From c42c404c84e001c7c7f39a1b5afaaeec3bd5912d Mon Sep 17 00:00:00 2001 From: "yizhi.fzh" Date: Tue, 2 Apr 2024 09:39:17 +0800 Subject: [PATCH v1 2/8] Remove estimiatedcluases and varRelid arguments comments and Assert around the changes provides more information. --- src/backend/optimizer/path/clausesel.c | 16 ++++++++++------ src/backend/statistics/extended_stats.c | 24 ++++++++++-------------- src/include/statistics/statistics.h | 4 +--- 3 files changed, 21 insertions(+), 23 deletions(-) diff --git a/src/backend/optimizer/path/clausesel.c b/src/backend/optimizer/path/clausesel.c index bedf76edae..e1683febf6 100644 --- a/src/backend/optimizer/path/clausesel.c +++ b/src/backend/optimizer/path/clausesel.c @@ -204,14 +204,18 @@ clauselist_selectivity_ext(PlannerInfo *root, * do to detect when this makes sense, but we can check that there are * join clauses, and that at least some of the rels have stats. * - * XXX Isn't this mutually exclusive with the preceding block which - * calculates estimates for a single relation? + * rel != NULL can't grantee the clause is not a join clause, for example + * t1 left join t2 ON t1.a = 3, but it can grantee we can't use extended + * statistics for estimation since it has only 1 relid. + * + * XXX: so we can grantee estimatedclauses == NULL now, so estimatedclauses + * in statext_try_join_estimates is removed. */ - if (use_extended_stats && - statext_try_join_estimates(root, clauses, varRelid, jointype, sjinfo, - estimatedclauses)) + if (use_extended_stats && rel == NULL && + statext_try_join_estimates(root, clauses, varRelid, jointype, sjinfo)) { - s1 *= statext_clauselist_join_selectivity(root, clauses, varRelid, + Assert(varRelid == 0); + s1 *= statext_clauselist_join_selectivity(root, clauses, jointype, sjinfo, &estimatedclauses); } diff --git a/src/backend/statistics/extended_stats.c b/src/backend/statistics/extended_stats.c index 183a8af07b..519c367dee 100644 --- a/src/backend/statistics/extended_stats.c +++ b/src/backend/statistics/extended_stats.c @@ -2804,8 +2804,7 @@ statext_determine_join_restrictions(PlannerInfo *root, RelOptInfo *rel, * on the conditions. */ static bool -statext_is_supported_join_clause(PlannerInfo *root, Node *clause, - int varRelid, SpecialJoinInfo *sjinfo) +statext_is_supported_join_clause(PlannerInfo *root, Node *clause, SpecialJoinInfo *sjinfo) { Oid oprsel; RestrictInfo *rinfo; @@ -2817,7 +2816,9 @@ statext_is_supported_join_clause(PlannerInfo *root, Node *clause, * * XXX See treat_as_join_clause. */ - if ((varRelid != 0) || (sjinfo == NULL)) + + /* duplicated with statext_try_join_estimates */ + if (sjinfo == NULL) return false; /* XXX Can we rely on always getting RestrictInfo here? */ @@ -2901,8 +2902,7 @@ statext_is_supported_join_clause(PlannerInfo *root, Node *clause, */ bool statext_try_join_estimates(PlannerInfo *root, List *clauses, int varRelid, - JoinType jointype, SpecialJoinInfo *sjinfo, - Bitmapset *estimatedclauses) + JoinType jointype, SpecialJoinInfo *sjinfo) { int listidx; int k; @@ -2939,15 +2939,11 @@ statext_try_join_estimates(PlannerInfo *root, List *clauses, int varRelid, /* needs to happen before skipping any clauses */ listidx++; - /* Skip clauses that were already estimated. */ - if (bms_is_member(listidx, estimatedclauses)) - continue; - /* * Skip clauses that are not join clauses or that we don't know * how to handle estimate using extended statistics. */ - if (!statext_is_supported_join_clause(root, clause, varRelid, sjinfo)) + if (!statext_is_supported_join_clause(root, clause, sjinfo)) continue; /* @@ -3017,7 +3013,7 @@ typedef struct JoinPairInfo * with F_EQJOINSEL selectivity function at the moment). */ static JoinPairInfo * -statext_build_join_pairs(PlannerInfo *root, List *clauses, int varRelid, +statext_build_join_pairs(PlannerInfo *root, List *clauses, JoinType jointype, SpecialJoinInfo *sjinfo, Bitmapset *estimatedclauses, int *npairs) { @@ -3053,7 +3049,7 @@ statext_build_join_pairs(PlannerInfo *root, List *clauses, int varRelid, * the moment we support just (Expr op Expr) clauses with each * side referencing just a single relation). */ - if (!statext_is_supported_join_clause(root, clause, varRelid, sjinfo)) + if (!statext_is_supported_join_clause(root, clause, sjinfo)) continue; /* statext_is_supported_join_clause guarantees RestrictInfo */ @@ -3241,7 +3237,7 @@ get_expression_for_rel(PlannerInfo *root, RelOptInfo *rel, Node *clause) * statistics in that case yet). */ Selectivity -statext_clauselist_join_selectivity(PlannerInfo *root, List *clauses, int varRelid, +statext_clauselist_join_selectivity(PlannerInfo *root, List *clauses, JoinType jointype, SpecialJoinInfo *sjinfo, Bitmapset **estimatedclauses) { @@ -3256,7 +3252,7 @@ statext_clauselist_join_selectivity(PlannerInfo *root, List *clauses, int varRel return 1.0; /* extract pairs of joined relations from the list of clauses */ - info = statext_build_join_pairs(root, clauses, varRelid, jointype, sjinfo, + info = statext_build_join_pairs(root, clauses, jointype, sjinfo, *estimatedclauses, &ninfo); /* no useful join pairs */ diff --git a/src/include/statistics/statistics.h b/src/include/statistics/statistics.h index 60b222028d..4f70034983 100644 --- a/src/include/statistics/statistics.h +++ b/src/include/statistics/statistics.h @@ -131,11 +131,9 @@ extern StatisticExtInfo *statext_find_matching_mcv(PlannerInfo *root, RelOptInfo Bitmapset *attnums, List *exprs); extern bool statext_try_join_estimates(PlannerInfo *root, List *clauses, int varRelid, - JoinType jointype, SpecialJoinInfo *sjinfo, - Bitmapset *estimatedclauses); + JoinType jointype, SpecialJoinInfo *sjinfo); extern Selectivity statext_clauselist_join_selectivity(PlannerInfo *root, List *clauses, - int varRelid, JoinType jointype, SpecialJoinInfo *sjinfo, Bitmapset **estimatedclauses); -- 2.34.1