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 Subst ( Subst, SubstResult(..), mkSubst, mkSubst, extendTvSubstList,
18 simplBndr, simplBndrs, substTy,
19 substAndCloneId, substAndCloneIds, substAndCloneRecIds,
22 import Var ( zapSpecPragmaId )
26 import CoreUtils ( applyTypeToArgs, mkPiTypes )
27 import CoreFVs ( exprFreeVars, exprsFreeVars )
28 import CoreTidy ( pprTidyIdRules )
29 import CoreLint ( showPass, endPass )
30 import Rules ( addIdSpecialisations, lookupRule )
32 import UniqSupply ( UniqSupply,
33 UniqSM, initUs_, thenUs, returnUs, getUniqueUs,
36 import Name ( nameOccName, mkSpecOcc, getSrcLoc )
37 import MkId ( voidArgId, realWorldPrimId )
39 import Maybes ( catMaybes, maybeToBool )
40 import ErrUtils ( dumpIfSet_dyn )
41 import BasicTypes ( Activation( AlwaysActive ) )
43 import List ( partition )
44 import Util ( zipEqual, zipWithEqual, cmpList, lengthIs,
45 equalLength, lengthAtLeast, notNull )
52 %************************************************************************
54 \subsection[notes-Specialise]{Implementation notes [SLPJ, Aug 18 1993]}
56 %************************************************************************
58 These notes describe how we implement specialisation to eliminate
61 The specialisation pass works on Core
62 syntax, complete with all the explicit dictionary application,
63 abstraction and construction as added by the type checker. The
64 existing type checker remains largely as it is.
66 One important thought: the {\em types} passed to an overloaded
67 function, and the {\em dictionaries} passed are mutually redundant.
68 If the same function is applied to the same type(s) then it is sure to
69 be applied to the same dictionary(s)---or rather to the same {\em
70 values}. (The arguments might look different but they will evaluate
73 Second important thought: we know that we can make progress by
74 treating dictionary arguments as static and worth specialising on. So
75 we can do without binding-time analysis, and instead specialise on
76 dictionary arguments and no others.
85 and suppose f is overloaded.
87 STEP 1: CALL-INSTANCE COLLECTION
89 We traverse <body>, accumulating all applications of f to types and
92 (Might there be partial applications, to just some of its types and
93 dictionaries? In principle yes, but in practice the type checker only
94 builds applications of f to all its types and dictionaries, so partial
95 applications could only arise as a result of transformation, and even
96 then I think it's unlikely. In any case, we simply don't accumulate such
97 partial applications.)
102 So now we have a collection of calls to f:
106 Notice that f may take several type arguments. To avoid ambiguity, we
107 say that f is called at type t1/t2 and t3/t4.
109 We take equivalence classes using equality of the *types* (ignoring
110 the dictionary args, which as mentioned previously are redundant).
112 STEP 3: SPECIALISATION
114 For each equivalence class, choose a representative (f t1 t2 d1 d2),
115 and create a local instance of f, defined thus:
117 f@t1/t2 = <f_rhs> t1 t2 d1 d2
119 f_rhs presumably has some big lambdas and dictionary lambdas, so lots
120 of simplification will now result. However we don't actually *do* that
121 simplification. Rather, we leave it for the simplifier to do. If we
122 *did* do it, though, we'd get more call instances from the specialised
123 RHS. We can work out what they are by instantiating the call-instance
124 set from f's RHS with the types t1, t2.
126 Add this new id to f's IdInfo, to record that f has a specialised version.
128 Before doing any of this, check that f's IdInfo doesn't already
129 tell us about an existing instance of f at the required type/s.
130 (This might happen if specialisation was applied more than once, or
131 it might arise from user SPECIALIZE pragmas.)
135 Wait a minute! What if f is recursive? Then we can't just plug in
136 its right-hand side, can we?
138 But it's ok. The type checker *always* creates non-recursive definitions
139 for overloaded recursive functions. For example:
141 f x = f (x+x) -- Yes I know its silly
145 f a (d::Num a) = let p = +.sel a d
147 letrec fl (y::a) = fl (p y y)
151 We still have recusion for non-overloaded functions which we
152 speciailise, but the recursive call should get specialised to the
153 same recursive version.
159 All this is crystal clear when the function is applied to *constant
160 types*; that is, types which have no type variables inside. But what if
161 it is applied to non-constant types? Suppose we find a call of f at type
162 t1/t2. There are two possibilities:
164 (a) The free type variables of t1, t2 are in scope at the definition point
165 of f. In this case there's no problem, we proceed just as before. A common
166 example is as follows. Here's the Haskell:
171 After typechecking we have
173 g a (d::Num a) (y::a) = let f b (d'::Num b) (x::b) = +.sel b d' x x
174 in +.sel a d (f a d y) (f a d y)
176 Notice that the call to f is at type type "a"; a non-constant type.
177 Both calls to f are at the same type, so we can specialise to give:
179 g a (d::Num a) (y::a) = let f@a (x::a) = +.sel a d x x
180 in +.sel a d (f@a y) (f@a y)
183 (b) The other case is when the type variables in the instance types
184 are *not* in scope at the definition point of f. The example we are
185 working with above is a good case. There are two instances of (+.sel a d),
186 but "a" is not in scope at the definition of +.sel. Can we do anything?
187 Yes, we can "common them up", a sort of limited common sub-expression deal.
190 g a (d::Num a) (y::a) = let +.sel@a = +.sel a d
191 f@a (x::a) = +.sel@a x x
192 in +.sel@a (f@a y) (f@a y)
194 This can save work, and can't be spotted by the type checker, because
195 the two instances of +.sel weren't originally at the same type.
199 * There are quite a few variations here. For example, the defn of
200 +.sel could be floated ouside the \y, to attempt to gain laziness.
201 It certainly mustn't be floated outside the \d because the d has to
204 * We don't want to inline f_rhs in this case, because
205 that will duplicate code. Just commoning up the call is the point.
207 * Nothing gets added to +.sel's IdInfo.
209 * Don't bother unless the equivalence class has more than one item!
211 Not clear whether this is all worth it. It is of course OK to
212 simply discard call-instances when passing a big lambda.
214 Polymorphism 2 -- Overloading
216 Consider a function whose most general type is
218 f :: forall a b. Ord a => [a] -> b -> b
220 There is really no point in making a version of g at Int/Int and another
221 at Int/Bool, because it's only instancing the type variable "a" which
222 buys us any efficiency. Since g is completely polymorphic in b there
223 ain't much point in making separate versions of g for the different
226 That suggests that we should identify which of g's type variables
227 are constrained (like "a") and which are unconstrained (like "b").
228 Then when taking equivalence classes in STEP 2, we ignore the type args
229 corresponding to unconstrained type variable. In STEP 3 we make
230 polymorphic versions. Thus:
232 f@t1/ = /\b -> <f_rhs> t1 b d1 d2
241 f a (d::Num a) = let g = ...
243 ...(let d1::Ord a = Num.Ord.sel a d in g a d1)...
245 Here, g is only called at one type, but the dictionary isn't in scope at the
246 definition point for g. Usually the type checker would build a
247 definition for d1 which enclosed g, but the transformation system
248 might have moved d1's defn inward. Solution: float dictionary bindings
249 outwards along with call instances.
253 f x = let g p q = p==q
259 Before specialisation, leaving out type abstractions we have
261 f df x = let g :: Eq a => a -> a -> Bool
263 h :: Num a => a -> a -> (a, Bool)
264 h dh r s = let deq = eqFromNum dh
265 in (+ dh r s, g deq r s)
269 After specialising h we get a specialised version of h, like this:
271 h' r s = let deq = eqFromNum df
272 in (+ df r s, g deq r s)
274 But we can't naively make an instance for g from this, because deq is not in scope
275 at the defn of g. Instead, we have to float out the (new) defn of deq
276 to widen its scope. Notice that this floating can't be done in advance -- it only
277 shows up when specialisation is done.
279 User SPECIALIZE pragmas
280 ~~~~~~~~~~~~~~~~~~~~~~~
281 Specialisation pragmas can be digested by the type checker, and implemented
282 by adding extra definitions along with that of f, in the same way as before
284 f@t1/t2 = <f_rhs> t1 t2 d1 d2
286 Indeed the pragmas *have* to be dealt with by the type checker, because
287 only it knows how to build the dictionaries d1 and d2! For example
289 g :: Ord a => [a] -> [a]
290 {-# SPECIALIZE f :: [Tree Int] -> [Tree Int] #-}
292 Here, the specialised version of g is an application of g's rhs to the
293 Ord dictionary for (Tree Int), which only the type checker can conjure
294 up. There might not even *be* one, if (Tree Int) is not an instance of
295 Ord! (All the other specialision has suitable dictionaries to hand
298 Problem. The type checker doesn't have to hand a convenient <f_rhs>, because
299 it is buried in a complex (as-yet-un-desugared) binding group.
302 f@t1/t2 = f* t1 t2 d1 d2
304 where f* is the Id f with an IdInfo which says "inline me regardless!".
305 Indeed all the specialisation could be done in this way.
306 That in turn means that the simplifier has to be prepared to inline absolutely
307 any in-scope let-bound thing.
310 Again, the pragma should permit polymorphism in unconstrained variables:
312 h :: Ord a => [a] -> b -> b
313 {-# SPECIALIZE h :: [Int] -> b -> b #-}
315 We *insist* that all overloaded type variables are specialised to ground types,
316 (and hence there can be no context inside a SPECIALIZE pragma).
317 We *permit* unconstrained type variables to be specialised to
319 - or left as a polymorphic type variable
320 but nothing in between. So
322 {-# SPECIALIZE h :: [Int] -> [c] -> [c] #-}
324 is *illegal*. (It can be handled, but it adds complication, and gains the
328 SPECIALISING INSTANCE DECLARATIONS
329 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
332 instance Foo a => Foo [a] where
334 {-# SPECIALIZE instance Foo [Int] #-}
336 The original instance decl creates a dictionary-function
339 dfun.Foo.List :: forall a. Foo a -> Foo [a]
341 The SPECIALIZE pragma just makes a specialised copy, just as for
342 ordinary function definitions:
344 dfun.Foo.List@Int :: Foo [Int]
345 dfun.Foo.List@Int = dfun.Foo.List Int dFooInt
347 The information about what instance of the dfun exist gets added to
348 the dfun's IdInfo in the same way as a user-defined function too.
351 Automatic instance decl specialisation?
352 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
353 Can instance decls be specialised automatically? It's tricky.
354 We could collect call-instance information for each dfun, but
355 then when we specialised their bodies we'd get new call-instances
356 for ordinary functions; and when we specialised their bodies, we might get
357 new call-instances of the dfuns, and so on. This all arises because of
358 the unrestricted mutual recursion between instance decls and value decls.
360 Still, there's no actual problem; it just means that we may not do all
361 the specialisation we could theoretically do.
363 Furthermore, instance decls are usually exported and used non-locally,
364 so we'll want to compile enough to get those specialisations done.
366 Lastly, there's no such thing as a local instance decl, so we can
367 survive solely by spitting out *usage* information, and then reading that
368 back in as a pragma when next compiling the file. So for now,
369 we only specialise instance decls in response to pragmas.
372 SPITTING OUT USAGE INFORMATION
373 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
375 To spit out usage information we need to traverse the code collecting
376 call-instance information for all imported (non-prelude?) functions
377 and data types. Then we equivalence-class it and spit it out.
379 This is done at the top-level when all the call instances which escape
380 must be for imported functions and data types.
382 *** Not currently done ***
385 Partial specialisation by pragmas
386 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
387 What about partial specialisation:
389 k :: (Ord a, Eq b) => [a] -> b -> b -> [a]
390 {-# SPECIALIZE k :: Eq b => [Int] -> b -> b -> [a] #-}
394 {-# SPECIALIZE k :: Eq b => [Int] -> [b] -> [b] -> [a] #-}
396 Seems quite reasonable. Similar things could be done with instance decls:
398 instance (Foo a, Foo b) => Foo (a,b) where
400 {-# SPECIALIZE instance Foo a => Foo (a,Int) #-}
401 {-# SPECIALIZE instance Foo b => Foo (Int,b) #-}
403 Ho hum. Things are complex enough without this. I pass.
406 Requirements for the simplifer
407 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
408 The simplifier has to be able to take advantage of the specialisation.
410 * When the simplifier finds an application of a polymorphic f, it looks in
411 f's IdInfo in case there is a suitable instance to call instead. This converts
413 f t1 t2 d1 d2 ===> f_t1_t2
415 Note that the dictionaries get eaten up too!
417 * Dictionary selection operations on constant dictionaries must be
420 +.sel Int d ===> +Int
422 The obvious way to do this is in the same way as other specialised
423 calls: +.sel has inside it some IdInfo which tells that if it's applied
424 to the type Int then it should eat a dictionary and transform to +Int.
426 In short, dictionary selectors need IdInfo inside them for constant
429 * Exactly the same applies if a superclass dictionary is being
432 Eq.sel Int d ===> dEqInt
434 * Something similar applies to dictionary construction too. Suppose
435 dfun.Eq.List is the function taking a dictionary for (Eq a) to
436 one for (Eq [a]). Then we want
438 dfun.Eq.List Int d ===> dEq.List_Int
440 Where does the Eq [Int] dictionary come from? It is built in
441 response to a SPECIALIZE pragma on the Eq [a] instance decl.
443 In short, dfun Ids need IdInfo with a specialisation for each
444 constant instance of their instance declaration.
446 All this uses a single mechanism: the SpecEnv inside an Id
449 What does the specialisation IdInfo look like?
450 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
452 The SpecEnv of an Id maps a list of types (the template) to an expression
456 For example, if f has this SpecInfo:
458 [Int, a] -> \d:Ord Int. f' a
460 it means that we can replace the call
462 f Int t ===> (\d. f' t)
464 This chucks one dictionary away and proceeds with the
465 specialised version of f, namely f'.
468 What can't be done this way?
469 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
470 There is no way, post-typechecker, to get a dictionary for (say)
471 Eq a from a dictionary for Eq [a]. So if we find
475 we can't transform to
480 eqList :: (a->a->Bool) -> [a] -> [a] -> Bool
482 Of course, we currently have no way to automatically derive
483 eqList, nor to connect it to the Eq [a] instance decl, but you
484 can imagine that it might somehow be possible. Taking advantage
485 of this is permanently ruled out.
487 Still, this is no great hardship, because we intend to eliminate
488 overloading altogether anyway!
492 A note about non-tyvar dictionaries
493 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
494 Some Ids have types like
496 forall a,b,c. Eq a -> Ord [a] -> tau
498 This seems curious at first, because we usually only have dictionary
499 args whose types are of the form (C a) where a is a type variable.
500 But this doesn't hold for the functions arising from instance decls,
501 which sometimes get arguements with types of form (C (T a)) for some
504 Should we specialise wrt this compound-type dictionary? We used to say
506 "This is a heuristic judgement, as indeed is the fact that we
507 specialise wrt only dictionaries. We choose *not* to specialise
508 wrt compound dictionaries because at the moment the only place
509 they show up is in instance decls, where they are simply plugged
510 into a returned dictionary. So nothing is gained by specialising
513 But it is simpler and more uniform to specialise wrt these dicts too;
514 and in future GHC is likely to support full fledged type signatures
516 f ;: Eq [(a,b)] => ...
519 %************************************************************************
521 \subsubsection{The new specialiser}
523 %************************************************************************
525 Our basic game plan is this. For let(rec) bound function
526 f :: (C a, D c) => (a,b,c,d) -> Bool
528 * Find any specialised calls of f, (f ts ds), where
529 ts are the type arguments t1 .. t4, and
530 ds are the dictionary arguments d1 .. d2.
532 * Add a new definition for f1 (say):
534 f1 = /\ b d -> (..body of f..) t1 b t3 d d1 d2
536 Note that we abstract over the unconstrained type arguments.
540 [t1,b,t3,d] |-> \d1 d2 -> f1 b d
542 to the specialisations of f. This will be used by the
543 simplifier to replace calls
544 (f t1 t2 t3 t4) da db
546 (\d1 d1 -> f1 t2 t4) da db
548 All the stuff about how many dictionaries to discard, and what types
549 to apply the specialised function to, are handled by the fact that the
550 SpecEnv contains a template for the result of the specialisation.
552 We don't build *partial* specialisations for f. For example:
554 f :: Eq a => a -> a -> Bool
555 {-# SPECIALISE f :: (Eq b, Eq c) => (b,c) -> (b,c) -> Bool #-}
557 Here, little is gained by making a specialised copy of f.
558 There's a distinct danger that the specialised version would
559 first build a dictionary for (Eq b, Eq c), and then select the (==)
560 method from it! Even if it didn't, not a great deal is saved.
562 We do, however, generate polymorphic, but not overloaded, specialisations:
564 f :: Eq a => [a] -> b -> b -> b
565 {#- SPECIALISE f :: [Int] -> b -> b -> b #-}
567 Hence, the invariant is this:
569 *** no specialised version is overloaded ***
572 %************************************************************************
574 \subsubsection{The exported function}
576 %************************************************************************
579 specProgram :: DynFlags -> UniqSupply -> [CoreBind] -> IO [CoreBind]
580 specProgram dflags us binds
582 showPass dflags "Specialise"
584 let binds' = initSM us (go binds `thenSM` \ (binds', uds') ->
585 returnSM (dumpAllDictBinds uds' binds'))
587 endPass dflags "Specialise" Opt_D_dump_spec binds'
589 dumpIfSet_dyn dflags Opt_D_dump_rules "Top-level specialisations"
590 (vcat (map pprTidyIdRules (concat (map bindersOf binds'))))
594 -- We need to start with a Subst that knows all the things
595 -- that are in scope, so that the substitution engine doesn't
596 -- accidentally re-use a unique that's already in use
597 -- Easiest thing is to do it all at once, as if all the top-level
598 -- decls were mutually recursive
599 top_subst = mkSubst (mkInScopeSet (mkVarSet (bindersOfBinds binds)))
601 go [] = returnSM ([], emptyUDs)
602 go (bind:binds) = go binds `thenSM` \ (binds', uds) ->
603 specBind top_subst bind uds `thenSM` \ (bind', uds') ->
604 returnSM (bind' ++ binds', uds')
607 %************************************************************************
609 \subsubsection{@specExpr@: the main function}
611 %************************************************************************
614 specVar :: Subst -> Id -> CoreExpr
615 specVar subst v = case substId subst v of
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) = returnSM (Type (substTy subst ty), emptyUDs)
628 specExpr subst (Var v) = returnSM (specVar subst v, emptyUDs)
629 specExpr subst (Lit lit) = returnSM (Lit lit, emptyUDs)
631 specExpr subst (Note note body)
632 = specExpr subst body `thenSM` \ (body', uds) ->
633 returnSM (Note (specNote subst note) body', uds)
636 ---------------- Applications might generate a call instance --------------------
637 specExpr subst expr@(App fun arg)
640 go (App fun arg) args = specExpr subst arg `thenSM` \ (arg', uds_arg) ->
641 go fun (arg':args) `thenSM` \ (fun', uds_app) ->
642 returnSM (App fun' arg', uds_arg `plusUDs` uds_app)
644 go (Var f) args = case specVar subst f of
645 Var f' -> returnSM (Var f', mkCallUDs subst f' args)
646 e' -> returnSM (e', emptyUDs) -- I don't expect this!
647 go other args = specExpr subst other
649 ---------------- Lambda/case require dumping of usage details --------------------
650 specExpr subst e@(Lam _ _)
651 = specExpr subst' body `thenSM` \ (body', uds) ->
653 (filtered_uds, body'') = dumpUDs bndrs' uds body'
655 returnSM (mkLams bndrs' body'', filtered_uds)
657 (bndrs, body) = collectBinders e
658 (subst', bndrs') = simplBndrs 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
663 specExpr subst (Case scrut case_bndr ty alts)
664 = specExpr subst scrut `thenSM` \ (scrut', uds_scrut) ->
665 mapAndCombineSM spec_alt alts `thenSM` \ (alts', uds_alts) ->
666 returnSM (Case scrut' case_bndr' (substTy subst ty) alts', uds_scrut `plusUDs` uds_alts)
668 (subst_alt, case_bndr') = simplBndr subst case_bndr
669 -- No need to clone case binder; it can't float like a let(rec)
671 spec_alt (con, args, rhs)
672 = specExpr subst_rhs rhs `thenSM` \ (rhs', uds) ->
674 (uds', rhs'') = dumpUDs args uds rhs'
676 returnSM ((con, args', rhs''), uds')
678 (subst_rhs, args') = simplBndrs subst_alt args
680 ---------------- Finally, let is the interesting case --------------------
681 specExpr subst (Let bind body)
683 cloneBindSM subst bind `thenSM` \ (rhs_subst, body_subst, bind') ->
685 -- Deal with the body
686 specExpr body_subst body `thenSM` \ (body', body_uds) ->
688 -- Deal with the bindings
689 specBind rhs_subst bind' body_uds `thenSM` \ (binds', uds) ->
692 returnSM (foldr Let body' binds', uds)
694 -- Must apply the type substitution to coerceions
695 specNote subst (Coerce t1 t2) = Coerce (substTy subst t1) (substTy subst t2)
696 specNote subst note = note
699 %************************************************************************
701 \subsubsection{Dealing with a binding}
703 %************************************************************************
706 specBind :: Subst -- Use this for RHSs
708 -> UsageDetails -- Info on how the scope of the binding
709 -> SpecM ([CoreBind], -- New bindings
710 UsageDetails) -- And info to pass upstream
712 specBind rhs_subst bind body_uds
713 = specBindItself rhs_subst bind (calls body_uds) `thenSM` \ (bind', bind_uds) ->
715 bndrs = bindersOf bind
716 all_uds = zapCalls bndrs (body_uds `plusUDs` bind_uds)
717 -- It's important that the `plusUDs` is this way round,
718 -- because body_uds may bind dictionaries that are
719 -- used in the calls passed to specDefn. So the
720 -- dictionary bindings in bind_uds may mention
721 -- dictionaries bound in body_uds.
723 case splitUDs bndrs all_uds of
725 (_, ([],[])) -- This binding doesn't bind anything needed
726 -- in the UDs, so put the binding here
727 -- This is the case for most non-dict bindings, except
728 -- for the few that are mentioned in a dict binding
729 -- that is floating upwards in body_uds
730 -> returnSM ([bind'], all_uds)
732 (float_uds, (dict_binds, calls)) -- This binding is needed in the UDs, so float it out
733 -> returnSM ([], float_uds `plusUDs` mkBigUD bind' dict_binds calls)
736 -- A truly gruesome function
737 mkBigUD bind@(NonRec _ _) dbs calls
738 = -- Common case: non-recursive and no specialisations
739 -- (if there were any specialistions it would have been made recursive)
740 MkUD { dict_binds = listToBag (mkDB bind : dbs),
741 calls = listToCallDetails calls }
743 mkBigUD bind dbs calls
745 MkUD { dict_binds = unitBag (mkDB (Rec (bind_prs bind ++ dbsToPairs dbs))),
747 calls = listToCallDetails calls }
749 bind_prs (NonRec b r) = [(b,r)]
750 bind_prs (Rec prs) = prs
753 dbsToPairs ((bind,_):dbs) = bind_prs bind ++ dbsToPairs dbs
755 -- specBindItself deals with the RHS, specialising it according
756 -- to the calls found in the body (if any)
757 specBindItself rhs_subst (NonRec bndr rhs) call_info
758 = specDefn rhs_subst call_info (bndr,rhs) `thenSM` \ ((bndr',rhs'), spec_defns, spec_uds) ->
760 new_bind | null spec_defns = NonRec bndr' rhs'
761 | otherwise = Rec ((bndr',rhs'):spec_defns)
762 -- bndr' mentions the spec_defns in its SpecEnv
763 -- Not sure why we couln't just put the spec_defns first
765 returnSM (new_bind, spec_uds)
767 specBindItself rhs_subst (Rec pairs) call_info
768 = mapSM (specDefn rhs_subst call_info) pairs `thenSM` \ stuff ->
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')
775 returnSM (new_bind, spec_uds)
778 specDefn :: Subst -- Subst to use for RHS
779 -> CallDetails -- Info on how it is used in its scope
780 -> (Id, CoreExpr) -- The thing being bound and its un-processed RHS
781 -> SpecM ((Id, CoreExpr), -- The thing and its processed RHS
782 -- the Id may now have specialisations attached
783 [(Id,CoreExpr)], -- Extra, specialised bindings
784 UsageDetails -- Stuff to fling upwards from the RHS and its
785 ) -- specialised versions
787 specDefn subst calls (fn, rhs)
788 -- The first case is the interesting one
789 | rhs_tyvars `lengthIs` n_tyvars -- Rhs of fn's defn has right number of big lambdas
790 && rhs_bndrs `lengthAtLeast` n_dicts -- and enough dict args
791 && notNull calls_for_me -- And there are some calls to specialise
793 -- At one time I tried not specialising small functions
794 -- but sometimes there are big functions marked INLINE
795 -- that we'd like to specialise. In particular, dictionary
796 -- functions, which Marcin is keen to inline
797 -- && not (certainlyWillInline fn) -- And it's not small
798 -- If it's small, it's better just to inline
799 -- it than to construct lots of specialisations
800 = -- Specialise the body of the function
801 specExpr subst rhs `thenSM` \ (rhs', rhs_uds) ->
803 -- Make a specialised version for each call in calls_for_me
804 mapSM spec_call calls_for_me `thenSM` \ stuff ->
806 (spec_defns, spec_uds, spec_rules) = unzip3 stuff
808 fn' = addIdSpecialisations zapped_fn spec_rules
810 returnSM ((fn',rhs'),
812 rhs_uds `plusUDs` plusUDList spec_uds)
814 | otherwise -- No calls or RHS doesn't fit our preconceptions
815 = specExpr subst rhs `thenSM` \ (rhs', rhs_uds) ->
816 returnSM ((zapped_fn, rhs'), [], rhs_uds)
819 zapped_fn = zapSpecPragmaId fn
820 -- If the fn is a SpecPragmaId, make it discardable
821 -- It's role as a holder for a call instance is o'er
822 -- But it might be alive for some other reason by now.
825 (tyvars, theta, _) = tcSplitSigmaTy fn_type
826 n_tyvars = length tyvars
827 n_dicts = length theta
829 (rhs_tyvars, rhs_ids, rhs_body)
830 = collectTyAndValBinders (dropInline rhs)
831 -- It's important that we "see past" any INLINE pragma
832 -- else we'll fail to specialise an INLINE thing
834 rhs_dicts = take n_dicts rhs_ids
835 rhs_bndrs = rhs_tyvars ++ rhs_dicts
836 body = mkLams (drop n_dicts rhs_ids) rhs_body
837 -- Glue back on the non-dict lambdas
839 calls_for_me = case lookupFM calls fn of
841 Just cs -> fmToList cs
843 ----------------------------------------------------------
844 -- Specialise to one particular call pattern
845 spec_call :: (CallKey, ([DictExpr], VarSet)) -- Call instance
846 -> SpecM ((Id,CoreExpr), -- Specialised definition
847 UsageDetails, -- Usage details from specialised body
848 CoreRule) -- Info for the Id's SpecEnv
849 spec_call (CallKey call_ts, (call_ds, call_fvs))
850 = ASSERT( call_ts `lengthIs` n_tyvars && call_ds `lengthIs` n_dicts )
851 -- Calls are only recorded for properly-saturated applications
853 -- Suppose f's defn is f = /\ a b c d -> \ d1 d2 -> rhs
854 -- Supppose the call is for f [Just t1, Nothing, Just t3, Nothing] [dx1, dx2]
856 -- Construct the new binding
857 -- f1 = SUBST[a->t1,c->t3, d1->d1', d2->d2'] (/\ b d -> rhs)
858 -- PLUS the usage-details
859 -- { d1' = dx1; d2' = dx2 }
860 -- where d1', d2' are cloned versions of d1,d2, with the type substitution applied.
862 -- Note that the substitution is applied to the whole thing.
863 -- This is convenient, but just slightly fragile. Notably:
864 -- * There had better be no name clashes in a/b/c/d
867 -- poly_tyvars = [b,d] in the example above
868 -- spec_tyvars = [a,c]
869 -- ty_args = [t1,b,t3,d]
870 poly_tyvars = [tv | (tv, Nothing) <- rhs_tyvars `zip` call_ts]
871 spec_tyvars = [tv | (tv, Just _) <- rhs_tyvars `zip` call_ts]
872 ty_args = zipWithEqual "spec_call" mk_ty_arg rhs_tyvars call_ts
874 mk_ty_arg rhs_tyvar Nothing = Type (mkTyVarTy rhs_tyvar)
875 mk_ty_arg rhs_tyvar (Just ty) = Type ty
876 rhs_subst = extendTvSubstList subst (spec_tyvars `zip` [ty | Just ty <- call_ts])
878 cloneBinders rhs_subst rhs_dicts `thenSM` \ (rhs_subst', rhs_dicts') ->
880 inst_args = ty_args ++ map Var rhs_dicts'
882 -- Figure out the type of the specialised function
883 body_ty = applyTypeToArgs rhs fn_type inst_args
884 (lam_args, app_args) -- Add a dummy argument if body_ty is unlifted
885 | isUnLiftedType body_ty -- C.f. WwLib.mkWorkerArgs
886 = (poly_tyvars ++ [voidArgId], poly_tyvars ++ [realWorldPrimId])
887 | otherwise = (poly_tyvars, poly_tyvars)
888 spec_id_ty = mkPiTypes lam_args body_ty
890 newIdSM fn spec_id_ty `thenSM` \ spec_f ->
891 specExpr rhs_subst' (mkLams lam_args body) `thenSM` \ (spec_rhs, rhs_uds) ->
893 -- The rule to put in the function's specialisation is:
894 -- forall b,d, d1',d2'. f t1 b t3 d d1' d2' = f1 b d
895 spec_env_rule = Rule (mkFastString ("SPEC " ++ showSDoc (ppr fn)))
897 (poly_tyvars ++ rhs_dicts')
899 (mkVarApps (Var spec_f) app_args)
901 -- Add the { d1' = dx1; d2' = dx2 } usage stuff
902 final_uds = foldr addDictBind rhs_uds (my_zipEqual "spec_call" rhs_dicts' call_ds)
904 -- NOTE: we don't add back in any INLINE pragma on the RHS, so even if
905 -- the original function said INLINE, the specialised copies won't.
906 -- The idea is that the point of inlining was precisely to specialise
907 -- the function at its call site, and that's not so important for the
908 -- specialised copies. But it still smells like an ad hoc decision.
911 returnSM ((spec_f, spec_rhs),
916 my_zipEqual doc xs ys
917 | not (equalLength xs ys) = pprPanic "my_zipEqual" (ppr xs $$ ppr ys $$ (ppr fn <+> ppr call_ts) $$ ppr rhs)
918 | otherwise = zipEqual doc xs ys
920 dropInline :: CoreExpr -> CoreExpr
921 dropInline (Note InlineMe rhs) = rhs
925 %************************************************************************
927 \subsubsection{UsageDetails and suchlike}
929 %************************************************************************
934 dict_binds :: !(Bag DictBind),
935 -- Floated dictionary bindings
936 -- The order is important;
937 -- in ds1 `union` ds2, bindings in ds2 can depend on those in ds1
938 -- (Remember, Bags preserve order in GHC.)
940 calls :: !CallDetails
943 type DictBind = (CoreBind, VarSet)
944 -- The set is the free vars of the binding
945 -- both tyvars and dicts
947 type DictExpr = CoreExpr
949 emptyUDs = MkUD { dict_binds = emptyBag, calls = emptyFM }
951 type ProtoUsageDetails = ([DictBind],
952 [(Id, CallKey, ([DictExpr], VarSet))]
955 ------------------------------------------------------------
956 type CallDetails = FiniteMap Id CallInfo
957 newtype CallKey = CallKey [Maybe Type] -- Nothing => unconstrained type argument
958 type CallInfo = FiniteMap CallKey
959 ([DictExpr], VarSet) -- Dict args and the vars of the whole
960 -- call (including tyvars)
961 -- [*not* include the main id itself, of course]
962 -- The finite maps eliminate duplicates
963 -- The list of types and dictionaries is guaranteed to
964 -- match the type of f
966 -- Type isn't an instance of Ord, so that we can control which
967 -- instance we use. That's tiresome here. Oh well
968 instance Eq CallKey where
969 k1 == k2 = case k1 `compare` k2 of { EQ -> True; other -> False }
971 instance Ord CallKey where
972 compare (CallKey k1) (CallKey k2) = cmpList cmp k1 k2
974 cmp Nothing Nothing = EQ
975 cmp Nothing (Just t2) = LT
976 cmp (Just t1) Nothing = GT
977 cmp (Just t1) (Just t2) = tcCmpType t1 t2
979 unionCalls :: CallDetails -> CallDetails -> CallDetails
980 unionCalls c1 c2 = plusFM_C plusFM c1 c2
982 singleCall :: Id -> [Maybe Type] -> [DictExpr] -> CallDetails
983 singleCall id tys dicts
984 = unitFM id (unitFM (CallKey tys) (dicts, call_fvs))
986 call_fvs = exprsFreeVars dicts `unionVarSet` tys_fvs
987 tys_fvs = tyVarsOfTypes (catMaybes tys)
988 -- The type args (tys) are guaranteed to be part of the dictionary
989 -- types, because they are just the constrained types,
990 -- and the dictionary is therefore sure to be bound
991 -- inside the binding for any type variables free in the type;
992 -- hence it's safe to neglect tyvars free in tys when making
993 -- the free-var set for this call
994 -- BUT I don't trust this reasoning; play safe and include tys_fvs
996 -- We don't include the 'id' itself.
998 listToCallDetails calls
999 = foldr (unionCalls . mk_call) emptyFM calls
1001 mk_call (id, tys, dicts_w_fvs) = unitFM id (unitFM tys dicts_w_fvs)
1002 -- NB: the free vars of the call are provided
1004 callDetailsToList calls = [ (id,tys,dicts)
1005 | (id,fm) <- fmToList calls,
1006 (tys, dicts) <- fmToList fm
1009 mkCallUDs subst f args
1011 || not (all isClassPred theta)
1012 -- Only specialise if all overloading is on class params.
1013 -- In ptic, with implicit params, the type args
1014 -- *don't* say what the value of the implicit param is!
1015 || not (spec_tys `lengthIs` n_tyvars)
1016 || not ( dicts `lengthIs` n_dicts)
1017 || maybeToBool (lookupRule (\act -> True) (substInScope subst) f args)
1018 -- There's already a rule covering this call. A typical case
1019 -- is where there's an explicit user-provided rule. Then
1020 -- we don't want to create a specialised version
1021 -- of the function that overlaps.
1022 = emptyUDs -- Not overloaded, or no specialisation wanted
1025 = MkUD {dict_binds = emptyBag,
1026 calls = singleCall f spec_tys dicts
1029 (tyvars, theta, _) = tcSplitSigmaTy (idType f)
1030 constrained_tyvars = tyVarsOfTheta theta
1031 n_tyvars = length tyvars
1032 n_dicts = length theta
1034 spec_tys = [mk_spec_ty tv ty | (tv, Type ty) <- tyvars `zip` args]
1035 dicts = [dict_expr | (_, dict_expr) <- theta `zip` (drop n_tyvars args)]
1038 | tyvar `elemVarSet` constrained_tyvars = Just ty
1039 | otherwise = Nothing
1041 ------------------------------------------------------------
1042 plusUDs :: UsageDetails -> UsageDetails -> UsageDetails
1043 plusUDs (MkUD {dict_binds = db1, calls = calls1})
1044 (MkUD {dict_binds = db2, calls = calls2})
1045 = MkUD {dict_binds = d, calls = c}
1047 d = db1 `unionBags` db2
1048 c = calls1 `unionCalls` calls2
1050 plusUDList = foldr plusUDs emptyUDs
1052 -- zapCalls deletes calls to ids from uds
1053 zapCalls ids uds = uds {calls = delListFromFM (calls uds) ids}
1055 mkDB bind = (bind, bind_fvs bind)
1057 bind_fvs (NonRec bndr rhs) = exprFreeVars rhs
1058 bind_fvs (Rec prs) = foldl delVarSet rhs_fvs bndrs
1061 rhs_fvs = unionVarSets [exprFreeVars rhs | (bndr,rhs) <- prs]
1063 addDictBind (dict,rhs) uds = uds { dict_binds = mkDB (NonRec dict rhs) `consBag` dict_binds uds }
1065 dumpAllDictBinds (MkUD {dict_binds = dbs}) binds
1066 = foldrBag add binds dbs
1068 add (bind,_) binds = bind : binds
1070 dumpUDs :: [CoreBndr]
1071 -> UsageDetails -> CoreExpr
1072 -> (UsageDetails, CoreExpr)
1073 dumpUDs bndrs uds body
1074 = (free_uds, foldr add_let body dict_binds)
1076 (free_uds, (dict_binds, _)) = splitUDs bndrs uds
1077 add_let (bind,_) body = Let bind body
1079 splitUDs :: [CoreBndr]
1081 -> (UsageDetails, -- These don't mention the binders
1082 ProtoUsageDetails) -- These do
1084 splitUDs bndrs uds@(MkUD {dict_binds = orig_dbs,
1085 calls = orig_calls})
1087 = if isEmptyBag dump_dbs && null dump_calls then
1088 -- Common case: binder doesn't affect floats
1092 -- Binders bind some of the fvs of the floats
1093 (MkUD {dict_binds = free_dbs,
1094 calls = listToCallDetails free_calls},
1095 (bagToList dump_dbs, dump_calls)
1099 bndr_set = mkVarSet bndrs
1101 (free_dbs, dump_dbs, dump_idset)
1102 = foldlBag dump_db (emptyBag, emptyBag, bndr_set) orig_dbs
1103 -- Important that it's foldl not foldr;
1104 -- we're accumulating the set of dumped ids in dump_set
1106 -- Filter out any calls that mention things that are being dumped
1107 orig_call_list = callDetailsToList orig_calls
1108 (dump_calls, free_calls) = partition captured orig_call_list
1109 captured (id,tys,(dicts, fvs)) = fvs `intersectsVarSet` dump_idset
1110 || id `elemVarSet` dump_idset
1112 dump_db (free_dbs, dump_dbs, dump_idset) db@(bind, fvs)
1113 | dump_idset `intersectsVarSet` fvs -- Dump it
1114 = (free_dbs, dump_dbs `snocBag` db,
1115 extendVarSetList dump_idset (bindersOf bind))
1117 | otherwise -- Don't dump it
1118 = (free_dbs `snocBag` db, dump_dbs, dump_idset)
1122 %************************************************************************
1124 \subsubsection{Boring helper functions}
1126 %************************************************************************
1129 type SpecM a = UniqSM a
1133 getUniqSM = getUniqueUs
1137 mapAndCombineSM f [] = returnSM ([], emptyUDs)
1138 mapAndCombineSM f (x:xs) = f x `thenSM` \ (y, uds1) ->
1139 mapAndCombineSM f xs `thenSM` \ (ys, uds2) ->
1140 returnSM (y:ys, uds1 `plusUDs` uds2)
1142 cloneBindSM :: Subst -> CoreBind -> SpecM (Subst, Subst, CoreBind)
1143 -- Clone the binders of the bind; return new bind with the cloned binders
1144 -- Return the substitution to use for RHSs, and the one to use for the body
1145 cloneBindSM subst (NonRec bndr rhs)
1146 = getUs `thenUs` \ us ->
1148 (subst', bndr') = substAndCloneId subst us bndr
1150 returnUs (subst, subst', NonRec bndr' rhs)
1152 cloneBindSM subst (Rec pairs)
1153 = getUs `thenUs` \ us ->
1155 (subst', bndrs') = substAndCloneRecIds subst us (map fst pairs)
1157 returnUs (subst', subst', Rec (bndrs' `zip` map snd pairs))
1159 cloneBinders subst bndrs
1160 = getUs `thenUs` \ us ->
1161 returnUs (substAndCloneIds subst us bndrs)
1163 newIdSM old_id new_ty
1164 = getUniqSM `thenSM` \ uniq ->
1166 -- Give the new Id a similar occurrence name to the old one
1167 name = idName old_id
1168 new_id = mkUserLocal (mkSpecOcc (nameOccName name)) uniq new_ty (getSrcLoc name)
1174 Old (but interesting) stuff about unboxed bindings
1175 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1177 What should we do when a value is specialised to a *strict* unboxed value?
1179 map_*_* f (x:xs) = let h = f x
1183 Could convert let to case:
1185 map_*_Int# f (x:xs) = case f x of h# ->
1189 This may be undesirable since it forces evaluation here, but the value
1190 may not be used in all branches of the body. In the general case this
1191 transformation is impossible since the mutual recursion in a letrec
1192 cannot be expressed as a case.
1194 There is also a problem with top-level unboxed values, since our
1195 implementation cannot handle unboxed values at the top level.
1197 Solution: Lift the binding of the unboxed value and extract it when it
1200 map_*_Int# f (x:xs) = let h = case (f x) of h# -> _Lift h#
1205 Now give it to the simplifier and the _Lifting will be optimised away.
1207 The benfit is that we have given the specialised "unboxed" values a
1208 very simplep lifted semantics and then leave it up to the simplifier to
1209 optimise it --- knowing that the overheads will be removed in nearly
1212 In particular, the value will only be evaluted in the branches of the
1213 program which use it, rather than being forced at the point where the
1214 value is bound. For example:
1216 filtermap_*_* p f (x:xs)
1223 filtermap_*_Int# p f (x:xs)
1224 = let h = case (f x) of h# -> _Lift h#
1227 True -> case h of _Lift h#
1231 The binding for h can still be inlined in the one branch and the
1232 _Lifting eliminated.
1235 Question: When won't the _Lifting be eliminated?
1237 Answer: When they at the top-level (where it is necessary) or when
1238 inlining would duplicate work (or possibly code depending on
1239 options). However, the _Lifting will still be eliminated if the
1240 strictness analyser deems the lifted binding strict.