std.allocator: false pointers

[via Digitalmars-d]

On Saturday, 3 May 2014 at 23:41:02 UTC, safety0ff wrote:
>> Well, in a 64-bit address space, the false pointer issue is 
>> almost
>> mute, the issue comes in when you try to apply this design to 
>> 32-bit,

That assumes that the heap is located at a high address.

> I think false pointers must be addressed to make claims that D 
> is well-behaved on 32-bit systems.

Which means a precise collector, which means faster stack meta 
info access, which means faster exception handling? Then use a 
single pass mark-and-don't-sweep collector/allocator? (invert the 
meaning of the bit for every other full collection cycle)

A conservative collector will never be acceptable in a language 
which is primarily GC based. It is something you can bolt on to a 
RC system to collrct cycles, IMHO.