[GHC] #11735: Optimize coercionKind

[GHC]

#11735: Optimize coercionKind
-------------------------------------+-------------------------------------
        Reporter:  goldfire          |                Owner:  (none)
            Type:  task              |               Status:  new
        Priority:  normal            |            Milestone:
       Component:  Compiler          |              Version:  7.10.3
      Resolution:                    |             Keywords:
Operating System:  Unknown/Multiple  |         Architecture:
 Type of failure:  Compile-time      |  Unknown/Multiple
  performance bug                    |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:  #14683 #14975     |  Differential Rev(s):  D4394 D4395
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by tdammers):

 Profile after applying D4395 (coercionKind / Role refactoring):

 {{{
         Mon Mar 26 15:24 2018 Time and Allocation Profiling Report
 (Final)

            ghc-stage2 +RTS -p -RTS -B/home/tobias/well-
 typed/devel/ghc/D4395-modified/inplace/lib ./cases/Grammar.hs -o ./a
 -fforce-recomp

         total time  =       19.66 secs   (19655 ticks @ 1000 us, 1
 processor)
         total alloc = 24,638,084,488 bytes  (excludes profiling overheads)

 COST CENTRE     MODULE     SRC
 %time %alloc

 simplCast       Simplify
 compiler/simplCore/Simplify.hs:(1213,5)-(1215,37)    73.8   76.0
 tc_rn_src_decls TcRnDriver
 compiler/typecheck/TcRnDriver.hs:(494,4)-(556,7)      8.3    8.4
 CoreTidy        HscMain    compiler/main/HscMain.hs:1253:27-67
 2.8    2.2
 SimplTopBinds   SimplCore  compiler/simplCore/SimplCore.hs:770:39-74
 2.7    1.9
 coercionKind    Coercion   compiler/types/Coercion.hs:1698:3-7
 1.8    3.4
 zonkTopDecls    TcRnDriver
 compiler/typecheck/TcRnDriver.hs:(445,16)-(446,43)    1.3    1.3
 deSugar         HscMain    compiler/main/HscMain.hs:511:7-44
 1.1    0.8
 }}}

 This means that performance is on par, but `simplCast` still scores very
 high, so we should do more digging there.

-- 
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11735#comment:82>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler