Re: Hash tables in dynamic shared memory

On Mon, Dec 19, 2016 at 12:33 PM, Thomas Munro
<thomas(dot)munro(at)enterprisedb(dot)com> wrote:
> Since Dilip Kumar's Parallel Bitmap Heap Scan project is no longer
> using this, I think I should park it here unless/until another
> potential use case pops up. Some interesting candidates have been
> mentioned already, and I'm fairly sure there are other uses too, but I
> don't expect anyone to be interested in committing this patch until
> something concrete shows up, so I'll go work on other things until
> then.

Here is a rebased version. Changes since v2:

1. Out-of-memory conditions now raise errors (following
dsa_allocate's change in behaviour to match palloc).
2. Moved into 'lib' where other reusable data structures live.

There are still a couple of things I'd like to adjust in this code
(including feedback from John and improvements to the iterator code
which is overcomplicated) but I'm posting it today as infrastructure
for another patch.

--
Thomas Munro
http://www.enterprisedb.com