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

[Simon Marlow]

Ian Lynagh wrote:
> Thu May  3 15:38:33 PDT 2007  Ian Lynagh <igloo at earth.li>
>   * Fix truncate on amd64 NCG; fixes arith005.
>   cvts[sd]2siq? ->
>   cvtts[sd]2siq?
> 
>     M ./compiler/nativeGen/MachCodeGen.hs -1 +3
>     M ./compiler/nativeGen/MachInstrs.hs -2 +2
>     M ./compiler/nativeGen/PprMach.hs -6 +6
>     M ./compiler/nativeGen/RegAllocInfo.hs -4 +4

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.  Looks 
like you fixed the primop to have truncating semantics for the x86_64 NCG, but 
it looks it still has rounding semantics for other platforms: the x86 NCG uses 
fistpl, which rounds by default, and when going via C we use a cast, which also 
rounds.

Who is right?  I'd have thought the primop should round by default, but then the 
RULE is wrong.

Cheers,
	Simon