On Wed, Aug 16, 2017 at 12:49 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Robert Haas <robertmhaas@gmail.com> writes:
>> On Wed, Aug 16, 2017 at 11:56 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
>>> I can get on board with that statement. Can you draft a better wording?
>
>> Here is an attempt. Feel free to edit.
>
> I think s/plan/query/ in the last bit would be better. Perhaps
>
> + * However, if force_parallel_mode = on or force_parallel_mode = regress,
> + * then we impose parallel mode whenever it's safe to do so, even if the
> + * final plan doesn't use parallelism. It's not safe to do so if the query
> + * contains anything parallel-unsafe; parallelModeOK will be false in that
> + * case. Otherwise, everything in the query is either parallel-safe or
> + * parallel-restricted, and in either case it should be OK to impose
> + * parallel-mode restrictions. If that ends up breaking something, then
> + * either some function the user included in the query is incorrectly
> + * labelled as parallel-safe or parallel-restricted when in reality it's
> + * parallel-unsafe, or else the query planner itself has a bug.
> */
Works for me. I'm happy to phrase this in any way that makes it clear
to you, 'cuz it's already clear to me. :-)
You want to push something, or should I do it?
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company