(j3.2006) More on "contiguous"

Aleksandar Donev donev1
Tue May 15 19:24:31 EDT 2007


OK, Bob Corbett already said this, but I typed it before getting his 
message :-)

> Does "contiguous" make any sense and confer any benefit for other than
> the first dimension?
Of course it makes sense. And yes, it confers some benefits. For example,


would not generate any loops and run at memory bandwidth speed.

Other benefits include making such arrays "interoperable" since one can take 
C_LOC of them.

On Tuesday 15 May 2007 14:59, Bill Long wrote:
> know that the leftmost dimension of the array D is contiguous, but other
> dimensions might not qualify, then you could associate a contiguous
> pointer with D(:,i) and reference the pointer.
We killed that horse. One can only point contiguous pointers to "simply 
contiguous" targets, which D(:,i) is not unless D is known to be a contiguous 
array at compile-time.

We should certainly not change anything about contiguous at this point in 
time. There are better ways to convey contiguity/small dimensionality/fixed 
dimensionality for specific dimensions than the attribute. In fact, the whole 
attribute looks not so well-designed now...


More information about the J3 mailing list