On Sat, Feb 29, 2020 at 8:18 AM Tom Lane <t...@sss.pgh.pa.us> wrote: > Amit Langote <amitlangot...@gmail.com> writes: > > On Wed, Nov 6, 2019 at 2:00 AM Tom Lane <t...@sss.pgh.pa.us> wrote: > >> Just to leave a breadcrumb in this thread --- the planner failure > >> induced by d25ea01275 has been fixed in 529ebb20a. The difficulty > >> with multiway full joins that Amit started this thread with remains > >> open, but I imagine the posted patches will need rebasing over > >> 529ebb20a. > > > Here are the rebased patches. > > The cfbot shows these patches as failing regression tests. I think > it is just cosmetic fallout from 6ef77cf46 having changed EXPLAIN's > choices of table alias names; but please look closer to confirm, > and post updated patches.
Thanks for notifying. Checked and indeed fallout from 6ef77cf46 seems to be the reason a test is failing. Updated patches attached. Thanks, Amit
v5-0002-Move-some-code-from-joinrel.c-to-relnode.c.patch
Description: Binary data
v5-0001-Some-cosmetic-improvements-to-partitionwise-join-.patch
Description: Binary data
v5-0003-Fix-partitionwise-join-to-handle-FULL-JOINs-corre.patch
Description: Binary data
