While i think we can get somewhere by breaking things into service and
using appropriate allocations in each service ( like ref counting or static
allocation )  this is mainly for other purposes like reuse/ resiliance .  I
really don't see what the issue is with having 10-15 % of a runtime code
base in C++  or asm   ( after all we do this with SSE  anyway )  ?   The
cost of  that 15% is probably much lower than trying to make a GC code base
work and even today many C based OS have asm chunks anyway

On Thu, Apr 12, 2012 at 8:57 AM, David Jeske <[email protected]> wrote:

> On Tue, Apr 10, 2012 at 12:26 PM, Jonathan S. Shapiro <[email protected]>wrote:
>
>> It's also worth noting that *malloc* can be expensive!
>>>
>>
> ...even an expensive malloc does not stop all threads.
>
> I'm sad to see folks still continuing this discussion of how such-and-such
> pause is okay in so-and-so situation. This might be okay for a
> whole-program Haskell compile, however, this doesn't fly in an industry
> wide modular systems runtime.
>
_______________________________________________
bitc-dev mailing list
[email protected]
http://www.coyotos.org/mailman/listinfo/bitc-dev

Reply via email to