#2132: Optimise nested comparisons
---------------------------------+------------------------------------------
Reporter: simonpj | Owner: simonpj
Type: bug | Status: patch
Priority: low | Milestone: 7.4.1
Component: Compiler | Version: 6.8.2
Keywords: | Testcase:
Blockedby: | Difficulty: Unknown
Os: Unknown/Multiple | Blocking:
Architecture: Unknown/Multiple | Failure: Runtime performance bug
---------------------------------+------------------------------------------
Comment(by basvandijk):
Replying to [comment:18 michalt]:
> Note that this optimization goes a bit further than your example. For
> instance, once it knows `x ># y` and `y >=# z` it can optimize away `x
<# z`.
This is possible with sub-rules:
{{{
{-# RULES
forall x y t f.
(case (x ># y) of
True -> t
False -> f
)
=
(case (x ># y) of
True -> t {{ forall z t2 f2.
(case (y >=# z) of
True -> t2
False -> f2
)
=
(case (y >=# z) of
True -> t2 {{(x <# z)=False}}
False -> f2
)
}}
False -> f
)
#-}
}}}
> Similarly knowing `x ># 100` and `y <# 10` it can get rid of `x ==# y`.
> That would be probably quite difficult to do with rules (even extended).
Encoding it for the concrete `100` and `10` is easy. However encoding it
for all numbers is not possible without having the ability to reason about
numbers in the language of RULES.
I'm sure your patch is way more powerful since I expect it will have this
ability. I do wonder if sub-rules have other useful applications?
--
Ticket URL: <http://hackage.haskell.org/trac/ghc/ticket/2132#comment:19>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler
_______________________________________________
Glasgow-haskell-bugs mailing list
[email protected]
http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs
