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 : 246618 : process_equivalence(PlannerInfo *root,
181 : : RestrictInfo **p_restrictinfo,
182 : : JoinDomain *jdomain)
183 : : {
184 : 246618 : RestrictInfo *restrictinfo = *p_restrictinfo;
185 : 246618 : 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 [ + + + + ]: 246618 : if (restrictinfo->security_level > 0 && !restrictinfo->leakproof)
208 : 172 : return false;
209 : :
210 : : /* Extract info from given clause */
211 : : Assert(is_opclause(clause));
212 : 246446 : opno = ((OpExpr *) clause)->opno;
213 : 246446 : collation = ((OpExpr *) clause)->inputcollid;
214 : 246446 : item1 = (Expr *) get_leftop(clause);
215 : 246446 : item2 = (Expr *) get_rightop(clause);
216 : 246446 : item1_relids = restrictinfo->left_relids;
217 : 246446 : 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 : 246446 : item1 = canonicalize_ec_expression(item1,
224 : : exprType((Node *) item1),
225 : : collation);
226 : 246446 : 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 [ + + ]: 246446 : 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 : 246401 : op_input_types(opno, &item1_type, &item2_type);
282 : :
283 : 246401 : 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 : 246401 : ec1 = ec2 = NULL;
304 : 246401 : em1 = em2 = NULL;
305 : 246401 : ec2_idx = -1;
306 [ + + + + : 425792 : foreach(lc1, root->eq_classes)
+ + ]
307 : : {
308 : 179494 : EquivalenceClass *cur_ec = (EquivalenceClass *) lfirst(lc1);
309 : : ListCell *lc2;
310 : :
311 : : /* Never match to a volatile EC */
312 [ - + ]: 179494 : 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 [ + + ]: 179494 : if (collation != cur_ec->ec_collation)
320 : 18507 : 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 [ + + ]: 160987 : if (!equal(opfamilies, cur_ec->ec_opfamilies))
328 : 42514 : continue;
329 : :
330 : : /* We don't expect any children yet */
331 : : Assert(cur_ec->ec_childmembers == NULL);
332 : :
333 [ + - + + : 353366 : foreach(lc2, cur_ec->ec_members)
+ + ]
334 : : {
335 : 234996 : 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 [ + + + + ]: 234996 : if (cur_em->em_is_const && cur_em->em_jdomain != jdomain)
345 : 3013 : continue;
346 : :
347 [ + + ]: 231983 : if (!ec1 &&
348 [ + + + + ]: 443784 : item1_type == cur_em->em_datatype &&
349 : 221739 : equal(item1, cur_em->em_expr))
350 : : {
351 : 13921 : ec1 = cur_ec;
352 : 13921 : em1 = cur_em;
353 [ + + ]: 13921 : if (ec2)
354 : 60 : break;
355 : : }
356 : :
357 [ + + ]: 231923 : if (!ec2 &&
358 [ + + + + ]: 459413 : item2_type == cur_em->em_datatype &&
359 : 229401 : 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 : 43 : break;
366 : : }
367 : : }
368 : :
369 [ + + + + ]: 118473 : if (ec1 && ec2)
370 : 103 : break;
371 : : }
372 : :
373 : : /* Sweep finished, what did we find? */
374 : :
375 [ + + + + ]: 246401 : if (ec1 && ec2)
376 : : {
377 : : /* If case 1, nothing to do, except add to sources */
378 [ + + ]: 103 : if (ec1 == ec2)
379 : : {
380 : 70 : ec1->ec_sources = lappend(ec1->ec_sources, restrictinfo);
381 : 70 : ec1->ec_min_security = Min(ec1->ec_min_security,
382 : : restrictinfo->security_level);
383 : 70 : ec1->ec_max_security = Max(ec1->ec_max_security,
384 : : restrictinfo->security_level);
385 : : /* mark the RI as associated with this eclass */
386 : 70 : restrictinfo->left_ec = ec1;
387 : 70 : restrictinfo->right_ec = ec1;
388 : : /* mark the RI as usable with this pair of EMs */
389 : 70 : restrictinfo->left_em = em1;
390 : 70 : restrictinfo->right_em = em2;
391 : 70 : 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 [ + + ]: 246298 : else if (ec1)
445 : : {
446 : : /* Case 3: add item2 to ec1 */
447 : 13818 : em2 = add_eq_member(ec1, item2, item2_relids,
448 : : jdomain, item2_type);
449 : 13818 : ec1->ec_sources = lappend(ec1->ec_sources, restrictinfo);
450 : 13818 : ec1->ec_min_security = Min(ec1->ec_min_security,
451 : : restrictinfo->security_level);
452 : 13818 : ec1->ec_max_security = Max(ec1->ec_max_security,
453 : : restrictinfo->security_level);
454 : : /* mark the RI as associated with this eclass */
455 : 13818 : restrictinfo->left_ec = ec1;
456 : 13818 : restrictinfo->right_ec = ec1;
457 : : /* mark the RI as usable with this pair of EMs */
458 : 13818 : restrictinfo->left_em = em1;
459 : 13818 : restrictinfo->right_em = em2;
460 : : }
461 [ + + ]: 232480 : else if (ec2)
462 : : {
463 : : /* Case 3: add item1 to ec2 */
464 : 2876 : em1 = add_eq_member(ec2, item1, item1_relids,
465 : : jdomain, item1_type);
466 : 2876 : ec2->ec_sources = lappend(ec2->ec_sources, restrictinfo);
467 : 2876 : ec2->ec_min_security = Min(ec2->ec_min_security,
468 : : restrictinfo->security_level);
469 : 2876 : ec2->ec_max_security = Max(ec2->ec_max_security,
470 : : restrictinfo->security_level);
471 : : /* mark the RI as associated with this eclass */
472 : 2876 : restrictinfo->left_ec = ec2;
473 : 2876 : restrictinfo->right_ec = ec2;
474 : : /* mark the RI as usable with this pair of EMs */
475 : 2876 : restrictinfo->left_em = em1;
476 : 2876 : restrictinfo->right_em = em2;
477 : : }
478 : : else
479 : : {
480 : : /* Case 4: make a new, two-entry EC */
481 : 229604 : EquivalenceClass *ec = makeNode(EquivalenceClass);
482 : :
483 : 229604 : ec->ec_opfamilies = opfamilies;
484 : 229604 : ec->ec_collation = collation;
485 : 229604 : ec->ec_childmembers_size = 0;
486 : 229604 : ec->ec_members = NIL;
487 : 229604 : ec->ec_childmembers = NULL;
488 : 229604 : ec->ec_sources = list_make1(restrictinfo);
489 : 229604 : ec->ec_derives_list = NIL;
490 : 229604 : ec->ec_derives_hash = NULL;
491 : 229604 : ec->ec_relids = NULL;
492 : 229604 : ec->ec_has_const = false;
493 : 229604 : ec->ec_has_volatile = false;
494 : 229604 : ec->ec_broken = false;
495 : 229604 : ec->ec_sortref = 0;
496 : 229604 : ec->ec_min_security = restrictinfo->security_level;
497 : 229604 : ec->ec_max_security = restrictinfo->security_level;
498 : 229604 : ec->ec_merged = NULL;
499 : 229604 : em1 = add_eq_member(ec, item1, item1_relids,
500 : : jdomain, item1_type);
501 : 229604 : em2 = add_eq_member(ec, item2, item2_relids,
502 : : jdomain, item2_type);
503 : :
504 : 229604 : root->eq_classes = lappend(root->eq_classes, ec);
505 : :
506 : : /* mark the RI as associated with this eclass */
507 : 229604 : restrictinfo->left_ec = ec;
508 : 229604 : restrictinfo->right_ec = ec;
509 : : /* mark the RI as usable with this pair of EMs */
510 : 229604 : restrictinfo->left_em = em1;
511 : 229604 : restrictinfo->right_em = em2;
512 : : }
513 : :
514 : 246331 : 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 : 2378552 : canonicalize_ec_expression(Expr *expr, Oid req_type, Oid req_collation)
547 : : {
548 : 2378552 : 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 [ + - + + : 2378552 : if (IsPolymorphicType(req_type) || req_type == RECORDOID)
+ - + + +
+ + + + -
+ - + - +
- + - +
+ ]
555 : 9440 : req_type = expr_type;
556 : :
557 : : /*
558 : : * No work if the expression exposes the right type/collation already.
559 : : */
560 [ + + + + ]: 4694393 : if (expr_type != req_type ||
561 : 2315841 : 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 [ + + ]: 64523 : if (expr_type != req_type)
571 : 62711 : req_typmod = -1;
572 : : else
573 : 1812 : 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 : 64523 : expr = (Expr *) applyRelabelType((Node *) expr,
580 : : req_type, req_typmod, req_collation,
581 : : COERCE_IMPLICIT_CAST, -1, false);
582 : : }
583 : :
584 : 2378552 : 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 : 783832 : make_eq_member(EquivalenceClass *ec, Expr *expr, Relids relids,
594 : : JoinDomain *jdomain, EquivalenceMember *parent, Oid datatype)
595 : : {
596 : 783832 : EquivalenceMember *em = makeNode(EquivalenceMember);
597 : :
598 : 783832 : em->em_expr = expr;
599 : 783832 : em->em_relids = relids;
600 : 783832 : em->em_is_const = false;
601 : 783832 : em->em_is_child = (parent != NULL);
602 : 783832 : em->em_datatype = datatype;
603 : 783832 : em->em_jdomain = jdomain;
604 : 783832 : em->em_parent = parent;
605 : :
606 [ + + ]: 783832 : 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 : 174135 : em->em_is_const = true;
618 : 174135 : ec->ec_has_const = true;
619 : : /* it can't affect ec_relids */
620 : : }
621 : :
622 : 783832 : 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 : 699658 : add_eq_member(EquivalenceClass *ec, Expr *expr, Relids relids,
630 : : JoinDomain *jdomain, Oid datatype)
631 : : {
632 : 699658 : EquivalenceMember *em = make_eq_member(ec, expr, relids, jdomain,
633 : : NULL, datatype);
634 : :
635 : : /* add to the members list */
636 : 699658 : ec->ec_members = lappend(ec->ec_members, em);
637 : :
638 : : /* record the relids for parent members */
639 : 699658 : ec->ec_relids = bms_add_members(ec->ec_relids, relids);
640 : :
641 : 699658 : 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 : 84174 : 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 [ + + ]: 84174 : if (unlikely(ec->ec_childmembers_size < root->simple_rel_array_size))
676 : : {
677 [ + - ]: 23784 : if (ec->ec_childmembers == NULL)
678 : 23784 : 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 : 23784 : ec->ec_childmembers_size = root->simple_rel_array_size;
685 : : }
686 : :
687 : 84174 : 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 : 84174 : ec->ec_childmembers[child_relid] = lappend(ec->ec_childmembers[child_relid], em);
691 : :
692 : : /* Record this EC index for the child rel */
693 [ + + ]: 84174 : if (ec_index >= 0)
694 : : {
695 : 51253 : RelOptInfo *child_rel = root->simple_rel_array[child_relid];
696 : :
697 : 51253 : child_rel->eclass_indexes =
698 : 51253 : bms_add_member(child_rel->eclass_indexes, ec_index);
699 : : }
700 : :
701 : 84174 : 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 : 1757294 : get_eclass_for_sort_expr(PlannerInfo *root,
738 : : Expr *expr,
739 : : List *opfamilies,
740 : : Oid opcintype,
741 : : Oid collation,
742 : : Index sortref,
743 : : Relids rel,
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 : 1757294 : 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 : 1757294 : jdomain = linitial_node(JoinDomain, root->join_domains);
763 : :
764 : : /*
765 : : * Scan through the existing EquivalenceClasses for a match
766 : : */
767 [ + + + + : 5742619 : foreach(lc1, root->eq_classes)
+ + ]
768 : : {
769 : 5004488 : 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 [ + + + + ]: 5004488 : if (cur_ec->ec_has_volatile &&
778 [ + + ]: 28 : (sortref == 0 || sortref != cur_ec->ec_sortref))
779 : 2116051 : continue;
780 : :
781 [ + + ]: 5003978 : if (collation != cur_ec->ec_collation)
782 : 1400585 : continue;
783 [ + + ]: 3603393 : if (!equal(opfamilies, cur_ec->ec_opfamilies))
784 : 714956 : continue;
785 : :
786 : 2888437 : setup_eclass_member_iterator(&it, cur_ec, rel);
787 [ + + ]: 6418628 : while ((cur_em = eclass_member_iterator_next(&it)) != NULL)
788 : : {
789 : : /*
790 : : * Ignore child members unless they match the request.
791 : : */
792 [ + + ]: 4549354 : if (cur_em->em_is_child &&
793 [ - + ]: 84926 : !bms_equal(cur_em->em_relids, rel))
794 : 0 : continue;
795 : :
796 : : /*
797 : : * Match constants only within the same JoinDomain (see
798 : : * optimizer/README).
799 : : */
800 [ + + + + ]: 4549354 : if (cur_em->em_is_const && cur_em->em_jdomain != jdomain)
801 : 57322 : continue;
802 : :
803 [ + + + + ]: 8950599 : if (opcintype == cur_em->em_datatype &&
804 : 4458567 : equal(expr, cur_em->em_expr))
805 : 1019163 : return cur_ec; /* Match! */
806 : : }
807 : : }
808 : :
809 : : /* No match; does caller want a NULL result? */
810 [ + + ]: 738131 : if (!create_it)
811 : 514375 : 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 : 223756 : oldcontext = MemoryContextSwitchTo(root->planner_cxt);
819 : :
820 : 223756 : newec = makeNode(EquivalenceClass);
821 : 223756 : newec->ec_opfamilies = list_copy(opfamilies);
822 : 223756 : newec->ec_collation = collation;
823 : 223756 : newec->ec_childmembers_size = 0;
824 : 223756 : newec->ec_members = NIL;
825 : 223756 : newec->ec_childmembers = NULL;
826 : 223756 : newec->ec_sources = NIL;
827 : 223756 : newec->ec_derives_list = NIL;
828 : 223756 : newec->ec_derives_hash = NULL;
829 : 223756 : newec->ec_relids = NULL;
830 : 223756 : newec->ec_has_const = false;
831 : 223756 : newec->ec_has_volatile = contain_volatile_functions((Node *) expr);
832 : 223756 : newec->ec_broken = false;
833 : 223756 : newec->ec_sortref = sortref;
834 : 223756 : newec->ec_min_security = UINT_MAX;
835 : 223756 : newec->ec_max_security = 0;
836 : 223756 : newec->ec_merged = NULL;
837 : :
838 [ + + - + ]: 223756 : 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 : 223756 : expr_relids = pull_varnos(root, (Node *) expr);
845 : :
846 : 223756 : 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 [ + + ]: 223756 : if (newec->ec_has_const)
856 : : {
857 [ + + + + ]: 15370 : if (newec->ec_has_volatile ||
858 [ + + ]: 15142 : expression_returns_set((Node *) expr) ||
859 [ + + ]: 14871 : contain_agg_clause((Node *) expr) ||
860 : 7360 : contain_window_function((Node *) expr))
861 : : {
862 : 384 : newec->ec_has_const = false;
863 : 384 : newem->em_is_const = false;
864 : : }
865 : : }
866 : :
867 : 223756 : 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 [ + + ]: 223756 : if (root->ec_merging_done)
874 : : {
875 : 130415 : int ec_index = list_length(root->eq_classes) - 1;
876 : 130415 : int i = -1;
877 : :
878 [ + + ]: 251722 : while ((i = bms_next_member(newec->ec_relids, i)) > 0)
879 : : {
880 : 121307 : RelOptInfo *rel = root->simple_rel_array[i];
881 : :
882 : : /* ignore the RTE_GROUP RTE */
883 [ + + ]: 121307 : if (i == root->group_rtindex)
884 : 584 : continue;
885 : :
886 [ + + ]: 120723 : if (rel == NULL) /* must be an outer join */
887 : : {
888 : : Assert(bms_is_member(i, root->outer_join_rels));
889 : 5446 : continue;
890 : : }
891 : :
892 : : Assert(rel->reloptkind == RELOPT_BASEREL);
893 : :
894 : 115277 : rel->eclass_indexes = bms_add_member(rel->eclass_indexes,
895 : : ec_index);
896 : : }
897 : : }
898 : :
899 : 223756 : MemoryContextSwitchTo(oldcontext);
900 : :
901 : 223756 : 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 : 275612 : 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 [ + - + + ]: 298035 : while (expr && IsA(expr, RelabelType))
926 : 22423 : expr = ((RelabelType *) expr)->arg;
927 : :
928 : 275612 : setup_eclass_member_iterator(&it, ec, relids);
929 [ + + ]: 464101 : 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 [ - + ]: 307429 : if (em->em_is_const)
938 : 0 : continue;
939 : :
940 : : /*
941 : : * Ignore child members unless they belong to the requested rel.
942 : : */
943 [ + + ]: 307429 : if (em->em_is_child &&
944 [ + + ]: 10050 : !bms_is_subset(em->em_relids, relids))
945 : 3396 : continue;
946 : :
947 : : /*
948 : : * Match if same expression (after stripping relabel).
949 : : */
950 : 304033 : emexpr = em->em_expr;
951 [ + - + + ]: 310853 : while (emexpr && IsA(emexpr, RelabelType))
952 : 6820 : emexpr = ((RelabelType *) emexpr)->arg;
953 : :
954 [ + + ]: 304033 : if (equal(emexpr, expr))
955 : 118940 : return em;
956 : : }
957 : :
958 : 156672 : 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 : 7333 : 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 : 7333 : exprvars = pull_var_clause((Node *) exprs,
1009 : : PVC_INCLUDE_AGGREGATES |
1010 : : PVC_INCLUDE_WINDOWFUNCS |
1011 : : PVC_INCLUDE_PLACEHOLDERS);
1012 : :
1013 : 7333 : setup_eclass_member_iterator(&it, ec, relids);
1014 [ + + ]: 14663 : 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 [ - + ]: 7801 : if (em->em_is_const)
1024 : 0 : continue;
1025 : :
1026 : : /*
1027 : : * Ignore child members unless they belong to the requested rel.
1028 : : */
1029 [ + + ]: 7801 : 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 : 7691 : emvars = pull_var_clause((Node *) em->em_expr,
1037 : : PVC_INCLUDE_AGGREGATES |
1038 : : PVC_INCLUDE_WINDOWFUNCS |
1039 : : PVC_INCLUDE_PLACEHOLDERS);
1040 [ + + + + : 8618 : foreach(lc2, emvars)
+ + ]
1041 : : {
1042 [ + + ]: 8027 : if (!list_member(exprvars, lfirst(lc2)))
1043 : 7100 : break;
1044 : : }
1045 : 7691 : list_free(emvars);
1046 [ + + ]: 7691 : if (lc2)
1047 : 7100 : 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 [ + + ]: 591 : if (require_parallel_safe &&
1054 [ + + ]: 281 : !is_parallel_safe(root, (Node *) em->em_expr))
1055 : 120 : continue;
1056 : :
1057 : 471 : return em; /* found usable expression */
1058 : : }
1059 : :
1060 : 6862 : 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 : 16620 : relation_can_be_sorted_early(PlannerInfo *root, RelOptInfo *rel,
1078 : : EquivalenceClass *ec, bool require_parallel_safe)
1079 : : {
1080 : 16620 : 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 [ + + ]: 16620 : if (ec->ec_has_volatile)
1088 : 60 : return false;
1089 : :
1090 : : /*
1091 : : * Try to find an EM directly matching some reltarget member.
1092 : : */
1093 [ + + + + : 38366 : foreach(lc, target->exprs)
+ + ]
1094 : : {
1095 : 31343 : Expr *targetexpr = (Expr *) lfirst(lc);
1096 : :
1097 : 31343 : em = find_ec_member_matching_expr(ec, targetexpr, rel->relids);
1098 [ + + ]: 31343 : if (!em)
1099 : 21806 : 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 [ - + ]: 9537 : 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 [ + - ]: 9537 : if (require_parallel_safe &&
1115 [ - + ]: 9537 : !is_parallel_safe(root, (Node *) em->em_expr))
1116 : 0 : continue;
1117 : :
1118 : 9537 : return true;
1119 : : }
1120 : :
1121 : : /*
1122 : : * Try to find an expression computable from the reltarget.
1123 : : */
1124 : 7023 : em = find_computable_ec_member(root, ec, target->exprs, rel->relids,
1125 : : require_parallel_safe);
1126 [ + + ]: 7023 : if (!em)
1127 : 6862 : 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 [ + + ]: 161 : if (expression_returns_set((Node *) em->em_expr))
1136 : 10 : return false;
1137 : :
1138 : 151 : 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 : 257712 : 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 : 257712 : root->ec_merging_done = true;
1201 : :
1202 : 257712 : ec_index = 0;
1203 [ + + + + : 580624 : foreach(lc, root->eq_classes)
+ + ]
1204 : : {
1205 : 322912 : EquivalenceClass *ec = (EquivalenceClass *) lfirst(lc);
1206 : 322912 : 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 [ + + ]: 322912 : if (list_length(ec->ec_members) > 1)
1218 : : {
1219 [ + + ]: 231243 : if (ec->ec_has_const)
1220 : 166256 : generate_base_implied_equalities_const(root, ec);
1221 : : else
1222 : 64987 : generate_base_implied_equalities_no_const(root, ec);
1223 : :
1224 : : /* Recover if we failed to generate required derived clauses */
1225 [ + + ]: 231243 : if (ec->ec_broken)
1226 : 25 : generate_base_implied_equalities_broken(root, ec);
1227 : :
1228 : : /* Detect whether this EC might generate join clauses */
1229 : 231243 : can_generate_joinclause =
1230 : 231243 : (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 : 322912 : i = -1;
1241 [ + + ]: 726990 : while ((i = bms_next_member(ec->ec_relids, i)) > 0)
1242 : : {
1243 : 404078 : RelOptInfo *rel = root->simple_rel_array[i];
1244 : :
1245 : : /* ignore the RTE_GROUP RTE */
1246 [ - + ]: 404078 : if (i == root->group_rtindex)
1247 : 0 : continue;
1248 : :
1249 [ + + ]: 404078 : if (rel == NULL) /* must be an outer join */
1250 : : {
1251 : : Assert(bms_is_member(i, root->outer_join_rels));
1252 : 3349 : continue;
1253 : : }
1254 : :
1255 : : Assert(rel->reloptkind == RELOPT_BASEREL);
1256 : :
1257 : 400729 : rel->eclass_indexes = bms_add_member(rel->eclass_indexes,
1258 : : ec_index);
1259 : :
1260 [ + + ]: 400729 : if (can_generate_joinclause)
1261 : 154422 : rel->has_eclass_joins = true;
1262 : : }
1263 : :
1264 : 322912 : ec_index++;
1265 : : }
1266 : 257712 : }
1267 : :
1268 : : /*
1269 : : * generate_base_implied_equalities when EC contains pseudoconstant(s)
1270 : : */
1271 : : static void
1272 : 166256 : generate_base_implied_equalities_const(PlannerInfo *root,
1273 : : EquivalenceClass *ec)
1274 : : {
1275 : 166256 : 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 [ + + + + ]: 320026 : if (list_length(ec->ec_members) == 2 &&
1286 : 153770 : list_length(ec->ec_sources) == 1)
1287 : : {
1288 : 153735 : RestrictInfo *restrictinfo = (RestrictInfo *) linitial(ec->ec_sources);
1289 : :
1290 : 153735 : distribute_restrictinfo_to_rels(root, restrictinfo);
1291 : 153735 : 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 [ + - + + : 30068 : foreach(lc, ec->ec_members)
+ + ]
1304 : : {
1305 : 29974 : EquivalenceMember *cur_em = (EquivalenceMember *) lfirst(lc);
1306 : :
1307 [ + + ]: 29974 : if (cur_em->em_is_const)
1308 : : {
1309 : 12526 : const_em = cur_em;
1310 [ + + ]: 12526 : if (IsA(cur_em->em_expr, Const))
1311 : 12427 : break;
1312 : : }
1313 : : }
1314 : : Assert(const_em != NULL);
1315 : :
1316 : : /* Generate a derived equality against each other member */
1317 [ + - + + : 50165 : foreach(lc, ec->ec_members)
+ + ]
1318 : : {
1319 : 37669 : 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 [ + + ]: 37669 : if (cur_em == const_em)
1326 : 12501 : continue;
1327 : 25168 : eq_op = select_equality_operator(ec,
1328 : : cur_em->em_datatype,
1329 : : const_em->em_datatype);
1330 [ + + ]: 25168 : 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 : 25143 : rinfo = process_implied_equality(root, eq_op, ec->ec_collation,
1343 : : cur_em->em_expr, const_em->em_expr,
1344 : 25143 : const_em->em_jdomain->jd_relids,
1345 : : ec->ec_min_security,
1346 : 25143 : 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 [ + - + + ]: 25143 : if (rinfo && rinfo->mergeopfamilies)
1357 : : {
1358 : : /* it's not redundant, so don't set parent_ec */
1359 : 25008 : rinfo->left_ec = rinfo->right_ec = ec;
1360 : 25008 : rinfo->left_em = cur_em;
1361 : 25008 : rinfo->right_em = const_em;
1362 : 25008 : 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 : 64987 : 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 : 64987 : 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 [ + - + + : 197387 : foreach(lc, ec->ec_members)
+ + ]
1391 : : {
1392 : 132400 : 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 [ + + ]: 132400 : if (!bms_get_singleton_member(cur_em->em_relids, &relid))
1399 : 185 : continue;
1400 : : Assert(relid < root->simple_rel_array_size);
1401 : :
1402 [ + + ]: 132215 : if (prev_ems[relid] != NULL)
1403 : : {
1404 : 429 : EquivalenceMember *prev_em = prev_ems[relid];
1405 : : Oid eq_op;
1406 : : RestrictInfo *rinfo;
1407 : :
1408 : 429 : eq_op = select_equality_operator(ec,
1409 : : prev_em->em_datatype,
1410 : : cur_em->em_datatype);
1411 [ - + ]: 429 : 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 : 429 : 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 [ + - + - ]: 429 : if (rinfo && rinfo->mergeopfamilies)
1437 : : {
1438 : : /* it's not redundant, so don't set parent_ec */
1439 : 429 : rinfo->left_ec = rinfo->right_ec = ec;
1440 : 429 : rinfo->left_em = prev_em;
1441 : 429 : rinfo->right_em = cur_em;
1442 : : }
1443 : : }
1444 : 132215 : prev_ems[relid] = cur_em;
1445 : : }
1446 : :
1447 : 64987 : 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 [ + - + + : 197387 : foreach(lc, ec->ec_members)
+ + ]
1458 : : {
1459 : 132400 : EquivalenceMember *cur_em = (EquivalenceMember *) lfirst(lc);
1460 : 132400 : List *vars = pull_var_clause((Node *) cur_em->em_expr,
1461 : : PVC_RECURSE_AGGREGATES |
1462 : : PVC_RECURSE_WINDOWFUNCS |
1463 : : PVC_INCLUDE_PLACEHOLDERS);
1464 : :
1465 : 132400 : add_vars_to_targetlist(root, vars, ec->ec_relids);
1466 : 132400 : list_free(vars);
1467 : : }
1468 : 64987 : }
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 : 451646 : generate_join_implied_equalities(PlannerInfo *root,
1549 : : Relids join_relids,
1550 : : Relids outer_relids,
1551 : : RelOptInfo *inner_rel,
1552 : : SpecialJoinInfo *sjinfo)
1553 : : {
1554 : 451646 : List *result = NIL;
1555 : 451646 : 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 [ + + + + : 451646 : 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 : 6332 : nominal_inner_relids = inner_rel->top_parent_relids;
1568 : : /* ECs will be marked with the parent's relid, not the child's */
1569 : 6332 : nominal_join_relids = bms_union(outer_relids, nominal_inner_relids);
1570 : 6332 : nominal_join_relids = add_outer_joins_to_relids(root,
1571 : : nominal_join_relids,
1572 : : sjinfo,
1573 : : NULL);
1574 : : }
1575 : : else
1576 : : {
1577 : 445314 : nominal_inner_relids = inner_relids;
1578 : 445314 : 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 [ + + + + ]: 451646 : if (sjinfo && sjinfo->ojrelid != 0)
1595 : 63005 : matching_ecs = get_eclass_indexes_for_relids(root, nominal_join_relids);
1596 : : else
1597 : 388641 : matching_ecs = get_common_eclass_indexes(root, nominal_inner_relids,
1598 : : outer_relids);
1599 : :
1600 : 451646 : i = -1;
1601 [ + + ]: 1198600 : while ((i = bms_next_member(matching_ecs, i)) >= 0)
1602 : : {
1603 : 746954 : EquivalenceClass *ec = (EquivalenceClass *) list_nth(root->eq_classes, i);
1604 : 746954 : List *sublist = NIL;
1605 : :
1606 : : /* ECs containing consts do not need any further enforcement */
1607 [ + + ]: 746954 : if (ec->ec_has_const)
1608 : 88119 : continue;
1609 : :
1610 : : /* Single-member ECs won't generate any deductions */
1611 [ + + ]: 658835 : if (list_length(ec->ec_members) <= 1)
1612 : 301990 : continue;
1613 : :
1614 : : /* Sanity check that this eclass overlaps the join */
1615 : : Assert(bms_overlap(ec->ec_relids, nominal_join_relids));
1616 : :
1617 [ + + ]: 356845 : if (!ec->ec_broken)
1618 : 356575 : 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 [ + + ]: 356845 : 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 : 356845 : result = list_concat(result, sublist);
1634 : : }
1635 : :
1636 : 451646 : 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 : 5909 : 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 : 5909 : List *result = NIL;
1655 : 5909 : 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 [ + + + - : 5909 : 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 : 20 : nominal_inner_relids = inner_rel->top_parent_relids;
1667 : : /* ECs will be marked with the parent's relid, not the child's */
1668 : 20 : nominal_join_relids = bms_union(outer_relids, nominal_inner_relids);
1669 : : }
1670 : : else
1671 : : {
1672 : 5889 : nominal_inner_relids = inner_relids;
1673 : 5889 : nominal_join_relids = join_relids;
1674 : : }
1675 : :
1676 [ + - + + : 12231 : foreach(lc, eclasses)
+ + ]
1677 : : {
1678 : 6322 : EquivalenceClass *ec = (EquivalenceClass *) lfirst(lc);
1679 : 6322 : List *sublist = NIL;
1680 : :
1681 : : /* ECs containing consts do not need any further enforcement */
1682 [ - + ]: 6322 : if (ec->ec_has_const)
1683 : 0 : continue;
1684 : :
1685 : : /* Single-member ECs won't generate any deductions */
1686 [ - + ]: 6322 : 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 [ - + ]: 6322 : if (!bms_overlap(ec->ec_relids, nominal_join_relids))
1691 : 0 : continue;
1692 : :
1693 [ + - ]: 6322 : if (!ec->ec_broken)
1694 : 6322 : 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 [ - + ]: 6322 : 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 : 6322 : result = list_concat(result, sublist);
1710 : : }
1711 : :
1712 : 5909 : return result;
1713 : : }
1714 : :
1715 : : /*
1716 : : * generate_join_implied_equalities for a still-valid EC
1717 : : */
1718 : : static List *
1719 : 362897 : generate_join_implied_equalities_normal(PlannerInfo *root,
1720 : : EquivalenceClass *ec,
1721 : : Relids join_relids,
1722 : : Relids outer_relids,
1723 : : Relids inner_relids)
1724 : : {
1725 : 362897 : List *result = NIL;
1726 : 362897 : List *new_members = NIL;
1727 : 362897 : List *outer_members = NIL;
1728 : 362897 : 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 : 362897 : setup_eclass_member_iterator(&it, ec, join_relids);
1742 [ + + ]: 1160942 : 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 [ + + ]: 798045 : if (!bms_is_subset(cur_em->em_relids, join_relids))
1750 : 66896 : continue; /* not computable yet, or wrong child */
1751 : :
1752 [ + + ]: 731149 : if (bms_is_subset(cur_em->em_relids, outer_relids))
1753 : 408897 : outer_members = lappend(outer_members, cur_em);
1754 [ + + ]: 322252 : else if (bms_is_subset(cur_em->em_relids, inner_relids))
1755 : 320422 : inner_members = lappend(inner_members, cur_em);
1756 : : else
1757 : 1830 : 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 [ + + + + ]: 362897 : if (outer_members && inner_members)
1770 : : {
1771 : 308298 : EquivalenceMember *best_outer_em = NULL;
1772 : 308298 : EquivalenceMember *best_inner_em = NULL;
1773 : 308298 : Oid best_eq_op = InvalidOid;
1774 : 308298 : int best_score = -1;
1775 : : RestrictInfo *rinfo;
1776 : : ListCell *lc1;
1777 : :
1778 [ + - + + : 321144 : foreach(lc1, outer_members)
+ + ]
1779 : : {
1780 : 308363 : EquivalenceMember *outer_em = (EquivalenceMember *) lfirst(lc1);
1781 : : ListCell *lc2;
1782 : :
1783 [ + - + + : 321229 : foreach(lc2, inner_members)
+ + ]
1784 : : {
1785 : 308383 : EquivalenceMember *inner_em = (EquivalenceMember *) lfirst(lc2);
1786 : : Oid eq_op;
1787 : : int score;
1788 : :
1789 : 308383 : eq_op = select_equality_operator(ec,
1790 : : outer_em->em_datatype,
1791 : : inner_em->em_datatype);
1792 [ + + ]: 308383 : if (!OidIsValid(eq_op))
1793 : 30 : continue;
1794 : 308353 : score = 0;
1795 [ + + ]: 308353 : if (IsA(outer_em->em_expr, Var) ||
1796 [ + + ]: 17119 : (IsA(outer_em->em_expr, RelabelType) &&
1797 [ + + ]: 6316 : IsA(((RelabelType *) outer_em->em_expr)->arg, Var)))
1798 : 297429 : score++;
1799 [ + + ]: 308353 : if (IsA(inner_em->em_expr, Var) ||
1800 [ + + ]: 12405 : (IsA(inner_em->em_expr, RelabelType) &&
1801 [ + + ]: 10226 : IsA(((RelabelType *) inner_em->em_expr)->arg, Var)))
1802 : 306098 : score++;
1803 [ + + ]: 308353 : if (op_hashjoinable(eq_op,
1804 : 308353 : exprType((Node *) outer_em->em_expr)))
1805 : 308294 : score++;
1806 [ + + ]: 308353 : if (score > best_score)
1807 : : {
1808 : 308268 : best_outer_em = outer_em;
1809 : 308268 : best_inner_em = inner_em;
1810 : 308268 : best_eq_op = eq_op;
1811 : 308268 : best_score = score;
1812 [ + + ]: 308268 : if (best_score == 3)
1813 : 295517 : break; /* no need to look further */
1814 : : }
1815 : : }
1816 [ + + ]: 308363 : if (best_score == 3)
1817 : 295517 : break; /* no need to look further */
1818 : : }
1819 [ + + ]: 308298 : 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 : 308268 : rinfo = create_join_clause(root, ec, best_eq_op,
1831 : : best_outer_em, best_inner_em,
1832 : : ec);
1833 : :
1834 : 308268 : 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 [ + + ]: 362867 : if (new_members)
1847 : : {
1848 : 1800 : List *old_members = list_concat(outer_members, inner_members);
1849 : 1800 : 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 [ + + ]: 1800 : if (old_members)
1855 : 1515 : new_members = lappend(new_members, linitial(old_members));
1856 : :
1857 [ + - + + : 5145 : foreach(lc1, new_members)
+ + ]
1858 : : {
1859 : 3345 : cur_em = (EquivalenceMember *) lfirst(lc1);
1860 : :
1861 [ + + ]: 3345 : if (prev_em != NULL)
1862 : : {
1863 : : Oid eq_op;
1864 : :
1865 : 1545 : eq_op = select_equality_operator(ec,
1866 : : prev_em->em_datatype,
1867 : : cur_em->em_datatype);
1868 [ - + ]: 1545 : 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 : 1545 : rinfo = create_join_clause(root, ec, eq_op,
1876 : : prev_em, cur_em,
1877 : : NULL);
1878 : :
1879 : 1545 : result = lappend(result, rinfo);
1880 : : }
1881 : 3345 : prev_em = cur_em;
1882 : : }
1883 : : }
1884 : :
1885 : 362867 : 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 : 439009 : select_equality_operator(EquivalenceClass *ec, Oid lefttype, Oid righttype)
1947 : : {
1948 : : ListCell *lc;
1949 : :
1950 [ + - + + : 439064 : foreach(lc, ec->ec_opfamilies)
+ + ]
1951 : : {
1952 : 439009 : Oid opfamily = lfirst_oid(lc);
1953 : : Oid opno;
1954 : :
1955 : 439009 : opno = get_opfamily_member_for_cmptype(opfamily, lefttype, righttype, COMPARE_EQ);
1956 [ + + ]: 439009 : if (!OidIsValid(opno))
1957 : 55 : continue;
1958 : : /* If no barrier quals in query, don't worry about leaky operators */
1959 [ + + ]: 438954 : if (ec->ec_max_security == 0)
1960 : 438954 : return opno;
1961 : : /* Otherwise, insist that selected operators be leakproof */
1962 [ + - ]: 839 : if (get_func_leakproof(get_opcode(opno)))
1963 : 839 : 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 : 415274 : 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 : 415274 : RestrictInfo *parent_rinfo = NULL;
1989 : : MemoryContext oldcontext;
1990 : :
1991 : 415274 : rinfo = ec_search_clause_for_ems(root, ec, leftem, rightem, parent_ec);
1992 [ + + ]: 415274 : if (rinfo)
1993 : 345704 : 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 : 69570 : 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 [ + + + + ]: 69570 : if (leftem->em_is_child || rightem->em_is_child)
2006 : : {
2007 [ + + ]: 3679 : EquivalenceMember *leftp = leftem->em_parent ? leftem->em_parent : leftem;
2008 [ + + ]: 3679 : EquivalenceMember *rightp = rightem->em_parent ? rightem->em_parent : rightem;
2009 : :
2010 : 3679 : parent_rinfo = create_join_clause(root, ec, opno,
2011 : : leftp, rightp,
2012 : : parent_ec);
2013 : : }
2014 : :
2015 : 69570 : rinfo = build_implied_join_equality(root,
2016 : : opno,
2017 : : ec->ec_collation,
2018 : : leftem->em_expr,
2019 : : rightem->em_expr,
2020 : 69570 : bms_union(leftem->em_relids,
2021 : 69570 : rightem->em_relids),
2022 : : ec->ec_min_security);
2023 : :
2024 : : /*
2025 : : * If either EM is a child, force the clause's clause_relids to include
2026 : : * the relid(s) of the child rel. In normal cases it would already, but
2027 : : * not if we are considering appendrel child relations with pseudoconstant
2028 : : * translated variables (i.e., UNION ALL sub-selects with constant output
2029 : : * items). We must do this so that join_clause_is_movable_into() will
2030 : : * think that the clause should be evaluated at the correct place.
2031 : : */
2032 [ + + ]: 69570 : if (leftem->em_is_child)
2033 : 3208 : rinfo->clause_relids = bms_add_members(rinfo->clause_relids,
2034 : 3208 : leftem->em_relids);
2035 [ + + ]: 69570 : if (rightem->em_is_child)
2036 : 471 : rinfo->clause_relids = bms_add_members(rinfo->clause_relids,
2037 : 471 : rightem->em_relids);
2038 : :
2039 : : /* If it's a child clause, copy the parent's rinfo_serial */
2040 [ + + ]: 69570 : if (parent_rinfo)
2041 : 3679 : rinfo->rinfo_serial = parent_rinfo->rinfo_serial;
2042 : :
2043 : : /* Mark the clause as redundant, or not */
2044 : 69570 : rinfo->parent_ec = parent_ec;
2045 : :
2046 : : /*
2047 : : * We know the correct values for left_ec/right_ec, ie this particular EC,
2048 : : * so we can just set them directly instead of forcing another lookup.
2049 : : */
2050 : 69570 : rinfo->left_ec = ec;
2051 : 69570 : rinfo->right_ec = ec;
2052 : :
2053 : : /* Mark it as usable with these EMs */
2054 : 69570 : rinfo->left_em = leftem;
2055 : 69570 : rinfo->right_em = rightem;
2056 : : /* and save it for possible re-use */
2057 : 69570 : ec_add_derived_clause(ec, rinfo);
2058 : :
2059 : 69570 : MemoryContextSwitchTo(oldcontext);
2060 : :
2061 : 69570 : return rinfo;
2062 : : }
2063 : :
2064 : :
2065 : : /*
2066 : : * reconsider_outer_join_clauses
2067 : : * Re-examine any outer-join clauses that were set aside by
2068 : : * distribute_qual_to_rels(), and see if we can derive any
2069 : : * EquivalenceClasses from them. Then, if they were not made
2070 : : * redundant, push them out into the regular join-clause lists.
2071 : : *
2072 : : * When we have mergejoinable clauses A = B that are outer-join clauses,
2073 : : * we can't blindly combine them with other clauses A = C to deduce B = C,
2074 : : * since in fact the "equality" A = B won't necessarily hold above the
2075 : : * outer join (one of the variables might be NULL instead). Nonetheless
2076 : : * there are cases where we can add qual clauses using transitivity.
2077 : : *
2078 : : * One case that we look for here is an outer-join clause OUTERVAR = INNERVAR
2079 : : * for which there is also an equivalence clause OUTERVAR = CONSTANT.
2080 : : * It is safe and useful to push a clause INNERVAR = CONSTANT into the
2081 : : * evaluation of the inner (nullable) relation, because any inner rows not
2082 : : * meeting this condition will not contribute to the outer-join result anyway.
2083 : : * (Any outer rows they could join to will be eliminated by the pushed-down
2084 : : * equivalence clause.)
2085 : : *
2086 : : * Note that the above rule does not work for full outer joins; nor is it
2087 : : * very interesting to consider cases where the generated equivalence clause
2088 : : * would involve relations outside the outer join, since such clauses couldn't
2089 : : * be pushed into the inner side's scan anyway. So the restriction to
2090 : : * outervar = pseudoconstant is not really giving up anything.
2091 : : *
2092 : : * For full-join cases, we can only do something useful if it's a FULL JOIN
2093 : : * USING and a merged column has an equivalence MERGEDVAR = CONSTANT.
2094 : : * By the time it gets here, the merged column will look like
2095 : : * COALESCE(LEFTVAR, RIGHTVAR)
2096 : : * and we will have a full-join clause LEFTVAR = RIGHTVAR that we can match
2097 : : * the COALESCE expression to. In this situation we can push LEFTVAR = CONSTANT
2098 : : * and RIGHTVAR = CONSTANT into the input relations, since any rows not
2099 : : * meeting these conditions cannot contribute to the join result.
2100 : : *
2101 : : * Again, there isn't any traction to be gained by trying to deal with
2102 : : * clauses comparing a mergedvar to a non-pseudoconstant. So we can make
2103 : : * use of the EquivalenceClasses to search for matching variables that were
2104 : : * equivalenced to constants. The interesting outer-join clauses were
2105 : : * accumulated for us by distribute_qual_to_rels.
2106 : : *
2107 : : * When we find one of these cases, we implement the changes we want by
2108 : : * generating a new equivalence clause INNERVAR = CONSTANT (or LEFTVAR, etc)
2109 : : * and pushing it into the EquivalenceClass structures. This is because we
2110 : : * may already know that INNERVAR is equivalenced to some other var(s), and
2111 : : * we'd like the constant to propagate to them too. Note that it would be
2112 : : * unsafe to merge any existing EC for INNERVAR with the OUTERVAR's EC ---
2113 : : * that could result in propagating constant restrictions from
2114 : : * INNERVAR to OUTERVAR, which would be very wrong.
2115 : : *
2116 : : * It's possible that the INNERVAR is also an OUTERVAR for some other
2117 : : * outer-join clause, in which case the process can be repeated. So we repeat
2118 : : * looping over the lists of clauses until no further deductions can be made.
2119 : : * Whenever we do make a deduction, we remove the generating clause from the
2120 : : * lists, since we don't want to make the same deduction twice.
2121 : : *
2122 : : * If we don't find any match for a set-aside outer join clause, we must
2123 : : * throw it back into the regular joinclause processing by passing it to
2124 : : * distribute_restrictinfo_to_rels(). If we do generate a derived clause,
2125 : : * however, the outer-join clause is redundant. We must still put some
2126 : : * clause into the regular processing, because otherwise the join will be
2127 : : * seen as a clauseless join and avoided during join order searching.
2128 : : * We handle this by generating a constant-TRUE clause that is marked with
2129 : : * the same required_relids etc as the removed outer-join clause, thus
2130 : : * making it a join clause between the correct relations.
2131 : : */
2132 : : void
2133 : 257712 : reconsider_outer_join_clauses(PlannerInfo *root)
2134 : : {
2135 : : bool found;
2136 : : ListCell *cell;
2137 : :
2138 : : /* Outer loop repeats until we find no more deductions */
2139 : : do
2140 : : {
2141 : 259364 : found = false;
2142 : :
2143 : : /* Process the LEFT JOIN clauses */
2144 [ + + + + : 281787 : foreach(cell, root->left_join_clauses)
+ + ]
2145 : : {
2146 : 22423 : OuterJoinClauseInfo *ojcinfo = (OuterJoinClauseInfo *) lfirst(cell);
2147 : :
2148 [ + + ]: 22423 : if (reconsider_outer_join_clause(root, ojcinfo, true))
2149 : : {
2150 : 517 : RestrictInfo *rinfo = ojcinfo->rinfo;
2151 : :
2152 : 517 : found = true;
2153 : : /* remove it from the list */
2154 : 517 : root->left_join_clauses =
2155 : 517 : foreach_delete_current(root->left_join_clauses, cell);
2156 : : /* throw back a dummy replacement clause (see notes above) */
2157 : 517 : rinfo = make_restrictinfo(root,
2158 : 517 : (Expr *) makeBoolConst(true, false),
2159 : 517 : rinfo->is_pushed_down,
2160 : 517 : rinfo->has_clone,
2161 : 517 : rinfo->is_clone,
2162 : : false, /* pseudoconstant */
2163 : : 0, /* security_level */
2164 : : rinfo->required_relids,
2165 : : rinfo->incompatible_relids,
2166 : : rinfo->outer_relids);
2167 : 517 : distribute_restrictinfo_to_rels(root, rinfo);
2168 : : }
2169 : : }
2170 : :
2171 : : /* Process the RIGHT JOIN clauses */
2172 [ + + + + : 288518 : foreach(cell, root->right_join_clauses)
+ + ]
2173 : : {
2174 : 29154 : OuterJoinClauseInfo *ojcinfo = (OuterJoinClauseInfo *) lfirst(cell);
2175 : :
2176 [ + + ]: 29154 : if (reconsider_outer_join_clause(root, ojcinfo, false))
2177 : : {
2178 : 1140 : RestrictInfo *rinfo = ojcinfo->rinfo;
2179 : :
2180 : 1140 : found = true;
2181 : : /* remove it from the list */
2182 : 1140 : root->right_join_clauses =
2183 : 1140 : foreach_delete_current(root->right_join_clauses, cell);
2184 : : /* throw back a dummy replacement clause (see notes above) */
2185 : 1140 : rinfo = make_restrictinfo(root,
2186 : 1140 : (Expr *) makeBoolConst(true, false),
2187 : 1140 : rinfo->is_pushed_down,
2188 : 1140 : rinfo->has_clone,
2189 : 1140 : rinfo->is_clone,
2190 : : false, /* pseudoconstant */
2191 : : 0, /* security_level */
2192 : : rinfo->required_relids,
2193 : : rinfo->incompatible_relids,
2194 : : rinfo->outer_relids);
2195 : 1140 : distribute_restrictinfo_to_rels(root, rinfo);
2196 : : }
2197 : : }
2198 : :
2199 : : /* Process the FULL JOIN clauses */
2200 [ + + + + : 260409 : foreach(cell, root->full_join_clauses)
+ + ]
2201 : : {
2202 : 1045 : OuterJoinClauseInfo *ojcinfo = (OuterJoinClauseInfo *) lfirst(cell);
2203 : :
2204 [ + + ]: 1045 : if (reconsider_full_join_clause(root, ojcinfo))
2205 : : {
2206 : 5 : RestrictInfo *rinfo = ojcinfo->rinfo;
2207 : :
2208 : 5 : found = true;
2209 : : /* remove it from the list */
2210 : 5 : root->full_join_clauses =
2211 : 5 : foreach_delete_current(root->full_join_clauses, cell);
2212 : : /* throw back a dummy replacement clause (see notes above) */
2213 : 5 : rinfo = make_restrictinfo(root,
2214 : 5 : (Expr *) makeBoolConst(true, false),
2215 : 5 : rinfo->is_pushed_down,
2216 : 5 : rinfo->has_clone,
2217 : 5 : rinfo->is_clone,
2218 : : false, /* pseudoconstant */
2219 : : 0, /* security_level */
2220 : : rinfo->required_relids,
2221 : : rinfo->incompatible_relids,
2222 : : rinfo->outer_relids);
2223 : 5 : distribute_restrictinfo_to_rels(root, rinfo);
2224 : : }
2225 : : }
2226 [ + + ]: 259364 : } while (found);
2227 : :
2228 : : /* Now, any remaining clauses have to be thrown back */
2229 [ + + + + : 279302 : foreach(cell, root->left_join_clauses)
+ + ]
2230 : : {
2231 : 21590 : OuterJoinClauseInfo *ojcinfo = (OuterJoinClauseInfo *) lfirst(cell);
2232 : :
2233 : 21590 : distribute_restrictinfo_to_rels(root, ojcinfo->rinfo);
2234 : : }
2235 [ + + + + : 284539 : foreach(cell, root->right_join_clauses)
+ + ]
2236 : : {
2237 : 26827 : OuterJoinClauseInfo *ojcinfo = (OuterJoinClauseInfo *) lfirst(cell);
2238 : :
2239 : 26827 : distribute_restrictinfo_to_rels(root, ojcinfo->rinfo);
2240 : : }
2241 [ + + + + : 258752 : foreach(cell, root->full_join_clauses)
+ + ]
2242 : : {
2243 : 1040 : OuterJoinClauseInfo *ojcinfo = (OuterJoinClauseInfo *) lfirst(cell);
2244 : :
2245 : 1040 : distribute_restrictinfo_to_rels(root, ojcinfo->rinfo);
2246 : : }
2247 : 257712 : }
2248 : :
2249 : : /*
2250 : : * reconsider_outer_join_clauses for a single LEFT/RIGHT JOIN clause
2251 : : *
2252 : : * Returns true if we were able to propagate a constant through the clause.
2253 : : */
2254 : : static bool
2255 : 51577 : reconsider_outer_join_clause(PlannerInfo *root, OuterJoinClauseInfo *ojcinfo,
2256 : : bool outer_on_left)
2257 : : {
2258 : 51577 : RestrictInfo *rinfo = ojcinfo->rinfo;
2259 : 51577 : SpecialJoinInfo *sjinfo = ojcinfo->sjinfo;
2260 : : Expr *outervar,
2261 : : *innervar;
2262 : : Oid opno,
2263 : : collation,
2264 : : left_type,
2265 : : right_type,
2266 : : inner_datatype;
2267 : : Relids inner_relids;
2268 : : ListCell *lc1;
2269 : :
2270 : : Assert(is_opclause(rinfo->clause));
2271 : 51577 : opno = ((OpExpr *) rinfo->clause)->opno;
2272 : 51577 : collation = ((OpExpr *) rinfo->clause)->inputcollid;
2273 : :
2274 : : /* Extract needed info from the clause */
2275 : 51577 : op_input_types(opno, &left_type, &right_type);
2276 [ + + ]: 51577 : if (outer_on_left)
2277 : : {
2278 : 22423 : outervar = (Expr *) get_leftop(rinfo->clause);
2279 : 22423 : innervar = (Expr *) get_rightop(rinfo->clause);
2280 : 22423 : inner_datatype = right_type;
2281 : 22423 : inner_relids = rinfo->right_relids;
2282 : : }
2283 : : else
2284 : : {
2285 : 29154 : outervar = (Expr *) get_rightop(rinfo->clause);
2286 : 29154 : innervar = (Expr *) get_leftop(rinfo->clause);
2287 : 29154 : inner_datatype = left_type;
2288 : 29154 : inner_relids = rinfo->left_relids;
2289 : : }
2290 : :
2291 : : /* Scan EquivalenceClasses for a match to outervar */
2292 [ + - + + : 313975 : foreach(lc1, root->eq_classes)
+ + ]
2293 : : {
2294 : 264055 : EquivalenceClass *cur_ec = (EquivalenceClass *) lfirst(lc1);
2295 : : bool match;
2296 : : ListCell *lc2;
2297 : :
2298 : : /* We don't expect any children yet */
2299 : : Assert(cur_ec->ec_childmembers == NULL);
2300 : :
2301 : : /* Ignore EC unless it contains pseudoconstants */
2302 [ + + ]: 264055 : if (!cur_ec->ec_has_const)
2303 : 204909 : continue;
2304 : : /* Never match to a volatile EC */
2305 [ - + ]: 59146 : if (cur_ec->ec_has_volatile)
2306 : 0 : continue;
2307 : : /* It has to match the outer-join clause as to semantics, too */
2308 [ + + ]: 59146 : if (collation != cur_ec->ec_collation)
2309 : 3860 : continue;
2310 [ + + ]: 55286 : if (!equal(rinfo->mergeopfamilies, cur_ec->ec_opfamilies))
2311 : 10132 : continue;
2312 : : /* Does it contain a match to outervar? */
2313 : 45154 : match = false;
2314 [ + - + + : 143738 : foreach(lc2, cur_ec->ec_members)
+ + ]
2315 : : {
2316 : 100241 : EquivalenceMember *cur_em = (EquivalenceMember *) lfirst(lc2);
2317 : :
2318 : : /* Child members should not exist in ec_members */
2319 : : Assert(!cur_em->em_is_child);
2320 [ + + ]: 100241 : if (equal(outervar, cur_em->em_expr))
2321 : : {
2322 : 1657 : match = true;
2323 : 1657 : break;
2324 : : }
2325 : : }
2326 [ + + ]: 45154 : if (!match)
2327 : 43497 : continue; /* no match, so ignore this EC */
2328 : :
2329 : : /*
2330 : : * Yes it does! Try to generate a clause INNERVAR = CONSTANT for each
2331 : : * CONSTANT in the EC. Note that we must succeed with at least one
2332 : : * constant before we can decide to throw away the outer-join clause.
2333 : : */
2334 : 1657 : match = false;
2335 [ + - + + : 5864 : foreach(lc2, cur_ec->ec_members)
+ + ]
2336 : : {
2337 : 4207 : EquivalenceMember *cur_em = (EquivalenceMember *) lfirst(lc2);
2338 : : Oid eq_op;
2339 : : RestrictInfo *newrinfo;
2340 : : JoinDomain *jdomain;
2341 : :
2342 [ + + ]: 4207 : if (!cur_em->em_is_const)
2343 : 2515 : continue; /* ignore non-const members */
2344 : 1692 : eq_op = select_equality_operator(cur_ec,
2345 : : inner_datatype,
2346 : : cur_em->em_datatype);
2347 [ - + ]: 1692 : if (!OidIsValid(eq_op))
2348 : 0 : continue; /* can't generate equality */
2349 : 1692 : newrinfo = build_implied_join_equality(root,
2350 : : eq_op,
2351 : : cur_ec->ec_collation,
2352 : : innervar,
2353 : : cur_em->em_expr,
2354 : : bms_copy(inner_relids),
2355 : : cur_ec->ec_min_security);
2356 : : /* This equality holds within the OJ's child JoinDomain */
2357 : 1692 : jdomain = find_join_domain(root, sjinfo->syn_righthand);
2358 [ + - ]: 1692 : if (process_equivalence(root, &newrinfo, jdomain))
2359 : 1692 : match = true;
2360 : : }
2361 : :
2362 : : /*
2363 : : * If we were able to equate INNERVAR to any constant, report success.
2364 : : * Otherwise, fall out of the search loop, since we know the OUTERVAR
2365 : : * appears in at most one EC.
2366 : : */
2367 [ + - ]: 1657 : if (match)
2368 : 1657 : return true;
2369 : : else
2370 : 0 : break;
2371 : : }
2372 : :
2373 : 49920 : return false; /* failed to make any deduction */
2374 : : }
2375 : :
2376 : : /*
2377 : : * reconsider_outer_join_clauses for a single FULL JOIN clause
2378 : : *
2379 : : * Returns true if we were able to propagate a constant through the clause.
2380 : : */
2381 : : static bool
2382 : 1045 : reconsider_full_join_clause(PlannerInfo *root, OuterJoinClauseInfo *ojcinfo)
2383 : : {
2384 : 1045 : RestrictInfo *rinfo = ojcinfo->rinfo;
2385 : 1045 : SpecialJoinInfo *sjinfo = ojcinfo->sjinfo;
2386 : 1045 : Relids fjrelids = bms_make_singleton(sjinfo->ojrelid);
2387 : : Expr *leftvar;
2388 : : Expr *rightvar;
2389 : : Oid opno,
2390 : : collation,
2391 : : left_type,
2392 : : right_type;
2393 : : Relids left_relids,
2394 : : right_relids;
2395 : : ListCell *lc1;
2396 : :
2397 : : /* Extract needed info from the clause */
2398 : : Assert(is_opclause(rinfo->clause));
2399 : 1045 : opno = ((OpExpr *) rinfo->clause)->opno;
2400 : 1045 : collation = ((OpExpr *) rinfo->clause)->inputcollid;
2401 : 1045 : op_input_types(opno, &left_type, &right_type);
2402 : 1045 : leftvar = (Expr *) get_leftop(rinfo->clause);
2403 : 1045 : rightvar = (Expr *) get_rightop(rinfo->clause);
2404 : 1045 : left_relids = rinfo->left_relids;
2405 : 1045 : right_relids = rinfo->right_relids;
2406 : :
2407 [ + - + + : 5305 : foreach(lc1, root->eq_classes)
+ + ]
2408 : : {
2409 : 4265 : EquivalenceClass *cur_ec = (EquivalenceClass *) lfirst(lc1);
2410 : 4265 : EquivalenceMember *coal_em = NULL;
2411 : : bool match;
2412 : : bool matchleft;
2413 : : bool matchright;
2414 : : ListCell *lc2;
2415 : 4265 : int coal_idx = -1;
2416 : :
2417 : : /* We don't expect any children yet */
2418 : : Assert(cur_ec->ec_childmembers == NULL);
2419 : :
2420 : : /* Ignore EC unless it contains pseudoconstants */
2421 [ + + ]: 4265 : if (!cur_ec->ec_has_const)
2422 : 4015 : continue;
2423 : : /* Never match to a volatile EC */
2424 [ - + ]: 250 : if (cur_ec->ec_has_volatile)
2425 : 0 : continue;
2426 : : /* It has to match the outer-join clause as to semantics, too */
2427 [ + + ]: 250 : if (collation != cur_ec->ec_collation)
2428 : 30 : continue;
2429 [ - + ]: 220 : if (!equal(rinfo->mergeopfamilies, cur_ec->ec_opfamilies))
2430 : 0 : continue;
2431 : :
2432 : : /*
2433 : : * Does it contain a COALESCE(leftvar, rightvar) construct?
2434 : : *
2435 : : * We can assume the COALESCE() inputs are in the same order as the
2436 : : * join clause, since both were automatically generated in the cases
2437 : : * we care about.
2438 : : *
2439 : : * XXX currently this may fail to match in cross-type cases because
2440 : : * the COALESCE will contain typecast operations while the join clause
2441 : : * may not (if there is a cross-type mergejoin operator available for
2442 : : * the two column types). Is it OK to strip implicit coercions from
2443 : : * the COALESCE arguments?
2444 : : */
2445 : 220 : match = false;
2446 [ + - + + : 645 : foreach(lc2, cur_ec->ec_members)
+ + ]
2447 : : {
2448 : 430 : coal_em = (EquivalenceMember *) lfirst(lc2);
2449 : :
2450 : : /* Child members should not exist in ec_members */
2451 : : Assert(!coal_em->em_is_child);
2452 [ + + ]: 430 : if (IsA(coal_em->em_expr, CoalesceExpr))
2453 : : {
2454 : 15 : CoalesceExpr *cexpr = (CoalesceExpr *) coal_em->em_expr;
2455 : : Node *cfirst;
2456 : : Node *csecond;
2457 : :
2458 [ - + ]: 15 : if (list_length(cexpr->args) != 2)
2459 : 0 : continue;
2460 : 15 : cfirst = (Node *) linitial(cexpr->args);
2461 : 15 : csecond = (Node *) lsecond(cexpr->args);
2462 : :
2463 : : /*
2464 : : * The COALESCE arguments will be marked as possibly nulled by
2465 : : * the full join, while we wish to generate clauses that apply
2466 : : * to the join's inputs. So we must strip the join from the
2467 : : * nullingrels fields of cfirst/csecond before comparing them
2468 : : * to leftvar/rightvar. (Perhaps with a less hokey
2469 : : * representation for FULL JOIN USING output columns, this
2470 : : * wouldn't be needed?)
2471 : : */
2472 : 15 : cfirst = remove_nulling_relids(cfirst, fjrelids, NULL);
2473 : 15 : csecond = remove_nulling_relids(csecond, fjrelids, NULL);
2474 : :
2475 [ + + + - ]: 15 : if (equal(leftvar, cfirst) && equal(rightvar, csecond))
2476 : : {
2477 : 5 : coal_idx = foreach_current_index(lc2);
2478 : 5 : match = true;
2479 : 5 : break;
2480 : : }
2481 : : }
2482 : : }
2483 [ + + ]: 220 : if (!match)
2484 : 215 : continue; /* no match, so ignore this EC */
2485 : :
2486 : : /*
2487 : : * Yes it does! Try to generate clauses LEFTVAR = CONSTANT and
2488 : : * RIGHTVAR = CONSTANT for each CONSTANT in the EC. Note that we must
2489 : : * succeed with at least one constant for each var before we can
2490 : : * decide to throw away the outer-join clause.
2491 : : */
2492 : 5 : matchleft = matchright = false;
2493 [ + - + + : 15 : foreach(lc2, cur_ec->ec_members)
+ + ]
2494 : : {
2495 : 10 : EquivalenceMember *cur_em = (EquivalenceMember *) lfirst(lc2);
2496 : : Oid eq_op;
2497 : : RestrictInfo *newrinfo;
2498 : : JoinDomain *jdomain;
2499 : :
2500 [ + + ]: 10 : if (!cur_em->em_is_const)
2501 : 5 : continue; /* ignore non-const members */
2502 : 5 : eq_op = select_equality_operator(cur_ec,
2503 : : left_type,
2504 : : cur_em->em_datatype);
2505 [ + - ]: 5 : if (OidIsValid(eq_op))
2506 : : {
2507 : 5 : newrinfo = build_implied_join_equality(root,
2508 : : eq_op,
2509 : : cur_ec->ec_collation,
2510 : : leftvar,
2511 : : cur_em->em_expr,
2512 : : bms_copy(left_relids),
2513 : : cur_ec->ec_min_security);
2514 : : /* This equality holds within the lefthand child JoinDomain */
2515 : 5 : jdomain = find_join_domain(root, sjinfo->syn_lefthand);
2516 [ + - ]: 5 : if (process_equivalence(root, &newrinfo, jdomain))
2517 : 5 : matchleft = true;
2518 : : }
2519 : 5 : eq_op = select_equality_operator(cur_ec,
2520 : : right_type,
2521 : : cur_em->em_datatype);
2522 [ + - ]: 5 : if (OidIsValid(eq_op))
2523 : : {
2524 : 5 : newrinfo = build_implied_join_equality(root,
2525 : : eq_op,
2526 : : cur_ec->ec_collation,
2527 : : rightvar,
2528 : : cur_em->em_expr,
2529 : : bms_copy(right_relids),
2530 : : cur_ec->ec_min_security);
2531 : : /* This equality holds within the righthand child JoinDomain */
2532 : 5 : jdomain = find_join_domain(root, sjinfo->syn_righthand);
2533 [ + - ]: 5 : if (process_equivalence(root, &newrinfo, jdomain))
2534 : 5 : matchright = true;
2535 : : }
2536 : : }
2537 : :
2538 : : /*
2539 : : * If we were able to equate both vars to constants, we're done, and
2540 : : * we can throw away the full-join clause as redundant. Moreover, we
2541 : : * can remove the COALESCE entry from the EC, since the added
2542 : : * restrictions ensure it will always have the expected value. (We
2543 : : * don't bother trying to update ec_relids or ec_sources.)
2544 : : */
2545 [ + - + - ]: 5 : if (matchleft && matchright)
2546 : : {
2547 : 5 : cur_ec->ec_members = list_delete_nth_cell(cur_ec->ec_members, coal_idx);
2548 : 5 : return true;
2549 : : }
2550 : :
2551 : : /*
2552 : : * Otherwise, fall out of the search loop, since we know the COALESCE
2553 : : * appears in at most one EC (XXX might stop being true if we allow
2554 : : * stripping of coercions above?)
2555 : : */
2556 : 0 : break;
2557 : : }
2558 : :
2559 : 1040 : return false; /* failed to make any deduction */
2560 : : }
2561 : :
2562 : : /*
2563 : : * find_join_domain
2564 : : * Find the highest JoinDomain enclosed within the given relid set.
2565 : : *
2566 : : * (We could avoid this search at the cost of complicating APIs elsewhere,
2567 : : * which doesn't seem worth it.)
2568 : : */
2569 : : static JoinDomain *
2570 : 1702 : find_join_domain(PlannerInfo *root, Relids relids)
2571 : : {
2572 : : ListCell *lc;
2573 : :
2574 [ + - + - : 3491 : foreach(lc, root->join_domains)
+ - ]
2575 : : {
2576 : 3491 : JoinDomain *jdomain = (JoinDomain *) lfirst(lc);
2577 : :
2578 [ + + ]: 3491 : if (bms_is_subset(jdomain->jd_relids, relids))
2579 : 1702 : return jdomain;
2580 : : }
2581 [ # # ]: 0 : elog(ERROR, "failed to find appropriate JoinDomain");
2582 : : return NULL; /* keep compiler quiet */
2583 : : }
2584 : :
2585 : :
2586 : : /*
2587 : : * exprs_known_equal
2588 : : * Detect whether two expressions are known equal due to equivalence
2589 : : * relationships.
2590 : : *
2591 : : * If opfamily is given, the expressions must be known equal per the semantics
2592 : : * of that opfamily (note it has to be a btree opfamily, since those are the
2593 : : * only opfamilies equivclass.c deals with). If opfamily is InvalidOid, we'll
2594 : : * return true if they're equal according to any opfamily, which is fuzzy but
2595 : : * OK for estimation purposes.
2596 : : *
2597 : : * Note: does not bother to check for "equal(item1, item2)"; caller must
2598 : : * check that case if it's possible to pass identical items.
2599 : : */
2600 : : bool
2601 : 25396 : exprs_known_equal(PlannerInfo *root, Node *item1, Node *item2, Oid opfamily)
2602 : : {
2603 : : ListCell *lc1;
2604 : :
2605 [ + + + + : 195323 : foreach(lc1, root->eq_classes)
+ + ]
2606 : : {
2607 : 172763 : EquivalenceClass *ec = (EquivalenceClass *) lfirst(lc1);
2608 : 172763 : bool item1member = false;
2609 : 172763 : bool item2member = false;
2610 : : ListCell *lc2;
2611 : :
2612 : : /* Never match to a volatile EC */
2613 [ - + ]: 172763 : if (ec->ec_has_volatile)
2614 : 0 : continue;
2615 : :
2616 : : /*
2617 : : * It's okay to consider ec_broken ECs here. Brokenness just means we
2618 : : * couldn't derive all the implied clauses we'd have liked to; it does
2619 : : * not invalidate our knowledge that the members are equal.
2620 : : */
2621 : :
2622 : : /* Ignore if this EC doesn't use specified opfamily */
2623 [ + + ]: 172763 : if (OidIsValid(opfamily) &&
2624 [ + + ]: 550 : !list_member_oid(ec->ec_opfamilies, opfamily))
2625 : 190 : continue;
2626 : :
2627 : : /* Ignore children here */
2628 [ + - + + : 413209 : foreach(lc2, ec->ec_members)
+ + ]
2629 : : {
2630 : 243472 : EquivalenceMember *em = (EquivalenceMember *) lfirst(lc2);
2631 : :
2632 : : /* Child members should not exist in ec_members */
2633 : : Assert(!em->em_is_child);
2634 [ + + ]: 243472 : if (equal(item1, em->em_expr))
2635 : 12976 : item1member = true;
2636 [ + + ]: 230496 : else if (equal(item2, em->em_expr))
2637 : 20959 : item2member = true;
2638 : : /* Exit as soon as equality is proven */
2639 [ + + + + ]: 243472 : if (item1member && item2member)
2640 : 2836 : return true;
2641 : : }
2642 : : }
2643 : 22560 : return false;
2644 : : }
2645 : :
2646 : :
2647 : : /*
2648 : : * match_eclasses_to_foreign_key_col
2649 : : * See whether a foreign key column match is proven by any eclass.
2650 : : *
2651 : : * If the referenced and referencing Vars of the fkey's colno'th column are
2652 : : * known equal due to any eclass, return that eclass; otherwise return NULL.
2653 : : * (In principle there might be more than one matching eclass if multiple
2654 : : * collations are involved, but since collation doesn't matter for equality,
2655 : : * we ignore that fine point here.) This is much like exprs_known_equal,
2656 : : * except for the format of the input.
2657 : : *
2658 : : * On success, we also set fkinfo->eclass[colno] to the matching eclass,
2659 : : * and set fkinfo->fk_eclass_member[colno] to the eclass member for the
2660 : : * referencing Var.
2661 : : */
2662 : : EquivalenceClass *
2663 : 2191 : match_eclasses_to_foreign_key_col(PlannerInfo *root,
2664 : : ForeignKeyOptInfo *fkinfo,
2665 : : int colno)
2666 : : {
2667 : 2191 : Index var1varno = fkinfo->con_relid;
2668 : 2191 : AttrNumber var1attno = fkinfo->conkey[colno];
2669 : 2191 : Index var2varno = fkinfo->ref_relid;
2670 : 2191 : AttrNumber var2attno = fkinfo->confkey[colno];
2671 : 2191 : Oid eqop = fkinfo->conpfeqop[colno];
2672 : 2191 : RelOptInfo *rel1 = root->simple_rel_array[var1varno];
2673 : 2191 : RelOptInfo *rel2 = root->simple_rel_array[var2varno];
2674 : 2191 : List *opfamilies = NIL; /* compute only if needed */
2675 : : Bitmapset *matching_ecs;
2676 : : int i;
2677 : :
2678 : : /* Consider only eclasses mentioning both relations */
2679 : : Assert(root->ec_merging_done);
2680 : : Assert(IS_SIMPLE_REL(rel1));
2681 : : Assert(IS_SIMPLE_REL(rel2));
2682 : 2191 : matching_ecs = bms_intersect(rel1->eclass_indexes,
2683 : 2191 : rel2->eclass_indexes);
2684 : :
2685 : 2191 : i = -1;
2686 [ + + ]: 2271 : while ((i = bms_next_member(matching_ecs, i)) >= 0)
2687 : : {
2688 : 653 : EquivalenceClass *ec = (EquivalenceClass *) list_nth(root->eq_classes,
2689 : : i);
2690 : 653 : EquivalenceMember *item1_em = NULL;
2691 : 653 : EquivalenceMember *item2_em = NULL;
2692 : : ListCell *lc2;
2693 : :
2694 : : /* Never match to a volatile EC */
2695 [ - + ]: 653 : if (ec->ec_has_volatile)
2696 : 0 : continue;
2697 : :
2698 : : /*
2699 : : * It's okay to consider "broken" ECs here, see exprs_known_equal.
2700 : : * Ignore children here.
2701 : : */
2702 [ + - + + : 1511 : foreach(lc2, ec->ec_members)
+ + ]
2703 : : {
2704 : 1431 : EquivalenceMember *em = (EquivalenceMember *) lfirst(lc2);
2705 : : Var *var;
2706 : :
2707 : : /* Child members should not exist in ec_members */
2708 : : Assert(!em->em_is_child);
2709 : :
2710 : : /* EM must be a Var, possibly with RelabelType */
2711 : 1431 : var = (Var *) em->em_expr;
2712 [ + - - + ]: 1431 : while (var && IsA(var, RelabelType))
2713 : 0 : var = (Var *) ((RelabelType *) var)->arg;
2714 [ + - + + ]: 1431 : if (!(var && IsA(var, Var)))
2715 : 15 : continue;
2716 : :
2717 : : /* Match? */
2718 [ + + + + ]: 1416 : if (var->varno == var1varno && var->varattno == var1attno)
2719 : 573 : item1_em = em;
2720 [ + + + + ]: 843 : else if (var->varno == var2varno && var->varattno == var2attno)
2721 : 573 : item2_em = em;
2722 : :
2723 : : /* Have we found both PK and FK column in this EC? */
2724 [ + + + + ]: 1416 : if (item1_em && item2_em)
2725 : : {
2726 : : /*
2727 : : * Succeed if eqop matches EC's opfamilies. We could test
2728 : : * this before scanning the members, but it's probably cheaper
2729 : : * to test for member matches first.
2730 : : */
2731 [ + - ]: 573 : if (opfamilies == NIL) /* compute if we didn't already */
2732 : 573 : opfamilies = get_mergejoin_opfamilies(eqop);
2733 [ + - ]: 573 : if (equal(opfamilies, ec->ec_opfamilies))
2734 : : {
2735 : 573 : fkinfo->eclass[colno] = ec;
2736 : 573 : fkinfo->fk_eclass_member[colno] = item2_em;
2737 : 573 : return ec;
2738 : : }
2739 : : /* Otherwise, done with this EC, move on to the next */
2740 : 0 : break;
2741 : : }
2742 : : }
2743 : : }
2744 : 1618 : return NULL;
2745 : : }
2746 : :
2747 : : /*
2748 : : * find_derived_clause_for_ec_member
2749 : : * Search for a previously-derived clause mentioning the given EM.
2750 : : *
2751 : : * The eclass should be an ec_has_const EC, of which the EM is a non-const
2752 : : * member. This should ensure there is just one derived clause mentioning
2753 : : * the EM (and equating it to a constant).
2754 : : * Returns NULL if no such clause can be found.
2755 : : */
2756 : : RestrictInfo *
2757 : 5 : find_derived_clause_for_ec_member(PlannerInfo *root,
2758 : : EquivalenceClass *ec,
2759 : : EquivalenceMember *em)
2760 : : {
2761 : : Assert(ec->ec_has_const);
2762 : : Assert(!em->em_is_const);
2763 : :
2764 : 5 : return ec_search_derived_clause_for_ems(root, ec, em, NULL, NULL);
2765 : : }
2766 : :
2767 : :
2768 : : /*
2769 : : * add_child_rel_equivalences
2770 : : * Search for EC members that reference the root parent of child_rel, and
2771 : : * add transformed members referencing the child_rel.
2772 : : *
2773 : : * Note that this function won't be called at all unless we have at least some
2774 : : * reason to believe that the EC members it generates will be useful.
2775 : : *
2776 : : * parent_rel and child_rel could be derived from appinfo, but since the
2777 : : * caller has already computed them, we might as well just pass them in.
2778 : : *
2779 : : * The passed-in AppendRelInfo is not used when the parent_rel is not a
2780 : : * top-level baserel, since it shows the mapping from the parent_rel but
2781 : : * we need to translate EC expressions that refer to the top-level parent.
2782 : : * Using it is faster than using adjust_appendrel_attrs_multilevel(), though,
2783 : : * so we prefer it when we can.
2784 : : */
2785 : : void
2786 : 28738 : add_child_rel_equivalences(PlannerInfo *root,
2787 : : AppendRelInfo *appinfo,
2788 : : RelOptInfo *parent_rel,
2789 : : RelOptInfo *child_rel)
2790 : : {
2791 : 28738 : Relids top_parent_relids = child_rel->top_parent_relids;
2792 : 28738 : Relids child_relids = child_rel->relids;
2793 : : int i;
2794 : :
2795 : : /*
2796 : : * EC merging should be complete already, so we can use the parent rel's
2797 : : * eclass_indexes to avoid searching all of root->eq_classes.
2798 : : */
2799 : : Assert(root->ec_merging_done);
2800 : : Assert(IS_SIMPLE_REL(parent_rel));
2801 : :
2802 : 28738 : i = -1;
2803 [ + + ]: 81900 : while ((i = bms_next_member(parent_rel->eclass_indexes, i)) >= 0)
2804 : : {
2805 : 53162 : EquivalenceClass *cur_ec = (EquivalenceClass *) list_nth(root->eq_classes, i);
2806 : :
2807 : : /*
2808 : : * If this EC contains a volatile expression, then generating child
2809 : : * EMs would be downright dangerous, so skip it. We rely on a
2810 : : * volatile EC having only one EM.
2811 : : */
2812 [ - + ]: 53162 : if (cur_ec->ec_has_volatile)
2813 : 0 : continue;
2814 : :
2815 : : /* Sanity check eclass_indexes only contain ECs for parent_rel */
2816 : : Assert(bms_is_subset(top_parent_relids, cur_ec->ec_relids));
2817 : :
2818 [ + - + + : 182062 : foreach_node(EquivalenceMember, cur_em, cur_ec->ec_members)
+ + ]
2819 : : {
2820 [ + + ]: 75738 : if (cur_em->em_is_const)
2821 : 2762 : continue; /* ignore consts here */
2822 : :
2823 : : /* Child members should not exist in ec_members */
2824 : : Assert(!cur_em->em_is_child);
2825 : :
2826 : : /*
2827 : : * Consider only members that reference and can be computed at
2828 : : * child's topmost parent rel. In particular we want to exclude
2829 : : * parent-rel Vars that have nonempty varnullingrels. Translating
2830 : : * those might fail, if the transformed expression wouldn't be a
2831 : : * simple Var; and in any case it wouldn't produce a member that
2832 : : * has any use in creating plans for the child rel.
2833 : : */
2834 [ + + ]: 72976 : if (bms_is_subset(cur_em->em_relids, top_parent_relids) &&
2835 [ + - ]: 51253 : !bms_is_empty(cur_em->em_relids))
2836 : : {
2837 : : /* OK, generate transformed child version */
2838 : : Expr *child_expr;
2839 : : Relids new_relids;
2840 : :
2841 [ + + ]: 51253 : if (parent_rel->reloptkind == RELOPT_BASEREL)
2842 : : {
2843 : : /* Simple single-level transformation */
2844 : : child_expr = (Expr *)
2845 : 44056 : adjust_appendrel_attrs(root,
2846 : 44056 : (Node *) cur_em->em_expr,
2847 : : 1, &appinfo);
2848 : : }
2849 : : else
2850 : : {
2851 : : /* Must do multi-level transformation */
2852 : : child_expr = (Expr *)
2853 : 7197 : adjust_appendrel_attrs_multilevel(root,
2854 : 7197 : (Node *) cur_em->em_expr,
2855 : : child_rel,
2856 : 7197 : child_rel->top_parent);
2857 : : }
2858 : :
2859 : : /*
2860 : : * Transform em_relids to match. Note we do *not* do
2861 : : * pull_varnos(child_expr) here, as for example the
2862 : : * transformation might have substituted a constant, but we
2863 : : * don't want the child member to be marked as constant.
2864 : : */
2865 : 51253 : new_relids = bms_difference(cur_em->em_relids,
2866 : : top_parent_relids);
2867 : 51253 : new_relids = bms_add_members(new_relids, child_relids);
2868 : :
2869 : 51253 : add_child_eq_member(root,
2870 : : cur_ec,
2871 : : i,
2872 : : child_expr,
2873 : : new_relids,
2874 : : cur_em->em_jdomain,
2875 : : cur_em,
2876 : : cur_em->em_datatype,
2877 : : child_rel->relid);
2878 : : }
2879 : : }
2880 : : }
2881 : 28738 : }
2882 : :
2883 : : /*
2884 : : * add_child_join_rel_equivalences
2885 : : * Like add_child_rel_equivalences(), but for joinrels
2886 : : *
2887 : : * Here we find the ECs relevant to the top parent joinrel and add transformed
2888 : : * member expressions that refer to this child joinrel.
2889 : : *
2890 : : * Note that this function won't be called at all unless we have at least some
2891 : : * reason to believe that the EC members it generates will be useful.
2892 : : */
2893 : : void
2894 : 14885 : add_child_join_rel_equivalences(PlannerInfo *root,
2895 : : int nappinfos, AppendRelInfo **appinfos,
2896 : : RelOptInfo *parent_joinrel,
2897 : : RelOptInfo *child_joinrel)
2898 : : {
2899 : 14885 : Relids top_parent_relids = child_joinrel->top_parent_relids;
2900 : 14885 : Relids child_relids = child_joinrel->relids;
2901 : : Bitmapset *matching_ecs;
2902 : : MemoryContext oldcontext;
2903 : : int i;
2904 : :
2905 : : Assert(IS_JOIN_REL(child_joinrel) && IS_JOIN_REL(parent_joinrel));
2906 : :
2907 : : /* We need consider only ECs that mention the parent joinrel */
2908 : 14885 : matching_ecs = get_eclass_indexes_for_relids(root, top_parent_relids);
2909 : :
2910 : : /*
2911 : : * If we're being called during GEQO join planning, we still have to
2912 : : * create any new EC members in the main planner context, to avoid having
2913 : : * a corrupt EC data structure after the GEQO context is reset. This is
2914 : : * problematic since we'll leak memory across repeated GEQO cycles. For
2915 : : * now, though, bloat is better than crash. If it becomes a real issue
2916 : : * we'll have to do something to avoid generating duplicate EC members.
2917 : : */
2918 : 14885 : oldcontext = MemoryContextSwitchTo(root->planner_cxt);
2919 : :
2920 : 14885 : i = -1;
2921 [ + + ]: 39801 : while ((i = bms_next_member(matching_ecs, i)) >= 0)
2922 : : {
2923 : 24916 : EquivalenceClass *cur_ec = (EquivalenceClass *) list_nth(root->eq_classes, i);
2924 : :
2925 : : /*
2926 : : * If this EC contains a volatile expression, then generating child
2927 : : * EMs would be downright dangerous, so skip it. We rely on a
2928 : : * volatile EC having only one EM.
2929 : : */
2930 [ - + ]: 24916 : if (cur_ec->ec_has_volatile)
2931 : 0 : continue;
2932 : :
2933 : : /* Sanity check on get_eclass_indexes_for_relids result */
2934 : : Assert(bms_overlap(top_parent_relids, cur_ec->ec_relids));
2935 : :
2936 [ + - + + : 92622 : foreach_node(EquivalenceMember, cur_em, cur_ec->ec_members)
+ + ]
2937 : : {
2938 [ + + ]: 42790 : if (cur_em->em_is_const)
2939 : 1898 : continue; /* ignore consts here */
2940 : :
2941 : : /* Child members should not exist in ec_members */
2942 : : Assert(!cur_em->em_is_child);
2943 : :
2944 : : /*
2945 : : * We may ignore expressions that reference a single baserel,
2946 : : * because add_child_rel_equivalences should have handled them.
2947 : : */
2948 [ + + ]: 40892 : if (bms_membership(cur_em->em_relids) != BMS_MULTIPLE)
2949 : 38715 : continue;
2950 : :
2951 : : /* Does this member reference child's topmost parent rel? */
2952 [ + - ]: 2177 : if (bms_overlap(cur_em->em_relids, top_parent_relids))
2953 : : {
2954 : : /* Yes, generate transformed child version */
2955 : : Expr *child_expr;
2956 : : Relids new_relids;
2957 : :
2958 [ + + ]: 2177 : if (parent_joinrel->reloptkind == RELOPT_JOINREL)
2959 : : {
2960 : : /* Simple single-level transformation */
2961 : : child_expr = (Expr *)
2962 : 2097 : adjust_appendrel_attrs(root,
2963 : 2097 : (Node *) cur_em->em_expr,
2964 : : nappinfos, appinfos);
2965 : : }
2966 : : else
2967 : : {
2968 : : /* Must do multi-level transformation */
2969 : : Assert(parent_joinrel->reloptkind == RELOPT_OTHER_JOINREL);
2970 : : child_expr = (Expr *)
2971 : 80 : adjust_appendrel_attrs_multilevel(root,
2972 : 80 : (Node *) cur_em->em_expr,
2973 : : child_joinrel,
2974 : 80 : child_joinrel->top_parent);
2975 : : }
2976 : :
2977 : : /*
2978 : : * Transform em_relids to match. Note we do *not* do
2979 : : * pull_varnos(child_expr) here, as for example the
2980 : : * transformation might have substituted a constant, but we
2981 : : * don't want the child member to be marked as constant.
2982 : : */
2983 : 2177 : new_relids = bms_difference(cur_em->em_relids,
2984 : : top_parent_relids);
2985 : 2177 : new_relids = bms_add_members(new_relids, child_relids);
2986 : :
2987 : : /*
2988 : : * Add new child member to the EquivalenceClass. Because this
2989 : : * is a RELOPT_OTHER_JOINREL which has multiple component
2990 : : * relids, there is no ideal place to store these members in
2991 : : * the class. Ordinarily, child members are stored in the
2992 : : * ec_childmembers[] array element corresponding to their
2993 : : * relid, however, here we have multiple component relids, so
2994 : : * there's no single ec_childmembers[] array element to store
2995 : : * this member. So that we still correctly find this member
2996 : : * in loops iterating over an EquivalenceMemberIterator, we
2997 : : * opt to store the member in the ec_childmembers array in
2998 : : * only the first component relid slot of the array. This
2999 : : * allows the member to be found, providing callers of
3000 : : * setup_eclass_member_iterator() specify all the component
3001 : : * relids for the RELOPT_OTHER_JOINREL, which they do. If we
3002 : : * opted to store the member in each ec_childmembers[] element
3003 : : * for all the component relids, then that would just result
3004 : : * in eclass_member_iterator_next() finding the member
3005 : : * multiple times, which is a waste of effort.
3006 : : */
3007 : 2177 : add_child_eq_member(root,
3008 : : cur_ec,
3009 : : -1,
3010 : : child_expr,
3011 : : new_relids,
3012 : : cur_em->em_jdomain,
3013 : : cur_em,
3014 : : cur_em->em_datatype,
3015 : 2177 : bms_next_member(child_joinrel->relids, -1));
3016 : : }
3017 : : }
3018 : : }
3019 : :
3020 : 14885 : MemoryContextSwitchTo(oldcontext);
3021 : 14885 : }
3022 : :
3023 : : /*
3024 : : * add_setop_child_rel_equivalences
3025 : : * Add equivalence members for each non-resjunk target in 'child_tlist'
3026 : : * to the EquivalenceClass in the corresponding setop_pathkey's pk_eclass.
3027 : : *
3028 : : * 'root' is the PlannerInfo belonging to the top-level set operation.
3029 : : * 'child_rel' is the RelOptInfo of the child relation we're adding
3030 : : * EquivalenceMembers for.
3031 : : * 'child_tlist' is the target list for the setop child relation. The target
3032 : : * list expressions are what we add as EquivalenceMembers.
3033 : : * 'setop_pathkeys' is a list of PathKeys which must contain an entry for each
3034 : : * non-resjunk target in 'child_tlist'.
3035 : : */
3036 : : void
3037 : 10328 : add_setop_child_rel_equivalences(PlannerInfo *root, RelOptInfo *child_rel,
3038 : : List *child_tlist, List *setop_pathkeys)
3039 : : {
3040 : : ListCell *lc;
3041 : 10328 : ListCell *lc2 = list_head(setop_pathkeys);
3042 : :
3043 [ + - + + : 41072 : foreach(lc, child_tlist)
+ + ]
3044 : : {
3045 : 30744 : TargetEntry *tle = lfirst_node(TargetEntry, lc);
3046 : : EquivalenceMember *parent_em;
3047 : : PathKey *pk;
3048 : :
3049 [ - + ]: 30744 : if (tle->resjunk)
3050 : 0 : continue;
3051 : :
3052 [ - + ]: 30744 : if (lc2 == NULL)
3053 [ # # ]: 0 : elog(ERROR, "too few pathkeys for set operation");
3054 : :
3055 : 30744 : pk = lfirst_node(PathKey, lc2);
3056 : 30744 : parent_em = linitial(pk->pk_eclass->ec_members);
3057 : :
3058 : : /*
3059 : : * We can safely pass the parent member as the first member in the
3060 : : * ec_members list as this is added first in generate_union_paths,
3061 : : * likewise, the JoinDomain can be that of the initial member of the
3062 : : * Pathkey's EquivalenceClass. We pass -1 for ec_index since we
3063 : : * maintain the eclass_indexes for the child_rel after the loop.
3064 : : */
3065 : 30744 : add_child_eq_member(root,
3066 : : pk->pk_eclass,
3067 : : -1,
3068 : : tle->expr,
3069 : : child_rel->relids,
3070 : : parent_em->em_jdomain,
3071 : : parent_em,
3072 : 30744 : exprType((Node *) tle->expr),
3073 : : child_rel->relid);
3074 : :
3075 : 30744 : lc2 = lnext(setop_pathkeys, lc2);
3076 : : }
3077 : :
3078 : : /*
3079 : : * transformSetOperationStmt() ensures that the targetlist never contains
3080 : : * any resjunk columns, so all eclasses that exist in 'root' must have
3081 : : * received a new member in the loop above. Add them to the child_rel's
3082 : : * eclass_indexes.
3083 : : */
3084 : 10328 : child_rel->eclass_indexes = bms_add_range(child_rel->eclass_indexes, 0,
3085 : 10328 : list_length(root->eq_classes) - 1);
3086 : 10328 : }
3087 : :
3088 : : /*
3089 : : * setup_eclass_member_iterator
3090 : : * Setup an EquivalenceMemberIterator 'it' to iterate over all parent
3091 : : * EquivalenceMembers and child members belonging to the given 'ec'.
3092 : : *
3093 : : * This iterator returns:
3094 : : * - All parent members stored directly in ec_members for 'ec', and;
3095 : : * - Any child member added to the given ec by add_child_eq_member() where
3096 : : * the child_relid specified in the add_child_eq_member() call is a member
3097 : : * of the 'child_relids' parameter.
3098 : : *
3099 : : * Note:
3100 : : * The given 'child_relids' must remain allocated and not be changed for the
3101 : : * lifetime of the iterator.
3102 : : *
3103 : : * Parameters:
3104 : : * 'it' is a pointer to the iterator to set up. Normally stack allocated.
3105 : : * 'ec' is the EquivalenceClass from which to iterate members for.
3106 : : * 'child_relids' is the relids to return child members for.
3107 : : */
3108 : : void
3109 : 4000962 : setup_eclass_member_iterator(EquivalenceMemberIterator *it,
3110 : : EquivalenceClass *ec, Relids child_relids)
3111 : : {
3112 : 4000962 : it->ec = ec;
3113 : : /* no need to set this if the class has no child members array set */
3114 [ + + ]: 4000962 : it->child_relids = ec->ec_childmembers != NULL ? child_relids : NULL;
3115 : 4000962 : it->current_relid = -1;
3116 : 4000962 : it->current_list = ec->ec_members;
3117 : 4000962 : it->current_cell = list_head(it->current_list);
3118 : 4000962 : }
3119 : :
3120 : : /*
3121 : : * eclass_member_iterator_next
3122 : : * Get the next EquivalenceMember from the EquivalenceMemberIterator 'it',
3123 : : * as setup by setup_eclass_member_iterator(). NULL is returned if there
3124 : : * are no members left, after which callers must not call
3125 : : * eclass_member_iterator_next() again for the given iterator.
3126 : : */
3127 : : EquivalenceMember *
3128 : 9341757 : eclass_member_iterator_next(EquivalenceMemberIterator *it)
3129 : : {
3130 [ + - ]: 9341757 : while (it->current_list != NULL)
3131 : : {
3132 [ + + ]: 9341757 : while (it->current_cell != NULL)
3133 : : {
3134 : : EquivalenceMember *em;
3135 : :
3136 : 6469383 : nextcell:
3137 : 6578065 : em = lfirst_node(EquivalenceMember, it->current_cell);
3138 : 6578065 : it->current_cell = lnext(it->current_list, it->current_cell);
3139 : 6578065 : return em;
3140 : : }
3141 : :
3142 : : /* Search for the next list to return members from */
3143 [ + + ]: 2988248 : while ((it->current_relid = bms_next_member(it->child_relids, it->current_relid)) > 0)
3144 : : {
3145 : : /*
3146 : : * Be paranoid in case we're given relids above what we've sized
3147 : : * the ec_childmembers array to.
3148 : : */
3149 [ - + ]: 224556 : if (it->current_relid >= it->ec->ec_childmembers_size)
3150 : 0 : return NULL;
3151 : :
3152 : 224556 : it->current_list = it->ec->ec_childmembers[it->current_relid];
3153 : :
3154 : : /* If there are members in this list, use it. */
3155 [ + + ]: 224556 : if (it->current_list != NIL)
3156 : : {
3157 : : /* point current_cell to the head of this list */
3158 : 108682 : it->current_cell = list_head(it->current_list);
3159 : 108682 : goto nextcell;
3160 : : }
3161 : : }
3162 : 2763692 : return NULL;
3163 : : }
3164 : :
3165 : 0 : return NULL;
3166 : : }
3167 : :
3168 : : /*
3169 : : * generate_implied_equalities_for_column
3170 : : * Create EC-derived joinclauses usable with a specific column.
3171 : : *
3172 : : * This is used by indxpath.c to extract potentially indexable joinclauses
3173 : : * from ECs, and can be used by foreign data wrappers for similar purposes.
3174 : : * We assume that only expressions in Vars of a single table are of interest,
3175 : : * but the caller provides a callback function to identify exactly which
3176 : : * such expressions it would like to know about.
3177 : : *
3178 : : * We assume that any given table/index column could appear in only one EC.
3179 : : * (This should be true in all but the most pathological cases, and if it
3180 : : * isn't, we stop on the first match anyway.) Therefore, what we return
3181 : : * is a redundant list of clauses equating the table/index column to each of
3182 : : * the other-relation values it is known to be equal to. Any one of
3183 : : * these clauses can be used to create a parameterized path, and there
3184 : : * is no value in using more than one. (But it *is* worthwhile to create
3185 : : * a separate parameterized path for each one, since that leads to different
3186 : : * join orders.)
3187 : : *
3188 : : * The caller can pass a Relids set of rels we aren't interested in joining
3189 : : * to, so as to save the work of creating useless clauses.
3190 : : */
3191 : : List *
3192 : 471280 : generate_implied_equalities_for_column(PlannerInfo *root,
3193 : : RelOptInfo *rel,
3194 : : ec_matches_callback_type callback,
3195 : : void *callback_arg,
3196 : : Relids prohibited_rels)
3197 : : {
3198 : 471280 : List *result = NIL;
3199 : 471280 : bool is_child_rel = (rel->reloptkind == RELOPT_OTHER_MEMBER_REL);
3200 : : Relids parent_relids;
3201 : : int i;
3202 : :
3203 : : /* Should be OK to rely on eclass_indexes */
3204 : : Assert(root->ec_merging_done);
3205 : :
3206 : : /* Indexes are available only on base or "other" member relations. */
3207 : : Assert(IS_SIMPLE_REL(rel));
3208 : :
3209 : : /* If it's a child rel, we'll need to know what its parent(s) are */
3210 [ + + ]: 471280 : if (is_child_rel)
3211 : 10172 : parent_relids = find_childrel_parents(root, rel);
3212 : : else
3213 : 461108 : parent_relids = NULL; /* not used, but keep compiler quiet */
3214 : :
3215 : 471280 : i = -1;
3216 [ + + ]: 1353638 : while ((i = bms_next_member(rel->eclass_indexes, i)) >= 0)
3217 : : {
3218 : 978939 : EquivalenceClass *cur_ec = (EquivalenceClass *) list_nth(root->eq_classes, i);
3219 : : EquivalenceMemberIterator it;
3220 : : EquivalenceMember *cur_em;
3221 : : ListCell *lc2;
3222 : :
3223 : : /* Sanity check eclass_indexes only contain ECs for rel */
3224 : : Assert(is_child_rel || bms_is_subset(rel->relids, cur_ec->ec_relids));
3225 : :
3226 : : /*
3227 : : * Won't generate joinclauses if const or single-member (the latter
3228 : : * test covers the volatile case too)
3229 : : */
3230 [ + + + + ]: 978939 : if (cur_ec->ec_has_const || list_length(cur_ec->ec_members) <= 1)
3231 : 882198 : continue;
3232 : :
3233 : : /*
3234 : : * Scan members, looking for a match to the target column. Note that
3235 : : * child EC members are considered, but only when they belong to the
3236 : : * target relation. (Unlike regular members, the same expression
3237 : : * could be a child member of more than one EC. Therefore, it's
3238 : : * potentially order-dependent which EC a child relation's target
3239 : : * column gets matched to. This is annoying but it only happens in
3240 : : * corner cases, so for now we live with just reporting the first
3241 : : * match. See also get_eclass_for_sort_expr.)
3242 : : */
3243 : 463494 : setup_eclass_member_iterator(&it, cur_ec, rel->relids);
3244 [ + + ]: 1741483 : while ((cur_em = eclass_member_iterator_next(&it)) != NULL)
3245 : : {
3246 [ + + + + ]: 1374986 : if (bms_equal(cur_em->em_relids, rel->relids) &&
3247 : 463750 : callback(root, rel, cur_ec, cur_em, callback_arg))
3248 : 96741 : break;
3249 : : }
3250 : :
3251 [ + + ]: 463494 : if (!cur_em)
3252 : 366753 : continue;
3253 : :
3254 : : /*
3255 : : * Found our match. Scan the other EC members and attempt to generate
3256 : : * joinclauses. Ignore children here.
3257 : : */
3258 [ + - + + : 295549 : foreach(lc2, cur_ec->ec_members)
+ + ]
3259 : : {
3260 : 198808 : EquivalenceMember *other_em = (EquivalenceMember *) lfirst(lc2);
3261 : : Oid eq_op;
3262 : : RestrictInfo *rinfo;
3263 : :
3264 : : /* Child members should not exist in ec_members */
3265 : : Assert(!other_em->em_is_child);
3266 : :
3267 : : /* Make sure it'll be a join to a different rel */
3268 [ + + + + ]: 303602 : if (other_em == cur_em ||
3269 : 104794 : bms_overlap(other_em->em_relids, rel->relids))
3270 : 94099 : continue;
3271 : :
3272 : : /* Forget it if caller doesn't want joins to this rel */
3273 [ + + ]: 104709 : if (bms_overlap(other_em->em_relids, prohibited_rels))
3274 : 130 : continue;
3275 : :
3276 : : /*
3277 : : * Also, if this is a child rel, avoid generating a useless join
3278 : : * to its parent rel(s).
3279 : : */
3280 [ + + + + ]: 110650 : if (is_child_rel &&
3281 : 6071 : bms_overlap(parent_relids, other_em->em_relids))
3282 : 2797 : continue;
3283 : :
3284 : 101782 : eq_op = select_equality_operator(cur_ec,
3285 : : cur_em->em_datatype,
3286 : : other_em->em_datatype);
3287 [ - + ]: 101782 : if (!OidIsValid(eq_op))
3288 : 0 : continue;
3289 : :
3290 : : /* set parent_ec to mark as redundant with other joinclauses */
3291 : 101782 : rinfo = create_join_clause(root, cur_ec, eq_op,
3292 : : cur_em, other_em,
3293 : : cur_ec);
3294 : :
3295 : 101782 : result = lappend(result, rinfo);
3296 : : }
3297 : :
3298 : : /*
3299 : : * If somehow we failed to create any join clauses, we might as well
3300 : : * keep scanning the ECs for another match. But if we did make any,
3301 : : * we're done, because we don't want to return non-redundant clauses.
3302 : : */
3303 [ + + ]: 96741 : if (result)
3304 : 96581 : break;
3305 : : }
3306 : :
3307 : 471280 : return result;
3308 : : }
3309 : :
3310 : : /*
3311 : : * have_relevant_eclass_joinclause
3312 : : * Detect whether there is an EquivalenceClass that could produce
3313 : : * a joinclause involving the two given relations.
3314 : : *
3315 : : * This is essentially a very cut-down version of
3316 : : * generate_join_implied_equalities(). Note it's OK to occasionally say "yes"
3317 : : * incorrectly. Hence we don't bother with details like whether the lack of a
3318 : : * cross-type operator might prevent the clause from actually being generated.
3319 : : * False negatives are not always fatal either: they will discourage, but not
3320 : : * completely prevent, investigation of particular join pathways.
3321 : : */
3322 : : bool
3323 : 170450 : have_relevant_eclass_joinclause(PlannerInfo *root,
3324 : : RelOptInfo *rel1, RelOptInfo *rel2)
3325 : : {
3326 : : Bitmapset *matching_ecs;
3327 : : int i;
3328 : :
3329 : : /*
3330 : : * Examine only eclasses mentioning both rel1 and rel2.
3331 : : *
3332 : : * Note that we do not consider the possibility of an eclass generating
3333 : : * "join" clauses that mention just one of the rels plus an outer join
3334 : : * that could be formed from them. Although such clauses must be
3335 : : * correctly enforced when we form the outer join, they don't seem like
3336 : : * sufficient reason to prioritize this join over other ones. The join
3337 : : * ordering rules will force the join to be made when necessary.
3338 : : */
3339 : 170450 : matching_ecs = get_common_eclass_indexes(root, rel1->relids,
3340 : : rel2->relids);
3341 : :
3342 : 170450 : i = -1;
3343 [ + + ]: 170540 : while ((i = bms_next_member(matching_ecs, i)) >= 0)
3344 : : {
3345 : 145506 : EquivalenceClass *ec = (EquivalenceClass *) list_nth(root->eq_classes,
3346 : : i);
3347 : :
3348 : : /*
3349 : : * Sanity check that get_common_eclass_indexes gave only ECs
3350 : : * containing both rels.
3351 : : */
3352 : : Assert(bms_overlap(rel1->relids, ec->ec_relids));
3353 : : Assert(bms_overlap(rel2->relids, ec->ec_relids));
3354 : :
3355 : : /*
3356 : : * Won't generate joinclauses if single-member (this test covers the
3357 : : * volatile case too)
3358 : : */
3359 [ + + ]: 145506 : if (list_length(ec->ec_members) <= 1)
3360 : 90 : continue;
3361 : :
3362 : : /*
3363 : : * We do not need to examine the individual members of the EC, because
3364 : : * all that we care about is whether each rel overlaps the relids of
3365 : : * at least one member, and get_common_eclass_indexes() and the single
3366 : : * member check above are sufficient to prove that. (As with
3367 : : * have_relevant_joinclause(), it is not necessary that the EC be able
3368 : : * to form a joinclause relating exactly the two given rels, only that
3369 : : * it be able to form a joinclause mentioning both, and this will
3370 : : * surely be true if both of them overlap ec_relids.)
3371 : : *
3372 : : * Note we don't test ec_broken; if we did, we'd need a separate code
3373 : : * path to look through ec_sources. Checking the membership anyway is
3374 : : * OK as a possibly-overoptimistic heuristic.
3375 : : *
3376 : : * We don't test ec_has_const either, even though a const eclass won't
3377 : : * generate real join clauses. This is because if we had "WHERE a.x =
3378 : : * b.y and a.x = 42", it is worth considering a join between a and b,
3379 : : * since the join result is likely to be small even though it'll end
3380 : : * up being an unqualified nestloop.
3381 : : */
3382 : :
3383 : 145416 : return true;
3384 : : }
3385 : :
3386 : 25034 : return false;
3387 : : }
3388 : :
3389 : :
3390 : : /*
3391 : : * has_relevant_eclass_joinclause
3392 : : * Detect whether there is an EquivalenceClass that could produce
3393 : : * a joinclause involving the given relation and anything else.
3394 : : *
3395 : : * This is the same as have_relevant_eclass_joinclause with the other rel
3396 : : * implicitly defined as "everything else in the query".
