More precise GC

[Robert Jacques]

On Mon, 29 Mar 2010 08:09:09 -0300, Steven Schveighoffer  
<schveiguy at yahoo.com> wrote:

> On Sun, 28 Mar 2010 23:30:32 -0400, Robert Jacques <sandford at jhu.edu>  
> wrote:
>
>> On Sun, 28 Mar 2010 16:16:41 -0300, Steven Schveighoffer  
>> <schveiguy at yahoo.com> wrote:
>>> The current GC has a simple "type info" if you will -- contains  
>>> pointers or doesn't contain pointers.  It doesn't mean we cannot add  
>>> to that.  In fact, I think dsimcha has provided a way to have precise  
>>> scanning for heap-allocated types.  I don't think a reasonably precise  
>>> GC is out of the question.  However, it may be too much to require the  
>>> GC to do semantic analysis of enums for unions.  Not impossible, but  
>>> probably not worth the effort and restrictions necessary.
>>>
>>
>> Also, don't forget that classes have a bunch of runtime type info.
>
> But the GC doesn't use/need this info (except to call the destructors).   
> At least the mark/sweep portion doesn't.
>
> -Steve

Sorry, my comment was more for a D in general than the GC. GC's in general  
don't know or need anything beyond a pointer mask and whether to finalize  
or not.