More embarrassing microbenchmars for D's GC.

[Marius Muja]

Leandro Lucarella wrote:
> 
> Considering that the version with GC disabled is almost as fast as the
> malloc version, and that the time grows exponentially (graph attached)
> with the number of nodes, it looks like the problem is the GC
> collection cycles to me. Is the GC algorithm exponential?
> 

The graph may look exponential but that's often misleading (it may very 
well be a polynomial). To tell for sure plot the data on a graph with a 
logarithmic Y axis and then on a graph with both axes logarithmic. If 
the plot is a straight line on the semilogy graph then it's exponential. 
If it's a straight line on the loglog graph then it's polinomial.