[GHC] #11060: Failing tests on 32 bit platforms

[GHC]

#11060: Failing tests on 32 bit platforms
-------------------------------------+-------------------------------------
        Reporter:  erikd             |                Owner:
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:  8.0.1
       Component:  Test Suite        |              Version:  7.11
      Resolution:                    |             Keywords:
Operating System:  Unknown/Multiple  |         Architecture:
                                     |  Unknown/Multiple
 Type of failure:  Other             |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:                    |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by bgamari):

 The relevant part of the expected output of this test prior to being
 mangled by the testsuite driver is,

 {{{#!hs
 T8274.$tcN :: TyCon
 [GblId[ReflectionId],
  Caf=NoCafRefs,
  Str=DmdType m,
  Unf=Unf{Src=<vanilla>, TopLvl=True, Value=True, ConLike=True,
          WorkFree=True, Expandable=True, Guidance=IF_ARGS [] 10 50}]
 T8274.$tcN =
   TyCon
     7479687563082171902##
     17616649989360543185##
     T8274.$trModule

 }}}

 So, it appears that the difference here arises from the fact that
 something formatting a `Data.Typeable.Internal.TyCon` is including an
 extra `L` while formatting the `Word64#` fingerprints on 32-bit machines.
 Given that this isn't valid Haskell on 32- nor 64-bit machines, this
 should likely be fixed. However, I can't for the life of me figure out
 where this is happening.

 It would appear that the `Show TyCon` instance is defined in
 `libraries,base/GHC/Show.hs`,
 {{{#!hs
 instance Show TyCon where
   showsPrec p (TyCon _ _ _ tc_name) = showsPrec p tc_name
 }}}
 which clearly doesn't match the output we are seeing here. I've tried
 looking for `Outputable` instances but the only one I've been able to find
 is for `typecheck/TyCon.hs`.

 Perhaps this is being derived by `TcGenDeriv` but I've not seen any logic
 in there that might be adding this errant `L`. It's all quite perplexing.

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11060#comment:1>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler