2 % (c) The GRASP/AQUA Project, Glasgow University, 1993-1998
4 \section[Specialise]{Stamping out overloading, and (optionally) polymorphism}
7 module Specialise ( specProgram ) where
9 #include "HsVersions.h"
11 import DynFlags ( DynFlags, DynFlag(..) )
12 import Id ( Id, idName, idType, mkUserLocal )
13 import TcType ( Type, mkTyVarTy, tcSplitSigmaTy,
14 tyVarsOfTypes, tyVarsOfTheta, isClassPred,
15 tcCmpType, isUnLiftedType
17 import CoreSubst ( Subst, mkEmptySubst, extendTvSubstList, lookupIdSubst,
18 substBndr, substBndrs, substTy, substInScope,
19 cloneIdBndr, cloneIdBndrs, cloneRecIdBndrs
24 import CoreUtils ( applyTypeToArgs, mkPiTypes )
25 import CoreFVs ( exprFreeVars, exprsFreeVars, idRuleVars )
26 import CoreTidy ( tidyRules )
27 import CoreLint ( showPass, endPass )
28 import Rules ( addIdSpecialisations, mkLocalRule, lookupRule, emptyRuleBase, rulesOfBinds )
29 import PprCore ( pprRules )
30 import UniqSupply ( UniqSupply,
31 UniqSM, initUs_, thenUs, returnUs, getUniqueUs,
34 import Name ( nameOccName, mkSpecOcc, getSrcLoc )
35 import MkId ( voidArgId, realWorldPrimId )
37 import Maybes ( catMaybes, maybeToBool )
38 import ErrUtils ( dumpIfSet_dyn )
39 import BasicTypes ( Activation( AlwaysActive ) )
41 import List ( partition )
42 import Util ( zipEqual, zipWithEqual, cmpList, lengthIs,
43 equalLength, lengthAtLeast, notNull )
50 %************************************************************************
52 \subsection[notes-Specialise]{Implementation notes [SLPJ, Aug 18 1993]}
54 %************************************************************************
56 These notes describe how we implement specialisation to eliminate
59 The specialisation pass works on Core
60 syntax, complete with all the explicit dictionary application,
61 abstraction and construction as added by the type checker. The
62 existing type checker remains largely as it is.
64 One important thought: the {\em types} passed to an overloaded
65 function, and the {\em dictionaries} passed are mutually redundant.
66 If the same function is applied to the same type(s) then it is sure to
67 be applied to the same dictionary(s)---or rather to the same {\em
68 values}. (The arguments might look different but they will evaluate
71 Second important thought: we know that we can make progress by
72 treating dictionary arguments as static and worth specialising on. So
73 we can do without binding-time analysis, and instead specialise on
74 dictionary arguments and no others.
83 and suppose f is overloaded.
85 STEP 1: CALL-INSTANCE COLLECTION
87 We traverse <body>, accumulating all applications of f to types and
90 (Might there be partial applications, to just some of its types and
91 dictionaries? In principle yes, but in practice the type checker only
92 builds applications of f to all its types and dictionaries, so partial
93 applications could only arise as a result of transformation, and even
94 then I think it's unlikely. In any case, we simply don't accumulate such
95 partial applications.)
100 So now we have a collection of calls to f:
104 Notice that f may take several type arguments. To avoid ambiguity, we
105 say that f is called at type t1/t2 and t3/t4.
107 We take equivalence classes using equality of the *types* (ignoring
108 the dictionary args, which as mentioned previously are redundant).
110 STEP 3: SPECIALISATION
112 For each equivalence class, choose a representative (f t1 t2 d1 d2),
113 and create a local instance of f, defined thus:
115 f@t1/t2 = <f_rhs> t1 t2 d1 d2
117 f_rhs presumably has some big lambdas and dictionary lambdas, so lots
118 of simplification will now result. However we don't actually *do* that
119 simplification. Rather, we leave it for the simplifier to do. If we
120 *did* do it, though, we'd get more call instances from the specialised
121 RHS. We can work out what they are by instantiating the call-instance
122 set from f's RHS with the types t1, t2.
124 Add this new id to f's IdInfo, to record that f has a specialised version.
126 Before doing any of this, check that f's IdInfo doesn't already
127 tell us about an existing instance of f at the required type/s.
128 (This might happen if specialisation was applied more than once, or
129 it might arise from user SPECIALIZE pragmas.)
133 Wait a minute! What if f is recursive? Then we can't just plug in
134 its right-hand side, can we?
136 But it's ok. The type checker *always* creates non-recursive definitions
137 for overloaded recursive functions. For example:
139 f x = f (x+x) -- Yes I know its silly
143 f a (d::Num a) = let p = +.sel a d
145 letrec fl (y::a) = fl (p y y)
149 We still have recusion for non-overloaded functions which we
150 speciailise, but the recursive call should get specialised to the
151 same recursive version.
157 All this is crystal clear when the function is applied to *constant
158 types*; that is, types which have no type variables inside. But what if
159 it is applied to non-constant types? Suppose we find a call of f at type
160 t1/t2. There are two possibilities:
162 (a) The free type variables of t1, t2 are in scope at the definition point
163 of f. In this case there's no problem, we proceed just as before. A common
164 example is as follows. Here's the Haskell:
169 After typechecking we have
171 g a (d::Num a) (y::a) = let f b (d'::Num b) (x::b) = +.sel b d' x x
172 in +.sel a d (f a d y) (f a d y)
174 Notice that the call to f is at type type "a"; a non-constant type.
175 Both calls to f are at the same type, so we can specialise to give:
177 g a (d::Num a) (y::a) = let f@a (x::a) = +.sel a d x x
178 in +.sel a d (f@a y) (f@a y)
181 (b) The other case is when the type variables in the instance types
182 are *not* in scope at the definition point of f. The example we are
183 working with above is a good case. There are two instances of (+.sel a d),
184 but "a" is not in scope at the definition of +.sel. Can we do anything?
185 Yes, we can "common them up", a sort of limited common sub-expression deal.
188 g a (d::Num a) (y::a) = let +.sel@a = +.sel a d
189 f@a (x::a) = +.sel@a x x
190 in +.sel@a (f@a y) (f@a y)
192 This can save work, and can't be spotted by the type checker, because
193 the two instances of +.sel weren't originally at the same type.
197 * There are quite a few variations here. For example, the defn of
198 +.sel could be floated ouside the \y, to attempt to gain laziness.
199 It certainly mustn't be floated outside the \d because the d has to
202 * We don't want to inline f_rhs in this case, because
203 that will duplicate code. Just commoning up the call is the point.
205 * Nothing gets added to +.sel's IdInfo.
207 * Don't bother unless the equivalence class has more than one item!
209 Not clear whether this is all worth it. It is of course OK to
210 simply discard call-instances when passing a big lambda.
212 Polymorphism 2 -- Overloading
214 Consider a function whose most general type is
216 f :: forall a b. Ord a => [a] -> b -> b
218 There is really no point in making a version of g at Int/Int and another
219 at Int/Bool, because it's only instancing the type variable "a" which
220 buys us any efficiency. Since g is completely polymorphic in b there
221 ain't much point in making separate versions of g for the different
224 That suggests that we should identify which of g's type variables
225 are constrained (like "a") and which are unconstrained (like "b").
226 Then when taking equivalence classes in STEP 2, we ignore the type args
227 corresponding to unconstrained type variable. In STEP 3 we make
228 polymorphic versions. Thus:
230 f@t1/ = /\b -> <f_rhs> t1 b d1 d2
239 f a (d::Num a) = let g = ...
241 ...(let d1::Ord a = Num.Ord.sel a d in g a d1)...
243 Here, g is only called at one type, but the dictionary isn't in scope at the
244 definition point for g. Usually the type checker would build a
245 definition for d1 which enclosed g, but the transformation system
246 might have moved d1's defn inward. Solution: float dictionary bindings
247 outwards along with call instances.
251 f x = let g p q = p==q
257 Before specialisation, leaving out type abstractions we have
259 f df x = let g :: Eq a => a -> a -> Bool
261 h :: Num a => a -> a -> (a, Bool)
262 h dh r s = let deq = eqFromNum dh
263 in (+ dh r s, g deq r s)
267 After specialising h we get a specialised version of h, like this:
269 h' r s = let deq = eqFromNum df
270 in (+ df r s, g deq r s)
272 But we can't naively make an instance for g from this, because deq is not in scope
273 at the defn of g. Instead, we have to float out the (new) defn of deq
274 to widen its scope. Notice that this floating can't be done in advance -- it only
275 shows up when specialisation is done.
277 User SPECIALIZE pragmas
278 ~~~~~~~~~~~~~~~~~~~~~~~
279 Specialisation pragmas can be digested by the type checker, and implemented
280 by adding extra definitions along with that of f, in the same way as before
282 f@t1/t2 = <f_rhs> t1 t2 d1 d2
284 Indeed the pragmas *have* to be dealt with by the type checker, because
285 only it knows how to build the dictionaries d1 and d2! For example
287 g :: Ord a => [a] -> [a]
288 {-# SPECIALIZE f :: [Tree Int] -> [Tree Int] #-}
290 Here, the specialised version of g is an application of g's rhs to the
291 Ord dictionary for (Tree Int), which only the type checker can conjure
292 up. There might not even *be* one, if (Tree Int) is not an instance of
293 Ord! (All the other specialision has suitable dictionaries to hand
296 Problem. The type checker doesn't have to hand a convenient <f_rhs>, because
297 it is buried in a complex (as-yet-un-desugared) binding group.
300 f@t1/t2 = f* t1 t2 d1 d2
302 where f* is the Id f with an IdInfo which says "inline me regardless!".
303 Indeed all the specialisation could be done in this way.
304 That in turn means that the simplifier has to be prepared to inline absolutely
305 any in-scope let-bound thing.
308 Again, the pragma should permit polymorphism in unconstrained variables:
310 h :: Ord a => [a] -> b -> b
311 {-# SPECIALIZE h :: [Int] -> b -> b #-}
313 We *insist* that all overloaded type variables are specialised to ground types,
314 (and hence there can be no context inside a SPECIALIZE pragma).
315 We *permit* unconstrained type variables to be specialised to
317 - or left as a polymorphic type variable
318 but nothing in between. So
320 {-# SPECIALIZE h :: [Int] -> [c] -> [c] #-}
322 is *illegal*. (It can be handled, but it adds complication, and gains the
326 SPECIALISING INSTANCE DECLARATIONS
327 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
330 instance Foo a => Foo [a] where
332 {-# SPECIALIZE instance Foo [Int] #-}
334 The original instance decl creates a dictionary-function
337 dfun.Foo.List :: forall a. Foo a -> Foo [a]
339 The SPECIALIZE pragma just makes a specialised copy, just as for
340 ordinary function definitions:
342 dfun.Foo.List@Int :: Foo [Int]
343 dfun.Foo.List@Int = dfun.Foo.List Int dFooInt
345 The information about what instance of the dfun exist gets added to
346 the dfun's IdInfo in the same way as a user-defined function too.
349 Automatic instance decl specialisation?
350 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
351 Can instance decls be specialised automatically? It's tricky.
352 We could collect call-instance information for each dfun, but
353 then when we specialised their bodies we'd get new call-instances
354 for ordinary functions; and when we specialised their bodies, we might get
355 new call-instances of the dfuns, and so on. This all arises because of
356 the unrestricted mutual recursion between instance decls and value decls.
358 Still, there's no actual problem; it just means that we may not do all
359 the specialisation we could theoretically do.
361 Furthermore, instance decls are usually exported and used non-locally,
362 so we'll want to compile enough to get those specialisations done.
364 Lastly, there's no such thing as a local instance decl, so we can
365 survive solely by spitting out *usage* information, and then reading that
366 back in as a pragma when next compiling the file. So for now,
367 we only specialise instance decls in response to pragmas.
370 SPITTING OUT USAGE INFORMATION
371 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
373 To spit out usage information we need to traverse the code collecting
374 call-instance information for all imported (non-prelude?) functions
375 and data types. Then we equivalence-class it and spit it out.
377 This is done at the top-level when all the call instances which escape
378 must be for imported functions and data types.
380 *** Not currently done ***
383 Partial specialisation by pragmas
384 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
385 What about partial specialisation:
387 k :: (Ord a, Eq b) => [a] -> b -> b -> [a]
388 {-# SPECIALIZE k :: Eq b => [Int] -> b -> b -> [a] #-}
392 {-# SPECIALIZE k :: Eq b => [Int] -> [b] -> [b] -> [a] #-}
394 Seems quite reasonable. Similar things could be done with instance decls:
396 instance (Foo a, Foo b) => Foo (a,b) where
398 {-# SPECIALIZE instance Foo a => Foo (a,Int) #-}
399 {-# SPECIALIZE instance Foo b => Foo (Int,b) #-}
401 Ho hum. Things are complex enough without this. I pass.
404 Requirements for the simplifer
405 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
406 The simplifier has to be able to take advantage of the specialisation.
408 * When the simplifier finds an application of a polymorphic f, it looks in
409 f's IdInfo in case there is a suitable instance to call instead. This converts
411 f t1 t2 d1 d2 ===> f_t1_t2
413 Note that the dictionaries get eaten up too!
415 * Dictionary selection operations on constant dictionaries must be
418 +.sel Int d ===> +Int
420 The obvious way to do this is in the same way as other specialised
421 calls: +.sel has inside it some IdInfo which tells that if it's applied
422 to the type Int then it should eat a dictionary and transform to +Int.
424 In short, dictionary selectors need IdInfo inside them for constant
427 * Exactly the same applies if a superclass dictionary is being
430 Eq.sel Int d ===> dEqInt
432 * Something similar applies to dictionary construction too. Suppose
433 dfun.Eq.List is the function taking a dictionary for (Eq a) to
434 one for (Eq [a]). Then we want
436 dfun.Eq.List Int d ===> dEq.List_Int
438 Where does the Eq [Int] dictionary come from? It is built in
439 response to a SPECIALIZE pragma on the Eq [a] instance decl.
441 In short, dfun Ids need IdInfo with a specialisation for each
442 constant instance of their instance declaration.
444 All this uses a single mechanism: the SpecEnv inside an Id
447 What does the specialisation IdInfo look like?
448 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
450 The SpecEnv of an Id maps a list of types (the template) to an expression
454 For example, if f has this SpecInfo:
456 [Int, a] -> \d:Ord Int. f' a
458 it means that we can replace the call
460 f Int t ===> (\d. f' t)
462 This chucks one dictionary away and proceeds with the
463 specialised version of f, namely f'.
466 What can't be done this way?
467 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
468 There is no way, post-typechecker, to get a dictionary for (say)
469 Eq a from a dictionary for Eq [a]. So if we find
473 we can't transform to
478 eqList :: (a->a->Bool) -> [a] -> [a] -> Bool
480 Of course, we currently have no way to automatically derive
481 eqList, nor to connect it to the Eq [a] instance decl, but you
482 can imagine that it might somehow be possible. Taking advantage
483 of this is permanently ruled out.
485 Still, this is no great hardship, because we intend to eliminate
486 overloading altogether anyway!
490 A note about non-tyvar dictionaries
491 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
492 Some Ids have types like
494 forall a,b,c. Eq a -> Ord [a] -> tau
496 This seems curious at first, because we usually only have dictionary
497 args whose types are of the form (C a) where a is a type variable.
498 But this doesn't hold for the functions arising from instance decls,
499 which sometimes get arguements with types of form (C (T a)) for some
502 Should we specialise wrt this compound-type dictionary? We used to say
504 "This is a heuristic judgement, as indeed is the fact that we
505 specialise wrt only dictionaries. We choose *not* to specialise
506 wrt compound dictionaries because at the moment the only place
507 they show up is in instance decls, where they are simply plugged
508 into a returned dictionary. So nothing is gained by specialising
511 But it is simpler and more uniform to specialise wrt these dicts too;
512 and in future GHC is likely to support full fledged type signatures
514 f ;: Eq [(a,b)] => ...
517 %************************************************************************
519 \subsubsection{The new specialiser}
521 %************************************************************************
523 Our basic game plan is this. For let(rec) bound function
524 f :: (C a, D c) => (a,b,c,d) -> Bool
526 * Find any specialised calls of f, (f ts ds), where
527 ts are the type arguments t1 .. t4, and
528 ds are the dictionary arguments d1 .. d2.
530 * Add a new definition for f1 (say):
532 f1 = /\ b d -> (..body of f..) t1 b t3 d d1 d2
534 Note that we abstract over the unconstrained type arguments.
538 [t1,b,t3,d] |-> \d1 d2 -> f1 b d
540 to the specialisations of f. This will be used by the
541 simplifier to replace calls
542 (f t1 t2 t3 t4) da db
544 (\d1 d1 -> f1 t2 t4) da db
546 All the stuff about how many dictionaries to discard, and what types
547 to apply the specialised function to, are handled by the fact that the
548 SpecEnv contains a template for the result of the specialisation.
550 We don't build *partial* specialisations for f. For example:
552 f :: Eq a => a -> a -> Bool
553 {-# SPECIALISE f :: (Eq b, Eq c) => (b,c) -> (b,c) -> Bool #-}
555 Here, little is gained by making a specialised copy of f.
556 There's a distinct danger that the specialised version would
557 first build a dictionary for (Eq b, Eq c), and then select the (==)
558 method from it! Even if it didn't, not a great deal is saved.
560 We do, however, generate polymorphic, but not overloaded, specialisations:
562 f :: Eq a => [a] -> b -> b -> b
563 {#- SPECIALISE f :: [Int] -> b -> b -> b #-}
565 Hence, the invariant is this:
567 *** no specialised version is overloaded ***
570 %************************************************************************
572 \subsubsection{The exported function}
574 %************************************************************************
577 specProgram :: DynFlags -> UniqSupply -> [CoreBind] -> IO [CoreBind]
578 specProgram dflags us binds
580 showPass dflags "Specialise"
582 let binds' = initSM us (go binds `thenSM` \ (binds', uds') ->
583 returnSM (dumpAllDictBinds uds' binds'))
585 endPass dflags "Specialise" Opt_D_dump_spec binds'
587 dumpIfSet_dyn dflags Opt_D_dump_rules "Top-level specialisations"
588 (pprRules (tidyRules emptyTidyEnv (rulesOfBinds binds')))
592 -- We need to start with a Subst that knows all the things
593 -- that are in scope, so that the substitution engine doesn't
594 -- accidentally re-use a unique that's already in use
595 -- Easiest thing is to do it all at once, as if all the top-level
596 -- decls were mutually recursive
597 top_subst = mkEmptySubst (mkInScopeSet (mkVarSet (bindersOfBinds binds)))
599 go [] = returnSM ([], emptyUDs)
600 go (bind:binds) = go binds `thenSM` \ (binds', uds) ->
601 specBind top_subst bind uds `thenSM` \ (bind', uds') ->
602 returnSM (bind' ++ binds', uds')
605 %************************************************************************
607 \subsubsection{@specExpr@: the main function}
609 %************************************************************************
612 specVar :: Subst -> Id -> CoreExpr
613 specVar subst v = lookupIdSubst subst v
615 specExpr :: Subst -> CoreExpr -> SpecM (CoreExpr, UsageDetails)
616 -- We carry a substitution down:
617 -- a) we must clone any binding that might flaot outwards,
618 -- to avoid name clashes
619 -- b) we carry a type substitution to use when analysing
620 -- the RHS of specialised bindings (no type-let!)
622 ---------------- First the easy cases --------------------
623 specExpr subst (Type ty) = returnSM (Type (substTy subst ty), emptyUDs)
624 specExpr subst (Var v) = returnSM (specVar subst v, emptyUDs)
625 specExpr subst (Lit lit) = returnSM (Lit lit, emptyUDs)
627 specExpr subst (Note note body)
628 = specExpr subst body `thenSM` \ (body', uds) ->
629 returnSM (Note (specNote subst note) body', uds)
632 ---------------- Applications might generate a call instance --------------------
633 specExpr subst expr@(App fun arg)
636 go (App fun arg) args = specExpr subst arg `thenSM` \ (arg', uds_arg) ->
637 go fun (arg':args) `thenSM` \ (fun', uds_app) ->
638 returnSM (App fun' arg', uds_arg `plusUDs` uds_app)
640 go (Var f) args = case specVar subst f of
641 Var f' -> returnSM (Var f', mkCallUDs subst f' args)
642 e' -> returnSM (e', emptyUDs) -- I don't expect this!
643 go other args = specExpr subst other
645 ---------------- Lambda/case require dumping of usage details --------------------
646 specExpr subst e@(Lam _ _)
647 = specExpr subst' body `thenSM` \ (body', uds) ->
649 (filtered_uds, body'') = dumpUDs bndrs' uds body'
651 returnSM (mkLams bndrs' body'', filtered_uds)
653 (bndrs, body) = collectBinders e
654 (subst', bndrs') = substBndrs subst bndrs
655 -- More efficient to collect a group of binders together all at once
656 -- and we don't want to split a lambda group with dumped bindings
658 specExpr subst (Case scrut case_bndr ty alts)
659 = specExpr subst scrut `thenSM` \ (scrut', uds_scrut) ->
660 mapAndCombineSM spec_alt alts `thenSM` \ (alts', uds_alts) ->
661 returnSM (Case scrut' case_bndr' (substTy subst ty) alts', uds_scrut `plusUDs` uds_alts)
663 (subst_alt, case_bndr') = substBndr subst case_bndr
664 -- No need to clone case binder; it can't float like a let(rec)
666 spec_alt (con, args, rhs)
667 = specExpr subst_rhs rhs `thenSM` \ (rhs', uds) ->
669 (uds', rhs'') = dumpUDs args uds rhs'
671 returnSM ((con, args', rhs''), uds')
673 (subst_rhs, args') = substBndrs subst_alt args
675 ---------------- Finally, let is the interesting case --------------------
676 specExpr subst (Let bind body)
678 cloneBindSM subst bind `thenSM` \ (rhs_subst, body_subst, bind') ->
680 -- Deal with the body
681 specExpr body_subst body `thenSM` \ (body', body_uds) ->
683 -- Deal with the bindings
684 specBind rhs_subst bind' body_uds `thenSM` \ (binds', uds) ->
687 returnSM (foldr Let body' binds', uds)
689 -- Must apply the type substitution to coerceions
690 specNote subst (Coerce t1 t2) = Coerce (substTy subst t1) (substTy subst t2)
691 specNote subst note = note
694 %************************************************************************
696 \subsubsection{Dealing with a binding}
698 %************************************************************************
701 specBind :: Subst -- Use this for RHSs
703 -> UsageDetails -- Info on how the scope of the binding
704 -> SpecM ([CoreBind], -- New bindings
705 UsageDetails) -- And info to pass upstream
707 specBind rhs_subst bind body_uds
708 = specBindItself rhs_subst bind (calls body_uds) `thenSM` \ (bind', bind_uds) ->
710 bndrs = bindersOf bind
711 all_uds = zapCalls bndrs (body_uds `plusUDs` bind_uds)
712 -- It's important that the `plusUDs` is this way round,
713 -- because body_uds may bind dictionaries that are
714 -- used in the calls passed to specDefn. So the
715 -- dictionary bindings in bind_uds may mention
716 -- dictionaries bound in body_uds.
718 case splitUDs bndrs all_uds of
720 (_, ([],[])) -- This binding doesn't bind anything needed
721 -- in the UDs, so put the binding here
722 -- This is the case for most non-dict bindings, except
723 -- for the few that are mentioned in a dict binding
724 -- that is floating upwards in body_uds
725 -> returnSM ([bind'], all_uds)
727 (float_uds, (dict_binds, calls)) -- This binding is needed in the UDs, so float it out
728 -> returnSM ([], float_uds `plusUDs` mkBigUD bind' dict_binds calls)
731 -- A truly gruesome function
732 mkBigUD bind@(NonRec _ _) dbs calls
733 = -- Common case: non-recursive and no specialisations
734 -- (if there were any specialistions it would have been made recursive)
735 MkUD { dict_binds = listToBag (mkDB bind : dbs),
736 calls = listToCallDetails calls }
738 mkBigUD bind dbs calls
740 MkUD { dict_binds = unitBag (mkDB (Rec (bind_prs bind ++ dbsToPairs dbs))),
742 calls = listToCallDetails calls }
744 bind_prs (NonRec b r) = [(b,r)]
745 bind_prs (Rec prs) = prs
748 dbsToPairs ((bind,_):dbs) = bind_prs bind ++ dbsToPairs dbs
750 -- specBindItself deals with the RHS, specialising it according
751 -- to the calls found in the body (if any)
752 specBindItself rhs_subst (NonRec bndr rhs) call_info
753 = specDefn rhs_subst call_info (bndr,rhs) `thenSM` \ ((bndr',rhs'), spec_defns, spec_uds) ->
755 new_bind | null spec_defns = NonRec bndr' rhs'
756 | otherwise = Rec ((bndr',rhs'):spec_defns)
757 -- bndr' mentions the spec_defns in its SpecEnv
758 -- Not sure why we couln't just put the spec_defns first
760 returnSM (new_bind, spec_uds)
762 specBindItself rhs_subst (Rec pairs) call_info
763 = mapSM (specDefn rhs_subst call_info) pairs `thenSM` \ stuff ->
765 (pairs', spec_defns_s, spec_uds_s) = unzip3 stuff
766 spec_defns = concat spec_defns_s
767 spec_uds = plusUDList spec_uds_s
768 new_bind = Rec (spec_defns ++ pairs')
770 returnSM (new_bind, spec_uds)
773 specDefn :: Subst -- Subst to use for RHS
774 -> CallDetails -- Info on how it is used in its scope
775 -> (Id, CoreExpr) -- The thing being bound and its un-processed RHS
776 -> SpecM ((Id, CoreExpr), -- The thing and its processed RHS
777 -- the Id may now have specialisations attached
778 [(Id,CoreExpr)], -- Extra, specialised bindings
779 UsageDetails -- Stuff to fling upwards from the RHS and its
780 ) -- specialised versions
782 specDefn subst calls (fn, rhs)
783 -- The first case is the interesting one
784 | rhs_tyvars `lengthIs` n_tyvars -- Rhs of fn's defn has right number of big lambdas
785 && rhs_bndrs `lengthAtLeast` n_dicts -- and enough dict args
786 && notNull calls_for_me -- And there are some calls to specialise
788 -- At one time I tried not specialising small functions
789 -- but sometimes there are big functions marked INLINE
790 -- that we'd like to specialise. In particular, dictionary
791 -- functions, which Marcin is keen to inline
792 -- && not (certainlyWillInline fn) -- And it's not small
793 -- If it's small, it's better just to inline
794 -- it than to construct lots of specialisations
795 = -- Specialise the body of the function
796 specExpr subst rhs `thenSM` \ (rhs', rhs_uds) ->
798 -- Make a specialised version for each call in calls_for_me
799 mapSM spec_call calls_for_me `thenSM` \ stuff ->
801 (spec_defns, spec_uds, spec_rules) = unzip3 stuff
803 fn' = addIdSpecialisations fn spec_rules
805 returnSM ((fn',rhs'),
807 rhs_uds `plusUDs` plusUDList spec_uds)
809 | otherwise -- No calls or RHS doesn't fit our preconceptions
810 = specExpr subst rhs `thenSM` \ (rhs', rhs_uds) ->
811 returnSM ((fn, rhs'), [], rhs_uds)
815 (tyvars, theta, _) = tcSplitSigmaTy fn_type
816 n_tyvars = length tyvars
817 n_dicts = length theta
819 (rhs_tyvars, rhs_ids, rhs_body)
820 = collectTyAndValBinders (dropInline rhs)
821 -- It's important that we "see past" any INLINE pragma
822 -- else we'll fail to specialise an INLINE thing
824 rhs_dicts = take n_dicts rhs_ids
825 rhs_bndrs = rhs_tyvars ++ rhs_dicts
826 body = mkLams (drop n_dicts rhs_ids) rhs_body
827 -- Glue back on the non-dict lambdas
829 calls_for_me = case lookupFM calls fn of
831 Just cs -> fmToList cs
833 ----------------------------------------------------------
834 -- Specialise to one particular call pattern
835 spec_call :: (CallKey, ([DictExpr], VarSet)) -- Call instance
836 -> SpecM ((Id,CoreExpr), -- Specialised definition
837 UsageDetails, -- Usage details from specialised body
838 CoreRule) -- Info for the Id's SpecEnv
839 spec_call (CallKey call_ts, (call_ds, call_fvs))
840 = ASSERT( call_ts `lengthIs` n_tyvars && call_ds `lengthIs` n_dicts )
841 -- Calls are only recorded for properly-saturated applications
843 -- Suppose f's defn is f = /\ a b c d -> \ d1 d2 -> rhs
844 -- Supppose the call is for f [Just t1, Nothing, Just t3, Nothing] [dx1, dx2]
846 -- Construct the new binding
847 -- f1 = SUBST[a->t1,c->t3, d1->d1', d2->d2'] (/\ b d -> rhs)
848 -- PLUS the usage-details
849 -- { d1' = dx1; d2' = dx2 }
850 -- where d1', d2' are cloned versions of d1,d2, with the type substitution applied.
852 -- Note that the substitution is applied to the whole thing.
853 -- This is convenient, but just slightly fragile. Notably:
854 -- * There had better be no name clashes in a/b/c/d
857 -- poly_tyvars = [b,d] in the example above
858 -- spec_tyvars = [a,c]
859 -- ty_args = [t1,b,t3,d]
860 poly_tyvars = [tv | (tv, Nothing) <- rhs_tyvars `zip` call_ts]
861 spec_tyvars = [tv | (tv, Just _) <- rhs_tyvars `zip` call_ts]
862 ty_args = zipWithEqual "spec_call" mk_ty_arg rhs_tyvars call_ts
864 mk_ty_arg rhs_tyvar Nothing = Type (mkTyVarTy rhs_tyvar)
865 mk_ty_arg rhs_tyvar (Just ty) = Type ty
866 rhs_subst = extendTvSubstList subst (spec_tyvars `zip` [ty | Just ty <- call_ts])
868 cloneBinders rhs_subst rhs_dicts `thenSM` \ (rhs_subst', rhs_dicts') ->
870 inst_args = ty_args ++ map Var rhs_dicts'
872 -- Figure out the type of the specialised function
873 body_ty = applyTypeToArgs rhs fn_type inst_args
874 (lam_args, app_args) -- Add a dummy argument if body_ty is unlifted
875 | isUnLiftedType body_ty -- C.f. WwLib.mkWorkerArgs
876 = (poly_tyvars ++ [voidArgId], poly_tyvars ++ [realWorldPrimId])
877 | otherwise = (poly_tyvars, poly_tyvars)
878 spec_id_ty = mkPiTypes lam_args body_ty
880 newIdSM fn spec_id_ty `thenSM` \ spec_f ->
881 specExpr rhs_subst' (mkLams lam_args body) `thenSM` \ (spec_rhs, rhs_uds) ->
883 -- The rule to put in the function's specialisation is:
884 -- forall b,d, d1',d2'. f t1 b t3 d d1' d2' = f1 b d
885 spec_env_rule = mkLocalRule (mkFastString ("SPEC " ++ showSDoc (ppr fn)))
886 AlwaysActive (idName fn)
887 (poly_tyvars ++ rhs_dicts')
889 (mkVarApps (Var spec_f) app_args)
891 -- Add the { d1' = dx1; d2' = dx2 } usage stuff
892 final_uds = foldr addDictBind rhs_uds (my_zipEqual "spec_call" rhs_dicts' call_ds)
894 -- NOTE: we don't add back in any INLINE pragma on the RHS, so even if
895 -- the original function said INLINE, the specialised copies won't.
896 -- The idea is that the point of inlining was precisely to specialise
897 -- the function at its call site, and that's not so important for the
898 -- specialised copies. But it still smells like an ad hoc decision.
901 returnSM ((spec_f, spec_rhs),
906 my_zipEqual doc xs ys
907 | not (equalLength xs ys) = pprPanic "my_zipEqual" (ppr xs $$ ppr ys $$ (ppr fn <+> ppr call_ts) $$ ppr rhs)
908 | otherwise = zipEqual doc xs ys
910 dropInline :: CoreExpr -> CoreExpr
911 dropInline (Note InlineMe rhs) = rhs
915 %************************************************************************
917 \subsubsection{UsageDetails and suchlike}
919 %************************************************************************
924 dict_binds :: !(Bag DictBind),
925 -- Floated dictionary bindings
926 -- The order is important;
927 -- in ds1 `union` ds2, bindings in ds2 can depend on those in ds1
928 -- (Remember, Bags preserve order in GHC.)
930 calls :: !CallDetails
933 type DictBind = (CoreBind, VarSet)
934 -- The set is the free vars of the binding
935 -- both tyvars and dicts
937 type DictExpr = CoreExpr
939 emptyUDs = MkUD { dict_binds = emptyBag, calls = emptyFM }
941 type ProtoUsageDetails = ([DictBind],
942 [(Id, CallKey, ([DictExpr], VarSet))]
945 ------------------------------------------------------------
946 type CallDetails = FiniteMap Id CallInfo
947 newtype CallKey = CallKey [Maybe Type] -- Nothing => unconstrained type argument
948 type CallInfo = FiniteMap CallKey
949 ([DictExpr], VarSet) -- Dict args and the vars of the whole
950 -- call (including tyvars)
951 -- [*not* include the main id itself, of course]
952 -- The finite maps eliminate duplicates
953 -- The list of types and dictionaries is guaranteed to
954 -- match the type of f
956 -- Type isn't an instance of Ord, so that we can control which
957 -- instance we use. That's tiresome here. Oh well
958 instance Eq CallKey where
959 k1 == k2 = case k1 `compare` k2 of { EQ -> True; other -> False }
961 instance Ord CallKey where
962 compare (CallKey k1) (CallKey k2) = cmpList cmp k1 k2
964 cmp Nothing Nothing = EQ
965 cmp Nothing (Just t2) = LT
966 cmp (Just t1) Nothing = GT
967 cmp (Just t1) (Just t2) = tcCmpType t1 t2
969 unionCalls :: CallDetails -> CallDetails -> CallDetails
970 unionCalls c1 c2 = plusFM_C plusFM c1 c2
972 singleCall :: Id -> [Maybe Type] -> [DictExpr] -> CallDetails
973 singleCall id tys dicts
974 = unitFM id (unitFM (CallKey tys) (dicts, call_fvs))
976 call_fvs = exprsFreeVars dicts `unionVarSet` tys_fvs
977 tys_fvs = tyVarsOfTypes (catMaybes tys)
978 -- The type args (tys) are guaranteed to be part of the dictionary
979 -- types, because they are just the constrained types,
980 -- and the dictionary is therefore sure to be bound
981 -- inside the binding for any type variables free in the type;
982 -- hence it's safe to neglect tyvars free in tys when making
983 -- the free-var set for this call
984 -- BUT I don't trust this reasoning; play safe and include tys_fvs
986 -- We don't include the 'id' itself.
988 listToCallDetails calls
989 = foldr (unionCalls . mk_call) emptyFM calls
991 mk_call (id, tys, dicts_w_fvs) = unitFM id (unitFM tys dicts_w_fvs)
992 -- NB: the free vars of the call are provided
994 callDetailsToList calls = [ (id,tys,dicts)
995 | (id,fm) <- fmToList calls,
996 (tys, dicts) <- fmToList fm
999 mkCallUDs subst f args
1001 || not (all isClassPred theta)
1002 -- Only specialise if all overloading is on class params.
1003 -- In ptic, with implicit params, the type args
1004 -- *don't* say what the value of the implicit param is!
1005 || not (spec_tys `lengthIs` n_tyvars)
1006 || not ( dicts `lengthIs` n_dicts)
1007 || maybeToBool (lookupRule (\act -> True) (substInScope subst) emptyRuleBase f args)
1008 -- There's already a rule covering this call. A typical case
1009 -- is where there's an explicit user-provided rule. Then
1010 -- we don't want to create a specialised version
1011 -- of the function that overlaps.
1012 = emptyUDs -- Not overloaded, or no specialisation wanted
1015 = MkUD {dict_binds = emptyBag,
1016 calls = singleCall f spec_tys dicts
1019 (tyvars, theta, _) = tcSplitSigmaTy (idType f)
1020 constrained_tyvars = tyVarsOfTheta theta
1021 n_tyvars = length tyvars
1022 n_dicts = length theta
1024 spec_tys = [mk_spec_ty tv ty | (tv, Type ty) <- tyvars `zip` args]
1025 dicts = [dict_expr | (_, dict_expr) <- theta `zip` (drop n_tyvars args)]
1028 | tyvar `elemVarSet` constrained_tyvars = Just ty
1029 | otherwise = Nothing
1031 ------------------------------------------------------------
1032 plusUDs :: UsageDetails -> UsageDetails -> UsageDetails
1033 plusUDs (MkUD {dict_binds = db1, calls = calls1})
1034 (MkUD {dict_binds = db2, calls = calls2})
1035 = MkUD {dict_binds = d, calls = c}
1037 d = db1 `unionBags` db2
1038 c = calls1 `unionCalls` calls2
1040 plusUDList = foldr plusUDs emptyUDs
1042 -- zapCalls deletes calls to ids from uds
1043 zapCalls ids uds = uds {calls = delListFromFM (calls uds) ids}
1045 mkDB bind = (bind, bind_fvs bind)
1047 bind_fvs (NonRec bndr rhs) = pair_fvs (bndr,rhs)
1048 bind_fvs (Rec prs) = foldl delVarSet rhs_fvs bndrs
1051 rhs_fvs = unionVarSets (map pair_fvs prs)
1053 pair_fvs (bndr, rhs) = exprFreeVars rhs `unionVarSet` idRuleVars bndr
1054 -- Don't forget variables mentioned in the
1055 -- rules of the bndr. C.f. OccAnal.addRuleUsage
1058 addDictBind (dict,rhs) uds = uds { dict_binds = mkDB (NonRec dict rhs) `consBag` dict_binds uds }
1060 dumpAllDictBinds (MkUD {dict_binds = dbs}) binds
1061 = foldrBag add binds dbs
1063 add (bind,_) binds = bind : binds
1065 dumpUDs :: [CoreBndr]
1066 -> UsageDetails -> CoreExpr
1067 -> (UsageDetails, CoreExpr)
1068 dumpUDs bndrs uds body
1069 = (free_uds, foldr add_let body dict_binds)
1071 (free_uds, (dict_binds, _)) = splitUDs bndrs uds
1072 add_let (bind,_) body = Let bind body
1074 splitUDs :: [CoreBndr]
1076 -> (UsageDetails, -- These don't mention the binders
1077 ProtoUsageDetails) -- These do
1079 splitUDs bndrs uds@(MkUD {dict_binds = orig_dbs,
1080 calls = orig_calls})
1082 = if isEmptyBag dump_dbs && null dump_calls then
1083 -- Common case: binder doesn't affect floats
1087 -- Binders bind some of the fvs of the floats
1088 (MkUD {dict_binds = free_dbs,
1089 calls = listToCallDetails free_calls},
1090 (bagToList dump_dbs, dump_calls)
1094 bndr_set = mkVarSet bndrs
1096 (free_dbs, dump_dbs, dump_idset)
1097 = foldlBag dump_db (emptyBag, emptyBag, bndr_set) orig_dbs
1098 -- Important that it's foldl not foldr;
1099 -- we're accumulating the set of dumped ids in dump_set
1101 -- Filter out any calls that mention things that are being dumped
1102 orig_call_list = callDetailsToList orig_calls
1103 (dump_calls, free_calls) = partition captured orig_call_list
1104 captured (id,tys,(dicts, fvs)) = fvs `intersectsVarSet` dump_idset
1105 || id `elemVarSet` dump_idset
1107 dump_db (free_dbs, dump_dbs, dump_idset) db@(bind, fvs)
1108 | dump_idset `intersectsVarSet` fvs -- Dump it
1109 = (free_dbs, dump_dbs `snocBag` db,
1110 extendVarSetList dump_idset (bindersOf bind))
1112 | otherwise -- Don't dump it
1113 = (free_dbs `snocBag` db, dump_dbs, dump_idset)
1117 %************************************************************************
1119 \subsubsection{Boring helper functions}
1121 %************************************************************************
1124 type SpecM a = UniqSM a
1128 getUniqSM = getUniqueUs
1132 mapAndCombineSM f [] = returnSM ([], emptyUDs)
1133 mapAndCombineSM f (x:xs) = f x `thenSM` \ (y, uds1) ->
1134 mapAndCombineSM f xs `thenSM` \ (ys, uds2) ->
1135 returnSM (y:ys, uds1 `plusUDs` uds2)
1137 cloneBindSM :: Subst -> CoreBind -> SpecM (Subst, Subst, CoreBind)
1138 -- Clone the binders of the bind; return new bind with the cloned binders
1139 -- Return the substitution to use for RHSs, and the one to use for the body
1140 cloneBindSM subst (NonRec bndr rhs)
1141 = getUs `thenUs` \ us ->
1143 (subst', bndr') = cloneIdBndr subst us bndr
1145 returnUs (subst, subst', NonRec bndr' rhs)
1147 cloneBindSM subst (Rec pairs)
1148 = getUs `thenUs` \ us ->
1150 (subst', bndrs') = cloneRecIdBndrs subst us (map fst pairs)
1152 returnUs (subst', subst', Rec (bndrs' `zip` map snd pairs))
1154 cloneBinders subst bndrs
1155 = getUs `thenUs` \ us ->
1156 returnUs (cloneIdBndrs subst us bndrs)
1158 newIdSM old_id new_ty
1159 = getUniqSM `thenSM` \ uniq ->
1161 -- Give the new Id a similar occurrence name to the old one
1162 name = idName old_id
1163 new_id = mkUserLocal (mkSpecOcc (nameOccName name)) uniq new_ty (getSrcLoc name)
1169 Old (but interesting) stuff about unboxed bindings
1170 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1172 What should we do when a value is specialised to a *strict* unboxed value?
1174 map_*_* f (x:xs) = let h = f x
1178 Could convert let to case:
1180 map_*_Int# f (x:xs) = case f x of h# ->
1184 This may be undesirable since it forces evaluation here, but the value
1185 may not be used in all branches of the body. In the general case this
1186 transformation is impossible since the mutual recursion in a letrec
1187 cannot be expressed as a case.
1189 There is also a problem with top-level unboxed values, since our
1190 implementation cannot handle unboxed values at the top level.
1192 Solution: Lift the binding of the unboxed value and extract it when it
1195 map_*_Int# f (x:xs) = let h = case (f x) of h# -> _Lift h#
1200 Now give it to the simplifier and the _Lifting will be optimised away.
1202 The benfit is that we have given the specialised "unboxed" values a
1203 very simplep lifted semantics and then leave it up to the simplifier to
1204 optimise it --- knowing that the overheads will be removed in nearly
1207 In particular, the value will only be evaluted in the branches of the
1208 program which use it, rather than being forced at the point where the
1209 value is bound. For example:
1211 filtermap_*_* p f (x:xs)
1218 filtermap_*_Int# p f (x:xs)
1219 = let h = case (f x) of h# -> _Lift h#
1222 True -> case h of _Lift h#
1226 The binding for h can still be inlined in the one branch and the
1227 _Lifting eliminated.
1230 Question: When won't the _Lifting be eliminated?
1232 Answer: When they at the top-level (where it is necessary) or when
1233 inlining would duplicate work (or possibly code depending on
1234 options). However, the _Lifting will still be eliminated if the
1235 strictness analyser deems the lifted binding strict.