[Haskell-cafe] Re: MD5 performance optimizations

Don Stewart dons at galois.com
Tue May 20 17:02:49 EDT 2008


bulat.ziganshin:
> Hello Andrew,
> 
> Wednesday, May 21, 2008, 12:09:39 AM, you wrote:
> 
> > Right, OK - so strict evaluation, but some bitpattern in RAM.
> 
> yes. in particular this allows to designate fields that should de
> initialized:
> 
> data A = A {a :: Int, b :: !Int}
> main = print A{}
> 
> >> it is not enabled by default, because for *non-primitive* datatypes
> >> such as B below automatic unboxing of strict fields of this type may
> >> decrease sharing and thus memory/performance.
> 
> > Yeah, I can imagine...
> 
> sometime i've seriously optimized my program by removing this switch
> and adding manual unboxing directives to the sources. it will be great
> if ghc will catch jhc here

Would you like to file a feature request?

    http://hackage.haskell.org/trac/ghc/newticket?type=bug

Particularly if you have an example where JHC has done a better job
optimising some simple example. So far I've seen no ticke

Unless tickets are opened, progress will not be made.

-- Don


More information about the Haskell-Cafe mailing list