Re: Parallel safety of contrib extensions

2020-06-18 Thread Michael Paquier
On Thu, Jun 18, 2020 at 02:26:04PM +, Winfield, Steven wrote: > Many thanks for the pointers - I've submitted a patch. Thanks Steve, I have noticed the patch. For the sake of the archives, it is here: https://www.postgresql.org/message-id/AM5PR0901MB1587E47B1ACF23C6089DFCA3FD9B0%40AM5PR0901MB

Re: Parallel safety of contrib extensions

2020-06-18 Thread Winfield, Steven
> This would not be a complicated change as it requires creating a new > version script for those modules. You can look at commit 20eb273 as > an example, and the work is even simpler now that we just need update > scripts when bumping a module's version (those named foo--1.0--1.1.sql > and not fo

Re: Parallel safety of contrib extensions

2020-06-11 Thread Michael Paquier
On Wed, Jun 10, 2020 at 12:40:54PM -0400, Tom Lane wrote: > "Winfield, Steven" writes: >> There was a thread about this back in 2016[1], but I've just been >> bitten by it and wondered if any (more) extensions, particularly >> btree_gist, will have their operators/functions verified and marked as

Re: Parallel safety of contrib extensions

2020-06-10 Thread Tom Lane
"Winfield, Steven" writes: > There was a thread about this back in 2016[1], but I've just been bitten by > it and wondered if any (more) extensions, particularly btree_gist, will have > their operators/functions verified and marked as parallel-safe? Whenever somebody does the legwork and sends

Parallel safety of contrib extensions

2020-06-10 Thread Winfield, Steven
Hi all, There was a thread about this back in 2016[1], but I've just been bitten by it and wondered if any (more) extensions, particularly btree_gist, will have their operators/functions verified and marked as parallel-safe? In our case, we're interested in the float8 <-> float8 operator. Altho