2 % (c) The GRASP/AQUA Project, Glasgow University, 1993-1998
4 \section[Specialise]{Stamping out overloading, and (optionally) polymorphism}
7 -- The above warning supression flag is a temporary kludge.
8 -- While working on this module you are encouraged to remove it and fix
9 -- any warnings in the module. See
10 -- http://hackage.haskell.org/trac/ghc/wiki/Commentary/CodingStyle#Warnings
13 module Specialise ( specProgram ) where
15 #include "HsVersions.h"
17 import DynFlags ( DynFlags, DynFlag(..) )
18 import Id ( Id, idName, idType, mkUserLocal,
19 idInlinePragma, setInlinePragma )
20 import TcType ( Type, mkTyVarTy, tcSplitSigmaTy,
21 tyVarsOfTypes, tyVarsOfTheta, isClassPred,
22 tcCmpType, isUnLiftedType
24 import CoreSubst ( Subst, mkEmptySubst, extendTvSubstList, lookupIdSubst,
25 substBndr, substBndrs, substTy, substInScope,
26 cloneIdBndr, cloneIdBndrs, cloneRecIdBndrs
31 import CoreUtils ( applyTypeToArgs, mkPiTypes )
32 import CoreFVs ( exprFreeVars, exprsFreeVars, idFreeVars )
33 import CoreTidy ( tidyRules )
34 import CoreLint ( showPass, endPass )
35 import Rules ( addIdSpecialisations, mkLocalRule, lookupRule, emptyRuleBase, rulesOfBinds )
36 import PprCore ( pprRules )
37 import UniqSupply ( UniqSupply,
42 import MkId ( voidArgId, realWorldPrimId )
44 import Maybes ( catMaybes, maybeToBool )
45 import ErrUtils ( dumpIfSet_dyn )
53 %************************************************************************
55 \subsection[notes-Specialise]{Implementation notes [SLPJ, Aug 18 1993]}
57 %************************************************************************
59 These notes describe how we implement specialisation to eliminate
62 The specialisation pass works on Core
63 syntax, complete with all the explicit dictionary application,
64 abstraction and construction as added by the type checker. The
65 existing type checker remains largely as it is.
67 One important thought: the {\em types} passed to an overloaded
68 function, and the {\em dictionaries} passed are mutually redundant.
69 If the same function is applied to the same type(s) then it is sure to
70 be applied to the same dictionary(s)---or rather to the same {\em
71 values}. (The arguments might look different but they will evaluate
74 Second important thought: we know that we can make progress by
75 treating dictionary arguments as static and worth specialising on. So
76 we can do without binding-time analysis, and instead specialise on
77 dictionary arguments and no others.
86 and suppose f is overloaded.
88 STEP 1: CALL-INSTANCE COLLECTION
90 We traverse <body>, accumulating all applications of f to types and
93 (Might there be partial applications, to just some of its types and
94 dictionaries? In principle yes, but in practice the type checker only
95 builds applications of f to all its types and dictionaries, so partial
96 applications could only arise as a result of transformation, and even
97 then I think it's unlikely. In any case, we simply don't accumulate such
98 partial applications.)
103 So now we have a collection of calls to f:
107 Notice that f may take several type arguments. To avoid ambiguity, we
108 say that f is called at type t1/t2 and t3/t4.
110 We take equivalence classes using equality of the *types* (ignoring
111 the dictionary args, which as mentioned previously are redundant).
113 STEP 3: SPECIALISATION
115 For each equivalence class, choose a representative (f t1 t2 d1 d2),
116 and create a local instance of f, defined thus:
118 f@t1/t2 = <f_rhs> t1 t2 d1 d2
120 f_rhs presumably has some big lambdas and dictionary lambdas, so lots
121 of simplification will now result. However we don't actually *do* that
122 simplification. Rather, we leave it for the simplifier to do. If we
123 *did* do it, though, we'd get more call instances from the specialised
124 RHS. We can work out what they are by instantiating the call-instance
125 set from f's RHS with the types t1, t2.
127 Add this new id to f's IdInfo, to record that f has a specialised version.
129 Before doing any of this, check that f's IdInfo doesn't already
130 tell us about an existing instance of f at the required type/s.
131 (This might happen if specialisation was applied more than once, or
132 it might arise from user SPECIALIZE pragmas.)
136 Wait a minute! What if f is recursive? Then we can't just plug in
137 its right-hand side, can we?
139 But it's ok. The type checker *always* creates non-recursive definitions
140 for overloaded recursive functions. For example:
142 f x = f (x+x) -- Yes I know its silly
146 f a (d::Num a) = let p = +.sel a d
148 letrec fl (y::a) = fl (p y y)
152 We still have recusion for non-overloaded functions which we
153 speciailise, but the recursive call should get specialised to the
154 same recursive version.
160 All this is crystal clear when the function is applied to *constant
161 types*; that is, types which have no type variables inside. But what if
162 it is applied to non-constant types? Suppose we find a call of f at type
163 t1/t2. There are two possibilities:
165 (a) The free type variables of t1, t2 are in scope at the definition point
166 of f. In this case there's no problem, we proceed just as before. A common
167 example is as follows. Here's the Haskell:
172 After typechecking we have
174 g a (d::Num a) (y::a) = let f b (d'::Num b) (x::b) = +.sel b d' x x
175 in +.sel a d (f a d y) (f a d y)
177 Notice that the call to f is at type type "a"; a non-constant type.
178 Both calls to f are at the same type, so we can specialise to give:
180 g a (d::Num a) (y::a) = let f@a (x::a) = +.sel a d x x
181 in +.sel a d (f@a y) (f@a y)
184 (b) The other case is when the type variables in the instance types
185 are *not* in scope at the definition point of f. The example we are
186 working with above is a good case. There are two instances of (+.sel a d),
187 but "a" is not in scope at the definition of +.sel. Can we do anything?
188 Yes, we can "common them up", a sort of limited common sub-expression deal.
191 g a (d::Num a) (y::a) = let +.sel@a = +.sel a d
192 f@a (x::a) = +.sel@a x x
193 in +.sel@a (f@a y) (f@a y)
195 This can save work, and can't be spotted by the type checker, because
196 the two instances of +.sel weren't originally at the same type.
200 * There are quite a few variations here. For example, the defn of
201 +.sel could be floated ouside the \y, to attempt to gain laziness.
202 It certainly mustn't be floated outside the \d because the d has to
205 * We don't want to inline f_rhs in this case, because
206 that will duplicate code. Just commoning up the call is the point.
208 * Nothing gets added to +.sel's IdInfo.
210 * Don't bother unless the equivalence class has more than one item!
212 Not clear whether this is all worth it. It is of course OK to
213 simply discard call-instances when passing a big lambda.
215 Polymorphism 2 -- Overloading
217 Consider a function whose most general type is
219 f :: forall a b. Ord a => [a] -> b -> b
221 There is really no point in making a version of g at Int/Int and another
222 at Int/Bool, because it's only instancing the type variable "a" which
223 buys us any efficiency. Since g is completely polymorphic in b there
224 ain't much point in making separate versions of g for the different
227 That suggests that we should identify which of g's type variables
228 are constrained (like "a") and which are unconstrained (like "b").
229 Then when taking equivalence classes in STEP 2, we ignore the type args
230 corresponding to unconstrained type variable. In STEP 3 we make
231 polymorphic versions. Thus:
233 f@t1/ = /\b -> <f_rhs> t1 b d1 d2
242 f a (d::Num a) = let g = ...
244 ...(let d1::Ord a = Num.Ord.sel a d in g a d1)...
246 Here, g is only called at one type, but the dictionary isn't in scope at the
247 definition point for g. Usually the type checker would build a
248 definition for d1 which enclosed g, but the transformation system
249 might have moved d1's defn inward. Solution: float dictionary bindings
250 outwards along with call instances.
254 f x = let g p q = p==q
260 Before specialisation, leaving out type abstractions we have
262 f df x = let g :: Eq a => a -> a -> Bool
264 h :: Num a => a -> a -> (a, Bool)
265 h dh r s = let deq = eqFromNum dh
266 in (+ dh r s, g deq r s)
270 After specialising h we get a specialised version of h, like this:
272 h' r s = let deq = eqFromNum df
273 in (+ df r s, g deq r s)
275 But we can't naively make an instance for g from this, because deq is not in scope
276 at the defn of g. Instead, we have to float out the (new) defn of deq
277 to widen its scope. Notice that this floating can't be done in advance -- it only
278 shows up when specialisation is done.
280 User SPECIALIZE pragmas
281 ~~~~~~~~~~~~~~~~~~~~~~~
282 Specialisation pragmas can be digested by the type checker, and implemented
283 by adding extra definitions along with that of f, in the same way as before
285 f@t1/t2 = <f_rhs> t1 t2 d1 d2
287 Indeed the pragmas *have* to be dealt with by the type checker, because
288 only it knows how to build the dictionaries d1 and d2! For example
290 g :: Ord a => [a] -> [a]
291 {-# SPECIALIZE f :: [Tree Int] -> [Tree Int] #-}
293 Here, the specialised version of g is an application of g's rhs to the
294 Ord dictionary for (Tree Int), which only the type checker can conjure
295 up. There might not even *be* one, if (Tree Int) is not an instance of
296 Ord! (All the other specialision has suitable dictionaries to hand
299 Problem. The type checker doesn't have to hand a convenient <f_rhs>, because
300 it is buried in a complex (as-yet-un-desugared) binding group.
303 f@t1/t2 = f* t1 t2 d1 d2
305 where f* is the Id f with an IdInfo which says "inline me regardless!".
306 Indeed all the specialisation could be done in this way.
307 That in turn means that the simplifier has to be prepared to inline absolutely
308 any in-scope let-bound thing.
311 Again, the pragma should permit polymorphism in unconstrained variables:
313 h :: Ord a => [a] -> b -> b
314 {-# SPECIALIZE h :: [Int] -> b -> b #-}
316 We *insist* that all overloaded type variables are specialised to ground types,
317 (and hence there can be no context inside a SPECIALIZE pragma).
318 We *permit* unconstrained type variables to be specialised to
320 - or left as a polymorphic type variable
321 but nothing in between. So
323 {-# SPECIALIZE h :: [Int] -> [c] -> [c] #-}
325 is *illegal*. (It can be handled, but it adds complication, and gains the
329 SPECIALISING INSTANCE DECLARATIONS
330 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
333 instance Foo a => Foo [a] where
335 {-# SPECIALIZE instance Foo [Int] #-}
337 The original instance decl creates a dictionary-function
340 dfun.Foo.List :: forall a. Foo a -> Foo [a]
342 The SPECIALIZE pragma just makes a specialised copy, just as for
343 ordinary function definitions:
345 dfun.Foo.List@Int :: Foo [Int]
346 dfun.Foo.List@Int = dfun.Foo.List Int dFooInt
348 The information about what instance of the dfun exist gets added to
349 the dfun's IdInfo in the same way as a user-defined function too.
352 Automatic instance decl specialisation?
353 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
354 Can instance decls be specialised automatically? It's tricky.
355 We could collect call-instance information for each dfun, but
356 then when we specialised their bodies we'd get new call-instances
357 for ordinary functions; and when we specialised their bodies, we might get
358 new call-instances of the dfuns, and so on. This all arises because of
359 the unrestricted mutual recursion between instance decls and value decls.
361 Still, there's no actual problem; it just means that we may not do all
362 the specialisation we could theoretically do.
364 Furthermore, instance decls are usually exported and used non-locally,
365 so we'll want to compile enough to get those specialisations done.
367 Lastly, there's no such thing as a local instance decl, so we can
368 survive solely by spitting out *usage* information, and then reading that
369 back in as a pragma when next compiling the file. So for now,
370 we only specialise instance decls in response to pragmas.
373 SPITTING OUT USAGE INFORMATION
374 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
376 To spit out usage information we need to traverse the code collecting
377 call-instance information for all imported (non-prelude?) functions
378 and data types. Then we equivalence-class it and spit it out.
380 This is done at the top-level when all the call instances which escape
381 must be for imported functions and data types.
383 *** Not currently done ***
386 Partial specialisation by pragmas
387 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
388 What about partial specialisation:
390 k :: (Ord a, Eq b) => [a] -> b -> b -> [a]
391 {-# SPECIALIZE k :: Eq b => [Int] -> b -> b -> [a] #-}
395 {-# SPECIALIZE k :: Eq b => [Int] -> [b] -> [b] -> [a] #-}
397 Seems quite reasonable. Similar things could be done with instance decls:
399 instance (Foo a, Foo b) => Foo (a,b) where
401 {-# SPECIALIZE instance Foo a => Foo (a,Int) #-}
402 {-# SPECIALIZE instance Foo b => Foo (Int,b) #-}
404 Ho hum. Things are complex enough without this. I pass.
407 Requirements for the simplifer
408 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
409 The simplifier has to be able to take advantage of the specialisation.
411 * When the simplifier finds an application of a polymorphic f, it looks in
412 f's IdInfo in case there is a suitable instance to call instead. This converts
414 f t1 t2 d1 d2 ===> f_t1_t2
416 Note that the dictionaries get eaten up too!
418 * Dictionary selection operations on constant dictionaries must be
421 +.sel Int d ===> +Int
423 The obvious way to do this is in the same way as other specialised
424 calls: +.sel has inside it some IdInfo which tells that if it's applied
425 to the type Int then it should eat a dictionary and transform to +Int.
427 In short, dictionary selectors need IdInfo inside them for constant
430 * Exactly the same applies if a superclass dictionary is being
433 Eq.sel Int d ===> dEqInt
435 * Something similar applies to dictionary construction too. Suppose
436 dfun.Eq.List is the function taking a dictionary for (Eq a) to
437 one for (Eq [a]). Then we want
439 dfun.Eq.List Int d ===> dEq.List_Int
441 Where does the Eq [Int] dictionary come from? It is built in
442 response to a SPECIALIZE pragma on the Eq [a] instance decl.
444 In short, dfun Ids need IdInfo with a specialisation for each
445 constant instance of their instance declaration.
447 All this uses a single mechanism: the SpecEnv inside an Id
450 What does the specialisation IdInfo look like?
451 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
453 The SpecEnv of an Id maps a list of types (the template) to an expression
457 For example, if f has this SpecInfo:
459 [Int, a] -> \d:Ord Int. f' a
461 it means that we can replace the call
463 f Int t ===> (\d. f' t)
465 This chucks one dictionary away and proceeds with the
466 specialised version of f, namely f'.
469 What can't be done this way?
470 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
471 There is no way, post-typechecker, to get a dictionary for (say)
472 Eq a from a dictionary for Eq [a]. So if we find
476 we can't transform to
481 eqList :: (a->a->Bool) -> [a] -> [a] -> Bool
483 Of course, we currently have no way to automatically derive
484 eqList, nor to connect it to the Eq [a] instance decl, but you
485 can imagine that it might somehow be possible. Taking advantage
486 of this is permanently ruled out.
488 Still, this is no great hardship, because we intend to eliminate
489 overloading altogether anyway!
493 A note about non-tyvar dictionaries
494 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
495 Some Ids have types like
497 forall a,b,c. Eq a -> Ord [a] -> tau
499 This seems curious at first, because we usually only have dictionary
500 args whose types are of the form (C a) where a is a type variable.
501 But this doesn't hold for the functions arising from instance decls,
502 which sometimes get arguements with types of form (C (T a)) for some
505 Should we specialise wrt this compound-type dictionary? We used to say
507 "This is a heuristic judgement, as indeed is the fact that we
508 specialise wrt only dictionaries. We choose *not* to specialise
509 wrt compound dictionaries because at the moment the only place
510 they show up is in instance decls, where they are simply plugged
511 into a returned dictionary. So nothing is gained by specialising
514 But it is simpler and more uniform to specialise wrt these dicts too;
515 and in future GHC is likely to support full fledged type signatures
517 f ;: Eq [(a,b)] => ...
520 %************************************************************************
522 \subsubsection{The new specialiser}
524 %************************************************************************
526 Our basic game plan is this. For let(rec) bound function
527 f :: (C a, D c) => (a,b,c,d) -> Bool
529 * Find any specialised calls of f, (f ts ds), where
530 ts are the type arguments t1 .. t4, and
531 ds are the dictionary arguments d1 .. d2.
533 * Add a new definition for f1 (say):
535 f1 = /\ b d -> (..body of f..) t1 b t3 d d1 d2
537 Note that we abstract over the unconstrained type arguments.
541 [t1,b,t3,d] |-> \d1 d2 -> f1 b d
543 to the specialisations of f. This will be used by the
544 simplifier to replace calls
545 (f t1 t2 t3 t4) da db
547 (\d1 d1 -> f1 t2 t4) da db
549 All the stuff about how many dictionaries to discard, and what types
550 to apply the specialised function to, are handled by the fact that the
551 SpecEnv contains a template for the result of the specialisation.
553 We don't build *partial* specialisations for f. For example:
555 f :: Eq a => a -> a -> Bool
556 {-# SPECIALISE f :: (Eq b, Eq c) => (b,c) -> (b,c) -> Bool #-}
558 Here, little is gained by making a specialised copy of f.
559 There's a distinct danger that the specialised version would
560 first build a dictionary for (Eq b, Eq c), and then select the (==)
561 method from it! Even if it didn't, not a great deal is saved.
563 We do, however, generate polymorphic, but not overloaded, specialisations:
565 f :: Eq a => [a] -> b -> b -> b
566 {#- SPECIALISE f :: [Int] -> b -> b -> b #-}
568 Hence, the invariant is this:
570 *** no specialised version is overloaded ***
573 %************************************************************************
575 \subsubsection{The exported function}
577 %************************************************************************
580 specProgram :: DynFlags -> UniqSupply -> [CoreBind] -> IO [CoreBind]
581 specProgram dflags us binds = do
583 showPass dflags "Specialise"
585 let binds' = initSM us (do (binds', uds') <- go binds
586 return (dumpAllDictBinds uds' binds'))
588 endPass dflags "Specialise" Opt_D_dump_spec binds'
590 dumpIfSet_dyn dflags Opt_D_dump_rules "Top-level specialisations"
591 (withPprStyle defaultUserStyle $
592 pprRules (tidyRules emptyTidyEnv (rulesOfBinds binds')))
596 -- We need to start with a Subst that knows all the things
597 -- that are in scope, so that the substitution engine doesn't
598 -- accidentally re-use a unique that's already in use
599 -- Easiest thing is to do it all at once, as if all the top-level
600 -- decls were mutually recursive
601 top_subst = mkEmptySubst (mkInScopeSet (mkVarSet (bindersOfBinds binds)))
603 go [] = return ([], emptyUDs)
604 go (bind:binds) = do (binds', uds) <- go binds
605 (bind', uds') <- specBind top_subst bind uds
606 return (bind' ++ binds', uds')
609 %************************************************************************
611 \subsubsection{@specExpr@: the main function}
613 %************************************************************************
616 specVar :: Subst -> Id -> CoreExpr
617 specVar subst v = lookupIdSubst subst v
619 specExpr :: Subst -> CoreExpr -> SpecM (CoreExpr, UsageDetails)
620 -- We carry a substitution down:
621 -- a) we must clone any binding that might flaot outwards,
622 -- to avoid name clashes
623 -- b) we carry a type substitution to use when analysing
624 -- the RHS of specialised bindings (no type-let!)
626 ---------------- First the easy cases --------------------
627 specExpr subst (Type ty) = return (Type (substTy subst ty), emptyUDs)
628 specExpr subst (Var v) = return (specVar subst v, emptyUDs)
629 specExpr _ (Lit lit) = return (Lit lit, emptyUDs)
630 specExpr subst (Cast e co) = do
631 (e', uds) <- specExpr subst e
632 return ((Cast e' (substTy subst co)), uds)
633 specExpr subst (Note note body) = do
634 (body', uds) <- specExpr subst body
635 return (Note (specNote subst note) body', uds)
638 ---------------- Applications might generate a call instance --------------------
639 specExpr subst expr@(App {})
642 go (App fun arg) args = do (arg', uds_arg) <- specExpr subst arg
643 (fun', uds_app) <- go fun (arg':args)
644 return (App fun' arg', uds_arg `plusUDs` uds_app)
646 go (Var f) args = case specVar subst f of
647 Var f' -> return (Var f', mkCallUDs subst f' args)
648 e' -> return (e', emptyUDs) -- I don't expect this!
649 go other _ = specExpr subst other
651 ---------------- Lambda/case require dumping of usage details --------------------
652 specExpr subst e@(Lam _ _) = do
653 (body', uds) <- specExpr subst' body
654 let (filtered_uds, body'') = dumpUDs bndrs' uds body'
655 return (mkLams bndrs' body'', filtered_uds)
657 (bndrs, body) = collectBinders e
658 (subst', bndrs') = substBndrs subst bndrs
659 -- More efficient to collect a group of binders together all at once
660 -- and we don't want to split a lambda group with dumped bindings
662 specExpr subst (Case scrut case_bndr ty alts) = do
663 (scrut', uds_scrut) <- specExpr subst scrut
664 (alts', uds_alts) <- mapAndCombineSM spec_alt alts
665 return (Case scrut' case_bndr' (substTy subst ty) alts', uds_scrut `plusUDs` uds_alts)
667 (subst_alt, case_bndr') = substBndr subst case_bndr
668 -- No need to clone case binder; it can't float like a let(rec)
670 spec_alt (con, args, rhs) = do
671 (rhs', uds) <- specExpr subst_rhs rhs
672 let (uds', rhs'') = dumpUDs args uds rhs'
673 return ((con, args', rhs''), uds')
675 (subst_rhs, args') = substBndrs subst_alt args
677 ---------------- Finally, let is the interesting case --------------------
678 specExpr subst (Let bind body) = do
680 (rhs_subst, body_subst, bind') <- cloneBindSM subst bind
682 -- Deal with the body
683 (body', body_uds) <- specExpr body_subst body
685 -- Deal with the bindings
686 (binds', uds) <- specBind rhs_subst bind' body_uds
689 return (foldr Let body' binds', uds)
691 -- Must apply the type substitution to coerceions
692 specNote :: Subst -> Note -> Note
693 specNote _ note = note
696 %************************************************************************
698 \subsubsection{Dealing with a binding}
700 %************************************************************************
703 specBind :: Subst -- Use this for RHSs
705 -> UsageDetails -- Info on how the scope of the binding
706 -> SpecM ([CoreBind], -- New bindings
707 UsageDetails) -- And info to pass upstream
709 specBind rhs_subst bind body_uds
710 = do { (bind', bind_uds) <- specBindItself rhs_subst bind (calls body_uds)
711 ; return (finishSpecBind bind' bind_uds body_uds) }
713 finishSpecBind :: CoreBind -> UsageDetails -> UsageDetails -> ([CoreBind], UsageDetails)
715 (MkUD { dict_binds = rhs_dbs, calls = rhs_calls, ud_fvs = rhs_fvs })
716 (MkUD { dict_binds = body_dbs, calls = body_calls, ud_fvs = body_fvs })
717 | not (mkVarSet bndrs `intersectsVarSet` all_fvs)
718 -- Common case 1: the bound variables are not
719 -- mentioned in the dictionary bindings
720 = ([bind], MkUD { dict_binds = body_dbs `unionBags` rhs_dbs
721 -- It's important that the `unionBags` is this way round,
722 -- because body_uds may bind dictionaries that are
723 -- used in the calls passed to specDefn. So the
724 -- dictionary bindings in rhs_uds may mention
725 -- dictionaries bound in body_uds.
