boringIdInfo (IdInfo UnknownArity - UnknownDemand - specenv - strictness - unfolding - NoUpdateInfo - Don'tDeforest - _ {- arg_usage: currently no interface effect -} - _ {- no f/b w/w -} - _ {- src_loc: no effect on interfaces-} - ) - | null (mEnvToList specenv) - && boring_strictness strictness - && boring_unfolding unfolding - = True - where - boring_strictness NoStrictnessInfo = True - boring_strictness BottomGuaranteed = False - boring_strictness (StrictnessInfo wrap_args _) = all_present_WwLazies wrap_args - boring_unfolding NoUnfoldingDetails = True - boring_unfolding _ = False +import CoreSyn +import PrimOp ( PrimOp ) +import Var ( Id ) +import BasicTypes ( OccInfo(..), isFragileOcc, isDeadOcc, seqOccInfo, isLoopBreaker, + InsideLam, insideLam, notInsideLam, + OneBranch, oneBranch, notOneBranch, + Arity + ) +import DataCon ( DataCon ) +import FieldLabel ( FieldLabel ) +import Demand -- Lots of stuff +import Outputable +import Maybe ( isJust ) + +infixl 1 `setDemandInfo`, + `setStrictnessInfo`, + `setSpecInfo`, + `setArityInfo`, + `setInlinePragInfo`, + `setUnfoldingInfo`, + `setCprInfo`, + `setWorkerInfo`, + `setCafInfo`, + `setOccInfo` + -- infixl so you can say (id `set` a `set` b) +\end{code} -boringIdInfo _ = False +An @IdInfo@ gives {\em optional} information about an @Id@. If +present it never lies, but it may not be present, in which case there +is always a conservative assumption which can be made. -pp_NONE = ppPStr SLIT("_N_") -\end{code} + There is one exception: the 'flavour' is *not* optional. + You must not discard it. + It used to be in Var.lhs, but that seems unclean. -Simply turgid. But BE CAREFUL: don't @apply_to_Id@ if that @Id@ -will in turn @apply_to_IdInfo@ of the self-same @IdInfo@. (A very -nasty loop, friends...) -\begin{code} -apply_to_IdInfo ty_fn idinfo@(IdInfo arity demand spec strictness unfold - update deforest arg_usage fb_ww srcloc) - | isEmptyMEnv spec - = idinfo - | otherwise - = panic "IdInfo:apply_to_IdInfo" -{- LATER: - let - new_spec = apply_spec spec - - -- NOT a good idea: - -- apply_strict strictness `thenLft` \ new_strict -> - -- apply_wrap wrap `thenLft` \ new_wrap -> - in - IdInfo arity demand new_spec strictness unfold - update deforest arg_usage fb_ww srcloc - where - apply_spec (SpecEnv is) - = SpecEnv (map do_one is) - where - do_one (SpecInfo ty_maybes ds spec_id) - = --apply_to_Id ty_fn spec_id `thenLft` \ new_spec_id -> - SpecInfo (map apply_to_maybe ty_maybes) ds spec_id - where - apply_to_maybe Nothing = Nothing - apply_to_maybe (Just ty) = Just (ty_fn ty) --} - -{- NOT a good idea; - apply_strict info@NoStrictnessInfo = returnLft info - apply_strict BottomGuaranteed = ??? - apply_strict (StrictnessInfo wrap_arg_info id_maybe) - = (case id_maybe of - Nothing -> returnLft Nothing - Just xx -> applySubstToId subst xx `thenLft` \ new_xx -> - returnLft (Just new_xx) - ) `thenLft` \ new_id_maybe -> - returnLft (StrictnessInfo wrap_arg_info new_id_maybe) --} -\end{code} +Two @Id@s may have different info even though they have the same +@Unique@ (and are hence the same @Id@); for example, one might lack +the properties attached to the other. -Variant of the same thing for the typechecker. -\begin{code} -applySubstToIdInfo s0 (IdInfo arity demand spec strictness unfold - update deforest arg_usage fb_ww srcloc) - = panic "IdInfo:applySubstToIdInfo" -{- LATER: - case (apply_spec s0 spec) of { (s1, new_spec) -> - (s1, IdInfo arity demand new_spec strictness unfold update deforest arg_usage fb_ww srcloc) } - where - apply_spec s0 (SpecEnv is) - = case (mapAccumL do_one s0 is) of { (s1, new_is) -> - (s1, SpecEnv new_is) } - where - do_one s0 (SpecInfo ty_maybes ds spec_id) - = case (mapAccumL apply_to_maybe s0 ty_maybes) of { (s1, new_maybes) -> - (s1, SpecInfo new_maybes ds spec_id) } - where - apply_to_maybe s0 Nothing = (s0, Nothing) - apply_to_maybe s0 (Just ty) - = case (applySubstToTy s0 ty) of { (s1, new_ty) -> - (s1, Just new_ty) } --} -\end{code} +The @IdInfo@ gives information about the value, or definition, of the +@Id@. It does {\em not} contain information about the @Id@'s usage +(except for @DemandInfo@? ToDo). (@lbvarInfo@ is also a marginal +case. KSW 1999-04). \begin{code} -ppIdInfo :: PprStyle - -> Id -- The Id for which we're printing this IdInfo - -> Bool -- True <=> print specialisations, please - -> (Id -> Id) -- to look up "better Ids" w/ better IdInfos; - -> IdEnv UnfoldingDetails - -- inlining info for top-level fns in this module - -> IdInfo -- see MkIface notes - -> Pretty - -ppIdInfo sty for_this_id specs_please better_id_fn inline_env - i@(IdInfo arity demand specenv strictness unfold update deforest arg_usage fbtype srcloc) - | boringIdInfo i - = ppPStr SLIT("_NI_") +data IdInfo + = IdInfo { + flavourInfo :: IdFlavour, -- NOT OPTIONAL + arityInfo :: ArityInfo, -- Its arity + demandInfo :: Demand, -- Whether or not it is definitely demanded + specInfo :: CoreRules, -- Specialisations of this function which exist + strictnessInfo :: StrictnessInfo, -- Strictness properties + workerInfo :: WorkerInfo, -- Pointer to Worker Function + unfoldingInfo :: Unfolding, -- Its unfolding + cafInfo :: CafInfo, + cprInfo :: CprInfo, -- Function always constructs