[GHC] #9630: compile-time performance regression (probably due to Generics)

[GHC]

#9630: compile-time performance regression (probably due to Generics)
-------------------------------------+-------------------------------------
        Reporter:  hvr               |                Owner:  simonpj
            Type:  bug               |               Status:  patch
        Priority:  high              |            Milestone:  8.0.1
       Component:  Compiler          |              Version:  7.9
      Resolution:                    |             Keywords:  deriving-perf
Operating System:  Unknown/Multiple  |         Architecture:
 Type of failure:  Compile-time      |  Unknown/Multiple
  performance bug                    |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:  #9583, #10293     |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by simonpj):

 Splitting `binary` as done in comment:23 will improve matters.  As you say
 it doesn't fix GHC; indeed it'll make the problem less egregious.

 Just to check, then: does comment:22 give a standalone case, independent
 of the `binary` class, that demonstrates the effect of splitting the
 `Binary` class into two?

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9630#comment:26>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler