#5068: Possible loss of sharing: big performance change
---------------------------------+------------------------------------------
Reporter: simonpj | Owner:
Type: bug | Status: new
Priority: normal | Milestone: 7.4.1
Component: Compiler | Version: 7.0.3
Keywords: | Testcase:
Blockedby: | Difficulty:
Os: Unknown/Multiple | Blocking:
Architecture: Unknown/Multiple | Failure: None/Unknown
---------------------------------+------------------------------------------
Comment(by bos):
Replying to [comment:1 simonmar]:
> If you say `INLINE` on a redex, GHC will happily do as you ask, won't
it? I took a quick look at the diff that Bryan pointed to, and it looks
like the problem is `ratios` being inlined in `loop`, which is a monadic
computation and therefore probably a lambda.
I agree. Although I didn't think to check when I first wrote the code,
when it occurred to me that there might be a performance problem, the
computation of `ratios` was the very first thing I thought to look at. In
other words, GHC's behaviour here was maybe a little unfortunate, but it
wasn't unexpected (admittedly by a fairly sophisticated user).
I can submit a test case if that would be useful.
--
Ticket URL: <http://hackage.haskell.org/trac/ghc/ticket/5068#comment:3>
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
