Wow ... thanks Serguey .... that gets rid of DatatypeContexts as well!

Regards,
Kashyap


On Mon, Apr 1, 2013 at 9:12 PM, Serguey Zefirov <sergu...@gmail.com> wrote:

> You have fixed the type of list by move RAX RAX. Now it has type
> Instruction SNDREG SNDREG
>
> Make your Instruction a GADT and require that MOV should have appropriate
> constraints:
>
> {-# LANGUAGE DatatypeContexts, GADTs #-}
>
>
> data SREG = RIP
> data DREG = RBX
> data SNDREG = RAX
>
>
> data Instruction where
>         MOV :: (Source s, Destination d) => s -> d -> Instruction
>
>
>
> class Source a
> class Destination a
>
> instance Source SREG
> instance Source SNDREG
>
> instance Destination DREG
> instance Destination SNDREG
>
>
> move :: (Source s, Destination d) => s -> d -> Instruction
> move s d = MOV s d
>
> hello = [move RAX RAX, move RAX RAX]
>
> hello2 = [move RAX RAX, move RAX RBX] -- this is still not allowed.
>
>
>
>
> 2013/4/1 C K Kashyap <ckkash...@gmail.com>
>
>> Hi Cafe,
>> I am trying to embed x86 assembly in Haskell. I'd like the EDSL to not
>> allow invalid movements into registers - for example, should not allow
>> moving into RIP. I was not able to get it to work. I ended up using
>> DataTypeContexts - which is considered misfeature anyway. I was wondering
>> if I could get some suggestions.
>>
>> {-# LANGUAGE DatatypeContexts #-}
>>
>> data SREG = RIP
>> data DREG = RBX
>> data SNDREG = RAX
>>
>>
>> data (Source s, Destination d) => Instruction s d = MOV s d
>>
>>
>> class Source a
>> class Destination a
>>
>> instance Source SREG
>> instance Source SNDREG
>>
>> instance Destination DREG
>> instance Destination SNDREG
>>
>>
>> move :: (Source s, Destination d) => s -> d -> Instruction s d
>> move s d = MOV s d
>>
>> hello = [move RAX RAX, move RAX RAX]
>>
>> hello = [move RAX RAX, move RAX RBX] -- this is still not allowed.
>>
>> Regards,
>> Kashyap
>>
>>
>> _______________________________________________
>> Haskell-Cafe mailing list
>> Haskell-Cafe@haskell.org
>> http://www.haskell.org/mailman/listinfo/haskell-cafe
>>
>>
>
_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe

Reply via email to