[Haskell-cafe] Re: MD5 performance optimizations

[Andrew Coppin]

Bulat Ziganshin wrote:
> Hello Andrew,
>
> Tuesday, May 20, 2008, 11:05:52 PM, you wrote:
>
>   
>>> -funbox-strict-fields.
>>>   
>>>       
>
>   
>> I did try that, but it didn't seem to make any difference for me. [Maybe
>>     
>
> it may be that ghc just not recompiled program when you supplied this
> switch.
>   

Possibly. I notice that adding *some* switches makes GHC recompile 
everything, while others don't. [It's probably documented in the manual 
somewhere...]

> and, without this switch representation for !Int32 is the same as for
> Int32 - only difference is that when data is assigned to such field
> they are evaluated first (and then boxed)
>   

Right, OK - so strict evaluation, but some bitpattern in RAM.

> 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...