PostgreSQL Source Code  git master
selfuncs.h
Go to the documentation of this file.
1 /*-------------------------------------------------------------------------
2  *
3  * selfuncs.h
4  * Selectivity functions for standard operators, and assorted
5  * infrastructure for selectivity and cost estimation.
6  *
7  *
8  * Portions Copyright (c) 1996-2024, PostgreSQL Global Development Group
9  * Portions Copyright (c) 1994, Regents of the University of California
10  *
11  * src/include/utils/selfuncs.h
12  *
13  *-------------------------------------------------------------------------
14  */
15 #ifndef SELFUNCS_H
16 #define SELFUNCS_H
17 
18 #include "access/htup.h"
19 #include "fmgr.h"
20 #include "nodes/pathnodes.h"
21 
22 
23 /*
24  * Note: the default selectivity estimates are not chosen entirely at random.
25  * We want them to be small enough to ensure that indexscans will be used if
26  * available, for typical table densities of ~100 tuples/page. Thus, for
27  * example, 0.01 is not quite small enough, since that makes it appear that
28  * nearly all pages will be hit anyway. Also, since we sometimes estimate
29  * eqsel as 1/num_distinct, we probably want DEFAULT_NUM_DISTINCT to equal
30  * 1/DEFAULT_EQ_SEL.
31  */
32 
33 /* default selectivity estimate for equalities such as "A = b" */
34 #define DEFAULT_EQ_SEL 0.005
35 
36 /* default selectivity estimate for inequalities such as "A < b" */
37 #define DEFAULT_INEQ_SEL 0.3333333333333333
38 
39 /* default selectivity estimate for range inequalities "A > b AND A < c" */
40 #define DEFAULT_RANGE_INEQ_SEL 0.005
41 
42 /* default selectivity estimate for multirange inequalities "A > b AND A < c" */
43 #define DEFAULT_MULTIRANGE_INEQ_SEL 0.005
44 
45 /* default selectivity estimate for pattern-match operators such as LIKE */
46 #define DEFAULT_MATCH_SEL 0.005
47 
48 /* default selectivity estimate for other matching operators */
49 #define DEFAULT_MATCHING_SEL 0.010
50 
51 /* default number of distinct values in a table */
52 #define DEFAULT_NUM_DISTINCT 200
53 
54 /* default selectivity estimate for boolean and null test nodes */
55 #define DEFAULT_UNK_SEL 0.005
56 #define DEFAULT_NOT_UNK_SEL (1.0 - DEFAULT_UNK_SEL)
57 
58 
59 /*
60  * Clamp a computed probability estimate (which may suffer from roundoff or
61  * estimation errors) to valid range. Argument must be a float variable.
62  */
63 #define CLAMP_PROBABILITY(p) \
64  do { \
65  if (p < 0.0) \
66  p = 0.0; \
67  else if (p > 1.0) \
68  p = 1.0; \
69  } while (0)
70 
71 /*
72  * A set of flags which some selectivity estimation functions can pass back to
73  * callers to provide further details about some assumptions which were made
74  * during the estimation.
75  */
76 #define SELFLAG_USED_DEFAULT (1 << 0) /* Estimation fell back on one
77  * of the DEFAULTs as defined
78  * above. */
79 
80 typedef struct EstimationInfo
81 {
82  uint32 flags; /* Flags, as defined above to mark special
83  * properties of the estimation. */
85 
86 /* Return data from examine_variable and friends */
87 typedef struct VariableStatData
88 {
89  Node *var; /* the Var or expression tree */
90  RelOptInfo *rel; /* Relation, or NULL if not identifiable */
91  HeapTuple statsTuple; /* pg_statistic tuple, or NULL if none */
92  /* NB: if statsTuple!=NULL, it must be freed when caller is done */
93  void (*freefunc) (HeapTuple tuple); /* how to free statsTuple */
94  Oid vartype; /* exposed type of expression */
95  Oid atttype; /* actual type (after stripping relabel) */
96  int32 atttypmod; /* actual typmod (after stripping relabel) */
97  bool isunique; /* matches unique index or DISTINCT clause */
98  bool acl_ok; /* result of ACL check on table or column */
100 
101 #define ReleaseVariableStats(vardata) \
102  do { \
103  if (HeapTupleIsValid((vardata).statsTuple)) \
104  (vardata).freefunc((vardata).statsTuple); \
105  } while(0)
106 
107 
108 /*
109  * genericcostestimate is a general-purpose estimator that can be used for
110  * most index types. In some cases we use genericcostestimate as the base
111  * code and then incorporate additional index-type-specific knowledge in
112  * the type-specific calling function. To avoid code duplication, we make
113  * genericcostestimate return a number of intermediate values as well as
114  * its preliminary estimates of the output cost values. The GenericCosts
115  * struct includes all these values.
116  *
117  * Callers should initialize all fields of GenericCosts to zero. In addition,
118  * they can set numIndexTuples to some positive value if they have a better
119  * than default way of estimating the number of leaf index tuples visited.
120  * Similarly, they can set num_sa_scans to some value >= 1 for an index AM
121  * that doesn't necessarily perform exactly one primitive index scan per
122  * distinct combination of ScalarArrayOp array elements.
123  */
124 typedef struct
125 {
126  /* These are the values the cost estimator must return to the planner */
127  Cost indexStartupCost; /* index-related startup cost */
128  Cost indexTotalCost; /* total index-related scan cost */
129  Selectivity indexSelectivity; /* selectivity of index */
130  double indexCorrelation; /* order correlation of index */
131 
132  /* Intermediate values we obtain along the way */
133  double numIndexPages; /* number of leaf pages visited */
134  double numIndexTuples; /* number of leaf tuples visited */
135  double spc_random_page_cost; /* relevant random_page_cost value */
136  double num_sa_scans; /* # indexscans from ScalarArrayOpExprs */
137 } GenericCosts;
138 
139 /* Hooks for plugins to get control when we ask for stats */
141  RangeTblEntry *rte,
143  VariableStatData *vardata);
146  Oid indexOid,
147  AttrNumber indexattnum,
148  VariableStatData *vardata);
150 
151 /* Functions in selfuncs.c */
152 
153 extern void examine_variable(PlannerInfo *root, Node *node, int varRelid,
154  VariableStatData *vardata);
155 extern bool statistic_proc_security_check(VariableStatData *vardata, Oid func_oid);
157  int varRelid,
158  VariableStatData *vardata, Node **other,
159  bool *varonleft);
161  SpecialJoinInfo *sjinfo,
162  VariableStatData *vardata1,
163  VariableStatData *vardata2,
164  bool *join_is_reversed);
165 extern double get_variable_numdistinct(VariableStatData *vardata,
166  bool *isdefault);
167 extern double mcv_selectivity(VariableStatData *vardata,
168  FmgrInfo *opproc, Oid collation,
169  Datum constval, bool varonleft,
170  double *sumcommonp);
171 extern double histogram_selectivity(VariableStatData *vardata,
172  FmgrInfo *opproc, Oid collation,
173  Datum constval, bool varonleft,
174  int min_hist_size, int n_skip,
175  int *hist_size);
177  Oid oproid, Oid collation,
178  List *args, int varRelid,
179  double default_selectivity);
181  VariableStatData *vardata,
182  Oid opoid, FmgrInfo *opproc,
183  bool isgt, bool iseq,
184  Oid collation,
185  Datum constval, Oid consttype);
186 extern double var_eq_const(VariableStatData *vardata,
187  Oid oproid, Oid collation,
188  Datum constval, bool constisnull,
189  bool varonleft, bool negate);
190 extern double var_eq_non_const(VariableStatData *vardata,
191  Oid oproid, Oid collation,
192  Node *other,
193  bool varonleft, bool negate);
194 
195 extern Selectivity boolvarsel(PlannerInfo *root, Node *arg, int varRelid);
196 extern Selectivity booltestsel(PlannerInfo *root, BoolTestType booltesttype,
197  Node *arg, int varRelid,
198  JoinType jointype, SpecialJoinInfo *sjinfo);
199 extern Selectivity nulltestsel(PlannerInfo *root, NullTestType nulltesttype,
200  Node *arg, int varRelid,
201  JoinType jointype, SpecialJoinInfo *sjinfo);
203  ScalarArrayOpExpr *clause,
