Branch data Line data Source code
1 : : /*-------------------------------------------------------------------------
2 : : *
3 : : * equivclass.c
4 : : * Routines for managing EquivalenceClasses
5 : : *
6 : : * See src/backend/optimizer/README for discussion of EquivalenceClasses.
7 : : *
8 : : *
9 : : * Portions Copyright (c) 1996-2026, PostgreSQL Global Development Group
10 : : * Portions Copyright (c) 1994, Regents of the University of California
11 : : *
12 : : * IDENTIFICATION
13 : : * src/backend/optimizer/path/equivclass.c
14 : : *
15 : : *-------------------------------------------------------------------------
16 : : */
17 : : #include "postgres.h"
18 : :
19 : : #include <limits.h>
20 : :
21 : : #include "access/stratnum.h"
22 : : #include "catalog/pg_type.h"
23 : : #include "common/hashfn.h"
24 : : #include "nodes/makefuncs.h"
25 : : #include "nodes/nodeFuncs.h"
26 : : #include "optimizer/appendinfo.h"
27 : : #include "optimizer/clauses.h"
28 : : #include "optimizer/optimizer.h"
29 : : #include "optimizer/pathnode.h"
30 : : #include "optimizer/paths.h"
31 : : #include "optimizer/planmain.h"
32 : : #include "optimizer/restrictinfo.h"
33 : : #include "rewrite/rewriteManip.h"
34 : : #include "utils/lsyscache.h"
35 : :
36 : :
37 : : static EquivalenceMember *make_eq_member(EquivalenceClass *ec,
38 : : Expr *expr, Relids relids,
39 : : JoinDomain *jdomain,
40 : : EquivalenceMember *parent,
41 : : Oid datatype);
42 : : static EquivalenceMember *add_eq_member(EquivalenceClass *ec,
43 : : Expr *expr, Relids relids,
44 : : JoinDomain *jdomain,
45 : : Oid datatype);
46 : : static EquivalenceMember *add_child_eq_member(PlannerInfo *root,
47 : : EquivalenceClass *ec,
48 : : int ec_index, Expr *expr,
49 : : Relids relids,
50 : : JoinDomain *jdomain,
51 : : EquivalenceMember *parent_em,
52 : : Oid datatype,
53 : : Index child_relid);
54 : : static void generate_base_implied_equalities_const(PlannerInfo *root,
55 : : EquivalenceClass *ec);
56 : : static void generate_base_implied_equalities_no_const(PlannerInfo *root,
57 : : EquivalenceClass *ec);
58 : : static void generate_base_implied_equalities_broken(PlannerInfo *root,
59 : : EquivalenceClass *ec);
60 : : static List *generate_join_implied_equalities_normal(PlannerInfo *root,
61 : : EquivalenceClass *ec,
62 : : Relids join_relids,
63 : : Relids outer_relids,
64 : : Relids inner_relids);
65 : : static List *generate_join_implied_equalities_broken(PlannerInfo *root,
66 : : EquivalenceClass *ec,
67 : : Relids nominal_join_relids,
68 : : Relids outer_relids,
69 : : Relids nominal_inner_relids,
70 : : RelOptInfo *inner_rel);
71 : : static Oid select_equality_operator(EquivalenceClass *ec,
72 : : Oid lefttype, Oid righttype);
73 : : static RestrictInfo *create_join_clause(PlannerInfo *root,
74 : : EquivalenceClass *ec, Oid opno,
75 : : EquivalenceMember *leftem,
76 : : EquivalenceMember *rightem,
77 : : EquivalenceClass *parent_ec);
78 : : static bool reconsider_outer_join_clause(PlannerInfo *root,
79 : : OuterJoinClauseInfo *ojcinfo,
80 : : bool outer_on_left);
81 : : static bool reconsider_full_join_clause(PlannerInfo *root,
82 : : OuterJoinClauseInfo *ojcinfo);
83 : : static JoinDomain *find_join_domain(PlannerInfo *root, Relids relids);
84 : : static Bitmapset *get_eclass_indexes_for_relids(PlannerInfo *root,
85 : : Relids relids);
86 : : static Bitmapset *get_common_eclass_indexes(PlannerInfo *root, Relids relids1,
87 : : Relids relids2);
88 : : static void ec_build_derives_hash(PlannerInfo *root, EquivalenceClass *ec);
89 : : static void ec_add_derived_clauses(EquivalenceClass *ec, List *clauses);
90 : : static void ec_add_derived_clause(EquivalenceClass *ec, RestrictInfo *clause);
91 : : static void ec_add_clause_to_derives_hash(EquivalenceClass *ec, RestrictInfo *rinfo);
92 : : static void ec_clear_derived_clauses(EquivalenceClass *ec);
93 : : static RestrictInfo *ec_search_clause_for_ems(PlannerInfo *root, EquivalenceClass *ec,
94 : : EquivalenceMember *leftem,
95 : : EquivalenceMember *rightem,
96 : : EquivalenceClass *parent_ec);
97 : : static RestrictInfo *ec_search_derived_clause_for_ems(PlannerInfo *root,
98 : : EquivalenceClass *ec,
99 : : EquivalenceMember *leftem,
100 : : EquivalenceMember *rightem,
101 : : EquivalenceClass *parent_ec);
102 : :
103 : : /*
104 : : * Hash key identifying a derived clause.
105 : : *
106 : : * This structure should not be filled manually. Use fill_ec_derives_key() to
107 : : * set it up in canonical form.
108 : : */
109 : : typedef struct
110 : : {
111 : : EquivalenceMember *em1;
112 : : EquivalenceMember *em2;
113 : : EquivalenceClass *parent_ec;
114 : : } ECDerivesKey;
115 : :
116 : : /* Hash table entry in ec_derives_hash. */
117 : : typedef struct
118 : : {
119 : : uint32 status;
120 : : ECDerivesKey key;
121 : : RestrictInfo *rinfo;
122 : : } ECDerivesEntry;
123 : :
124 : : /* Threshold for switching from list to hash table */
125 : : #define EC_DERIVES_HASH_THRESHOLD 32
126 : :
127 : : #define SH_PREFIX derives
128 : : #define SH_ELEMENT_TYPE ECDerivesEntry
129 : : #define SH_KEY_TYPE ECDerivesKey
130 : : #define SH_KEY key
131 : : #define SH_HASH_KEY(tb, key) \
132 : : hash_bytes((const unsigned char *) &(key), sizeof(ECDerivesKey))
133 : : #define SH_EQUAL(tb, a, b) \
134 : : ((a).em1 == (b).em1 && (a).em2 == (b).em2 && (a).parent_ec == (b).parent_ec)
135 : : #define SH_SCOPE static inline
136 : : #define SH_DECLARE
137 : : #define SH_DEFINE
138 : : #include "lib/simplehash.h"
139 : :
140 : : /*
141 : : * process_equivalence
142 : : * The given clause has a mergejoinable operator and is not an outer-join
143 : : * qualification, so its two sides can be considered equal
144 : : * anywhere they are both computable; moreover that equality can be
145 : : * extended transitively. Record this knowledge in the EquivalenceClass
146 : : * data structure, if applicable. Returns true if successful, false if not
147 : : * (in which case caller should treat the clause as ordinary, not an
148 : : * equivalence).
149 : : *
150 : : * In some cases, although we cannot convert a clause into EquivalenceClass
151 : : * knowledge, we can still modify it to a more useful form than the original.
152 : : * Then, *p_restrictinfo will be replaced by a new RestrictInfo, which is what
153 : : * the caller should use for further processing.
154 : : *
155 : : * jdomain is the join domain within which the given clause was found.
156 : : * This limits the applicability of deductions from the EquivalenceClass,
157 : : * as described in optimizer/README.
158 : : *
159 : : * We reject proposed equivalence clauses if they contain leaky functions
160 : : * and have security_level above zero. The EC evaluation rules require us to
161 : : * apply certain tests at certain joining levels, and we can't tolerate
162 : : * delaying any test on security_level grounds. By rejecting candidate clauses
163 : : * that might require security delays, we ensure it's safe to apply an EC
164 : : * clause as soon as it's supposed to be applied.
165 : : *
166 : : * On success return, we have also initialized the clause's left_ec/right_ec
167 : : * fields to point to the EquivalenceClass representing it. This saves lookup
168 : : * effort later.
169 : : *
170 : : * Note: constructing merged EquivalenceClasses is a standard UNION-FIND
171 : : * problem, for which there exist better data structures than simple lists.
172 : : * If this code ever proves to be a bottleneck then it could be sped up ---
173 : : * but for now, simple is beautiful.
174 : : *
175 : : * Note: this is only called during planner startup, not during GEQO
176 : : * exploration, so we need not worry about whether we're in the right
177 : : * memory context.
178 : : */
179 : : bool
180 : 242065 : process_equivalence(PlannerInfo *root,
181 : : RestrictInfo **p_restrictinfo,
182 : : JoinDomain *jdomain)
183 : : {
184 : 242065 : RestrictInfo *restrictinfo = *p_restrictinfo;
185 : 242065 : Expr *clause = restrictinfo->clause;
186 : : Oid opno,
187 : : collation,
188 : : item1_type,
189 : : item2_type;
190 : : Expr *item1;
191 : : Expr *item2;
192 : : Relids item1_relids,
193 : : item2_relids;
194 : : List *opfamilies;
195 : : EquivalenceClass *ec1,
196 : : *ec2;
197 : : EquivalenceMember *em1,
198 : : *em2;
199 : : ListCell *lc1;
200 : : int ec2_idx;
201 : :
202 : : /* Should not already be marked as having generated an eclass */
203 : : Assert(restrictinfo->left_ec == NULL);
204 : : Assert(restrictinfo->right_ec == NULL);
205 : :
206 : : /* Reject if it is potentially postponable by security considerations */
207 [ + + + + ]: 242065 : if (restrictinfo->security_level > 0 && !restrictinfo->leakproof)
208 : 172 : return false;
209 : :
210 : : /* Extract info from given clause */
211 : : Assert(is_opclause(clause));
212 : 241893 : opno = ((OpExpr *) clause)->opno;
213 : 241893 : collation = ((OpExpr *) clause)->inputcollid;
214 : 241893 : item1 = (Expr *) get_leftop(clause);
215 : 241893 : item2 = (Expr *) get_rightop(clause);
216 : 241893 : item1_relids = restrictinfo->left_relids;
217 : 241893 : item2_relids = restrictinfo->right_relids;
218 : :
219 : : /*
220 : : * Ensure both input expressions expose the desired collation (their types
221 : : * should be OK already); see comments for canonicalize_ec_expression.
222 : : */
223 : 241893 : item1 = canonicalize_ec_expression(item1,
224 : : exprType((Node *) item1),
225 : : collation);
226 : 241893 : item2 = canonicalize_ec_expression(item2,
227 : : exprType((Node *) item2),
228 : : collation);
229 : :
230 : : /*
231 : : * Clauses of the form X=X cannot be translated into EquivalenceClasses.
232 : : * We'd either end up with a single-entry EC, losing the knowledge that
233 : : * the clause was present at all, or else make an EC with duplicate
234 : : * entries, causing other issues.
235 : : */
236 [ + + ]: 241893 : if (equal(item1, item2))
237 : : {
238 : : /*
239 : : * If the operator is strict, then the clause can be treated as just
240 : : * "X IS NOT NULL". (Since we know we are considering a top-level
241 : : * qual, we can ignore the difference between FALSE and NULL results.)
242 : : * It's worth making the conversion because we'll typically get a much
243 : : * better selectivity estimate than we would for X=X.
244 : : *
245 : : * If the operator is not strict, we can't be sure what it will do
246 : : * with NULLs, so don't attempt to optimize it.
247 : : */
248 : 45 : set_opfuncid((OpExpr *) clause);
249 [ + - ]: 45 : if (func_strict(((OpExpr *) clause)->opfuncid))
250 : : {
251 : 45 : NullTest *ntest = makeNode(NullTest);
252 : :
253 : 45 : ntest->arg = item1;
254 : 45 : ntest->nulltesttype = IS_NOT_NULL;
255 : 45 : ntest->argisrow = false; /* correct even if composite arg */
256 : 45 : ntest->location = -1;
257 : :
258 : 45 : *p_restrictinfo =
259 : 45 : make_restrictinfo(root,
260 : : (Expr *) ntest,
261 : 45 : restrictinfo->is_pushed_down,
262 : 45 : restrictinfo->has_clone,
263 : 45 : restrictinfo->is_clone,
264 : 45 : restrictinfo->pseudoconstant,
265 : : restrictinfo->security_level,
266 : : NULL,
267 : : restrictinfo->incompatible_relids,
268 : : restrictinfo->outer_relids);
269 : : }
270 : 45 : return false;
271 : : }
272 : :
273 : : /*
274 : : * We use the declared input types of the operator, not exprType() of the
275 : : * inputs, as the nominal datatypes for opfamily lookup. This presumes
276 : : * that btree operators are always registered with amoplefttype and
277 : : * amoprighttype equal to their declared input types. We will need this
278 : : * info anyway to build EquivalenceMember nodes, and by extracting it now
279 : : * we can use type comparisons to short-circuit some equal() tests.
280 : : */
281 : 241848 : op_input_types(opno, &item1_type, &item2_type);
282 : :
283 : 241848 : opfamilies = restrictinfo->mergeopfamilies;
284 : :
285 : : /*
286 : : * Sweep through the existing EquivalenceClasses looking for matches to
287 : : * item1 and item2. These are the possible outcomes:
288 : : *
289 : : * 1. We find both in the same EC. The equivalence is already known, so
290 : : * there's nothing to do.
291 : : *
292 : : * 2. We find both in different ECs. Merge the two ECs together.
293 : : *
294 : : * 3. We find just one. Add the other to its EC.
295 : : *
296 : : * 4. We find neither. Make a new, two-entry EC.
297 : : *
298 : : * Note: since all ECs are built through this process or the similar
299 : : * search in get_eclass_for_sort_expr(), it's impossible that we'd match
300 : : * an item in more than one existing nonvolatile EC. So it's okay to stop
301 : : * at the first match.
302 : : */
303 : 241848 : ec1 = ec2 = NULL;
304 : 241848 : em1 = em2 = NULL;
305 : 241848 : ec2_idx = -1;
306 [ + + + + : 416023 : foreach(lc1, root->eq_classes)
+ + ]
307 : : {
308 : 174253 : EquivalenceClass *cur_ec = (EquivalenceClass *) lfirst(lc1);
309 : : ListCell *lc2;
310 : :
311 : : /* Never match to a volatile EC */
312 [ - + ]: 174253 : if (cur_ec->ec_has_volatile)
313 : 0 : continue;
314 : :
315 : : /*
316 : : * The collation has to match; check this first since it's cheaper
317 : : * than the opfamily comparison.
318 : : */
319 [ + + ]: 174253 : if (collation != cur_ec->ec_collation)
320 : 17668 : continue;
321 : :
322 : : /*
323 : : * A "match" requires matching sets of btree opfamilies. Use of
324 : : * equal() for this test has implications discussed in the comments
325 : : * for get_mergejoin_opfamilies().
326 : : */
327 [ + + ]: 156585 : if (!equal(opfamilies, cur_ec->ec_opfamilies))
328 : 42622 : continue;
329 : :
330 : : /* We don't expect any children yet */
331 : : Assert(cur_ec->ec_childmembers == NULL);
332 : :
333 [ + - + + : 339101 : foreach(lc2, cur_ec->ec_members)
+ + ]
334 : : {
335 : 225216 : EquivalenceMember *cur_em = (EquivalenceMember *) lfirst(lc2);
336 : :
337 : : /* Child members should not exist in ec_members */
338 : : Assert(!cur_em->em_is_child);
339 : :
340 : : /*
341 : : * Match constants only within the same JoinDomain (see
342 : : * optimizer/README).
343 : : */
344 [ + + + + ]: 225216 : if (cur_em->em_is_const && cur_em->em_jdomain != jdomain)
345 : 3057 : continue;
346 : :
347 [ + + ]: 222159 : if (!ec1 &&
348 [ + + + + ]: 427773 : item1_type == cur_em->em_datatype &&
349 : 213718 : equal(item1, cur_em->em_expr))
350 : : {
351 : 12995 : ec1 = cur_ec;
352 : 12995 : em1 = cur_em;
353 [ + + ]: 12995 : if (ec2)
354 : 60 : break;
355 : : }
356 : :
357 [ + + ]: 222099 : if (!ec2 &&
358 [ + + + + ]: 439656 : item2_type == cur_em->em_datatype &&
359 : 219507 : equal(item2, cur_em->em_expr))
360 : : {
361 : 2979 : ec2 = cur_ec;
362 : 2979 : ec2_idx = foreach_current_index(lc1);
363 : 2979 : em2 = cur_em;
364 [ + + ]: 2979 : if (ec1)
365 : 18 : break;
366 : : }
367 : : }
368 : :
369 [ + + + + ]: 113963 : if (ec1 && ec2)
370 : 78 : break;
371 : : }
372 : :
373 : : /* Sweep finished, what did we find? */
374 : :
375 [ + + + + ]: 241848 : if (ec1 && ec2)
376 : : {
377 : : /* If case 1, nothing to do, except add to sources */
378 [ + + ]: 78 : if (ec1 == ec2)
379 : : {
380 : 45 : ec1->ec_sources = lappend(ec1->ec_sources, restrictinfo);
381 : 45 : ec1->ec_min_security = Min(ec1->ec_min_security,
382 : : restrictinfo->security_level);
383 : 45 : ec1->ec_max_security = Max(ec1->ec_max_security,
384 : : restrictinfo->security_level);
385 : : /* mark the RI as associated with this eclass */
386 : 45 : restrictinfo->left_ec = ec1;
387 : 45 : restrictinfo->right_ec = ec1;
388 : : /* mark the RI as usable with this pair of EMs */
389 : 45 : restrictinfo->left_em = em1;
390 : 45 : restrictinfo->right_em = em2;
391 : 45 : return true;
392 : : }
393 : :
394 : : /*
395 : : * Case 2: need to merge ec1 and ec2. This should never happen after
396 : : * the ECs have reached canonical state; otherwise, pathkeys could be
397 : : * rendered non-canonical by the merge, and relation eclass indexes
398 : : * would get broken by removal of an eq_classes list entry.
399 : : */
400 [ - + ]: 33 : if (root->ec_merging_done)
401 [ # # ]: 0 : elog(ERROR, "too late to merge equivalence classes");
402 : :
403 : : /*
404 : : * We add ec2's items to ec1, then set ec2's ec_merged link to point
405 : : * to ec1 and remove ec2 from the eq_classes list. We cannot simply
406 : : * delete ec2 because that could leave dangling pointers in existing
407 : : * PathKeys. We leave it behind with a link so that the merged EC can
408 : : * be found.
409 : : */
410 : 33 : ec1->ec_members = list_concat(ec1->ec_members, ec2->ec_members);
411 : 33 : ec1->ec_sources = list_concat(ec1->ec_sources, ec2->ec_sources);
412 : :
413 : : /*
414 : : * Appends ec2's derived clauses to ec1->ec_derives_list and adds them
415 : : * to ec1->ec_derives_hash if present.
416 : : */
417 : 33 : ec_add_derived_clauses(ec1, ec2->ec_derives_list);
418 : 33 : ec1->ec_relids = bms_join(ec1->ec_relids, ec2->ec_relids);
419 : 33 : ec1->ec_has_const |= ec2->ec_has_const;
420 : : /* can't need to set has_volatile */
421 : 33 : ec1->ec_min_security = Min(ec1->ec_min_security,
422 : : ec2->ec_min_security);
423 : 33 : ec1->ec_max_security = Max(ec1->ec_max_security,
424 : : ec2->ec_max_security);
425 : 33 : ec2->ec_merged = ec1;
426 : 33 : root->eq_classes = list_delete_nth_cell(root->eq_classes, ec2_idx);
427 : : /* just to avoid debugging confusion w/ dangling pointers: */
428 : 33 : ec2->ec_members = NIL;
429 : 33 : ec2->ec_sources = NIL;
430 : 33 : ec_clear_derived_clauses(ec2);
431 : 33 : ec2->ec_relids = NULL;
432 : 33 : ec1->ec_sources = lappend(ec1->ec_sources, restrictinfo);
433 : 33 : ec1->ec_min_security = Min(ec1->ec_min_security,
434 : : restrictinfo->security_level);
435 : 33 : ec1->ec_max_security = Max(ec1->ec_max_security,
436 : : restrictinfo->security_level);
437 : : /* mark the RI as associated with this eclass */
438 : 33 : restrictinfo->left_ec = ec1;
439 : 33 : restrictinfo->right_ec = ec1;
440 : : /* mark the RI as usable with this pair of EMs */
441 : 33 : restrictinfo->left_em = em1;
442 : 33 : restrictinfo->right_em = em2;
443 : : }
444 [ + + ]: 241770 : else if (ec1)
445 : : {
446 : : /* Case 3: add item2 to ec1 */
447 : 12917 : em2 = add_eq_member(ec1, item2, item2_relids,
448 : : jdomain, item2_type);
449 : 12917 : ec1->ec_sources = lappend(ec1->ec_sources, restrictinfo);
450 : 12917 : ec1->ec_min_security = Min(ec1->ec_min_security,
451 : : restrictinfo->security_level);
452 : 12917 : ec1->ec_max_security = Max(ec1->ec_max_security,
453 : : restrictinfo->security_level);
454 : : /* mark the RI as associated with this eclass */
455 : 12917 : restrictinfo->left_ec = ec1;
456 : 12917 : restrictinfo->right_ec = ec1;
457 : : /* mark the RI as usable with this pair of EMs */
458 : 12917 : restrictinfo->left_em = em1;
459 : 12917 : restrictinfo->right_em = em2;
460 : : }
461 [ + + ]: 228853 : else if (ec2)
462 : : {
463 : : /* Case 3: add item1 to ec2 */
464 : 2901 : em1 = add_eq_member(ec2, item1, item1_relids,
465 : : jdomain, item1_type);
466 : 2901 : ec2->ec_sources = lappend(ec2->ec_sources, restrictinfo);
467 : 2901 : ec2->ec_min_security = Min(ec2->ec_min_security,
468 : : restrictinfo->security_level);
469 : 2901 : ec2->ec_max_security = Max(ec2->ec_max_security,
470 : : restrictinfo->security_level);
471 : : /* mark the RI as associated with this eclass */
472 : 2901 : restrictinfo->left_ec = ec2;
473 : 2901 : restrictinfo->right_ec = ec2;
474 : : /* mark the RI as usable with this pair of EMs */
475 : 2901 : restrictinfo->left_em = em1;
476 : 2901 : restrictinfo->right_em = em2;
477 : : }
478 : : else
479 : : {
480 : : /* Case 4: make a new, two-entry EC */
481 : 225952 : EquivalenceClass *ec = makeNode(EquivalenceClass);
482 : :
483 : 225952 : ec->ec_opfamilies = opfamilies;
484 : 225952 : ec->ec_collation = collation;
485 : 225952 : ec->ec_childmembers_size = 0;
486 : 225952 : ec->ec_members = NIL;
487 : 225952 : ec->ec_childmembers = NULL;
488 : 225952 : ec->ec_sources = list_make1(restrictinfo);
489 : 225952 : ec->ec_derives_list = NIL;
490 : 225952 : ec->ec_derives_hash = NULL;
491 : 225952 : ec->ec_relids = NULL;
492 : 225952 : ec->ec_has_const = false;
493 : 225952 : ec->ec_has_volatile = false;
494 : 225952 : ec->ec_broken = false;
495 : 225952 : ec->ec_sortref = 0;
496 : 225952 : ec->ec_min_security = restrictinfo->security_level;
497 : 225952 : ec->ec_max_security = restrictinfo->security_level;
498 : 225952 : ec->ec_merged = NULL;
499 : 225952 : em1 = add_eq_member(ec, item1, item1_relids,
500 : : jdomain, item1_type);
501 : 225952 : em2 = add_eq_member(ec, item2, item2_relids,
502 : : jdomain, item2_type);
503 : :
504 : 225952 : root->eq_classes = lappend(root->eq_classes, ec);
505 : :
506 : : /* mark the RI as associated with this eclass */
507 : 225952 : restrictinfo->left_ec = ec;
508 : 225952 : restrictinfo->right_ec = ec;
509 : : /* mark the RI as usable with this pair of EMs */
510 : 225952 : restrictinfo->left_em = em1;
511 : 225952 : restrictinfo->right_em = em2;
512 : : }
513 : :
514 : 241803 : return true;
515 : : }
516 : :
517 : : /*
518 : : * canonicalize_ec_expression
519 : : *
520 : : * This function ensures that the expression exposes the expected type and
521 : : * collation, so that it will be equal() to other equivalence-class expressions
522 : : * that it ought to be equal() to.
