[GHC] #15646: ghci takes super long time to find the type of large fractional number

[GHC]

#15646: ghci takes super long time to find the type of large fractional number
-------------------------------------+-------------------------------------
        Reporter:  Johannkokos       |                Owner:
                                     |  JulianLeviston
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:  8.6.1
       Component:  GHCi              |              Version:  8.4.3
      Resolution:                    |             Keywords:  newcomer
Operating System:  Unknown/Multiple  |         Architecture:
 Type of failure:  Compile-time      |  Unknown/Multiple
  performance bug                    |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:  #5692             |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by osa1):

 > I can't print out litE and litI because they don't have a show instance
 for Core

 In GHC you should be using `Outputable` instances (with `pprTrace` and
 frieds).
 Most GHC types don't have a `Show` instance, but most have `Outputable`.
 In this
 case `pprTrace "..." (text "litI:" <+> ppr litI <+> "litE:" <+> ppr litE)`
 should work.

 > So... yeah, this is what happens when I run it:

 Right, so the problem is in your `mkRational` name definition:

 {{{
 mkRationalName      = tcQual  gHC_REAL (fsLit "mkRational")
 mkRationalIdKey
 }}}

 You're saying that `mkRational` lives in `GHC.Real`, but that's not the
 case. It
 actually lives in the GHC module `BasicTypes`.

 We need to put `mkRational` to somehwere in `base`. `GHC.Real` is a good
 place I
 think.

 Secondly, I see some uses of `mkRational` in `BasicTypes`. Those also need
 to be
 moved, because when booting GHC, when building stage 1, we'll be using a
 GHC
 that won't have `mkRational` in `GHC.Real`. So all those definitions that
 use
 `mkRational` need to be moved to `GHC.Real` too.

 Does this make sense?

-- 
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15646#comment:60>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler