Simplify.lhs 93.8 KB
Newer Older
1
%
2
% (c) The AQUA Project, Glasgow University, 1993-1998
3 4 5 6
%
\section[Simplify]{The main module of the simplifier}

\begin{code}
7
module Simplify ( simplTopBinds, simplExpr ) where
8

9
#include "HsVersions.h"
10

simonpj@microsoft.com's avatar
Wibble  
simonpj@microsoft.com committed
11
import DynFlags
12
import SimplMonad
13
import Type hiding      ( substTy, extendTvSubst, substTyVar )
Ian Lynagh's avatar
Ian Lynagh committed
14
import SimplEnv
15
import SimplUtils
16
import FamInstEnv	( FamInstEnv )
17
import Id
18 19
import MkId		( seqId, realWorldPrimId )
import MkCore		( mkImpossibleExpr )
20
import Var
21
import IdInfo
22
import Name		( mkSystemVarName, isExternalName )
23
import Coercion
24
import OptCoercion	( optCoercion )
Ian Lynagh's avatar
Ian Lynagh committed
25
import FamInstEnv       ( topNormaliseType )
26
import DataCon          ( DataCon, dataConWorkId, dataConRepStrictness )
27
import CoreMonad	( SimplifierSwitch(..), Tick(..) )
28
import CoreSyn
29
import Demand           ( isStrictDmd, splitStrictSig )
Ian Lynagh's avatar
Ian Lynagh committed
30
import PprCore          ( pprParendExpr, pprCoreExpr )
31 32 33
import CoreUnfold       ( mkUnfolding, mkCoreUnfolding
                        , mkInlineUnfolding, mkSimpleUnfolding
                        , exprIsConApp_maybe, callSiteInline, CallCtxt(..) )
34
import CoreUtils
35
import qualified CoreSubst
36
import CoreArity	( exprArity )
37
import Rules            ( lookupRule, getRules )
38
import BasicTypes       ( isMarkedStrict, Arity )
39
import CostCentre       ( currentCCS, pushCCisNop )
Ian Lynagh's avatar
Ian Lynagh committed
40
import TysPrim          ( realWorldStatePrimTy )
41
import BasicTypes       ( TopLevelFlag(..), isTopLevel, RecFlag(..) )
42
import MonadUtils	( foldlM, mapAccumLM )
Ian Lynagh's avatar
Ian Lynagh committed
43 44
import Maybes           ( orElse )
import Data.List        ( mapAccumL )
45
import Outputable
46
import FastString
47 48 49
\end{code}


50 51
The guts of the simplifier is in this module, but the driver loop for
the simplifier is in SimplCore.lhs.
52 53


54
-----------------------------------------
Ian Lynagh's avatar
Ian Lynagh committed
55
        *** IMPORTANT NOTE ***
56 57 58 59 60 61
-----------------------------------------
The simplifier used to guarantee that the output had no shadowing, but
it does not do so any more.   (Actually, it never did!)  The reason is
documented with simplifyArgs.


62
-----------------------------------------
Ian Lynagh's avatar
Ian Lynagh committed
63
        *** IMPORTANT NOTE ***
64 65 66 67 68 69 70 71 72 73
-----------------------------------------
Many parts of the simplifier return a bunch of "floats" as well as an
expression. This is wrapped as a datatype SimplUtils.FloatsWith.

All "floats" are let-binds, not case-binds, but some non-rec lets may
be unlifted (with RHS ok-for-speculation).



-----------------------------------------
Ian Lynagh's avatar
Ian Lynagh committed
74
        ORGANISATION OF FUNCTIONS
75 76 77 78 79 80
-----------------------------------------
simplTopBinds
  - simplify all top-level binders
  - for NonRec, call simplRecOrTopPair
  - for Rec,    call simplRecBind

Ian Lynagh's avatar
Ian Lynagh committed
81 82 83

        ------------------------------
simplExpr (applied lambda)      ==> simplNonRecBind
84 85 86
simplExpr (Let (NonRec ...) ..) ==> simplNonRecBind
simplExpr (Let (Rec ...)    ..) ==> simplify binders; simplRecBind

Ian Lynagh's avatar
Ian Lynagh committed
87 88
        ------------------------------
simplRecBind    [binders already simplfied]
89 90 91 92
  - use simplRecOrTopPair on each pair in turn

simplRecOrTopPair [binder already simplified]
  Used for: recursive bindings (top level and nested)
Ian Lynagh's avatar
Ian Lynagh committed
93 94
            top-level non-recursive bindings
  Returns:
95 96 97 98 99
  - check for PreInlineUnconditionally
  - simplLazyBind

simplNonRecBind
  Used for: non-top-level non-recursive bindings
Ian Lynagh's avatar
Ian Lynagh committed
100 101 102
            beta reductions (which amount to the same thing)
  Because it can deal with strict arts, it takes a
        "thing-inside" and returns an expression
103 104 105 106

  - check for PreInlineUnconditionally
  - simplify binder, including its IdInfo
  - if strict binding
Ian Lynagh's avatar
Ian Lynagh committed
107 108 109
        simplStrictArg
        mkAtomicArgs
        completeNonRecX
110
    else
Ian Lynagh's avatar
Ian Lynagh committed
111 112
        simplLazyBind
        addFloats
113

Ian Lynagh's avatar
Ian Lynagh committed
114
simplNonRecX:   [given a *simplified* RHS, but an *unsimplified* binder]
115 116 117 118
  Used for: binding case-binder and constr args in a known-constructor case
  - check for PreInLineUnconditionally
  - simplify binder
  - completeNonRecX
Ian Lynagh's avatar
Ian Lynagh committed
119 120 121

        ------------------------------
simplLazyBind:  [binder already simplified, RHS not]
122
  Used for: recursive bindings (top level and nested)
Ian Lynagh's avatar
Ian Lynagh committed
123 124 125
            top-level non-recursive bindings
            non-top-level, but *lazy* non-recursive bindings
        [must not be strict or unboxed]
126
  Returns floats + an augmented environment, not an expression
Ian Lynagh's avatar
Ian Lynagh committed
127 128
  - substituteIdInfo and add result to in-scope
        [so that rules are available in rec rhs]
129 130 131
  - simplify rhs
  - mkAtomicArgs
  - float if exposes constructor or PAP
132
  - completeBind
133 134


Ian Lynagh's avatar
Ian Lynagh committed
135
completeNonRecX:        [binder and rhs both simplified]
136
  - if the the thing needs case binding (unlifted and not ok-for-spec)
Ian Lynagh's avatar
Ian Lynagh committed
137
        build a Case
138
   else
Ian Lynagh's avatar
Ian Lynagh committed
139 140
        completeBind
        addFloats
141

Ian Lynagh's avatar
Ian Lynagh committed
142 143
completeBind:   [given a simplified RHS]
        [used for both rec and non-rec bindings, top level and not]
144 145 146 147 148 149 150 151
  - try PostInlineUnconditionally
  - add unfolding [this is the only place we add an unfolding]
  - add arity



Right hand sides and arguments
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Ian Lynagh's avatar
Ian Lynagh committed
152 153 154
In many ways we want to treat
        (a) the right hand side of a let(rec), and
        (b) a function argument
155 156 157
in the same way.  But not always!  In particular, we would
like to leave these arguments exactly as they are, so they
will match a RULE more easily.
Ian Lynagh's avatar
Ian Lynagh committed
158 159 160

        f (g x, h x)
        g (+ x)
161 162 163 164

It's harder to make the rule match if we ANF-ise the constructor,
or eta-expand the PAP:

Ian Lynagh's avatar
Ian Lynagh committed
165 166
        f (let { a = g x; b = h x } in (a,b))
        g (\y. + x y)
167 168 169

On the other hand if we see the let-defns

Ian Lynagh's avatar
Ian Lynagh committed
170 171
        p = (g x, h x)
        q = + x
172 173

then we *do* want to ANF-ise and eta-expand, so that p and q
Ian Lynagh's avatar
Ian Lynagh committed
174
can be safely inlined.
175 176 177 178 179

Even floating lets out is a bit dubious.  For let RHS's we float lets
out if that exposes a value, so that the value can be inlined more vigorously.
For example

Ian Lynagh's avatar
Ian Lynagh committed
180
        r = let x = e in (x,x)
181 182 183 184 185 186 187 188 189 190 191 192 193 194

Here, if we float the let out we'll expose a nice constructor. We did experiments
that showed this to be a generally good thing.  But it was a bad thing to float
lets out unconditionally, because that meant they got allocated more often.

For function arguments, there's less reason to expose a constructor (it won't
get inlined).  Just possibly it might make a rule match, but I'm pretty skeptical.
So for the moment we don't float lets out of function arguments either.


Eta expansion
~~~~~~~~~~~~~~
For eta expansion, we want to catch things like

Ian Lynagh's avatar
Ian Lynagh committed
195
        case e of (a,b) -> \x -> case a of (p,q) -> \y -> r
196 197 198 199 200

If the \x was on the RHS of a let, we'd eta expand to bring the two
lambdas together.  And in general that's a good thing to do.  Perhaps
we should eta expand wherever we find a (value) lambda?  Then the eta
expansion at a let RHS can concentrate solely on the PAP case.
201 202


203
%************************************************************************
Ian Lynagh's avatar
Ian Lynagh committed
204
%*                                                                      *
205
\subsection{Bindings}
Ian Lynagh's avatar
Ian Lynagh committed
206
%*                                                                      *
207 208 209
%************************************************************************

\begin{code}
210
simplTopBinds :: SimplEnv -> [InBind] -> SimplM SimplEnv
211

Ian Lynagh's avatar
Ian Lynagh committed
212
simplTopBinds env0 binds0
Ian Lynagh's avatar
Ian Lynagh committed
213 214 215 216
  = do  {       -- Put all the top-level binders into scope at the start
                -- so that if a transformation rule has unexpectedly brought
                -- anything into scope, then we don't get a complaint about that.
                -- It's rather as if the top-level binders were imported.
Ian Lynagh's avatar
Ian Lynagh committed
217
        ; env1 <- simplRecBndrs env0 (bindersOfBinds binds0)
Ian Lynagh's avatar
Ian Lynagh committed
218
        ; dflags <- getDOptsSmpl
219
        ; let dump_flag = dopt Opt_D_verbose_core2core dflags
Ian Lynagh's avatar
Ian Lynagh committed
220
        ; env2 <- simpl_binds dump_flag env1 binds0
Ian Lynagh's avatar
Ian Lynagh committed
221
        ; freeTick SimplifierDone
222
        ; return env2 }
223
  where
Ian Lynagh's avatar
Ian Lynagh committed
224 225 226 227 228 229
        -- We need to track the zapped top-level binders, because
        -- they should have their fragile IdInfo zapped (notably occurrence info)
        -- That's why we run down binds and bndrs' simultaneously.
        --
        -- The dump-flag emits a trace for each top-level binding, which
        -- helps to locate the tracing for inlining and rule firing
230
    simpl_binds :: Bool -> SimplEnv -> [InBind] -> SimplM SimplEnv
Ian Lynagh's avatar
Ian Lynagh committed
231 232
    simpl_binds _    env []           = return env
    simpl_binds dump env (bind:binds) = do { env' <- trace_bind dump bind $
Ian Lynagh's avatar
Ian Lynagh committed
233 234
                                                     simpl_bind env bind
                                           ; simpl_binds dump env' binds }
235

Ian Lynagh's avatar
Ian Lynagh committed
236 237
    trace_bind True  bind = pprTrace "SimplBind" (ppr (bindersOf bind))
    trace_bind False _    = \x -> x
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
238

239 240
    simpl_bind env (Rec pairs)  = simplRecBind      env  TopLevel pairs
    simpl_bind env (NonRec b r) = simplRecOrTopPair env' TopLevel b b' r
Ian Lynagh's avatar
Ian Lynagh committed
241 242
        where
          (env', b') = addBndrRules env b (lookupRecBndr env b)
243 244 245 246
\end{code}


%************************************************************************
Ian Lynagh's avatar
Ian Lynagh committed
247
%*                                                                      *
248
\subsection{Lazy bindings}
Ian Lynagh's avatar
Ian Lynagh committed
249
%*                                                                      *
250 251 252
%************************************************************************

simplRecBind is used for
Ian Lynagh's avatar
Ian Lynagh committed
253
        * recursive bindings only
254 255 256

\begin{code}
simplRecBind :: SimplEnv -> TopLevelFlag
Ian Lynagh's avatar
Ian Lynagh committed
257 258
             -> [(InId, InExpr)]
             -> SimplM SimplEnv
Ian Lynagh's avatar
Ian Lynagh committed
259 260 261 262 263
simplRecBind env0 top_lvl pairs0
  = do  { let (env_with_info, triples) = mapAccumL add_rules env0 pairs0
        ; env1 <- go (zapFloats env_with_info) triples
        ; return (env0 `addRecFloats` env1) }
        -- addFloats adds the floats from env1,
Thomas Schilling's avatar
Thomas Schilling committed
264
        -- _and_ updates env0 with the in-scope set from env1
265
  where
266
    add_rules :: SimplEnv -> (InBndr,InExpr) -> (SimplEnv, (InBndr, OutBndr, InExpr))
Ian Lynagh's avatar
Ian Lynagh committed
267
        -- Add the (substituted) rules to the binder
268
    add_rules env (bndr, rhs) = (env', (bndr, bndr', rhs))
Ian Lynagh's avatar
Ian Lynagh committed
269 270
        where
          (env', bndr') = addBndrRules env bndr (lookupRecBndr env bndr)
271

272
    go env [] = return env
Ian Lynagh's avatar
Ian Lynagh committed
273

274
    go env ((old_bndr, new_bndr, rhs) : pairs)
Ian Lynagh's avatar
Ian Lynagh committed
275 276
        = do { env' <- simplRecOrTopPair env top_lvl old_bndr new_bndr rhs
             ; go env' pairs }
277 278
\end{code}

279
simplOrTopPair is used for
Ian Lynagh's avatar
Ian Lynagh committed
280 281
        * recursive bindings (whether top level or not)
        * top-level non-recursive bindings
282 283 284 285 286

It assumes the binder has already been simplified, but not its IdInfo.

\begin{code}
simplRecOrTopPair :: SimplEnv
Ian Lynagh's avatar
Ian Lynagh committed
287 288 289
                  -> TopLevelFlag
                  -> InId -> OutBndr -> InExpr  -- Binder and rhs
                  -> SimplM SimplEnv    -- Returns an env that includes the binding
290

291
simplRecOrTopPair env top_lvl old_bndr new_bndr rhs
Ian Lynagh's avatar
Ian Lynagh committed
292 293 294
  | preInlineUnconditionally env top_lvl old_bndr rhs   -- Check for unconditional inline
  = do  { tick (PreInlineUnconditionally old_bndr)
        ; return (extendIdSubst env old_bndr (mkContEx env rhs)) }
295 296

  | otherwise
Ian Lynagh's avatar
Ian Lynagh committed
297 298
  = simplLazyBind env top_lvl Recursive old_bndr new_bndr rhs env
        -- May not actually be recursive, but it doesn't matter
299 300 301 302
\end{code}


simplLazyBind is used for
303 304
  * [simplRecOrTopPair] recursive bindings (whether top level or not)
  * [simplRecOrTopPair] top-level non-recursive bindings
Ian Lynagh's avatar
Ian Lynagh committed
305
  * [simplNonRecE]      non-top-level *lazy* non-recursive bindings
306 307

Nota bene:
Ian Lynagh's avatar
Ian Lynagh committed
308
    1. It assumes that the binder is *already* simplified,
309
       and is in scope, and its IdInfo too, except unfolding
310 311 312 313 314 315 316 317

    2. It assumes that the binder type is lifted.

    3. It does not check for pre-inline-unconditionallly;
       that should have been done already.

\begin{code}
simplLazyBind :: SimplEnv
Ian Lynagh's avatar
Ian Lynagh committed
318 319 320 321 322
              -> TopLevelFlag -> RecFlag
              -> InId -> OutId          -- Binder, both pre-and post simpl
                                        -- The OutId has IdInfo, except arity, unfolding
              -> InExpr -> SimplEnv     -- The RHS and its environment
              -> SimplM SimplEnv
323

324
simplLazyBind env top_lvl is_rec bndr bndr1 rhs rhs_se
325 326
  = -- pprTrace "simplLazyBind" ((ppr bndr <+> ppr bndr1) $$ ppr rhs $$ ppr (seIdSubst rhs_se)) $
    do  { let   rhs_env     = rhs_se `setInScope` env
327 328 329 330 331 332 333 334 335 336
		(tvs, body) = case collectTyBinders rhs of
			        (tvs, body) | not_lam body -> (tvs,body)
					    | otherwise	   -> ([], rhs)
		not_lam (Lam _ _) = False
		not_lam _	  = True
			-- Do not do the "abstract tyyvar" thing if there's
			-- a lambda inside, becuase it defeats eta-reduction
			--    f = /\a. \x. g a x  
			-- should eta-reduce

Ian Lynagh's avatar
Ian Lynagh committed
337
        ; (body_env, tvs') <- simplBinders rhs_env tvs
338
                -- See Note [Floating and type abstraction] in SimplUtils
Ian Lynagh's avatar
Ian Lynagh committed
339

340
        -- Simplify the RHS
341
        ; (body_env1, body1) <- simplExprF body_env body mkRhsStop
Ian Lynagh's avatar
Ian Lynagh committed
342
        -- ANF-ise a constructor or PAP rhs
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
343
        ; (body_env2, body2) <- prepareRhs top_lvl body_env1 bndr1 body1
Ian Lynagh's avatar
Ian Lynagh committed
344 345 346

        ; (env', rhs')
            <-  if not (doFloatFromRhs top_lvl is_rec False body2 body_env2)
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
347 348
                then                            -- No floating, revert to body1
                     do { rhs' <- mkLam env tvs' (wrapFloats body_env1 body1)
Ian Lynagh's avatar
Ian Lynagh committed
349 350 351 352 353 354 355 356 357
                        ; return (env, rhs') }

                else if null tvs then           -- Simple floating
                     do { tick LetFloatFromLet
                        ; return (addFloats env body_env2, body2) }

                else                            -- Do type-abstraction first
                     do { tick LetFloatFromLet
                        ; (poly_binds, body3) <- abstractFloats tvs' body_env2 body2
358
                        ; rhs' <- mkLam env tvs' body3
359
                        ; env' <- foldlM (addPolyBind top_lvl) env poly_binds
360
                        ; return (env', rhs') }
Ian Lynagh's avatar
Ian Lynagh committed
361 362

        ; completeBind env' top_lvl bndr bndr1 rhs' }
363
\end{code}
364

Ian Lynagh's avatar
Ian Lynagh committed
365
A specialised variant of simplNonRec used when the RHS is already simplified,
366 367 368 369
notably in knownCon.  It uses case-binding where necessary.

\begin{code}
simplNonRecX :: SimplEnv
Ian Lynagh's avatar
Ian Lynagh committed
370 371 372
             -> InId            -- Old binder
             -> OutExpr         -- Simplified RHS
             -> SimplM SimplEnv
373 374

simplNonRecX env bndr new_rhs
375 376 377
  | isDeadBinder bndr	-- Not uncommon; e.g. case (a,b) of b { (p,q) -> p }
  = return env		-- 		 Here b is dead, and we avoid creating
  | otherwise		--		 the binding b = (a,b)
Ian Lynagh's avatar
Ian Lynagh committed
378
  = do  { (env', bndr') <- simplBinder env bndr
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
379 380
        ; completeNonRecX NotTopLevel env' (isStrictId bndr) bndr bndr' new_rhs }
		-- simplNonRecX is only used for NotTopLevel things
381

simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
382
completeNonRecX :: TopLevelFlag -> SimplEnv
383
                -> Bool
Ian Lynagh's avatar
Ian Lynagh committed
384 385 386 387
                -> InId                 -- Old binder
                -> OutId                -- New binder
                -> OutExpr              -- Simplified RHS
                -> SimplM SimplEnv
388

simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
389 390
completeNonRecX top_lvl env is_strict old_bndr new_bndr new_rhs
  = do  { (env1, rhs1) <- prepareRhs top_lvl (zapFloats env) new_bndr new_rhs
391
        ; (env2, rhs2) <- 
392
                if doFloatFromRhs NotTopLevel NonRecursive is_strict rhs1 env1
Ian Lynagh's avatar
Ian Lynagh committed
393 394 395 396
                then do { tick LetFloatFromLet
                        ; return (addFloats env env1, rhs1) }   -- Add the floats to the main env
                else return (env, wrapFloats env1 rhs1)         -- Wrap the floats around the RHS
        ; completeBind env2 NotTopLevel old_bndr new_bndr rhs2 }
397 398 399 400
\end{code}

{- No, no, no!  Do not try preInlineUnconditionally in completeNonRecX
   Doing so risks exponential behaviour, because new_rhs has been simplified once already
Ian Lynagh's avatar
Ian Lynagh committed
401
   In the cases described by the folowing commment, postInlineUnconditionally will
402
   catch many of the relevant cases.
Ian Lynagh's avatar
Ian Lynagh committed
403 404 405 406 407 408 409 410
        -- This happens; for example, the case_bndr during case of
        -- known constructor:  case (a,b) of x { (p,q) -> ... }
        -- Here x isn't mentioned in the RHS, so we don't want to
        -- create the (dead) let-binding  let x = (a,b) in ...
        --
        -- Similarly, single occurrences can be inlined vigourously
        -- e.g.  case (f x, g y) of (a,b) -> ....
        -- If a,b occur once we can avoid constructing the let binding for them.
411

412
   Furthermore in the case-binding case preInlineUnconditionally risks extra thunks
Ian Lynagh's avatar
Ian Lynagh committed
413 414 415 416 417 418
        -- Consider     case I# (quotInt# x y) of
        --                I# v -> let w = J# v in ...
        -- If we gaily inline (quotInt# x y) for v, we end up building an
        -- extra thunk:
        --                let w = J# (quotInt# x y) in ...
        -- because quotInt# can fail.
419

420 421 422 423
  | preInlineUnconditionally env NotTopLevel bndr new_rhs
  = thing_inside (extendIdSubst env bndr (DoneEx new_rhs))
-}

424
----------------------------------
425
prepareRhs takes a putative RHS, checks whether it's a PAP or
Ian Lynagh's avatar
Ian Lynagh committed
426
constructor application and, if so, converts it to ANF, so that the
427
resulting thing can be inlined more easily.  Thus
Ian Lynagh's avatar
Ian Lynagh committed
428
        x = (f a, g b)
429
becomes
Ian Lynagh's avatar
Ian Lynagh committed
430 431 432
        t1 = f a
        t2 = g b
        x = (t1,t2)
433

434
We also want to deal well cases like this
Ian Lynagh's avatar
Ian Lynagh committed
435
        v = (f e1 `cast` co) e2
436
Here we want to make e1,e2 trivial and get
Ian Lynagh's avatar
Ian Lynagh committed
437
        x1 = e1; x2 = e2; v = (f x1 `cast` co) v2
438 439
That's what the 'go' loop in prepareRhs does

440
\begin{code}
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
441
prepareRhs :: TopLevelFlag -> SimplEnv -> OutId -> OutExpr -> SimplM (SimplEnv, OutExpr)
442
-- Adds new floats to the env iff that allows us to return a good RHS
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
443
prepareRhs top_lvl env id (Cast rhs co)    -- Note [Float coercions]
Ian Lynagh's avatar
Ian Lynagh committed
444
  | (ty1, _ty2) <- coercionKind co       -- Do *not* do this if rhs has an unlifted type
Ian Lynagh's avatar
Ian Lynagh committed
445
  , not (isUnLiftedType ty1)            -- see Note [Float coercions (unlifted)]
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
446
  = do  { (env', rhs') <- makeTrivialWithInfo top_lvl env sanitised_info rhs
Ian Lynagh's avatar
Ian Lynagh committed
447
        ; return (env', Cast rhs' co) }
448
  where
449 450
    sanitised_info = vanillaIdInfo `setStrictnessInfo` strictnessInfo info
                                   `setDemandInfo`     demandInfo info
451
    info = idInfo id
452

simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
453
prepareRhs top_lvl env0 _ rhs0
454
  = do  { (_is_exp, env1, rhs1) <- go 0 env0 rhs0
Ian Lynagh's avatar
Ian Lynagh committed
455
        ; return (env1, rhs1) }
456
  where
457
    go n_val_args env (Cast rhs co)
458 459
        = do { (is_exp, env', rhs') <- go n_val_args env rhs
             ; return (is_exp, env', Cast rhs' co) }
460
    go n_val_args env (App fun (Type ty))
461 462
        = do { (is_exp, env', rhs') <- go n_val_args env fun
             ; return (is_exp, env', App rhs' (Type ty)) }
463
    go n_val_args env (App fun arg)
464 465
        = do { (is_exp, env', fun') <- go (n_val_args+1) env fun
             ; case is_exp of
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
466
                True -> do { (env'', arg') <- makeTrivial top_lvl env' arg
Ian Lynagh's avatar
Ian Lynagh committed
467 468
                           ; return (True, env'', App fun' arg') }
                False -> return (False, env, App fun arg) }
469
    go n_val_args env (Var fun)
470
        = return (is_exp, env, Var fun)
Ian Lynagh's avatar
Ian Lynagh committed
471
        where
472 473 474 475 476
          is_exp = isExpandableApp fun n_val_args   -- The fun a constructor or PAP
		        -- See Note [CONLIKE pragma] in BasicTypes
			-- The definition of is_exp should match that in
	                -- OccurAnal.occAnalApp

Ian Lynagh's avatar
Ian Lynagh committed
477
    go _ env other
Ian Lynagh's avatar
Ian Lynagh committed
478
        = return (False, env, other)
479 480
\end{code}

481

482 483 484
Note [Float coercions]
~~~~~~~~~~~~~~~~~~~~~~
When we find the binding
Ian Lynagh's avatar
Ian Lynagh committed
485
        x = e `cast` co
486
we'd like to transform it to
Ian Lynagh's avatar
Ian Lynagh committed
487 488
        x' = e
        x = x `cast` co         -- A trivial binding
489 490 491 492 493 494 495 496 497 498 499 500 501
There's a chance that e will be a constructor application or function, or something
like that, so moving the coerion to the usage site may well cancel the coersions
and lead to further optimisation.  Example:

     data family T a :: *
     data instance T Int = T Int

     foo :: Int -> Int -> Int
     foo m n = ...
        where
          x = T m
          go 0 = 0
          go n = case x of { T m -> go (n-m) }
Ian Lynagh's avatar
Ian Lynagh committed
502
                -- This case should optimise
503

504 505 506 507 508 509 510 511 512 513 514
Note [Preserve strictness when floating coercions]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In the Note [Float coercions] transformation, keep the strictness info.
Eg
	f = e `cast` co	   -- f has strictness SSL
When we transform to
        f' = e		   -- f' also has strictness SSL
        f = f' `cast` co   -- f still has strictness SSL

Its not wrong to drop it on the floor, but better to keep it.

515 516
Note [Float coercions (unlifted)]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Ian Lynagh's avatar
Ian Lynagh committed
517
BUT don't do [Float coercions] if 'e' has an unlifted type.
518 519
This *can* happen:

Ian Lynagh's avatar
Ian Lynagh committed
520 521
     foo :: Int = (error (# Int,Int #) "urk")
                  `cast` CoUnsafe (# Int,Int #) Int
522 523 524

If do the makeTrivial thing to the error call, we'll get
    foo = case error (# Int,Int #) "urk" of v -> v `cast` ...
Ian Lynagh's avatar
Ian Lynagh committed
525
But 'v' isn't in scope!
526 527

These strange casts can happen as a result of case-of-case
Ian Lynagh's avatar
Ian Lynagh committed
528 529
        bar = case (case x of { T -> (# 2,3 #); F -> error "urk" }) of
                (# p,q #) -> p+q
530

531 532

\begin{code}
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
533
makeTrivial :: TopLevelFlag -> SimplEnv -> OutExpr -> SimplM (SimplEnv, OutExpr)
534
-- Binds the expression to a variable, if it's not trivial, returning the variable
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
535
makeTrivial top_lvl env expr = makeTrivialWithInfo top_lvl env vanillaIdInfo expr
536

simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
537 538
makeTrivialWithInfo :: TopLevelFlag -> SimplEnv -> IdInfo 
                    -> OutExpr -> SimplM (SimplEnv, OutExpr)
539 540
-- Propagate strictness and demand info to the new binder
-- Note [Preserve strictness when floating coercions]
541
-- Returned SimplEnv has same substitution as incoming one
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
542 543 544 545
makeTrivialWithInfo top_lvl env info expr
  | exprIsTrivial expr 				-- Already trivial
  || not (bindingOk top_lvl expr expr_ty)	-- Cannot trivialise
						--   See Note [Cannot trivialise]
546
  = return (env, expr)
Ian Lynagh's avatar
Ian Lynagh committed
547
  | otherwise           -- See Note [Take care] below
548 549
  = do  { uniq <- getUniqueM
        ; let name = mkSystemVarName uniq (fsLit "a")
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
550
              var = mkLocalIdWithInfo name expr_ty info
551
        ; env'  <- completeNonRecX top_lvl env False var var expr
552 553 554
	; expr' <- simplVar env' var
        ; return (env', expr') }
	-- The simplVar is needed becase we're constructing a new binding
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
555 556 557 558 559 560
	--     a = rhs
	-- And if rhs is of form (rhs1 |> co), then we might get
	--     a1 = rhs1
	--     a = a1 |> co
	-- and now a's RHS is trivial and can be substituted out, and that
	-- is what completeNonRecX will do
561 562
	-- To put it another way, it's as if we'd simplified
	--    let var = e in var
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
563 564 565 566
  where
    expr_ty = exprType expr

bindingOk :: TopLevelFlag -> CoreExpr -> Type -> Bool
567
-- True iff we can have a binding of this expression at this level
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
568 569 570 571
-- Precondition: the type is the type of the expression
bindingOk top_lvl _ expr_ty
  | isTopLevel top_lvl = not (isUnLiftedType expr_ty) 
  | otherwise          = True
572
\end{code}
573

simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594
Note [Cannot trivialise]
~~~~~~~~~~~~~~~~~~~~~~~~
Consider tih
   f :: Int -> Addr#
   
   foo :: Bar
   foo = Bar (f 3)

Then we can't ANF-ise foo, even though we'd like to, because
we can't make a top-level binding for the Addr# (f 3). And if
so we don't want to turn it into
   foo = let x = f 3 in Bar x
because we'll just end up inlining x back, and that makes the
simplifier loop.  Better not to ANF-ise it at all.

A case in point is literal strings (a MachStr is not regarded as
trivial):

   foo = Ptr "blob"#

We don't want to ANF-ise this.
595

596
%************************************************************************
Ian Lynagh's avatar
Ian Lynagh committed
597
%*                                                                      *
598
\subsection{Completing a lazy binding}
Ian Lynagh's avatar
Ian Lynagh committed
599
%*                                                                      *
600 601
%************************************************************************

602 603 604 605 606
completeBind
  * deals only with Ids, not TyVars
  * takes an already-simplified binder and RHS
  * is used for both recursive and non-recursive bindings
  * is used for both top-level and non-top-level bindings
607 608 609 610 611 612 613 614

It does the following:
  - tries discarding a dead binding
  - tries PostInlineUnconditionally
  - add unfolding [this is the only place we add an unfolding]
  - add arity

It does *not* attempt to do let-to-case.  Why?  Because it is used for
Ian Lynagh's avatar
Ian Lynagh committed
615
  - top-level bindings (when let-to-case is impossible)
616
  - many situations where the "rhs" is known to be a WHNF
Ian Lynagh's avatar
Ian Lynagh committed
617
                (so let-to-case is inappropriate).
618

619 620
Nor does it do the atomic-argument thing

621
\begin{code}
622
completeBind :: SimplEnv
Ian Lynagh's avatar
Ian Lynagh committed
623 624 625 626 627 628 629
             -> TopLevelFlag            -- Flag stuck into unfolding
             -> InId                    -- Old binder
             -> OutId -> OutExpr        -- New binder and RHS
             -> SimplM SimplEnv
-- completeBind may choose to do its work
--      * by extending the substitution (e.g. let x = y in ...)
--      * or by adding to the floats in the envt
630 631

completeBind env top_lvl old_bndr new_bndr new_rhs
632 633 634
  = do	{ let old_info = idInfo old_bndr
	      old_unf  = unfoldingInfo old_info
	      occ_info = occInfo old_info
635

636 637 638 639 640
	; new_unfolding <- simplUnfolding env top_lvl old_bndr occ_info new_rhs old_unf

	; if postInlineUnconditionally env top_lvl new_bndr occ_info new_rhs new_unfolding
	                -- Inline and discard the binding
	  then do  { tick (PostInlineUnconditionally old_bndr)
641 642
	           ; -- pprTrace "postInlineUnconditionally" (ppr old_bndr <+> equals <+> ppr new_rhs) $
                     return (extendIdSubst env old_bndr (DoneEx new_rhs)) }
643 644 645 646 647 648 649
	        -- Use the substitution to make quite, quite sure that the
	        -- substitution will happen, since we are going to discard the binding

	  else return (addNonRecWithUnf env new_bndr new_rhs new_unfolding) }

------------------------------
addPolyBind :: TopLevelFlag -> SimplEnv -> OutBind -> SimplM SimplEnv
650 651 652 653 654 655 656 657 658 659 660 661
-- Add a new binding to the environment, complete with its unfolding
-- but *do not* do postInlineUnconditionally, because we have already
-- processed some of the scope of the binding
-- We still want the unfolding though.  Consider
--	let 
--	      x = /\a. let y = ... in Just y
--	in body
-- Then we float the y-binding out (via abstractFloats and addPolyBind)
-- but 'x' may well then be inlined in 'body' in which case we'd like the 
-- opportunity to inline 'y' too.

addPolyBind top_lvl env (NonRec poly_id rhs)
662 663 664 665
  = do	{ unfolding <- simplUnfolding env top_lvl poly_id NoOccInfo rhs noUnfolding
    	  		-- Assumes that poly_id did not have an INLINE prag
			-- which is perhaps wrong.  ToDo: think about this
        ; return (addNonRecWithUnf env poly_id rhs unfolding) }
666

667
addPolyBind _ env bind@(Rec _) = return (extendFloats env bind)
668 669 670 671
		-- Hack: letrecs are more awkward, so we extend "by steam"
		-- without adding unfoldings etc.  At worst this leads to
		-- more simplifier iterations

672
------------------------------
673
addNonRecWithUnf :: SimplEnv
674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689
             	 -> OutId -> OutExpr    -- New binder and RHS
	     	 -> Unfolding		-- New unfolding
             	 -> SimplEnv
addNonRecWithUnf env new_bndr new_rhs new_unfolding
  = let new_arity = exprArity new_rhs
	old_arity = idArity new_bndr
        info1 = idInfo new_bndr `setArityInfo` new_arity
	
              -- Unfolding info: Note [Setting the new unfolding]
	info2 = info1 `setUnfoldingInfo` new_unfolding

        -- Demand info: Note [Setting the demand info]
        info3 | isEvaldUnfolding new_unfolding = zapDemandInfo info2 `orElse` info2
              | otherwise                      = info2

        final_id = new_bndr `setIdInfo` info3
690
	dmd_arity = length $ fst $ splitStrictSig $ idStrictness new_bndr
691 692
    in
    ASSERT( isId new_bndr )
693
    WARN( new_arity < old_arity || new_arity < dmd_arity, 
694 695
          (ptext (sLit "Arity decrease:") <+> (ppr final_id <+> ppr old_arity
		<+> ppr new_arity <+> ppr dmd_arity) $$ ppr new_rhs) )
696
	-- Note [Arity decrease]
Simon Marlow's avatar
Simon Marlow committed
697

698 699 700 701 702 703 704 705
    final_id `seq`   -- This seq forces the Id, and hence its IdInfo,
	             -- and hence any inner substitutions
	    -- pprTrace "Binding" (ppr final_id <+> ppr unfolding) $
    addNonRec env final_id new_rhs
		-- The addNonRec adds it to the in-scope set too

------------------------------
simplUnfolding :: SimplEnv-> TopLevelFlag
706
	       -> Id
707 708 709
	       -> OccInfo -> OutExpr
	       -> Unfolding -> SimplM Unfolding
-- Note [Setting the new unfolding]
710 711
simplUnfolding env _ _ _ _ (DFunUnfolding ar con ops)
  = return (DFunUnfolding ar con ops')
712
  where
713
    ops' = map (substExpr (text "simplUnfolding") env) ops
714

715
simplUnfolding env top_lvl id _ _ 
716
    (CoreUnfolding { uf_tmpl = expr, uf_arity = arity
717
                   , uf_src = src, uf_guidance = guide })
718
  | isStableSource src
719 720
  = do { expr' <- simplExpr rule_env expr
       ; let src' = CoreSubst.substUnfoldingSource (mkCoreSubst (text "inline-unf") env) src
721 722 723
             is_top_lvl = isTopLevel top_lvl
       ; case guide of
           UnfIfGoodArgs{} ->
724 725 726 727
              -- We need to force bottoming, or the new unfolding holds
              -- on to the old unfolding (which is part of the id).
              let bottoming = isBottomingId id
              in bottoming `seq` return (mkUnfolding src' is_top_lvl bottoming expr')
728 729 730 731 732
                -- If the guidance is UnfIfGoodArgs, this is an INLINABLE
                -- unfolding, and we need to make sure the guidance is kept up
                -- to date with respect to any changes in the unfolding.
           _other -> 
              return (mkCoreUnfolding src' is_top_lvl expr' arity guide)
733
		-- See Note [Top-level flag on inline rules] in CoreUnfold
734
       }
735
  where
736 737
    act      = idInlineActivation id
    rule_env = updMode (updModeForInlineRules act) env
738
       	       -- See Note [Simplifying inside InlineRules] in SimplUtils
739

740
simplUnfolding _ top_lvl id _occ_info new_rhs _
741 742 743 744
  = -- We need to force bottoming, or the new unfolding holds
    -- on to the old unfolding (which is part of the id).
    let bottoming = isBottomingId id
    in bottoming `seq` return (mkUnfolding InlineRhs (isTopLevel top_lvl) bottoming new_rhs)
745 746 747 748 749 750
  -- We make an  unfolding *even for loop-breakers*.
  -- Reason: (a) It might be useful to know that they are WHNF
  -- 	     (b) In TidyPgm we currently assume that, if we want to
  --	         expose the unfolding then indeed we *have* an unfolding
  --		 to expose.  (We could instead use the RHS, but currently
  --		 we don't.)  The simple thing is always to have one.
SamB's avatar
SamB committed
751
\end{code}
752

753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774
Note [Arity decrease]
~~~~~~~~~~~~~~~~~~~~~
Generally speaking the arity of a binding should not decrease.  But it *can* 
legitimately happen becuase of RULES.  Eg
	f = g Int
where g has arity 2, will have arity 2.  But if there's a rewrite rule
	g Int --> h
where h has arity 1, then f's arity will decrease.  Here's a real-life example,
which is in the output of Specialise:

     Rec {
	$dm {Arity 2} = \d.\x. op d
	{-# RULES forall d. $dm Int d = $s$dm #-}
	
	dInt = MkD .... opInt ...
	opInt {Arity 1} = $dm dInt

	$s$dm {Arity 0} = \x. op dInt }

Here opInt has arity 1; but when we apply the rule its arity drops to 0.
That's why Specialise goes to a little trouble to pin the right arity
on specialised functions too.
775

776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807
Note [Setting the new unfolding]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
* If there's an INLINE pragma, we simplify the RHS gently.  Maybe we
  should do nothing at all, but simplifying gently might get rid of 
  more crap.

* If not, we make an unfolding from the new RHS.  But *only* for
  non-loop-breakers. Making loop breakers not have an unfolding at all
  means that we can avoid tests in exprIsConApp, for example.  This is
  important: if exprIsConApp says 'yes' for a recursive thing, then we
  can get into an infinite loop

If there's an InlineRule on a loop breaker, we hang on to the inlining.
It's pretty dodgy, but the user did say 'INLINE'.  May need to revisit
this choice.

Note [Setting the demand info]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
If the unfolding is a value, the demand info may
go pear-shaped, so we nuke it.  Example:
     let x = (a,b) in
     case x of (p,q) -> h p q x
Here x is certainly demanded. But after we've nuked
the case, we'll get just
     let x = (a,b) in h a b x
and now x is not demanded (I'm assuming h is lazy)
This really happens.  Similarly
     let f = \x -> e in ...f..f...
After inlining f at some of its call sites the original binding may
(for example) be no longer strictly demanded.
The solution here is a bit ad hoc...

808

809
%************************************************************************
Ian Lynagh's avatar
Ian Lynagh committed
810
%*                                                                      *
811
\subsection[Simplify-simplExpr]{The main function: simplExpr}
Ian Lynagh's avatar
Ian Lynagh committed
812
%*                                                                      *
813 814
%************************************************************************

815 816 817 818 819 820
The reason for this OutExprStuff stuff is that we want to float *after*
simplifying a RHS, not before.  If we do so naively we get quadratic
behaviour as things float out.

To see why it's important to do it after, consider this (real) example:

Ian Lynagh's avatar
Ian Lynagh committed
821 822
        let t = f x
        in fst t
823
==>
Ian Lynagh's avatar
Ian Lynagh committed
824 825 826 827
        let t = let a = e1
                    b = e2
                in (a,b)
        in fst t
828
==>
Ian Lynagh's avatar
Ian Lynagh committed
829 830 831 832 833
        let a = e1
            b = e2
            t = (a,b)
        in
        a       -- Can't inline a this round, cos it appears twice
834
==>
Ian Lynagh's avatar
Ian Lynagh committed
835
        e1
836 837 838 839

Each of the ==> steps is a round of simplification.  We'd save a
whole round if we float first.  This can cascade.  Consider

Ian Lynagh's avatar
Ian Lynagh committed
840 841
        let f = g d
        in \x -> ...f...
842
==>
Ian Lynagh's avatar
Ian Lynagh committed
843 844
        let f = let d1 = ..d.. in \y -> e
        in \x -> ...f...
845
==>
Ian Lynagh's avatar
Ian Lynagh committed
846 847
        let d1 = ..d..
        in \x -> ...(\y ->e)...
848

Ian Lynagh's avatar
Ian Lynagh committed
849
Only in this second round can the \y be applied, and it
850 851 852
might do the same again.


853
\begin{code}
854
simplExpr :: SimplEnv -> CoreExpr -> SimplM CoreExpr
855
simplExpr env expr = simplExprC env expr mkBoringStop
856

857
simplExprC :: SimplEnv -> CoreExpr -> SimplCont -> SimplM CoreExpr
Ian Lynagh's avatar
Ian Lynagh committed
858 859
        -- Simplify an expression, given a continuation
simplExprC env expr cont
860
  = -- pprTrace "simplExprC" (ppr expr $$ ppr cont {- $$ ppr (seIdSubst env) -} $$ ppr (seFloats env) ) $
Ian Lynagh's avatar
Ian Lynagh committed
861 862 863 864
    do  { (env', expr') <- simplExprF (zapFloats env) expr cont
        ; -- pprTrace "simplExprC ret" (ppr expr $$ ppr expr') $
          -- pprTrace "simplExprC ret3" (ppr (seInScope env')) $
          -- pprTrace "simplExprC ret4" (ppr (seFloats env')) $
865 866 867 868
          return (wrapFloats env' expr') }

--------------------------------------------------
simplExprF :: SimplEnv -> InExpr -> SimplCont
Ian Lynagh's avatar
Ian Lynagh committed
869
           -> SimplM (SimplEnv, OutExpr)
870

Ian Lynagh's avatar
Ian Lynagh committed
871
simplExprF env e cont
872 873
  = -- pprTrace "simplExprF" (ppr e $$ ppr cont $$ ppr (seTvSubst env) $$ ppr (seIdSubst env) {- $$ ppr (seFloats env) -} ) $
    simplExprF' env e cont
Ian Lynagh's avatar
Ian Lynagh committed
874

Ian Lynagh's avatar
Ian Lynagh committed
875 876
simplExprF' :: SimplEnv -> InExpr -> SimplCont
            -> SimplM (SimplEnv, OutExpr)
877
simplExprF' env (Var v)        cont = simplVarF env v cont
878 879 880 881
simplExprF' env (Lit lit)      cont = rebuild env (Lit lit) cont
simplExprF' env (Note n expr)  cont = simplNote env n expr cont
simplExprF' env (Cast body co) cont = simplCast env body co cont
simplExprF' env (App fun arg)  cont = simplExprF env fun $
Ian Lynagh's avatar
Ian Lynagh committed
882
                                      ApplyTo NoDup arg env cont
883

Ian Lynagh's avatar
Ian Lynagh committed
884
simplExprF' env expr@(Lam _ _) cont
885
  = simplLam env (map zap bndrs) body cont
Ian Lynagh's avatar
Ian Lynagh committed
886 887 888 889 890 891
        -- The main issue here is under-saturated lambdas
        --   (\x1. \x2. e) arg1
        -- Here x1 might have "occurs-once" occ-info, because occ-info
        -- is computed assuming that a group of lambdas is applied
        -- all at once.  If there are too few args, we must zap the
        -- occ-info.
892 893 894 895
  where
    n_args   = countArgs cont
    n_params = length bndrs
    (bndrs, body) = collectBinders expr
Ian Lynagh's avatar
Ian Lynagh committed
896
    zap | n_args >= n_params = \b -> b
897
        | otherwise          = \b -> if isTyCoVar b then b
Ian Lynagh's avatar
Ian Lynagh committed
898 899 900
                                     else zapLamIdInfo b
        -- NB: we count all the args incl type args
        -- so we must count all the binders (incl type lambdas)
901

902
simplExprF' env (Type ty) cont
903
  = ASSERT( contIsRhsOrArg cont )
904
    do  { ty' <- simplCoercion env ty
Ian Lynagh's avatar
Ian Lynagh committed
905
        ; rebuild env (Type ty') cont }
906

907
simplExprF' env (Case scrut bndr _ alts) cont
908
  | not (switchIsOn (getSwitchChecker env) NoCaseOfCase)
Ian Lynagh's avatar
Ian Lynagh committed
909
  =     -- Simplify the scrutinee with a Select continuation
910
    simplExprF env scrut (Select NoDup bndr alts env cont)
911

912
  | otherwise
Ian Lynagh's avatar
Ian Lynagh committed
913 914
  =     -- If case-of-case is off, simply simplify the case expression
        -- in a vanilla Stop context, and rebuild the result around it
915 916
    do  { case_expr' <- simplExprC env scrut
                             (Select NoDup bndr alts env mkBoringStop)
Ian Lynagh's avatar
Ian Lynagh committed
917
        ; rebuild env case_expr' cont }
918

919
simplExprF' env (Let (Rec pairs) body) cont
Ian Lynagh's avatar
Ian Lynagh committed
920
  = do  { env' <- simplRecBndrs env (map fst pairs)
Ian Lynagh's avatar
Ian Lynagh committed
921 922
                -- NB: bndrs' don't have unfoldings or rules
                -- We add them as we go down
923

Ian Lynagh's avatar
Ian Lynagh committed
924 925
        ; env'' <- simplRecBind env' NotTopLevel pairs
        ; simplExprF env'' body cont }
926

927 928
simplExprF' env (Let (NonRec bndr rhs) body) cont
  = simplNonRecE env bndr (rhs, env) ([], body) cont
929 930

---------------------------------