🌐 AI搜索 & 代理 主页
Skip to content

Commit 4c2777d

Browse files
tglsfdctglsfdc
committed
Change get_variable_numdistinct's API to flag default estimates explicitly.
Formerly, callers tested for DEFAULT_NUM_DISTINCT, which had the problem that a perfectly solid estimate might be mistaken for a content-free default.
1 parent 1cb108e commit 4c2777d

File tree

2 files changed

+45
-28
lines changed

2 files changed

+45
-28
lines changed

‎src/backend/utils/adt/selfuncs.c‎

Lines changed: 43 additions & 27 deletions
Original file line numberDiff line numberDiff line change
@@ -244,6 +244,7 @@ var_eq_const(VariableStatData *vardata, Oid operator,
244244
bool varonleft)
245245
{
246246
double selec;
247+
bool isdefault;
247248

248249
/*
249250
* If the constant is NULL, assume operator is strict and return zero, ie,
@@ -344,7 +345,7 @@ var_eq_const(VariableStatData *vardata, Oid operator,
344345
* all the not-common values share this remaining fraction
345346
* equally, so we divide by the number of other distinct values.
346347
*/
347-
otherdistinct = get_variable_numdistinct(vardata) - nnumbers;
348+
otherdistinct = get_variable_numdistinct(vardata, &isdefault) - nnumbers;
348349
if (otherdistinct > 1)
349350
selec /= otherdistinct;
350351

@@ -366,7 +367,7 @@ var_eq_const(VariableStatData *vardata, Oid operator,
366367
* of distinct values and assuming they are equally common. (The guess
367368
* is unlikely to be very good, but we do know a few special cases.)
368369
*/
369-
selec = 1.0 / get_variable_numdistinct(vardata);
370+
selec = 1.0 / get_variable_numdistinct(vardata, &isdefault);
370371
}
371372

372373
/* result should be in range, but make sure... */
@@ -384,6 +385,7 @@ var_eq_non_const(VariableStatData *vardata, Oid operator,
384385
bool varonleft)
385386
{
386387
double selec;
388+
bool isdefault;
387389

388390
/*
389391
* If we matched the var to a unique index, assume there is exactly one
@@ -414,7 +416,7 @@ var_eq_non_const(VariableStatData *vardata, Oid operator,
414416
* idea?)
415417
*/
416418
selec = 1.0 - stats->stanullfrac;
417-
ndistinct = get_variable_numdistinct(vardata);
419+
ndistinct = get_variable_numdistinct(vardata, &isdefault);
418420
if (ndistinct > 1)
419421
selec /= ndistinct;
420422

@@ -441,7 +443,7 @@ var_eq_non_const(VariableStatData *vardata, Oid operator,
441443
* of distinct values and assuming they are equally common. (The guess
442444
* is unlikely to be very good, but we do know a few special cases.)
443445
*/
444-
selec = 1.0 / get_variable_numdistinct(vardata);
446+
selec = 1.0 / get_variable_numdistinct(vardata, &isdefault);
445447
}
446448

447449
/* result should be in range, but make sure... */
@@ -2071,6 +2073,8 @@ eqjoinsel_inner(Oid operator,
20712073
double selec;
20722074
double nd1;
20732075
double nd2;
2076+
bool isdefault1;
2077+
bool isdefault2;
20742078
Form_pg_statistic stats1 = NULL;
20752079
Form_pg_statistic stats2 = NULL;
20762080
bool have_mcvs1 = false;
@@ -2084,8 +2088,8 @@ eqjoinsel_inner(Oid operator,
20842088
float4 *numbers2 = NULL;
20852089
int nnumbers2 = 0;
20862090

2087-
nd1 = get_variable_numdistinct(vardata1);
2088-
nd2 = get_variable_numdistinct(vardata2);
2091+
nd1 = get_variable_numdistinct(vardata1, &isdefault1);
2092+
nd2 = get_variable_numdistinct(vardata2, &isdefault2);
20892093

20902094
if (HeapTupleIsValid(vardata1->statsTuple))
20912095
{
@@ -2296,6 +2300,8 @@ eqjoinsel_semi(Oid operator,
22962300
double selec;
22972301
double nd1;
22982302
double nd2;
2303+
bool isdefault1;
2304+
bool isdefault2;
22992305
Form_pg_statistic stats1 = NULL;
23002306
bool have_mcvs1 = false;
23012307
Datum *values1 = NULL;
@@ -2308,8 +2314,8 @@ eqjoinsel_semi(Oid operator,
23082314
float4 *numbers2 = NULL;
23092315
int nnumbers2 = 0;
23102316

2311-
nd1 = get_variable_numdistinct(vardata1);
2312-
nd2 = get_variable_numdistinct(vardata2);
2317+
nd1 = get_variable_numdistinct(vardata1, &isdefault1);
2318+
nd2 = get_variable_numdistinct(vardata2, &isdefault2);
23132319

23142320
/*
23152321
* We clamp nd2 to be not more than what we estimate the inner relation's
@@ -2441,7 +2447,7 @@ eqjoinsel_semi(Oid operator,
24412447
* nd2 is default, punt and assume half of the uncertain rows have
24422448
* join partners.
24432449
*/
2444-
if (nd1 != DEFAULT_NUM_DISTINCT && nd2 != DEFAULT_NUM_DISTINCT)
2450+
if (!isdefault1 && !isdefault2)
24452451
{
24462452
nd1 -= nmatches;
24472453
nd2 -= nmatches;
@@ -2464,7 +2470,7 @@ eqjoinsel_semi(Oid operator,
24642470
*/
24652471
double nullfrac1 = stats1 ? stats1->stanullfrac : 0.0;
24662472