a product result + lbvarInfo :: LBVarInfo, -- Info about a lambda-bound variable + inlinePragInfo :: InlinePragInfo, -- Inline pragma + occInfo :: OccInfo -- How it occurs + } + +seqIdInfo :: IdInfo -> () +seqIdInfo (IdInfo {}) = () + +megaSeqIdInfo :: IdInfo -> () +megaSeqIdInfo info + = seqFlavour (flavourInfo info) `seq` + seqArity (arityInfo info) `seq` + seqDemand (demandInfo info) `seq` + seqRules (specInfo info) `seq` + seqStrictnessInfo (strictnessInfo info) `seq` + seqWorker (workerInfo info) `seq` + +-- seqUnfolding (unfoldingInfo info) `seq` +-- Omitting this improves runtimes a little, presumably because +-- some unfoldings are not calculated at all + + seqCaf (cafInfo info) `seq` + seqCpr (cprInfo info) `seq` + seqLBVar (lbvarInfo info) `seq` + seqOccInfo (occInfo info) +\end{code} + +Setters + +\begin{code} +setWorkerInfo info wk = wk `seq` info { workerInfo = wk } +setSpecInfo info sp = PSEQ sp (info { specInfo = sp }) +setInlinePragInfo info pr = pr `seq` info { inlinePragInfo = pr } +setOccInfo info oc = oc `seq` info { occInfo = oc } +setStrictnessInfo info st = st `seq` info { strictnessInfo = st } + -- Try to avoid spack leaks by seq'ing + +setUnfoldingInfo info uf + | isEvaldUnfolding uf && isStrict (demandInfo info) + -- If the unfolding is a value, the demand info may + -- go pear-shaped, so we nuke it. Example: + -- let x = (a,b) in + -- case x of (p,q) -> h p q x + -- Here x is certainly demanded. But after we've nuked + -- the case, we'll get just + -- let x = (a,b) in h a b x + -- and now x is not demanded (I'm assuming h is lazy) + -- This really happens. The solution here is a bit ad hoc... + = info { unfoldingInfo = uf, demandInfo = wwLazy } | otherwise - = let - stuff = ppCat [ - -- order is important!: - ppInfo sty better_id_fn arity, - ppInfo sty better_id_fn update, - ppInfo sty better_id_fn deforest, - - pp_strictness sty (Just for_this_id) - better_id_fn inline_env strictness, - - if bottomIsGuaranteed strictness - then pp_NONE - else pp_unfolding sty for_this_id inline_env unfold, - - if specs_please - then ppSpecs sty (not (isDataCon for_this_id)) - better_id_fn inline_env (mEnvToList specenv) - else pp_NONE, - - -- DemandInfo needn't be printed since it has no effect on interfaces - ppInfo sty better_id_fn demand, - ppInfo sty better_id_fn fbtype - ] - in - case sty of - PprInterface -> if opt_OmitInterfacePragmas - then ppNil - else stuff - _ -> stuff + -- We do *not* seq on the unfolding info, For some reason, doing so + -- actually increases residency significantly. + = info { unfoldingInfo = uf } + +setDemandInfo info dd = info { demandInfo = dd } +setArityInfo info ar = info { arityInfo = ar } +setCafInfo info cf = info { cafInfo = cf } +setCprInfo info cp = info { cprInfo = cp } +setLBVarInfo info lb = info { lbvarInfo = lb } + +setNoDiscardInfo info = case flavourInfo info of + VanillaId -> info { flavourInfo = NoDiscardId } + other -> info +zapSpecPragInfo info = case flavourInfo info of + SpecPragmaId -> info { flavourInfo = VanillaId } + other -> info \end{code} -%************************************************************************ -%* * -\subsection[OptIdInfo-class]{The @OptIdInfo@ class (keeps things tidier)} -%* * -%************************************************************************ \begin{code} -class OptIdInfo a where - noInfo :: a - getInfo :: IdInfo -> a - addInfo :: IdInfo -> a -> IdInfo - -- By default, "addInfo" will not overwrite - -- "info" with "non-info"; look at any instance - -- to see an example. - ppInfo :: PprStyle -> (Id -> Id) -> a -> Pretty +vanillaIdInfo :: IdInfo +vanillaIdInfo = mkIdInfo VanillaId + +mkIdInfo :: IdFlavour -> IdInfo +mkIdInfo flv = IdInfo { + flavourInfo = flv, + arityInfo = UnknownArity, + demandInfo = wwLazy, + specInfo = emptyCoreRules, + workerInfo = NoWorker, + strictnessInfo = NoStrictnessInfo, + unfoldingInfo = noUnfolding, + cafInfo = MayHaveCafRefs, + cprInfo = NoCPRInfo, + lbvarInfo = NoLBVarInfo, + inlinePragInfo = NoInlinePragInfo, + occInfo = NoOccInfo + } \end{code} + %************************************************************************ %* * -\subsection[srcloc-IdInfo]{Source-location info in an @IdInfo@} +\subsection{Flavour} %* * %************************************************************************ -Not used much, but... \begin{code} -getSrcLocIdInfo (IdInfo _ _ _ _ _ _ _ _ _ src_loc) = src_loc -\end{code} +data IdFlavour + = VanillaId -- Most Ids are like this + | DataConId DataCon -- The Id for a data constructor *worker* + | DataConWrapId DataCon -- The Id for a data constructor *wrapper* + -- [the only reasons we need to know is so that + -- a) we can suppress printing a definition in the interface file + -- b) when typechecking a pattern we can get from the + -- Id back to the data con] + | PrimOpId PrimOp -- The Id for a primitive operator + | RecordSelId FieldLabel -- The Id for a record selector + | SpecPragmaId -- Don't discard these + | NoDiscardId -- Don't discard these either + +ppFlavourInfo :: IdFlavour -> SDoc +ppFlavourInfo VanillaId = empty +ppFlavourInfo (DataConId _) = ptext SLIT("[DataCon]") +ppFlavourInfo (DataConWrapId _) = ptext SLIT("[DataConWrapper]") +ppFlavourInfo (PrimOpId _) = ptext SLIT("[PrimOp]") +ppFlavourInfo (RecordSelId _) = ptext SLIT("[RecSel]") +ppFlavourInfo SpecPragmaId = ptext SLIT("[SpecPrag]") +ppFlavourInfo NoDiscardId = ptext SLIT("[NoDiscard]") + +seqFlavour :: IdFlavour -> () +seqFlavour f = f `seq` () +\end{code} + +The @SpecPragmaId@ exists only to make Ids that are +on the *LHS* of bindings created by SPECIALISE pragmas; +eg: s = f Int d +The SpecPragmaId is never itself mentioned; it +exists solely so that the specialiser will find +the call to f, and make specialised version of it. +The SpecPragmaId binding is discarded by the specialiser +when it gathers up overloaded calls. +Meanwhile, it is not discarded as dead code. + %************************************************************************ %* * @@ -325,632 +266,401 @@ getSrcLocIdInfo (IdInfo _ _ _ _ _ _ _ _ _ src_loc) = src_loc %* * %************************************************************************ -\begin{code} -data ArityInfo - = UnknownArity -- no idea - | ArityExactly Int -- arity is exactly this -\end{code} - -\begin{code} -mkArityInfo = ArityExactly -unknownArity = UnknownArity - -arityMaybe :: ArityInfo -> Maybe Int - -arityMaybe UnknownArity = Nothing -arityMaybe (ArityExactly i) = Just i -\end{code} +For locally-defined Ids, the code generator maintains its own notion +of their arities; so it should not be asking... (but other things +besides the code-generator need arity info!) \begin{code} -instance OptIdInfo ArityInfo where - noInfo = UnknownArity - - getInfo (IdInfo arity _ _ _ _ _ _ _ _ _) = arity +data ArityInfo + = UnknownArity -- No idea - addInfo id_info UnknownArity = id_info - addInfo (IdInfo _ a c d e f g h i j) arity = IdInfo arity a c d e f g h i j + | ArityExactly Arity -- Arity is exactly this. We use this when importing a + -- function; it's already been compiled and we know its + -- arity for sure. - ppInfo sty _ UnknownArity = ifPprInterface sty pp_NONE - ppInfo sty _ (ArityExactly arity) = ppCat [ppPStr SLIT("_A_"), ppInt arity] -\end{code} + | ArityAtLeast Arity -- A partial application of this Id to up to n-1 value arguments + -- does essentially no work. That is not necessarily the + -- same as saying that it has n leading lambdas, because coerces + -- may get in the way. -%************************************************************************ -%* * -\subsection[demand-IdInfo]{Demand info about an @Id@} -%* * -%************************************************************************ + -- functions in the module being compiled. Their arity + -- might increase later in the compilation process, if + -- an extra lambda floats up to the binding site. + deriving( Eq ) -Whether a value is certain to be demanded or not. (This is the -information that is computed by the ``front-end'' of the strictness -analyser.) +seqArity :: ArityInfo -> () +seqArity a = arityLowerBound a `seq` () -This information is only used within a module, it is not exported -(obviously). +exactArity = ArityExactly +atLeastArity = ArityAtLeast +unknownArity = UnknownArity -\begin{code} -data DemandInfo - = UnknownDemand - | DemandedAsPer Demand -\end{code} +arityLowerBound :: ArityInfo -> Arity +arityLowerBound UnknownArity = 0 +arityLowerBound (ArityAtLeast n) = n +arityLowerBound (ArityExactly n) = n -\begin{code} -mkDemandInfo :: Demand -> DemandInfo -mkDemandInfo demand = DemandedAsPer demand +hasArity :: ArityInfo -> Bool +hasArity UnknownArity = False +hasArity other = True -willBeDemanded :: DemandInfo -> Bool -willBeDemanded (DemandedAsPer demand) = isStrict demand -willBeDemanded _ = False -\end{code} - -\begin{code} -instance OptIdInfo DemandInfo where - noInfo = UnknownDemand - - getInfo (IdInfo _ demand _ _ _ _ _ _ _ _) = demand - -{- DELETED! If this line is in, there is no way to - nuke a DemandInfo, and we have to be able to do that - when floating let-bindings around - addInfo id_info UnknownDemand = id_info --} - addInfo (IdInfo a _ c d e f g h i j) demand = IdInfo a demand c d e f g h i j - - ppInfo PprInterface _ _ = ppNil - ppInfo sty _ UnknownDemand = ppStr "{-# L #-}" - ppInfo sty _ (DemandedAsPer info) - = ppCat [ppStr "{-#", ppStr (showList [info] ""), ppStr "#-}"] +ppArityInfo UnknownArity = empty +ppArityInfo (ArityExactly arity) = hsep [ptext SLIT("__A"), int arity] +ppArityInfo (ArityAtLeast arity) = hsep [ptext SLIT("__AL"), int arity] \end{code} %************************************************************************ %* * -\subsection[specialisation-IdInfo]{Specialisation info about an @Id@} +\subsection{Inline-pragma information} %* * %************************************************************************ -See SpecEnv.lhs - \begin{code} -instance OptIdInfo (MatchEnv [Type] CoreExpr) where - noInfo = nullMEnv +data InlinePragInfo + = NoInlinePragInfo + | IMustNotBeINLINEd Bool -- True <=> came from an INLINE prag, False <=> came from a NOINLINE prag + (Maybe Int) -- Phase number from pragma, if any + deriving( Eq ) + -- The True, Nothing case doesn't need to be recorded + + -- SEE COMMENTS WITH CoreUnfold.blackListed on the + -- exact significance of the IMustNotBeINLINEd pragma - getInfo (IdInfo _ _ spec _ _ _ _ _ _ _) = spec +isNeverInlinePrag :: InlinePragInfo -> Bool +isNeverInlinePrag (IMustNotBeINLINEd _ Nothing) = True +isNeverInlinePrag other = False - addInfo id_info spec | null (mEnvToList spec) = id_info - addInfo (IdInfo a b _ d e f g h i j) spec = IdInfo a b spec d e f g h i j +neverInlinePrag :: InlinePragInfo +neverInlinePrag = IMustNotBeINLINEd True{-should be False? --SDM -} Nothing - ppInfo sty better_id_fn spec - = ppSpecs sty True better_id_fn nullIdEnv (mEnvToList spec) +instance Outputable InlinePragInfo where + -- This is now parsed in interface files + ppr NoInlinePragInfo = empty + ppr other_prag = ptext SLIT("__U") <> pprInlinePragInfo other_prag -ppSpecs sty print_spec_id_info better_id_fn inline_env spec_env - = panic "IdInfo:ppSpecs" +pprInlinePragInfo NoInlinePragInfo = empty +pprInlinePragInfo (IMustNotBeINLINEd True Nothing) = empty +pprInlinePragInfo (IMustNotBeINLINEd True (Just n)) = brackets (int n) +pprInlinePragInfo (IMustNotBeINLINEd False Nothing) = brackets (char '!') +pprInlinePragInfo (IMustNotBeINLINEd False (Just n)) = brackets (char '!' <> int n) + +instance Show InlinePragInfo where + showsPrec p prag = showsPrecSDoc p (ppr prag) \end{code} + %************************************************************************ %* * -\subsection[strictness-IdInfo]{Strictness info about an @Id@} +\subsection[worker-IdInfo]{Worker info about an @Id@} %* * %************************************************************************ -We specify the strictness of a function by giving information about -each of the ``wrapper's'' arguments (see the description about -worker/wrapper-style transformations in the PJ/Launchbury paper on -unboxed types). - -The list of @Demands@ specifies: (a)~the strictness properties -of a function's arguments; (b)~the {\em existence} of a ``worker'' -version of the function; and (c)~the type signature of that worker (if -it exists); i.e. its calling convention. - -\begin{code} -data StrictnessInfo - = NoStrictnessInfo - - | BottomGuaranteed -- This Id guarantees never to return; - -- it is bottom regardless of its arguments. - -- Useful for "error" and other disguised - -- variants thereof. - - | StrictnessInfo [Demand] -- the main stuff; see below. - (Maybe Id) -- worker's Id, if applicable. -\end{code} +If this Id has a worker then we store a reference to it. Worker +functions are generated by the worker/wrapper pass. This uses +information from the strictness and CPR analyses. -This type is also actually used in the strictness analyser: -\begin{code} -data Demand - = WwLazy -- Argument is lazy as far as we know - MaybeAbsent -- (does not imply worker's existence [etc]). - -- If MaybeAbsent == True, then it is - -- *definitely* lazy. (NB: Absence implies - -- a worker...) - - | WwStrict -- Argument is strict but that's all we know - -- (does not imply worker's existence or any - -- calling-convention magic) - - | WwUnpack -- Argument is strict & a single-constructor - [Demand] -- type; its constituent parts (whose StrictInfos - -- are in the list) should be passed - -- as arguments to the worker. - - | WwPrim -- Argument is of primitive type, therefore - -- strict; doesn't imply existence of a worker; - -- argument should be passed as is to worker. - - | WwEnum -- Argument is strict & an enumeration type; - -- an Int# representing the tag (start counting - -- at zero) should be passed to the worker. - deriving (Eq, Ord) - -- we need Eq/Ord to cross-chk update infos in interfaces - -type MaybeAbsent = Bool -- True <=> not even used - --- versions that don't worry about Absence: -wwLazy = WwLazy False -wwStrict = WwStrict -wwUnpack xs = WwUnpack xs -wwPrim = WwPrim -wwEnum = WwEnum -\end{code} +There might not be a worker, even for a strict function, because: +(a) the function might be small enough to inline, so no need + for w/w split +(b) the strictness info might be "SSS" or something, so no w/w split. \begin{code} -mkStrictnessInfo :: [Demand] -> Maybe Id -> StrictnessInfo - -mkStrictnessInfo [] _ = NoStrictnessInfo -mkStrictnessInfo xs wrkr = StrictnessInfo xs wrkr - -mkBottomStrictnessInfo = BottomGuaranteed - -bottomIsGuaranteed BottomGuaranteed = True -bottomIsGuaranteed other = False - -getWrapperArgTypeCategories - :: Type -- wrapper's type - -> StrictnessInfo -- strictness info about its args - -> Maybe String - -getWrapperArgTypeCategories _ NoStrictnessInfo = Nothing -getWrapperArgTypeCategories _ BottomGuaranteed - = trace "getWrapperArgTypeCategories:BottomGuaranteed!" Nothing -- wrong -getWrapperArgTypeCategories _ (StrictnessInfo [] _) = Nothing -getWrapperArgTypeCategories ty (StrictnessInfo arg_info _) - = Just (mkWrapperArgTypeCategories ty arg_info) +data WorkerInfo = NoWorker + | HasWorker Id Arity + -- The Arity is the arity of the *wrapper* at the moment of the + -- w/w split. See comments in MkIface.ifaceId, with the 'Worker' code. -workerExists :: StrictnessInfo -> Bool -workerExists (StrictnessInfo _ (Just worker_id)) = True -workerExists other = False +seqWorker :: WorkerInfo -> () +seqWorker (HasWorker id _) = id `seq` () +seqWorker NoWorker = () -getWorkerId :: StrictnessInfo -> Id +ppWorkerInfo NoWorker = empty +ppWorkerInfo (HasWorker wk_id _) = ptext SLIT("__P") <+> ppr wk_id -getWorkerId (StrictnessInfo _ (Just worker_id)) = worker_id -#ifdef DEBUG -getWorkerId junk = pprPanic "getWorkerId: " (ppInfo PprDebug (\x->x) junk) -#endif -\end{code} - -\begin{code} -isStrict :: Demand -> Bool +noWorkerInfo = NoWorker -isStrict WwStrict = True -isStrict (WwUnpack _) = True -isStrict WwPrim = True -isStrict WwEnum = True -isStrict _ = False +workerExists :: WorkerInfo -> Bool +workerExists NoWorker = False +workerExists (HasWorker _ _) = True -nonAbsentArgs :: [Demand] -> Int - -nonAbsentArgs cmpts - = foldr tick_non 0 cmpts - where - tick_non (WwLazy True) acc = acc - tick_non other acc = acc + 1 +workerId :: WorkerInfo -> Id +workerId (HasWorker id _) = id -all_present_WwLazies :: [Demand] -> Bool -all_present_WwLazies infos - = and (map is_L infos) - where - is_L (WwLazy False) = True -- False <=> "Absent" args do *not* count! - is_L _ = False -- (as they imply a worker) +wrapperArity :: WorkerInfo -> Arity +wrapperArity (HasWorker _ a) = a \end{code} -WDP 95/04: It is no longer enough to look at a list of @Demands@ for -an ``Unpack'' or an ``Absent'' and declare a worker. We also have to -check that @mAX_WORKER_ARGS@ hasn't been exceeded. Therefore, -@indicatesWorker@ mirrors the process used in @mk_ww_arg_processing@ -in \tr{WwLib.lhs}. A worker is ``indicated'' when we hit an Unpack -or an Absent {\em that we accept}. -\begin{code} -indicatesWorker :: [Demand] -> Bool - -indicatesWorker dems - = fake_mk_ww (_trace "mAX_WORKER_ARGS" 6 - nonAbsentArgs dems) dems - where - fake_mk_ww _ [] = False - fake_mk_ww _ (WwLazy True : _) = True -- we accepted an Absent - fake_mk_ww extra_args (WwUnpack cmpnts : dems) - | extra_args_now > 0 = True -- we accepted an Unpack - where - extra_args_now = extra_args + 1 - nonAbsentArgs cmpnts - - fake_mk_ww extra_args (_ : dems) - = fake_mk_ww extra_args dems -\end{code} - -\begin{code} -mkWrapperArgTypeCategories - :: Type -- wrapper's type - -> [Demand] -- info about its arguments - -> String -- a string saying lots about the args - -mkWrapperArgTypeCategories wrapper_ty wrap_info - = case (splitTypeWithDictsAsArgs wrapper_ty) of { (_,arg_tys,_) -> - map do_one (wrap_info `zip` (map showTypeCategory arg_tys)) - } - where - -- ToDo: this needs FIXING UP (it was a hack anyway...) - do_one (WwPrim, _) = 'P' - do_one (WwEnum, _) = 'E' - do_one (WwStrict, arg_ty_char) = arg_ty_char - do_one (WwUnpack _, arg_ty_char) - = if arg_ty_char `elem` "CIJFDTS" - then toLower arg_ty_char - else if arg_ty_char == '+' then 't' - else trace ("mkWrapp..:funny char:"++[arg_ty_char]) '-' - do_one (other_wrap_info, _) = '-' -\end{code} - -Whether a worker exists depends on whether the worker has an -absent argument, a @WwUnpack@ argument, (or @WwEnum@ ToDo???) arguments. - -If a @WwUnpack@ argument is for an {\em abstract} type (or one that -will be abstract outside this module), which might happen for an -imported function, then we can't (or don't want to...) unpack the arg -as the worker requires. Hence we have to give up altogether, and call -the wrapper only; so under these circumstances we return \tr{False}. - -\begin{code} -instance Text Demand where - readList str = read_em [{-acc-}] str - where - read_em acc [] = [(reverse acc, "")] - -- lower case indicates absence... - read_em acc ('L' : xs) = read_em (WwLazy False : acc) xs - read_em acc ('A' : xs) = read_em (WwLazy True : acc) xs - read_em acc ('S' : xs) = read_em (WwStrict : acc) xs - read_em acc ('P' : xs) = read_em (WwPrim : acc) xs - read_em acc ('E' : xs) = read_em (WwEnum : acc) xs - - read_em acc (')' : xs) = [(reverse acc, xs)] - read_em acc ( 'U' : '(' : xs) - = case (read_em [] xs) of - [(stuff, rest)] -> read_em (WwUnpack stuff : acc) rest - _ -> panic ("Text.Demand:"++str++"::"++xs) - - read_em acc other = panic ("IdInfo.readem:"++other) - - showList wrap_args rest = (concat (map show1 wrap_args)) ++ rest - where - show1 (WwLazy False) = "L" - show1 (WwLazy True) = "A" - show1 WwStrict = "S" - show1 WwPrim = "P" - show1 WwEnum = "E" - show1 (WwUnpack args)= "U(" ++ (concat (map show1 args)) ++ ")" - -instance Outputable Demand where - ppr sty si = ppStr (showList [si] "") - -instance OptIdInfo StrictnessInfo where - noInfo = NoStrictnessInfo - - getInfo (IdInfo _ _ _ strict _ _ _ _ _ _) = strict - - addInfo id_info NoStrictnessInfo = id_info - addInfo (IdInfo a b d _ e f g h i j) strict = IdInfo a b d strict e f g h i j - - ppInfo sty better_id_fn strictness_info - = pp_strictness sty Nothing better_id_fn nullIdEnv strictness_info -\end{code} - -We'll omit the worker info if the thing has an explicit unfolding -already. -\begin{code} -pp_strictness sty _ _ _ NoStrictnessInfo = ifPprInterface sty pp_NONE - -pp_strictness sty _ _ _ BottomGuaranteed = ppPStr SLIT("_S_ _!_") - -pp_strictness sty for_this_id_maybe better_id_fn inline_env - info@(StrictnessInfo wrapper_args wrkr_maybe) - = let - (have_wrkr, wrkr_id) = case wrkr_maybe of - Nothing -> (False, panic "ppInfo(Strictness)") - Just xx -> (True, xx) - - wrkr_to_print = better_id_fn wrkr_id - wrkr_info = getIdInfo wrkr_to_print - - -- if we aren't going to be able to *read* the strictness info - -- in TcPragmas, we need not even print it. - wrapper_args_to_use - = if not (indicatesWorker wrapper_args) then - wrapper_args -- no worker/wrappering in any case - else - case for_this_id_maybe of - Nothing -> wrapper_args - Just id -> if externallyVisibleId id - && (unfoldingUnfriendlyId id || not have_wrkr) then - -- pprTrace "IdInfo: unworker-ising:" (ppCat [ppr PprDebug have_wrkr, ppr PprDebug id]) $ - map un_workerise wrapper_args - else - wrapper_args - - id_is_worker - = case for_this_id_maybe of - Nothing -> False - Just id -> isWorkerId id - - am_printing_iface = case sty of { PprInterface -> True ; _ -> False } - - pp_basic_info - = ppBesides [ppStr "_S_ \"", - ppStr (showList wrapper_args_to_use ""), ppStr "\""] - - pp_with_worker - = ppBesides [ ppSP, ppChar '{', - ppIdInfo sty wrkr_to_print True{-wrkr specs, yes!-} better_id_fn inline_env wrkr_info, - ppChar '}' ] - in - if all_present_WwLazies wrapper_args_to_use then -- too boring - ifPprInterface sty pp_NONE - - else if id_is_worker && am_printing_iface then - pp_NONE -- we don't put worker strictness in interfaces - -- (it can be deduced) - - else if not (indicatesWorker wrapper_args_to_use) - || not have_wrkr - || boringIdInfo wrkr_info then - ppBeside pp_basic_info ppNil - else - ppBeside pp_basic_info pp_with_worker - where - un_workerise (WwLazy _) = WwLazy False -- avoid absence - un_workerise (WwUnpack _) = WwStrict - un_workerise other = other -\end{code} %************************************************************************ %* * -\subsection[unfolding-IdInfo]{Unfolding info about an @Id@} +\subsection[CAF-IdInfo]{CAF-related information} %* * %************************************************************************ -\begin{code} -mkUnfolding guide expr - = GenForm False (mkFormSummary NoStrictnessInfo expr) - (BSCC("OccurExpr") occurAnalyseGlobalExpr expr ESCC) - guide -\end{code} +This information is used to build Static Reference Tables (see +simplStg/ComputeSRT.lhs). \begin{code} -noInfo_UF = NoUnfoldingDetails - -getInfo_UF (IdInfo _ _ _ _ unfolding _ _ _ _ _) - = case unfolding of - GenForm _ _ _ BadUnfolding -> NoUnfoldingDetails - unfolding_as_was -> unfolding_as_was - --- getInfo_UF ensures that any BadUnfoldings are never returned --- We had to delay the test required in TcPragmas until now due --- to strictness constraints in TcPragmas - -addInfo_UF id_info@(IdInfo a b c d e f g h i j) NoUnfoldingDetails = id_info -addInfo_UF (IdInfo a b d e _ f g h i j) uf = IdInfo a b d e uf f g h i j -\end{code} +data CafInfo + = MayHaveCafRefs -- either: + -- (1) A function or static constructor + -- that refers to one or more CAFs, + -- (2) A real live CAF -\begin{code} -pp_unfolding sty for_this_id inline_env uf_details - = case (lookupIdEnv inline_env for_this_id) of - Nothing -> pp uf_details - Just dt -> pp dt - where - pp NoUnfoldingDetails = pp_NONE + | NoCafRefs -- A function or static constructor + -- that refers to no CAFs. - pp (MagicForm tag _) - = ppCat [ppPStr SLIT("_MF_"), ppPStr tag] +-- LATER: not sure how easy this is... +-- | OneCafRef Id - pp (GenForm _ _ _ BadUnfolding) = pp_NONE - pp (GenForm _ _ template guide) - = let - untagged = unTagBinders template - in - if untagged `isWrapperFor` for_this_id - then -- pprTrace "IdInfo:isWrapperFor:" (ppAbove (ppr PprDebug for_this_id) (ppr PprDebug untagged)) - pp_NONE - else ppCat [ppPStr SLIT("_F_"), ppr sty guide, pprCoreUnfolding untagged] +seqCaf c = c `seq` () +ppCafInfo NoCafRefs = ptext SLIT("__C") +ppCafInfo MayHaveCafRefs = empty \end{code} + %************************************************************************ %* * -\subsection[update-IdInfo]{Update-analysis info about an @Id@} +\subsection[cpr-IdInfo]{Constructed Product Result info about an @Id@} %* * %************************************************************************ -\begin{code} -data UpdateInfo - = NoUpdateInfo - | SomeUpdateInfo UpdateSpec - deriving (Eq, Ord) - -- we need Eq/Ord to cross-chk update infos in interfaces - --- the form in which we pass update-analysis info between modules: -type UpdateSpec = [Int] -\end{code} - -\begin{code} -mkUpdateInfo = SomeUpdateInfo +If the @Id@ is a function then it may have CPR info. A CPR analysis +phase detects whether: -updateInfoMaybe NoUpdateInfo = Nothing -updateInfoMaybe (SomeUpdateInfo []) = Nothing -updateInfoMaybe (SomeUpdateInfo u) = Just u -\end{code} +\begin{enumerate} +\item +The function's return value has a product type, i.e. an algebraic type +with a single constructor. Examples of such types are tuples and boxed +primitive values. +\item +The function always 'constructs' the value that it is returning. It +must do this on every path through, and it's OK if it calls another +function which constructs the result. +\end{enumerate} -Text instance so that the update annotations can be read in. +If this is the case then we store a template which tells us the +function has the CPR property and which components of the result are +also CPRs. \begin{code} -instance Text UpdateInfo where - readsPrec p s | null s = panic "IdInfo: empty update pragma?!" - | otherwise = [(SomeUpdateInfo (map ok_digit s),"")] - where - ok_digit c | c >= '0' && c <= '2' = ord c - ord '0' - | otherwise = panic "IdInfo: not a digit while reading update pragma" - -instance OptIdInfo UpdateInfo where - noInfo = NoUpdateInfo +data CprInfo + = NoCPRInfo + | ReturnsCPR -- Yes, this function returns a constructed product + -- Implicitly, this means "after the function has been applied + -- to all its arguments", so the worker/wrapper builder in + -- WwLib.mkWWcpr checks that that it is indeed saturated before + -- making use of the CPR info - getInfo (IdInfo _ _ _ _ _ update _ _ _ _) = update - - addInfo id_info NoUpdateInfo = id_info - addInfo (IdInfo a b d e f _ g h i j) upd_info = IdInfo a b d e f upd_info g h i j - - ppInfo sty better_id_fn NoUpdateInfo = ifPprInterface sty pp_NONE - ppInfo sty better_id_fn (SomeUpdateInfo []) = ifPprInterface sty pp_NONE - ppInfo sty better_id_fn (SomeUpdateInfo spec) - = ppBeside (ppPStr SLIT("_U_ ")) (ppBesides (map ppInt spec)) + -- We used to keep nested info about sub-components, but + -- we never used it so I threw it away \end{code} -%************************************************************************ -%* * -\subsection[deforest-IdInfo]{Deforestation info about an @Id@} -%* * -%************************************************************************ - -The deforest info says whether this Id is to be unfolded during -deforestation. Therefore, when the deforest pragma is true, we must -also have the unfolding information available for this Id. - \begin{code} -data DeforestInfo - = Don'tDeforest -- just a bool, might extend this - | DoDeforest -- later. - -- deriving (Eq, Ord) -\end{code} +seqCpr :: CprInfo -> () +seqCpr ReturnsCPR = () +seqCpr NoCPRInfo = () -\begin{code} -instance OptIdInfo DeforestInfo where - noInfo = Don'tDeforest +noCprInfo = NoCPRInfo - getInfo (IdInfo _ _ _ _ _ _ deforest _ _ _) = deforest +ppCprInfo NoCPRInfo = empty +ppCprInfo ReturnsCPR = ptext SLIT("__M") - addInfo id_info Don'tDeforest = id_info - addInfo (IdInfo a b d e f g _ h i j) deforest = - IdInfo a b d e f g deforest h i j +instance Outputable CprInfo where + ppr = ppCprInfo - ppInfo sty better_id_fn Don'tDeforest - = ifPprInterface sty pp_NONE - ppInfo sty better_id_fn DoDeforest - = ppPStr SLIT("_DEFOREST_") +instance Show CprInfo where + showsPrec p c = showsPrecSDoc p (ppr c) \end{code} + %************************************************************************ %* * -\subsection[argUsage-IdInfo]{Argument Usage info about an @Id@} +\subsection[lbvar-IdInfo]{Lambda-bound var info about an @Id@} %* * %************************************************************************ -\begin{code} -data ArgUsageInfo - = NoArgUsageInfo - | SomeArgUsageInfo ArgUsageType - -- ??? deriving (Eq, Ord) - -data ArgUsage = ArgUsage Int -- number of arguments (is linear!) - | UnknownArgUsage -type ArgUsageType = [ArgUsage] -- c_1 -> ... -> BLOB -\end{code} +If the @Id@ is a lambda-bound variable then it may have lambda-bound +var info. The usage analysis (UsageSP) detects whether the lambda +binding this var is a ``one-shot'' lambda; that is, whether it is +applied at most once. + +This information may be useful in optimisation, as computations may +safely be floated inside such a lambda without risk of duplicating +work. \begin{code} -mkArgUsageInfo = SomeArgUsageInfo +data LBVarInfo + = NoLBVarInfo + + | IsOneShotLambda -- The lambda that binds this Id is applied + -- at most once + -- HACK ALERT! placing this info here is a short-term hack, + -- but it minimises changes to the rest of the compiler. + -- Hack agreed by SLPJ/KSW 1999-04. -getArgUsage :: ArgUsageInfo -> ArgUsageType -getArgUsage NoArgUsageInfo = [] -getArgUsage (SomeArgUsageInfo u) = u +seqLBVar l = l `seq` () \end{code} \begin{code} -instance OptIdInfo ArgUsageInfo where - noInfo = NoArgUsageInfo +noLBVarInfo = NoLBVarInfo - getInfo (IdInfo _ _ _ _ _ _ _ au _ _) = au +-- not safe to print or parse LBVarInfo because it is not really a +-- property of the definition, but a property of the context. +pprLBVarInfo NoLBVarInfo = empty +pprLBVarInfo IsOneShotLambda = getPprStyle $ \ sty -> + if ifaceStyle sty then empty + else ptext SLIT("OneShot") - addInfo id_info NoArgUsageInfo = id_info - addInfo (IdInfo a b d e f g h _ i j) au_info = IdInfo a b d e f g h au_info i j - - ppInfo sty better_id_fn NoArgUsageInfo = ifPprInterface sty pp_NONE - ppInfo sty better_id_fn (SomeArgUsageInfo []) = ifPprInterface sty pp_NONE - ppInfo sty better_id_fn (SomeArgUsageInfo aut) - = ppBeside (ppPStr SLIT("_L_ ")) (ppArgUsageType aut) +instance Outputable LBVarInfo where + ppr = pprLBVarInfo +instance Show LBVarInfo where + showsPrec p c = showsPrecSDoc p (ppr c) +\end{code} -ppArgUsage (ArgUsage n) = ppInt n -ppArgUsage (UnknownArgUsage) = ppChar '-' -ppArgUsageType aut = ppBesides - [ ppChar '"' , - ppIntersperse ppComma (map ppArgUsage aut), - ppChar '"' ] -\end{code} %************************************************************************ %* * -\subsection[FBType-IdInfo]{Type of an expression through Foldr/build's eyes} +\subsection{Bulk operations on IdInfo} %* * %************************************************************************ -\begin{code} -data FBTypeInfo - = NoFBTypeInfo - | SomeFBTypeInfo FBType - -- ??? deriving (Eq, Ord) - -data FBType = FBType [FBConsum] FBProd deriving (Eq) +zapFragileInfo is used when cloning binders, mainly in the +simplifier. We must forget about used-once information because that +isn't necessarily correct in the transformed program. +Also forget specialisations and unfoldings because they would need +substitution to be correct. (They get pinned back on separately.) + +\begin{code} +zapFragileInfo :: IdInfo -> Maybe IdInfo +zapFragileInfo info@(IdInfo {occInfo = occ, + workerInfo = wrkr, + specInfo = rules, + unfoldingInfo = unfolding}) + | not (isFragileOcc occ) + -- We must forget about whether it was marked safe-to-inline, + -- because that isn't necessarily true in the simplified expression. + -- This is important because expressions may be re-simplified + -- We don't zap deadness or loop-breaker-ness. + -- The latter is important because it tells MkIface not to + -- spit out an inlining for the thing. The former doesn't + -- seem so important, but there's no harm. + + && isEmptyCoreRules rules + -- Specialisations would need substituting. They get pinned + -- back on separately. + + && not (workerExists wrkr) + + && not (hasUnfolding unfolding) + -- This is very important; occasionally a let-bound binder is used + -- as a binder in some lambda, in which case its unfolding is utterly + -- bogus. Also the unfolding uses old binders so if we left it we'd + -- have to substitute it. Much better simply to give the Id a new + -- unfolding each time, which is what the simplifier does. + = Nothing -data FBConsum = FBGoodConsum | FBBadConsum deriving(Eq) -data FBProd = FBGoodProd | FBBadProd deriving(Eq) + | otherwise + = Just (info {occInfo = robust_occ_info, + workerInfo = noWorkerInfo, + specInfo = emptyCoreRules, + unfoldingInfo = noUnfolding}) + where + -- It's important to keep the loop-breaker info, + -- because the substitution doesn't remember it. + robust_occ_info = case occ of + OneOcc _ _ -> NoOccInfo + other -> occ \end{code} -\begin{code} -mkFBTypeInfo = SomeFBTypeInfo - -getFBType :: FBTypeInfo -> Maybe FBType -getFBType NoFBTypeInfo = Nothing -getFBType (SomeFBTypeInfo u) = Just u -\end{code} +@zapLamInfo@ is used for lambda binders that turn out to to be +part of an unsaturated lambda \begin{code} -instance OptIdInfo FBTypeInfo where - noInfo = NoFBTypeInfo - - getInfo (IdInfo _ _ _ _ _ _ _ _ fb _) = fb - - addInfo id_info NoFBTypeInfo = id_info - addInfo (IdInfo a b d e f g h i _ j) fb_info = IdInfo a b d e f g h i fb_info j - - ppInfo PprInterface _ NoFBTypeInfo = ppNil - ppInfo sty _ NoFBTypeInfo = ifPprInterface sty pp_NONE - ppInfo sty _ (SomeFBTypeInfo (FBType cons prod)) - = ppBeside (ppPStr SLIT("_F_ ")) (ppFBType cons prod) - ---ppFBType (FBType n) = ppBesides [ppInt n] ---ppFBType (UnknownFBType) = ppBesides [ppStr "-"] --- - -ppFBType cons prod = ppBesides - ([ ppChar '"' ] ++ map ppCons cons ++ [ ppChar '-', ppProd prod, ppChar '"' ]) +zapLamInfo :: IdInfo -> Maybe IdInfo +zapLamInfo info@(IdInfo {occInfo = occ, demandInfo = demand}) + | is_safe_occ && not (isStrict demand) + = Nothing + | otherwise + = Just (info {occInfo = safe_occ, + demandInfo = wwLazy}) where - ppCons FBGoodConsum = ppChar 'G' - ppCons FBBadConsum = ppChar 'B' - ppProd FBGoodProd = ppChar 'G' - ppProd FBBadProd = ppChar 'B' + -- The "unsafe" occ info is the ones that say I'm not in a lambda + -- because that might not be true for an unsaturated lambda + is_safe_occ = case occ of + OneOcc in_lam once -> in_lam + other -> True + + safe_occ = case occ of + OneOcc _ once -> OneOcc insideLam once + other -> occ +\end{code} + + +copyIdInfo is used when shorting out a top-level binding + f_local = BIG + f = f_local +where f is exported. We are going to swizzle it around to + f = BIG + f_local = f + +BUT (a) we must be careful about messing up rules + (b) we must ensure f's IdInfo ends up right + +(a) Messing up the rules +~~~~~~~~~~~~~~~~~~~~ +The example that went bad on me was this one: + + iterate :: (a -> a) -> a -> [a] + iterate = iterateList + + iterateFB c f x = x `c` iterateFB c f (f x) + iterateList f x = x : iterateList f (f x) + + {-# RULES + "iterate" forall f x. iterate f x = build (\c _n -> iterateFB c f x) + "iterateFB" iterateFB (:) = iterateList + #-} + +This got shorted out to: + + iterateList :: (a -> a) -> a -> [a] + iterateList = iterate + + iterateFB c f x = x `c` iterateFB c f (f x) + iterate f x = x : iterate f (f x) + + {-# RULES + "iterate" forall f x. iterate f x = build (\c _n -> iterateFB c f x) + "iterateFB" iterateFB (:) = iterate + #-} + +And now we get an infinite loop in the rule system + iterate f x -> build (\cn -> iterateFB c f x + -> iterateFB (:) f x + -> iterate f x + +Tiresome solution: don't do shorting out if f has rewrite rules. +Hence shortableIdInfo. + +(b) Keeping the IdInfo right +~~~~~~~~~~~~~~~~~~~~~~~~ +We want to move strictness/worker info from f_local to f, but keep the rest. +Hence copyIdInfo. + +\begin{code} +shortableIdInfo :: IdInfo -> Bool +shortableIdInfo info = isEmptyCoreRules (specInfo info) + +copyIdInfo :: IdInfo -- f_local + -> IdInfo -- f (the exported one) + -> IdInfo -- New info for f +copyIdInfo f_local f = f { strictnessInfo = strictnessInfo f_local, + workerInfo = workerInfo f_local, + cprInfo = cprInfo f_local + } \end{code}