TcCanonical.hs 77.1 KB
Newer Older
1 2
{-# LANGUAGE CPP #-}

3
module TcCanonical(
4 5
     canonicalize,
     unifyDerived,
6
     makeSuperClasses, maybeSym,
7 8
     StopOrContinue(..), stopWith, continueWith
  ) where
9 10 11 12

#include "HsVersions.h"

import TcRnTypes
13
import TcUnify( swapOverTyVars, metaTyVarUpdateOK )
14
import TcType
15
import Type
16 17
import TcFlatten
import TcSMonad
18
import TcEvidence
19 20
import Class
import TyCon
21
import TyCoRep   -- cleverly decomposes types, good for completeness checking
22 23 24
import Coercion
import FamInstEnv ( FamInstEnvs )
import FamInst ( tcTopNormaliseNewTypeTF_maybe )
25 26
import Var
import Outputable
27
import DynFlags( DynFlags )
28
import NameSet
29
import RdrName
30

31
import Pair
32
import Util
33
import Bag
34 35 36
import MonadUtils
import Control.Monad
import Data.List  ( zip4, foldl' )
37
import BasicTypes
38

39 40
import Data.Bifunctor ( bimap )

Austin Seipp's avatar
Austin Seipp committed
41 42 43 44 45 46
{-
************************************************************************
*                                                                      *
*                      The Canonicaliser                               *
*                                                                      *
************************************************************************
47

48 49
Note [Canonicalization]
~~~~~~~~~~~~~~~~~~~~~~~
50

51
Canonicalization converts a simple constraint to a canonical form. It is
52 53 54
unary (i.e. treats individual constraints one at a time), does not do
any zonking, but lives in TcS monad because it needs to create fresh
variables (for flattening) and consult the inerts (for efficiency).
55

56
The execution plan for canonicalization is the following:
Simon Peyton Jones's avatar
Simon Peyton Jones committed
57 58

  1) Decomposition of equalities happens as necessary until we reach a
59
     variable or type family in one side. There is no decomposition step
Simon Peyton Jones's avatar
Simon Peyton Jones committed
60
     for other forms of constraints.
61

Simon Peyton Jones's avatar
Simon Peyton Jones committed
62 63 64 65
  2) If, when we decompose, we discover a variable on the head then we
     look at inert_eqs from the current inert for a substitution for this
     variable and contine decomposing. Hence we lazily apply the inert
     substitution if it is needed.
66

67 68
  3) If no more decomposition is possible, we deeply apply the substitution
     from the inert_eqs and continue with flattening.
69

Simon Peyton Jones's avatar
Simon Peyton Jones committed
70 71 72 73 74
  4) During flattening, we examine whether we have already flattened some
     function application by looking at all the CTyFunEqs with the same
     function in the inert set. The reason for deeply applying the inert
     substitution at step (3) is to maximise our chances of matching an
     already flattened family application in the inert.
75

Simon Peyton Jones's avatar
Simon Peyton Jones committed
76 77
The net result is that a constraint coming out of the canonicalization
phase cannot be rewritten any further from the inerts (but maybe /it/ can
78 79
rewrite an inert or still interact with an inert in a further phase in the
simplifier.
dimitris's avatar
dimitris committed
80

81
Note [Caching for canonicals]
Simon Peyton Jones's avatar
Simon Peyton Jones committed
82
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
83 84 85 86
Our plan with pre-canonicalization is to be able to solve a constraint
really fast from existing bindings in TcEvBinds. So one may think that
the condition (isCNonCanonical) is not necessary.  However consider
the following setup:
87

Simon Peyton Jones's avatar
Simon Peyton Jones committed
88 89
InertSet = { [W] d1 : Num t }
WorkList = { [W] d2 : Num t, [W] c : t ~ Int}
90

91 92 93 94 95
Now, we prioritize equalities, but in our concrete example
(should_run/mc17.hs) the first (d2) constraint is dealt with first,
because (t ~ Int) is an equality that only later appears in the
worklist since it is pulled out from a nested implication
constraint. So, let's examine what happens:
Simon Peyton Jones's avatar
Simon Peyton Jones committed
96

97 98
   - We encounter work item (d2 : Num t)

Simon Peyton Jones's avatar
Simon Peyton Jones committed
99
   - Nothing is yet in EvBinds, so we reach the interaction with inerts
100
     and set:
Simon Peyton Jones's avatar
Simon Peyton Jones committed
101
              d2 := d1
102 103
    and we discard d2 from the worklist. The inert set remains unaffected.

104 105 106
   - Now the equation ([W] c : t ~ Int) is encountered and kicks-out
     (d1 : Num t) from the inerts.  Then that equation gets
     spontaneously solved, perhaps. We end up with:
107
        InertSet : { [G] c : t ~ Int }
Simon Peyton Jones's avatar
Simon Peyton Jones committed
108
        WorkList : { [W] d1 : Num t}
109

110 111
   - Now we examine (d1), we observe that there is a binding for (Num
     t) in the evidence binds and we set:
Simon Peyton Jones's avatar
Simon Peyton Jones committed
112
             d1 := d2
113 114
     and end up in a loop!

115 116 117 118 119 120 121 122
Now, the constraints that get kicked out from the inert set are always
Canonical, so by restricting the use of the pre-canonicalizer to
NonCanonical constraints we eliminate this danger. Moreover, for
canonical constraints we already have good caching mechanisms
(effectively the interaction solver) and we are interested in reducing
things like superclasses of the same non-canonical constraint being
generated hence I don't expect us to lose a lot by introducing the
(isCNonCanonical) restriction.
123

124 125 126 127 128 129 130
A similar situation can arise in TcSimplify, at the end of the
solve_wanteds function, where constraints from the inert set are
returned as new work -- our substCt ensures however that if they are
not rewritten by subst, they remain canonical and hence we will not
attempt to solve them from the EvBinds. If on the other hand they did
get rewritten and are now non-canonical they will still not match the
EvBinds, so we are again good.
Austin Seipp's avatar
Austin Seipp committed
131
-}
132

133 134 135
-- Top-level canonicalization
-- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

136
canonicalize :: Ct -> TcS (StopOrContinue Ct)
137
canonicalize ct@(CNonCanonical { cc_ev = ev })
138
  = do { traceTcS "canonicalize (non-canonical)" (ppr ct)
139
       ; {-# SCC "canEvVar" #-}
140
         canEvNC ev }
141

142 143
canonicalize (CDictCan { cc_ev = ev, cc_class  = cls
                       , cc_tyargs = xis, cc_pend_sc = pend_sc })
144
  = {-# SCC "canClass" #-}
145 146
    canClass ev cls xis pend_sc

147
canonicalize (CTyEqCan { cc_ev = ev
148
                       , cc_tyvar  = tv
149 150
                       , cc_rhs    = xi
                       , cc_eq_rel = eq_rel })
151
  = {-# SCC "canEqLeafTyVarEq" #-}
152 153 154
    canEqNC ev eq_rel (mkTyVarTy tv) xi
      -- NB: Don't use canEqTyVar because that expects flattened types,
      -- and tv and xi may not be flat w.r.t. an updated inert set
155

156
canonicalize (CFunEqCan { cc_ev = ev
157 158
                        , cc_fun    = fn
                        , cc_tyargs = xis1
159
                        , cc_fsk    = fsk })
Simon Peyton Jones's avatar
Simon Peyton Jones committed
160
  = {-# SCC "canEqLeafFunEq" #-}
161
    canCFunEqCan ev fn xis1 fsk
162

163 164
canonicalize (CIrredEvCan { cc_ev = ev })
  = canIrred ev
165 166
canonicalize (CHoleCan { cc_ev = ev, cc_hole = hole })
  = canHole ev hole
167

168
canEvNC :: CtEvidence -> TcS (StopOrContinue Ct)
Simon Peyton Jones's avatar
Simon Peyton Jones committed
169
-- Called only for non-canonical EvVars
170
canEvNC ev
171
  = case classifyPredType (ctEvPred ev) of
172 173 174 175 176 177
      ClassPred cls tys     -> do traceTcS "canEvNC:cls" (ppr cls <+> ppr tys)
                                  canClassNC ev cls tys
      EqPred eq_rel ty1 ty2 -> do traceTcS "canEvNC:eq" (ppr ty1 $$ ppr ty2)
                                  canEqNC    ev eq_rel ty1 ty2
      IrredPred {}          -> do traceTcS "canEvNC:irred" (ppr (ctEvPred ev))
                                  canIrred   ev
Austin Seipp's avatar
Austin Seipp committed
178 179 180 181 182 183 184
{-
************************************************************************
*                                                                      *
*                      Class Canonicalization
*                                                                      *
************************************************************************
-}
185

186
canClassNC :: CtEvidence -> Class -> [Type] -> TcS (StopOrContinue Ct)
187 188
-- "NC" means "non-canonical"; that is, we have got here
-- from a NonCanonical constrataint, not from a CDictCan
Simon Peyton Jones's avatar
Simon Peyton Jones committed
189
-- Precondition: EvVar is class evidence
190 191 192 193 194 195 196
canClassNC ev cls tys
  | isGiven ev  -- See Note [Eagerly expand given superclasses]
  = do { sc_cts <- mkStrictSuperClasses ev cls tys
       ; emitWork sc_cts
       ; canClass ev cls tys False }
  | otherwise
  = canClass ev cls tys (has_scs cls)
197 198
  where
    has_scs cls = not (null (classSCTheta cls))
199

200 201 202 203
canClass :: CtEvidence
         -> Class -> [Type]
         -> Bool            -- True <=> un-explored superclasses
         -> TcS (StopOrContinue Ct)
204
-- Precondition: EvVar is class evidence
205

206
canClass ev cls tys pend_sc
207 208
  =   -- all classes do *nominal* matching
    ASSERT2( ctEvRole ev == Nominal, ppr ev $$ ppr cls $$ ppr tys )
209
    do { (xis, cos) <- flattenManyNom ev tys
Joachim Breitner's avatar
Joachim Breitner committed
210
       ; let co = mkTcTyConAppCo Nominal (classTyCon cls) cos
211
             xi = mkClassPred cls xis
212
             mk_ct new_ev = CDictCan { cc_ev = new_ev
213 214 215
                                     , cc_tyargs = xis
                                     , cc_class = cls
                                     , cc_pend_sc = pend_sc }
216
       ; mb <- rewriteEvidence ev xi co
217
       ; traceTcS "canClass" (vcat [ ppr ev
Simon Peyton Jones's avatar
Simon Peyton Jones committed
218
                                   , ppr xi, ppr mb ])
219
       ; return (fmap mk_ct mb) }
dimitris's avatar
dimitris committed
220

221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261
{- Note [The superclass story]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
We need to add superclass constraints for two reasons:

* For givens, they give us a route to to proof.  E.g.
    f :: Ord a => a -> Bool
    f x = x == x
  We get a Wanted (Eq a), which can only be solved from the superclass
  of the Given (Ord a).

* For wanteds, they may give useful functional dependencies.  E.g.
     class C a b | a -> b where ...
     class C a b => D a b where ...
  Now a Wanted constraint (D Int beta) has (C Int beta) as a superclass
  and that might tell us about beta, via C's fundeps.  We can get this
  by generateing a Derived (C Int beta) constraint.  It's derived because
  we don't actually have to cough up any evidence for it; it's only there
  to generate fundep equalities.

See Note [Why adding superclasses can help].

For these reasons we want to generate superclass constraints for both
Givens and Wanteds. But:

* (Minor) they are often not needed, so generating them aggressively
  is a waste of time.

* (Major) if we want recursive superclasses, there would be an infinite
  number of them.  Here is a real-life example (Trac #10318);

     class (Frac (Frac a) ~ Frac a,
            Fractional (Frac a),
            IntegralDomain (Frac a))
         => IntegralDomain a where
      type Frac a :: *

  Notice that IntegralDomain has an associated type Frac, and one
  of IntegralDomain's superclasses is another IntegralDomain constraint.

So here's the plan:

262 263 264 265 266 267 268
1. Eagerly generate superclasses for given (but not wanted)
   constraints; see Note [Eagerly expand given superclasses].
   This is done in canClassNC, when we take a non-canonical constraint
   and cannonicalise it.

   However stop if you encounter the same class twice.  That is,
   expand eagerly, but have a conservative termination condition: see
269 270
   Note [Expanding superclasses] in TcType.

271 272 273 274 275 276 277 278 279
2. Solve the wanteds as usual, but do no further expansion of
   superclasses for canonical CDictCans in solveSimpleGivens or
   solveSimpleWanteds; Note [Danger of adding superclasses during solving]

   However, /do/ continue to eagerly expand superlasses for /given/
   non-canonical constraints (canClassNC does this).  As Trac #12175
   showed, a type-family application can expand to a class constraint,
   and we want to see its superclasses for just the same reason as
   Note [Eagerly expand given superclasses].
280

281 282 283 284 285 286
3. If we have any remaining unsolved wanteds
        (see Note [When superclasses help] in TcRnTypes)
   try harder: take both the Givens and Wanteds, and expand
   superclasses again.  This may succeed in generating (a finite
   number of) extra Givens, and extra Deriveds. Both may help the
   proof.  This is done in TcSimplify.expandSuperClasses.
287 288 289 290 291 292 293 294 295 296 297 298 299

4. Go round to (2) again.  This loop (2,3,4) is implemented
   in TcSimplify.simpl_loop.

We try to terminate the loop by flagging which class constraints
(given or wanted) are potentially un-expanded.  This is what the
cc_pend_sc flag is for in CDictCan.  So in Step 3 we only expand
superclasses for constraints with cc_pend_sc set to true (i.e.
isPendingScDict holds).

When we take a CNonCanonical or CIrredCan, but end up classifying it
as a CDictCan, we set the cc_pend_sc flag to False.

300 301 302 303 304
Note [Eagerly expand given superclasses]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In step (1) of Note [The superclass story], why do we eagerly expand
Given superclasses by one layer?  Mainly because of some very obscure
cases like this:
305 306 307 308 309 310 311 312 313 314 315

   instance Bad a => Eq (T a)

   f :: (Ord (T a)) => blah
   f x = ....needs Eq (T a), Ord (T a)....

Here if we can't satisfy (Eq (T a)) from the givens we'll use the
instance declaration; but then we are stuck with (Bad a).  Sigh.
This is really a case of non-confluent proofs, but to stop our users
complaining we expand one layer in advance.

316 317 318 319 320 321 322 323 324 325 326 327 328
Note [Instance and Given overlap] in TcInteract.

We also want to do this if we have

   f :: F (T a) => blah

where
   type instance F (T a) = Ord (T a)

So we may need to do a little work on the givens to expose the
class that has the superclasses.  That's why the superclass
expansion for Givens happens in canClassNC.

329 330 331
Note [Why adding superclasses can help]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Examples of how adding superclasses can help:
332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349

    --- Example 1
        class C a b | a -> b
    Suppose we want to solve
         [G] C a b
         [W] C a beta
    Then adding [D] beta~b will let us solve it.

    -- Example 2 (similar but using a type-equality superclass)
        class (F a ~ b) => C a b
    And try to sllve:
         [G] C a b
         [W] C a beta
    Follow the superclass rules to add
         [G] F a ~ b
         [D] F a ~ beta
    Now we we get [D] beta ~ b, and can solve that.

350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367
    -- Example (tcfail138)
      class L a b | a -> b
      class (G a, L a b) => C a b

      instance C a b' => G (Maybe a)
      instance C a b  => C (Maybe a) a
      instance L (Maybe a) a

    When solving the superclasses of the (C (Maybe a) a) instance, we get
      [G] C a b, and hance by superclasses, [G] G a, [G] L a b
      [W] G (Maybe a)
    Use the instance decl to get
      [W] C a beta
    Generate its derived superclass
      [D] L a beta.  Now using fundeps, combine with [G] L a b to get
      [D] beta ~ b
    which is what we want.

368 369
Note [Danger of adding superclasses during solving]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
370
Here's a serious, but now out-dated example, from Trac #4497:
Simon Peyton Jones's avatar
Simon Peyton Jones committed
371

372 373 374
   class Num (RealOf t) => Normed t
   type family RealOf x

Simon Peyton Jones's avatar
Simon Peyton Jones committed
375
Assume the generated wanted constraint is:
376 377 378
   [W] RealOf e ~ e
   [W] Normed e

Simon Peyton Jones's avatar
Simon Peyton Jones committed
379
If we were to be adding the superclasses during simplification we'd get:
380 381 382 383
   [W] RealOf e ~ e
   [W] Normed e
   [D] RealOf e ~ fuv
   [D] Num fuv
Simon Peyton Jones's avatar
Simon Peyton Jones committed
384
==>
385
   e := fuv, Num fuv, Normed fuv, RealOf fuv ~ fuv
Simon Peyton Jones's avatar
Simon Peyton Jones committed
386

387 388 389
While looks exactly like our original constraint. If we add the
superclass of (Normed fuv) again we'd loop.  By adding superclasses
definitely only once, during canonicalisation, this situation can't
390
happen.
391 392 393 394

Mind you, now that Wanteds cannot rewrite Derived, I think this particular
situation can't happen.
  -}
395

396 397 398 399
makeSuperClasses :: [Ct] -> TcS [Ct]
-- Returns strict superclasses, transitively, see Note [The superclasses story]
-- See Note [The superclass story]
-- The loop-breaking here follows Note [Expanding superclasses] in TcType
400 401 402 403 404 405 406 407 408 409
-- Specifically, for an incoming (C t) constraint, we return all of (C t)'s
--    superclasses, up to /and including/ the first repetition of C
--
-- Example:  class D a => C a
--           class C [a] => D a
-- makeSuperClasses (C x) will return (D x, C [x])
--
-- NB: the incoming constraints have had their cc_pend_sc flag already
--     flipped to False, by isPendingScDict, so we are /obliged/ to at
--     least produce the immediate superclasses
410 411 412
makeSuperClasses cts = concatMapM go cts
  where
    go (CDictCan { cc_ev = ev, cc_class = cls, cc_tyargs = tys })
413
          = mkStrictSuperClasses ev cls tys
414 415
    go ct = pprPanic "makeSuperClasses" (ppr ct)

416 417 418 419 420
mkStrictSuperClasses :: CtEvidence -> Class -> [Type] -> TcS [Ct]
-- Return constraints for the strict superclasses of (c tys)
mkStrictSuperClasses ev cls tys
  = mk_strict_superclasses (unitNameSet (className cls)) ev cls tys

421 422 423 424 425 426 427 428 429 430
mk_superclasses :: NameSet -> CtEvidence -> TcS [Ct]
-- Return this constraint, plus its superclasses, if any
mk_superclasses rec_clss ev
  | ClassPred cls tys <- classifyPredType (ctEvPred ev)
  = mk_superclasses_of rec_clss ev cls tys

  | otherwise   -- Superclass is not a class predicate
  = return [mkNonCanonical ev]

mk_superclasses_of :: NameSet -> CtEvidence -> Class -> [Type] -> TcS [Ct]
431 432
-- Always return this class constraint,
-- and expand its superclasses
433
mk_superclasses_of rec_clss ev cls tys
434 435 436 437 438 439 440
  | loop_found = do { traceTcS "mk_superclasses_of: loop" (ppr cls <+> ppr tys)
                    ; return [this_ct] }  -- cc_pend_sc of this_ct = True
  | otherwise  = do { traceTcS "mk_superclasses_of" (vcat [ ppr cls <+> ppr tys
                                                          , ppr (isCTupleClass cls)
                                                          , ppr rec_clss
                                                          ])
                    ; sc_cts <- mk_strict_superclasses rec_clss' ev cls tys
441 442
                    ; return (this_ct : sc_cts) }
                                   -- cc_pend_sc of this_ct = False