727 , ud_fvs = all_fvs })
729 | case bind of { NonRec {} -> True; Rec {} -> False }
730 -- Common case 2: no specialisation happened, and binding
731 -- is non-recursive. But the binding may be
732 -- mentioned in body_dbs, so we should put it first
733 = ([], MkUD { dict_binds = rhs_dbs `unionBags` ((bind, b_fvs) `consBag` body_dbs)
735 , ud_fvs = all_fvs `unionVarSet` b_fvs })
737 | otherwise -- General case: make a huge Rec (sigh)
738 = ([], MkUD { dict_binds = unitBag (Rec all_db_prs, all_db_fvs)
740 , ud_fvs = all_fvs `unionVarSet` b_fvs })
742 all_fvs = rhs_fvs `unionVarSet` body_fvs
743 all_calls = zapCalls bndrs (rhs_calls `unionCalls` body_calls)
745 bndrs = bindersOf bind
746 b_fvs = bind_fvs bind
748 (all_db_prs, all_db_fvs) = add (bind, b_fvs) $
749 foldrBag add ([], emptyVarSet) $
750 rhs_dbs `unionBags` body_dbs
751 add (NonRec b r, b_fvs) (prs, fvs) = ((b,r) : prs, b_fvs `unionVarSet` fvs)
752 add (Rec b_prs, b_fvs) (prs, fvs) = (b_prs ++ prs, b_fvs `unionVarSet` fvs)
754 specBindItself :: Subst -> CoreBind -> CallDetails -> SpecM (CoreBind, UsageDetails)
756 -- specBindItself deals with the RHS, specialising it according
757 -- to the calls found in the body (if any)
758 specBindItself rhs_subst (NonRec bndr rhs) call_info = do
759 ((bndr',rhs'), spec_defns, spec_uds) <- specDefn rhs_subst call_info (bndr,rhs)
761 new_bind | null spec_defns = NonRec bndr' rhs'
762 | otherwise = Rec ((bndr',rhs'):spec_defns)
763 -- bndr' mentions the spec_defns in its SpecEnv
764 -- Not sure why we couln't just put the spec_defns first
765 return (new_bind, spec_uds)
767 specBindItself rhs_subst (Rec pairs) call_info = do
768 stuff <- mapM (specDefn rhs_subst call_info) pairs
770 (pairs', spec_defns_s, spec_uds_s) = unzip3 stuff
771 spec_defns = concat spec_defns_s
772 spec_uds = plusUDList spec_uds_s
773 new_bind = Rec (spec_defns ++ pairs')
774 return (new_bind, spec_uds)
777 specDefn :: Subst -- Subst to use for RHS
778 -> CallDetails -- Info on how it is used in its scope
779 -> (Id, CoreExpr) -- The thing being bound and its un-processed RHS
780 -> SpecM ((Id, CoreExpr), -- The thing and its processed RHS
781 -- the Id may now have specialisations attached
782 [(Id,CoreExpr)], -- Extra, specialised bindings
783 UsageDetails -- Stuff to fling upwards from the RHS and its
784 ) -- specialised versions
786 specDefn subst calls (fn, rhs)
787 -- The first case is the interesting one
788 | rhs_tyvars `lengthIs` n_tyvars -- Rhs of fn's defn has right number of big lambdas
789 && rhs_ids `lengthAtLeast` n_dicts -- and enough dict args
790 && notNull calls_for_me -- And there are some calls to specialise
792 -- && not (certainlyWillInline (idUnfolding fn)) -- And it's not small
793 -- See Note [Inline specialisation] for why we do not
794 -- switch off specialisation for inline functions = do
796 -- Specialise the body of the function
797 (rhs', rhs_uds) <- specExpr subst rhs
799 -- Make a specialised version for each call in calls_for_me
800 stuff <- mapM spec_call calls_for_me
802 (spec_defns, spec_uds, spec_rules) = unzip3 stuff
804 fn' = addIdSpecialisations fn spec_rules
808 rhs_uds `plusUDs` plusUDList spec_uds)
810 | otherwise -- No calls or RHS doesn't fit our preconceptions
811 = WARN( notNull calls_for_me, ptext (sLit "Missed specialisation opportunity for") <+> ppr fn )
812 -- Note [Specialisation shape]
813 (do { (rhs', rhs_uds) <- specExpr subst rhs
814 ; return ((fn, rhs'), [], rhs_uds) })
818 (tyvars, theta, _) = tcSplitSigmaTy fn_type
819 n_tyvars = length tyvars
820 n_dicts = length theta
821 inline_prag = idInlinePragma fn
823 -- It's important that we "see past" any INLINE pragma
824 -- else we'll fail to specialise an INLINE thing
825 (inline_rhs, rhs_inside) = dropInline rhs
826 (rhs_tyvars, rhs_ids, rhs_body) = collectTyAndValBinders rhs_inside
828 rhs_dicts = take n_dicts rhs_ids
829 body = mkLams (drop n_dicts rhs_ids) rhs_body
830 -- Glue back on the non-dict lambdas
832 calls_for_me = case lookupFM calls fn of
834 Just cs -> fmToList cs
836 ----------------------------------------------------------
837 -- Specialise to one particular call pattern
838 spec_call :: (CallKey, ([DictExpr], VarSet)) -- Call instance
839 -> SpecM ((Id,CoreExpr), -- Specialised definition
840 UsageDetails, -- Usage details from specialised body
841 CoreRule) -- Info for the Id's SpecEnv
842 spec_call (CallKey call_ts, (call_ds, _))
843 = ASSERT( call_ts `lengthIs` n_tyvars && call_ds `lengthIs` n_dicts ) do
844 -- Calls are only recorded for properly-saturated applications
846 -- Suppose f's defn is f = /\ a b c d -> \ d1 d2 -> rhs
847 -- Supppose the call is for f [Just t1, Nothing, Just t3, Nothing] [dx1, dx2]
849 -- Construct the new binding
850 -- f1 = SUBST[a->t1,c->t3, d1->d1', d2->d2'] (/\ b d -> rhs)
851 -- PLUS the usage-details
852 -- { d1' = dx1; d2' = dx2 }
853 -- where d1', d2' are cloned versions of d1,d2, with the type substitution applied.
855 -- Note that the substitution is applied to the whole thing.
