patch applied (ghc): Fix truncate on amd64 NCG; fixes arith005.

[Brian Alliet]

On Mon, May 07, 2007 at 08:58:33AM +0100, Simon Marlow wrote:
> Something still looks odd here.  We have a RULE in GHC/Float.hs mapping 
> truncate on Float->Int to the float2Int# primop (similarly for Double), 
> which means that either float2Int# should have truncating semantics, or the 
> RULE is wrong.  

There are also an issue with overflow with float2Int# on powerpc.

http://hackage.haskell.org/trac/ghc/ticket/1254

Probably makes sense to look into this at the same time.

-Brian