On 4/11/12 10:30 PM, Jonathan S. Shapiro wrote:
> On Wed, Apr 11, 2012 at 3:24 PM, wren ng thornton<[email protected]> 
wrote:
>
>> Nope. For instance, the places where GC can happen are fully defined in
>> the Glasgow Haskell Compiler....
>
> This, of course, is insufficient for multithreaded programs.

How's that now? GHC has extremely good multithreaded support, and if all
active threads are in non-allocating loops then what I said holds...


>> ...you can write tight loops which are strict
>> enough to guarantee no allocation and therefore no GC.
>
> My sense, as yet undemonstrated, is that it may be a bit more scalable in
> something like BitC, where unboxed types are available.

Surely. There are unboxed types in GHC though they're not first-class, as
they would be in BitC. I wasn't claiming that Haskell solves this problem,
far from it; I was merely pointing to GHC as a constructive proof that
there are shipping garbage collectors which are not unpredictable in the
way claimed by whoever I was replying to.

-- 
Live well,
~wren

_______________________________________________
bitc-dev mailing list
[email protected]
http://www.coyotos.org/mailman/listinfo/bitc-dev

Reply via email to