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

28 Nov 2018

      #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 JulianLeviston):

 So I tried to print out the values you suggested. I can't print out `litE`
 and `litI` because they don't have a show instance for Core, but I tried
 to print out the `Integer` values `fl_signi` and `fl_exp` that I use to
 build those Core expressions with:

 {{{
     import IOEnv (liftIO)
     ...
     HsRat _ fl _     -> case fl of
                           FL { fl_signi = fl_signi, fl_exp = fl_exp } ->
 do
                             mkRational <- dsLookupGlobalId mkRationalName
                             litI <- mkIntegerExpr fl_signi
                             litE <- mkIntegerExpr fl_exp
                             -- temporary logging
                             liftIO $ putStrLn $ "fl_signi: " ++ show
 fl_signi
                             liftIO $ putStrLn $ "fl_exp: " ++ show fl_exp
                             return ((Var mkRational) `App` litI `App`
 litE)
 }}}

 That compiles, but unfortunately when I try to run it, it blows up with
 the same interface file stuff. I guess this is because I renumbered
 `RubbishLit` here: https://github.com/JulianLeviston/ghc/pull/1/files
 #diff-78b3c572cf078eb00f09974b18f3eedfR214 — I'm really interested in
 suggestions about a better way to do that. It didn't seem good to put it
 after (it's out of order then). But I'm not sure what changes I need to
 make if I'm going to adjust the interface file(s) as I seem to have done
 by doing that? (bit lost there).

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

 {{{
 itsy:ghc julianleviston$ ./inplace/bin/ghc-stage2 -e ":t 1e302"
 1e302 :: Fractional p => p
 itsy:ghc julianleviston$ ./inplace/bin/ghc-stage2 -e "1e302"
 GHC error in desugarer lookup in Ghci1:
   Can't find interface-file declaration for type constructor or class
 mkRational
     Probable cause: bug in .hi-boot file, or inconsistent .hi file
     Use -ddump-if-trace to get an idea of which file caused the error
 }}}

-- 
Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/15646#comment:59
GHC http://www.haskell.org/ghc/
The Glasgow Haskell Compiler