-- The Name type
Name, -- Abstract
mkLocalName, mkSysLocalName, mkCCallName,
- mkTopName, mkIPName,
+ mkIPName,
mkDerivedName, mkGlobalName, mkKnownKeyGlobal, mkWiredInName,
nameUnique, setNameUnique,
- tidyTopName,
nameOccName, nameModule, nameModule_maybe,
setNameOcc, nameRdrName, setNameModuleAndLoc,
- toRdrName, hashName,
+ toRdrName, hashName,
+ globaliseName, localiseName,
nameSrcLoc, nameIsLocallyDefined, isDllName, nameIsFrom, nameIsLocalOrFrom,
-- Environment
NameEnv, mkNameEnv,
emptyNameEnv, unitNameEnv, nameEnvElts,
- extendNameEnv_C, extendNameEnv, foldNameEnv,
+ extendNameEnv_C, extendNameEnv, foldNameEnv, filterNameEnv,
plusNameEnv, plusNameEnv_C, extendNameEnv, extendNameEnvList,
lookupNameEnv, lookupNameEnv_NF, delFromNameEnv, elemNameEnv,
mkGlobalName :: Unique -> Module -> OccName -> SrcLoc -> Name
mkGlobalName uniq mod occ loc = Name { n_uniq = uniq, n_sort = Global mod,
n_occ = occ, n_loc = loc }
-
mkKnownKeyGlobal :: RdrName -> Unique -> Name
mkKnownKeyGlobal rdr_name uniq
setNameUnique name uniq = name {n_uniq = uniq}
setNameOcc :: Name -> OccName -> Name
- -- Give the thing a new OccName, *and*
- -- record that it's no longer a sys-local
- -- This is used by the tidy-up pass
setNameOcc name occ = name {n_occ = occ}
+globaliseName :: Name -> Module -> Name
+globaliseName n mod = n { n_sort = Global mod }
+
+localiseName :: Name -> Name
+localiseName n = n { n_sort = Local }
+
setNameModuleAndLoc :: Name -> Module -> SrcLoc -> Name
setNameModuleAndLoc name mod loc = name {n_sort = set (n_sort name), n_loc = loc}
where
%************************************************************************
%* *
-\subsection{Tidying a name}
-%* *
-%************************************************************************
-
-tidyTopName is applied to top-level names in the final program
-
-For top-level things,
- it globalises Local names
- (if all top-level things should be visible)
- and localises non-exported Global names
- (if only exported things should be visible)
-
-In all cases except an exported global, it gives it a new occurrence name.
-
-The "visibility" here concerns whether the .o file's symbol table
-mentions the thing; if so, it needs a module name in its symbol.
-The Global things are "visible" and the Local ones are not
-
-Why should things be "visible"? Certainly they must be if they
-are exported. But also:
-
-(a) In certain (prelude only) modules we split up the .hc file into
- lots of separate little files, which are separately compiled by the C
- compiler. That gives lots of little .o files. The idea is that if
- you happen to mention one of them you don't necessarily pull them all
- in. (Pulling in a piece you don't need can be v bad, because it may
- mention other pieces you don't need either, and so on.)
-
- Sadly, splitting up .hc files means that local names (like s234) are
- now globally visible, which can lead to clashes between two .hc
- files. So unlocaliseWhatnot goes through making all the local things
- into global things, essentially by giving them full names so when they
- are printed they'll have their module name too. Pretty revolting
- really.
-
-(b) When optimisation is on we want to make all the internal
- top-level defns externally visible
-
-\begin{code}
-tidyTopName :: Module -> TidyOccEnv -> Bool -> Name -> (TidyOccEnv, Name)
-tidyTopName mod env is_exported
- name@(Name { n_occ = occ, n_sort = sort, n_uniq = uniq, n_loc = loc })
- = case sort of
- Global _ | is_exported -> (env, name)
- | otherwise -> (env, name { n_sort = new_sort })
- -- Don't change the occurrnce names of globals, because many of them
- -- are bound by either a class declaration or a data declaration
- -- or an explicit user export.
-
- other | is_exported -> (env', name { n_sort = Global mod, n_occ = occ' })
- | otherwise -> (env', name { n_sort = new_sort, n_occ = occ' })
- where
- (env', occ') = tidyOccName env occ
- new_sort = mkLocalTopSort mod
-
-mkTopName :: Unique -> Module -> FAST_STRING -> Name
- -- Make a top-level name; make it Global if top-level
- -- things should be externally visible; Local otherwise
- -- This chap is only used *after* the tidyCore phase
- -- Notably, it is used during STG lambda lifting
- --
- -- We have to make sure that the name is globally unique
- -- and we don't have tidyCore to help us. So we append
- -- the unique. Hack! Hack!
- -- (Used only by the STG lambda lifter.)
-mkTopName uniq mod fs
- = Name { n_uniq = uniq,
- n_sort = mkLocalTopSort mod,
- n_occ = mkVarOcc (_PK_ ((_UNPK_ fs) ++ show uniq)),
- n_loc = noSrcLoc }
-
-mkLocalTopSort :: Module -> NameSort
-mkLocalTopSort mod
- | all_toplev_ids_visible = Global mod
- | otherwise = Local
-
-all_toplev_ids_visible
- = not opt_OmitInterfacePragmas || -- Pragmas can make them visible
- opt_EnsureSplittableC -- Splitting requires visiblilty
-\end{code}
-
-
-
-%************************************************************************
-%* *
\subsection{Predicates and selectors}
%* *
%************************************************************************
isTyVarName :: Name -> Bool
isTyVarName name = isTvOcc (nameOccName name)
-
\end{code}
lookupNameEnv_NF :: NameEnv a -> Name -> a
mapNameEnv :: (a->b) -> NameEnv a -> NameEnv b
foldNameEnv :: (a -> b -> b) -> b -> NameEnv a -> b
+filterNameEnv :: (elt -> Bool) -> NameEnv elt -> NameEnv elt
emptyNameEnv = emptyUFM
foldNameEnv = foldUFM
elemNameEnv = elemUFM
mapNameEnv = mapUFM
unitNameEnv = unitUFM
+filterNameEnv = filterUFM
lookupNameEnv = lookupUFM
lookupNameEnv_NF env n = expectJust "lookupNameEnv_NF" (lookupUFM env n)
#include "HsVersions.h"
-import CmdLineOpts ( DynFlags, DynFlag(..), opt_UsageSPOn, dopt )
+import CmdLineOpts ( DynFlags, DynFlag(..), opt_OmitInterfacePragmas, dopt )
import CoreSyn
-import CoreUnfold ( noUnfolding )
+import CoreUnfold ( noUnfolding, mkTopUnfolding, okToUnfoldInHiFile )
+import CoreFVs ( ruleSomeFreeVars, exprSomeFreeVars )
import CoreLint ( showPass, endPass )
-import UsageSPInf ( doUsageSPInf )
import VarEnv
import VarSet
import Var ( Id, Var )
import Id ( idType, idInfo, idName, isExportedId,
- mkVanillaId, mkId,
- idStrictness, setIdStrictness,
- idDemandInfo, setIdDemandInfo,
+ mkVanillaId, mkId, isLocalId,
+ setIdStrictness, setIdDemandInfo,
)
-import IdInfo ( specInfo, setSpecInfo,
- setUnfoldingInfo, setDemandInfo,
+import IdInfo ( constantIdInfo,
+ specInfo, setSpecInfo,
+ cprInfo, setCprInfo,
+ inlinePragInfo, setInlinePragInfo, isNeverInlinePrag,
+ strictnessInfo, setStrictnessInfo, isBottomingStrictness,
+ unfoldingInfo, setUnfoldingInfo,
+ demandInfo,
+ occInfo, isLoopBreaker,
workerInfo, setWorkerInfo, WorkerInfo(..)
)
-import Demand ( wwLazy )
-import Name ( getOccName, tidyTopName, mkLocalName, isGlobalName )
-import OccName ( initTidyOccEnv, tidyOccName )
+import Name ( getOccName, nameOccName, globaliseName, setNameOcc,
+ localiseName, mkLocalName, isGlobalName
+ )
+import OccName ( TidyOccEnv, initTidyOccEnv, tidyOccName )
import Type ( tidyTopType, tidyType, tidyTyVar )
-import Module ( Module )
-import UniqSupply ( mkSplitUniqSupply )
+import Module ( Module, moduleName )
+import HscTypes ( PersistentCompilerState( pcs_PRS ), PersistentRenamerState( prsOrig ),
+ OrigNameEnv( origNames ), OrigNameNameEnv
+ )
import Unique ( Uniquable(..) )
+import FiniteMap ( lookupFM, addToFM )
+import Maybes ( maybeToBool, orElse )
import ErrUtils ( showPass )
import SrcLoc ( noSrcLoc )
+import UniqFM ( mapUFM )
+import Outputable
+import List ( partition )
import Util ( mapAccumL )
\end{code}
%************************************************************************
-%* *
-\subsection{Tidying core}
-%* *
+%* *
+\subsection{What goes on}
+%* *
%************************************************************************
-Several tasks are done by @tidyCorePgm@
-
-1. If @opt_UsageSPOn@ then compute usage information (which is
- needed by Core2Stg). ** NOTE _scc_ HERE **
- Do this first, because it may introduce new binders.
-
-2. Make certain top-level bindings into Globals. The point is that
- Global things get externally-visible labels at code generation
- time
-
-
-3. Give all binders a nice print-name. Their uniques aren't changed;
- rather we give them lexically unique occ-names, so that we can
- safely print the OccNae only in the interface file. [Bad idea to
- change the uniques, because the code generator makes global labels
- from the uniques for local thunks etc.]
+[SLPJ: 19 Nov 00]
+
+The plan is this.
+
+Step 1: Figure out external Ids
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+First we figure out which Ids are "external" Ids. An
+"external" Id is one that is visible from outside the compilation
+unit. These are
+ a) the user exported ones
+ b) ones mentioned in the unfoldings, workers,
+ or rules of externally-visible ones
+This exercise takes a sweep of the bindings bottom to top. Actually,
+in Step 2 we're also going to need to know which Ids should be
+exported with their unfoldings, so we produce not an IdSet but an
+IdEnv Bool
+
+
+
+Step 2: Tidy the program
+~~~~~~~~~~~~~~~~~~~~~~~~
+Next we traverse the bindings top to bottom. For each top-level
+binder
+
+ - Make all external Ids have Global names and vice versa
+ This is used by the code generator to decide whether
+ to make the label externally visible
+
+ - Give external ids a "tidy" occurrence name. This means
+ we can print them in interface files without confusing
+ "x" (unique 5) with "x" (unique 10).
+
+ - Give external Ids the same Unique as they had before
+ if the name is in the renamer's name cache
+
+ - Give the Id its final IdInfo; in ptic,
+ * Its flavour becomes ConstantId, reflecting the fact that
+ from now on we regard it as a constant, not local, Id
+ * its unfolding, if it should have one
+
+Finally, substitute these new top-level binders consistently
+throughout, including in unfoldings. We also tidy binders in
+RHSs, so that they print nicely in interfaces.
\begin{code}
tidyCorePgm :: DynFlags -> Module
+ -> PersistentCompilerState
-> [CoreBind] -> [IdCoreRule]
- -> IO ([CoreBind], [IdCoreRule])
-tidyCorePgm dflags module_name binds_in orphans_in
- = do
- us <- mkSplitUniqSupply 'u'
+ -> IO (PersistentCompilerState, [CoreBind], [IdCoreRule])
+tidyCorePgm dflags mod pcs binds_in orphans_in
+ = do { showPass dflags "Tidy Core"
+
+ ; let ext_ids = findExternalSet binds_in orphans_in
- showPass dflags "Tidy Core"
+ ; let ((orig_env', occ_env, subst_env), binds_out)
+ = mapAccumL (tidyTopBind mod ext_ids) init_tidy_env binds_in
- binds_in1 <- if opt_UsageSPOn
- then _scc_ "CoreUsageSPInf"
- doUsageSPInf dflags us binds_in
- else return binds_in
+ ; let orphans_out = tidyIdRules (occ_env,subst_env) orphans_in
- let (tidy_env1, binds_out) = mapAccumL (tidyBind (Just module_name))
- init_tidy_env binds_in1
- orphans_out = tidyIdRules tidy_env1 orphans_in
+ ; let pcs' = pcs { pcs_PRS = prs { prsOrig = orig { origNames = orig_env' }}}
- endPass dflags "Tidy Core" (dopt Opt_D_dump_simpl dflags ||
- dopt Opt_D_verbose_core2core dflags)
- binds_out
+ ; endPass dflags "Tidy Core" (dopt Opt_D_dump_simpl dflags ||
+ dopt Opt_D_verbose_core2core dflags)
+ binds_out
- return (binds_out, orphans_out)
+ ; return (pcs', binds_out, orphans_out)
+ }
where
-- We also make sure to avoid any exported binders. Consider
-- f{-u1-} = 1 -- Local decl
-- The second exported decl must 'get' the name 'f', so we
-- have to put 'f' in the avoids list before we get to the first
-- decl. tidyTopId then does a no-op on exported binders.
- init_tidy_env = (initTidyOccEnv avoids, emptyVarEnv)
+ prs = pcs_PRS pcs
+ orig = prsOrig prs
+ orig_env = origNames orig
+
+ init_tidy_env = (orig_env, initTidyOccEnv avoids, emptyVarEnv)
avoids = [getOccName bndr | bndr <- bindersOfBinds binds_in,
isGlobalName (idName bndr)]
+\end{code}
+
+
+%************************************************************************
+%* *
+\subsection{Step 1: finding externals}
+%* *
+%************************************************************************
+
+\begin{code}
+findExternalSet :: [CoreBind] -> [IdCoreRule]
+ -> IdEnv Bool -- True <=> show unfolding
+ -- Step 1 from the notes above
+findExternalSet binds orphan_rules
+ = foldr find init_needed binds
+ where
+ orphan_rule_ids :: IdSet
+ orphan_rule_ids = unionVarSets [ ruleSomeFreeVars isIdAndLocal rule
+ | (_, rule) <- orphan_rules]
+ init_needed :: IdEnv Bool
+ init_needed = mapUFM (\_ -> False) orphan_rule_ids
+ -- The mapUFM is a bit cheesy. It is a cheap way
+ -- to turn the set of orphan_rule_ids, which we use to initialise
+ -- the sweep, into a mapping saying 'don't expose unfolding'
+ -- (When we come to the binding site we may change our mind, of course.)
+
+ find (NonRec id rhs) needed
+ | need_id needed id = addExternal (id,rhs) needed
+ | otherwise = needed
+ find (Rec prs) needed = find_prs prs needed
+
+ -- For a recursive group we have to look for a fixed point
+ find_prs prs needed
+ | null needed_prs = needed
+ | otherwise = find_prs other_prs new_needed
+ where
+ (needed_prs, other_prs) = partition (need_pr needed) prs
+ new_needed = foldr addExternal needed needed_prs
+
+ -- The 'needed' set contains the Ids that are needed by earlier
+ -- interface file emissions. If the Id isn't in this set, and isn't
+ -- exported, there's no need to emit anything
+ need_id needed_set id = id `elemVarEnv` needed_set || isExportedId id
+ need_pr needed_set (id,rhs) = need_id needed_set id
+
+isIdAndLocal id = isId id && isLocalId id
+
+addExternal :: (Id,CoreExpr) -> IdEnv Bool -> IdEnv Bool
+-- The Id is needed; extend the needed set
+-- with it and its dependents (free vars etc)
+addExternal (id,rhs) needed
+ = extendVarEnv (foldVarSet add_occ needed new_needed_ids)
+ id show_unfold
+ where
+ add_occ id needed = extendVarEnv needed id False
+ -- "False" because we don't know we need the Id's unfolding
+ -- We'll override it later when we find the binding site
+
+ new_needed_ids | opt_OmitInterfacePragmas = emptyVarSet
+ | otherwise = worker_ids `unionVarSet`
+ unfold_ids `unionVarSet`
+ spec_ids
+
+ idinfo = idInfo id
+ dont_inline = isNeverInlinePrag (inlinePragInfo idinfo)
+ loop_breaker = isLoopBreaker (occInfo idinfo)
+ bottoming_fn = isBottomingStrictness (strictnessInfo idinfo)
+ spec_ids = rulesRhsFreeVars (specInfo idinfo)
+ worker_info = workerInfo idinfo
+
+ -- Stuff to do with the Id's unfolding
+ -- The simplifier has put an up-to-date unfolding
+ -- in the IdInfo, but the RHS will do just as well
+ unfolding = unfoldingInfo idinfo
+ rhs_is_small = not (neverUnfold unfolding)
+
+ -- We leave the unfolding there even if there is a worker
+ -- In GHCI the unfolding is used by importers
+ -- When writing an interface file, we omit the unfolding
+ -- if there is a worker
+ show_unfold = not bottoming_fn && -- Not necessary
+ not dont_inline &&
+ not loop_breaker &&
+ rhs_is_small && -- Small enough
+ okToUnfoldInHiFile rhs -- No casms etc
+
+ unfold_ids | show_unfold = exprSomeFreeVars isIdAndLocal rhs
+ | otherwise = emptyVarSet
+
+ worker_ids = case worker_info of
+ HasWorker work_id _ -> unitVarSet work_id
+ otherwise -> emptyVarSet
+\end{code}
+
+
+%************************************************************************
+%* *
+\subsection{Step 2: top-level tidying}
+%* *
+%************************************************************************
+
+
+\begin{code}
+type TopTidyEnv = (OrigNameNameEnv, TidyOccEnv, VarEnv Var)
+
+-- TopTidyEnv: when tidying we need to know
+-- * orig_env: Any pre-ordained Names. These may have arisen because the
+-- renamer read in an interface file mentioning M.$wf, say,
+-- and assigned it unique r77. If, on this compilation, we've
+-- invented an Id whose name is $wf (but with a different unique)
+-- we want to rename it to have unique r77, so that we can do easy
+-- comparisons with stuff from the interface file
-tidyBind :: Maybe Module -- (Just m) for top level, Nothing for nested
- -> TidyEnv
+-- * occ_env: The TidyOccEnv, which tells us which local occurrences are 'used'
+
+-- * subst_env: A Var->Var mapping that substitutes the new Var for the old
+\end{code}
+
+
+\begin{code}
+tidyTopBind :: Module
+ -> IdEnv Bool -- Domain = Ids that should be exernal
+ -- True <=> their unfolding is external too
+ -> TopTidyEnv -> CoreBind
+ -> (TopTidyEnv, CoreBind)
+
+tidyTopBind mod ext_ids env (NonRec bndr rhs)
+ = (env', NonRec bndr' rhs')
+ where
+ rhs' = tidyTopRhs env rhs
+ (env', bndr') = tidyTopBinder mod ext_ids env rhs' env bndr
+
+tidyTopBind mod ext_ids env (Rec prs)
+ = (final_env, Rec prs')
+ where
+ (final_env, prs') = mapAccumL do_one env prs
+ do_one env (bndr,rhs) = (env', (bndr', rhs'))
+ where
+ rhs' = tidyTopRhs final_env rhs
+ (env', bndr') = tidyTopBinder mod ext_ids final_env
+ rhs env bndr
+
+tidyTopRhs :: TopTidyEnv -> CoreExpr -> CoreExpr
+ -- Just an impedence matcher
+tidyTopRhs (_, occ_env, subst_env) rhs = tidyExpr (occ_env, subst_env) rhs
+
+tidyTopBinder :: Module -> IdEnv Bool
+ -> TopTidyEnv -> CoreExpr
+ -> TopTidyEnv -> Id -> (TopTidyEnv, Id)
+tidyTopBinder mod ext_ids env_idinfo rhs (orig_env, occ_env, subst_env) id
+ -- This function is the heart of Step 2
+ -- The second env is the one to use for the IdInfo
+ -- It's necessary because when we are dealing with a recursive
+ -- group, a variable late in the group might be mentioned
+ -- in the IdInfo of one early in the group
+
+ -- The rhs is already tidied
+
+ = ((orig_env', occ_env', subst_env'), id')
+ where
+ (orig_env', occ_env', name') = tidyTopName mod orig_env occ_env
+ is_external
+ (idName id)
+ ty' = tidyTopType (idType id)
+ idinfo' = tidyIdInfo env_idinfo is_external unfold_info id
+ id' = mkId name' ty' idinfo'
+ subst_env' = extendVarEnv subst_env id id'
+
+ maybe_external = lookupVarEnv ext_ids id
+ is_external = maybeToBool maybe_external
+
+ -- Expose an unfolding if ext_ids tells us to
+ show_unfold = maybe_external `orElse` False
+ unfold_info | show_unfold = mkTopUnfolding rhs
+ | otherwise = noUnfolding
+
+tidyIdInfo (_, occ_env, subst_env) is_external unfold_info id
+
+ | opt_OmitInterfacePragmas || not is_external
+ -- No IdInfo if the Id isn't
+ = constantIdInfo
+
+ | otherwise
+ = constantIdInfo `setCprInfo` cprInfo core_idinfo
+ `setStrictnessInfo` strictnessInfo core_idinfo
+ `setInlinePragInfo` inlinePragInfo core_idinfo
+ `setUnfoldingInfo` unfold_info
+ `setWorkerInfo` tidyWorker tidy_env (workerInfo core_idinfo)
+ `setSpecInfo` tidyRules tidy_env (specInfo core_idinfo)
+ where
+ tidy_env = (occ_env, subst_env)
+ core_idinfo = idInfo id
+
+tidyTopName mod orig_env occ_env external name
+ | global && internal = (orig_env, occ_env, localiseName name)
+ | local && internal = (orig_env, occ_env', setNameOcc name occ')
+ | global && external = (orig_env, occ_env, name)
+ | local && external = globalise
+ where
+ -- If we want to globalise a currently-local name, check
+ -- whether we have already assigned a unique for it.
+ -- If so, use it; if not, extend the table
+ globalise = case lookupFM orig_env key of
+ Just orig -> (orig_env, occ_env', orig)
+ Nothing -> (addToFM orig_env key global_name, occ_env', global_name)
+
+ (occ_env', occ') = tidyOccName occ_env (nameOccName name)
+ key = (moduleName mod, occ')
+ global_name = globaliseName (setNameOcc name occ') mod
+ global = isGlobalName name
+ local = not global
+ internal = not external
+
+tidyIdRules :: TidyEnv -> [IdCoreRule] -> [IdCoreRule]
+tidyIdRules env rules
+ = [ (tidyVarOcc env fn, tidyRule env rule) | (fn,rule) <- rules ]
+
+
+tidyWorker tidy_env (HasWorker work_id wrap_arity)
+ = HasWorker (tidyVarOcc tidy_env work_id) wrap_arity
+tidyWorker tidy_env NoWorker
+ = NoWorker
+
+tidyRules :: TidyEnv -> CoreRules -> CoreRules
+tidyRules env (Rules rules fvs)
+ = Rules (map (tidyRule env) rules)
+ (foldVarSet tidy_set_elem emptyVarSet fvs)
+ where
+ tidy_set_elem var new_set = extendVarSet new_set (tidyVarOcc env var)
+
+tidyRule :: TidyEnv -> CoreRule -> CoreRule
+tidyRule env rule@(BuiltinRule _) = rule
+tidyRule env (Rule name vars tpl_args rhs)
+ = (Rule name vars' (map (tidyExpr env') tpl_args) (tidyExpr env' rhs))
+ where
+ (env', vars') = tidyBndrs env vars
+\end{code}
+
+
+%************************************************************************
+%* *
+\subsection{Step 2: inner tidying
+%* *
+%************************************************************************
+
+\begin{code}
+tidyBind :: TidyEnv
-> CoreBind
-> (TidyEnv, CoreBind)
-tidyBind maybe_mod env (NonRec bndr rhs)
+tidyBind env (NonRec bndr rhs)
= let
- (env', bndr') = tidy_bndr maybe_mod env' env bndr
+ (env', bndr') = tidyBndr env bndr
rhs' = tidyExpr env' rhs
-- We use env' when tidying the RHS even though it's not
- -- strictly necessary; it makes the code pretty hard to read
- -- if we don't!
+ -- strictly necessary; it makes the tidied code pretty
+ -- hard to read if we don't!
in
(env', NonRec bndr' rhs')
-tidyBind maybe_mod env (Rec pairs)
- = let
- -- We use env' when tidying the rhss
- -- When tidying the binder itself we may tidy it's
- -- specialisations; if any of these mention other binders
- -- in the group we should really feed env' to them too;
- -- but that seems (a) unlikely and (b) a bit tiresome.
- -- So I left it out for now
-
- (bndrs, rhss) = unzip pairs
- (env', bndrs') = mapAccumL (tidy_bndr maybe_mod env') env bndrs
- rhss' = map (tidyExpr env') rhss
- in
- (env', Rec (zip bndrs' rhss'))
+tidyBind env (Rec prs)
+ = (final_env, Rec prs')
+ where
+ (final_env, prs') = mapAccumL do_one env prs
+ do_one env (bndr,rhs) = (env', (bndr', rhs'))
+ where
+ (env', bndr') = tidyBndr env bndr
+ rhs' = tidyExpr final_env rhs
tidyExpr env (Type ty) = Type (tidyType env ty)
tidyExpr env (Lit lit) = Lit lit
tidyExpr env (Let b e) = Let b' (tidyExpr env' e)
where
- (env', b') = tidyBind Nothing env b
+ (env', b') = tidyBind env b
tidyExpr env (Case e b alts) = Case (tidyExpr env e) b' (map (tidyAlt env') alts)
where
tidyNote env (Coerce t1 t2) = Coerce (tidyType env t1) (tidyType env t2)
tidyNote env note = note
-
-tidyVarOcc (_, var_env) v = case lookupVarEnv var_env v of
- Just v' -> v'
- Nothing -> v
\end{code}
-\begin{code}
-tidy_bndr (Just mod) env_idinfo env var = tidyTopId mod env env_idinfo var
-tidy_bndr Nothing env_idinfo env var = tidyBndr env var
-\end{code}
-
-
%************************************************************************
%* *
-\subsection{Tidying up a binder}
+\subsection{Tidying up non-top-level binders}
%* *
%************************************************************************
\begin{code}
+tidyVarOcc (_, var_env) v = case lookupVarEnv var_env v of
+ Just v' -> v'
+ Nothing -> v
+
tidyBndr :: TidyEnv -> Var -> (TidyEnv, Var)
tidyBndr env var | isTyVar var = tidyTyVar env var
| otherwise = tidyId env var
name' = mkLocalName (getUnique id) occ' noSrcLoc
(tidy_env', occ') = tidyOccName tidy_env (getOccName id)
ty' = tidyType env (idType id)
+ idinfo = idInfo id
id' = mkVanillaId name' ty'
- `setIdStrictness` idStrictness id
- `setIdDemandInfo` idDemandInfo id
+ `setIdStrictness` strictnessInfo idinfo
+ `setIdDemandInfo` demandInfo idinfo
-- NB: This throws away the IdInfo of the Id, which we
-- no longer need. That means we don't need to
-- run over it with env, nor renumber it.
var_env' = extendVarEnv var_env id id'
in
((tidy_env', var_env'), id')
-
-tidyTopId :: Module -> TidyEnv -> TidyEnv -> Id -> (TidyEnv, Id)
- -- The second env is the one to use for the IdInfo
- -- It's necessary because when we are dealing with a recursive
- -- group, a variable late in the group might be mentioned
- -- in the IdInfo of one early in the group
-tidyTopId mod env@(tidy_env, var_env) env_idinfo id
- = -- Top level variables
- let
- (tidy_env', name') = tidyTopName mod tidy_env (isExportedId id) (idName id)
- ty' = tidyTopType (idType id)
- idinfo' = tidyIdInfo env_idinfo (idInfo id)
- id' = mkId name' ty' idinfo'
- var_env' = extendVarEnv var_env id id'
- in
- ((tidy_env', var_env'), id')
-\end{code}
-
-\begin{code}
--- tidyIdInfo does these things:
--- a) tidy the specialisation info and worker info (if any)
--- b) zap the unfolding and demand info
--- The latter two are to avoid space leaks
-
-tidyIdInfo env info
- = info5
- where
- rules = specInfo info
-
- info2 | isEmptyCoreRules rules = info
- | otherwise = info `setSpecInfo` tidyRules env rules
-
- info3 = info2 `setUnfoldingInfo` noUnfolding
- info4 = info3 `setDemandInfo` wwLazy
-
- info5 = case workerInfo info of
- NoWorker -> info4
- HasWorker w a -> info4 `setWorkerInfo` HasWorker (tidyVarOcc env w) a
-
-tidyIdRules :: TidyEnv -> [IdCoreRule] -> [IdCoreRule]
-tidyIdRules env rules
- = [ (tidyVarOcc env fn, tidyRule env rule) | (fn,rule) <- rules ]
-
-tidyRules :: TidyEnv -> CoreRules -> CoreRules
-tidyRules env (Rules rules fvs)
- = Rules (map (tidyRule env) rules)
- (foldVarSet tidy_set_elem emptyVarSet fvs)
- where
- tidy_set_elem var new_set = extendVarSet new_set (tidyVarOcc env var)
-
-tidyRule :: TidyEnv -> CoreRule -> CoreRule
-tidyRule env rule@(BuiltinRule _) = rule
-tidyRule env (Rule name vars tpl_args rhs)
- = (Rule name vars' (map (tidyExpr env') tpl_args) (tidyExpr env' rhs))
- where
- (env', vars') = tidyBndrs env vars
\end{code}
\section[GHC_Main]{Main driver for Glasgow Haskell compiler}
\begin{code}
-module HscMain ( HscResult(..), hscMain, hscExpr, hscTypeExpr,
+module HscMain ( HscResult(..), hscMain,
+#ifdef GHCI
+ hscExpr, hscTypeExpr,
+#endif
initPersistentCompilerState ) where
#include "HsVersions.h"
import InstEnv ( emptyInstEnv )
import Desugar
import SimplCore
-import OccurAnal ( occurAnalyseBinds )
import CoreUtils ( coreBindsSize )
import CoreTidy ( tidyCorePgm )
import CoreToStg ( topCoreBindsToStg )
-- We grab the the unfoldings at this point.
; simpl_result <- dsThenSimplThenTidy dflags pcs_tc hst this_mod
print_unqualified is_exported tc_result
- ; let (tidy_binds, orphan_rules, foreign_stuff) = simpl_result
+ ; let (pcs_simpl, tidy_binds, orphan_rules, foreign_stuff) = simpl_result
-------------------
-- CONVERT TO STG
-------------------
- ; (stg_binds, oa_tidy_binds, cost_centre_info, top_level_ids)
+ ; (stg_binds, cost_centre_info, top_level_ids)
<- myCoreToStg dflags this_mod tidy_binds
; (maybe_stub_h_filename, maybe_stub_c_filename, maybe_ibinds)
<- restOfCodeGeneration dflags toInterp this_mod
(map ideclName (hsModuleImports rdr_module))
- cost_centre_info foreign_stuff env_tc stg_binds oa_tidy_binds
- hit (pcs_PIT pcs_tc)
+ cost_centre_info foreign_stuff env_tc stg_binds tidy_binds
+ hit (pcs_PIT pcs_simpl)
-- and the answer is ...
- ; return (HscRecomp pcs_tc new_details final_iface
+ ; return (HscRecomp pcs_simpl new_details final_iface
maybe_stub_h_filename maybe_stub_c_filename
maybe_ibinds)
}}}}}}}
restOfCodeGeneration dflags toInterp this_mod imported_module_names cost_centre_info
- foreign_stuff env_tc stg_binds oa_tidy_binds
+ foreign_stuff env_tc stg_binds tidy_binds
hit pit -- these last two for mapping ModNames to Modules
| toInterp
= do (ibinds,itbl_env)
-- _scc_ "CodeOutput"
(maybe_stub_h_name, maybe_stub_c_name)
<- codeOutput dflags this_mod local_tycons
- oa_tidy_binds stg_binds
+ tidy_binds stg_binds
c_code h_code abstractC
return (maybe_stub_h_name, maybe_stub_c_name, Nothing)
<- core2core dflags pcs hst is_exported desugared rules
-- Do the final tidy-up
- (tidy_binds, tidy_orphan_rules)
- <- tidyCorePgm dflags this_mod simplified orphan_rules
+ (pcs', tidy_binds, tidy_orphan_rules)
+ <- tidyCorePgm dflags this_mod pcs simplified orphan_rules
- return (tidy_binds, tidy_orphan_rules, (fe_binders,h_code,c_code))
+ return (pcs', tidy_binds, tidy_orphan_rules, (fe_binders,h_code,c_code))
myCoreToStg dflags this_mod tidy_binds
= do
- st_uniqs <- mkSplitUniqSupply 'g'
- let occ_anal_tidy_binds = occurAnalyseBinds tidy_binds
-
() <- coreBindsSize occ_anal_tidy_binds `seq` return ()
-- TEMP: the above call zaps some space usage allocated by the
-- simplifier, which for reasons I don't understand, persists
-- _scc_ "Core2Stg"
stg_binds <- topCoreBindsToStg dflags occ_anal_tidy_binds
- showPass dflags "Stg2Stg"
-- _scc_ "Stg2Stg"
(stg_binds2, cost_centre_info) <- stg2stg dflags this_mod st_uniqs stg_binds
let final_ids = collectFinalStgBinders (map fst stg_binds2)
- return (stg_binds2, occ_anal_tidy_binds, cost_centre_info, final_ids)
+ return (stg_binds2, cost_centre_info, final_ids)
\end{code}
IfaceDecls, mkIfaceDecls, dcl_tycl, dcl_rules, dcl_insts,
TyThing(..), DFunId, TypeEnv, isTyClThing, Avails,
WhatsImported(..), GenAvailInfo(..),
- ImportVersion, AvailInfo, Deprecations(..)
+ ImportVersion, AvailInfo, Deprecations(..),
+ extendTypeEnvList
)
import CmdLineOpts
-import Id ( Id, idType, idInfo, omitIfaceSigForId, isExportedId, hasNoBinding,
- idSpecialisation, idName, setIdInfo
+import Id ( Id, idType, idInfo, omitIfaceSigForId,
+ idSpecialisation, setIdInfo, isLocalId
)
import Var ( isId )
import VarSet
import DataCon ( StrictnessMark(..), dataConSig, dataConFieldLabels, dataConStrictMarks )
import IdInfo -- Lots
-import CoreSyn ( CoreExpr, CoreBind, Bind(..), CoreRule(..), IdCoreRule,
- isBuiltinRule, rulesRules, rulesRhsFreeVars, emptyCoreRules,
- bindersOfBinds
+import CoreSyn ( CoreBind, CoreRule(..), IdCoreRule,
+ isBuiltinRule, rulesRules,
+ bindersOf, bindersOfBinds
)
-import CoreFVs ( exprSomeFreeVars, ruleSomeLhsFreeVars, ruleSomeFreeVars, mustHaveLocalBinding )
-import CoreUnfold ( okToUnfoldInHiFile, mkTopUnfolding, neverUnfold, unfoldingTemplate, noUnfolding )
+import CoreFVs ( ruleSomeLhsFreeVars, ruleSomeFreeVars )
+import CoreUnfold ( neverUnfold, unfoldingTemplate )
import Name ( getName, nameModule, Name, NamedThing(..) )
import Name -- Env
import OccName ( pprOccName )
import Outputable
import Module ( ModuleName )
-import List ( partition )
import IO ( IOMode(..), openFile, hClose )
\end{code}
-- However, we do keep things like constructors, which should not appear
-- in interface files, because they are needed by importing modules when
-- using the compilation manager
- new_type_env = mkNameEnv [(getName tycl, tycl) | tycl <- orig_type_env, isTyClThing tycl]
- `plusNameEnv`
- mkNameEnv [(idName id, AnId id) | id <- final_ids]
+ new_type_env = extendTypeEnvList (filterNameEnv isTyClThing type_env)
+ (map AnId final_ids)
- orig_type_env = nameEnvElts type_env
+ stg_id_set = mkVarSet stg_ids
+ final_ids = [addStgInfo stg_id_set id | bind <- tidy_binds
+ , id <- bindersOf bind
+ , isGlobalName (idName id)]
- final_ids = bindsToIds (mkVarSet dfun_ids `unionVarSet` orphan_rule_ids)
- (mkVarSet stg_ids)
- tidy_binds
-- The complete rules are gotten by combining
-- a) the orphan rules
rule_dcls | opt_OmitInterfacePragmas = []
| otherwise = getRules orphan_rules tidy_binds (mkVarSet final_ids)
- orphan_rule_ids = unionVarSets [ ruleSomeFreeVars interestingId rule
- | (_, rule) <- orphan_rules]
-
-
-- This version is used when we are re-linking a module
-- so we've only run the type checker on its previous interface
mkModDetailsFromIface :: TypeEnv -> [DFunId] -- From typechecker
where
rule_dcls = [(id,rule) | IfaceRuleOut id rule <- rules]
-- All the rules from an interface are of the IfaceRuleOut form
+\end{code}
+
+
+We have to add on the arity and CAF info computed by the code generator
+This is also the moment at which we may forget that this function has
+a worker: see the comments below
+
+\begin{code}
+addStgInfo :: IdSet -- Ids used at code-gen time; they have better pragma info!
+ -> Id -> Id
+addStgInfo stg_ids id
+ = id `setIdInfo` final_idinfo
+ where
+ idinfo = idInfo id
+ idinfo' = idinfo `setArityInfo` stg_arity
+ `setCafInfo` cafInfo stg_idinfo
+ final_idinfo | worker_ok = idinfo'
+ | otherwise = idinfo' `setWorkerInfo` NoWorker
+
+ stg_idinfo = case lookupVarSet stg_ids id of
+ Just id' -> idInfo id'
+ Nothing -> pprTrace "ifaceBinds not found:" (ppr id) $
+ idInfo id
+
+ stg_arity = arityInfo stg_idinfo
+
+ ------------ Worker --------------
+ -- We only treat a function as having a worker if
+ -- the exported arity (which is now the number of visible lambdas)
+ -- is the same as the arity at the moment of the w/w split
+ -- If so, we can safely omit the unfolding inside the wrapper, and
+ -- instead re-generate it from the type/arity/strictness info
+ -- But if the arity has changed, we just take the simple path and
+ -- put the unfolding into the interface file, forgetting the fact
+ -- that it's a wrapper.
+ --
+ -- How can this happen? Sometimes we get
+ -- f = coerce t (\x y -> $wf x y)
+ -- at the moment of w/w split; but the eta reducer turns it into
+ -- f = coerce t $wf
+ -- which is perfectly fine except that the exposed arity so far as
+ -- the code generator is concerned (zero) differs from the arity
+ -- when we did the split (2).
+ --
+ -- All this arises because we use 'arity' to mean "exactly how many
+ -- top level lambdas are there" in interface files; but during the
+ -- compilation of this module it means "how many things can I apply
+ -- this to".
+ worker_ok = case workerInfo idinfo of
+ NoWorker -> True
+ HasWorker work_id wrap_arity -> wrap_arity == arityLowerBound stg_arity
+\end{code}
+
+
+\begin{code}
+getRules :: [IdCoreRule] -- Orphan rules
+ -> [CoreBind] -- Bindings, with rules in the top-level Ids
+ -> IdSet -- Ids that are exported, so we need their rules
+ -> [IdCoreRule]
+getRules orphan_rules binds emitted
+ = orphan_rules ++ local_rules
+ where
+ local_rules = [ (fn, rule)
+ | fn <- bindersOfBinds binds,
+ fn `elemVarSet` emitted,
+ rule <- rulesRules (idSpecialisation fn),
+ not (isBuiltinRule rule),
+ -- We can't print builtin rules in interface files
+ -- Since they are built in, an importing module
+ -- will have access to them anyway
+ -- Sept 00: I've disabled this test. It doesn't stop many, if any, rules
+ -- from coming out, and to make it work properly we need to add ????
+ -- (put it back in for now)
+ all (`elemVarSet` emitted) (varSetElems (ruleSomeLhsFreeVars interestingId rule))
+ -- Spit out a rule only if all its lhs free vars are emitted
+ -- This is a good reason not to do it when we emit the Id itself
+ ]
+
+interestingId id = isId id && isLocalId id
+\end{code}
+
+%************************************************************************
+%* *
+\subsection{Completing an interface}
+%* *
+%************************************************************************
+
+\begin{code}
completeIface :: Maybe ModIface -- The old interface, if we have it
-> ModIface -- The new one, minus the decls and versions
-> ModDetails -- The ModDetails for this module
\end{code}
-%************************************************************************
-%* *
-\subsection{Types and classes}
-%* *
-%************************************************************************
-
\begin{code}
ifaceTyCls :: TyThing -> [RenamedTyClDecl] -> [RenamedTyClDecl]
ifaceTyCls (AClass clas) so_far
------------ Worker --------------
- wrkr_hsinfo = case workerInfo id_info of
+ work_info = workerInfo id_info
+ has_worker = case work_info of { HasWorker _ _ -> True; other -> False }
+ wrkr_hsinfo = case work_info of
HasWorker work_id wrap_arity -> [HsWorker (getName work_id)]
NoWorker -> []
------------ Unfolding --------------
+ -- The unfolding is redundant if there is a worker
unfold_info = unfoldingInfo id_info
inline_prag = inlinePragInfo id_info
rhs = unfoldingTemplate unfold_info
- unfold_hsinfo | neverUnfold unfold_info = []
- | otherwise = [HsUnfold inline_prag (toUfExpr rhs)]
+ unfold_hsinfo | neverUnfold unfold_info
+ || has_worker = []
+ | otherwise = [HsUnfold inline_prag (toUfExpr rhs)]
\end{code}
-
-%************************************************************************
-%* *
-\subsection{Instances and rules}
-%* *
-%************************************************************************
-
\begin{code}
ifaceInstance :: DFunId -> RenamedInstDecl
ifaceInstance dfun_id
%************************************************************************
%* *
-\subsection{Compute final Ids}
-%* *
-%************************************************************************
-
-A "final Id" has exactly the IdInfo for going into an interface file, or
-exporting to another module.
-
-\begin{code}
-bindsToIds :: IdSet -- These Ids are needed already
- -> IdSet -- Ids used at code-gen time; they have better pragma info!
- -> [CoreBind] -- In dependency order, later depend on earlier
- -> [Id] -- Set of Ids actually spat out, complete with exactly the IdInfo
- -- they need for exporting to another module
-
-bindsToIds needed_ids codegen_ids binds
- = go needed_ids (reverse binds) []
- -- Reverse so that later things will
- -- provoke earlier ones to be emitted
- where
- -- The 'needed' set contains the Ids that are needed by earlier
- -- interface file emissions. If the Id isn't in this set, and isn't
- -- exported, there's no need to emit anything
- need_id needed_set id = id `elemVarSet` needed_set || isExportedId id
-
- go needed [] emitted
- | not (isEmptyVarSet needed) = pprTrace "ifaceBinds: free vars:"
- (sep (map ppr (varSetElems needed)))
- emitted
- | otherwise = emitted
-
- go needed (NonRec id rhs : binds) emitted
- | need_id needed id = go new_needed binds (new_id:emitted)
- | otherwise = go needed binds emitted
- where
- (new_id, extras) = mkFinalId codegen_ids False id rhs
- new_needed = (needed `unionVarSet` extras) `delVarSet` id
-
- -- Recursive groups are a bit more of a pain. We may only need one to
- -- start with, but it may call out the next one, and so on. So we
- -- have to look for a fixed point. We don't want necessarily them all,
- -- because without -O we may only need the first one (if we don't emit
- -- its unfolding)
- go needed (Rec pairs : binds) emitted
- = go needed' binds emitted'
- where
- (new_emitted, extras) = go_rec needed pairs
- needed' = (needed `unionVarSet` extras) `minusVarSet` mkVarSet (map fst pairs)
- emitted' = new_emitted ++ emitted
-
- go_rec :: IdSet -> [(Id,CoreExpr)] -> ([Id], IdSet)
- go_rec needed pairs
- | null needed_prs = ([], emptyVarSet)
- | otherwise = (emitted ++ more_emitted,
- extras `unionVarSet` more_extras)
- where
- (needed_prs,leftover_prs) = partition is_needed pairs
- (emitted, extras_s) = unzip [ mkFinalId codegen_ids True id rhs
- | (id,rhs) <- needed_prs, not (omitIfaceSigForId id)]
- extras = unionVarSets extras_s
- (more_emitted, more_extras) = go_rec extras leftover_prs
-
- is_needed (id,_) = need_id needed id
-\end{code}
-
-
-
-\begin{code}
-mkFinalId :: IdSet -- The Ids with arity info from the code generator
- -> Bool -- True <=> recursive, so don't include unfolding
- -> Id
- -> CoreExpr -- The Id's right hand side
- -> (Id, IdSet) -- The emitted id, plus any *extra* needed Ids
-
-mkFinalId codegen_ids is_rec id rhs
- | omitIfaceSigForId id
- = (id, emptyVarSet) -- An optimisation for top-level constructors and suchlike
- | otherwise
- = (id `setIdInfo` new_idinfo, new_needed_ids)
- where
- core_idinfo = idInfo id
- stg_idinfo = case lookupVarSet codegen_ids id of
- Just id' -> idInfo id'
- Nothing -> pprTrace "ifaceBinds not found:" (ppr id) $
- idInfo id
-
- new_idinfo | opt_OmitInterfacePragmas
- = constantIdInfo
- | otherwise
- = core_idinfo `setArityInfo` arity_info
- `setCafInfo` cafInfo stg_idinfo
- `setUnfoldingInfo` unfold_info
- `setWorkerInfo` worker_info
- `setSpecInfo` emptyCoreRules
- -- We zap the specialisations because they are
- -- passed on separately through the modules IdCoreRules
-
- ------------ Arity --------------
- arity_info = arityInfo stg_idinfo
- stg_arity = arityLowerBound arity_info
-
- ------------ Worker --------------
- -- We only treat a function as having a worker if
- -- the exported arity (which is now the number of visible lambdas)
- -- is the same as the arity at the moment of the w/w split
- -- If so, we can safely omit the unfolding inside the wrapper, and
- -- instead re-generate it from the type/arity/strictness info
- -- But if the arity has changed, we just take the simple path and
- -- put the unfolding into the interface file, forgetting the fact
- -- that it's a wrapper.
- --
- -- How can this happen? Sometimes we get
- -- f = coerce t (\x y -> $wf x y)
- -- at the moment of w/w split; but the eta reducer turns it into
- -- f = coerce t $wf
- -- which is perfectly fine except that the exposed arity so far as
- -- the code generator is concerned (zero) differs from the arity
- -- when we did the split (2).
- --
- -- All this arises because we use 'arity' to mean "exactly how many
- -- top level lambdas are there" in interface files; but during the
- -- compilation of this module it means "how many things can I apply
- -- this to".
- worker_info = case workerInfo core_idinfo of
- info@(HasWorker work_id wrap_arity)
- | wrap_arity == stg_arity -> info
- | otherwise -> pprTrace "ifaceId: arity change:" (ppr id)
- NoWorker
- NoWorker -> NoWorker
-
- has_worker = case worker_info of
- HasWorker _ _ -> True
- other -> False
-
- HasWorker work_id _ = worker_info
-
- ------------ Unfolding --------------
- inline_pragma = inlinePragInfo core_idinfo
- dont_inline = isNeverInlinePrag inline_pragma
- loop_breaker = isLoopBreaker (occInfo core_idinfo)
- bottoming_fn = isBottomingStrictness (strictnessInfo core_idinfo)
-
- unfolding = mkTopUnfolding rhs
- rhs_is_small = not (neverUnfold unfolding)
-
- unfold_info | show_unfold = unfolding
- | otherwise = noUnfolding
-
- show_unfold = not has_worker && -- Not unnecessary
- not bottoming_fn && -- Not necessary
- not dont_inline &&
- not loop_breaker &&
- rhs_is_small && -- Small enough
- okToUnfoldInHiFile rhs -- No casms etc
-
-
- ------------ Extra free Ids --------------
- new_needed_ids | opt_OmitInterfacePragmas = emptyVarSet
- | otherwise = worker_ids `unionVarSet`
- unfold_ids `unionVarSet`
- spec_ids
-
- spec_ids = filterVarSet interestingId (rulesRhsFreeVars (specInfo core_idinfo))
-
- worker_ids | has_worker && interestingId work_id = unitVarSet work_id
- -- Conceivably, the worker might come from
- -- another module
- | otherwise = emptyVarSet
-
- unfold_ids | show_unfold = find_fvs rhs
- | otherwise = emptyVarSet
-
- find_fvs expr = exprSomeFreeVars interestingId expr
-
-interestingId id = isId id && mustHaveLocalBinding id
-\end{code}
-
-
-\begin{code}
-getRules :: [IdCoreRule] -- Orphan rules
- -> [CoreBind] -- Bindings, with rules in the top-level Ids
- -> IdSet -- Ids that are exported, so we need their rules
- -> [IdCoreRule]
-getRules orphan_rules binds emitted
- = orphan_rules ++ local_rules
- where
- local_rules = [ (fn, rule)
- | fn <- bindersOfBinds binds,
- fn `elemVarSet` emitted,
- rule <- rulesRules (idSpecialisation fn),
- not (isBuiltinRule rule),
- -- We can't print builtin rules in interface files
- -- Since they are built in, an importing module
- -- will have access to them anyway
-
- -- Sept 00: I've disabled this test. It doesn't stop many, if any, rules
- -- from coming out, and to make it work properly we need to add ????
- -- (put it back in for now)
- all (`elemVarSet` emitted) (varSetElems (ruleSomeLhsFreeVars interestingId rule))
- -- Spit out a rule only if all its lhs free vars are emitted
- -- This is a good reason not to do it when we emit the Id itself
- ]
-\end{code}
-
-
-%************************************************************************
-%* *
\subsection{Checking if the new interface is up to date
%* *
%************************************************************************
diff ok_so_far pp new_vers old [] = (False, pp, new_vers)
diff ok_so_far pp new_vers [] (nd:nds) = diff False (pp $$ only_new nd) new_vers [] nds
diff ok_so_far pp new_vers (od:ods) (nd:nds)
- = case nameOccName od_name `compare` nameOccName nd_name of
+ = case od_name `compare` nd_name of
LT -> diff False (pp $$ only_old od) new_vers ods (nd:nds)
GT -> diff False (pp $$ only_new nd) new_vers (od:ods) nds
EQ | od `eq_tc` nd -> diff ok_so_far pp new_vers ods nds
import StgSyn
+import CmdLineOpts ( opt_EnsureSplittableC )
import Bag ( Bag, emptyBag, unionBags, unitBag, snocBag, bagToList )
import Id ( mkVanillaId, idType, setIdArityInfo, Id )
import VarSet
import VarEnv
import IdInfo ( exactArity )
import Module ( Module )
-import Name ( mkTopName )
+import Name ( Name, mkGlobalName, mkLocalName )
+import OccName ( mkVarOcc )
import Type ( splitForAllTys, mkForAllTys, mkFunTys, Type )
+import Unique ( Unique )
import UniqSupply ( uniqFromSupply, splitUniqSupply, UniqSupply )
import Util ( zipEqual )
+import SrcLoc ( noSrcLoc )
import Panic ( panic, assertPanic )
\end{code}
where
uniq = uniqFromSupply us
+
+mkTopName :: Unique -> Module -> FAST_STRING -> Name
+ -- Make a top-level name; make it Global if top-level
+ -- things should be externally visible; Local otherwise
+ -- This chap is only used *after* the tidyCore phase
+ -- Notably, it is used during STG lambda lifting
+ --
+ -- We have to make sure that the name is globally unique
+ -- and we don't have tidyCore to help us. So we append
+ -- the unique. Hack! Hack!
+ -- (Used only by the STG lambda lifter.)
+mkTopName uniq mod fs
+ | opt_EnsureSplittableC = mkGlobalName uniq mod occ noSrcLoc
+ | otherwise = mkLocalName uniq occ noSrcLoc
+ where
+ occ = mkVarOcc (_PK_ ((_UNPK_ fs) ++ show uniq))
+
lookUp :: Id -> LiftM (Id,[Id])
lookUp v mod ci us idenv
= case (lookupVarEnv idenv v) of
are globally unique, not simply not-in-scope, which is all that
the simplifier ensures.
+4. If we are going to do object-file splitting, we make ALL top-level
+ names into Globals. Why?
+
+ In certain (prelude only) modules we split up the .hc file into
+ lots of separate little files, which are separately compiled by the C
+ compiler. That gives lots of little .o files. The idea is that if
+ you happen to mention one of them you don't necessarily pull them all
+ in. (Pulling in a piece you don't need can be v bad, because it may
+ mention other pieces you don't need either, and so on.)
+
+ Sadly, splitting up .hc files means that local names (like s234) are
+ now globally visible, which can lead to clashes between two .hc
+ files. So we make them all Global, so they are printed complete
+ with their module name.
+
+ We don't want to do this in CoreTidy, because at that stage we use
+ Global to mean "external" and hence "should appear in interface files".
+ This object-file splitting thing is a code generator matter that we
+ don't want to pollute earlier phases.
NOTE THAT:
newLocalIds :: TopLevelFlag -> StgEnv -> [Id] -> UniqSM (StgEnv, [Id])
newLocalIds top_lev env []
= returnUs (env, [])
+
newLocalIds top_lev env (b:bs)
= newLocalId top_lev env b `thenUs` \ (env', b') ->
newLocalIds top_lev env' bs `thenUs` \ (env'', bs') ->
projects / ghc-hetmet.git / commitdiff