204  bool is_join_clause,
205  int varRelid, JoinType jointype, SpecialJoinInfo *sjinfo);
206 extern double estimate_array_length(PlannerInfo *root, Node *arrayexpr);
208  RowCompareExpr *clause,
209  int varRelid, JoinType jointype, SpecialJoinInfo *sjinfo);
210 
211 extern void mergejoinscansel(PlannerInfo *root, Node *clause,
212  Oid opfamily, int strategy, bool nulls_first,
213  Selectivity *leftstart, Selectivity *leftend,
214  Selectivity *rightstart, Selectivity *rightend);
215 
216 extern double estimate_num_groups(PlannerInfo *root, List *groupExprs,
217  double input_rows, List **pgset,
218  EstimationInfo *estinfo);
219 
221  Node *hashkey, double nbuckets,
222  Selectivity *mcv_freq,
223  Selectivity *bucketsize_frac);
224 extern double estimate_hashagg_tablesize(PlannerInfo *root, Path *path,
225  const AggClauseCosts *agg_costs,
226  double dNumGroups);
227 
228 extern List *get_quals_from_indexclauses(List *indexclauses);
230  List *indexquals);
232  List *indexQuals);
233 extern void genericcostestimate(PlannerInfo *root, IndexPath *path,
234  double loop_count,
235  GenericCosts *costs);
236 
237 /* Functions in array_selfuncs.c */
238 
240  Node *leftop, Node *rightop,
241  Oid elemtype, bool isEquality, bool useOr,
242  int varRelid);
243 
244 #endif /* SELFUNCS_H */
int16 AttrNumber
Definition: attnum.h:21
unsigned int uint32
Definition: c.h:506
#define PGDLLIMPORT
Definition: c.h:1316
signed int int32
Definition: c.h:494
unsigned char bool
Definition: c.h:456
double Cost
Definition: nodes.h:251
double Selectivity
Definition: nodes.h:250
JoinType
Definition: nodes.h:288
int16 attnum
Definition: pg_attribute.h:74
void * arg
uintptr_t Datum
Definition: postgres.h:64
unsigned int Oid
Definition: postgres_ext.h:31
BoolTestType
Definition: primnodes.h:1977
NullTestType
Definition: primnodes.h:1953
tree ctl root
Definition: radixtree.h:1886
Selectivity scalararraysel_containment(PlannerInfo *root, Node *leftop, Node *rightop, Oid elemtype, bool isEquality, bool useOr, int varRelid)
bool get_restriction_variable(PlannerInfo *root, List *args, int varRelid, VariableStatData *vardata, Node **other, bool *varonleft)
Definition: selfuncs.c:4883
List * get_quals_from_indexclauses(List *indexclauses)
Definition: selfuncs.c:6460
double var_eq_const(VariableStatData *vardata, Oid oproid, Oid collation, Datum constval, bool constisnull, bool varonleft, bool negate)
Definition: selfuncs.c:295
double generic_restriction_selectivity(PlannerInfo *root, Oid oproid, Oid collation, List *args, int varRelid, double default_selectivity)
Definition: selfuncs.c:914
Selectivity booltestsel(PlannerInfo *root, BoolTestType booltesttype, Node *arg, int varRelid, JoinType jointype, SpecialJoinInfo *sjinfo)
Definition: selfuncs.c:1540
double estimate_array_length(PlannerInfo *root, Node *arrayexpr)
Definition: selfuncs.c:2136
double mcv_selectivity(VariableStatData *vardata, FmgrInfo *opproc, Oid collation, Datum constval, bool varonleft, double *sumcommonp)
Definition: selfuncs.c:732
Selectivity nulltestsel(PlannerInfo *root, NullTestType nulltesttype, Node *arg, int varRelid, JoinType jointype, SpecialJoinInfo *sjinfo)
Definition: selfuncs.c:1698
PGDLLIMPORT get_relation_stats_hook_type get_relation_stats_hook
Definition: selfuncs.c:147
struct VariableStatData VariableStatData
void examine_variable(PlannerInfo *root, Node *node, int varRelid, VariableStatData *vardata)
Definition: selfuncs.c:5012
void mergejoinscansel(PlannerInfo *root, Node *clause, Oid opfamily, int strategy, bool nulls_first, Selectivity *leftstart, Selectivity *leftend, Selectivity *rightstart, Selectivity *rightend)
Definition: selfuncs.c:2952
bool(* get_relation_stats_hook_type)(PlannerInfo *root, RangeTblEntry *rte, AttrNumber attnum, VariableStatData *vardata)
Definition: selfuncs.h:138
struct EstimationInfo EstimationInfo
double estimate_num_groups(PlannerInfo *root, List *groupExprs, double input_rows, List **pgset, EstimationInfo *estinfo)
Definition: selfuncs.c:3416
double ineq_histogram_selectivity(PlannerInfo *root, VariableStatData *vardata, Oid opoid, FmgrInfo *opproc, bool isgt, bool iseq, Oid collation, Datum constval, Oid consttype)
Definition: selfuncs.c:1041
void genericcostestimate(PlannerInfo *root, IndexPath *path, double loop_count, GenericCosts *costs)
Definition: selfuncs.c:6544
bool(* get_index_stats_hook_type)(PlannerInfo *root, Oid indexOid, AttrNumber indexattnum, VariableStatData *vardata)
Definition: selfuncs.h:143
List * add_predicate_to_index_quals(IndexOptInfo *index, List *indexQuals)
Definition: selfuncs.c:6767
double histogram_selectivity(VariableStatData *vardata, FmgrInfo *opproc, Oid collation, Datum constval, bool varonleft, int min_hist_size, int n_skip, int *hist_size)
Definition: selfuncs.c:823
Selectivity boolvarsel(PlannerInfo *root, Node *arg, int varRelid)
Definition: selfuncs.c:1512
Selectivity scalararraysel(PlannerInfo *root, ScalarArrayOpExpr *clause, bool is_join_clause, int varRelid, JoinType jointype, SpecialJoinInfo *sjinfo)
Definition: selfuncs.c:1816
double var_eq_non_const(VariableStatData *vardata, Oid oproid, Oid collation, Node *other, bool varonleft, bool negate)
Definition: selfuncs.c:466
double get_variable_numdistinct(VariableStatData *vardata, bool *isdefault)
Definition: selfuncs.c:5764
PGDLLIMPORT get_index_stats_hook_type get_index_stats_hook
Definition: selfuncs.c:148
bool statistic_proc_security_check(VariableStatData *vardata, Oid func_oid)
Definition: selfuncs.c:5735
double estimate_hashagg_tablesize(PlannerInfo *root, Path *path, const AggClauseCosts *agg_costs, double dNumGroups)
Definition: selfuncs.c:3917
void estimate_hash_bucket_stats(PlannerInfo *root, Node *hashkey, double nbuckets, Selectivity *mcv_freq, Selectivity *bucketsize_frac)
Definition: selfuncs.c:3798
Cost index_other_operands_eval_cost(PlannerInfo *root, List *indexquals)
Definition: selfuncs.c:6490
Selectivity rowcomparesel(PlannerInfo *root, RowCompareExpr *clause, int varRelid, JoinType jointype, SpecialJoinInfo *sjinfo)
Definition: selfuncs.c:2202
void get_join_variables(PlannerInfo *root, List *args, SpecialJoinInfo *sjinfo, VariableStatData *vardata1, VariableStatData *vardata2, bool *join_is_reversed)
Definition: selfuncs.c:4943
uint32 flags
Definition: selfuncs.h:80
Definition: fmgr.h:57
Definition: pg_list.h:54
Definition: nodes.h:129
HeapTuple statsTuple
Definition: selfuncs.h:89
int32 atttypmod
Definition: selfuncs.h:94
RelOptInfo * rel
Definition: selfuncs.h:88
void(* freefunc)(HeapTuple tuple)
Definition: selfuncs.h:91
Definition: type.h:95