[Haskell-cafe] Support of multi-constructor types by UNPACK pragma

[Carter Schonwald]

Woops. I misread the original email then!

Yeah, unpacking sums could have an exponential blowup right?

On Sunday, June 15, 2014, Johan Tibell <johan.tibell at gmail.com> wrote:

> GHC doesn't unpack sums, like B in the example.
>
>
> On Sun, Jun 15, 2014 at 10:30 PM, Carter Schonwald <
> carter.schonwald at gmail.com
> <javascript:_e(%7B%7D,'cvml','carter.schonwald at gmail.com');>> wrote:
>
>> Are you sure that's not supported? That seems like a bug if so.  I'm away
>> from my computer right now so I can't check.
>>
>> What ghc version? How are you checking if the unpack is firing?
>>
>> Ghc 7.8 should by default be unpacking ALL strict small fields.
>>
>>
>> On Sunday, June 15, 2014, Nikita Volkov <nikita.y.volkov at mail.ru
>> <javascript:_e(%7B%7D,'cvml','nikita.y.volkov at mail.ru');>> wrote:
>>
>>> Currently the pragma only supports single-constructor types. So in the
>>> following example it will simply be ignored:
>>>
>>> data A =
>>>   A1 Char |
>>>   A2 {-# UNPACK #-} !B
>>>
>>> data B =
>>>   B1 Int |
>>>   B2 Bool
>>>
>>> However it seems to be easily solvable by changing the type A to
>>> something like the following during unpacking:
>>>
>>> data A =
>>>   A1 Char |
>>>   A2_1 Int |    -- from B1
>>>   A2_2 Bool     -- from B2
>>>
>>> Am I missing something? Why is it not implemented?
>>> ​
>>>
>>>
>> _______________________________________________
>> Haskell-Cafe mailing list
>> Haskell-Cafe at haskell.org
>> <javascript:_e(%7B%7D,'cvml','Haskell-Cafe at haskell.org');>
>> http://www.haskell.org/mailman/listinfo/haskell-cafe
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/haskell-cafe/attachments/20140615/665783fa/attachment.html>