[GHC] #2664: type family + data family + typeclass + type error causes GHC to diverge

[GHC]

#2664: type family + data family + typeclass + type error causes GHC to diverge
---------------------------------+------------------------------------------
    Reporter:  ryani             |        Owner:  chak            
        Type:  bug               |       Status:  new             
    Priority:  normal            |    Milestone:  6.12 branch     
   Component:  Compiler          |      Version:  6.11            
    Severity:  normal            |   Resolution:                  
    Keywords:                    |   Difficulty:  Unknown         
    Testcase:                    |           Os:  Unknown/Multiple
Architecture:  Unknown/Multiple  |  
---------------------------------+------------------------------------------
Comment (by chak):

 Replying to [comment:6 simonpj]:
 > HEAD gives
 {{{
 WARNING: file compiler/types/Coercion.lhs line 465
 Strange! Type mismatch in trans coercion, probably a bug
 ...plus some more...
 }}}
 > But Manuel thinks the WARNING has false positives. (Maybe we should nuke
 it.)

 One case of false positives is when one of the compared types contains an
 unzonked type variable.  I believe this situation arises during solving
 quality constraints.  In most other places where trans coercions are made,
 the test may be accurate.  So, if we knew the caller, it would be easier
 to say whether a particular warning may potentially be a false positive.

-- 
Ticket URL: <http://hackage.haskell.org/trac/ghc/ticket/2664#comment:7>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler