mirrors/postgres
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

pull_up_subqueries() should copy the subquery before starting to modify

Browse Source 
it.  Not sure why I'd thought it would be a good idea to do differently
way back when, but Greg Stark exposed the folly of doing so ...
This commit is contained in:
Tom Lane 2003-10-13 23:48:16 +00:00
parent 4a39057e59
commit d18ba3f452
1 changed files with 16 additions and 19 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

35
src/backend/optimizer/prep/prepjointree.c
View File

@ -16,7 +16,7 @@
* *
* *
* IDENTIFICATION * IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/optimizer/prep/prepjointree.c,v 1.12 2003/09/25 06:58:00 petere Exp $ * $Header: /cvsroot/pgsql/src/backend/optimizer/prep/prepjointree.c,v 1.13 2003/10/13 23:48:16 tgl Exp $
* *
*------------------------------------------------------------------------- *-------------------------------------------------------------------------
*/ */
@ -148,9 +148,10 @@ pull_up_subqueries(Query *parse, Node *jtnode, bool below_outer_join)
* If we are inside an outer join, only pull up subqueries whose * If we are inside an outer join, only pull up subqueries whose
* targetlists are nullable --- otherwise substituting their tlist * targetlists are nullable --- otherwise substituting their tlist
* entries for upper Var references would do the wrong thing (the * entries for upper Var references would do the wrong thing (the
* results wouldn't become NULL when they're supposed to). XXX * results wouldn't become NULL when they're supposed to).
* This could be improved by generating pseudo-variables for such *
* expressions; we'd have to figure out how to get the pseudo- * XXX This could be improved by generating pseudo-variables for
* such expressions; we'd have to figure out how to get the pseudo-
* variables evaluated at the right place in the modified plan * variables evaluated at the right place in the modified plan
* tree. Fix it someday. * tree. Fix it someday.
* *
@ -167,7 +168,15 @@ pull_up_subqueries(Query *parse, Node *jtnode, bool below_outer_join)
List *rt; List *rt;
/* /*
* First, pull up any IN clauses within the subquery's WHERE, * First make a modifiable copy of the subquery. This avoids
* problems if the same subquery is referenced from multiple
* jointree items (which can't happen normally, but might after
* rule rewriting).
*/
subquery = copyObject(subquery);
/*
* Pull up any IN clauses within the subquery's WHERE,
* so that we don't leave unoptimized INs behind. * so that we don't leave unoptimized INs behind.
*/ */
if (subquery->hasSubLinks) if (subquery->hasSubLinks)
@ -175,15 +184,9 @@ pull_up_subqueries(Query *parse, Node *jtnode, bool below_outer_join)
subquery->jointree->quals); subquery->jointree->quals);
/* /*
* Now, recursively pull up the subquery's subqueries, so that * Recursively pull up the subquery's subqueries, so that
* this routine's processing is complete for its jointree and * this routine's processing is complete for its jointree and
* rangetable. NB: if the same subquery is referenced from * rangetable.
* multiple jointree items (which can't happen normally, but
* might after rule rewriting), then we will invoke this
* processing multiple times on that subquery. OK because
* nothing will happen after the first time. We do have to be
* careful to copy everything we pull up, however, or risk
* having chunks of structure multiply linked.
* *
* Note: 'false' is correct here even if we are within an outer * Note: 'false' is correct here even if we are within an outer
* join in the upper query; the lower query starts with a * join in the upper query; the lower query starts with a
@ -193,12 +196,6 @@ pull_up_subqueries(Query *parse, Node *jtnode, bool below_outer_join)
pull_up_subqueries(subquery, (Node *) subquery->jointree, pull_up_subqueries(subquery, (Node *) subquery->jointree,
false); false);
/*
* Now make a modifiable copy of the subquery that we can run
* OffsetVarNodes and IncrementVarSublevelsUp on.
*/
subquery = copyObject(subquery);
/* /*
* Adjust level-0 varnos in subquery so that we can append its * Adjust level-0 varnos in subquery so that we can append its
* rangetable to upper query's. * rangetable to upper query's.