Re: WIP: [[Parallel] Shared] Hash

On Sat, Dec 31, 2016 at 2:52 AM, Thomas Munro
<thomas(dot)munro(at)enterprisedb(dot)com> wrote:
> Unfortunately it's been a bit trickier than I anticipated to get the
> interprocess batch file sharing and hash table shrinking working
> correctly and I don't yet have a new patch in good enough shape to
> post in time for the January CF. More soon.

I noticed a bug in your latest revision:

> + /*
> + * In HJ_NEED_NEW_OUTER, we already selected the current inner batch for
> + * reading from. If there is a shared hash table, we may have already
> + * partially loaded the hash table in ExecHashJoinPreloadNextBatch.
> + */
> + Assert(hashtable->batch_reader.batchno = curbatch);
> + Assert(hashtable->batch_reader.inner);

Obviously this isn't supposed to be an assignment.

--
Peter Geoghegan