Fix filtering of "cloned" outer-join quals some more.

We've had multiple issues with the clause_is_computable_at logic that
I introduced in 2489d76c4: it's been known to accept more than one
clone of the same qual at the same plan node, and also to accept no
clones at all.  It's looking impractical to get it 100% right on the
basis of the currently-stored information, so fix it by introducing a
new RestrictInfo field "incompatible_relids" that explicitly shows
which outer joins a given clone mustn't be pushed above.

In principle we could populate this field in every RestrictInfo, but
that would cost space and there doesn't presently seem to be a need
for it in general.  Also, while deconstruct_distribute_oj_quals can
easily fill the field with the remaining members of the commutative
join set that it's considering, computing it in the general case
seems again pretty complicated.  So for now, just fill it for
clone quals.

Along the way, fix a bug that may or may not be only latent:
equivclass.c was generating replacement clauses with is_pushed_down
and has_clone/is_clone markings that didn't match their
required_relids.  This led me to conclude that leaving the clone flags
out of make_restrictinfo's purview wasn't such a great idea after all,
so add them.

Per report from Richard Guo.

Discussion: https://postgr.es/m/CAMbWs48EYi_9-pSd0ORes1kTmTeAjT4Q3gu49hJtYCbSn2JyeA@mail.gmail.com

1 files changed, 12 insertions, 0 deletions

diff --git a/src/include/nodes/pathnodes.h b/src/include/nodes/pathnodes.h
index 23dd671bf4e..c17b53f7adb 100644
--- a/src/include/nodes/pathnodes.h
+++ b/src/include/nodes/pathnodes.h
@@ -2430,6 +2430,15 @@ typedef struct LimitPath
  * conditions.  Possibly we should rename it to reflect that meaning?  But
  * see also the comments for RINFO_IS_PUSHED_DOWN, below.)
  *
+ * There is also an incompatible_relids field, which is a set of outer-join
+ * relids above which we cannot evaluate the clause (because they might null
+ * Vars it uses that should not be nulled yet).  In principle this could be
+ * filled in any RestrictInfo as the set of OJ relids that appear above the
+ * clause and null Vars that it uses.  In practice we only bother to populate
+ * it for "clone" clauses, as it's currently only needed to prevent multiple
+ * clones of the same clause from being accepted for evaluation at the same
+ * join level.
+ *
  * There is also an outer_relids field, which is NULL except for outer join
  * clauses; for those, it is the set of relids on the outer side of the
  * clause's outer join.  (These are rels that the clause cannot be applied to
@@ -2537,6 +2546,9 @@ typedef struct RestrictInfo
 	/* The set of relids required to evaluate the clause: */
 	Relids		required_relids;
 
+	/* Relids above which we cannot evaluate the clause (see comment above) */
+	Relids		incompatible_relids;
+
 	/* If an outer-join clause, the outer-side relations, else NULL: */
 	Relids		outer_relids;