523 : : *
524 : : * The rule for datatypes is that the exposed type should match what it would
525 : : * be for an input to an operator of the EC's opfamilies; which is usually
526 : : * the declared input type of the operator, but in the case of polymorphic
527 : : * operators no relabeling is wanted (compare the behavior of parse_coerce.c).
528 : : * Expressions coming in from quals will generally have the right type
529 : : * already, but expressions coming from indexkeys may not (because they are
530 : : * represented without any explicit relabel in pg_index), and the same problem
531 : : * occurs for sort expressions (because the parser is likewise cavalier about
532 : : * putting relabels on them). Such cases will be binary-compatible with the
533 : : * real operators, so adding a RelabelType is sufficient.
534 : : *
535 : : * Also, the expression's exposed collation must match the EC's collation.
536 : : * This is important because in comparisons like "foo < bar COLLATE baz",
537 : : * only one of the expressions has the correct exposed collation as we receive
538 : : * it from the parser. Forcing both of them to have it ensures that all
539 : : * variant spellings of such a construct behave the same. Again, we can
540 : : * stick on a RelabelType to force the right exposed collation. (It might
541 : : * work to not label the collation at all in EC members, but this is risky
542 : : * since some parts of the system expect exprCollation() to deliver the
543 : : * right answer for a sort key.)
544 : : */
545 : : Expr *
546 : 2390612 : canonicalize_ec_expression(Expr *expr, Oid req_type, Oid req_collation)
547 : : {
548 : 2390612 : Oid expr_type = exprType((Node *) expr);
549 : :
550 : : /*
551 : : * For a polymorphic-input-type opclass, just keep the same exposed type.
552 : : * RECORD opclasses work like polymorphic-type ones for this purpose.
553 : : */
554 [ + - + + : 2390612 : if (IsPolymorphicType(req_type) || req_type == RECORDOID)
+ - + + +
+ + + + -
+ - + - +
- + - +
+ ]
555 : 9658 : req_type = expr_type;
556 : :
557 : : /*
558 : : * No work if the expression exposes the right type/collation already.
559 : : */
560 [ + + + + ]: 4700281 : if (expr_type != req_type ||
561 : 2309669 : exprCollation((Node *) expr) != req_collation)
562 : : {
563 : : /*
564 : : * If we have to change the type of the expression, set typmod to -1,
565 : : * since the new type may not have the same typmod interpretation.
566 : : * When we only have to change collation, preserve the exposed typmod.
567 : : */
568 : : int32 req_typmod;
569 : :
570 [ + + ]: 83127 : if (expr_type != req_type)
571 : 80943 : req_typmod = -1;
572 : : else
573 : 2184 : req_typmod = exprTypmod((Node *) expr);
574 : :
575 : : /*
576 : : * Use applyRelabelType so that we preserve const-flatness. This is
577 : : * important since eval_const_expressions has already been applied.
578 : : */
579 : 83127 : expr = (Expr *) applyRelabelType((Node *) expr,
580 : : req_type, req_typmod, req_collation,
581 : : COERCE_IMPLICIT_CAST, -1, false);
582 : : }
583 : :
584 : 2390612 : return expr;
585 : : }
586 : :
587 : : /*
588 : : * make_eq_member
589 : : * Build a new EquivalenceMember without adding it to an EC. If 'parent'
590 : : * is NULL, the result will be a parent member, otherwise a child member.
591 : : */
592 : : static EquivalenceMember *
593 : 776321 : make_eq_member(EquivalenceClass *ec, Expr *expr, Relids relids,
594 : : JoinDomain *jdomain, EquivalenceMember *parent, Oid datatype)
595 : : {
596 : 776321 : EquivalenceMember *em = makeNode(EquivalenceMember);
597 : :
598 : 776321 : em->em_expr = expr;
599 : 776321 : em->em_relids = relids;
600 : 776321 : em->em_is_const = false;
601 : 776321 : em->em_is_child = (parent != NULL);
602 : 776321 : em->em_datatype = datatype;
603 : 776321 : em->em_jdomain = jdomain;
604 : 776321 : em->em_parent = parent;
605 : :
606 [ + + ]: 776321 : if (bms_is_empty(relids))
607 : : {
608 : : /*
609 : : * No Vars, assume it's a pseudoconstant. This is correct for entries
610 : : * generated from process_equivalence(), because a WHERE clause can't
611 : : * contain aggregates or SRFs, and non-volatility was checked before
612 : : * process_equivalence() ever got called. But
613 : : * get_eclass_for_sort_expr() has to work harder. We put the tests
614 : : * there not here to save cycles in the equivalence case.
615 : : */
616 : : Assert(!parent);
617 : 172631 : em->em_is_const = true;
618 : 172631 : ec->ec_has_const = true;
619 : : /* it can't affect ec_relids */
620 : : }
621 : :
622 : 776321 : return em;
623 : : }
624 : :
625 : : /*
626 : : * add_eq_member - build a new non-child EquivalenceMember and add it to 'ec'.
627 : : */
628 : : static EquivalenceMember *
629 : 692368 : add_eq_member(EquivalenceClass *ec, Expr *expr, Relids relids,
630 : : JoinDomain *jdomain, Oid datatype)
631 : : {
632 : 692368 : EquivalenceMember *em = make_eq_member(ec, expr, relids, jdomain,
633 : : NULL, datatype);
634 : :
635 : : /* add to the members list */
636 : 692368 : ec->ec_members = lappend(ec->ec_members, em);
637 : :
638 : : /* record the relids for parent members */
639 : 692368 : ec->ec_relids = bms_add_members(ec->ec_relids, relids);
640 : :
641 : 692368 : return em;
642 : : }
643 : :
644 : : /*
645 : : * add_child_eq_member
646 : : * Create an em_is_child=true EquivalenceMember and add it to 'ec'.
647 : : *
648 : : * 'root' is the PlannerInfo that 'ec' belongs to.
649 : : * 'ec' is the EquivalenceClass to add the child member to.
650 : : * 'ec_index' the index of 'ec' within root->eq_classes, or -1 if maintaining
651 : : * the RelOptInfo.eclass_indexes isn't needed.
652 : : * 'expr' is the em_expr for the new member.
653 : : * 'relids' is the 'em_relids' for the new member.
654 : : * 'jdomain' is the 'em_jdomain' for the new member.
655 : : * 'parent_em' is the parent member of the child to create.
656 : : * 'datatype' is the em_datatype of the new member.
657 : : * 'child_relid' defines which element of ec_childmembers to add this member
658 : : * to. This is generally a RELOPT_OTHER_MEMBER_REL, but for set operations
659 : : * can be a RELOPT_BASEREL representing the set-op children.
660 : : */
661 : : static EquivalenceMember *
662 : 83953 : add_child_eq_member(PlannerInfo *root, EquivalenceClass *ec, int ec_index,
663 : : Expr *expr, Relids relids, JoinDomain *jdomain,
664 : : EquivalenceMember *parent_em, Oid datatype,
665 : : Index child_relid)
666 : : {
667 : : EquivalenceMember *em;
668 : :
669 : : Assert(parent_em != NULL);
670 : :
671 : : /*
672 : : * Allocate the array to store child members; an array of Lists indexed by
673 : : * relid, or expand the existing one, if necessary.
674 : : */
675 [ + + ]: 83953 : if (unlikely(ec->ec_childmembers_size < root->simple_rel_array_size))
676 : : {
677 [ + - ]: 23840 : if (ec->ec_childmembers == NULL)
678 : 23840 : ec->ec_childmembers = palloc0_array(List *, root->simple_rel_array_size);
679 : : else
680 : 0 : ec->ec_childmembers = repalloc0_array(ec->ec_childmembers, List *,
681 : : ec->ec_childmembers_size,
682 : : root->simple_rel_array_size);
683 : :
684 : 23840 : ec->ec_childmembers_size = root->simple_rel_array_size;
685 : : }
686 : :
687 : 83953 : em = make_eq_member(ec, expr, relids, jdomain, parent_em, datatype);
688 : :
689 : : /* add member to the ec_childmembers List for the given child_relid */
690 : 83953 : ec->ec_childmembers[child_relid] = lappend(ec->ec_childmembers[child_relid], em);
691 : :
692 : : /* Record this EC index for the child rel */
693 [ + + ]: 83953 : if (ec_index >= 0)
694 : : {
695 : 49663 : RelOptInfo *child_rel = root->simple_rel_array[child_relid];
696 : :
697 : 49663 : child_rel->eclass_indexes =
698 : 49663 : bms_add_member(child_rel->eclass_indexes, ec_index);
699 : : }
700 : :
701 : 83953 : return em;
702 : : }
703 : :
704 : :
705 : : /*
706 : : * get_eclass_for_sort_expr
707 : : * Given an expression and opfamily/collation info, find an existing
708 : : * equivalence class it is a member of; if none, optionally build a new
709 : : * single-member EquivalenceClass for it.
710 : : *
711 : : * sortref is the SortGroupRef of the originating SortGroupClause, if any,
712 : : * or zero if not. (It should never be zero if the expression is volatile!)
713 : : *
714 : : * If rel is not NULL, it identifies a specific relation we're considering
715 : : * a path for, and indicates that child EC members for that relation can be
716 : : * considered. Otherwise child members are ignored. (Note: since child EC
717 : : * members aren't guaranteed unique, a non-NULL value means that there could
718 : : * be more than one EC that matches the expression; if so it's order-dependent
719 : : * which one you get. This is annoying but it only happens in corner cases,
720 : : * so for now we live with just reporting the first match. See also
721 : : * generate_implied_equalities_for_column and match_pathkeys_to_index.)
722 : : *
723 : : * If create_it is true, we'll build a new EquivalenceClass when there is no
724 : : * match. If create_it is false, we just return NULL when no match.
725 : : *
726 : : * This can be used safely both before and after EquivalenceClass merging;
727 : : * since it never causes merging it does not invalidate any existing ECs
728 : : * or PathKeys. However, ECs added after path generation has begun are
729 : : * of limited usefulness, so usually it's best to create them beforehand.
730 : : *
731 : : * Note: opfamilies must be chosen consistently with the way
732 : : * process_equivalence() would do; that is, generated from a mergejoinable
733 : : * equality operator. Else we might fail to detect valid equivalences,
734 : : * generating poor (but not incorrect) plans.
735 : : */
736 : : EquivalenceClass *
737 : 1749453 : get_eclass_for_sort_expr(PlannerInfo *root,
738 : : Expr *expr,
739 : : List *opfamilies,
740 : : Oid opcintype,
741 : : Oid collation,
742 : : Index sortref,
743 : : Relids relids,
744 : : bool create_it)
745 : : {
746 : : JoinDomain *jdomain;
747 : : Relids expr_relids;
748 : : EquivalenceClass *newec;
749 : : EquivalenceMember *newem;
750 : : ListCell *lc1;
751 : : MemoryContext oldcontext;
752 : :
753 : : /*
754 : : * Ensure the expression exposes the correct type and collation.
755 : : */
756 : 1749453 : expr = canonicalize_ec_expression(expr, opcintype, collation);
757 : :
758 : : /*
759 : : * Since SortGroupClause nodes are top-level expressions (GROUP BY, ORDER
760 : : * BY, etc), they can be presumed to belong to the top JoinDomain.
761 : : */
762 : 1749453 : jdomain = linitial_node(JoinDomain, root->join_domains);
763 : :
764 : : /*
765 : : * Scan through the existing EquivalenceClasses for a match
766 : : */
767 [ + + + + : 5822583 : foreach(lc1, root->eq_classes)
+ + ]
768 : : {
769 : 5085095 : EquivalenceClass *cur_ec = (EquivalenceClass *) lfirst(lc1);
770 : : EquivalenceMemberIterator it;
771 : : EquivalenceMember *cur_em;
772 : :
773 : : /*
774 : : * Never match to a volatile EC, except when we are looking at another
775 : : * reference to the same volatile SortGroupClause.
776 : : */
777 [ + + + + ]: 5085095 : if (cur_ec->ec_has_volatile &&
778 [ + + ]: 28 : (sortref == 0 || sortref != cur_ec->ec_sortref))
779 : 2198200 : continue;
780 : :
781 [ + + ]: 5084585 : if (collation != cur_ec->ec_collation)
782 : 1430369 : continue;
783 [ + + ]: 3654216 : if (!equal(opfamilies, cur_ec->ec_opfamilies))
784 : 767321 : continue;
785 : :
786 : 2886895 : setup_eclass_member_iterator(&it, cur_ec, relids);
787 [ + + ]: 6411038 : while ((cur_em = eclass_member_iterator_next(&it)) != NULL)
788 : : {
789 : : /*
790 : : * Ignore child members unless they match the request.
791 : : */
792 [ + + ]: 4536108 : if (cur_em->em_is_child &&
793 [ - + ]: 89186 : !bms_equal(cur_em->em_relids, relids))
794 : 0 : continue;
795 : :
796 : : /*
797 : : * Match constants only within the same JoinDomain (see
798 : : * optimizer/README).
799 : : */
800 [ + + + + ]: 4536108 : if (cur_em->em_is_const && cur_em->em_jdomain != jdomain)
801 : 57397 : continue;
802 : :
803 [ + + + + ]: 8921614 : if (opcintype == cur_em->em_datatype &&
804 : 4442903 : equal(expr, cur_em->em_expr))
805 : 1011965 : return cur_ec; /* Match! */
806 : : }
807 : : }
808 : :
809 : : /* No match; does caller want a NULL result? */
810 [ + + ]: 737488 : if (!create_it)
811 : 512842 : return NULL;
812 : :
813 : : /*
814 : : * OK, build a new single-member EC
815 : : *
816 : : * Here, we must be sure that we construct the EC in the right context.
817 : : */
818 : 224646 : oldcontext = MemoryContextSwitchTo(root->planner_cxt);
819 : :
820 : 224646 : newec = makeNode(EquivalenceClass);
821 : 224646 : newec->ec_opfamilies = list_copy(opfamilies);
822 : 224646 : newec->ec_collation = collation;
823 : 224646 : newec->ec_childmembers_size = 0;
824 : 224646 : newec->ec_members = NIL;
825 : 224646 : newec->ec_childmembers = NULL;
826 : 224646 : newec->ec_sources = NIL;
827 : 224646 : newec->ec_derives_list = NIL;
828 : 224646 : newec->ec_derives_hash = NULL;
829 : 224646 : newec->ec_relids = NULL;
830 : 224646 : newec->ec_has_const = false;
831 : 224646 : newec->ec_has_volatile = contain_volatile_functions((Node *) expr);
832 : 224646 : newec->ec_broken = false;
833 : 224646 : newec->ec_sortref = sortref;
834 : 224646 : newec->ec_min_security = UINT_MAX;
835 : 224646 : newec->ec_max_security = 0;
836 : 224646 : newec->ec_merged = NULL;
837 : :
838 [ + + - + ]: 224646 : if (newec->ec_has_volatile && sortref == 0) /* should not happen */
839 [ # # ]: 0 : elog(ERROR, "volatile EquivalenceClass has no sortref");
840 : :
841 : : /*
842 : : * Get the precise set of relids appearing in the expression.
843 : : */
844 : 224646 : expr_relids = pull_varnos(root, (Node *) expr);
845 : :
846 : 224646 : newem = add_eq_member(newec, copyObject(expr), expr_relids,
847 : : jdomain, opcintype);
848 : :
849 : : /*
850 : : * add_eq_member doesn't check for volatile functions, set-returning
851 : : * functions, aggregates, or window functions, but such could appear in
852 : : * sort expressions; so we have to check whether its const-marking was
853 : : * correct.
854 : : */
855 [ + + ]: 224646 : if (newec->ec_has_const)
856 : : {
857 [ + + + + ]: 15402 : if (newec->ec_has_volatile ||
858 [ + + ]: 15174 : expression_returns_set((Node *) expr) ||
859 [ + + ]: 14903 : contain_agg_clause((Node *) expr) ||
860 : 7376 : contain_window_function((Node *) expr))
861 : : {
862 : 384 : newec->ec_has_const = false;
863 : 384 : newem->em_is_const = false;
864 : : }
865 : : }
866 : :
867 : 224646 : root->eq_classes = lappend(root->eq_classes, newec);
868 : :
869 : : /*
870 : : * If EC merging is already complete, we have to mop up by adding the new
871 : : * EC to the eclass_indexes of the relation(s) mentioned in it.
872 : : */
873 [ + + ]: 224646 : if (root->ec_merging_done)
874 : : {
875 : 130948 : int ec_index = list_length(root->eq_classes) - 1;
876 : 130948 : int i = -1;
877 : :
878 [ + + ]: 252202 : while ((i = bms_next_member(newec->ec_relids, i)) > 0)
879 : : {
880 : 121254 : RelOptInfo *rel = root->simple_rel_array[i];
881 : :
882 : : /* ignore the RTE_GROUP RTE */
883 [ + + ]: 121254 : if (i == root->group_rtindex)
884 : 584 : continue;
885 : :
886 [ + + ]: 120670 : if (rel == NULL) /* must be an outer join */
887 : : {
888 : : Assert(bms_is_member(i, root->outer_join_rels));
889 : 5432 : continue;
890 : : }
891 : :
892 : : Assert(rel->reloptkind == RELOPT_BASEREL);
893 : :
894 : 115238 : rel->eclass_indexes = bms_add_member(rel->eclass_indexes,
895 : : ec_index);
896 : : }
897 : : }
898 : :
899 : 224646 : MemoryContextSwitchTo(oldcontext);
900 : :
901 : 224646 : return newec;
902 : : }
903 : :
904 : : /*
905 : : * find_ec_member_matching_expr
906 : : * Locate an EquivalenceClass member matching the given expr, if any;
907 : : * return NULL if no match.
908 : : *
909 : : * "Matching" is defined as "equal after stripping RelabelTypes".
910 : : * This is used for identifying sort expressions, and we need to allow
911 : : * binary-compatible relabeling for some cases involving binary-compatible
912 : : * sort operators.
913 : : *
914 : : * Child EC members are ignored unless they belong to given 'relids'.
915 : : */
916 : : EquivalenceMember *
917 : 263897 : find_ec_member_matching_expr(EquivalenceClass *ec,
918 : : Expr *expr,
919 : : Relids relids)
920 : : {
921 : : EquivalenceMemberIterator it;
922 : : EquivalenceMember *em;
923 : :
924 : : /* We ignore binary-compatible relabeling on both ends */
925 [ + - + + ]: 285404 : while (expr && IsA(expr, RelabelType))
926 : 21507 : expr = ((RelabelType *) expr)->arg;
927 : :
928 : 263897 : setup_eclass_member_iterator(&it, ec, relids);
929 [ + + ]: 442168 : while ((em = eclass_member_iterator_next(&it)) != NULL)
930 : : {
931 : : Expr *emexpr;
932 : :
933 : : /*
934 : : * We shouldn't be trying to sort by an equivalence class that
935 : : * contains a constant, so no need to consider such cases any further.
936 : : */
937 [ - + ]: 293795 : if (em->em_is_const)
938 : 0 : continue;
939 : :
940 : : /*
941 : : * Ignore child members unless they belong to the requested rel.
942 : : */
943 [ + + ]: 293795 : if (em->em_is_child &&
944 [ + + ]: 10717 : !bms_is_subset(em->em_relids, relids))
945 : 3576 : continue;
946 : :
947 : : /*
948 : : * Match if same expression (after stripping relabel).
949 : : */
950 : 290219 : emexpr = em->em_expr;
951 [ + - + + ]: 295182 : while (emexpr && IsA(emexpr, RelabelType))
952 : 4963 : emexpr = ((RelabelType *) emexpr)->arg;
953 : :
954 [ + + ]: 290219 : if (equal(emexpr, expr))
955 : 115524 : return em;
956 : : }
957 : :
958 : 148373 : return NULL;
959 : : }
960 : :
961 : : /*
962 : : * find_computable_ec_member
963 : : * Locate an EquivalenceClass member that can be computed from the
964 : : * expressions appearing in "exprs"; return NULL if no match.
965 : : *
966 : : * "exprs" can be either a list of bare expression trees, or a list of
967 : : * TargetEntry nodes. Typically it will contain Vars and possibly Aggrefs
968 : : * and WindowFuncs; however, when considering an appendrel member the list
969 : : * could contain arbitrary expressions. We consider an EC member to be
970 : : * computable if all the Vars, PlaceHolderVars, Aggrefs, and WindowFuncs
971 : : * it needs are present in "exprs".
972 : : *
973 : : * There is some subtlety in that definition: for example, if an EC member is
974 : : * Var_A + 1 while what is in "exprs" is Var_A + 2, it's still computable.
975 : : * This works because in the final plan tree, the EC member's expression will
976 : : * be computed as part of the same plan node targetlist that is currently
977 : : * represented by "exprs". So if we have Var_A available for the existing
978 : : * tlist member, it must be OK to use it in the EC expression too.
979 : : *
980 : : * Unlike find_ec_member_matching_expr, there's no special provision here
981 : : * for binary-compatible relabeling. This is intentional: if we have to
982 : : * compute an expression in this way, setrefs.c is going to insist on exact
983 : : * matches of Vars to the source tlist.
984 : : *
985 : : * Child EC members are ignored unless they belong to given 'relids'.
986 : : * Also, non-parallel-safe expressions are ignored if 'require_parallel_safe'.
987 : : *
988 : : * Note: some callers pass root == NULL for notational reasons. This is OK
989 : : * when require_parallel_safe is false.
990 : : */
991 : : EquivalenceMember *
992 : 6927 : find_computable_ec_member(PlannerInfo *root,
993 : : EquivalenceClass *ec,
994 : : List *exprs,
995 : : Relids relids,
996 : : bool require_parallel_safe)
997 : : {
998 : : List *exprvars;
999 : : EquivalenceMemberIterator it;
1000 : : EquivalenceMember *em;
1001 : :
1002 : : /*
1003 : : * Pull out the Vars and quasi-Vars present in "exprs". In the typical
1004 : : * non-appendrel case, this is just another representation of the same
1005 : : * list. However, it does remove the distinction between the case of a
1006 : : * list of plain expressions and a list of TargetEntrys.
1007 : : */
1008 : 6927 : exprvars = pull_var_clause((Node *) exprs,
1009 : : PVC_INCLUDE_AGGREGATES |
1010 : : PVC_INCLUDE_WINDOWFUNCS |
1011 : : PVC_INCLUDE_PLACEHOLDERS);
1012 : :
1013 : 6927 : setup_eclass_member_iterator(&it, ec, relids);
1014 [ + + ]: 13952 : while ((em = eclass_member_iterator_next(&it)) != NULL)
1015 : : {
1016 : : List *emvars;
1017 : : ListCell *lc2;
1018 : :
1019 : : /*
1020 : : * We shouldn't be trying to sort by an equivalence class that
1021 : : * contains a constant, so no need to consider such cases any further.
1022 : : */
1023 [ - + ]: 7415 : if (em->em_is_const)
1024 : 0 : continue;
1025 : :
1026 : : /*
1027 : : * Ignore child members unless they belong to the requested rel.
1028 : : */
1029 [ + + ]: 7415 : if (em->em_is_child &&
1030 [ + + ]: 270 : !bms_is_subset(em->em_relids, relids))
1031 : 110 : continue;
1032 : :
1033 : : /*
1034 : : * Match if all Vars and quasi-Vars are present in "exprs".
1035 : : */
1036 : 7305 : emvars = pull_var_clause((Node *) em->em_expr,
1037 : : PVC_INCLUDE_AGGREGATES |
1038 : : PVC_INCLUDE_WINDOWFUNCS |
1039 : : PVC_INCLUDE_PLACEHOLDERS);
1040 [ + + + + : 8071 : foreach(lc2, emvars)
+ + ]
1041 : : {
1042 [ + + ]: 7641 : if (!list_member(exprvars, lfirst(lc2)))
1043 : 6875 : break;
1044 : : }
1045 : 7305 : list_free(emvars);
1046 [ + + ]: 7305 : if (lc2)
1047 : 6875 : continue; /* we hit a non-available Var */
1048 : :
1049 : : /*
1050 : : * If requested, reject expressions that are not parallel-safe. We
1051 : : * check this last because it's a rather expensive test.
1052 : : */
1053 [ + + ]: 430 : if (require_parallel_safe &&
1054 [ + + ]: 121 : !is_parallel_safe(root, (Node *) em->em_expr))
1055 : 40 : continue;
1056 : :
1057 : 390 : return em; /* found usable expression */
1058 : : }
1059 : :
1060 : 6537 : return NULL;
1061 : : }
1062 : :
1063 : : /*
1064 : : * relation_can_be_sorted_early
1065 : : * Can this relation be sorted on this EC before the final output step?
1066 : : *
1067 : : * To succeed, we must find an EC member that prepare_sort_from_pathkeys knows
1068 : : * how to sort on, given the rel's reltarget as input. There are also a few
1069 : : * additional constraints based on the fact that the desired sort will be done
1070 : : * "early", within the scan/join part of the plan. Also, non-parallel-safe
1071 : : * expressions are ignored if 'require_parallel_safe'.
1072 : : *
1073 : : * At some point we might want to return the identified EquivalenceMember,
1074 : : * but for now, callers only want to know if there is one.
1075 : : */
1076 : : bool
1077 : 15748 : relation_can_be_sorted_early(PlannerInfo *root, RelOptInfo *rel,
1078 : : EquivalenceClass *ec, bool require_parallel_safe)
1079 : : {
1080 : 15748 : PathTarget *target = rel->reltarget;
1081 : : EquivalenceMember *em;
1082 : : ListCell *lc;
1083 : :
1084 : : /*
1085 : : * Reject volatile ECs immediately; such sorts must always be postponed.
1086 : : */
1087 [ + + ]: 15748 : if (ec->ec_has_volatile)
1088 : 60 : return false;
1089 : :
1090 : : /*
1091 : : * Try to find an EM directly matching some reltarget member.
1092 : : */
1093 [ + + + + : 34612 : foreach(lc, target->exprs)
+ + ]
1094 : : {
1095 : 27994 : Expr *targetexpr = (Expr *) lfirst(lc);
1096 : :
1097 : 27994 : em = find_ec_member_matching_expr(ec, targetexpr, rel->relids);
1098 [ + + ]: 27994 : if (!em)
1099 : 18924 : continue;
1100 : :
1101 : : /*
1102 : : * Reject expressions involving set-returning functions, as those
1103 : : * can't be computed early either. (Note: this test and the following
1104 : : * one are effectively checking properties of targetexpr, so there's
1105 : : * no point in asking whether some other EC member would be better.)
1106 : : */
1107 [ - + ]: 9070 : if (expression_returns_set((Node *) em->em_expr))
1108 : 0 : continue;
1109 : :
1110 : : /*
1111 : : * If requested, reject expressions that are not parallel-safe. We
1112 : : * check this last because it's a rather expensive test.
1113 : : */
1114 [ + - ]: 9070 : if (require_parallel_safe &&
1115 [ - + ]: 9070 : !is_parallel_safe(root, (Node *) em->em_expr))
1116 : 0 : continue;
1117 : :
1118 : 9070 : return true;
1119 : : }
1120 : :
1121 : : /*
1122 : : * Try to find an expression computable from the reltarget.
1123 : : */
1124 : 6618 : em = find_computable_ec_member(root, ec, target->exprs, rel->relids,
1125 : : require_parallel_safe);
1126 [ + + ]: 6618 : if (!em)
1127 : 6537 : return false;
1128 : :
1129 : : /*
1130 : : * Reject expressions involving set-returning functions, as those can't be
1131 : : * computed early either. (There's no point in looking for another EC
1132 : : * member in this case; since SRFs can't appear in WHERE, they cannot
1133 : : * belong to multi-member ECs.)
1134 : : */
1135 [ + + ]: 81 : if (expression_returns_set((Node *) em->em_expr))
1136 : 10 : return false;
1137 : :
1138 : 71 : return true;
1139 : : }
1140 : :
1141 : : /*
1142 : : * generate_base_implied_equalities
1143 : : * Generate any restriction clauses that we can deduce from equivalence
1144 : : * classes.
1145 : : *
1146 : : * When an EC contains pseudoconstants, our strategy is to generate
1147 : : * "member = const1" clauses where const1 is the first constant member, for
1148 : : * every other member (including other constants). If we are able to do this
1149 : : * then we don't need any "var = var" comparisons because we've successfully
1150 : : * constrained all the vars at their points of creation. If we fail to
1151 : : * generate any of these clauses due to lack of cross-type operators, we fall
1152 : : * back to the "ec_broken" strategy described below. (XXX if there are
1153 : : * multiple constants of different types, it's possible that we might succeed
1154 : : * in forming all the required clauses if we started from a different const
1155 : : * member; but this seems a sufficiently hokey corner case to not be worth
1156 : : * spending lots of cycles on.)
1157 : : *
1158 : : * For ECs that contain no pseudoconstants, we generate derived clauses
1159 : : * "member1 = member2" for each pair of members belonging to the same base
1160 : : * relation (actually, if there are more than two for the same base relation,
1161 : : * we only need enough clauses to link each to each other). This provides
1162 : : * the base case for the recursion: each row emitted by a base relation scan
1163 : : * will constrain all computable members of the EC to be equal. As each
1164 : : * join path is formed, we'll add additional derived clauses on-the-fly
1165 : : * to maintain this invariant (see generate_join_implied_equalities).
1166 : : *
1167 : : * If the opfamilies used by the EC do not provide complete sets of cross-type
1168 : : * equality operators, it is possible that we will fail to generate a clause
1169 : : * that must be generated to maintain the invariant. (An example: given
1170 : : * "WHERE a.x = b.y AND b.y = a.z", the scheme breaks down if we cannot
1171 : : * generate "a.x = a.z" as a restriction clause for A.) In this case we mark
1172 : : * the EC "ec_broken" and fall back to regurgitating its original source
1173 : : * RestrictInfos at appropriate times. We do not try to retract any derived
1174 : : * clauses already generated from the broken EC, so the resulting plan could
1175 : : * be poor due to bad selectivity estimates caused by redundant clauses. But
1176 : : * the correct solution to that is to fix the opfamilies ...
1177 : : *
1178 : : * Equality clauses derived by this function are passed off to
1179 : : * process_implied_equality (in plan/initsplan.c) to be inserted into the
1180 : : * restrictinfo datastructures. Note that this must be called after initial
1181 : : * scanning of the quals and before Path construction begins.
1182 : : *
1183 : : * We make no attempt to avoid generating duplicate RestrictInfos here: we
1184 : : * don't search existing source or derived clauses in the EC for matches. It
1185 : : * doesn't really seem worth the trouble to do so.
1186 : : */
1187 : : void
1188 : 255010 : generate_base_implied_equalities(PlannerInfo *root)
1189 : : {
1190 : : int ec_index;
1191 : : ListCell *lc;
1192 : :
1193 : : /*
1194 : : * At this point, we're done absorbing knowledge of equivalences in the
1195 : : * query, so no further EC merging should happen, and ECs remaining in the
1196 : : * eq_classes list can be considered canonical. (But note that it's still
1197 : : * possible for new single-member ECs to be added through
1198 : : * get_eclass_for_sort_expr().)
1199 : : */
1200 : 255010 : root->ec_merging_done = true;
1201 : :
1202 : 255010 : ec_index = 0;
1203 [ + + + + : 574627 : foreach(lc, root->eq_classes)
+ + ]
1204 : : {
1205 : 319617 : EquivalenceClass *ec = (EquivalenceClass *) lfirst(lc);
1206 : 319617 : bool can_generate_joinclause = false;
1207 : : int i;
1208 : :
1209 : : Assert(ec->ec_merged == NULL); /* else shouldn't be in list */
1210 : : Assert(!ec->ec_broken); /* not yet anyway... */
1211 : :
1212 : : /*
1213 : : * Generate implied equalities that are restriction clauses.
1214 : : * Single-member ECs won't generate any deductions, either here or at
1215 : : * the join level.
1216 : : */
1217 [ + + ]: 319617 : if (list_length(ec->ec_members) > 1)
1218 : : {
1219 [ + + ]: 227600 : if (ec->ec_has_const)
1220 : 164736 : generate_base_implied_equalities_const(root, ec);
1221 : : else
1222 : 62864 : generate_base_implied_equalities_no_const(root, ec);
1223 : :
1224 : : /* Recover if we failed to generate required derived clauses */
1225 [ + + ]: 227600 : if (ec->ec_broken)
1226 : 25 : generate_base_implied_equalities_broken(root, ec);
1227 : :
1228 : : /* Detect whether this EC might generate join clauses */
1229 : 227600 : can_generate_joinclause =
1230 : 227600 : (bms_membership(ec->ec_relids) == BMS_MULTIPLE);
1231 : : }
1232 : :
1233 : : /*
1234 : : * Mark the base rels cited in each eclass (which should all exist by
1235 : : * now) with the eq_classes indexes of all eclasses mentioning them.
1236 : : * This will let us avoid searching in subsequent lookups. While
1237 : : * we're at it, we can mark base rels that have pending eclass joins;
1238 : : * this is a cheap version of has_relevant_eclass_joinclause().
1239 : : */
1240 : 319617 : i = -1;
1241 [ + + ]: 717538 : while ((i = bms_next_member(ec->ec_relids, i)) > 0)
1242 : : {
1243 : 397921 : RelOptInfo *rel = root->simple_rel_array[i];
1244 : :
1245 : : /* ignore the RTE_GROUP RTE */
1246 [ - + ]: 397921 : if (i == root->group_rtindex)
1247 : 0 : continue;
1248 : :
1249 [ + + ]: 397921 : if (rel == NULL) /* must be an outer join */
1250 : : {
1251 : : Assert(bms_is_member(i, root->outer_join_rels));
1252 : 3463 : continue;
1253 : : }
1254 : :
1255 : : Assert(rel->reloptkind == RELOPT_BASEREL);
1256 : :
1257 : 394458 : rel->eclass_indexes = bms_add_member(rel->eclass_indexes,
1258 : : ec_index);
1259 : :
1260 [ + + ]: 394458 : if (can_generate_joinclause)
1261 : 149052 : rel->has_eclass_joins = true;
1262 : : }
1263 : :
1264 : 319617 : ec_index++;
1265 : : }
1266 : 255010 : }
1267 : :
1268 : : /*
1269 : : * generate_base_implied_equalities when EC contains pseudoconstant(s)
1270 : : */
1271 : : static void
1272 : 164736 : generate_base_implied_equalities_const(PlannerInfo *root,
1273 : : EquivalenceClass *ec)
1274 : : {
1275 : 164736 : EquivalenceMember *const_em = NULL;
1276 : : ListCell *lc;
1277 : :
1278 : : /*
1279 : : * In the trivial case where we just had one "var = const" clause, push
1280 : : * the original clause back into the main planner machinery. There is
1281 : : * nothing to be gained by doing it differently, and we save the effort to
1282 : : * re-build and re-analyze an equality clause that will be exactly
1283 : : * equivalent to the old one.
1284 : : */
1285 [ + + + + ]: 317276 : if (list_length(ec->ec_members) == 2 &&
1286 : 152540 : list_length(ec->ec_sources) == 1)
1287 : : {
1288 : 152505 : RestrictInfo *restrictinfo = (RestrictInfo *) linitial(ec->ec_sources);
1289 : :
1290 : 152505 : distribute_restrictinfo_to_rels(root, restrictinfo);
1291 : 152505 : return;
1292 : : }
1293 : :
1294 : : /* We don't expect any children yet */
1295 : : Assert(ec->ec_childmembers == NULL);
1296 : :
1297 : : /*
1298 : : * Find the constant member to use. We prefer an actual constant to
1299 : : * pseudo-constants (such as Params), because the constraint exclusion
1300 : : * machinery might be able to exclude relations on the basis of generated
1301 : : * "var = const" equalities, but "var = param" won't work for that.
1302 : : */
1303 [ + - + + : 29457 : foreach(lc, ec->ec_members)
+ + ]
1304 : : {
1305 : 29373 : EquivalenceMember *cur_em = (EquivalenceMember *) lfirst(lc);
1306 : :
1307 [ + + ]: 29373 : if (cur_em->em_is_const)
1308 : : {
1309 : 12236 : const_em = cur_em;
1310 [ + + ]: 12236 : if (IsA(cur_em->em_expr, Const))
1311 : 12147 : break;
1312 : : }
1313 : : }
1314 : : Assert(const_em != NULL);
1315 : :
1316 : : /* Generate a derived equality against each other member */
1317 [ + - + + : 49005 : foreach(lc, ec->ec_members)
+ + ]
1318 : : {
1319 : 36799 : EquivalenceMember *cur_em = (EquivalenceMember *) lfirst(lc);
1320 : : Oid eq_op;
1321 : : RestrictInfo *rinfo;
1322 : :
1323 : : /* Child members should not exist in ec_members */
1324 : : Assert(!cur_em->em_is_child);
1325 [ + + ]: 36799 : if (cur_em == const_em)
1326 : 12211 : continue;
1327 : 24588 : eq_op = select_equality_operator(ec,
1328 : : cur_em->em_datatype,
1329 : : const_em->em_datatype);
1330 [ + + ]: 24588 : if (!OidIsValid(eq_op))
1331 : : {
1332 : : /* failed... */
1333 : 25 : ec->ec_broken = true;
1334 : 25 : break;
1335 : : }
1336 : :
1337 : : /*
1338 : : * We use the constant's em_jdomain as qualscope, so that if the
1339 : : * generated clause is variable-free (i.e, both EMs are consts) it
1340 : : * will be enforced at the join domain level.
1341 : : */
1342 : 24563 : rinfo = process_implied_equality(root, eq_op, ec->ec_collation,
1343 : : cur_em->em_expr, const_em->em_expr,
1344 : 24563 : const_em->em_jdomain->jd_relids,
1345 : : ec->ec_min_security,
1346 : 24563 : cur_em->em_is_const);
1347 : :
1348 : : /*
1349 : : * If the clause didn't degenerate to a constant, fill in the correct
1350 : : * markings for a mergejoinable clause, and save it as a derived
1351 : : * clause. (We will not re-use such clauses directly, but selectivity
1352 : : * estimation may consult those later. Note that this use of derived
1353 : : * clauses does not overlap with its use for join clauses, since we
1354 : : * never generate join clauses from an ec_has_const eclass.)
1355 : : */
1356 [ + - + + ]: 24563 : if (rinfo && rinfo->mergeopfamilies)
1357 : : {
1358 : : /* it's not redundant, so don't set parent_ec */
1359 : 24428 : rinfo->left_ec = rinfo->right_ec = ec;
1360 : 24428 : rinfo->left_em = cur_em;
1361 : 24428 : rinfo->right_em = const_em;
1362 : 24428 : ec_add_derived_clause(ec, rinfo);
1363 : : }
1364 : : }
1365 : : }
1366 : :
1367 : : /*
1368 : : * generate_base_implied_equalities when EC contains no pseudoconstants
1369 : : */
1370 : : static void
1371 : 62864 : generate_base_implied_equalities_no_const(PlannerInfo *root,
1372 : : EquivalenceClass *ec)
1373 : : {
1374 : : EquivalenceMember **prev_ems;
1375 : : ListCell *lc;
1376 : :
1377 : : /*
1378 : : * We scan the EC members once and track the last-seen member for each
1379 : : * base relation. When we see another member of the same base relation,
1380 : : * we generate "prev_em = cur_em". This results in the minimum number of
1381 : : * derived clauses, but it's possible that it will fail when a different
1382 : : * ordering would succeed. XXX FIXME: use a UNION-FIND algorithm similar
1383 : : * to the way we build merged ECs. (Use a list-of-lists for each rel.)
1384 : : */
1385 : 62864 : prev_ems = palloc0_array(EquivalenceMember *, root->simple_rel_array_size);
1386 : :
1387 : : /* We don't expect any children yet */
1388 : : Assert(ec->ec_childmembers == NULL);
1389 : :
1390 [ + - + + : 190423 : foreach(lc, ec->ec_members)
+ + ]
1391 : : {
1392 : 127559 : EquivalenceMember *cur_em = (EquivalenceMember *) lfirst(lc);
1393 : : int relid;
1394 : :
1395 : : /* Child members should not exist in ec_members */
1396 : : Assert(!cur_em->em_is_child);
1397 : :
1398 [ + + ]: 127559 : if (!bms_get_singleton_member(cur_em->em_relids, &relid))
1399 : 205 : continue;
1400 : : Assert(relid < root->simple_rel_array_size);
1401 : :
1402 [ + + ]: 127354 : if (prev_ems[relid] != NULL)
1403 : : {
1404 : 394 : EquivalenceMember *prev_em = prev_ems[relid];
1405 : : Oid eq_op;
1406 : : RestrictInfo *rinfo;
1407 : :
1408 : 394 : eq_op = select_equality_operator(ec,
1409 : : prev_em->em_datatype,
1410 : : cur_em->em_datatype);
1411 [ - + ]: 394 : if (!OidIsValid(eq_op))
1412 : : {
1413 : : /* failed... */
1414 : 0 : ec->ec_broken = true;
1415 : 0 : break;
1416 : : }
1417 : :
1418 : : /*
1419 : : * The expressions aren't constants, so the passed qualscope will
1420 : : * never be used to place the generated clause. We just need to
1421 : : * be sure it covers both expressions, which em_relids should do.
1422 : : */
1423 : 394 : rinfo = process_implied_equality(root, eq_op, ec->ec_collation,
1424 : : prev_em->em_expr, cur_em->em_expr,
1425 : : cur_em->em_relids,
1426 : : ec->ec_min_security,
1427 : : false);
1428 : :
1429 : : /*
1430 : : * If the clause didn't degenerate to a constant, fill in the
1431 : : * correct markings for a mergejoinable clause. We don't record
1432 : : * it as a derived clause, since we don't currently need to
1433 : : * re-find such clauses, and don't want to clutter the
1434 : : * derived-clause set with non-join clauses.
1435 : : */
1436 [ + - + - ]: 394 : if (rinfo && rinfo->mergeopfamilies)
1437 : : {
1438 : : /* it's not redundant, so don't set parent_ec */
1439 : 394 : rinfo->left_ec = rinfo->right_ec = ec;
1440 : 394 : rinfo->left_em = prev_em;
1441 : 394 : rinfo->right_em = cur_em;
1442 : : }
1443 : : }
1444 : 127354 : prev_ems[relid] = cur_em;
1445 : : }
1446 : :
1447 : 62864 : pfree(prev_ems);
1448 : :
1449 : : /*
1450 : : * We also have to make sure that all the Vars used in the member clauses
1451 : : * will be available at any join node we might try to reference them at.
1452 : : * For the moment we force all the Vars to be available at all join nodes
1453 : : * for this eclass. Perhaps this could be improved by doing some
1454 : : * pre-analysis of which members we prefer to join, but it's no worse than
1455 : : * what happened in the pre-8.3 code.
1456 : : */
1457 [ + - + + : 190423 : foreach(lc, ec->ec_members)
+ + ]
1458 : : {
1459 : 127559 : EquivalenceMember *cur_em = (EquivalenceMember *) lfirst(lc);
1460 : 127559 : List *vars = pull_var_clause((Node *) cur_em->em_expr,
1461 : : PVC_RECURSE_AGGREGATES |
1462 : : PVC_RECURSE_WINDOWFUNCS |
1463 : : PVC_INCLUDE_PLACEHOLDERS);
1464 : :
1465 : 127559 : add_vars_to_targetlist(root, vars, ec->ec_relids);
1466 : 127559 : list_free(vars);
1467 : : }
1468 : 62864 : }
1469 : :
1470 : : /*
1471 : : * generate_base_implied_equalities cleanup after failure
1472 : : *
1473 : : * What we must do here is push any zero- or one-relation source RestrictInfos
1474 : : * of the EC back into the main restrictinfo datastructures. Multi-relation
1475 : : * clauses will be regurgitated later by generate_join_implied_equalities().
1476 : : * (We do it this way to maintain continuity with the case that ec_broken
1477 : : * becomes set only after we've gone up a join level or two.) However, for
1478 : : * an EC that contains constants, we can adopt a simpler strategy and just
1479 : : * throw back all the source RestrictInfos immediately; that works because
1480 : : * we know that such an EC can't become broken later. (This rule justifies
1481 : : * ignoring ec_has_const ECs in generate_join_implied_equalities, even when
1482 : : * they are broken.)
1483 : : */
1484 : : static void
1485 : 25 : generate_base_implied_equalities_broken(PlannerInfo *root,
1486 : : EquivalenceClass *ec)
1487 : : {
1488 : : ListCell *lc;
1489 : :
1490 [ + - + + : 80 : foreach(lc, ec->ec_sources)
+ + ]
1491 : : {
1492 : 55 : RestrictInfo *restrictinfo = (RestrictInfo *) lfirst(lc);
1493 : :
1494 [ - + - - ]: 55 : if (ec->ec_has_const ||
1495 : 0 : bms_membership(restrictinfo->required_relids) != BMS_MULTIPLE)
1496 : 55 : distribute_restrictinfo_to_rels(root, restrictinfo);
1497 : : }
1498 : 25 : }
1499 : :
1500 : :
1501 : : /*
1502 : : * generate_join_implied_equalities
1503 : : * Generate any join clauses that we can deduce from equivalence classes.
1504 : : *
1505 : : * At a join node, we must enforce restriction clauses sufficient to ensure
1506 : : * that all equivalence-class members computable at that node are equal.
1507 : : * Since the set of clauses to enforce can vary depending on which subset
1508 : : * relations are the inputs, we have to compute this afresh for each join
1509 : : * relation pair. Hence a fresh List of RestrictInfo nodes is built and
1510 : : * passed back on each call.
1511 : : *
1512 : : * In addition to its use at join nodes, this can be applied to generate
1513 : : * eclass-based join clauses for use in a parameterized scan of a base rel.
1514 : : * The reason for the asymmetry of specifying the inner rel as a RelOptInfo
1515 : : * and the outer rel by Relids is that this usage occurs before we have
1516 : : * built any join RelOptInfos.
1517 : : *
1518 : : * An annoying special case for parameterized scans is that the inner rel can
1519 : : * be an appendrel child (an "other rel"). In this case we must generate
1520 : : * appropriate clauses using child EC members. add_child_rel_equivalences
1521 : : * must already have been done for the child rel.
1522 : : *
1523 : : * The results are sufficient for use in merge, hash, and plain nestloop join
1524 : : * methods. We do not worry here about selecting clauses that are optimal
1525 : : * for use in a parameterized indexscan. indxpath.c makes its own selections
1526 : : * of clauses to use, and if the ones we pick here are redundant with those,
1527 : : * the extras will be eliminated at createplan time, using the parent_ec
1528 : : * markers that we provide (see is_redundant_derived_clause()).
1529 : : *
1530 : : * Because the same join clauses are likely to be needed multiple times as
1531 : : * we consider different join paths, we avoid generating multiple copies:
1532 : : * whenever we select a particular pair of EquivalenceMembers to join,
1533 : : * we check to see if the pair matches any original clause (in ec_sources)
1534 : : * or previously-built derived clause. This saves memory and allows
1535 : : * re-use of information cached in RestrictInfos. We also avoid generating
1536 : : * commutative duplicates, i.e. if the algorithm selects "a.x = b.y" but
1537 : : * we already have "b.y = a.x", we return the existing clause.
1538 : : *
1539 : : * If we are considering an outer join, sjinfo is the associated OJ info,
1540 : : * otherwise it can be NULL.
1541 : : *
1542 : : * join_relids should always equal bms_union(outer_relids, inner_rel->relids)
1543 : : * plus whatever add_outer_joins_to_relids() would add. We could simplify
1544 : : * this function's API by computing it internally, but most callers have the
1545 : : * value at hand anyway.
1546 : : */
1547 : : List *
1548 : 434889 : generate_join_implied_equalities(PlannerInfo *root,
1549 : : Relids join_relids,
1550 : : Relids outer_relids,
1551 : : RelOptInfo *inner_rel,
1552 : : SpecialJoinInfo *sjinfo)
1553 : : {
1554 : 434889 : List *result = NIL;
1555 : 434889 : Relids inner_relids = inner_rel->relids;
1556 : : Relids nominal_inner_relids;
1557 : : Relids nominal_join_relids;
1558 : : Bitmapset *matching_ecs;
1559 : : int i;
1560 : :
1561 : : /* If inner rel is a child, extra setup work is needed */
1562 [ + + + + : 434889 : if (IS_OTHER_REL(inner_rel))
- + ]
1563 : : {
1564 : : Assert(!bms_is_empty(inner_rel->top_parent_relids));
1565 : :
1566 : : /* Fetch relid set for the topmost parent rel */
1567 : 6653 : nominal_inner_relids = inner_rel->top_parent_relids;
1568 : : /* ECs will be marked with the parent's relid, not the child's */
1569 : 6653 : nominal_join_relids = bms_union(outer_relids, nominal_inner_relids);
1570 : 6653 : nominal_join_relids = add_outer_joins_to_relids(root,
1571 : : nominal_join_relids,
1572 : : sjinfo,
1573 : : NULL);
1574 : : }
1575 : : else
1576 : : {
1577 : 428236 : nominal_inner_relids = inner_relids;
1578 : 428236 : nominal_join_relids = join_relids;
1579 : : }
1580 : :
1581 : : /*
1582 : : * Examine all potentially-relevant eclasses.
1583 : : *
1584 : : * If we are considering an outer join, we must include "join" clauses
1585 : : * that mention either input rel plus the outer join's relid; these
1586 : : * represent post-join filter clauses that have to be applied at this
1587 : : * join. We don't have infrastructure that would let us identify such
1588 : : * eclasses cheaply, so just fall back to considering all eclasses
1589 : : * mentioning anything in nominal_join_relids.
1590 : : *
1591 : : * At inner joins, we can be smarter: only consider eclasses mentioning
1592 : : * both input rels.
1593 : : */
1594 [ + + + + ]: 434889 : if (sjinfo && sjinfo->ojrelid != 0)
1595 : 63133 : matching_ecs = get_eclass_indexes_for_relids(root, nominal_join_relids);
1596 : : else
1597 : 371756 : matching_ecs = get_common_eclass_indexes(root, nominal_inner_relids,
1598 : : outer_relids);
1599 : :
1600 : 434889 : i = -1;
1601 [ + + ]: 1153579 : while ((i = bms_next_member(matching_ecs, i)) >= 0)
1602 : : {
1603 : 718690 : EquivalenceClass *ec = (EquivalenceClass *) list_nth(root->eq_classes, i);
1604 : 718690 : List *sublist = NIL;
1605 : :
1606 : : /* ECs containing consts do not need any further enforcement */
1607 [ + + ]: 718690 : if (ec->ec_has_const)
1608 : 87549 : continue;
1609 : :
1610 : : /* Single-member ECs won't generate any deductions */
1611 [ + + ]: 631141 : if (list_length(ec->ec_members) <= 1)
1612 : 301441 : continue;
1613 : :
1614 : : /* Sanity check that this eclass overlaps the join */
1615 : : Assert(bms_overlap(ec->ec_relids, nominal_join_relids));
1616 : :
1617 [ + + ]: 329700 : if (!ec->ec_broken)
1618 : 329430 : sublist = generate_join_implied_equalities_normal(root,
1619 : : ec,
1620 : : join_relids,
1621 : : outer_relids,
1622 : : inner_relids);
1623 : :
1624 : : /* Recover if we failed to generate required derived clauses */
1625 [ + + ]: 329700 : if (ec->ec_broken)
1626 : 300 : sublist = generate_join_implied_equalities_broken(root,
1627 : : ec,
1628 : : nominal_join_relids,
1629 : : outer_relids,
1630 : : nominal_inner_relids,
1631 : : inner_rel);
1632 : :
1633 : 329700 : result = list_concat(result, sublist);
1634 : : }
1635 : :
1636 : 434889 : return result;
1637 : : }
1638 : :
1639 : : /*
1640 : : * generate_join_implied_equalities_for_ecs
1641 : : * As above, but consider only the listed ECs.
1642 : : *
1643 : : * For the sole current caller, we can assume sjinfo == NULL, that is we are
1644 : : * not interested in outer-join filter clauses. This might need to change
1645 : : * in future.
1646 : : */
1647 : : List *
1648 : 4651 : generate_join_implied_equalities_for_ecs(PlannerInfo *root,
1649 : : List *eclasses,
1650 : : Relids join_relids,
1651 : : Relids outer_relids,
1652 : : RelOptInfo *inner_rel)
1653 : : {
1654 : 4651 : List *result = NIL;
1655 : 4651 : Relids inner_relids = inner_rel->relids;
1656 : : Relids nominal_inner_relids;
1657 : : Relids nominal_join_relids;
1658 : : ListCell *lc;
1659 : :
1660 : : /* If inner rel is a child, extra setup work is needed */
1661 [ + + + - : 4651 : if (IS_OTHER_REL(inner_rel))
- + ]
1662 : : {
1663 : : Assert(!bms_is_empty(inner_rel->top_parent_relids));
1664 : :
1665 : : /* Fetch relid set for the topmost parent rel */
1666 : 80 : nominal_inner_relids = inner_rel->top_parent_relids;
1667 : : /* ECs will be marked with the parent's relid, not the child's */
1668 : 80 : nominal_join_relids = bms_union(outer_relids, nominal_inner_relids);
1669 : : }
1670 : : else
1671 : : {
1672 : 4571 : nominal_inner_relids = inner_relids;
1673 : 4571 : nominal_join_relids = join_relids;
1674 : : }
1675 : :
1676 [ + - + + : 9730 : foreach(lc, eclasses)
+ + ]
1677 : : {
1678 : 5079 : EquivalenceClass *ec = (EquivalenceClass *) lfirst(lc);
1679 : 5079 : List *sublist = NIL;
1680 : :
1681 : : /* ECs containing consts do not need any further enforcement */
1682 [ - + ]: 5079 : if (ec->ec_has_const)
1683 : 0 : continue;
1684 : :
1685 : : /* Single-member ECs won't generate any deductions */
1686 [ - + ]: 5079 : if (list_length(ec->ec_members) <= 1)
1687 : 0 : continue;
1688 : :
1689 : : /* We can quickly ignore any that don't overlap the join, too */
1690 [ - + ]: 5079 : if (!bms_overlap(ec->ec_relids, nominal_join_relids))
1691 : 0 : continue;
1692 : :
1693 [ + - ]: 5079 : if (!ec->ec_broken)
1694 : 5079 : sublist = generate_join_implied_equalities_normal(root,
1695 : : ec,
1696 : : join_relids,
1697 : : outer_relids,
1698 : : inner_relids);
1699 : :
1700 : : /* Recover if we failed to generate required derived clauses */
1701 [ - + ]: 5079 : if (ec->ec_broken)
1702 : 0 : sublist = generate_join_implied_equalities_broken(root,
1703 : : ec,
1704 : : nominal_join_relids,
1705 : : outer_relids,
1706 : : nominal_inner_relids,
1707 : : inner_rel);
1708 : :
1709 : 5079 : result = list_concat(result, sublist);
1710 : : }
1711 : :
1712 : 4651 : return result;
1713 : : }
1714 : :
1715 : : /*
1716 : : * generate_join_implied_equalities for a still-valid EC
1717 : : */
1718 : : static List *
1719 : 334509 : generate_join_implied_equalities_normal(PlannerInfo *root,
1720 : : EquivalenceClass *ec,
1721 : : Relids join_relids,
1722 : : Relids outer_relids,
1723 : : Relids inner_relids)
1724 : : {
1725 : 334509 : List *result = NIL;
1726 : 334509 : List *new_members = NIL;
1727 : 334509 : List *outer_members = NIL;
1728 : 334509 : List *inner_members = NIL;
1729 : : EquivalenceMemberIterator it;
1730 : : EquivalenceMember *cur_em;
1731 : :
1732 : : /*
1733 : : * First, scan the EC to identify member values that are computable at the
1734 : : * outer rel, at the inner rel, or at this relation but not in either
1735 : : * input rel. The outer-rel members should already be enforced equal,
1736 : : * likewise for the inner-rel members. We'll need to create clauses to
1737 : : * enforce that any newly computable members are all equal to each other
1738 : : * as well as to at least one input member, plus enforce at least one
1739 : : * outer-rel member equal to at least one inner-rel member.
1740 : : */
1741 : 334509 : setup_eclass_member_iterator(&it, ec, join_relids);
1742 [ + + ]: 1062910 : while ((cur_em = eclass_member_iterator_next(&it)) != NULL)
1743 : : {
1744 : : /*
1745 : : * We don't need to check explicitly for child EC members. This test
1746 : : * against join_relids will cause them to be ignored except when
1747 : : * considering a child inner rel, which is what we want.
1748 : : */
1749 [ + + ]: 728401 : if (!bms_is_subset(cur_em->em_relids, join_relids))
1750 : 60334 : continue; /* not computable yet, or wrong child */
1751 : :
1752 [ + + ]: 668067 : if (bms_is_subset(cur_em->em_relids, outer_relids))
1753 : 374090 : outer_members = lappend(outer_members, cur_em);
1754 [ + + ]: 293977 : else if (bms_is_subset(cur_em->em_relids, inner_relids))
1755 : 291907 : inner_members = lappend(inner_members, cur_em);
1756 : : else
1757 : 2070 : new_members = lappend(new_members, cur_em);
1758 : : }
1759 : :
1760 : : /*
1761 : : * First, select the joinclause if needed. We can equate any one outer
1762 : : * member to any one inner member, but we have to find a datatype
1763 : : * combination for which an opfamily member operator exists. If we have
1764 : : * choices, we prefer simple Var members (possibly with RelabelType) since
1765 : : * these are (a) cheapest to compute at runtime and (b) most likely to
1766 : : * have useful statistics. Also, prefer operators that are also
1767 : : * hashjoinable.
1768 : : */
1769 [ + + + + ]: 334509 : if (outer_members && inner_members)
1770 : : {
1771 : 280996 : EquivalenceMember *best_outer_em = NULL;
1772 : 280996 : EquivalenceMember *best_inner_em = NULL;
1773 : 280996 : Oid best_eq_op = InvalidOid;
1774 : 280996 : int best_score = -1;
1775 : : RestrictInfo *rinfo;
1776 : : ListCell *lc1;
1777 : :
1778 [ + - + + : 293780 : foreach(lc1, outer_members)
+ + ]
1779 : : {
1780 : 281061 : EquivalenceMember *outer_em = (EquivalenceMember *) lfirst(lc1);
1781 : : ListCell *lc2;
1782 : :
1783 [ + - + + : 293865 : foreach(lc2, inner_members)
+ + ]
1784 : : {
1785 : 281081 : EquivalenceMember *inner_em = (EquivalenceMember *) lfirst(lc2);
1786 : : Oid eq_op;
1787 : : int score;
1788 : :
1789 : 281081 : eq_op = select_equality_operator(ec,
1790 : : outer_em->em_datatype,
1791 : : inner_em->em_datatype);
1792 [ + + ]: 281081 : if (!OidIsValid(eq_op))
1793 : 30 : continue;
1794 : 281051 : score = 0;
1795 [ + + ]: 281051 : if (IsA(outer_em->em_expr, Var) ||
1796 [ + + ]: 14252 : (IsA(outer_em->em_expr, RelabelType) &&
1797 [ + + ]: 3648 : IsA(((RelabelType *) outer_em->em_expr)->arg, Var)))
1798 : 270236 : score++;
1799 [ + + ]: 281051 : if (IsA(inner_em->em_expr, Var) ||
1800 [ + + ]: 9740 : (IsA(inner_em->em_expr, RelabelType) &&
1801 [ + + ]: 7564 : IsA(((RelabelType *) inner_em->em_expr)->arg, Var)))
1802 : 278749 : score++;
1803 [ + + ]: 281051 : if (op_hashjoinable(eq_op,
1804 : 281051 : exprType((Node *) outer_em->em_expr)))
1805 : 280992 : score++;
1806 [ + + ]: 281051 : if (score > best_score)
1807 : : {
1808 : 280966 : best_outer_em = outer_em;
1809 : 280966 : best_inner_em = inner_em;
1810 : 280966 : best_eq_op = eq_op;
1811 : 280966 : best_score = score;
1812 [ + + ]: 280966 : if (best_score == 3)
1813 : 268277 : break; /* no need to look further */
1814 : : }
1815 : : }
1816 [ + + ]: 281061 : if (best_score == 3)
1817 : 268277 : break; /* no need to look further */
1818 : : }
1819 [ + + ]: 280996 : if (best_score < 0)
1820 : : {
1821 : : /* failed... */
1822 : 30 : ec->ec_broken = true;
1823 : 30 : return NIL;
1824 : : }
1825 : :
1826 : : /*
1827 : : * Create clause, setting parent_ec to mark it as redundant with other
1828 : : * joinclauses
1829 : : */
1830 : 280966 : rinfo = create_join_clause(root, ec, best_eq_op,
1831 : : best_outer_em, best_inner_em,
1832 : : ec);
1833 : :
1834 : 280966 : result = lappend(result, rinfo);
1835 : : }
1836 : :
1837 : : /*
1838 : : * Now deal with building restrictions for any expressions that involve
1839 : : * Vars from both sides of the join. We have to equate all of these to
1840 : : * each other as well as to at least one old member (if any).
1841 : : *
1842 : : * XXX as in generate_base_implied_equalities_no_const, we could be a lot
1843 : : * smarter here to avoid unnecessary failures in cross-type situations.
1844 : : * For now, use the same left-to-right method used there.
1845 : : */
1846 [ + + ]: 334479 : if (new_members)
1847 : : {
1848 : 2040 : List *old_members = list_concat(outer_members, inner_members);
1849 : 2040 : EquivalenceMember *prev_em = NULL;
1850 : : RestrictInfo *rinfo;
1851 : : ListCell *lc1;
1852 : :
1853 : : /* For now, arbitrarily take the first old_member as the one to use */
1854 [ + + ]: 2040 : if (old_members)
1855 : 1683 : new_members = lappend(new_members, linitial(old_members));
1856 : :
1857 [ + - + + : 5793 : foreach(lc1, new_members)
+ + ]
1858 : : {
1859 : 3753 : cur_em = (EquivalenceMember *) lfirst(lc1);
1860 : :
1861 [ + + ]: 3753 : if (prev_em != NULL)
1862 : : {
1863 : : Oid eq_op;
1864 : :
1865 : 1713 : eq_op = select_equality_operator(ec,
1866 : : prev_em->em_datatype,
1867 : : cur_em->em_datatype);
1868 [ - + ]: 1713 : if (!OidIsValid(eq_op))
1869 : : {
1870 : : /* failed... */
1871 : 0 : ec->ec_broken = true;
1872 : 0 : return NIL;
1873 : : }
1874 : : /* do NOT set parent_ec, this qual is not redundant! */
1875 : 1713 : rinfo = create_join_clause(root, ec, eq_op,
1876 : : prev_em, cur_em,
1877 : : NULL);
1878 : :
1879 : 1713 : result = lappend(result, rinfo);
1880 : : }
1881 : 3753 : prev_em = cur_em;
1882 : : }
1883 : : }
1884 : :
1885 : 334479 : return result;
1886 : : }
1887 : :
1888 : : /*
1889 : : * generate_join_implied_equalities cleanup after failure
1890 : : *
1891 : : * Return any original RestrictInfos that are enforceable at this join.
1892 : : *
1893 : : * In the case of a child inner relation, we have to translate the
1894 : : * original RestrictInfos from parent to child Vars.
1895 : : */
1896 : : static List *
1897 : 300 : generate_join_implied_equalities_broken(PlannerInfo *root,
1898 : : EquivalenceClass *ec,
1899 : : Relids nominal_join_relids,
1900 : : Relids outer_relids,
1901 : : Relids nominal_inner_relids,
1902 : : RelOptInfo *inner_rel)
1903 : : {
1904 : 300 : List *result = NIL;
1905 : : ListCell *lc;
1906 : :
1907 [ + - + + : 820 : foreach(lc, ec->ec_sources)
+ + ]
1908 : : {
1909 : 520 : RestrictInfo *restrictinfo = (RestrictInfo *) lfirst(lc);
1910 : 520 : Relids clause_relids = restrictinfo->required_relids;
1911 : :
1912 [ + + ]: 520 : if (bms_is_subset(clause_relids, nominal_join_relids) &&
1913 [ + + ]: 280 : !bms_is_subset(clause_relids, outer_relids) &&
1914 [ + - ]: 260 : !bms_is_subset(clause_relids, nominal_inner_relids))
1915 : 260 : result = lappend(result, restrictinfo);
1916 : : }
1917 : :
1918 : : /*
1919 : : * If we have to translate, just brute-force apply adjust_appendrel_attrs
1920 : : * to all the RestrictInfos at once. This will result in returning
1921 : : * RestrictInfos that are not included in EC's derived clauses, but there
1922 : : * shouldn't be any duplication, and it's a sufficiently narrow corner
1923 : : * case that we shouldn't sweat too much over it anyway.
1924 : : *
1925 : : * Since inner_rel might be an indirect descendant of the baserel
1926 : : * mentioned in the ec_sources clauses, we have to be prepared to apply
1927 : : * multiple levels of Var translation.
1928 : : */
1929 [ + + + - : 300 : if (IS_OTHER_REL(inner_rel) && result != NIL)
- + + + ]
1930 : 135 : result = (List *) adjust_appendrel_attrs_multilevel(root,
1931 : : (Node *) result,
1932 : : inner_rel,
1933 : 135 : inner_rel->top_parent);
1934 : :
1935 : 300 : return result;
1936 : : }
1937 : :
1938 : :
1939 : : /*
1940 : : * select_equality_operator
1941 : : * Select a suitable equality operator for comparing two EC members
1942 : : *
1943 : : * Returns InvalidOid if no operator can be found for this datatype combination
1944 : : */
1945 : : static Oid
1946 : 410416 : select_equality_operator(EquivalenceClass *ec, Oid lefttype, Oid righttype)
1947 : : {
1948 : : ListCell *lc;
1949 : :
1950 [ + - + + : 410471 : foreach(lc, ec->ec_opfamilies)
+ + ]
1951 : : {
1952 : 410416 : Oid opfamily = lfirst_oid(lc);
1953 : : Oid opno;
1954 : :
1955 : 410416 : opno = get_opfamily_member_for_cmptype(opfamily, lefttype, righttype, COMPARE_EQ);
1956 [ + + ]: 410416 : if (!OidIsValid(opno))
1957 : 55 : continue;
1958 : : /* If no barrier quals in query, don't worry about leaky operators */
1959 [ + + ]: 410361 : if (ec->ec_max_security == 0)
1960 : 410361 : return opno;
1961 : : /* Otherwise, insist that selected operators be leakproof */
1962 [ + - ]: 509 : if (get_func_leakproof(get_opcode(opno)))
1963 : 509 : return opno;
1964 : : }
1965 : 55 : return InvalidOid;
1966 : : }
1967 : :
1968 : :
1969 : : /*
1970 : : * create_join_clause
1971 : : * Find or make a RestrictInfo comparing the two given EC members
1972 : : * with the given operator (or, possibly, its commutator, because
1973 : : * the ordering of the operands in the result is not guaranteed).
1974 : : *
1975 : : * parent_ec is either equal to ec (if the clause is a potentially-redundant
1976 : : * join clause) or NULL (if not). We have to treat this as part of the
1977 : : * match requirements --- it's possible that a clause comparing the same two
1978 : : * EMs is a join clause in one join path and a restriction clause in another.
1979 : : */
1980 : : static RestrictInfo *
1981 : 387696 : create_join_clause(PlannerInfo *root,
1982 : : EquivalenceClass *ec, Oid opno,
1983 : : EquivalenceMember *leftem,
1984 : : EquivalenceMember *rightem,
1985 : : EquivalenceClass *parent_ec)
1986 : : {
1987 : : RestrictInfo *rinfo;
1988 : 387696 : RestrictInfo *parent_rinfo = NULL;
1989 : : MemoryContext oldcontext;
1990 : :
1991 : 387696 : rinfo = ec_search_clause_for_ems(root, ec, leftem, rightem, parent_ec);
1992 [ + + ]: 387696 : if (rinfo)
1993 : 320981 : return rinfo;
1994 : :
1995 : : /*
1996 : : * Not there, so build it, in planner context so we can re-use it. (Not
1997 : : * important in normal planning, but definitely so in GEQO.)
1998 : : */
1999 : 66715 : oldcontext = MemoryContextSwitchTo(root->planner_cxt);
2000 : :
2001 : : /*
2002 : : * If either EM is a child, recursively create the corresponding
2003 : : * parent-to-parent clause, so that we can duplicate its rinfo_serial.
2004 : : */
2005 [ + + + + ]: 66715 : if (leftem->em_is_child || rightem->em_is_child)
2006 : : {
2007 [ + + ]: 4088 : EquivalenceMember *leftp = leftem->em_parent ? leftem->em_parent : leftem;
2008 [ + + ]: 4088 : EquivalenceMember *rightp = rightem->em_parent ? rightem->em_parent : rightem;
2009 : :
2010 : 4088 : parent_rinfo = create_join_clause(root, ec, opno,
2011 : : leftp, rightp,
2012 : : parent_ec);
2013 : : }
2014 : :
2015 : 66715 : rinfo = build_implied_join_equality(root,
2016 : : opno,
2017 : : ec->ec_collation,
2018 : : leftem->em_expr,
2019 : : rightem->em_expr,
2020 : 66715 : bms_union(leftem->em_relids,
2021 : 66715 : rightem->em_relids),
2022 : : ec->ec_min_security);
2023 : :
2024 : : /*
2025 : : * If either EM is a child, set the clause's clause_relids from the
2026 : : * members' em_relids rather than the relids found in the expressions.
2027 : : * These normally match, but not for UNION ALL sub-selects whose output
2028 : : * items are constants (mentioning no rels) or contain lateral references
2029 : : * (mentioning rels that the child's parameterization supplies). We must
2030 : : * do this so that join_clause_is_movable_into() will think that the
2031 : : * clause should be evaluated at the correct place.
2032 : : */
2033 [ + + + + ]: 66715 : if (leftem->em_is_child || rightem->em_is_child)
2034 : : {
2035 : : Relids baserels;
2036 : :
2037 : 8176 : rinfo->clause_relids = bms_union(leftem->em_relids,
2038 : 4088 : rightem->em_relids);
2039 : :
2040 : : /* keep num_base_rels in sync, as in make_restrictinfo() */
2041 : 4088 : baserels = bms_difference(rinfo->clause_relids,
2042 : 4088 : root->outer_join_rels);
2043 : 4088 : rinfo->num_base_rels = bms_num_members(baserels);
2044 : 4088 : bms_free(baserels);
2045 : : }
2046 : :
2047 : : /* If it's a child clause, copy the parent's rinfo_serial */
2048 [ + + ]: 66715 : if (parent_rinfo)
2049 : 4088 : rinfo->rinfo_serial = parent_rinfo->rinfo_serial;
2050 : : else
2051 : : {
2052 : : RestrictInfo *counterpart;
2053 : :
2054 : : /*
2055 : : * If a clause comparing the same two EMs already exists with the
2056 : : * opposite parent_ec marking, adopt its rinfo_serial: the two clauses
2057 : : * enforce the same condition, and they must share a serial number
2058 : : * lest we enforce that condition more than once in a plan.
2059 : : */
2060 [ + + ]: 62627 : counterpart = ec_search_clause_for_ems(root, ec, leftem, rightem,
2061 : : parent_ec ? NULL : ec);
2062 [ + + ]: 62627 : if (counterpart)
2063 : 60858 : rinfo->rinfo_serial = counterpart->rinfo_serial;
2064 : : }
2065 : :
2066 : : /* Mark the clause as redundant, or not */
2067 : 66715 : rinfo->parent_ec = parent_ec;
2068 : :
2069 : : /*
2070 : : * We know the correct values for left_ec/right_ec, ie this particular EC,
2071 : : * so we can just set them directly instead of forcing another lookup.
2072 : : */
2073 : 66715 : rinfo->left_ec = ec;
2074 : 66715 : rinfo->right_ec = ec;
2075 : :
2076 : : /* Mark it as usable with these EMs */
2077 : 66715 : rinfo->left_em = leftem;
2078 : 66715 : rinfo->right_em = rightem;
2079 : : /* and save it for possible re-use */
2080 : 66715 : ec_add_derived_clause(ec, rinfo);
2081 : :
2082 : 66715 : MemoryContextSwitchTo(oldcontext);
2083 : :
2084 : 66715 : return rinfo;
2085 : : }
2086 : :
2087 : :
2088 : : /*
2089 : : * reconsider_outer_join_clauses
2090 : : * Re-examine any outer-join clauses that were set aside by
2091 : : * distribute_qual_to_rels(), and see if we can derive any
2092 : : * EquivalenceClasses from them. Then, if they were not made
2093 : : * redundant, push them out into the regular join-clause lists.
2094 : : *
2095 : : * When we have mergejoinable clauses A = B that are outer-join clauses,
2096 : : * we can't blindly combine them with other clauses A = C to deduce B = C,
2097 : : * since in fact the "equality" A = B won't necessarily hold above the
2098 : : * outer join (one of the variables might be NULL instead). Nonetheless
2099 : : * there are cases where we can add qual clauses using transitivity.
2100 : : *
2101 : : * One case that we look for here is an outer-join clause OUTERVAR = INNERVAR
2102 : : * for which there is also an equivalence clause OUTERVAR = CONSTANT.
2103 : : * It is safe and useful to push a clause INNERVAR = CONSTANT into the
2104 : : * evaluation of the inner (nullable) relation, because any inner rows not
2105 : : * meeting this condition will not contribute to the outer-join result anyway.
2106 : : * (Any outer rows they could join to will be eliminated by the pushed-down
2107 : : * equivalence clause.)
2108 : : *
2109 : : * Note that the above rule does not work for full outer joins; nor is it
2110 : : * very interesting to consider cases where the generated equivalence clause
2111 : : * would involve relations outside the outer join, since such clauses couldn't
2112 : : * be pushed into the inner side's scan anyway. So the restriction to
2113 : : * outervar = pseudoconstant is not really giving up anything.
2114 : : *
2115 : : * For full-join cases, we can only do something useful if it's a FULL JOIN
2116 : : * USING and a merged column has an equivalence MERGEDVAR = CONSTANT.
2117 : : * By the time it gets here, the merged column will look like
2118 : : * COALESCE(LEFTVAR, RIGHTVAR)
2119 : : * and we will have a full-join clause LEFTVAR = RIGHTVAR that we can match
2120 : : * the COALESCE expression to. In this situation we can push LEFTVAR = CONSTANT
2121 : : * and RIGHTVAR = CONSTANT into the input relations, since any rows not
2122 : : * meeting these conditions cannot contribute to the join result.
2123 : : *
2124 : : * Again, there isn't any traction to be gained by trying to deal with
2125 : : * clauses comparing a mergedvar to a non-pseudoconstant. So we can make
2126 : : * use of the EquivalenceClasses to search for matching variables that were
2127 : : * equivalenced to constants. The interesting outer-join clauses were
2128 : : * accumulated for us by distribute_qual_to_rels.
2129 : : *
2130 : : * When we find one of these cases, we implement the changes we want by
2131 : : * generating a new equivalence clause INNERVAR = CONSTANT (or LEFTVAR, etc)
2132 : : * and pushing it into the EquivalenceClass structures. This is because we
2133 : : * may already know that INNERVAR is equivalenced to some other var(s), and
2134 : : * we'd like the constant to propagate to them too. Note that it would be
2135 : : * unsafe to merge any existing EC for INNERVAR with the OUTERVAR's EC ---
2136 : : * that could result in propagating constant restrictions from
2137 : : * INNERVAR to OUTERVAR, which would be very wrong.
2138 : : *
2139 : : * It's possible that the INNERVAR is also an OUTERVAR for some other
2140 : : * outer-join clause, in which case the process can be repeated. So we repeat
2141 : : * looping over the lists of clauses until no further deductions can be made.
2142 : : * Whenever we do make a deduction, we remove the generating clause from the
2143 : : * lists, since we don't want to make the same deduction twice.
2144 : : *
2145 : : * If we don't find any match for a set-aside outer join clause, we must
2146 : : * throw it back into the regular joinclause processing by passing it to
2147 : : * distribute_restrictinfo_to_rels(). If we do generate a derived clause,
2148 : : * however, the outer-join clause is redundant. We must still put some
2149 : : * clause into the regular processing, because otherwise the join will be
2150 : : * seen as a clauseless join and avoided during join order searching.
2151 : : * We handle this by generating a constant-TRUE clause that is marked with
2152 : : * the same required_relids etc as the removed outer-join clause, thus
2153 : : * making it a join clause between the correct relations.
2154 : : */
2155 : : void
2156 : 255011 : reconsider_outer_join_clauses(PlannerInfo *root)
2157 : : {
2158 : : bool found;
2159 : : ListCell *cell;
2160 : :
2161 : : /* Outer loop repeats until we find no more deductions */
2162 : : do
2163 : : {
2164 : 256672 : found = false;
2165 : :
2166 : : /* Process the LEFT JOIN clauses */
2167 [ + + + + : 279420 : foreach(cell, root->left_join_clauses)
+ + ]
2168 : : {
2169 : 22748 : OuterJoinClauseInfo *ojcinfo = (OuterJoinClauseInfo *) lfirst(cell);
2170 : :
2171 [ + + ]: 22748 : if (reconsider_outer_join_clause(root, ojcinfo, true))
2172 : : {
2173 : 522 : RestrictInfo *rinfo = ojcinfo->rinfo;
2174 : :
2175 : 522 : found = true;
2176 : : /* remove it from the list */
2177 : 522 : root->left_join_clauses =
2178 : 522 : foreach_delete_current(root->left_join_clauses, cell);
2179 : : /* throw back a dummy replacement clause (see notes above) */
2180 : 522 : rinfo = make_restrictinfo(root,
2181 : 522 : (Expr *) makeBoolConst(true, false),
2182 : 522 : rinfo->is_pushed_down,
2183 : 522 : rinfo->has_clone,
2184 : 522 : rinfo->is_clone,
2185 : : false, /* pseudoconstant */
2186 : : 0, /* security_level */
2187 : : rinfo->required_relids,
2188 : : rinfo->incompatible_relids,
2189 : : rinfo->outer_relids);
2190 : 522 : distribute_restrictinfo_to_rels(root, rinfo);
2191 : : }
2192 : : }
2193 : :
2194 : : /* Process the RIGHT JOIN clauses */
2195 [ + + + + : 285684 : foreach(cell, root->right_join_clauses)
+ + ]
2196 : : {
2197 : 29012 : OuterJoinClauseInfo *ojcinfo = (OuterJoinClauseInfo *) lfirst(cell);
2198 : :
2199 [ + + ]: 29012 : if (reconsider_outer_join_clause(root, ojcinfo, false))
2200 : : {
2201 : 1144 : RestrictInfo *rinfo = ojcinfo->rinfo;
2202 : :
2203 : 1144 : found = true;
2204 : : /* remove it from the list */
2205 : 1144 : root->right_join_clauses =
2206 : 1144 : foreach_delete_current(root->right_join_clauses, cell);
2207 : : /* throw back a dummy replacement clause (see notes above) */
2208 : 1144 : rinfo = make_restrictinfo(root,
2209 : 1144 : (Expr *) makeBoolConst(true, false),
2210 : 1144 : rinfo->is_pushed_down,
2211 : 1144 : rinfo->has_clone,
2212 : 1144 : rinfo->is_clone,
2213 : : false, /* pseudoconstant */
2214 : : 0, /* security_level */
2215 : : rinfo->required_relids,
2216 : : rinfo->incompatible_relids,
2217 : : rinfo->outer_relids);
2218 : 1144 : distribute_restrictinfo_to_rels(root, rinfo);
2219 : : }
2220 : : }
2221 : :
2222 : : /* Process the FULL JOIN clauses */
2223 [ + + + + : 257746 : foreach(cell, root->full_join_clauses)
+ + ]
2224 : : {
2225 : 1074 : OuterJoinClauseInfo *ojcinfo = (OuterJoinClauseInfo *) lfirst(cell);
2226 : :
2227 [ + + ]: 1074 : if (reconsider_full_join_clause(root, ojcinfo))
2228 : : {
2229 : 5 : RestrictInfo *rinfo = ojcinfo->rinfo;
2230 : :
2231 : 5 : found = true;
2232 : : /* remove it from the list */
2233 : 5 : root->full_join_clauses =
2234 : 5 : foreach_delete_current(root->full_join_clauses, cell);
2235 : : /* throw back a dummy replacement clause (see notes above) */
2236 : 5 : rinfo = make_restrictinfo(root,
2237 : 5 : (Expr *) makeBoolConst(true, false),
2238 : 5 : rinfo->is_pushed_down,
2239 : 5 : rinfo->has_clone,
2240 : 5 : rinfo->is_clone,
2241 : : false, /* pseudoconstant */
2242 : : 0, /* security_level */
2243 : : rinfo->required_relids,
2244 : : rinfo->incompatible_relids,
2245 : : rinfo->outer_relids);
2246 : 5 : distribute_restrictinfo_to_rels(root, rinfo);
2247 : : }
2248 : : }
2249 [ + + ]: 256672 : } while (found);
2250 : :
2251 : : /* Now, any remaining clauses have to be thrown back */
2252 [ + + + + : 276917 : foreach(cell, root->left_join_clauses)
+ + ]
2253 : : {
2254 : 21906 : OuterJoinClauseInfo *ojcinfo = (OuterJoinClauseInfo *) lfirst(cell);
2255 : :
2256 : 21906 : distribute_restrictinfo_to_rels(root, ojcinfo->rinfo);
2257 : : }
2258 [ + + + + : 281692 : foreach(cell, root->right_join_clauses)
+ + ]
2259 : : {
2260 : 26681 : OuterJoinClauseInfo *ojcinfo = (OuterJoinClauseInfo *) lfirst(cell);
2261 : :
2262 : 26681 : distribute_restrictinfo_to_rels(root, ojcinfo->rinfo);
2263 : : }
2264 [ + + + + : 256080 : foreach(cell, root->full_join_clauses)
+ + ]
2265 : : {
2266 : 1069 : OuterJoinClauseInfo *ojcinfo = (OuterJoinClauseInfo *) lfirst(cell);
2267 : :
2268 : 1069 : distribute_restrictinfo_to_rels(root, ojcinfo->rinfo);
2269 : : }
2270 : 255011 : }
2271 : :
2272 : : /*
2273 : : * reconsider_outer_join_clauses for a single LEFT/RIGHT JOIN clause
2274 : : *
2275 : : * Returns true if we were able to propagate a constant through the clause.
2276 : : */
2277 : : static bool
2278 : 51760 : reconsider_outer_join_clause(PlannerInfo *root, OuterJoinClauseInfo *ojcinfo,
2279 : : bool outer_on_left)
2280 : : {
2281 : 51760 : RestrictInfo *rinfo = ojcinfo->rinfo;
2282 : 51760 : SpecialJoinInfo *sjinfo = ojcinfo->sjinfo;
2283 : : Expr *outervar,
2284 : : *innervar;
2285 : : Oid opno,
2286 : : collation,
2287 : : left_type,
2288 : : right_type,
2289 : : inner_datatype;
2290 : : Relids inner_relids;
2291 : : ListCell *lc1;
2292 : :
2293 : : Assert(is_opclause(rinfo->clause));
2294 : 51760 : opno = ((OpExpr *) rinfo->clause)->opno;
2295 : 51760 : collation = ((OpExpr *) rinfo->clause)->inputcollid;
2296 : :
2297 : : /* Extract needed info from the clause */
2298 : 51760 : op_input_types(opno, &left_type, &right_type);
2299 [ + + ]: 51760 : if (outer_on_left)
2300 : : {
2301 : 22748 : outervar = (Expr *) get_leftop(rinfo->clause);
2302 : 22748 : innervar = (Expr *) get_rightop(rinfo->clause);
2303 : 22748 : inner_datatype = right_type;
2304 : 22748 : inner_relids = rinfo->right_relids;
2305 : : }
2306 : : else
2307 : : {
2308 : 29012 : outervar = (Expr *) get_rightop(rinfo->clause);
2309 : 29012 : innervar = (Expr *) get_leftop(rinfo->clause);
2310 : 29012 : inner_datatype = left_type;
2311 : 29012 : inner_relids = rinfo->left_relids;
2312 : : }
2313 : :
2314 : : /* Scan EquivalenceClasses for a match to outervar */
2315 [ + - + + : 315701 : foreach(lc1, root->eq_classes)
+ + ]
2316 : : {
2317 : 265607 : EquivalenceClass *cur_ec = (EquivalenceClass *) lfirst(lc1);
2318 : : bool match;
2319 : : ListCell *lc2;
2320 : :
2321 : : /* We don't expect any children yet */
2322 : : Assert(cur_ec->ec_childmembers == NULL);
2323 : :
2324 : : /* Ignore EC unless it contains pseudoconstants */
2325 [ + + ]: 265607 : if (!cur_ec->ec_has_const)
2326 : 206437 : continue;
2327 : : /* Never match to a volatile EC */
2328 [ - + ]: 59170 : if (cur_ec->ec_has_volatile)
2329 : 0 : continue;
2330 : : /* It has to match the outer-join clause as to semantics, too */
2331 [ + + ]: 59170 : if (collation != cur_ec->ec_collation)
2332 : 3968 : continue;
2333 [ + + ]: 55202 : if (!equal(rinfo->mergeopfamilies, cur_ec->ec_opfamilies))
2334 : 10173 : continue;
2335 : : /* Does it contain a match to outervar? */
2336 : 45029 : match = false;
2337 [ + - + + : 143324 : foreach(lc2, cur_ec->ec_members)
+ + ]
2338 : : {
2339 : 99961 : EquivalenceMember *cur_em = (EquivalenceMember *) lfirst(lc2);
2340 : :
2341 : : /* Child members should not exist in ec_members */
2342 : : Assert(!cur_em->em_is_child);
2343 [ + + ]: 99961 : if (equal(outervar, cur_em->em_expr))
2344 : : {
2345 : 1666 : match = true;
2346 : 1666 : break;
2347 : : }
2348 : : }
2349 [ + + ]: 45029 : if (!match)
2350 : 43363 : continue; /* no match, so ignore this EC */
2351 : :
2352 : : /*
2353 : : * Yes it does! Try to generate a clause INNERVAR = CONSTANT for each
2354 : : * CONSTANT in the EC. Note that we must succeed with at least one
2355 : : * constant before we can decide to throw away the outer-join clause.
2356 : : */
2357 : 1666 : match = false;
2358 [ + - + + : 5891 : foreach(lc2, cur_ec->ec_members)
+ + ]
2359 : : {
2360 : 4225 : EquivalenceMember *cur_em = (EquivalenceMember *) lfirst(lc2);
2361 : : Oid eq_op;
2362 : : RestrictInfo *newrinfo;
2363 : : JoinDomain *jdomain;
2364 : :
2365 [ + + ]: 4225 : if (!cur_em->em_is_const)
2366 : 2524 : continue; /* ignore non-const members */
2367 : 1701 : eq_op = select_equality_operator(cur_ec,
2368 : : inner_datatype,
2369 : : cur_em->em_datatype);
2370 [ - + ]: 1701 : if (!OidIsValid(eq_op))
2371 : 0 : continue; /* can't generate equality */
2372 : 1701 : newrinfo = build_implied_join_equality(root,
2373 : : eq_op,
2374 : : cur_ec->ec_collation,
2375 : : innervar,
2376 : : cur_em->em_expr,
2377 : : bms_copy(inner_relids),
2378 : : cur_ec->ec_min_security);
2379 : : /* This equality holds within the OJ's child JoinDomain */
2380 : 1701 : jdomain = find_join_domain(root, sjinfo->syn_righthand);
2381 [ + - ]: 1701 : if (process_equivalence(root, &newrinfo, jdomain))
2382 : 1701 : match = true;
2383 : : }
2384 : :
2385 : : /*
2386 : : * If we were able to equate INNERVAR to any constant, report success.
2387 : : * Otherwise, fall out of the search loop, since we know the OUTERVAR
2388 : : * appears in at most one EC.
2389 : : */
2390 [ + - ]: 1666 : if (match)
2391 : 1666 : return true;
2392 : : else
2393 : 0 : break;
2394 : : }
2395 : :
2396 : 50094 : return false; /* failed to make any deduction */
2397 : : }
2398 : :
2399 : : /*
2400 : : * reconsider_outer_join_clauses for a single FULL JOIN clause
2401 : : *
2402 : : * Returns true if we were able to propagate a constant through the clause.
2403 : : */
2404 : : static bool
2405 : 1074 : reconsider_full_join_clause(PlannerInfo *root, OuterJoinClauseInfo *ojcinfo)
2406 : : {
2407 : 1074 : RestrictInfo *rinfo = ojcinfo->rinfo;
2408 : 1074 : SpecialJoinInfo *sjinfo = ojcinfo->sjinfo;
2409 : 1074 : Relids fjrelids = bms_make_singleton(sjinfo->ojrelid);
2410 : : Expr *leftvar;
2411 : : Expr *rightvar;
2412 : : Oid opno,
2413 : : collation,
2414 : : left_type,
2415 : : right_type;
2416 : : Relids left_relids,
2417 : : right_relids;
2418 : : ListCell *lc1;
2419 : :
2420 : : /* Extract needed info from the clause */
2421 : : Assert(is_opclause(rinfo->clause));
2422 : 1074 : opno = ((OpExpr *) rinfo->clause)->opno;
2423 : 1074 : collation = ((OpExpr *) rinfo->clause)->inputcollid;
2424 : 1074 : op_input_types(opno, &left_type, &right_type);
2425 : 1074 : leftvar = (Expr *) get_leftop(rinfo->clause);
2426 : 1074 : rightvar = (Expr *) get_rightop(rinfo->clause);
2427 : 1074 : left_relids = rinfo->left_relids;
2428 : 1074 : right_relids = rinfo->right_relids;
2429 : :
2430 [ + - + + : 5404 : foreach(lc1, root->eq_classes)
+ + ]
2431 : : {
2432 : 4335 : EquivalenceClass *cur_ec = (EquivalenceClass *) lfirst(lc1);
2433 : 4335 : EquivalenceMember *coal_em = NULL;
2434 : : bool match;
2435 : : bool matchleft;
2436 : : bool matchright;
2437 : : ListCell *lc2;
2438 : 4335 : int coal_idx = -1;
2439 : :
2440 : : /* We don't expect any children yet */
2441 : : Assert(cur_ec->ec_childmembers == NULL);
2442 : :
2443 : : /* Ignore EC unless it contains pseudoconstants */
2444 [ + + ]: 4335 : if (!cur_ec->ec_has_const)
2445 : 4085 : continue;
2446 : : /* Never match to a volatile EC */
2447 [ - + ]: 250 : if (cur_ec->ec_has_volatile)
2448 : 0 : continue;
2449 : : /* It has to match the outer-join clause as to semantics, too */
2450 [ + + ]: 250 : if (collation != cur_ec->ec_collation)
2451 : 30 : continue;
2452 [ - + ]: 220 : if (!equal(rinfo->mergeopfamilies, cur_ec->ec_opfamilies))
2453 : 0 : continue;
2454 : :
2455 : : /*
2456 : : * Does it contain a COALESCE(leftvar, rightvar) construct?
2457 : : *
2458 : : * We can assume the COALESCE() inputs are in the same order as the
2459 : : * join clause, since both were automatically generated in the cases
2460 : : * we care about.
2461 : : *
2462 : : * XXX currently this may fail to match in cross-type cases because
2463 : : * the COALESCE will contain typecast operations while the join clause
2464 : : * may not (if there is a cross-type mergejoin operator available for
2465 : : * the two column types). Is it OK to strip implicit coercions from
2466 : : * the COALESCE arguments?
2467 : : */
2468 : 220 : match = false;
2469 [ + - + + : 645 : foreach(lc2, cur_ec->ec_members)
+ + ]
2470 : : {
2471 : 430 : coal_em = (EquivalenceMember *) lfirst(lc2);
2472 : :
2473 : : /* Child members should not exist in ec_members */
2474 : : Assert(!coal_em->em_is_child);
2475 [ + + ]: 430 : if (IsA(coal_em->em_expr, CoalesceExpr))
2476 : : {
2477 : 15 : CoalesceExpr *cexpr = (CoalesceExpr *) coal_em->em_expr;
2478 : : Node *cfirst;
2479 : : Node *csecond;
2480 : :
2481 [ - + ]: 15 : if (list_length(cexpr->args) != 2)
2482 : 0 : continue;
2483 : 15 : cfirst = (Node *) linitial(cexpr->args);
2484 : 15 : csecond = (Node *) lsecond(cexpr->args);
2485 : :
2486 : : /*
2487 : : * The COALESCE arguments will be marked as possibly nulled by
2488 : : * the full join, while we wish to generate clauses that apply
2489 : : * to the join's inputs. So we must strip the join from the
2490 : : * nullingrels fields of cfirst/csecond before comparing them
2491 : : * to leftvar/rightvar. (Perhaps with a less hokey
2492 : : * representation for FULL JOIN USING output columns, this
2493 : : * wouldn't be needed?)
2494 : : */
2495 : 15 : cfirst = remove_nulling_relids(cfirst, fjrelids, NULL);
2496 : 15 : csecond = remove_nulling_relids(csecond, fjrelids, NULL);
2497 : :
2498 [ + + + - ]: 15 : if (equal(leftvar, cfirst) && equal(rightvar, csecond))
2499 : : {
2500 : 5 : coal_idx = foreach_current_index(lc2);
2501 : 5 : match = true;
2502 : 5 : break;
2503 : : }
2504 : : }
2505 : : }
2506 [ + + ]: 220 : if (!match)
2507 : 215 : continue; /* no match, so ignore this EC */
2508 : :
2509 : : /*
2510 : : * Yes it does! Try to generate clauses LEFTVAR = CONSTANT and
2511 : : * RIGHTVAR = CONSTANT for each CONSTANT in the EC. Note that we must
2512 : : * succeed with at least one constant for each var before we can
2513 : : * decide to throw away the outer-join clause.
2514 : : */
2515 : 5 : matchleft = matchright = false;
2516 [ + - + + : 15 : foreach(lc2, cur_ec->ec_members)
+ + ]
2517 : : {
2518 : 10 : EquivalenceMember *cur_em = (EquivalenceMember *) lfirst(lc2);
2519 : : Oid eq_op;
2520 : : RestrictInfo *newrinfo;
2521 : : JoinDomain *jdomain;
2522 : :
2523 [ + + ]: 10 : if (!cur_em->em_is_const)
2524 : 5 : continue; /* ignore non-const members */
2525 : 5 : eq_op = select_equality_operator(cur_ec,
2526 : : left_type,
2527 : : cur_em->em_datatype);
2528 [ + - ]: 5 : if (OidIsValid(eq_op))
2529 : : {
2530 : 5 : newrinfo = build_implied_join_equality(root,
2531 : : eq_op,
2532 : : cur_ec->ec_collation,
2533 : : leftvar,
2534 : : cur_em->em_expr,
2535 : : bms_copy(left_relids),
2536 : : cur_ec->ec_min_security);
2537 : : /* This equality holds within the lefthand child JoinDomain */
2538 : 5 : jdomain = find_join_domain(root, sjinfo->syn_lefthand);
2539 [ + - ]: 5 : if (process_equivalence(root, &newrinfo, jdomain))
2540 : 5 : matchleft = true;
2541 : : }
2542 : 5 : eq_op = select_equality_operator(cur_ec,
2543 : : right_type,
2544 : : cur_em->em_datatype);
2545 [ + - ]: 5 : if (OidIsValid(eq_op))
2546 : : {
2547 : 5 : newrinfo = build_implied_join_equality(root,
2548 : : eq_op,
2549 : : cur_ec->ec_collation,
2550 : : rightvar,
2551 : : cur_em->em_expr,
2552 : : bms_copy(right_relids),
2553 : : cur_ec->ec_min_security);
2554 : : /* This equality holds within the righthand child JoinDomain */
2555 : 5 : jdomain = find_join_domain(root, sjinfo->syn_righthand);
2556 [ + - ]: 5 : if (process_equivalence(root, &newrinfo, jdomain))
2557 : 5 : matchright = true;
2558 : : }
2559 : : }
2560 : :
2561 : : /*
2562 : : * If we were able to equate both vars to constants, we're done, and
2563 : : * we can throw away the full-join clause as redundant. Moreover, we
2564 : : * can remove the COALESCE entry from the EC, since the added
2565 : : * restrictions ensure it will always have the expected value. (We
2566 : : * don't bother trying to update ec_relids or ec_sources.)
2567 : : */
2568 [ + - + - ]: 5 : if (matchleft && matchright)
2569 : : {
2570 : 5 : cur_ec->ec_members = list_delete_nth_cell(cur_ec->ec_members, coal_idx);
2571 : 5 : return true;
2572 : : }
2573 : :
2574 : : /*
2575 : : * Otherwise, fall out of the search loop, since we know the COALESCE
2576 : : * appears in at most one EC (XXX might stop being true if we allow
2577 : : * stripping of coercions above?)
2578 : : */
2579 : 0 : break;
2580 : : }
2581 : :
2582 : 1069 : return false; /* failed to make any deduction */
2583 : : }
2584 : :
2585 : : /*
2586 : : * find_join_domain
2587 : : * Find the highest JoinDomain enclosed within the given relid set.
2588 : : *
2589 : : * (We could avoid this search at the cost of complicating APIs elsewhere,
2590 : : * which doesn't seem worth it.)
2591 : : */
2592 : : static JoinDomain *
2593 : 1711 : find_join_domain(PlannerInfo *root, Relids relids)
2594 : : {
2595 : : ListCell *lc;
2596 : :
2597 [ + - + - : 3509 : foreach(lc, root->join_domains)
+ - ]
2598 : : {
2599 : 3509 : JoinDomain *jdomain = (JoinDomain *) lfirst(lc);
2600 : :
2601 [ + + ]: 3509 : if (bms_is_subset(jdomain->jd_relids, relids))
2602 : 1711 : return jdomain;
2603 : : }
2604 [ # # ]: 0 : elog(ERROR, "failed to find appropriate JoinDomain");
2605 : : return NULL; /* keep compiler quiet */
2606 : : }
2607 : :
2608 : :
2609 : : /*
2610 : : * exprs_known_equal
2611 : : * Detect whether two expressions are known equal due to equivalence
2612 : : * relationships.
2613 : : *
2614 : : * If opfamily is given, the expressions must be known equal per the semantics
2615 : : * of that opfamily (note it has to be a btree opfamily, since those are the
2616 : : * only opfamilies equivclass.c deals with). If opfamily is InvalidOid, we'll
2617 : : * return true if they're equal according to any opfamily, which is fuzzy but
2618 : : * OK for estimation purposes.
2619 : : *
2620 : : * Note: does not bother to check for "equal(item1, item2)"; caller must
2621 : : * check that case if it's possible to pass identical items.
2622 : : */
2623 : : bool
2624 : 25496 : exprs_known_equal(PlannerInfo *root, Node *item1, Node *item2, Oid opfamily)
2625 : : {
2626 : : ListCell *lc1;
2627 : :
2628 [ + + + + : 194398 : foreach(lc1, root->eq_classes)
+ + ]
2629 : : {
2630 : 171728 : EquivalenceClass *ec = (EquivalenceClass *) lfirst(lc1);
2631 : 171728 : bool item1member = false;
2632 : 171728 : bool item2member = false;
2633 : : ListCell *lc2;
2634 : :
2635 : : /* Never match to a volatile EC */
2636 [ - + ]: 171728 : if (ec->ec_has_volatile)
2637 : 0 : continue;
2638 : :
2639 : : /*
2640 : : * It's okay to consider ec_broken ECs here. Brokenness just means we
2641 : : * couldn't derive all the implied clauses we'd have liked to; it does
2642 : : * not invalidate our knowledge that the members are equal.
2643 : : */
2644 : :
2645 : : /* Ignore if this EC doesn't use specified opfamily */
2646 [ + + ]: 171728 : if (OidIsValid(opfamily) &&
2647 [ + + ]: 550 : !list_member_oid(ec->ec_opfamilies, opfamily))
2648 : 190 : continue;
2649 : :
2650 : : /* Ignore children here */
2651 [ + - + + : 410384 : foreach(lc2, ec->ec_members)
+ + ]
2652 : : {
2653 : 241672 : EquivalenceMember *em = (EquivalenceMember *) lfirst(lc2);
2654 : :
2655 : : /* Child members should not exist in ec_members */
2656 : : Assert(!em->em_is_child);
2657 [ + + ]: 241672 : if (equal(item1, em->em_expr))
2658 : 12808 : item1member = true;
2659 [ + + ]: 228864 : else if (equal(item2, em->em_expr))
2660 : 20908 : item2member = true;
2661 : : /* Exit as soon as equality is proven */
2662 [ + + + + ]: 241672 : if (item1member && item2member)
2663 : 2826 : return true;
2664 : : }
2665 : : }
2666 : 22670 : return false;
2667 : : }
2668 : :
2669 : :
2670 : : /*
2671 : : * match_eclasses_to_foreign_key_col
2672 : : * See whether a foreign key column match is proven by any eclass.
2673 : : *
2674 : : * If the referenced and referencing Vars of the fkey's colno'th column are
2675 : : * known equal due to any eclass, return that eclass; otherwise return NULL.
2676 : : * (In principle there might be more than one matching eclass if multiple
2677 : : * collations are involved, but since collation doesn't matter for equality,
2678 : : * we ignore that fine point here.) This is much like exprs_known_equal,
2679 : : * except for the format of the input.
2680 : : *
2681 : : * On success, we also set fkinfo->eclass[colno] to the matching eclass,
2682 : : * and set fkinfo->fk_eclass_member[colno] to the eclass member for the
2683 : : * referencing Var.
2684 : : */
2685 : : EquivalenceClass *
2686 : 1943 : match_eclasses_to_foreign_key_col(PlannerInfo *root,
2687 : : ForeignKeyOptInfo *fkinfo,
2688 : : int colno)
2689 : : {
2690 : 1943 : Index var1varno = fkinfo->con_relid;
2691 : 1943 : AttrNumber var1attno = fkinfo->conkey[colno];
2692 : 1943 : Index var2varno = fkinfo->ref_relid;
2693 : 1943 : AttrNumber var2attno = fkinfo->confkey[colno];
2694 : 1943 : Oid eqop = fkinfo->conpfeqop[colno];
2695 : 1943 : RelOptInfo *rel1 = root->simple_rel_array[var1varno];
2696 : 1943 : RelOptInfo *rel2 = root->simple_rel_array[var2varno];
2697 : 1943 : List *opfamilies = NIL; /* compute only if needed */
2698 : : Bitmapset *matching_ecs;
2699 : : int i;
2700 : :
2701 : : /* Consider only eclasses mentioning both relations */
2702 : : Assert(root->ec_merging_done);
2703 : : Assert(IS_SIMPLE_REL(rel1));
2704 : : Assert(IS_SIMPLE_REL(rel2));
2705 : 1943 : matching_ecs = bms_intersect(rel1->eclass_indexes,
2706 : 1943 : rel2->eclass_indexes);
2707 : :
2708 : 1943 : i = -1;
2709 [ + + ]: 2023 : while ((i = bms_next_member(matching_ecs, i)) >= 0)
2710 : : {
2711 : 425 : EquivalenceClass *ec = (EquivalenceClass *) list_nth(root->eq_classes,
2712 : : i);
2713 : 425 : EquivalenceMember *item1_em = NULL;
2714 : 425 : EquivalenceMember *item2_em = NULL;
2715 : : ListCell *lc2;
2716 : :
2717 : : /* Never match to a volatile EC */
2718 [ - + ]: 425 : if (ec->ec_has_volatile)
2719 : 0 : continue;
2720 : :
2721 : : /*
2722 : : * It's okay to consider "broken" ECs here, see exprs_known_equal.
2723 : : * Ignore children here.
2724 : : */
2725 [ + - + + : 1015 : foreach(lc2, ec->ec_members)
+ + ]
2726 : : {
2727 : 935 : EquivalenceMember *em = (EquivalenceMember *) lfirst(lc2);
2728 : : Var *var;
2729 : :
2730 : : /* Child members should not exist in ec_members */
2731 : : Assert(!em->em_is_child);
2732 : :
2733 : : /* EM must be a Var, possibly with RelabelType */
2734 : 935 : var = (Var *) em->em_expr;
2735 [ + - - + ]: 935 : while (var && IsA(var, RelabelType))
2736 : 0 : var = (Var *) ((RelabelType *) var)->arg;
2737 [ + - + + ]: 935 : if (!(var && IsA(var, Var)))
2738 : 5 : continue;
2739 : :
2740 : : /* Match? */
2741 [ + + + + ]: 930 : if (var->varno == var1varno && var->varattno == var1attno)
2742 : 345 : item1_em = em;
2743 [ + + + + ]: 585 : else if (var->varno == var2varno && var->varattno == var2attno)
2744 : 345 : item2_em = em;
2745 : :
2746 : : /* Have we found both PK and FK column in this EC? */
2747 [ + + + + ]: 930 : if (item1_em && item2_em)
2748 : : {
2749 : : /*
2750 : : * Succeed if eqop matches EC's opfamilies. We could test
2751 : : * this before scanning the members, but it's probably cheaper
2752 : : * to test for member matches first.
2753 : : */
2754 [ + - ]: 345 : if (opfamilies == NIL) /* compute if we didn't already */
2755 : 345 : opfamilies = get_mergejoin_opfamilies(eqop);
2756 [ + - ]: 345 : if (equal(opfamilies, ec->ec_opfamilies))
2757 : : {
2758 : 345 : fkinfo->eclass[colno] = ec;
2759 : 345 : fkinfo->fk_eclass_member[colno] = item2_em;
2760 : 345 : return ec;
2761 : : }
2762 : : /* Otherwise, done with this EC, move on to the next */
2763 : 0 : break;
2764 : : }
2765 : : }
2766 : : }
2767 : 1598 : return NULL;
2768 : : }
2769 : :
2770 : : /*
2771 : : * find_derived_clause_for_ec_member
2772 : : * Search for a previously-derived clause mentioning the given EM.
2773 : : *
2774 : : * The eclass should be an ec_has_const EC, of which the EM is a non-const
2775 : : * member. This should ensure there is just one derived clause mentioning
2776 : : * the EM (and equating it to a constant).
2777 : : * Returns NULL if no such clause can be found.
2778 : : */
2779 : : RestrictInfo *
2780 : 5 : find_derived_clause_for_ec_member(PlannerInfo *root,
2781 : : EquivalenceClass *ec,
2782 : : EquivalenceMember *em)
2783 : : {
2784 : : Assert(ec->ec_has_const);
2785 : : Assert(!em->em_is_const);
2786 : :
2787 : 5 : return ec_search_derived_clause_for_ems(root, ec, em, NULL, NULL);
2788 : : }
2789 : :
2790 : :
2791 : : /*
2792 : : * add_child_rel_equivalences
2793 : : * Search for EC members that reference the root parent of child_rel, and
2794 : : * add transformed members referencing the child_rel.
2795 : : *
2796 : : * Note that this function won't be called at all unless we have at least some
2797 : : * reason to believe that the EC members it generates will be useful.
2798 : : *
2799 : : * parent_rel and child_rel could be derived from appinfo, but since the
2800 : : * caller has already computed them, we might as well just pass them in.
2801 : : *
2802 : : * The passed-in AppendRelInfo is not used when the parent_rel is not a
2803 : : * top-level baserel, since it shows the mapping from the parent_rel but
2804 : : * we need to translate EC expressions that refer to the top-level parent.
2805 : : * Using it is faster than using adjust_appendrel_attrs_multilevel(), though,
2806 : : * so we prefer it when we can.
2807 : : */
2808 : : void
2809 : 28174 : add_child_rel_equivalences(PlannerInfo *root,
2810 : : AppendRelInfo *appinfo,
2811 : : RelOptInfo *parent_rel,
2812 : : RelOptInfo *child_rel)
2813 : : {
2814 : 28174 : Relids top_parent_relids = child_rel->top_parent_relids;
2815 : 28174 : Relids child_relids = child_rel->relids;
2816 : : int i;
2817 : :
2818 : : /*
2819 : : * EC merging should be complete already, so we can use the parent rel's
2820 : : * eclass_indexes to avoid searching all of root->eq_classes.
2821 : : */
2822 : : Assert(root->ec_merging_done);
2823 : : Assert(IS_SIMPLE_REL(parent_rel));
2824 : :
2825 : 28174 : i = -1;
2826 [ + + ]: 79806 : while ((i = bms_next_member(parent_rel->eclass_indexes, i)) >= 0)
2827 : : {
2828 : 51632 : EquivalenceClass *cur_ec = (EquivalenceClass *) list_nth(root->eq_classes, i);
2829 : :
2830 : : /*
2831 : : * If this EC contains a volatile expression, then generating child
2832 : : * EMs would be downright dangerous, so skip it. We rely on a
2833 : : * volatile EC having only one EM.
2834 : : */
2835 [ - + ]: 51632 : if (cur_ec->ec_has_volatile)
2836 : 0 : continue;
2837 : :
2838 : : /* Sanity check eclass_indexes only contain ECs for parent_rel */
2839 : : Assert(bms_is_subset(top_parent_relids, cur_ec->ec_relids));
2840 : :
2841 [ + - + + : 177546 : foreach_node(EquivalenceMember, cur_em, cur_ec->ec_members)
+ + ]
2842 : : {
2843 [ + + ]: 74282 : if (cur_em->em_is_const)
2844 : 2812 : continue; /* ignore consts here */
2845 : :
2846 : : /* Child members should not exist in ec_members */
2847 : : Assert(!cur_em->em_is_child);
2848 : :
2849 : : /*
2850 : : * Consider only members that reference and can be computed at
2851 : : * child's topmost parent rel. In particular we want to exclude
2852 : : * parent-rel Vars that have nonempty varnullingrels. Translating
2853 : : * those might fail, if the transformed expression wouldn't be a
2854 : : * simple Var; and in any case it wouldn't produce a member that
2855 : : * has any use in creating plans for the child rel.
2856 : : */
2857 [ + + ]: 71470 : if (bms_is_subset(cur_em->em_relids, top_parent_relids) &&
2858 [ + - ]: 49663 : !bms_is_empty(cur_em->em_relids))
2859 : : {
2860 : : /* OK, generate transformed child version */
2861 : : Expr *child_expr;
2862 : : Relids new_relids;
2863 : :
2864 [ + + ]: 49663 : if (parent_rel->reloptkind == RELOPT_BASEREL)
2865 : : {
2866 : : /* Simple single-level transformation */
2867 : : child_expr = (Expr *)
2868 : 42466 : adjust_appendrel_attrs(root,
2869 : 42466 : (Node *) cur_em->em_expr,
2870 : : 1, &appinfo);
2871 : : }
2872 : : else
2873 : : {
2874 : : /* Must do multi-level transformation */
2875 : : child_expr = (Expr *)
2876 : 7197 : adjust_appendrel_attrs_multilevel(root,
2877 : 7197 : (Node *) cur_em->em_expr,
2878 : : child_rel,
2879 : 7197 : child_rel->top_parent);
2880 : : }
2881 : :
2882 : : /*
2883 : : * Transform em_relids to match. Note we do *not* do
2884 : : * pull_varnos(child_expr) here, as for example the
2885 : : * transformation might have substituted a constant, but we
2886 : : * don't want the child member to be marked as constant.
2887 : : */
2888 : 49663 : new_relids = bms_difference(cur_em->em_relids,
2889 : : top_parent_relids);
2890 : 49663 : new_relids = bms_add_members(new_relids, child_relids);
2891 : :
2892 : 49663 : add_child_eq_member(root,
2893 : : cur_ec,
2894 : : i,
2895 : : child_expr,
2896 : : new_relids,
2897 : : cur_em->em_jdomain,
2898 : : cur_em,
2899 : : cur_em->em_datatype,
2900 : : child_rel->relid);
2901 : : }
2902 : : }
2903 : : }
2904 : 28174 : }
2905 : :
2906 : : /*
2907 : : * add_child_join_rel_equivalences
2908 : : * Like add_child_rel_equivalences(), but for joinrels
2909 : : *
2910 : : * Here we find the ECs relevant to the top parent joinrel and add transformed
2911 : : * member expressions that refer to this child joinrel.
2912 : : *
2913 : : * Note that this function won't be called at all unless we have at least some
2914 : : * reason to believe that the EC members it generates will be useful.
2915 : : */
2916 : : void
2917 : 15155 : add_child_join_rel_equivalences(PlannerInfo *root,
2918 : : int nappinfos, AppendRelInfo **appinfos,
2919 : : RelOptInfo *parent_joinrel,
2920 : : RelOptInfo *child_joinrel)
2921 : : {
2922 : 15155 : Relids top_parent_relids = child_joinrel->top_parent_relids;
2923 : 15155 : Relids child_relids = child_joinrel->relids;
2924 : : Bitmapset *matching_ecs;
2925 : : MemoryContext oldcontext;
2926 : : int i;
2927 : :
2928 : : Assert(IS_JOIN_REL(child_joinrel) && IS_JOIN_REL(parent_joinrel));
2929 : :
2930 : : /* We need consider only ECs that mention the parent joinrel */
2931 : 15155 : matching_ecs = get_eclass_indexes_for_relids(root, top_parent_relids);
2932 : :
2933 : : /*
2934 : : * If we're being called during GEQO join planning, we still have to
2935 : : * create any new EC members in the main planner context, to avoid having
2936 : : * a corrupt EC data structure after the GEQO context is reset. This is
2937 : : * problematic since we'll leak memory across repeated GEQO cycles. For
2938 : : * now, though, bloat is better than crash. If it becomes a real issue
2939 : : * we'll have to do something to avoid generating duplicate EC members.
2940 : : */
2941 : 15155 : oldcontext = MemoryContextSwitchTo(root->planner_cxt);
2942 : :
2943 : 15155 : i = -1;
2944 [ + + ]: 40881 : while ((i = bms_next_member(matching_ecs, i)) >= 0)
2945 : : {
2946 : 25726 : EquivalenceClass *cur_ec = (EquivalenceClass *) list_nth(root->eq_classes, i);
2947 : :
2948 : : /*
2949 : : * If this EC contains a volatile expression, then generating child
2950 : : * EMs would be downright dangerous, so skip it. We rely on a
2951 : : * volatile EC having only one EM.
2952 : : */
2953 [ - + ]: 25726 : if (cur_ec->ec_has_volatile)
2954 : 0 : continue;
2955 : :
2956 : : /* Sanity check on get_eclass_indexes_for_relids result */
2957 : : Assert(bms_overlap(top_parent_relids, cur_ec->ec_relids));
2958 : :
2959 [ + - + + : 95712 : foreach_node(EquivalenceMember, cur_em, cur_ec->ec_members)
+ + ]
2960 : : {
2961 [ + + ]: 44260 : if (cur_em->em_is_const)
2962 : 2018 : continue; /* ignore consts here */
2963 : :
2964 : : /* Child members should not exist in ec_members */
2965 : : Assert(!cur_em->em_is_child);
2966 : :
2967 : : /*
2968 : : * We may ignore expressions that reference a single baserel,
2969 : : * because add_child_rel_equivalences should have handled them.
2970 : : */
2971 [ + + ]: 42242 : if (bms_membership(cur_em->em_relids) != BMS_MULTIPLE)
2972 : 39915 : continue;
2973 : :
2974 : : /* Does this member reference child's topmost parent rel? */
2975 [ + - ]: 2327 : if (bms_overlap(cur_em->em_relids, top_parent_relids))
2976 : : {
2977 : : /* Yes, generate transformed child version */
2978 : : Expr *child_expr;
2979 : : Relids new_relids;
2980 : :
2981 [ + + ]: 2327 : if (parent_joinrel->reloptkind == RELOPT_JOINREL)
2982 : : {
2983 : : /* Simple single-level transformation */
2984 : : child_expr = (Expr *)
2985 : 2247 : adjust_appendrel_attrs(root,
2986 : 2247 : (Node *) cur_em->em_expr,
2987 : : nappinfos, appinfos);
2988 : : }
2989 : : else
2990 : : {
2991 : : /* Must do multi-level transformation */
2992 : : Assert(parent_joinrel->reloptkind == RELOPT_OTHER_JOINREL);
2993 : : child_expr = (Expr *)
2994 : 80 : adjust_appendrel_attrs_multilevel(root,
2995 : 80 : (Node *) cur_em->em_expr,
2996 : : child_joinrel,
2997 : 80 : child_joinrel->top_parent);
2998 : : }
2999 : :
3000 : : /*
3001 : : * Transform em_relids to match. Note we do *not* do
3002 : : * pull_varnos(child_expr) here, as for example the
3003 : : * transformation might have substituted a constant, but we
3004 : : * don't want the child member to be marked as constant.
3005 : : */
3006 : 2327 : new_relids = bms_difference(cur_em->em_relids,
3007 : : top_parent_relids);
3008 : 2327 : new_relids = bms_add_members(new_relids, child_relids);
3009 : :
3010 : : /*
3011 : : * Add new child member to the EquivalenceClass. Because this
3012 : : * is a RELOPT_OTHER_JOINREL which has multiple component
3013 : : * relids, there is no ideal place to store these members in
3014 : : * the class. Ordinarily, child members are stored in the
3015 : : * ec_childmembers[] array element corresponding to their
3016 : : * relid, however, here we have multiple component relids, so
3017 : : * there's no single ec_childmembers[] array element to store
3018 : : * this member. So that we still correctly find this member
3019 : : * in loops iterating over an EquivalenceMemberIterator, we
3020 : : * opt to store the member in the ec_childmembers array in
3021 : : * only the first component relid slot of the array. This
3022 : : * allows the member to be found, providing callers of
3023 : : * setup_eclass_member_iterator() specify all the component
3024 : : * relids for the RELOPT_OTHER_JOINREL, which they do. If we
3025 : : * opted to store the member in each ec_childmembers[] element
3026 : : * for all the component relids, then that would just result
3027 : : * in eclass_member_iterator_next() finding the member
3028 : : * multiple times, which is a waste of effort.
3029 : : */
3030 : 2327 : add_child_eq_member(root,
3031 : : cur_ec,
3032 : : -1,
3033 : : child_expr,
3034 : : new_relids,
3035 : : cur_em->em_jdomain,
3036 : : cur_em,
3037 : : cur_em->em_datatype,
3038 : 2327 : bms_next_member(child_joinrel->relids, -1));
3039 : : }
3040 : : }
3041 : : }
3042 : :
3043 : 15155 : MemoryContextSwitchTo(oldcontext);
3044 : 15155 : }
3045 : :
3046 : : /*
3047 : : * add_child_rel_pathkey_equivalences
3048 : : * Make sure the ECs of the given pathkeys have members for child_rel.
3049 : : *
3050 : : * An EC created after its relations' children were processed has no child
3051 : : * members, so child_rel could not be sorted by it. Add them here.
3052 : : */
3053 : : void
3054 : 590 : add_child_rel_pathkey_equivalences(PlannerInfo *root, RelOptInfo *child_rel,
3055 : : List *pathkeys)
3056 : : {
3057 : 590 : Relids top_parent_relids = child_rel->top_parent_relids;
3058 : : MemoryContext oldcontext;
3059 : : ListCell *lc;
3060 : :
3061 : : Assert(IS_OTHER_REL(child_rel));
3062 : :
3063 : : /* As in add_child_join_rel_equivalences, new members must survive GEQO */
3064 : 590 : oldcontext = MemoryContextSwitchTo(root->planner_cxt);
3065 : :
3066 [ + - + + : 1340 : foreach(lc, pathkeys)
+ + ]
3067 : : {
3068 : 750 : EquivalenceClass *ec = lfirst_node(PathKey, lc)->pk_eclass;
3069 : :
3070 [ - + ]: 750 : if (ec->ec_has_volatile)
3071 : 0 : continue;
3072 : :
3073 [ + - + + : 3040 : foreach_node(EquivalenceMember, cur_em, ec->ec_members)
+ + ]
3074 : : {
3075 : : EquivalenceMemberIterator it;
3076 : : EquivalenceMember *em;
3077 : : Expr *child_expr;
3078 : : Relids new_relids;
3079 : : int child_relid;
3080 : :
3081 : : /* Consider only members computable at the topmost parent */
3082 [ + - ]: 1540 : if (cur_em->em_is_const ||
3083 [ + + ]: 1540 : !bms_is_subset(cur_em->em_relids, top_parent_relids))
3084 : 1480 : continue;
3085 : :
3086 : : /* Skip members that already have a child version for this rel */
3087 : 950 : setup_eclass_member_iterator(&it, ec, child_rel->relids);
3088 [ + + ]: 4240 : while ((em = eclass_member_iterator_next(&it)) != NULL)
3089 : : {
3090 [ + + + - ]: 4120 : if (em->em_parent == cur_em &&
3091 : 890 : bms_is_subset(em->em_relids, child_rel->relids))
3092 : 890 : break;
3093 : : }
3094 [ + + ]: 950 : if (em != NULL)
3095 : 890 : continue;
3096 : :
3097 : 60 : new_relids = adjust_child_relids_multilevel(root,
3098 : : cur_em->em_relids,
3099 : : child_rel,
3100 : 60 : child_rel->top_parent);
3101 : :
3102 : : /* Store the member under one of its child relations */
3103 : 60 : child_relid = bms_next_member(bms_difference(new_relids,
3104 : : top_parent_relids),
3105 : : -1);
3106 [ - + ]: 60 : if (child_relid < 0)
3107 : 0 : continue;
3108 : :
3109 : : child_expr = (Expr *)
3110 : 60 : adjust_appendrel_attrs_multilevel(root,
3111 : 60 : (Node *) cur_em->em_expr,
3112 : : child_rel,
3113 : 60 : child_rel->top_parent);
3114 : :
3115 : 60 : add_child_eq_member(root, ec, -1, child_expr, new_relids,
3116 : : cur_em->em_jdomain, cur_em,
3117 : : cur_em->em_datatype, child_relid);
3118 : : }
3119 : : }
3120 : :
3121 : 590 : MemoryContextSwitchTo(oldcontext);
3122 : 590 : }
3123 : :
3124 : : /*
3125 : : * add_setop_child_rel_equivalences
3126 : : * Add equivalence members for each non-resjunk target in 'child_tlist'
3127 : : * to the EquivalenceClass in the corresponding setop_pathkey's pk_eclass.
3128 : : *
3129 : : * 'root' is the PlannerInfo belonging to the top-level set operation.
3130 : : * 'child_rel' is the RelOptInfo of the child relation we're adding
3131 : : * EquivalenceMembers for.
3132 : : * 'child_tlist' is the target list for the setop child relation. The target
3133 : : * list expressions are what we add as EquivalenceMembers.
3134 : : * 'setop_pathkeys' is a list of PathKeys which must contain an entry for each
3135 : : * non-resjunk target in 'child_tlist'.
3136 : : */
3137 : : void
3138 : 10387 : add_setop_child_rel_equivalences(PlannerInfo *root, RelOptInfo *child_rel,
3139 : : List *child_tlist, List *setop_pathkeys)
3140 : : {
3141 : : ListCell *lc;
3142 : 10387 : ListCell *lc2 = list_head(setop_pathkeys);
3143 : :
3144 [ + - + + : 42290 : foreach(lc, child_tlist)
+ + ]
3145 : : {
3146 : 31903 : TargetEntry *tle = lfirst_node(TargetEntry, lc);
3147 : : EquivalenceMember *parent_em;
3148 : : PathKey *pk;
3149 : :
3150 [ - + ]: 31903 : if (tle->resjunk)
3151 : 0 : continue;
3152 : :
3153 [ - + ]: 31903 : if (lc2 == NULL)
3154 [ # # ]: 0 : elog(ERROR, "too few pathkeys for set operation");
3155 : :
3156 : 31903 : pk = lfirst_node(PathKey, lc2);
3157 : 31903 : parent_em = linitial(pk->pk_eclass->ec_members);
3158 : :
3159 : : /*
3160 : : * We can safely pass the parent member as the first member in the
3161 : : * ec_members list as this is added first in generate_union_paths,
3162 : : * likewise, the JoinDomain can be that of the initial member of the
3163 : : * Pathkey's EquivalenceClass. We pass -1 for ec_index since we
3164 : : * maintain the eclass_indexes for the child_rel after the loop.
3165 : : */
3166 : 31903 : add_child_eq_member(root,
3167 : : pk->pk_eclass,
3168 : : -1,
3169 : : tle->expr,
3170 : : child_rel->relids,
3171 : : parent_em->em_jdomain,
3172 : : parent_em,
3173 : 31903 : exprType((Node *) tle->expr),
3174 : : child_rel->relid);
3175 : :
3176 : 31903 : lc2 = lnext(setop_pathkeys, lc2);
3177 : : }
3178 : :
3179 : : /*
3180 : : * transformSetOperationStmt() ensures that the targetlist never contains
3181 : : * any resjunk columns, so all eclasses that exist in 'root' must have
3182 : : * received a new member in the loop above. Add them to the child_rel's
3183 : : * eclass_indexes.
3184 : : */
3185 : 10387 : child_rel->eclass_indexes = bms_add_range(child_rel->eclass_indexes, 0,
3186 : 10387 : list_length(root->eq_classes) - 1);
3187 : 10387 : }
3188 : :
3189 : : /*
3190 : : * setup_eclass_member_iterator
3191 : : * Setup an EquivalenceMemberIterator 'it' to iterate over all parent
3192 : : * EquivalenceMembers and child members belonging to the given 'ec'.
3193 : : *
3194 : : * This iterator returns:
3195 : : * - All parent members stored directly in ec_members for 'ec', and;
3196 : : * - Any child member added to the given ec by add_child_eq_member() where
3197 : : * the child_relid specified in the add_child_eq_member() call is a member
3198 : : * of the 'child_relids' parameter.
3199 : : *
3200 : : * Note:
3201 : : * The given 'child_relids' must remain allocated and not be changed for the
3202 : : * lifetime of the iterator.
3203 : : *
3204 : : * Parameters:
3205 : : * 'it' is a pointer to the iterator to set up. Normally stack allocated.
3206 : : * 'ec' is the EquivalenceClass from which to iterate members for.
3207 : : * 'child_relids' is the relids to return child members for.
3208 : : */
3209 : : void
3210 : 3955534 : setup_eclass_member_iterator(EquivalenceMemberIterator *it,
3211 : : EquivalenceClass *ec, Relids child_relids)
3212 : : {
3213 : 3955534 : it->ec = ec;
3214 : : /* no need to set this if the class has no child members array set */
3215 [ + + ]: 3955534 : it->child_relids = ec->ec_childmembers != NULL ? child_relids : NULL;
3216 : 3955534 : it->current_relid = -1;
3217 : 3955534 : it->current_list = ec->ec_members;
3218 : 3955534 : it->current_cell = list_head(it->current_list);
3219 : 3955534 : }
3220 : :
3221 : : /*
3222 : : * eclass_member_iterator_next
3223 : : * Get the next EquivalenceMember from the EquivalenceMemberIterator 'it',
3224 : : * as setup by setup_eclass_member_iterator(). NULL is returned if there
3225 : : * are no members left, after which callers must not call
3226 : : * eclass_member_iterator_next() again for the given iterator.
3227 : : */
3228 : : EquivalenceMember *
3229 : 9205963 : eclass_member_iterator_next(EquivalenceMemberIterator *it)
3230 : : {
3231 [ + - ]: 9205963 : while (it->current_list != NULL)
3232 : : {
3233 [ + + ]: 9205963 : while (it->current_cell != NULL)
3234 : : {
3235 : : EquivalenceMember *em;
3236 : :
3237 : 6361731 : nextcell:
3238 : 6477010 : em = lfirst_node(EquivalenceMember, it->current_cell);
3239 : 6477010 : it->current_cell = lnext(it->current_list, it->current_cell);
3240 : 6477010 : return em;
3241 : : }
3242 : :
3243 : : /* Search for the next list to return members from */
3244 [ + + ]: 2951161 : while ((it->current_relid = bms_next_member(it->child_relids, it->current_relid)) > 0)
3245 : : {
3246 : : /*
3247 : : * Be paranoid in case we're given relids above what we've sized
3248 : : * the ec_childmembers array to.
3249 : : */
3250 [ - + ]: 222208 : if (it->current_relid >= it->ec->ec_childmembers_size)
3251 : 0 : return NULL;
3252 : :
3253 : 222208 : it->current_list = it->ec->ec_childmembers[it->current_relid];
3254 : :
3255 : : /* If there are members in this list, use it. */
3256 [ + + ]: 222208 : if (it->current_list != NIL)
3257 : : {
3258 : : /* point current_cell to the head of this list */
3259 : 115279 : it->current_cell = list_head(it->current_list);
3260 : 115279 : goto nextcell;
3261 : : }
3262 : : }
3263 : 2728953 : return NULL;
3264 : : }
3265 : :
3266 : 0 : return NULL;
3267 : : }
3268 : :
3269 : : /*
3270 : : * generate_implied_equalities_for_column
3271 : : * Create EC-derived joinclauses usable with a specific column.
3272 : : *
3273 : : * This is used by indxpath.c to extract potentially indexable joinclauses
3274 : : * from ECs, and can be used by foreign data wrappers for similar purposes.
3275 : : * We assume that only expressions in Vars of a single table are of interest,
3276 : : * but the caller provides a callback function to identify exactly which
3277 : : * such expressions it would like to know about.
3278 : : *
3279 : : * We assume that any given table/index column could appear in only one EC.
3280 : : * (This should be true in all but the most pathological cases, and if it
3281 : : * isn't, we stop on the first match anyway.) Therefore, what we return
3282 : : * is a redundant list of clauses equating the table/index column to each of
3283 : : * the other-relation values it is known to be equal to. Any one of
3284 : : * these clauses can be used to create a parameterized path, and there
3285 : : * is no value in using more than one. (But it *is* worthwhile to create
3286 : : * a separate parameterized path for each one, since that leads to different
3287 : : * join orders.)
3288 : : *
3289 : : * The caller can pass a Relids set of rels we aren't interested in joining
3290 : : * to, so as to save the work of creating useless clauses.
3291 : : */
3292 : : List *
3293 : 465781 : generate_implied_equalities_for_column(PlannerInfo *root,
3294 : : RelOptInfo *rel,
3295 : : ec_matches_callback_type callback,
3296 : : void *callback_arg,
3297 : : Relids prohibited_rels)
3298 : : {
3299 : 465781 : List *result = NIL;
3300 : 465781 : bool is_child_rel = (rel->reloptkind == RELOPT_OTHER_MEMBER_REL);
3301 : : Relids parent_relids;
3302 : : int i;
3303 : :
3304 : : /* Should be OK to rely on eclass_indexes */
3305 : : Assert(root->ec_merging_done);
3306 : :
3307 : : /* Indexes are available only on base or "other" member relations. */
3308 : : Assert(IS_SIMPLE_REL(rel));
3309 : :
3310 : : /* If it's a child rel, we'll need to know what its parent(s) are */
3311 [ + + ]: 465781 : if (is_child_rel)
3312 : 10192 : parent_relids = find_childrel_parents(root, rel);
3313 : : else
3314 : 455589 : parent_relids = NULL; /* not used, but keep compiler quiet */
3315 : :
3316 : 465781 : i = -1;
3317 [ + + ]: 1340221 : while ((i = bms_next_member(rel->eclass_indexes, i)) >= 0)
3318 : : {
3319 : 970129 : EquivalenceClass *cur_ec = (EquivalenceClass *) list_nth(root->eq_classes, i);
3320 : : EquivalenceMemberIterator it;
3321 : : EquivalenceMember *cur_em;
3322 : : ListCell *lc2;
3323 : :
3324 : : /* Sanity check eclass_indexes only contain ECs for rel */
3325 : : Assert(is_child_rel || bms_is_subset(rel->relids, cur_ec->ec_relids));
3326 : :
3327 : : /*
3328 : : * Won't generate joinclauses if const or single-member (the latter
3329 : : * test covers the volatile case too)
3330 : : */
3331 [ + + + + ]: 970129 : if (cur_ec->ec_has_const || list_length(cur_ec->ec_members) <= 1)
3332 : 874280 : continue;
3333 : :
3334 : : /*
3335 : : * Scan members, looking for a match to the target column. Note that
3336 : : * child EC members are considered, but only when they belong to the
3337 : : * target relation. (Unlike regular members, the same expression
3338 : : * could be a child member of more than one EC. Therefore, it's
3339 : : * potentially order-dependent which EC a child relation's target
3340 : : * column gets matched to. This is annoying but it only happens in
3341 : : * corner cases, so for now we live with just reporting the first
3342 : : * match. See also get_eclass_for_sort_expr.)
3343 : : */
3344 : 459659 : setup_eclass_member_iterator(&it, cur_ec, rel->relids);
3345 [ + + ]: 1727853 : while ((cur_em = eclass_member_iterator_next(&it)) != NULL)
3346 : : {
3347 [ + + + + ]: 1364213 : if (bms_equal(cur_em->em_relids, rel->relids) &&
3348 : 459829 : callback(root, rel, cur_ec, cur_em, callback_arg))
3349 : 95849 : break;
3350 : : }
3351 : :
3352 [ + + ]: 459659 : if (!cur_em)
3353 : 363810 : continue;
3354 : :
3355 : : /*
3356 : : * Found our match. Scan the other EC members and attempt to generate
3357 : : * joinclauses. Ignore children here.
3358 : : */
3359 [ + - + + : 292912 : foreach(lc2, cur_ec->ec_members)
+ + ]
3360 : : {
3361 : 197063 : EquivalenceMember *other_em = (EquivalenceMember *) lfirst(lc2);
3362 : : Oid eq_op;
3363 : : RestrictInfo *rinfo;
3364 : :
3365 : : /* Child members should not exist in ec_members */
3366 : : Assert(!other_em->em_is_child);
3367 : :
3368 : : /* Make sure it'll be a join to a different rel */
3369 [ + + + + ]: 301163 : if (other_em == cur_em ||
3370 : 104100 : bms_overlap(other_em->em_relids, rel->relids))
3371 : 93048 : continue;
3372 : :
3373 : : /* Forget it if caller doesn't want joins to this rel */
3374 [ + + ]: 104015 : if (bms_overlap(other_em->em_relids, prohibited_rels))
3375 : 130 : continue;
3376 : :
3377 : : /*
3378 : : * Also, if this is a child rel, avoid generating a useless join
3379 : : * to its parent rel(s).
3380 : : */
3381 [ + + + + ]: 110364 : if (is_child_rel &&
3382 : 6479 : bms_overlap(parent_relids, other_em->em_relids))
3383 : 2956 : continue;
3384 : :
3385 : 100929 : eq_op = select_equality_operator(cur_ec,
3386 : : cur_em->em_datatype,
3387 : : other_em->em_datatype);
3388 [ - + ]: 100929 : if (!OidIsValid(eq_op))
3389 : 0 : continue;
3390 : :
3391 : : /* set parent_ec to mark as redundant with other joinclauses */
3392 : 100929 : rinfo = create_join_clause(root, cur_ec, eq_op,
3393 : : cur_em, other_em,
3394 : : cur_ec);
3395 : :
3396 : 100929 : result = lappend(result, rinfo);
3397 : : }
3398 : :
3399 : : /*
3400 : : * If somehow we failed to create any join clauses, we might as well
3401 : : * keep scanning the ECs for another match. But if we did make any,
3402 : : * we're done, because we don't want to return non-redundant clauses.
3403 : : */
3404 [ + + ]: 95849 : if (result)
3405 : 95689 : break;
3406 : : }
3407 : :
3408 : 465781 : return result;
3409 : : }
3410 : :
3411 : : /*
3412 : : * have_relevant_eclass_joinclause
3413 : : * Detect whether there is an EquivalenceClass that could produce
3414 : : * a joinclause involving the two given relations.
3415 : : *
3416 : : * This is essentially a very cut-down version of
3417 : : * generate_join_implied_equalities(). Note it's OK to occasionally say "yes"
3418 : : * incorrectly. Hence we don't bother with details like whether the lack of a
3419 : : * cross-type operator might prevent the clause from actually being generated.
3420 : : * False negatives are not always fatal either: they will discourage, but not
3421 : : * completely prevent, investigation of particular join pathways.
3422 : : */
3423 : : bool
3424 : 152957 : have_relevant_eclass_joinclause(PlannerInfo *root,
3425 : : RelOptInfo *rel1, RelOptInfo *rel2)
3426 : : {
3427 : : Bitmapset *matching_ecs;
3428 : : int i;
3429 : :
3430 : : /*
3431 : : * Examine only eclasses mentioning both rel1 and rel2.
3432 : : *
3433 : : * Note that we do not consider the possibility of an eclass generating
3434 : : * "join" clauses that mention just one of the rels plus an outer join
3435 : : * that could be formed from them. Although such clauses must be
3436 : : * correctly enforced when we form the outer join, they don't seem like
3437 : : * sufficient reason to prioritize this join over other ones. The join
3438 : : * ordering rules will force the join to be made when necessary.
3439 : : */
3440 : 152957 : matching_ecs = get_common_eclass_indexes(root, rel1->relids,
3441 : : rel2->relids);
3442 : :
3443 : 152957 : i = -1;
3444 [ + + ]: 153017 : while ((i = bms_next_member(matching_ecs, i)) >= 0)
3445 : : {
3446 : 130523 : EquivalenceClass *ec = (EquivalenceClass *) list_nth(root->eq_classes,
3447 : : i);
3448 : :
3449 : : /*
3450 : : * Sanity check that get_common_eclass_indexes gave only ECs
3451 : : * containing both rels.
3452 : : */
3453 : : Assert(bms_overlap(rel1->relids, ec->ec_relids));
3454 : : Assert(bms_overlap(rel2->relids, ec->ec_relids));
3455 : :
3456 : : /*
3457 : : * Won't generate joinclauses if single-member (this test covers the
3458 : : * volatile case too)
3459 : : */
3460 [ + + ]: 130523 : if (list_length(ec->ec_members) <= 1)
3461 : 60 : continue;
3462 : :
3463 : : /*
3464 : : * We do not need to examine the individual members of the EC, because
3465 : : * all that we care about is whether each rel overlaps the relids of
3466 : : * at least one member, and get_common_eclass_indexes() and the single
3467 : : * member check above are sufficient to prove that. (As with
3468 : : * have_relevant_joinclause(), it is not necessary that the EC be able
3469 : : * to form a joinclause relating exactly the two given rels, only that
3470 : : * it be able to form a joinclause mentioning both, and this will
3471 : : * surely be true if both of them overlap ec_relids.)
3472 : : *
3473 : : * Note we don't test ec_broken; if we did, we'd need a separate code
3474 : : * path to look through ec_sources. Checking the membership anyway is
3475 : : * OK as a possibly-overoptimistic heuristic.
3476 : : *
3477 : : * We don't test ec_has_const either, even though a const eclass won't
3478 : : * generate real join clauses. This is because if we had "WHERE a.x =
3479 : : * b.y and a.x = 42", it is worth considering a join between a and b,
3480 : : * since the join result is likely to be small even though it'll end
3481 : : * up being an unqualified nestloop.
3482 : : */
3483 : :
3484 : 130463 : return true;
3485 : : }
3486 : :
3487 : 22494 : return false;
3488 : : }
3489 : :
3490 : :
3491 : : /*
3492 : : * has_relevant_eclass_joinclause
3493 : : * Detect whether there is an EquivalenceClass that could produce
3494 : : * a joinclause involving the given relation and anything else.
3495 : : *
3496 : : * This is the same as have_relevant_eclass_joinclause with the other rel
3497 : : * implicitly defined as "everything else in the query".
3498 : : */
3499 : : bool
3500 : 173720 : has_relevant_eclass_joinclause(PlannerInfo *root, RelOptInfo *rel1)
3501 : : {
3502 : : Bitmapset *matched_ecs;
3503 : : int i;
3504 : :
3505 : : /* Examine only eclasses mentioning rel1 */
3506 : 173720 : matched_ecs = get_eclass_indexes_for_relids(root, rel1->relids);
3507 : :
3508 : 173720 : i = -1;
3509 [ + + ]: 597112 : while ((i = bms_next_member(matched_ecs, i)) >= 0)
3510 : : {
3511 : 489568 : EquivalenceClass *ec = (EquivalenceClass *) list_nth(root->eq_classes,
3512 : : i);
3513 : :
3514 : : /*
3515 : : * Won't generate joinclauses if single-member (this test covers the
3516 : : * volatile case too)
3517 : : */
3518 [ + + ]: 489568 : if (list_length(ec->ec_members) <= 1)
3519 : 217850 : continue;
3520 : :
3521 : : /*
3522 : : * Per the comment in have_relevant_eclass_joinclause, it's sufficient
3523 : : * to find an EC that mentions both this rel and some other rel.
3524 : : */
3525 [ + + ]: 271718 : if (!bms_is_subset(ec->ec_relids, rel1->relids))
3526 : 66176 : return true;
3527 : : }
3528 : :
3529 : 107544 : return false;
3530 : : }
3531 : :
3532 : :
3533 : : /*
3534 : : * eclass_useful_for_merging
3535 : : * Detect whether the EC could produce any mergejoinable join clauses
3536 : : * against the specified relation.
3537 : : *
3538 : : * This is just a heuristic test and doesn't have to be exact; it's better
3539 : : * to say "yes" incorrectly than "no". Hence we don't bother with details
3540 : : * like whether the lack of a cross-type operator might prevent the clause
3541 : : * from actually being generated.
3542 : : */
3543 : : bool
3544 : 573346 : eclass_useful_for_merging(PlannerInfo *root,
3545 : : EquivalenceClass *eclass,
3546 : : RelOptInfo *rel)
3547 : : {
3548 : : Relids relids;
3549 : : ListCell *lc;
3550 : :
3551 : : Assert(!eclass->ec_merged);
3552 : :
3553 : : /*
3554 : : * Won't generate joinclauses if const or single-member (the latter test
3555 : : * covers the volatile case too)
3556 : : */
3557 [ + + + + ]: 573346 : if (eclass->ec_has_const || list_length(eclass->ec_members) <= 1)
3558 : 30724 : return false;
3559 : :
3560 : : /*
3561 : : * Note we don't test ec_broken; if we did, we'd need a separate code path
3562 : : * to look through ec_sources. Checking the members anyway is OK as a
3563 : : * possibly-overoptimistic heuristic.
3564 : : */
3565 : :
3566 : : /* If specified rel is a child, we must consider the topmost parent rel */
3567 [ + + + + : 542622 : if (IS_OTHER_REL(rel))
- + ]
3568 : : {
3569 : : Assert(!bms_is_empty(rel->top_parent_relids));
3570 : 5165 : relids = rel->top_parent_relids;
3571 : : }
3572 : : else
3573 : 537457 : relids = rel->relids;
3574 : :
3575 : : /* If rel already includes all members of eclass, no point in searching */
3576 [ + + ]: 542622 : if (bms_is_subset(eclass->ec_relids, relids))
3577 : 197619 : return false;
3578 : :
3579 : : /*
3580 : : * To join, we need a member not in the given rel. Ignore children here.
3581 : : */
3582 [ + - + + : 536554 : foreach(lc, eclass->ec_members)
+ + ]
3583 : : {
3584 : 535759 : EquivalenceMember *cur_em = (EquivalenceMember *) lfirst(lc);
3585 : :
3586 : : /* Child members should not exist in ec_members */
3587 : : Assert(!cur_em->em_is_child);
3588 : :
3589 [ + + ]: 535759 : if (!bms_overlap(cur_em->em_relids, relids))
3590 : 344208 : return true;
3591 : : }
3592 : :
3593 : 795 : return false;
3594 : : }
3595 : :
3596 : :
3597 : : /*
3598 : : * is_redundant_derived_clause
3599 : : * Test whether rinfo is derived from same EC as any clause in clauselist;
3600 : : * if so, it can be presumed to represent a condition that's redundant
3601 : : * with that member of the list.
3602 : : */
3603 : : bool
3604 : 70 : is_redundant_derived_clause(RestrictInfo *rinfo, List *clauselist)
3605 : : {
3606 : 70 : EquivalenceClass *parent_ec = rinfo->parent_ec;
3607 : : ListCell *lc;
3608 : :
3609 : : /* Fail if it's not a potentially-redundant clause from some EC */
3610 [ + - ]: 70 : if (parent_ec == NULL)
3611 : 70 : return false;
3612 : :
3613 [ # # # # : 0 : foreach(lc, clauselist)
# # ]
3614 : : {
3615 : 0 : RestrictInfo *otherrinfo = (RestrictInfo *) lfirst(lc);
3616 : :
3617 [ # # ]: 0 : if (otherrinfo->parent_ec == parent_ec)
3618 : 0 : return true;
3619 : : }
3620 : :
3621 : 0 : return false;
3622 : : }
3623 : :
3624 : : /*
3625 : : * is_redundant_with_indexclauses
3626 : : * Test whether rinfo is redundant with any clause in the IndexClause
3627 : : * list. Here, for convenience, we test both simple identity and
3628 : : * whether it is derived from the same EC as any member of the list.
3629 : : */
3630 : : bool
3631 : 1145252 : is_redundant_with_indexclauses(RestrictInfo *rinfo, List *indexclauses)
3632 : : {
3633 : 1145252 : EquivalenceClass *parent_ec = rinfo->parent_ec;
3634 : : ListCell *lc;
3635 : :
3636 [ + + + + : 1594514 : foreach(lc, indexclauses)
+ + ]
3637 : : {
3638 : 1158730 : IndexClause *iclause = lfirst_node(IndexClause, lc);
3639 : 1158730 : RestrictInfo *otherrinfo = iclause->rinfo;
3640 : :
3641 : : /* If indexclause is lossy, it won't enforce the condition exactly */
3642 [ + + ]: 1158730 : if (iclause->lossy)
3643 : 32467 : continue;
3644 : :
3645 : : /* Match if it's same clause (pointer equality should be enough) */
3646 [ + + ]: 1126263 : if (rinfo == otherrinfo)
3647 : 709468 : return true;
3648 : : /* Match if derived from same EC */
3649 [ + + + + ]: 417239 : if (parent_ec && otherrinfo->parent_ec == parent_ec)
3650 : 444 : return true;
3651 : :
3652 : : /*
3653 : : * No need to look at the derived clauses in iclause->indexquals; they
3654 : : * couldn't match if the parent clause didn't.
3655 : : */
3656 : : }
3657 : :
3658 : 435784 : return false;
3659 : : }
3660 : :
3661 : : /*
3662 : : * get_eclass_indexes_for_relids
3663 : : * Build and return a Bitmapset containing the indexes into root's
3664 : : * eq_classes list for all eclasses that mention any of these relids
3665 : : */
3666 : : static Bitmapset *
3667 : 897295 : get_eclass_indexes_for_relids(PlannerInfo *root, Relids relids)
3668 : : {
3669 : 897295 : Bitmapset *ec_indexes = NULL;
3670 : 897295 : int i = -1;
3671 : :
3672 : : /* Should be OK to rely on eclass_indexes */
3673 : : Assert(root->ec_merging_done);
3674 : :
3675 [ + + ]: 2766181 : while ((i = bms_next_member(relids, i)) > 0)
3676 : : {
3677 : 1868886 : RelOptInfo *rel = root->simple_rel_array[i];
3678 : :
3679 : : /* ignore the RTE_GROUP RTE */
3680 [ - + ]: 1868886 : if (i == root->group_rtindex)
3681 : 0 : continue;
3682 : :
3683 [ + + ]: 1868886 : if (rel == NULL) /* must be an outer join */
3684 : : {
3685 : : Assert(bms_is_member(i, root->outer_join_rels));
3686 : 215749 : continue;
3687 : : }
3688 : :
3689 : 1653137 : ec_indexes = bms_add_members(ec_indexes, rel->eclass_indexes);
3690 : : }
3691 : 897295 : return ec_indexes;
3692 : : }
3693 : :
3694 : : /*
3695 : : * get_common_eclass_indexes
3696 : : * Build and return a Bitmapset containing the indexes into root's
3697 : : * eq_classes list for all eclasses that mention rels in both
3698 : : * relids1 and relids2.
3699 : : */
3700 : : static Bitmapset *
3701 : 524713 : get_common_eclass_indexes(PlannerInfo *root, Relids relids1, Relids relids2)
3702 : : {
3703 : : Bitmapset *rel1ecs;
3704 : : Bitmapset *rel2ecs;
3705 : : int relid;
3706 : :
3707 : 524713 : rel1ecs = get_eclass_indexes_for_relids(root, relids1);
3708 : :
3709 : : /*
3710 : : * We can get away with just using the relation's eclass_indexes directly
3711 : : * when relids2 is a singleton set.
3712 : : */
3713 [ + + ]: 524713 : if (bms_get_singleton_member(relids2, &relid))
3714 : 404139 : rel2ecs = root->simple_rel_array[relid]->eclass_indexes;
3715 : : else
3716 : 120574 : rel2ecs = get_eclass_indexes_for_relids(root, relids2);
3717 : :
3718 : : /* Calculate and return the common EC indexes, recycling the left input. */
3719 : 524713 : return bms_int_members(rel1ecs, rel2ecs);
3720 : : }
3721 : :
3722 : : /*
3723 : : * ec_build_derives_hash
3724 : : * Construct the auxiliary hash table for derived clause lookups.
3725 : : */
3726 : : static void
3727 : 0 : ec_build_derives_hash(PlannerInfo *root, EquivalenceClass *ec)
3728 : : {
3729 : : Assert(!ec->ec_derives_hash);
3730 : :
3731 : : /*
3732 : : * Create the hash table.
3733 : : *
3734 : : * We pass list_length(ec->ec_derives_list) as the initial size.
3735 : : * Simplehash will divide this by the fillfactor (typically 0.9) and round
3736 : : * up to the next power of two, so this will usually give us at least 64
3737 : : * buckets around the threshold. That avoids immediate resizing without
3738 : : * hardcoding a specific size.
3739 : : */
3740 : 0 : ec->ec_derives_hash = derives_create(root->planner_cxt,
3741 : 0 : list_length(ec->ec_derives_list),
3742 : : NULL);
3743 : :
3744 [ # # # # : 0 : foreach_node(RestrictInfo, rinfo, ec->ec_derives_list)
# # ]
3745 : 0 : ec_add_clause_to_derives_hash(ec, rinfo);
3746 : 0 : }
3747 : :
3748 : : /*
3749 : : * ec_add_derived_clause
3750 : : * Add a clause to the set of derived clauses for the given
3751 : : * EquivalenceClass. Always appends to ec_derives_list; also adds
3752 : : * to ec_derives_hash if it exists.
3753 : : *
3754 : : * Also asserts expected invariants of derived clauses.
3755 : : */
3756 : : static void
3757 : 91143 : ec_add_derived_clause(EquivalenceClass *ec, RestrictInfo *clause)
3758 : : {
3759 : : /*
3760 : : * Constant, if present, is always placed on the RHS; see
3761 : : * generate_base_implied_equalities_const(). LHS is never a constant.
3762 : : */
3763 : : Assert(!clause->left_em->em_is_const);
3764 : :
3765 : : /*
3766 : : * Clauses containing a constant are never considered redundant, so
3767 : : * parent_ec is not set.
3768 : : */
3769 : : Assert(!clause->parent_ec || !clause->right_em->em_is_const);
3770 : :
3771 : 91143 : ec->ec_derives_list = lappend(ec->ec_derives_list, clause);
3772 [ - + ]: 91143 : if (ec->ec_derives_hash)
3773 : 0 : ec_add_clause_to_derives_hash(ec, clause);
3774 : 91143 : }
3775 : :
3776 : : /*
3777 : : * ec_add_derived_clauses
3778 : : * Add a list of clauses to the set of clauses derived from the given
3779 : : * EquivalenceClass; adding to the list and hash table if needed.
3780 : : *
3781 : : * This function is similar to ec_add_derived_clause() but optimized for adding
3782 : : * multiple clauses at a time to the ec_derives_list. The assertions from
3783 : : * ec_add_derived_clause() are not repeated here, as the input clauses are
3784 : : * assumed to have already been validated.
3785 : : */
3786 : : static void
3787 : 33 : ec_add_derived_clauses(EquivalenceClass *ec, List *clauses)
3788 : : {
3789 : 33 : ec->ec_derives_list = list_concat(ec->ec_derives_list, clauses);
3790 [ - + ]: 33 : if (ec->ec_derives_hash)
3791 [ # # # # : 0 : foreach_node(RestrictInfo, rinfo, clauses)
# # ]
3792 : 0 : ec_add_clause_to_derives_hash(ec, rinfo);
3793 : 33 : }
3794 : :
3795 : : /*
3796 : : * fill_ec_derives_key
3797 : : * Compute a canonical key for ec_derives_hash lookup or insertion.
3798 : : *
3799 : : * Derived clauses are looked up using a pair of EquivalenceMembers and a
3800 : : * parent EquivalenceClass. To avoid storing or searching for both EM orderings,
3801 : : * we canonicalize the key:
3802 : : *
3803 : : * - For clauses involving two non-constant EMs, em1 is set to the EM with lower
3804 : : * memory address and em2 is set to the other one.
3805 : : * - For clauses involving a constant EM, the caller must pass the non-constant
3806 : : * EM as leftem and NULL as rightem; we then set em1 = NULL and em2 = leftem.
3807 : : */
3808 : : static inline void
3809 : 0 : fill_ec_derives_key(ECDerivesKey *key,
3810 : : EquivalenceMember *leftem,
3811 : : EquivalenceMember *rightem,
3812 : : EquivalenceClass *parent_ec)
3813 : : {
3814 : : Assert(leftem); /* Always required for lookup or insertion */
3815 : :
3816 [ # # ]: 0 : if (rightem == NULL)
3817 : : {
3818 : 0 : key->em1 = NULL;
3819 : 0 : key->em2 = leftem;
3820 : : }
3821 [ # # ]: 0 : else if (leftem < rightem)
3822 : : {
3823 : 0 : key->em1 = leftem;
3824 : 0 : key->em2 = rightem;
3825 : : }
3826 : : else
3827 : : {
3828 : 0 : key->em1 = rightem;
3829 : 0 : key->em2 = leftem;
3830 : : }
3831 : 0 : key->parent_ec = parent_ec;
3832 : 0 : }
3833 : :
3834 : : /*
3835 : : * ec_add_clause_to_derives_hash
3836 : : * Add a derived clause to ec_derives_hash in the given EquivalenceClass.
3837 : : *
3838 : : * Each clause is associated with a canonicalized key. For constant-containing
3839 : : * clauses, only the non-constant EM is used for lookup; see comments in
3840 : : * fill_ec_derives_key().
3841 : : */
3842 : : static void
3843 : 0 : ec_add_clause_to_derives_hash(EquivalenceClass *ec, RestrictInfo *rinfo)
3844 : : {
3845 : : ECDerivesKey key;
3846 : : ECDerivesEntry *entry;
3847 : : bool found;
3848 : :
3849 : : /*
3850 : : * Constants are always placed on the RHS; see
3851 : : * generate_base_implied_equalities_const().
3852 : : */
3853 : : Assert(!rinfo->left_em->em_is_const);
3854 : :
3855 : : /*
3856 : : * Clauses containing a constant are never considered redundant, so
3857 : : * parent_ec is not set.
3858 : : */
3859 : : Assert(!rinfo->parent_ec || !rinfo->right_em->em_is_const);
3860 : :
3861 : : /*
3862 : : * See fill_ec_derives_key() for details: we use a canonicalized key to
3863 : : * avoid storing both EM orderings. For constant EMs, only the
3864 : : * non-constant EM is included in the key.
3865 : : */
3866 : 0 : fill_ec_derives_key(&key,
3867 : : rinfo->left_em,
3868 [ # # ]: 0 : rinfo->right_em->em_is_const ? NULL : rinfo->right_em,
3869 : : rinfo->parent_ec);
3870 : 0 : entry = derives_insert(ec->ec_derives_hash, key, &found);
3871 : : Assert(!found);
3872 : 0 : entry->rinfo = rinfo;
3873 : 0 : }
3874 : :
3875 : : /*
3876 : : * ec_clear_derived_clauses
3877 : : * Reset ec_derives_list and ec_derives_hash.
3878 : : *
3879 : : * We destroy the hash table explicitly, since it may consume significant
3880 : : * space. The list holds the same set of entries and can become equally large
3881 : : * when thousands of partitions are involved, so we free it as well -- even
3882 : : * though we do not typically free lists.
3883 : : */
3884 : : static void
3885 : 33 : ec_clear_derived_clauses(EquivalenceClass *ec)
3886 : : {
3887 : 33 : list_free(ec->ec_derives_list);
3888 : 33 : ec->ec_derives_list = NIL;
3889 : :
3890 [ - + ]: 33 : if (ec->ec_derives_hash)
3891 : : {
3892 : 0 : derives_destroy(ec->ec_derives_hash);
3893 : 0 : ec->ec_derives_hash = NULL;
3894 : : }
3895 : 33 : }
3896 : :
3897 : : /*
3898 : : * ec_search_clause_for_ems
3899 : : * Search for an existing RestrictInfo that equates the given pair
3900 : : * of EquivalenceMembers, either from ec_sources or ec_derives.
3901 : : *
3902 : : * Returns a clause with matching operands in either given order or commuted
3903 : : * order. We used to require matching operator OIDs, but dropped that since any
3904 : : * semantically different operator here would indicate a broken operator family.
3905 : : *
3906 : : * Returns NULL if no matching clause is found.
3907 : : */
3908 : : static RestrictInfo *
3909 : 450323 : ec_search_clause_for_ems(PlannerInfo *root, EquivalenceClass *ec,
3910 : : EquivalenceMember *leftem, EquivalenceMember *rightem,
3911 : : EquivalenceClass *parent_ec)
3912 : : {
3913 : : /* Check original source clauses */
3914 [ + - + + : 1289801 : foreach_node(RestrictInfo, rinfo, ec->ec_sources)
+ + ]
3915 : : {
3916 [ + + ]: 513769 : if (rinfo->left_em == leftem &&
3917 [ + + ]: 258781 : rinfo->right_em == rightem &&
3918 [ + + ]: 235526 : rinfo->parent_ec == parent_ec)
3919 : 62307 : return rinfo;
3920 [ + + ]: 470340 : if (rinfo->left_em == rightem &&
3921 [ + + ]: 210146 : rinfo->right_em == leftem &&
3922 [ + + ]: 191935 : rinfo->parent_ec == parent_ec)
3923 : 18878 : return rinfo;
3924 : : }
3925 : :
3926 : : /* Not found in ec_sources; search derived clauses */
3927 : 388016 : return ec_search_derived_clause_for_ems(root, ec, leftem, rightem,
3928 : : parent_ec);
3929 : : }
3930 : :
3931 : : /*
3932 : : * ec_search_derived_clause_for_ems
3933 : : * Search for an existing derived clause between two EquivalenceMembers.
3934 : : *
3935 : : * If the number of derived clauses exceeds a threshold, switch to hash table
3936 : : * lookup; otherwise, scan ec_derives_list linearly.
3937 : : *
3938 : : * Clauses involving constants are looked up by passing the non-constant EM
3939 : : * as leftem and setting rightem to NULL. In that case, we expect to find a
3940 : : * clause with a constant on the RHS.
3941 : : *
3942 : : * While searching the list, we compare each given EM with both sides of each
3943 : : * clause. But for hash table lookups, we construct a canonicalized key and
3944 : : * perform a single lookup.
3945 : : */
3946 : : static RestrictInfo *
3947 : 388021 : ec_search_derived_clause_for_ems(PlannerInfo *root, EquivalenceClass *ec,
3948 : : EquivalenceMember *leftem,
3949 : : EquivalenceMember *rightem,
3950 : : EquivalenceClass *parent_ec)
3951 : : {
3952 : : /* Switch to using hash lookup when list grows "too long". */
3953 [ + - - + ]: 776042 : if (!ec->ec_derives_hash &&
3954 : 388021 : list_length(ec->ec_derives_list) >= EC_DERIVES_HASH_THRESHOLD)
3955 : 0 : ec_build_derives_hash(root, ec);
3956 : :
3957 : : /* Perform hash table lookup if available */
3958 [ - + ]: 388021 : if (ec->ec_derives_hash)
3959 : : {
3960 : : ECDerivesKey key;
3961 : : RestrictInfo *rinfo;
3962 : : ECDerivesEntry *entry;
3963 : :
3964 : 0 : fill_ec_derives_key(&key, leftem, rightem, parent_ec);
3965 : 0 : entry = derives_lookup(ec->ec_derives_hash, key);
3966 [ # # ]: 0 : if (entry)
3967 : : {
3968 : 0 : rinfo = entry->rinfo;
3969 : : Assert(rinfo);
3970 : : Assert(rightem || rinfo->right_em->em_is_const);
3971 : 0 : return rinfo;
3972 : : }
3973 : : }
3974 : : else
3975 : : {
3976 : : /* Fallback to linear search over ec_derives_list */
3977 [ + + + + : 564639 : foreach_node(RestrictInfo, rinfo, ec->ec_derives_list)
+ + ]
3978 : : {
3979 : : /* Handle special case: lookup by non-const EM alone */
3980 [ + + ]: 427671 : if (!rightem &&
3981 [ + - ]: 5 : rinfo->left_em == leftem)
3982 : : {
3983 : : Assert(rinfo->right_em->em_is_const);
3984 : 319537 : return rinfo;
3985 : : }
3986 [ + + ]: 427666 : if (rinfo->left_em == leftem &&
3987 [ + + ]: 172131 : rinfo->right_em == rightem &&
3988 [ + + ]: 155783 : rinfo->parent_ec == parent_ec)
3989 : 155773 : return rinfo;
3990 [ + + ]: 271893 : if (rinfo->left_em == rightem &&
3991 [ + + ]: 173146 : rinfo->right_em == leftem &&
3992 [ + - ]: 163759 : rinfo->parent_ec == parent_ec)
3993 : 163759 : return rinfo;
3994 : : }
3995 : : }
3996 : :
3997 : 68484 : return NULL;
3998 : : }
|