443 444
  where
    cls_nm     = className cls
445
    loop_found = not (isCTupleClass cls) && cls_nm `elemNameSet` rec_clss
446
                 -- Tuples never contribute to recursion, and can be nested
447
    rec_clss'  = rec_clss `extendNameSet` cls_nm
448 449
    this_ct    = CDictCan { cc_ev = ev, cc_class = cls, cc_tyargs = tys
                          , cc_pend_sc = loop_found }
450 451
                 -- NB: If there is a loop, we cut off, so we have not
                 --     added the superclasses, hence cc_pend_sc = True
452 453

mk_strict_superclasses :: NameSet -> CtEvidence -> Class -> [Type] -> TcS [Ct]
454 455 456
-- Always return the immediate superclasses of (cls tys);
-- and expand their superclasses, provided none of them are in rec_clss
-- nor are repeated
457 458 459 460 461
mk_strict_superclasses rec_clss ev cls tys
  | CtGiven { ctev_evar = evar, ctev_loc = loc } <- ev
  = do { sc_evs <- newGivenEvVars (mk_given_loc loc)
                                  (mkEvScSelectors (EvId evar) cls tys)
       ; concatMapM (mk_superclasses rec_clss) sc_evs }
dimitris's avatar
dimitris committed
462

463
  | all noFreeVarsOfType tys
464
  = return [] -- Wanteds with no variables yield no deriveds.
465
              -- See Note [Improvement from Ground Wanteds]
466

467 468
  | otherwise -- Wanted/Derived case, just add Derived superclasses
              -- that can lead to improvement.
469 470 471
  = do { let loc = ctEvLoc ev
       ; sc_evs <- mapM (newDerivedNC loc) (immSuperClasses cls tys)
       ; concatMapM (mk_superclasses rec_clss) sc_evs }
472
  where
473
    size = sizeTypes tys
474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490
    mk_given_loc loc
       | isCTupleClass cls
       = loc   -- For tuple predicates, just take them apart, without
               -- adding their (large) size into the chain.  When we
               -- get down to a base predicate, we'll include its size.
               -- Trac #10335

       | GivenOrigin skol_info <- ctLocOrigin loc
         -- See Note [Solving superclass constraints] in TcInstDcls
         -- for explantation of this transformation for givens
       = case skol_info of
            InstSkol -> loc { ctl_origin = GivenOrigin (InstSC size) }
            InstSC n -> loc { ctl_origin = GivenOrigin (InstSC (n `max` size)) }
            _        -> loc

       | otherwise  -- Probably doesn't happen, since this function
       = loc        -- is only used for Givens, but does no harm
491

492

Austin Seipp's avatar
Austin Seipp committed
493 494 495 496 497 498 499
{-
************************************************************************
*                                                                      *
*                      Irreducibles canonicalization
*                                                                      *
************************************************************************
-}
500

501
canIrred :: CtEvidence -> TcS (StopOrContinue Ct)
502
-- Precondition: ty not a tuple and no other evidence form
503
canIrred old_ev
Simon Peyton Jones's avatar
Simon Peyton Jones committed
504 505
  = do { let old_ty = ctEvPred old_ev
       ; traceTcS "can_pred" (text "IrredPred = " <+> ppr old_ty)
506
       ; (xi,co) <- flatten FM_FlattenAll old_ev old_ty -- co :: xi ~ old_ty
507
       ; rewriteEvidence old_ev xi co `andWhenContinue` \ new_ev ->
508 509
    do { -- Re-classify, in case flattening has improved its shape
       ; case classifyPredType (ctEvPred new_ev) of
510 511 512 513
           ClassPred cls tys     -> canClassNC new_ev cls tys
           EqPred eq_rel ty1 ty2 -> canEqNC new_ev eq_rel ty1 ty2
           _                     -> continueWith $
                                    CIrredEvCan { cc_ev = new_ev } } }
514

515 516
canHole :: CtEvidence -> Hole -> TcS (StopOrContinue Ct)
canHole ev hole
517 518
  = do { let ty = ctEvPred ev
       ; (xi,co) <- flatten FM_SubstOnly ev ty -- co :: xi ~ ty
519 520
       ; rewriteEvidence ev xi co `andWhenContinue` \ new_ev ->
    do { emitInsoluble (CHoleCan { cc_ev = new_ev
521
                                 , cc_hole = hole })
522
       ; stopWith new_ev "Emit insoluble hole" } }
523

Austin Seipp's avatar
Austin Seipp committed
524 525 526 527 528 529
{-
************************************************************************
*                                                                      *
*        Equalities
*                                                                      *
************************************************************************
530 531 532 533 534 535 536 537 538 539 540 541 542

Note [Canonicalising equalities]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In order to canonicalise an equality, we look at the structure of the
two types at hand, looking for similarities. A difficulty is that the
types may look dissimilar before flattening but similar after flattening.
However, we don't just want to jump in and flatten right away, because
this might be wasted effort. So, after looking for similarities and failing,
we flatten and then try again. Of course, we don't want to loop, so we
track whether or not we've already flattened.

It is conceivable to do a better job at tracking whether or not a type
is flattened, but this is left as future work. (Mar '15)
Austin Seipp's avatar
Austin Seipp committed
543
-}
544

545 546
canEqNC :: CtEvidence -> EqRel -> Type -> Type -> TcS (StopOrContinue Ct)
canEqNC ev eq_rel ty1 ty2
547 548 549 550
  = do { result <- zonk_eq_types ty1 ty2
       ; case result of
           Left (Pair ty1' ty2') -> can_eq_nc False ev eq_rel ty1' ty1 ty2' ty2
           Right ty              -> canEqReflexive ev eq_rel ty }
551

552
can_eq_nc
553 554
   :: Bool            -- True => both types are flat
   -> CtEvidence
555
   -> EqRel
Austin Seipp's avatar
Austin Seipp committed
556 557
   -> Type -> Type    -- LHS, after and before type-synonym expansion, resp
   -> Type -> Type    -- RHS, after and before type-synonym expansion, resp
558
   -> TcS (StopOrContinue Ct)
559
can_eq_nc flat ev eq_rel ty1 ps_ty1 ty2 ps_ty2
Austin Seipp's avatar
Austin Seipp committed
560
  = do { traceTcS "can_eq_nc" $
561
         vcat [ ppr flat, ppr ev, ppr eq_rel, ppr ty1, ppr ps_ty1, ppr ty2, ppr ps_ty2 ]
562 563
       ; rdr_env <- getGlobalRdrEnvTcS
       ; fam_insts <- getFamInstEnvs
564
       ; can_eq_nc' flat rdr_env fam_insts ev eq_rel ty1 ps_ty1 ty2 ps_ty2 }
565 566

can_eq_nc'
567 568
   :: Bool           -- True => both input types are flattened
   -> GlobalRdrEnv   -- needed to see which newtypes are in scope
569 570 571 572 573 574
   -> FamInstEnvs    -- needed to unwrap data instances
   -> CtEvidence
   -> EqRel
   -> Type -> Type    -- LHS, after and before type-synonym expansion, resp
   -> Type -> Type    -- RHS, after and before type-synonym expansion, resp
   -> TcS (StopOrContinue Ct)
575 576

-- Expand synonyms first; see Note [Type synonyms and canonicalization]
577
can_eq_nc' flat _rdr_env _envs ev eq_rel ty1 ps_ty1 ty2 ps_ty2
578 579
  | Just ty1' <- coreView ty1 = can_eq_nc flat ev eq_rel ty1' ps_ty1 ty2  ps_ty2
  | Just ty2' <- coreView ty2 = can_eq_nc flat ev eq_rel ty1  ps_ty1 ty2' ps_ty2
580 581

-- need to check for reflexivity in the ReprEq case.
eir@cis.upenn.edu's avatar
eir@cis.upenn.edu committed
582
-- See Note [Eager reflexivity check]
583 584 585
-- Check only when flat because the zonk_eq_types check in canEqNC takes
-- care of the non-flat case.
can_eq_nc' True _rdr_env _envs ev ReprEq ty1 _ ty2 _
eir@cis.upenn.edu's avatar
eir@cis.upenn.edu committed
586
  | ty1 `tcEqType` ty2
587 588 589 590
  = canEqReflexive ev ReprEq ty1

-- When working with ReprEq, unwrap newtypes.
can_eq_nc' _flat rdr_env envs ev ReprEq ty1 _ ty2 ps_ty2
591 592
  | Just stuff1 <- tcTopNormaliseNewTypeTF_maybe envs rdr_env ty1
  = can_eq_newtype_nc ev NotSwapped ty1 stuff1 ty2 ps_ty2
593
can_eq_nc' _flat rdr_env envs ev ReprEq ty1 ps_ty1 ty2 _
594 595
  | Just stuff2 <- tcTopNormaliseNewTypeTF_maybe envs rdr_env ty2
  = can_eq_newtype_nc ev IsSwapped  ty2 stuff2 ty1 ps_ty1
596

597 598 599 600 601 602
-- Then, get rid of casts
can_eq_nc' flat _rdr_env _envs ev eq_rel (CastTy ty1 co1) _ ty2 ps_ty2
  = canEqCast flat ev eq_rel NotSwapped ty1 co1 ty2 ps_ty2
can_eq_nc' flat _rdr_env _envs ev eq_rel ty1 ps_ty1 (CastTy ty2 co2) _
  = canEqCast flat ev eq_rel IsSwapped ty2 co2 ty1 ps_ty1

603 604 605 606 607
----------------------
-- Otherwise try to decompose
----------------------

-- Literals
608
can_eq_nc' _flat _rdr_env _envs ev eq_rel ty1@(LitTy l1) _ (LitTy l2) _
609
 | l1 == l2
610
  = do { setEqIfWanted ev (mkReflCo (eqRelRole eq_rel) ty1)
611
       ; stopWith ev "Equal LitTy" }
612

Simon Peyton Jones's avatar
Simon Peyton Jones committed
613 614
-- Try to decompose type constructor applications
-- Including FunTy (s -> t)
615
can_eq_nc' _flat _rdr_env _envs ev eq_rel ty1 _ ty2 _
616 617
  | Just (tc1, tys1) <- tcRepSplitTyConApp_maybe ty1
  , Just (tc2, tys2) <- tcRepSplitTyConApp_maybe ty2
618 619
  , not (isTypeFamilyTyCon tc1)
  , not (isTypeFamilyTyCon tc2)
620
  = canTyConApp ev eq_rel tc1 tys1 tc2 tys2
621

622
can_eq_nc' _flat _rdr_env _envs ev eq_rel
623
           s1@(ForAllTy {}) _ s2@(ForAllTy {}) _
624
 | CtWanted { ctev_loc = loc, ctev_dest = orig_dest } <- ev
625 626
 = do { let (bndrs1,body1) = tcSplitForAllTyVarBndrs s1
            (bndrs2,body2) = tcSplitForAllTyVarBndrs s2
627
      ; if not (equalLength bndrs1 bndrs2)
628 629
        then do { traceTcS "Forall failure" $
                     vcat [ ppr s1, ppr s2, ppr bndrs1, ppr bndrs2
630 631
                          , ppr (map binderArgFlag bndrs1)
                          , ppr (map binderArgFlag bndrs2) ]
632
                ; canEqHardFailure ev s1 s2 }
633
        else
634
          do { traceTcS "Creating implication for polytype equality" $ ppr ev
635
             ; kind_cos <- zipWithM (unifyWanted loc Nominal)
636
                             (map binderKind bndrs1) (map binderKind bndrs2)
637 638 639
             ; all_co <- deferTcSForAllEq (eqRelRole eq_rel) loc
                                           kind_cos (bndrs1,body1) (bndrs2,body2)
             ; setWantedEq orig_dest all_co
640
             ; stopWith ev "Deferred polytype equality" } }
641
 | otherwise
642
 = do { traceTcS "Omitting decomposition of given polytype equality" $
643
        pprEq s1 s2    -- See Note [Do not decompose given polytype equalities]
644
      ; stopWith ev "Discard given polytype equality" }
645

646
-- See Note [Canonicalising type applications] about why we require flat types
eir@cis.upenn.edu's avatar
eir@cis.upenn.edu committed
647
can_eq_nc' True _rdr_env _envs ev eq_rel (AppTy t1 s1) _ ty2 _
648
  | Just (t2, s2) <- tcSplitAppTy_maybe ty2
eir@cis.upenn.edu's avatar
eir@cis.upenn.edu committed
649 650
  = can_eq_app ev eq_rel t1 s1 t2 s2
can_eq_nc' True _rdr_env _envs ev eq_rel ty1 _ (AppTy t2 s2) _
651
  | Just (t1, s1) <- tcSplitAppTy_maybe ty1
eir@cis.upenn.edu's avatar
eir@cis.upenn.edu committed
652
  = can_eq_app ev eq_rel t1 s1 t2 s2
653

654
-- No similarity in type structure detected. Flatten and try again.
655 656 657
can_eq_nc' False rdr_env envs ev eq_rel _ ps_ty1 _ ps_ty2
  = do { (xi1, co1) <- flatten FM_FlattenAll ev ps_ty1
       ; (xi2, co2) <- flatten FM_FlattenAll ev ps_ty2
658
       ; rewriteEqEvidence ev NotSwapped xi1 xi2 co1 co2
659
         `andWhenContinue` \ new_ev ->
660 661
         can_eq_nc' True rdr_env envs new_ev eq_rel xi1 xi1 xi2 xi2 }

662 663
-- Type variable on LHS or RHS are last.
-- NB: pattern match on True: we want only flat types sent to canEqTyVar.
664
-- See also Note [No top-level newtypes on RHS of representational equalities]
665 666 667 668
can_eq_nc' True _rdr_env _envs ev eq_rel (TyVarTy tv1) ps_ty1 ty2 ps_ty2
  = canEqTyVar ev eq_rel NotSwapped tv1 ps_ty1 ty2 ps_ty2
can_eq_nc' True _rdr_env _envs ev eq_rel ty1 ps_ty1 (TyVarTy tv2) ps_ty2
  = canEqTyVar ev eq_rel IsSwapped tv2 ps_ty2 ty1 ps_ty1
669 670

-- We've flattened and the types don't match. Give up.
671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701
can_eq_nc' True _rdr_env _envs ev _eq_rel _ ps_ty1 _ ps_ty2
  = do { traceTcS "can_eq_nc' catch-all case" (ppr ps_ty1 $$ ppr ps_ty2)
       ; canEqHardFailure ev ps_ty1 ps_ty2 }

---------------------------------
-- | Compare types for equality, while zonking as necessary. Gives up
-- as soon as it finds that two types are not equal.
-- This is quite handy when some unification has made two
-- types in an inert wanted to be equal. We can discover the equality without
-- flattening, which is sometimes very expensive (in the case of type functions).
-- In particular, this function makes a ~20% improvement in test case
-- perf/compiler/T5030.
--
-- Returns either the (partially zonked) types in the case of
-- inequality, or the one type in the case of equality. canEqReflexive is
-- a good next step in the 'Right' case. Returning 'Left' is always safe.
--
-- NB: This does *not* look through type synonyms. In fact, it treats type
-- synonyms as rigid constructors. In the future, it might be convenient
-- to look at only those arguments of type synonyms that actually appear
-- in the synonym RHS. But we're not there yet.
zonk_eq_types :: TcType -> TcType -> TcS (Either (Pair TcType) TcType)
zonk_eq_types = go
  where
    go (TyVarTy tv1) (TyVarTy tv2) = tyvar_tyvar tv1 tv2
    go (TyVarTy tv1) ty2           = tyvar NotSwapped tv1 ty2
    go ty1 (TyVarTy tv2)           = tyvar IsSwapped  tv2 ty1

    go ty1 ty2
      | Just (tc1, tys1) <- tcRepSplitTyConApp_maybe ty1
      , Just (tc2, tys2) <- tcRepSplitTyConApp_maybe ty2
702 703 704 705 706 707 708 709 710
      = if tc1 == tc2 && tys1 `equalLength` tys2
          -- Crucial to check for equal-length args, because
          -- we cannot assume that the two args to 'go' have
          -- the same kind.  E.g go (Proxy *      (Maybe Int))
          --                        (Proxy (*->*) Maybe)
          -- We'll call (go (Maybe Int) Maybe)
          -- See Trac #13083
        then tycon tc1 tys1 tys2
        else bale_out ty1 ty2
711 712 713 714 715 716 717 718 719 720 721 722

    go ty1 ty2
      | Just (ty1a, ty1b) <- tcRepSplitAppTy_maybe ty1
      , Just (ty2a, ty2b) <- tcRepSplitAppTy_maybe ty2
      = do { res_a <- go ty1a ty2a
           ; res_b <- go ty1b ty2b
           ; return $ combine_rev mkAppTy res_b res_a }

    go ty1@(LitTy lit1) (LitTy lit2)
      | lit1 == lit2
      = return (Right ty1)

723 724 725 726
    go ty1 ty2 = bale_out ty1 ty2
      -- We don't handle more complex forms here

    bale_out ty1 ty2 = return $ Left (Pair ty1 ty2)
727 728 729

    tyvar :: SwapFlag -> TcTyVar -> TcType
          -> TcS (Either (Pair TcType) TcType)
730
      -- Try to do as little as possible, as anything we do here is redundant
731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784
      -- with flattening. In particular, no need to zonk kinds. That's why
      -- we don't use the already-defined zonking functions
    tyvar swapped tv ty
      = case tcTyVarDetails tv of
          MetaTv { mtv_ref = ref }
            -> do { cts <- readTcRef ref
                  ; case cts of
                      Flexi        -> give_up
                      Indirect ty' -> unSwap swapped go ty' ty }
          _ -> give_up
      where
        give_up = return $ Left $ unSwap swapped Pair (mkTyVarTy tv) ty

    tyvar_tyvar tv1 tv2
      | tv1 == tv2 = return (Right (mkTyVarTy tv1))
      | otherwise  = do { (ty1', progress1) <- quick_zonk tv1
                        ; (ty2', progress2) <- quick_zonk tv2
                        ; if progress1 || progress2
                          then go ty1' ty2'
                          else return $ Left (Pair (TyVarTy tv1) (TyVarTy tv2)) }

    quick_zonk tv = case tcTyVarDetails tv of
      MetaTv { mtv_ref = ref }
        -> do { cts <- readTcRef ref
              ; case cts of
                  Flexi        -> return (TyVarTy tv, False)
                  Indirect ty' -> return (ty', True) }
      _ -> return (TyVarTy tv, False)

      -- This happens for type families, too. But recall that failure
      -- here just means to try harder, so it's OK if the type function
      -- isn't injective.
    tycon :: TyCon -> [TcType] -> [TcType]
          -> TcS (Either (Pair TcType) TcType)
    tycon tc tys1 tys2
      = do { results <- zipWithM go tys1 tys2
           ; return $ case combine_results results of
               Left tys  -> Left (mkTyConApp tc <$> tys)
               Right tys -> Right (mkTyConApp tc tys) }

    combine_results :: [Either (Pair TcType) TcType]
                    -> Either (Pair [TcType]) [TcType]
    combine_results = bimap (fmap reverse) reverse .
                      foldl' (combine_rev (:)) (Right [])

      -- combine (in reverse) a new result onto an already-combined result
    combine_rev :: (a -> b -> c)
                -> Either (Pair b) b
                -> Either (Pair a) a
                -> Either (Pair c) c
    combine_rev f (Left list) (Left elt) = Left (f <$> elt     <*> list)
    combine_rev f (Left list) (Right ty) = Left (f <$> pure ty <*> list)
    combine_rev f (Right tys) (Left elt) = Left (f <$> elt     <*> pure tys)
    combine_rev f (Right tys) (Right ty) = Right (f ty tys)
785

786
{-
787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804
Note [Newtypes can blow the stack]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Suppose we have

  newtype X = MkX (Int -> X)
  newtype Y = MkY (Int -> Y)

and now wish to prove

  [W] X ~R Y

This Wanted will loop, expanding out the newtypes ever deeper looking
for a solid match or a solid discrepancy. Indeed, there is something
appropriate to this looping, because X and Y *do* have the same representation,
in the limit -- they're both (Fix ((->) Int)). However, no finitely-sized
coercion will ever witness it. This loop won't actually cause GHC to hang,
though, because we check our depth when unwrapping newtypes.

805 806 807 808 809 810 811 812 813
Note [Eager reflexivity check]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Suppose we have

  newtype X = MkX (Int -> X)

and

  [W] X ~R X
814

815 816 817 818 819
Naively, we would start unwrapping X and end up in a loop. Instead,
we do this eager reflexivity check. This is necessary only for representational
equality because the flattener technology deals with the similar case