2467-
if (nd1 != DEFAULT_NUM_DISTINCT && nd2 != DEFAULT_NUM_DISTINCT)
2473+
if (!isdefault1 && !isdefault2)
24682474
{
24692475
if (nd1 <= nd2 || nd2 < 0)
24702476
selec = 1.0 - nullfrac1;
@@ -2955,9 +2961,10 @@ add_unique_group_var(PlannerInfo *root, List *varinfos,
29552961
{
29562962
GroupVarInfo *varinfo;
29572963
double ndistinct;
2964+
bool isdefault;
29582965
ListCell *lc;
29592966

2960-
ndistinct = get_variable_numdistinct(vardata);
2967+
ndistinct = get_variable_numdistinct(vardata, &isdefault);
29612968

29622969
/* cannot use foreach here because of possible list_delete */
29632970
lc = list_head(varinfos);
@@ -3292,14 +3299,23 @@ estimate_hash_bucketsize(PlannerInfo *root, Node *hashkey, double nbuckets)
32923299
stanullfrac,
32933300
mcvfreq,
32943301
avgfreq;
3302+
bool isdefault;
32953303
float4 *numbers;
32963304
int nnumbers;
32973305

32983306
examine_variable(root, hashkey, 0, &vardata);
32993307

3300-
/* Get number of distinct values and fraction that are null */
3301-
ndistinct = get_variable_numdistinct(&vardata);
3308+
/* Get number of distinct values */
3309+
ndistinct = get_variable_numdistinct(&vardata, &isdefault);
33023310

3311+
/* If ndistinct isn't real, punt and return 0.1, per comments above */
3312+
if (isdefault)
3313+
{
3314+
ReleaseVariableStats(vardata);
3315+
return (Selectivity) 0.1;
3316+
}
3317+
3318+
/* Get fraction that are null */
33033319
if (HeapTupleIsValid(vardata.statsTuple))
33043320
{
33053321
Form_pg_statistic stats;
@@ -3308,19 +3324,7 @@ estimate_hash_bucketsize(PlannerInfo *root, Node *hashkey, double nbuckets)
33083324
stanullfrac = stats->stanullfrac;
33093325
}
33103326
else
3311-
{
3312-
/*
3313-
* Believe a default ndistinct only if it came from stats. Otherwise
3314-
* punt and return 0.1, per comments above.
3315-
*/
3316-
if (ndistinct == DEFAULT_NUM_DISTINCT)
3317-
{
3318-
ReleaseVariableStats(vardata);
3319-
return (Selectivity) 0.1;
3320-
}
3321-
33223327
stanullfrac = 0.0;
3323-
}
33243328

33253329
/* Compute avg freq of all distinct data values in raw relation */
33263330
avgfreq = (1.0 - stanullfrac) / ndistinct;
@@ -4414,16 +4418,20 @@ examine_simple_variable(PlannerInfo *root, Var *var,
44144418
* Estimate the number of distinct values of a variable.
44154419
*
44164420
* vardata: results of examine_variable
4421+
* *isdefault: set to TRUE if the result is a default rather than based on
4422+
* anything meaningful.
44174423
*
44184424
* NB: be careful to produce an integral result, since callers may compare
44194425
* the result to exact integer counts.
44204426
*/
44214427
double
4422-
get_variable_numdistinct(VariableStatData *vardata)
4428+
get_variable_numdistinct(VariableStatData *vardata, bool *isdefault)
44234429
{
44244430
double stadistinct;
44254431
double ntuples;
44264432

4433+
*isdefault = false;
4434+
44274435
/*
44284436
* Determine the stadistinct value to use. There are cases where we can
44294437
* get an estimate even without a pg_statistic entry, or can get a better
@@ -4496,10 +4504,16 @@ get_variable_numdistinct(VariableStatData *vardata)
44964504
* Otherwise we need to get the relation size; punt if not available.
44974505
*/
44984506
if (vardata->rel == NULL)
4507+
{
4508+
*isdefault = true;
44994509
return DEFAULT_NUM_DISTINCT;
4510+
}
45004511
ntuples = vardata->rel->tuples;
45014512
if (ntuples <= 0.0)
4513+
{
4514+
*isdefault = true;
45024515
return DEFAULT_NUM_DISTINCT;
4516+
}
45034517

45044518
/*
45054519
* If we had a relative estimate, use that.
@@ -4509,11 +4523,13 @@ get_variable_numdistinct(VariableStatData *vardata)
45094523

45104524
/*
45114525
* With no data, estimate ndistinct = ntuples if the table is small, else
4512-
* use default.
4526+
* use default. We use DEFAULT_NUM_DISTINCT as the cutoff for "small"
4527+
* so that the behavior isn't discontinuous.
45134528
*/
45144529
if (ntuples < DEFAULT_NUM_DISTINCT)
45154530
return ntuples;
45164531

4532+
*isdefault = true;
45174533
return DEFAULT_NUM_DISTINCT;
45184534
}
45194535

‎src/include/utils/selfuncs.h‎

Lines changed: 2 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -121,7 +121,8 @@ extern void get_join_variables(PlannerInfo *root, List *args,
121121
VariableStatData *vardata1,
122122
VariableStatData *vardata2,
123123
bool *join_is_reversed);
124-
extern double get_variable_numdistinct(VariableStatData *vardata);
124+
extern double get_variable_numdistinct(VariableStatData *vardata,
125+
bool *isdefault);
125126
extern double mcv_selectivity(VariableStatData *vardata, FmgrInfo *opproc,
126127
Datum constval, bool varonleft,
127128
double *sumcommonp);

0 commit comments

Comments
 (0)