#5237: Inefficient code generated for x^2
--------------------------------------+-------------------------------------
Reporter: scpmw | Owner:
Type: bug | Status: new
Priority: normal | Milestone: 7.4.1
Component: libraries/base | Version: 7.0.3
Resolution: | Keywords:
Testcase: | Blockedby:
Difficulty: | Os: Linux
Blocking: | Architecture: x86_64 (amd64)
Failure: Runtime performance bug |
--------------------------------------+-------------------------------------
Comment(by daniel.is.fischer):
Yes. And with
{{{
{-# RULES "^^^2/Both" forall (x :: Integer). x ^^^ (2 :: Integer) = x
#-}
}}}
alongside the two others, that fires.
Re sidenote: yes, it should be for safety. That occurred to me today too.
I've done a few tests and so far GHC always was clever enough to share the
expression, but it's better not to rely on it.
With regard to the rule/specialisation competition, what if the `{-#
SPECIALISE #-}` pragma is removed? Due to the `{-# INLINABLE #-}`, with
optimisations, specialisations will (probably) be created at the call
sites, won't they? So then we wouldn't need the extra rules. However, I
think that the specialisations would be created in each module and not
once per programme/package, and that would mean code-bloat. So what would
be worse, more rules or the code-bloat?
--
Ticket URL: <http://hackage.haskell.org/trac/ghc/ticket/5237#comment:11>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler
_______________________________________________
Glasgow-haskell-bugs mailing list
Glasgow-haskell-bugs@haskell.org
http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs
