[GHC] #12603: INLINE and manually inlining produce different code

GHC ghc-devs at haskell.org
Mon Dec 5 12:35:27 UTC 2016


#12603: INLINE and manually inlining produce different code
-------------------------------------+-------------------------------------
        Reporter:  bgamari           |                Owner:  bgamari
            Type:  bug               |               Status:  new
        Priority:  high              |            Milestone:  8.2.1
       Component:  Compiler          |              Version:  8.0.1
      Resolution:                    |             Keywords:
Operating System:  Unknown/Multiple  |         Architecture:
 Type of failure:  Runtime           |  Unknown/Multiple
  performance bug                    |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:  #12747 #12781     |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by simonpj):

 It's all to do with ''interactions'' between multiple INLINEs.  If we did
 as you say, and inlined earlier, then we would ''also'' inline functions
 like `(+)` which also have INLINE pragmas on them.   So we might get
 {{{
   attrFromIntINLINE w = case w of I# w' ->
                         case 2# ^# 8# of lvl' ->
                         Attr (w' +# lvl')
 }}}
 which, as I say, is currently unfloatable.  You may say that the INLINE on
 `(+)` should be delayed and that would be one solution. But perhaps a
 better one would be to make floating more robust.

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12603#comment:28>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler


More information about the ghc-tickets mailing list