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 CmdLineOpts ( 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
21 import Var ( zapSpecPragmaId )
25 import CoreUtils ( applyTypeToArgs, mkPiTypes )
26 import CoreFVs ( exprFreeVars, exprsFreeVars )
27 import CoreTidy ( pprTidyIdRules )
28 import CoreLint ( showPass, endPass )
29 import Rules ( addIdSpecialisations, lookupRule, emptyRuleBase )
31 import UniqSupply ( UniqSupply,
32 UniqSM, initUs_, thenUs, returnUs, getUniqueUs,
35 import Name ( nameOccName, mkSpecOcc, getSrcLoc )
36 import MkId ( voidArgId, realWorldPrimId )
38 import Maybes ( catMaybes, maybeToBool )
39 import ErrUtils ( dumpIfSet_dyn )
40 import BasicTypes ( Activation( AlwaysActive ) )
42 import List ( partition )
43 import Util ( zipEqual, zipWithEqual, cmpList, lengthIs,
44 equalLength, lengthAtLeast, notNull )
51 %************************************************************************
53 \subsection[notes-Specialise]{Implementation notes [SLPJ, Aug 18 1993]}
55 %************************************************************************
57 These notes describe how we implement specialisation to eliminate
60 The specialisation pass works on Core
61 syntax, complete with all the explicit dictionary application,
62 abstraction and construction as added by the type checker. The
63 existing type checker remains largely as it is.
65 One important thought: the {\em types} passed to an overloaded
66 function, and the {\em dictionaries} passed are mutually redundant.
67 If the same function is applied to the same type(s) then it is sure to
68 be applied to the same dictionary(s)---or rather to the same {\em
69 values}. (The arguments might look different but they will evaluate
72 Second important thought: we know that we can make progress by
73 treating dictionary arguments as static and worth specialising on. So
74 we can do without binding-time analysis, and instead specialise on
75 dictionary arguments and no others.
84 and suppose f is overloaded.
86 STEP 1: CALL-INSTANCE COLLECTION
88 We traverse <body>, accumulating all applications of f to types and
91 (Might there be partial applications, to just some of its types and
92 dictionaries? In principle yes, but in practice the type checker only
93 builds applications of f to all its types and dictionaries, so partial
94 applications could only arise as a result of transformation, and even
95 then I think it's unlikely. In any case, we simply don't accumulate such
96 partial applications.)
101 So now we have a collection of calls to f:
105 Notice that f may take several type arguments. To avoid ambiguity, we
106 say that f is called at type t1/t2 and t3/t4.
108 We take equivalence classes using equality of the *types* (ignoring
109 the dictionary args, which as mentioned previously are redundant).
111 STEP 3: SPECIALISATION
113 For each equivalence class, choose a representative (f t1 t2 d1 d2),
114 and create a local instance of f, defined thus:
116 f@t1/t2 = <f_rhs> t1 t2 d1 d2
118 f_rhs presumably has some big lambdas and dictionary lambdas, so lots
119 of simplification will now result. However we don't actually *do* that
120 simplification. Rather, we leave it for the simplifier to do. If we
121 *did* do it, though, we'd get more call instances from the specialised
122 RHS. We can work out what they are by instantiating the call-instance
123 set from f's RHS with the types t1, t2.
125 Add this new id to f's IdInfo, to record that f has a specialised version.
127 Before doing any of this, check that f's IdInfo doesn't already
128 tell us about an existing instance of f at the required type/s.
129 (This might happen if specialisation was applied more than once, or
130 it might arise from user SPECIALIZE pragmas.)
134 Wait a minute! What if f is recursive? Then we can't just plug in
135 its right-hand side, can we?
137 But it's ok. The type checker *always* creates non-recursive definitions
138 for overloaded recursive functions. For example:
140 f x = f (x+x) -- Yes I know its silly
144 f a (d::Num a) = let p = +.sel a d
146 letrec fl (y::a) = fl (p y y)
150 We still have recusion for non-overloaded functions which we
151 speciailise, but the recursive call should get specialised to the
152 same recursive version.
158 All this is crystal clear when the function is applied to *constant
159 types*; that is, types which have no type variables inside. But what if
160 it is applied to non-constant types? Suppose we find a call of f at type
161 t1/t2. There are two possibilities:
163 (a) The free type variables of t1, t2 are in scope at the definition point
164 of f. In this case there's no problem, we proceed just as before. A common
165 example is as follows. Here's the Haskell:
170 After typechecking we have
172 g a (d::Num a) (y::a) = let f b (d'::Num b) (x::b) = +.sel b d' x x
173 in +.sel a d (f a d y) (f a d y)
175 Notice that the call to f is at type type "a"; a non-constant type.
176 Both calls to f are at the same type, so we can specialise to give:
178 g a (d::Num a) (y::a) = let f@a (x::a) = +.sel a d x x
179 in +.sel a d (f@a y) (f@a y)
182 (b) The other case is when the type variables in the instance types
183 are *not* in scope at the definition point of f. The example we are
184 working with above is a good case. There are two instances of (+.sel a d),
185 but "a" is not in scope at the definition of +.sel. Can we do anything?
186 Yes, we can "common them up", a sort of limited common sub-expression deal.
189 g a (d::Num a) (y::a) = let +.sel@a = +.sel a d
190 f@a (x::a) = +.sel@a x x
191 in +.sel@a (f@a y) (f@a y)
193 This can save work, and can't be spotted by the type checker, because
194 the two instances of +.sel weren't originally at the same type.
198 * There are quite a few variations here. For example, the defn of
199 +.sel could be floated ouside the \y, to attempt to gain laziness.
200 It certainly mustn't be floated outside the \d because the d has to
203 * We don't want to inline f_rhs in this case, because
204 that will duplicate code. Just commoning up the call is the point.
206 * Nothing gets added to +.sel's IdInfo.
208 * Don't bother unless the equivalence class has more than one item!
210 Not clear whether this is all worth it. It is of course OK to
211 simply discard call-instances when passing a big lambda.
213 Polymorphism 2 -- Overloading
215 Consider a function whose most general type is
217 f :: forall a b. Ord a => [a] -> b -> b
219 There is really no point in making a version of g at Int/Int and another
220 at Int/Bool, because it's only instancing the type variable "a" which
221 buys us any efficiency. Since g is completely polymorphic in b there
222 ain't much point in making separate versions of g for the different
225 That suggests that we should identify which of g's type variables
226 are constrained (like "a") and which are unconstrained (like "b").
227 Then when taking equivalence classes in STEP 2, we ignore the type args
228 corresponding to unconstrained type variable. In STEP 3 we make
229 polymorphic versions. Thus:
231 f@t1/ = /\b -> <f_rhs> t1 b d1 d2
240 f a (d::Num a) = let g = ...
242 ...(let d1::Ord a = Num.Ord.sel a d in g a d1)...
244 Here, g is only called at one type, but the dictionary isn't in scope at the
245 definition point for g. Usually the type checker would build a
246 definition for d1 which enclosed g, but the transformation system
247 might have moved d1's defn inward. Solution: float dictionary bindings
248 outwards along with call instances.
252 f x = let g p q = p==q
258 Before specialisation, leaving out type abstractions we have
260 f df x = let g :: Eq a => a -> a -> Bool
262 h :: Num a => a -> a -> (a, Bool)
263 h dh r s = let deq = eqFromNum dh
264 in (+ dh r s, g deq r s)
268 After specialising h we get a specialised version of h, like this:
270 h' r s = let deq = eqFromNum df
271 in (+ df r s, g deq r s)
273 But we can't naively make an instance for g from this, because deq is not in scope
274 at the defn of g. Instead, we have to float out the (new) defn of deq
275 to widen its scope. Notice that this floating can't be done in advance -- it only
276 shows up when specialisation is done.
278 User SPECIALIZE pragmas
279 ~~~~~~~~~~~~~~~~~~~~~~~
280 Specialisation pragmas can be digested by the type checker, and implemented
281 by adding extra definitions along with that of f, in the same way as before
283 f@t1/t2 = <f_rhs> t1 t2 d1 d2
285 Indeed the pragmas *have* to be dealt with by the type checker, because
286 only it knows how to build the dictionaries d1 and d2! For example
288 g :: Ord a => [a] -> [a]
289 {-# SPECIALIZE f :: [Tree Int] -> [Tree Int] #-}
291 Here, the specialised version of g is an application of g's rhs to the
292 Ord dictionary for (Tree Int), which only the type checker can conjure
293 up. There might not even *be* one, if (Tree Int) is not an instance of
294 Ord! (All the other specialision has suitable dictionaries to hand
297 Problem. The type checker doesn't have to hand a convenient <f_rhs>, because
298 it is buried in a complex (as-yet-un-desugared) binding group.
301 f@t1/t2 = f* t1 t2 d1 d2
303 where f* is the Id f with an IdInfo which says "inline me regardless!".
304 Indeed all the specialisation could be done in this way.
305 That in turn means that the simplifier has to be prepared to inline absolutely
306 any in-scope let-bound thing.
309 Again, the pragma should permit polymorphism in unconstrained variables:
311 h :: Ord a => [a] -> b -> b
312 {-# SPECIALIZE h :: [Int] -> b -> b #-}
314 We *insist* that all overloaded type variables are specialised to ground types,
315 (and hence there can be no context inside a SPECIALIZE pragma).
316 We *permit* unconstrained type variables to be specialised to
318 - or left as a polymorphic type variable
319 but nothing in between. So
321 {-# SPECIALIZE h :: [Int] -> [c] -> [c] #-}
323 is *illegal*. (It can be handled, but it adds complication, and gains the
327 SPECIALISING INSTANCE DECLARATIONS
328 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
331 instance Foo a => Foo [a] where
333 {-# SPECIALIZE instance Foo [Int] #-}
335 The original instance decl creates a dictionary-function
338 dfun.Foo.List :: forall a. Foo a -> Foo [a]
340 The SPECIALIZE pragma just makes a specialised copy, just as for
341 ordinary function definitions:
343 dfun.Foo.List@Int :: Foo [Int]
344 dfun.Foo.List@Int = dfun.Foo.List Int dFooInt
346 The information about what instance of the dfun exist gets added to
347 the dfun's IdInfo in the same way as a user-defined function too.
350 Automatic instance decl specialisation?
351 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
352 Can instance decls be specialised automatically? It's tricky.
353 We could collect call-instance information for each dfun, but
354 then when we specialised their bodies we'd get new call-instances
355 for ordinary functions; and when we specialised their bodies, we might get
356 new call-instances of the dfuns, and so on. This all arises because of
357 the unrestricted mutual recursion between instance decls and value decls.
359 Still, there's no actual problem; it just means that we may not do all
360 the specialisation we could theoretically do.
362 Furthermore, instance decls are usually exported and used non-locally,
363 so we'll want to compile enough to get those specialisations done.
365 Lastly, there's no such thing as a local instance decl, so we can
366 survive solely by spitting out *usage* information, and then reading that
367 back in as a pragma when next compiling the file. So for now,
368 we only specialise instance decls in response to pragmas.
371 SPITTING OUT USAGE INFORMATION
372 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
374 To spit out usage information we need to traverse the code collecting
375 call-instance information for all imported (non-prelude?) functions
376 and data types. Then we equivalence-class it and spit it out.
378 This is done at the top-level when all the call instances which escape
379 must be for imported functions and data types.
381 *** Not currently done ***
384 Partial specialisation by pragmas
385 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
386 What about partial specialisation:
388 k :: (Ord a, Eq b) => [a] -> b -> b -> [a]
389 {-# SPECIALIZE k :: Eq b => [Int] -> b -> b -> [a] #-}
393 {-# SPECIALIZE k :: Eq b => [Int] -> [b] -> [b] -> [a] #-}
395 Seems quite reasonable. Similar things could be done with instance decls:
397 instance (Foo a, Foo b) => Foo (a,b) where
399 {-# SPECIALIZE instance Foo a => Foo (a,Int) #-}
400 {-# SPECIALIZE instance Foo b => Foo (Int,b) #-}
402 Ho hum. Things are complex enough without this. I pass.
405 Requirements for the simplifer
406 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
407 The simplifier has to be able to take advantage of the specialisation.
409 * When the simplifier finds an application of a polymorphic f, it looks in
410 f's IdInfo in case there is a suitable instance to call instead. This converts
412 f t1 t2 d1 d2 ===> f_t1_t2
414 Note that the dictionaries get eaten up too!
416 * Dictionary selection operations on constant dictionaries must be
419 +.sel Int d ===> +Int
421 The obvious way to do this is in the same way as other specialised
422 calls: +.sel has inside it some IdInfo which tells that if it's applied
423 to the type Int then it should eat a dictionary and transform to +Int.
425 In short, dictionary selectors need IdInfo inside them for constant
428 * Exactly the same applies if a superclass dictionary is being
431 Eq.sel Int d ===> dEqInt
433 * Something similar applies to dictionary construction too. Suppose
434 dfun.Eq.List is the function taking a dictionary for (Eq a) to
435 one for (Eq [a]). Then we want
437 dfun.Eq.List Int d ===> dEq.List_Int
439 Where does the Eq [Int] dictionary come from? It is built in
440 response to a SPECIALIZE pragma on the Eq [a] instance decl.
442 In short, dfun Ids need IdInfo with a specialisation for each
443 constant instance of their instance declaration.
445 All this uses a single mechanism: the SpecEnv inside an Id
448 What does the specialisation IdInfo look like?
449 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
451 The SpecEnv of an Id maps a list of types (the template) to an expression
455 For example, if f has this SpecInfo:
457 [Int, a] -> \d:Ord Int. f' a
459 it means that we can replace the call
461 f Int t ===> (\d. f' t)
463 This chucks one dictionary away and proceeds with the
464 specialised version of f, namely f'.
467 What can't be done this way?
468 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
469 There is no way, post-typechecker, to get a dictionary for (say)
470 Eq a from a dictionary for Eq [a]. So if we find
474 we can't transform to
479 eqList :: (a->a->Bool) -> [a] -> [a] -> Bool
481 Of course, we currently have no way to automatically derive
482 eqList, nor to connect it to the Eq [a] instance decl, but you
483 can imagine that it might somehow be possible. Taking advantage
484 of this is permanently ruled out.
486 Still, this is no great hardship, because we intend to eliminate
487 overloading altogether anyway!
491 A note about non-tyvar dictionaries
492 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
493 Some Ids have types like
495 forall a,b,c. Eq a -> Ord [a] -> tau
497 This seems curious at first, because we usually only have dictionary
498 args whose types are of the form (C a) where a is a type variable.
499 But this doesn't hold for the functions arising from instance decls,
500 which sometimes get arguements with types of form (C (T a)) for some
503 Should we specialise wrt this compound-type dictionary? We used to say
505 "This is a heuristic judgement, as indeed is the fact that we
506 specialise wrt only dictionaries. We choose *not* to specialise
507 wrt compound dictionaries because at the moment the only place
508 they show up is in instance decls, where they are simply plugged
509 into a returned dictionary. So nothing is gained by specialising
512 But it is simpler and more uniform to specialise wrt these dicts too;
513 and in future GHC is likely to support full fledged type signatures
515 f ;: Eq [(a,b)] => ...
518 %************************************************************************
520 \subsubsection{The new specialiser}
522 %************************************************************************
524 Our basic game plan is this. For let(rec) bound function
525 f :: (C a, D c) => (a,b,c,d) -> Bool
527 * Find any specialised calls of f, (f ts ds), where
528 ts are the type arguments t1 .. t4, and
529 ds are the dictionary arguments d1 .. d2.
531 * Add a new definition for f1 (say):
533 f1 = /\ b d -> (..body of f..) t1 b t3 d d1 d2
535 Note that we abstract over the unconstrained type arguments.
539 [t1,b,t3,d] |-> \d1 d2 -> f1 b d
541 to the specialisations of f. This will be used by the
542 simplifier to replace calls
543 (f t1 t2 t3 t4) da db
545 (\d1 d1 -> f1 t2 t4) da db
547 All the stuff about how many dictionaries to discard, and what types
548 to apply the specialised function to, are handled by the fact that the
549 SpecEnv contains a template for the result of the specialisation.
551 We don't build *partial* specialisations for f. For example:
553 f :: Eq a => a -> a -> Bool
554 {-# SPECIALISE f :: (Eq b, Eq c) => (b,c) -> (b,c) -> Bool #-}
556 Here, little is gained by making a specialised copy of f.
557 There's a distinct danger that the specialised version would
558 first build a dictionary for (Eq b, Eq c), and then select the (==)
559 method from it! Even if it didn't, not a great deal is saved.
561 We do, however, generate polymorphic, but not overloaded, specialisations:
563 f :: Eq a => [a] -> b -> b -> b
564 {#- SPECIALISE f :: [Int] -> b -> b -> b #-}
566 Hence, the invariant is this:
568 *** no specialised version is overloaded ***
571 %************************************************************************
573 \subsubsection{The exported function}
575 %************************************************************************
578 specProgram :: DynFlags -> UniqSupply -> [CoreBind] -> IO [CoreBind]
579 specProgram dflags us binds
581 showPass dflags "Specialise"
583 let binds' = initSM us (go binds `thenSM` \ (binds', uds') ->
584 returnSM (dumpAllDictBinds uds' binds'))
586 endPass dflags "Specialise" Opt_D_dump_spec binds'
588 dumpIfSet_dyn dflags Opt_D_dump_rules "Top-level specialisations"
589 (vcat (map pprTidyIdRules (concat (map bindersOf binds'))))
593 -- We need to start with a Subst that knows all the things
594 -- that are in scope, so that the substitution engine doesn't
595 -- accidentally re-use a unique that's already in use
596 -- Easiest thing is to do it all at once, as if all the top-level
597 -- decls were mutually recursive
598 top_subst = mkEmptySubst (mkInScopeSet (mkVarSet (bindersOfBinds binds)))
600 go [] = returnSM ([], emptyUDs)
601 go (bind:binds) = go binds `thenSM` \ (binds', uds) ->
602 specBind top_subst bind uds `thenSM` \ (bind', uds') ->
603 returnSM (bind' ++ binds', uds')
606 %************************************************************************
608 \subsubsection{@specExpr@: the main function}
610 %************************************************************************
613 specVar :: Subst -> Id -> CoreExpr
614 specVar subst v = lookupIdSubst subst v
616 specExpr :: Subst -> CoreExpr -> SpecM (CoreExpr, UsageDetails)
617 -- We carry a substitution down:
618 -- a) we must clone any binding that might flaot outwards,
619 -- to avoid name clashes
620 -- b) we carry a type substitution to use when analysing
621 -- the RHS of specialised bindings (no type-let!)
623 ---------------- First the easy cases --------------------
624 specExpr subst (Type ty) = returnSM (Type (substTy subst ty), emptyUDs)
625 specExpr subst (Var v) = returnSM (specVar subst v, emptyUDs)
626 specExpr subst (Lit lit) = returnSM (Lit lit, emptyUDs)
628 specExpr subst (Note note body)
629 = specExpr subst body `thenSM` \ (body', uds) ->
630 returnSM (Note (specNote subst note) body', uds)
633 ---------------- Applications might generate a call instance --------------------
634 specExpr subst expr@(App fun arg)
637 go (App fun arg) args = specExpr subst arg `thenSM` \ (arg', uds_arg) ->
638 go fun (arg':args) `thenSM` \ (fun', uds_app) ->
639 returnSM (App fun' arg', uds_arg `plusUDs` uds_app)
641 go (Var f) args = case specVar subst f of
642 Var f' -> returnSM (Var f', mkCallUDs subst f' args)
643 e' -> returnSM (e', emptyUDs) -- I don't expect this!
644 go other args = specExpr subst other
646 ---------------- Lambda/case require dumping of usage details --------------------
647 specExpr subst e@(Lam _ _)
648 = specExpr subst' body `thenSM` \ (body', uds) ->
650 (filtered_uds, body'') = dumpUDs bndrs' uds body'
652 returnSM (mkLams bndrs' body'', filtered_uds)
654 (bndrs, body) = collectBinders e
655 (subst', bndrs') = substBndrs subst bndrs
656 -- More efficient to collect a group of binders together all at once
657 -- and we don't want to split a lambda group with dumped bindings
659 specExpr subst (Case scrut case_bndr ty alts)
660 = specExpr subst scrut `thenSM` \ (scrut', uds_scrut) ->
661 mapAndCombineSM spec_alt alts `thenSM` \ (alts', uds_alts) ->
662 returnSM (Case scrut' case_bndr' (substTy subst ty) alts', uds_scrut `plusUDs` uds_alts)
664 (subst_alt, case_bndr') = substBndr subst case_bndr
665 -- No need to clone case binder; it can't float like a let(rec)
667 spec_alt (con, args, rhs)
668 = specExpr subst_rhs rhs `thenSM` \ (rhs', uds) ->
670 (uds', rhs'') = dumpUDs args uds rhs'
672 returnSM ((con, args', rhs''), uds')
674 (subst_rhs, args') = substBndrs subst_alt args
676 ---------------- Finally, let is the interesting case --------------------
677 specExpr subst (Let bind body)
679 cloneBindSM subst bind `thenSM` \ (rhs_subst, body_subst, bind') ->
681 -- Deal with the body
682 specExpr body_subst body `thenSM` \ (body', body_uds) ->
684 -- Deal with the bindings
685 specBind rhs_subst bind' body_uds `thenSM` \ (binds', uds) ->
688 returnSM (foldr Let body' binds', uds)
690 -- Must apply the type substitution to coerceions
691 specNote subst (Coerce t1 t2) = Coerce (substTy subst t1) (substTy subst t2)
692 specNote subst note = note
695 %************************************************************************
697 \subsubsection{Dealing with a binding}
699 %************************************************************************
702 specBind :: Subst -- Use this for RHSs
704 -> UsageDetails -- Info on how the scope of the binding
705 -> SpecM ([CoreBind], -- New bindings
706 UsageDetails) -- And info to pass upstream
708 specBind rhs_subst bind body_uds
709 = specBindItself rhs_subst bind (calls body_uds) `thenSM` \ (bind', bind_uds) ->
711 bndrs = bindersOf bind
712 all_uds = zapCalls bndrs (body_uds `plusUDs` bind_uds)
713 -- It's important that the `plusUDs` is this way round,
714 -- because body_uds may bind dictionaries that are
715 -- used in the calls passed to specDefn. So the
716 -- dictionary bindings in bind_uds may mention
717 -- dictionaries bound in body_uds.
719 case splitUDs bndrs all_uds of
721 (_, ([],[])) -- This binding doesn't bind anything needed
722 -- in the UDs, so put the binding here
723 -- This is the case for most non-dict bindings, except
724 -- for the few that are mentioned in a dict binding
725 -- that is floating upwards in body_uds
726 -> returnSM ([bind'], all_uds)
728 (float_uds, (dict_binds, calls)) -- This binding is needed in the UDs, so float it out
729 -> returnSM ([], float_uds `plusUDs` mkBigUD bind' dict_binds calls)
732 -- A truly gruesome function
733 mkBigUD bind@(NonRec _ _) dbs calls
734 = -- Common case: non-recursive and no specialisations
735 -- (if there were any specialistions it would have been made recursive)
736 MkUD { dict_binds = listToBag (mkDB bind : dbs),
737 calls = listToCallDetails calls }
739 mkBigUD bind dbs calls
741 MkUD { dict_binds = unitBag (mkDB (Rec (bind_prs bind ++ dbsToPairs dbs))),
743 calls = listToCallDetails calls }
745 bind_prs (NonRec b r) = [(b,r)]
746 bind_prs (Rec prs) = prs
749 dbsToPairs ((bind,_):dbs) = bind_prs bind ++ dbsToPairs dbs
751 -- specBindItself deals with the RHS, specialising it according
752 -- to the calls found in the body (if any)
753 specBindItself rhs_subst (NonRec bndr rhs) call_info
754 = specDefn rhs_subst call_info (bndr,rhs) `thenSM` \ ((bndr',rhs'), spec_defns, spec_uds) ->
756 new_bind | null spec_defns = NonRec bndr' rhs'
757 | otherwise = Rec ((bndr',rhs'):spec_defns)
758 -- bndr' mentions the spec_defns in its SpecEnv
759 -- Not sure why we couln't just put the spec_defns first
761 returnSM (new_bind, spec_uds)
763 specBindItself rhs_subst (Rec pairs) call_info
764 = mapSM (specDefn rhs_subst call_info) pairs `thenSM` \ stuff ->
766 (pairs', spec_defns_s, spec_uds_s) = unzip3 stuff
767 spec_defns = concat spec_defns_s
768 spec_uds = plusUDList spec_uds_s
769 new_bind = Rec (spec_defns ++ pairs')
771 returnSM (new_bind, spec_uds)
774 specDefn :: Subst -- Subst to use for RHS
775 -> CallDetails -- Info on how it is used in its scope
776 -> (Id, CoreExpr) -- The thing being bound and its un-processed RHS
777 -> SpecM ((Id, CoreExpr), -- The thing and its processed RHS
778 -- the Id may now have specialisations attached
779 [(Id,CoreExpr)], -- Extra, specialised bindings
780 UsageDetails -- Stuff to fling upwards from the RHS and its
781 ) -- specialised versions
783 specDefn subst calls (fn, rhs)
784 -- The first case is the interesting one
785 | rhs_tyvars `lengthIs` n_tyvars -- Rhs of fn's defn has right number of big lambdas
786 && rhs_bndrs `lengthAtLeast` n_dicts -- and enough dict args
787 && notNull calls_for_me -- And there are some calls to specialise
789 -- At one time I tried not specialising small functions
790 -- but sometimes there are big functions marked INLINE
791 -- that we'd like to specialise. In particular, dictionary
792 -- functions, which Marcin is keen to inline
793 -- && not (certainlyWillInline fn) -- And it's not small
794 -- If it's small, it's better just to inline
795 -- it than to construct lots of specialisations
796 = -- Specialise the body of the function
797 specExpr subst rhs `thenSM` \ (rhs', rhs_uds) ->
799 -- Make a specialised version for each call in calls_for_me
800 mapSM spec_call calls_for_me `thenSM` \ stuff ->
802 (spec_defns, spec_uds, spec_rules) = unzip3 stuff
804 fn' = addIdSpecialisations zapped_fn spec_rules
806 returnSM ((fn',rhs'),
808 rhs_uds `plusUDs` plusUDList spec_uds)
810 | otherwise -- No calls or RHS doesn't fit our preconceptions
811 = specExpr subst rhs `thenSM` \ (rhs', rhs_uds) ->
812 returnSM ((zapped_fn, rhs'), [], rhs_uds)
815 zapped_fn = zapSpecPragmaId fn
816 -- If the fn is a SpecPragmaId, make it discardable
817 -- It's role as a holder for a call instance is o'er
818 -- But it might be alive for some other reason by now.
821 (tyvars, theta, _) = tcSplitSigmaTy fn_type
822 n_tyvars = length tyvars
823 n_dicts = length theta
825 (rhs_tyvars, rhs_ids, rhs_body)
826 = collectTyAndValBinders (dropInline rhs)
827 -- It's important that we "see past" any INLINE pragma
828 -- else we'll fail to specialise an INLINE thing
830 rhs_dicts = take n_dicts rhs_ids
831 rhs_bndrs = rhs_tyvars ++ rhs_dicts
832 body = mkLams (drop n_dicts rhs_ids) rhs_body
833 -- Glue back on the non-dict lambdas
835 calls_for_me = case lookupFM calls fn of
837 Just cs -> fmToList cs
839 ----------------------------------------------------------
840 -- Specialise to one particular call pattern
841 spec_call :: (CallKey, ([DictExpr], VarSet)) -- Call instance
842 -> SpecM ((Id,CoreExpr), -- Specialised definition
843 UsageDetails, -- Usage details from specialised body
844 CoreRule) -- Info for the Id's SpecEnv
845 spec_call (CallKey call_ts, (call_ds, call_fvs))
846 = ASSERT( call_ts `lengthIs` n_tyvars && call_ds `lengthIs` n_dicts )
847 -- Calls are only recorded for properly-saturated applications
849 -- Suppose f's defn is f = /\ a b c d -> \ d1 d2 -> rhs
850 -- Supppose the call is for f [Just t1, Nothing, Just t3, Nothing] [dx1, dx2]
852 -- Construct the new binding
853 -- f1 = SUBST[a->t1,c->t3, d1->d1', d2->d2'] (/\ b d -> rhs)
854 -- PLUS the usage-details
855 -- { d1' = dx1; d2' = dx2 }
856 -- where d1', d2' are cloned versions of d1,d2, with the type substitution applied.
858 -- Note that the substitution is applied to the whole thing.
859 -- This is convenient, but just slightly fragile. Notably:
860 -- * There had better be no name clashes in a/b/c/d
863 -- poly_tyvars = [b,d] in the example above
864 -- spec_tyvars = [a,c]
865 -- ty_args = [t1,b,t3,d]
866 poly_tyvars = [tv | (tv, Nothing) <- rhs_tyvars `zip` call_ts]
867 spec_tyvars = [tv | (tv, Just _) <- rhs_tyvars `zip` call_ts]
868 ty_args = zipWithEqual "spec_call" mk_ty_arg rhs_tyvars call_ts
870 mk_ty_arg rhs_tyvar Nothing = Type (mkTyVarTy rhs_tyvar)
871 mk_ty_arg rhs_tyvar (Just ty) = Type ty
872 rhs_subst = extendTvSubstList subst (spec_tyvars `zip` [ty | Just ty <- call_ts])
874 cloneBinders rhs_subst rhs_dicts `thenSM` \ (rhs_subst', rhs_dicts') ->
876 inst_args = ty_args ++ map Var rhs_dicts'
878 -- Figure out the type of the specialised function
879 body_ty = applyTypeToArgs rhs fn_type inst_args
880 (lam_args, app_args) -- Add a dummy argument if body_ty is unlifted
881 | isUnLiftedType body_ty -- C.f. WwLib.mkWorkerArgs
882 = (poly_tyvars ++ [voidArgId], poly_tyvars ++ [realWorldPrimId])
883 | otherwise = (poly_tyvars, poly_tyvars)
884 spec_id_ty = mkPiTypes lam_args body_ty
886 newIdSM fn spec_id_ty `thenSM` \ spec_f ->
887 specExpr rhs_subst' (mkLams lam_args body) `thenSM` \ (spec_rhs, rhs_uds) ->
889 -- The rule to put in the function's specialisation is:
890 -- forall b,d, d1',d2'. f t1 b t3 d d1' d2' = f1 b d
891 spec_env_rule = Rule (mkFastString ("SPEC " ++ showSDoc (ppr fn)))
893 (poly_tyvars ++ rhs_dicts')
895 (mkVarApps (Var spec_f) app_args)
897 -- Add the { d1' = dx1; d2' = dx2 } usage stuff
898 final_uds = foldr addDictBind rhs_uds (my_zipEqual "spec_call" rhs_dicts' call_ds)
900 -- NOTE: we don't add back in any INLINE pragma on the RHS, so even if
901 -- the original function said INLINE, the specialised copies won't.
902 -- The idea is that the point of inlining was precisely to specialise
903 -- the function at its call site, and that's not so important for the
904 -- specialised copies. But it still smells like an ad hoc decision.
907 returnSM ((spec_f, spec_rhs),
912 my_zipEqual doc xs ys
913 | not (equalLength xs ys) = pprPanic "my_zipEqual" (ppr xs $$ ppr ys $$ (ppr fn <+> ppr call_ts) $$ ppr rhs)
914 | otherwise = zipEqual doc xs ys
916 dropInline :: CoreExpr -> CoreExpr
917 dropInline (Note InlineMe rhs) = rhs
921 %************************************************************************
923 \subsubsection{UsageDetails and suchlike}
925 %************************************************************************
930 dict_binds :: !(Bag DictBind),
931 -- Floated dictionary bindings
932 -- The order is important;
933 -- in ds1 `union` ds2, bindings in ds2 can depend on those in ds1
934 -- (Remember, Bags preserve order in GHC.)
936 calls :: !CallDetails
939 type DictBind = (CoreBind, VarSet)
940 -- The set is the free vars of the binding
941 -- both tyvars and dicts
943 type DictExpr = CoreExpr
945 emptyUDs = MkUD { dict_binds = emptyBag, calls = emptyFM }
947 type ProtoUsageDetails = ([DictBind],
948 [(Id, CallKey, ([DictExpr], VarSet))]
951 ------------------------------------------------------------
952 type CallDetails = FiniteMap Id CallInfo
953 newtype CallKey = CallKey [Maybe Type] -- Nothing => unconstrained type argument
954 type CallInfo = FiniteMap CallKey
955 ([DictExpr], VarSet) -- Dict args and the vars of the whole
956 -- call (including tyvars)
957 -- [*not* include the main id itself, of course]
958 -- The finite maps eliminate duplicates
959 -- The list of types and dictionaries is guaranteed to
960 -- match the type of f
962 -- Type isn't an instance of Ord, so that we can control which
963 -- instance we use. That's tiresome here. Oh well
964 instance Eq CallKey where
965 k1 == k2 = case k1 `compare` k2 of { EQ -> True; other -> False }
967 instance Ord CallKey where
968 compare (CallKey k1) (CallKey k2) = cmpList cmp k1 k2
970 cmp Nothing Nothing = EQ
971 cmp Nothing (Just t2) = LT
972 cmp (Just t1) Nothing = GT
973 cmp (Just t1) (Just t2) = tcCmpType t1 t2
975 unionCalls :: CallDetails -> CallDetails -> CallDetails
976 unionCalls c1 c2 = plusFM_C plusFM c1 c2
978 singleCall :: Id -> [Maybe Type] -> [DictExpr] -> CallDetails
979 singleCall id tys dicts
980 = unitFM id (unitFM (CallKey tys) (dicts, call_fvs))
982 call_fvs = exprsFreeVars dicts `unionVarSet` tys_fvs
983 tys_fvs = tyVarsOfTypes (catMaybes tys)
984 -- The type args (tys) are guaranteed to be part of the dictionary
985 -- types, because they are just the constrained types,
986 -- and the dictionary is therefore sure to be bound
987 -- inside the binding for any type variables free in the type;
988 -- hence it's safe to neglect tyvars free in tys when making
989 -- the free-var set for this call
990 -- BUT I don't trust this reasoning; play safe and include tys_fvs
992 -- We don't include the 'id' itself.
994 listToCallDetails calls
995 = foldr (unionCalls . mk_call) emptyFM calls
997 mk_call (id, tys, dicts_w_fvs) = unitFM id (unitFM tys dicts_w_fvs)
998 -- NB: the free vars of the call are provided
1000 callDetailsToList calls = [ (id,tys,dicts)
1001 | (id,fm) <- fmToList calls,
1002 (tys, dicts) <- fmToList fm
1005 mkCallUDs subst f args
1007 || not (all isClassPred theta)
1008 -- Only specialise if all overloading is on class params.
1009 -- In ptic, with implicit params, the type args
1010 -- *don't* say what the value of the implicit param is!
1011 || not (spec_tys `lengthIs` n_tyvars)
1012 || not ( dicts `lengthIs` n_dicts)
1013 || maybeToBool (lookupRule (\act -> True) (substInScope subst) emptyRuleBase f args)
1014 -- There's already a rule covering this call. A typical case
1015 -- is where there's an explicit user-provided rule. Then
1016 -- we don't want to create a specialised version
1017 -- of the function that overlaps.
1018 = emptyUDs -- Not overloaded, or no specialisation wanted
1021 = MkUD {dict_binds = emptyBag,
1022 calls = singleCall f spec_tys dicts
1025 (tyvars, theta, _) = tcSplitSigmaTy (idType f)
1026 constrained_tyvars = tyVarsOfTheta theta
1027 n_tyvars = length tyvars
1028 n_dicts = length theta
1030 spec_tys = [mk_spec_ty tv ty | (tv, Type ty) <- tyvars `zip` args]
1031 dicts = [dict_expr | (_, dict_expr) <- theta `zip` (drop n_tyvars args)]
1034 | tyvar `elemVarSet` constrained_tyvars = Just ty
1035 | otherwise = Nothing
1037 ------------------------------------------------------------
1038 plusUDs :: UsageDetails -> UsageDetails -> UsageDetails
1039 plusUDs (MkUD {dict_binds = db1, calls = calls1})
1040 (MkUD {dict_binds = db2, calls = calls2})
1041 = MkUD {dict_binds = d, calls = c}
1043 d = db1 `unionBags` db2
1044 c = calls1 `unionCalls` calls2
1046 plusUDList = foldr plusUDs emptyUDs
1048 -- zapCalls deletes calls to ids from uds
1049 zapCalls ids uds = uds {calls = delListFromFM (calls uds) ids}
1051 mkDB bind = (bind, bind_fvs bind)
1053 bind_fvs (NonRec bndr rhs) = exprFreeVars rhs
1054 bind_fvs (Rec prs) = foldl delVarSet rhs_fvs bndrs
1057 rhs_fvs = unionVarSets [exprFreeVars rhs | (bndr,rhs) <- prs]
1059 addDictBind (dict,rhs) uds = uds { dict_binds = mkDB (NonRec dict rhs) `consBag` dict_binds uds }
1061 dumpAllDictBinds (MkUD {dict_binds = dbs}) binds
1062 = foldrBag add binds dbs
1064 add (bind,_) binds = bind : binds
1066 dumpUDs :: [CoreBndr]
1067 -> UsageDetails -> CoreExpr
1068 -> (UsageDetails, CoreExpr)
1069 dumpUDs bndrs uds body
1070 = (free_uds, foldr add_let body dict_binds)
1072 (free_uds, (dict_binds, _)) = splitUDs bndrs uds
1073 add_let (bind,_) body = Let bind body
1075 splitUDs :: [CoreBndr]
1077 -> (UsageDetails, -- These don't mention the binders
1078 ProtoUsageDetails) -- These do
1080 splitUDs bndrs uds@(MkUD {dict_binds = orig_dbs,
1081 calls = orig_calls})
1083 = if isEmptyBag dump_dbs && null dump_calls then
1084 -- Common case: binder doesn't affect floats
1088 -- Binders bind some of the fvs of the floats
1089 (MkUD {dict_binds = free_dbs,
1090 calls = listToCallDetails free_calls},
1091 (bagToList dump_dbs, dump_calls)
1095 bndr_set = mkVarSet bndrs
1097 (free_dbs, dump_dbs, dump_idset)
1098 = foldlBag dump_db (emptyBag, emptyBag, bndr_set) orig_dbs
1099 -- Important that it's foldl not foldr;
1100 -- we're accumulating the set of dumped ids in dump_set
1102 -- Filter out any calls that mention things that are being dumped
1103 orig_call_list = callDetailsToList orig_calls
1104 (dump_calls, free_calls) = partition captured orig_call_list
1105 captured (id,tys,(dicts, fvs)) = fvs `intersectsVarSet` dump_idset
1106 || id `elemVarSet` dump_idset
1108 dump_db (free_dbs, dump_dbs, dump_idset) db@(bind, fvs)
1109 | dump_idset `intersectsVarSet` fvs -- Dump it
1110 = (free_dbs, dump_dbs `snocBag` db,
1111 extendVarSetList dump_idset (bindersOf bind))
1113 | otherwise -- Don't dump it
1114 = (free_dbs `snocBag` db, dump_dbs, dump_idset)
1118 %************************************************************************
1120 \subsubsection{Boring helper functions}
1122 %************************************************************************
1125 type SpecM a = UniqSM a
1129 getUniqSM = getUniqueUs
1133 mapAndCombineSM f [] = returnSM ([], emptyUDs)
1134 mapAndCombineSM f (x:xs) = f x `thenSM` \ (y, uds1) ->
1135 mapAndCombineSM f xs `thenSM` \ (ys, uds2) ->
1136 returnSM (y:ys, uds1 `plusUDs` uds2)
1138 cloneBindSM :: Subst -> CoreBind -> SpecM (Subst, Subst, CoreBind)
1139 -- Clone the binders of the bind; return new bind with the cloned binders
1140 -- Return the substitution to use for RHSs, and the one to use for the body
1141 cloneBindSM subst (NonRec bndr rhs)
1142 = getUs `thenUs` \ us ->
1144 (subst', bndr') = cloneIdBndr subst us bndr
1146 returnUs (subst, subst', NonRec bndr' rhs)
1148 cloneBindSM subst (Rec pairs)
1149 = getUs `thenUs` \ us ->
1151 (subst', bndrs') = cloneRecIdBndrs subst us (map fst pairs)
1153 returnUs (subst', subst', Rec (bndrs' `zip` map snd pairs))
1155 cloneBinders subst bndrs
1156 = getUs `thenUs` \ us ->
1157 returnUs (cloneIdBndrs subst us bndrs)
1159 newIdSM old_id new_ty
1160 = getUniqSM `thenSM` \ uniq ->
1162 -- Give the new Id a similar occurrence name to the old one
1163 name = idName old_id
1164 new_id = mkUserLocal (mkSpecOcc (nameOccName name)) uniq new_ty (getSrcLoc name)
1170 Old (but interesting) stuff about unboxed bindings
1171 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1173 What should we do when a value is specialised to a *strict* unboxed value?
1175 map_*_* f (x:xs) = let h = f x
1179 Could convert let to case:
1181 map_*_Int# f (x:xs) = case f x of h# ->
1185 This may be undesirable since it forces evaluation here, but the value
1186 may not be used in all branches of the body. In the general case this
1187 transformation is impossible since the mutual recursion in a letrec
1188 cannot be expressed as a case.
1190 There is also a problem with top-level unboxed values, since our
1191 implementation cannot handle unboxed values at the top level.
1193 Solution: Lift the binding of the unboxed value and extract it when it
1196 map_*_Int# f (x:xs) = let h = case (f x) of h# -> _Lift h#
1201 Now give it to the simplifier and the _Lifting will be optimised away.
1203 The benfit is that we have given the specialised "unboxed" values a
1204 very simplep lifted semantics and then leave it up to the simplifier to
1205 optimise it --- knowing that the overheads will be removed in nearly
1208 In particular, the value will only be evaluted in the branches of the
1209 program which use it, rather than being forced at the point where the
1210 value is bound. For example:
1212 filtermap_*_* p f (x:xs)
1219 filtermap_*_Int# p f (x:xs)
1220 = let h = case (f x) of h# -> _Lift h#
1223 True -> case h of _Lift h#
1227 The binding for h can still be inlined in the one branch and the
1228 _Lifting eliminated.
1231 Question: When won't the _Lifting be eliminated?
1233 Answer: When they at the top-level (where it is necessary) or when
1234 inlining would duplicate work (or possibly code depending on
1235 options). However, the _Lifting will still be eliminated if the
1236 strictness analyser deems the lifted binding strict.