(recursive type families) for nominal equality.

820 821
Note that this check does not catch all cases, but it will catch the cases
we're most worried about, types like X above that are actually inhabited.
822

eir@cis.upenn.edu's avatar
eir@cis.upenn.edu committed
823
Here's another place where this reflexivity check is key:
824 825 826 827 828 829 830 831 832 833 834 835 836
Consider trying to prove (f a) ~R (f a). The AppTys in there can't
be decomposed, because representational equality isn't congruent with respect
to AppTy. So, when canonicalising the equality above, we get stuck and
would normally produce a CIrredEvCan. However, we really do want to
be able to solve (f a) ~R (f a). So, in the representational case only,
we do a reflexivity check.

(This would be sound in the nominal case, but unnecessary, and I [Richard
E.] am worried that it would slow down the common case.)
-}

------------------------
-- | We're able to unwrap a newtype. Update the bits accordingly.
837
can_eq_newtype_nc :: CtEvidence           -- ^ :: ty1 ~ ty2
838
                  -> SwapFlag
839 840
                  -> TcType                                    -- ^ ty1
                  -> ((Bag GlobalRdrElt, TcCoercion), TcType)  -- ^ :: ty1 ~ ty1'
841 842 843
                  -> TcType               -- ^ ty2
                  -> TcType               -- ^ ty2, with type synonyms
                  -> TcS (StopOrContinue Ct)
844
can_eq_newtype_nc ev swapped ty1 ((gres, co), ty1') ty2 ps_ty2
845
  = do { traceTcS "can_eq_newtype_nc" $
846
         vcat [ ppr ev, ppr swapped, ppr co, ppr gres, ppr ty1', ppr ty2 ]
847 848

         -- check for blowing our stack:
849 850
         -- See Note [Newtypes can blow the stack]
       ; checkReductionDepth (ctEvLoc ev) ty1
851
       ; addUsedGREs (bagToList gres)
852 853 854
           -- we have actually used the newtype constructor here, so
           -- make sure we don't warn about importing it!

855
       ; rewriteEqEvidence ev swapped ty1' ps_ty2
856 857
                           (mkTcSymCo co) (mkTcReflCo Representational ps_ty2)
         `andWhenContinue` \ new_ev ->
858
         can_eq_nc False new_ev ReprEq ty1' ty1' ty2 ps_ty2 }
859

860
---------
eir@cis.upenn.edu's avatar
eir@cis.upenn.edu committed
861
-- ^ Decompose a type application.
862
-- All input types must be flat. See Note [Canonicalising type applications]
eir@cis.upenn.edu's avatar
eir@cis.upenn.edu committed
863 864
can_eq_app :: CtEvidence       -- :: s1 t1 ~r s2 t2
           -> EqRel            -- r
865 866 867
           -> Xi -> Xi         -- s1 t1
           -> Xi -> Xi         -- s2 t2
           -> TcS (StopOrContinue Ct)
eir@cis.upenn.edu's avatar
eir@cis.upenn.edu committed
868 869

-- AppTys only decompose for nominal equality, so this case just leads
eir@cis.upenn.edu's avatar
eir@cis.upenn.edu committed
870
-- to an irreducible constraint; see typecheck/should_compile/T10494
Simon Peyton Jones's avatar
Simon Peyton Jones committed
871
-- See Note [Decomposing equality], note {4}
eir@cis.upenn.edu's avatar
eir@cis.upenn.edu committed
872 873 874
can_eq_app ev ReprEq _ _ _ _
  = do { traceTcS "failing to decompose representational AppTy equality" (ppr ev)
       ; continueWith (CIrredEvCan { cc_ev = ev }) }
eir@cis.upenn.edu's avatar
eir@cis.upenn.edu committed
875 876
          -- no need to call canEqFailure, because that flattens, and the
          -- types involved here are already flat
eir@cis.upenn.edu's avatar
eir@cis.upenn.edu committed
877 878

can_eq_app ev NomEq s1 t1 s2 t2
879
  | CtDerived { ctev_loc = loc } <- ev
880 881 882 883
  = do { unifyDeriveds loc [Nominal, Nominal] [s1, t1] [s2, t2]
       ; stopWith ev "Decomposed [D] AppTy" }
  | CtWanted { ctev_dest = dest, ctev_loc = loc } <- ev
  = do { co_s <- unifyWanted loc Nominal s1 s2
884
       ; co_t <- unifyWanted loc Nominal t1 t2
885 886 887
       ; let co = mkAppCo co_s co_t
       ; setWantedEq dest co
       ; stopWith ev "Decomposed [W] AppTy" }
888 889
  | CtGiven { ctev_evar = evar, ctev_loc = loc } <- ev
  = do { let co   = mkTcCoVarCo evar
890 891
             co_s = mkTcLRCo CLeft  co
             co_t = mkTcLRCo CRight co
892 893 894 895
       ; evar_s <- newGivenEvVar loc ( mkTcEqPredLikeEv ev s1 s2
                                     , EvCoercion co_s )
       ; evar_t <- newGivenEvVar loc ( mkTcEqPredLikeEv ev t1 t2
                                     , EvCoercion co_t )
896 897 898 899
       ; emitWorkNC [evar_t]
       ; canEqNC evar_s NomEq s1 s2 }
  | otherwise  -- Can't happen
  = error "can_eq_app"
900

901 902
-----------------------
-- | Break apart an equality over a casted type
Simon Peyton Jones's avatar
Simon Peyton Jones committed
903
-- looking like   (ty1 |> co1) ~ ty2   (modulo a swap-flag)
904 905 906 907
canEqCast :: Bool         -- are both types flat?
          -> CtEvidence
          -> EqRel
          -> SwapFlag
Simon Peyton Jones's avatar
Simon Peyton Jones committed
908 909
          -> TcType -> Coercion   -- LHS (res. RHS), ty1 |> co1
          -> TcType -> TcType     -- RHS (res. LHS), ty2 both normal and pretty
910 911 912 913 914 915 916 917 918 919 920 921 922 923
          -> TcS (StopOrContinue Ct)
canEqCast flat ev eq_rel swapped ty1 co1 ty2 ps_ty2
  = do { traceTcS "Decomposing cast" (vcat [ ppr ev
                                           , ppr ty1 <+> text "|>" <+> ppr co1
                                           , ppr ps_ty2 ])
       ; rewriteEqEvidence ev swapped ty1 ps_ty2
                           (mkTcReflCo role ty1
                              `mkTcCoherenceRightCo` co1)
                           (mkTcReflCo role ps_ty2)
         `andWhenContinue` \ new_ev ->
         can_eq_nc flat new_ev eq_rel ty1 ty1 ty2 ps_ty2 }
  where
    role = eqRelRole eq_rel

924
------------------------
925 926 927 928
canTyConApp :: CtEvidence -> EqRel
            -> TyCon -> [TcType]
            -> TyCon -> [TcType]
            -> TcS (StopOrContinue Ct)
929
-- See Note [Decomposing TyConApps]
930
canTyConApp ev eq_rel tc1 tys1 tc2 tys2
931
  | tc1 == tc2
932
  , length tys1 == length tys2
933
  = do { inerts <- getTcSInerts
934 935
       ; if can_decompose inerts
         then do { traceTcS "canTyConApp"
936 937 938 939
                       (ppr ev $$ ppr eq_rel $$ ppr tc1 $$ ppr tys1 $$ ppr tys2)
                 ; canDecomposableTyConAppOK ev eq_rel tc1 tys1 tys2
                 ; stopWith ev "Decomposed TyConApp" }
         else canEqFailure ev eq_rel ty1 ty2 }
940

941 942 943 944 945
  -- See Note [Skolem abstract data] (at SkolemAbstract)
  | isSkolemAbstractTyCon tc1 || isSkolemAbstractTyCon tc2
  = do { traceTcS "canTyConApp: skolem abstract" (ppr tc1 $$ ppr tc2)
       ; continueWith (CIrredEvCan { cc_ev = ev }) }

946 947
  -- Fail straight away for better error messages
  -- See Note [Use canEqFailure in canDecomposableTyConApp]
948 949
  | eq_rel == ReprEq && not (isGenerativeTyCon tc1 Representational &&
                             isGenerativeTyCon tc2 Representational)
950 951
  = canEqFailure ev eq_rel ty1 ty2
  | otherwise
952
  = canEqHardFailure ev ty1 ty2
953 954 955 956
  where
    ty1 = mkTyConApp tc1 tys1
    ty2 = mkTyConApp tc2 tys2

957 958 959
    loc  = ctEvLoc ev
    pred = ctEvPred ev

960 961 962 963 964
     -- See Note [Decomposing equality]
    can_decompose inerts
      =  isInjectiveTyCon tc1 (eqRelRole eq_rel)
      || (ctEvFlavour ev /= Given && isEmptyBag (matchableGivens loc pred inerts))

965 966 967 968 969
{-
Note [Use canEqFailure in canDecomposableTyConApp]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
We must use canEqFailure, not canEqHardFailure here, because there is
the possibility of success if working with a representational equality.
eir@cis.upenn.edu's avatar
eir@cis.upenn.edu committed
970
Here is one case:
971 972 973 974 975

  type family TF a where TF Char = Bool
  data family DF a
  newtype instance DF Bool = MkDF Int

eir@cis.upenn.edu's avatar
eir@cis.upenn.edu committed
976
Suppose we are canonicalising (Int ~R DF (TF a)), where we don't yet
977 978
know `a`. This is *not* a hard failure, because we might soon learn
that `a` is, in fact, Char, and then the equality succeeds.
979

eir@cis.upenn.edu's avatar
eir@cis.upenn.edu committed
980 981
Here is another case:

eir@cis.upenn.edu's avatar
eir@cis.upenn.edu committed
982
  [G] Age ~R Int
eir@cis.upenn.edu's avatar
eir@cis.upenn.edu committed
983 984 985 986

where Age's constructor is not in scope. We don't want to report
an "inaccessible code" error in the context of this Given!

eir@cis.upenn.edu's avatar
eir@cis.upenn.edu committed
987 988 989 990 991 992 993 994 995 996
For example, see typecheck/should_compile/T10493, repeated here:

  import Data.Ord (Down)  -- no constructor

  foo :: Coercible (Down Int) Int => Down Int -> Int
  foo = coerce

That should compile, but only because we use canEqFailure and not
canEqHardFailure.

997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035
Note [Decomposing equality]
~~~~~~~~~~~~~~~~~~~~~~~~~~~
If we have a constraint (of any flavour and role) that looks like
T tys1 ~ T tys2, what can we conclude about tys1 and tys2? The answer,
of course, is "it depends". This Note spells it all out.

In this Note, "decomposition" refers to taking the constraint
  [fl] (T tys1 ~X T tys2)
(for some flavour fl and some role X) and replacing it with
  [fls'] (tys1 ~Xs' tys2)
where that notation indicates a list of new constraints, where the
new constraints may have different flavours and different roles.

The key property to consider is injectivity. When decomposing a Given the
decomposition is sound if and only if T is injective in all of its type
arguments. When decomposing a Wanted, the decomposition is sound (assuming the
correct roles in the produced equality constraints), but it may be a guess --
that is, an unforced decision by the constraint solver. Decomposing Wanteds
over injective TyCons does not entail guessing. But sometimes we want to
decompose a Wanted even when the TyCon involved is not injective! (See below.)

So, in broad strokes, we want this rule:

(*) Decompose a constraint (T tys1 ~X T tys2) if and only if T is injective
at role X.

Pursuing the details requires exploring three axes:
* Flavour: Given vs. Derived vs. Wanted
* Role: Nominal vs. Representational
* TyCon species: datatype vs. newtype vs. data family vs. type family vs. type variable

(So a type variable isn't a TyCon, but it's convenient to put the AppTy case
in the same table.)

Right away, we can say that Derived behaves just as Wanted for the purposes
of decomposition. The difference between Derived and Wanted is the handling of
evidence. Since decomposition in these cases isn't a matter of soundness but of
guessing, we want the same behavior regardless of evidence.

Simon Peyton Jones's avatar
Simon Peyton Jones committed
1036 1037 1038 1039 1040 1041
Here is a table (discussion following) detailing where decomposition of
   (T s1 ... sn) ~r (T t1 .. tn)
is allowed.  The first four lines (Data types ... type family) refer
to TyConApps with various TyCons T; the last line is for AppTy, where
there is presumably a type variable at the head, so it's actually
   (s s1 ... sn) ~r (t t1 .. tn)
1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068

NOMINAL               GIVEN                       WANTED

Datatype               YES                         YES
Newtype                YES                         YES
Data family            YES                         YES
Type family            YES, in injective args{1}   YES, in injective args{1}
Type variable          YES                         YES

REPRESENTATIONAL      GIVEN                       WANTED

Datatype               YES                         YES
Newtype                NO{2}                      MAYBE{2}
Data family            NO{3}                      MAYBE{3}
Type family             NO                          NO
Type variable          NO{4}                       NO{4}

{1}: Type families can be injective in some, but not all, of their arguments,
so we want to do partial decomposition. This is quite different than the way
other decomposition is done, where the decomposed equalities replace the original
one. We thus proceed much like we do with superclasses: emitting new Givens
when "decomposing" a partially-injective type family Given and new Deriveds
when "decomposing" a partially-injective type family Wanted. (As of the time of
writing, 13 June 2015, the implementation of injective type families has not
been merged, but it should be soon. Please delete this parenthetical if the
implementation is indeed merged.)

Simon Peyton Jones's avatar
Simon Peyton Jones committed
1069
{2}: See Note [Decomposing newtypes at representational role]
1070

Simon Peyton Jones's avatar
Simon Peyton Jones committed
1071 1072 1073
{3}: Because of the possibility of newtype instances, we must treat
data families like newtypes. See also Note [Decomposing newtypes at
representational role]. See #10534 and test case
1074
typecheck/should_fail/T10534.
1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085

{4}: Because type variables can stand in for newtypes, we conservatively do not
decompose AppTys over representational equality.

In the implementation of can_eq_nc and friends, we don't directly pattern
match using lines like in the tables above, as those tables don't cover
all cases (what about PrimTyCon? tuples?). Instead we just ask about injectivity,
boiling the tables above down to rule (*). The exceptions to rule (*) are for
injective type families, which are handled separately from other decompositions,
and the MAYBE entries above.

Simon Peyton Jones's avatar
Simon Peyton Jones committed
1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117
Note [Decomposing newtypes at representational role]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
This note discusses the 'newtype' line in the REPRESENTATIONAL table
in Note [Decomposing equality]. (At nominal role, newtypes are fully
decomposable.)

Here is a representative example of why representational equality over
newtypes is tricky:

  newtype Nt a = Mk Bool         -- NB: a is not used in the RHS,
  type role Nt representational  -- but the user gives it an R role anyway

If we have [W] Nt alpha ~R Nt beta, we *don't* want to decompose to
[W] alpha ~R beta, because it's possible that alpha and beta aren't
representationally equal. Here's another example.

  newtype Nt a = MkNt (Id a)
  type family Id a where Id a = a

  [W] Nt Int ~R Nt Age

Because of its use of a type family, Nt's parameter will get inferred to have
a nominal role. Thus, decomposing the wanted will yield [W] Int ~N Age, which
is unsatisfiable. Unwrapping, though, leads to a solution.

Conclusion:
 * Unwrap newtypes before attempting to decompose them.
   This is done in can_eq_nc'.

It all comes from the fact that newtypes aren't necessarily injective
w.r.t. representational equality.

1118
Furthermore, as explained in Note [NthCo and newtypes] in TyCoRep, we can't use
Simon Peyton Jones's avatar
Simon Peyton Jones committed
1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145
NthCo on representational coercions over newtypes. NthCo comes into play
only when decomposing givens.

Conclusion:
 * Do not decompose [G] N s ~R N t

Is it sensible to decompose *Wanted* constraints over newtypes?  Yes!
It's the only way we could ever prove (IO Int ~R IO Age), recalling
that IO is a newtype.

However we must be careful.  Consider

  type role Nt representational

  [G] Nt a ~R Nt b       (1)
  [W] NT alpha ~R Nt b   (2)
  [W] alpha ~ a          (3)

If we focus on (3) first, we'll substitute in (2), and now it's
identical to the given (1), so we succeed.  But if we focus on (2)
first, and decompose it, we'll get (alpha ~R b), which is not soluble.
This is exactly like the question of overlapping Givens for class
constraints: see Note [Instance and Given overlap] in TcInteract.

Conclusion:
  * Decompose [W] N s ~R N t  iff there no given constraint that could
    later solve it.
1146 1147 1148
-}

canDecomposableTyConAppOK :: CtEvidence -> EqRel
1149
                          -> TyCon -> [TcType] -> [TcType]
1150 1151
                          -> TcS ()
-- Precondition: tys1 and tys2 are the same length, hence "OK"
1152
canDecomposableTyConAppOK ev eq_rel tc tys1 tys2
1153
  = case ev of
1154
     CtDerived {}
1155
        -> unifyDeriveds loc tc_roles tys1 tys2
1156

1157 1158 1159
     CtWanted { ctev_dest = dest }
        -> do { cos <- zipWith4M unifyWanted new_locs tc_roles tys1 tys2
              ; setWantedEq dest (mkTyConAppCo role tc cos) }
1160

1161 1162
     CtGiven { ctev_evar = evar }
        -> do { let ev_co = mkCoVarCo evar
1163
              ; given_evs <- newGivenEvVars loc $
1164 1165
                             [ ( mkPrimEqPredRole r ty1 ty2
                               , EvCoercion (mkNthCo i ev_co) )
1166
                             | (r, ty1, ty2, i) <- zip4 tc_roles tys1 tys2 [0..]
1167 1168
                             , r /= Phantom
                             , not (isCoercionTy ty1) && not (isCoercionTy ty2) ]
1169 1170
              ; emitWorkNC given_evs }
  where
1171 1172 1173 1174 1175 1176
    loc        = ctEvLoc ev
    role       = eqRelRole eq_rel
    tc_roles   = tyConRolesX role tc

      -- the following makes a better distinction between "kind" and "type"
      -- in error messages
1177
    bndrs      = tyConBinders tc
1178
    kind_loc   = toKindLoc loc
1179
    is_kinds   = map isNamedTyConBinder bndrs
1180 1181 1182 1183 1184
    new_locs | Just KindLevel <- ctLocTypeOrKind_maybe loc
             = repeat loc
             | otherwise
             = map (\is_kind -> if is_kind then kind_loc else loc) is_kinds

1185 1186 1187

-- | Call when canonicalizing an equality fails, but if the equality is
-- representational, there is some hope for the future.
1188
-- Examples in Note [Use canEqFailure in canDecomposableTyConApp]
1189 1190
canEqFailure :: CtEvidence -> EqRel
             -> TcType -> TcType -> TcS (StopOrContinue Ct)
1191
canEqFailure ev NomEq ty1 ty2
1192
  = canEqHardFailure ev ty1 ty2
1193
canEqFailure ev ReprEq ty1 ty2
1194
  = do { (xi1, co1) <- flatten FM_FlattenAll ev ty1
1195
       ; (xi2, co2) <- flatten FM_FlattenAll ev ty2
1196 1197 1198
            -- We must flatten the types before putting them in the
            -- inert set, so that we are sure to kick them out when
            -- new equalities become available
1199 1200
       ; traceTcS "canEqFailure with ReprEq" $
         vcat [ ppr ev, ppr ty1, ppr ty2, ppr xi1, ppr xi2 ]
1201
       ; rewriteEqEvidence ev NotSwapped xi1 xi2 co1 co2
eir@cis.upenn.edu's avatar
eir@cis.upenn.edu committed
1202 1203
         `andWhenContinue` \ new_ev ->
         continueWith (CIrredEvCan { cc_ev = new_ev }) }
1204 1205

-- | Call when canonicalizing an equality fails with utterly no hope.
1206
canEqHardFailure :: CtEvidence
1207
                 -> TcType -> TcType -> TcS (StopOrContinue Ct)
1208
-- See Note [Make sure that insolubles are fully rewritten]
1209
canEqHardFailure ev ty1 ty2
1210 1211
  = do { (s1, co1) <- flatten FM_SubstOnly ev ty1
       ; (s2, co2) <- flatten FM_SubstOnly ev ty2
1212
       ; rewriteEqEvidence ev NotSwapped s1 s2 co1 co2
1213 1214 1215
         `andWhenContinue` \ new_ev ->
    do { emitInsoluble (mkNonCanonical new_ev)
       ; stopWith new_ev "Definitely not equal" }}
1216

Austin Seipp's avatar
Austin Seipp committed
1217
{-
1218 1219 1220 1221 1222 1223
Note [Decomposing TyConApps]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
If we see (T s1 t1 ~ T s2 t2), then we can just decompose to
  (s1 ~ s2, t1 ~ t2)
and push those back into the work list.  But if
  s1 = K k1    s2 = K k2
Jan Stolarek's avatar
Jan Stolarek committed
1224
then we will just decomopose s1~s2, and it might be better to
1225 1226 1227 1228 1229 1230
do so on the spot.  An important special case is where s1=s2,
and we get just Refl.

So canDecomposableTyCon is a fast-path decomposition that uses
unifyWanted etc to short-cut that work.

1231 1232 1233
Note [Canonicalising type applications]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Given (s1 t1) ~ ty2, how should we proceed?
Austin Seipp's avatar
Austin Seipp committed
1234
The simple things is to see if ty2 is of form (s2 t2), and
1235
decompose.  By this time s1 and s2 can't be saturated type
Austin Seipp's avatar
Austin Seipp committed
1236 1237
function applications, because those have been dealt with
by an earlier equation in can_eq_nc, so it is always sound to
1238 1239
decompose.