856 -- This is convenient, but just slightly fragile. Notably:
857 -- * There had better be no name clashes in a/b/c/d
860 -- poly_tyvars = [b,d] in the example above
861 -- spec_tyvars = [a,c]
862 -- ty_args = [t1,b,t3,d]
863 poly_tyvars = [tv | (tv, Nothing) <- rhs_tyvars `zip` call_ts]
864 spec_tyvars = [tv | (tv, Just _) <- rhs_tyvars `zip` call_ts]
865 ty_args = zipWithEqual "spec_call" mk_ty_arg rhs_tyvars call_ts
867 mk_ty_arg rhs_tyvar Nothing = Type (mkTyVarTy rhs_tyvar)
868 mk_ty_arg _ (Just ty) = Type ty
869 rhs_subst = extendTvSubstList subst (spec_tyvars `zip` [ty | Just ty <- call_ts])
871 (rhs_subst', rhs_dicts') <- cloneBinders rhs_subst rhs_dicts
873 inst_args = ty_args ++ map Var rhs_dicts'
875 -- Figure out the type of the specialised function
876 body_ty = applyTypeToArgs rhs fn_type inst_args
877 (lam_args, app_args) -- Add a dummy argument if body_ty is unlifted
878 | isUnLiftedType body_ty -- C.f. WwLib.mkWorkerArgs
879 = (poly_tyvars ++ [voidArgId], poly_tyvars ++ [realWorldPrimId])
880 | otherwise = (poly_tyvars, poly_tyvars)
881 spec_id_ty = mkPiTypes lam_args body_ty
883 spec_f <- newIdSM fn spec_id_ty
884 (spec_rhs, rhs_uds) <- specExpr rhs_subst' (mkLams lam_args body)
886 -- The rule to put in the function's specialisation is:
887 -- forall b,d, d1',d2'. f t1 b t3 d d1' d2' = f1 b d
888 spec_env_rule = mkLocalRule (mkFastString ("SPEC " ++ showSDoc (ppr fn)))
889 inline_prag -- Note [Auto-specialisation and RULES]
891 (poly_tyvars ++ rhs_dicts')
893 (mkVarApps (Var spec_f) app_args)
895 -- Add the { d1' = dx1; d2' = dx2 } usage stuff
896 final_uds = foldr addDictBind rhs_uds (my_zipEqual "spec_call" rhs_dicts' call_ds)
898 spec_pr | inline_rhs = (spec_f `setInlinePragma` inline_prag, Note InlineMe spec_rhs)
899 | otherwise = (spec_f, spec_rhs)
901 return (spec_pr, final_uds, spec_env_rule)
904 my_zipEqual doc xs ys
905 | debugIsOn && not (equalLength xs ys)
906 = pprPanic "my_zipEqual" (vcat
908 , ppr fn <+> ppr call_ts
909 , ppr (idType fn), ppr theta
910 , ppr n_dicts, ppr rhs_dicts
912 | otherwise = zipEqual doc xs ys
915 Note [Auto-specialisation and RULES]
916 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
921 f :: (Int -> Int) -> Int
925 Suppose that auto-specialisation makes a specialised version of
926 g::Int->Int That version won't appear in the LHS of the RULE for f.
927 So if the specialisation rule fires too early, the rule for f may
930 It might be possible to add new rules, to "complete" the rewrite system.
932 RULE forall d. g Int d = g_spec
936 But that's a bit complicated. For now we ask the programmer's help,
937 by *copying the INLINE activation pragma* to the auto-specialised rule.
938 So if g says {-# NOINLINE[2] g #-}, then the auto-spec rule will also
939 not be active until phase 2.
942 Note [Specialisation shape]
943 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
944 We only specialise a function if it has visible top-level lambdas
945 corresponding to its overloading. E.g. if
946 f :: forall a. Eq a => ....
947 then its body must look like
950 Reason: when specialising the body for a call (f ty dexp), we want to
951 substitute dexp for d, and pick up specialised calls in the body of f.
953 This doesn't always work. One example I came across was htis:
954 newtype Gen a = MkGen{ unGen :: Int -> a }
956 choose :: Eq a => a -> Gen a
957 choose n = MkGen (\r -> n)
959 oneof = choose (1::Int)
961 It's a silly exapmle, but we get
962 choose = /\a. g `cast` co
963 where choose doesn't have any dict arguments. Thus far I have not
964 tried to fix this (wait till there's a real example).
967 Note [Inline specialisations]
968 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
969 We transfer to the specialised function any INLINE stuff from the
970 original. This means (a) the Activation in the IdInfo, and (b) any
973 This is a change (Jun06). Previously the idea is that the point of
974 inlining was precisely to specialise the function at its call site,
975 and that's not so important for the specialised copies. But
976 *pragma-directed* specialisation now takes place in the
977 typechecker/desugarer, with manually specified INLINEs. The
978 specialiation here is automatic. It'd be very odd if a function
979 marked INLINE was specialised (because of some local use), and then
980 forever after (including importing modules) the specialised version
981 wasn't INLINEd. After all, the programmer said INLINE!
983 You might wonder why we don't just not specialise INLINE functions.
984 It's because even INLINE functions are sometimes not inlined, when
985 they aren't applied to interesting arguments. But perhaps the type
986 arguments alone are enough to specialise (even though the args are too
987 boring to trigger inlining), and it's certainly better to call the
990 A case in point is dictionary functions, which are current marked
991 INLINE, but which are worth specialising.
994 dropInline :: CoreExpr -> (Bool, CoreExpr)
995 dropInline (Note InlineMe rhs) = (True, rhs)
996 dropInline rhs = (False, rhs)
999 %************************************************************************
1001 \subsubsection{UsageDetails and suchlike}
1003 %************************************************************************
1008 dict_binds :: !(Bag DictBind),
1009 -- Floated dictionary bindings
1010 -- The order is important;
1011 -- in ds1 `union` ds2, bindings in ds2 can depend on those in ds1
1012 -- (Remember, Bags preserve order in GHC.)
1014 calls :: !CallDetails,
1016 ud_fvs :: !VarSet -- A superset of the variables mentioned in
1017 -- either dict_binds or calls
1020 instance Outputable UsageDetails where
1021 ppr (MkUD { dict_binds = dbs, calls = calls, ud_fvs = fvs })
1022 = ptext (sLit "MkUD") <+> braces (sep (punctuate comma
1023 [ptext (sLit "binds") <+> equals <+> ppr dbs,
1024 ptext (sLit "calls") <+> equals <+> ppr calls,
1025 ptext (sLit "fvs") <+> equals <+> ppr fvs]))
1027 type DictBind = (CoreBind, VarSet)
1028 -- The set is the free vars of the binding
1029 -- both tyvars and dicts
1031 type DictExpr = CoreExpr
1033 emptyUDs :: UsageDetails
1034 emptyUDs = MkUD { dict_binds = emptyBag, calls = emptyFM, ud_fvs = emptyVarSet }
1036 ------------------------------------------------------------
1037 type CallDetails = FiniteMap Id CallInfo
1038 newtype CallKey = CallKey [Maybe Type] -- Nothing => unconstrained type argument
1039 type CallInfo = FiniteMap CallKey
1040 ([DictExpr], VarSet) -- Dict args and the vars of the whole
1041 -- call (including tyvars)
1042 -- [*not* include the main id itself, of course]
1043 -- The finite maps eliminate duplicates
1044 -- The list of types and dictionaries is guaranteed to
1045 -- match the type of f
1047 instance Outputable CallKey where
1048 ppr (CallKey ts) = ppr ts
1050 -- Type isn't an instance of Ord, so that we can control which
1051 -- instance we use. That's tiresome here. Oh well
1052 instance Eq CallKey where
1053 k1 == k2 = case k1 `compare` k2 of { EQ -> True; _ -> False }
1055 instance Ord CallKey where
1056 compare (CallKey k1) (CallKey k2) = cmpList cmp k1 k2
1058 cmp Nothing Nothing = EQ
1059 cmp Nothing (Just _) = LT
1060 cmp (Just _) Nothing = GT
1061 cmp (Just t1) (Just t2) = tcCmpType t1 t2
1063 unionCalls :: CallDetails -> CallDetails -> CallDetails
1064 unionCalls c1 c2 = plusFM_C plusFM c1 c2
1066 singleCall :: Id -> [Maybe Type] -> [DictExpr] -> UsageDetails
1067 singleCall id tys dicts
1068 = MkUD {dict_binds = emptyBag,
1069 calls = unitFM id (unitFM (CallKey tys) (dicts, call_fvs)),
1072 call_fvs = exprsFreeVars dicts `unionVarSet` tys_fvs
1073 tys_fvs = tyVarsOfTypes (catMaybes tys)
1074 -- The type args (tys) are guaranteed to be part of the dictionary
1075 -- types, because they are just the constrained types,
1076 -- and the dictionary is therefore sure to be bound
1077 -- inside the binding for any type variables free in the type;
1078 -- hence it's safe to neglect tyvars free in tys when making
1079 -- the free-var set for this call
1080 -- BUT I don't trust this reasoning; play safe and include tys_fvs
1082 -- We don't include the 'id' itself.
1084 mkCallUDs :: Subst -> Id -> [CoreExpr] -> UsageDetails
1085 mkCallUDs subst f args
1087 || not (all isClassPred theta)
1088 -- Only specialise if all overloading is on class params.
1089 -- In ptic, with implicit params, the type args
1090 -- *don't* say what the value of the implicit param is!
1091 || not (spec_tys `lengthIs` n_tyvars)
1092 || not ( dicts `lengthIs` n_dicts)
1093 || maybeToBool (lookupRule (\_act -> True) (substInScope subst) emptyRuleBase f args)
1094 -- There's already a rule covering this call. A typical case
1095 -- is where there's an explicit user-provided rule. Then
1096 -- we don't want to create a specialised version
1097 -- of the function that overlaps.
1098 = emptyUDs -- Not overloaded, or no specialisation wanted
1101 = singleCall f spec_tys dicts
1103 (tyvars, theta, _) = tcSplitSigmaTy (idType f)
1104 constrained_tyvars = tyVarsOfTheta theta
1105 n_tyvars = length tyvars
1106 n_dicts = length theta
1108 spec_tys = [mk_spec_ty tv ty | (tv, Type ty) <- tyvars `zip` args]
1109 dicts = [dict_expr | (_, dict_expr) <- theta `zip` (drop n_tyvars args)]
1112 | tyvar `elemVarSet` constrained_tyvars = Just ty
1113 | otherwise = Nothing
1115 ------------------------------------------------------------
1116 plusUDs :: UsageDetails -> UsageDetails -> UsageDetails
1117 plusUDs (MkUD {dict_binds = db1, calls = calls1, ud_fvs = fvs1})
1118 (MkUD {dict_binds = db2, calls = calls2, ud_fvs = fvs2})
1119 = MkUD {dict_binds = d, calls = c, ud_fvs = fvs1 `unionVarSet` fvs2}
1121 d = db1 `unionBags` db2
1122 c = calls1 `unionCalls` calls2
1124 plusUDList :: [UsageDetails] -> UsageDetails
1125 plusUDList = foldr plusUDs emptyUDs
1127 -- zapCalls deletes calls to ids from uds
1128 zapCalls :: [Id] -> CallDetails -> CallDetails
1129 zapCalls ids calls = delListFromFM calls ids
1131 mkDB :: CoreBind -> DictBind
1132 mkDB bind = (bind, bind_fvs bind)
1134 bind_fvs :: CoreBind -> VarSet
1135 bind_fvs (NonRec bndr rhs) = pair_fvs (bndr,rhs)
1136 bind_fvs (Rec prs) = foldl delVarSet rhs_fvs bndrs
1139 rhs_fvs = unionVarSets (map pair_fvs prs)
1141 pair_fvs :: (Id, CoreExpr) -> VarSet
1142 pair_fvs (bndr, rhs) = exprFreeVars rhs `unionVarSet` idFreeVars bndr
1143 -- Don't forget variables mentioned in the
1144 -- rules of the bndr. C.f. OccAnal.addRuleUsage
1145 -- Also tyvars mentioned in its type; they may not appear in the RHS
1149 addDictBind :: (Id,CoreExpr) -> UsageDetails -> UsageDetails
1150 addDictBind (dict,rhs) uds
1151 = uds { dict_binds = db `consBag` dict_binds uds
1152 , ud_fvs = ud_fvs uds `unionVarSet` fvs }
1154 db@(_, fvs) = mkDB (NonRec dict rhs)
1156 dumpAllDictBinds :: UsageDetails -> [CoreBind] -> [CoreBind]
1157 dumpAllDictBinds (MkUD {dict_binds = dbs}) binds
1158 = foldrBag add binds dbs
1160 add (bind,_) binds = bind : binds
1162 dumpUDs :: [CoreBndr]
1163 -> UsageDetails -> CoreExpr
1164 -> (UsageDetails, CoreExpr)
1165 dumpUDs bndrs (MkUD { dict_binds = orig_dbs
1166 , calls = orig_calls
1167 , ud_fvs = fvs}) body
1168 = (new_uds, foldrBag add_let body dump_dbs)
1169 -- This may delete fewer variables
1170 -- than in priciple possible
1173 MkUD { dict_binds = free_dbs
1174 , calls = free_calls
1175 , ud_fvs = fvs `minusVarSet` bndr_set}
1177 bndr_set = mkVarSet bndrs
1178 add_let (bind,_) body = Let bind body
1180 (free_dbs, dump_dbs, dump_set)
1181 = foldlBag dump_db (emptyBag, emptyBag, bndr_set) orig_dbs
1182 -- Important that it's foldl not foldr;
1183 -- we're accumulating the set of dumped ids in dump_set
1185 free_calls = filterCalls dump_set orig_calls
1187 dump_db (free_dbs, dump_dbs, dump_idset) db@(bind, fvs)
1188 | dump_idset `intersectsVarSet` fvs -- Dump it
1189 = (free_dbs, dump_dbs `snocBag` db,
1190 extendVarSetList dump_idset (bindersOf bind))
1192 | otherwise -- Don't dump it
1193 = (free_dbs `snocBag` db, dump_dbs, dump_idset)
1195 filterCalls :: VarSet -> CallDetails -> CallDetails
1196 -- Remove any calls that mention the variables
1197 filterCalls bs calls
1198 = mapFM (\_ cs -> filter_calls cs) $
1199 filterFM (\k _ -> not (k `elemVarSet` bs)) calls
1201 filter_calls :: CallInfo -> CallInfo
1202 filter_calls = filterFM (\_ (_, fvs) -> not (fvs `intersectsVarSet` bs))
1206 %************************************************************************
1208 \subsubsection{Boring helper functions}
1210 %************************************************************************
1213 type SpecM a = UniqSM a
1215 initSM :: UniqSupply -> SpecM a -> a
1218 mapAndCombineSM :: (a -> SpecM (b, UsageDetails)) -> [a] -> SpecM ([b], UsageDetails)
1219 mapAndCombineSM _ [] = return ([], emptyUDs)
1220 mapAndCombineSM f (x:xs) = do (y, uds1) <- f x
1221 (ys, uds2) <- mapAndCombineSM f xs
1222 return (y:ys, uds1 `plusUDs` uds2)
1224 cloneBindSM :: Subst -> CoreBind -> SpecM (Subst, Subst, CoreBind)
1225 -- Clone the binders of the bind; return new bind with the cloned binders
1226 -- Return the substitution to use for RHSs, and the one to use for the body
1227 cloneBindSM subst (NonRec bndr rhs) = do
1228 us <- getUniqueSupplyM
1229 let (subst', bndr') = cloneIdBndr subst us bndr
1230 return (subst, subst', NonRec bndr' rhs)
1232 cloneBindSM subst (Rec pairs) = do
1233 us <- getUniqueSupplyM
1234 let (subst', bndrs') = cloneRecIdBndrs subst us (map fst pairs)
1235 return (subst', subst', Rec (bndrs' `zip` map snd pairs))
1237 cloneBinders :: Subst -> [CoreBndr] -> SpecM (Subst, [CoreBndr])
1238 cloneBinders subst bndrs = do
1239 us <- getUniqueSupplyM
1240 return (cloneIdBndrs subst us bndrs)
1242 newIdSM :: Id -> Type -> SpecM Id
1243 newIdSM old_id new_ty = do
1246 -- Give the new Id a similar occurrence name to the old one
1247 name = idName old_id
1248 new_id = mkUserLocal (mkSpecOcc (nameOccName name)) uniq new_ty (getSrcSpan name)
1253 Old (but interesting) stuff about unboxed bindings
1254 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1256 What should we do when a value is specialised to a *strict* unboxed value?
1258 map_*_* f (x:xs) = let h = f x
1262 Could convert let to case:
1264 map_*_Int# f (x:xs) = case f x of h# ->
1268 This may be undesirable since it forces evaluation here, but the value
1269 may not be used in all branches of the body. In the general case this
1270 transformation is impossible since the mutual recursion in a letrec
1271 cannot be expressed as a case.
1273 There is also a problem with top-level unboxed values, since our
1274 implementation cannot handle unboxed values at the top level.
1276 Solution: Lift the binding of the unboxed value and extract it when it
1279 map_*_Int# f (x:xs) = let h = case (f x) of h# -> _Lift h#
1284 Now give it to the simplifier and the _Lifting will be optimised away.
1286 The benfit is that we have given the specialised "unboxed" values a
1287 very simplep lifted semantics and then leave it up to the simplifier to
1288 optimise it --- knowing that the overheads will be removed in nearly
1291 In particular, the value will only be evaluted in the branches of the
1292 program which use it, rather than being forced at the point where the
1293 value is bound. For example:
1295 filtermap_*_* p f (x:xs)
1302 filtermap_*_Int# p f (x:xs)
1303 = let h = case (f x) of h# -> _Lift h#
1306 True -> case h of _Lift h#
1310 The binding for h can still be inlined in the one branch and the
1311 _Lifting eliminated.
1314 Question: When won't the _Lifting be eliminated?
1316 Answer: When they at the top-level (where it is necessary) or when
1317 inlining would duplicate work (or possibly code depending on
1318 options). However, the _Lifting will still be eliminated if the
1319 strictness analyser deems the lifted binding strict.