3397 : : */
3398 : : bool
3399 : 179788 : has_relevant_eclass_joinclause(PlannerInfo *root, RelOptInfo *rel1)
3400 : : {
3401 : : Bitmapset *matched_ecs;
3402 : : int i;
3403 : :
3404 : : /* Examine only eclasses mentioning rel1 */
3405 : 179788 : matched_ecs = get_eclass_indexes_for_relids(root, rel1->relids);
3406 : :
3407 : 179788 : i = -1;
3408 [ + + ]: 609563 : while ((i = bms_next_member(matched_ecs, i)) >= 0)
3409 : : {
3410 : 501225 : EquivalenceClass *ec = (EquivalenceClass *) list_nth(root->eq_classes,
3411 : : i);
3412 : :
3413 : : /*
3414 : : * Won't generate joinclauses if single-member (this test covers the
3415 : : * volatile case too)
3416 : : */
3417 [ + + ]: 501225 : if (list_length(ec->ec_members) <= 1)
3418 : 218591 : continue;
3419 : :
3420 : : /*
3421 : : * Per the comment in have_relevant_eclass_joinclause, it's sufficient
3422 : : * to find an EC that mentions both this rel and some other rel.
3423 : : */
3424 [ + + ]: 282634 : if (!bms_is_subset(ec->ec_relids, rel1->relids))
3425 : 71450 : return true;
3426 : : }
3427 : :
3428 : 108338 : return false;
3429 : : }
3430 : :
3431 : :
3432 : : /*
3433 : : * eclass_useful_for_merging
3434 : : * Detect whether the EC could produce any mergejoinable join clauses
3435 : : * against the specified relation.
3436 : : *
3437 : : * This is just a heuristic test and doesn't have to be exact; it's better
3438 : : * to say "yes" incorrectly than "no". Hence we don't bother with details
3439 : : * like whether the lack of a cross-type operator might prevent the clause
3440 : : * from actually being generated.
3441 : : */
3442 : : bool
3443 : 630935 : eclass_useful_for_merging(PlannerInfo *root,
3444 : : EquivalenceClass *eclass,
3445 : : RelOptInfo *rel)
3446 : : {
3447 : : Relids relids;
3448 : : ListCell *lc;
3449 : :
3450 : : Assert(!eclass->ec_merged);
3451 : :
3452 : : /*
3453 : : * Won't generate joinclauses if const or single-member (the latter test
3454 : : * covers the volatile case too)
3455 : : */
3456 [ + + + + ]: 630935 : if (eclass->ec_has_const || list_length(eclass->ec_members) <= 1)
3457 : 30944 : return false;
3458 : :
3459 : : /*
3460 : : * Note we don't test ec_broken; if we did, we'd need a separate code path
3461 : : * to look through ec_sources. Checking the members anyway is OK as a
3462 : : * possibly-overoptimistic heuristic.
3463 : : */
3464 : :
3465 : : /* If specified rel is a child, we must consider the topmost parent rel */
3466 [ + + + + : 599991 : if (IS_OTHER_REL(rel))
- + ]
3467 : : {
3468 : : Assert(!bms_is_empty(rel->top_parent_relids));
3469 : 4692 : relids = rel->top_parent_relids;
3470 : : }
3471 : : else
3472 : 595299 : relids = rel->relids;
3473 : :
3474 : : /* If rel already includes all members of eclass, no point in searching */
3475 [ + + ]: 599991 : if (bms_is_subset(eclass->ec_relids, relids))
3476 : 227094 : return false;
3477 : :
3478 : : /*
3479 : : * To join, we need a member not in the given rel. Ignore children here.
3480 : : */
3481 [ + - + + : 587957 : foreach(lc, eclass->ec_members)
+ + ]
3482 : : {
3483 : 587432 : EquivalenceMember *cur_em = (EquivalenceMember *) lfirst(lc);
3484 : :
3485 : : /* Child members should not exist in ec_members */
3486 : : Assert(!cur_em->em_is_child);
3487 : :
3488 [ + + ]: 587432 : if (!bms_overlap(cur_em->em_relids, relids))
3489 : 372372 : return true;
3490 : : }
3491 : :
3492 : 525 : return false;
3493 : : }
3494 : :
3495 : :
3496 : : /*
3497 : : * is_redundant_derived_clause
3498 : : * Test whether rinfo is derived from same EC as any clause in clauselist;
3499 : : * if so, it can be presumed to represent a condition that's redundant
3500 : : * with that member of the list.
3501 : : */
3502 : : bool
3503 : 70 : is_redundant_derived_clause(RestrictInfo *rinfo, List *clauselist)
3504 : : {
3505 : 70 : EquivalenceClass *parent_ec = rinfo->parent_ec;
3506 : : ListCell *lc;
3507 : :
3508 : : /* Fail if it's not a potentially-redundant clause from some EC */
3509 [ + - ]: 70 : if (parent_ec == NULL)
3510 : 70 : return false;
3511 : :
3512 [ # # # # : 0 : foreach(lc, clauselist)
# # ]
3513 : : {
3514 : 0 : RestrictInfo *otherrinfo = (RestrictInfo *) lfirst(lc);
3515 : :
3516 [ # # ]: 0 : if (otherrinfo->parent_ec == parent_ec)
3517 : 0 : return true;
3518 : : }
3519 : :
3520 : 0 : return false;
3521 : : }
3522 : :
3523 : : /*
3524 : : * is_redundant_with_indexclauses
3525 : : * Test whether rinfo is redundant with any clause in the IndexClause
3526 : : * list. Here, for convenience, we test both simple identity and
3527 : : * whether it is derived from the same EC as any member of the list.
3528 : : */
3529 : : bool
3530 : 1158875 : is_redundant_with_indexclauses(RestrictInfo *rinfo, List *indexclauses)
3531 : : {
3532 : 1158875 : EquivalenceClass *parent_ec = rinfo->parent_ec;
3533 : : ListCell *lc;
3534 : :
3535 [ + + + + : 1612322 : foreach(lc, indexclauses)
+ + ]
3536 : : {
3537 : 1173728 : IndexClause *iclause = lfirst_node(IndexClause, lc);
3538 : 1173728 : RestrictInfo *otherrinfo = iclause->rinfo;
3539 : :
3540 : : /* If indexclause is lossy, it won't enforce the condition exactly */
3541 [ + + ]: 1173728 : if (iclause->lossy)
3542 : 33284 : continue;
3543 : :
3544 : : /* Match if it's same clause (pointer equality should be enough) */
3545 [ + + ]: 1140444 : if (rinfo == otherrinfo)
3546 : 720281 : return true;
3547 : : /* Match if derived from same EC */
3548 [ + + + + ]: 420490 : if (parent_ec && otherrinfo->parent_ec == parent_ec)
3549 : 327 : return true;
3550 : :
3551 : : /*
3552 : : * No need to look at the derived clauses in iclause->indexquals; they
3553 : : * couldn't match if the parent clause didn't.
3554 : : */
3555 : : }
3556 : :
3557 : 438594 : return false;
3558 : : }
3559 : :
3560 : : /*
3561 : : * get_eclass_indexes_for_relids
3562 : : * Build and return a Bitmapset containing the indexes into root's
3563 : : * eq_classes list for all eclasses that mention any of these relids
3564 : : */
3565 : : static Bitmapset *
3566 : 952211 : get_eclass_indexes_for_relids(PlannerInfo *root, Relids relids)
3567 : : {
3568 : 952211 : Bitmapset *ec_indexes = NULL;
3569 : 952211 : int i = -1;
3570 : :
3571 : : /* Should be OK to rely on eclass_indexes */
3572 : : Assert(root->ec_merging_done);
3573 : :
3574 [ + + ]: 2935830 : while ((i = bms_next_member(relids, i)) > 0)
3575 : : {
3576 : 1983619 : RelOptInfo *rel = root->simple_rel_array[i];
3577 : :
3578 : : /* ignore the RTE_GROUP RTE */
3579 [ - + ]: 1983619 : if (i == root->group_rtindex)
3580 : 0 : continue;
3581 : :
3582 [ + + ]: 1983619 : if (rel == NULL) /* must be an outer join */
3583 : : {
3584 : : Assert(bms_is_member(i, root->outer_join_rels));
3585 : 215649 : continue;
3586 : : }
3587 : :
3588 : 1767970 : ec_indexes = bms_add_members(ec_indexes, rel->eclass_indexes);
3589 : : }
3590 : 952211 : return ec_indexes;
3591 : : }
3592 : :
3593 : : /*
3594 : : * get_common_eclass_indexes
3595 : : * Build and return a Bitmapset containing the indexes into root's
3596 : : * eq_classes list for all eclasses that mention rels in both
3597 : : * relids1 and relids2.
3598 : : */
3599 : : static Bitmapset *
3600 : 559091 : get_common_eclass_indexes(PlannerInfo *root, Relids relids1, Relids relids2)
3601 : : {
3602 : : Bitmapset *rel1ecs;
3603 : : Bitmapset *rel2ecs;
3604 : : int relid;
3605 : :
3606 : 559091 : rel1ecs = get_eclass_indexes_for_relids(root, relids1);
3607 : :
3608 : : /*
3609 : : * We can get away with just using the relation's eclass_indexes directly
3610 : : * when relids2 is a singleton set.
3611 : : */
3612 [ + + ]: 559091 : if (bms_get_singleton_member(relids2, &relid))
3613 : 423649 : rel2ecs = root->simple_rel_array[relid]->eclass_indexes;
3614 : : else
3615 : 135442 : rel2ecs = get_eclass_indexes_for_relids(root, relids2);
3616 : :
3617 : : /* Calculate and return the common EC indexes, recycling the left input. */
3618 : 559091 : return bms_int_members(rel1ecs, rel2ecs);
3619 : : }
3620 : :
3621 : : /*
3622 : : * ec_build_derives_hash
3623 : : * Construct the auxiliary hash table for derived clause lookups.
3624 : : */
3625 : : static void
3626 : 0 : ec_build_derives_hash(PlannerInfo *root, EquivalenceClass *ec)
3627 : : {
3628 : : Assert(!ec->ec_derives_hash);
3629 : :
3630 : : /*
3631 : : * Create the hash table.
3632 : : *
3633 : : * We pass list_length(ec->ec_derives_list) as the initial size.
3634 : : * Simplehash will divide this by the fillfactor (typically 0.9) and round
3635 : : * up to the next power of two, so this will usually give us at least 64
3636 : : * buckets around the threshold. That avoids immediate resizing without
3637 : : * hardcoding a specific size.
3638 : : */
3639 : 0 : ec->ec_derives_hash = derives_create(root->planner_cxt,
3640 : 0 : list_length(ec->ec_derives_list),
3641 : : NULL);
3642 : :
3643 [ # # # # : 0 : foreach_node(RestrictInfo, rinfo, ec->ec_derives_list)
# # ]
3644 : 0 : ec_add_clause_to_derives_hash(ec, rinfo);
3645 : 0 : }
3646 : :
3647 : : /*
3648 : : * ec_add_derived_clause
3649 : : * Add a clause to the set of derived clauses for the given
3650 : : * EquivalenceClass. Always appends to ec_derives_list; also adds
3651 : : * to ec_derives_hash if it exists.
3652 : : *
3653 : : * Also asserts expected invariants of derived clauses.
3654 : : */
3655 : : static void
3656 : 94578 : ec_add_derived_clause(EquivalenceClass *ec, RestrictInfo *clause)
3657 : : {
3658 : : /*
3659 : : * Constant, if present, is always placed on the RHS; see
3660 : : * generate_base_implied_equalities_const(). LHS is never a constant.
3661 : : */
3662 : : Assert(!clause->left_em->em_is_const);
3663 : :
3664 : : /*
3665 : : * Clauses containing a constant are never considered redundant, so
3666 : : * parent_ec is not set.
3667 : : */
3668 : : Assert(!clause->parent_ec || !clause->right_em->em_is_const);
3669 : :
3670 : 94578 : ec->ec_derives_list = lappend(ec->ec_derives_list, clause);
3671 [ - + ]: 94578 : if (ec->ec_derives_hash)
3672 : 0 : ec_add_clause_to_derives_hash(ec, clause);
3673 : 94578 : }
3674 : :
3675 : : /*
3676 : : * ec_add_derived_clauses
3677 : : * Add a list of clauses to the set of clauses derived from the given
3678 : : * EquivalenceClass; adding to the list and hash table if needed.
3679 : : *
3680 : : * This function is similar to ec_add_derived_clause() but optimized for adding
3681 : : * multiple clauses at a time to the ec_derives_list. The assertions from
3682 : : * ec_add_derived_clause() are not repeated here, as the input clauses are
3683 : : * assumed to have already been validated.
3684 : : */
3685 : : static void
3686 : 33 : ec_add_derived_clauses(EquivalenceClass *ec, List *clauses)
3687 : : {
3688 : 33 : ec->ec_derives_list = list_concat(ec->ec_derives_list, clauses);
3689 [ - + ]: 33 : if (ec->ec_derives_hash)
3690 [ # # # # : 0 : foreach_node(RestrictInfo, rinfo, clauses)
# # ]
3691 : 0 : ec_add_clause_to_derives_hash(ec, rinfo);
3692 : 33 : }
3693 : :
3694 : : /*
3695 : : * fill_ec_derives_key
3696 : : * Compute a canonical key for ec_derives_hash lookup or insertion.
3697 : : *
3698 : : * Derived clauses are looked up using a pair of EquivalenceMembers and a
3699 : : * parent EquivalenceClass. To avoid storing or searching for both EM orderings,
3700 : : * we canonicalize the key:
3701 : : *
3702 : : * - For clauses involving two non-constant EMs, em1 is set to the EM with lower
3703 : : * memory address and em2 is set to the other one.
3704 : : * - For clauses involving a constant EM, the caller must pass the non-constant
3705 : : * EM as leftem and NULL as rightem; we then set em1 = NULL and em2 = leftem.
3706 : : */
3707 : : static inline void
3708 : 0 : fill_ec_derives_key(ECDerivesKey *key,
3709 : : EquivalenceMember *leftem,
3710 : : EquivalenceMember *rightem,
3711 : : EquivalenceClass *parent_ec)
3712 : : {
3713 : : Assert(leftem); /* Always required for lookup or insertion */
3714 : :
3715 [ # # ]: 0 : if (rightem == NULL)
3716 : : {
3717 : 0 : key->em1 = NULL;
3718 : 0 : key->em2 = leftem;
3719 : : }
3720 [ # # ]: 0 : else if (leftem < rightem)
3721 : : {
3722 : 0 : key->em1 = leftem;
3723 : 0 : key->em2 = rightem;
3724 : : }
3725 : : else
3726 : : {
3727 : 0 : key->em1 = rightem;
3728 : 0 : key->em2 = leftem;
3729 : : }
3730 : 0 : key->parent_ec = parent_ec;
3731 : 0 : }
3732 : :
3733 : : /*
3734 : : * ec_add_clause_to_derives_hash
3735 : : * Add a derived clause to ec_derives_hash in the given EquivalenceClass.
3736 : : *
3737 : : * Each clause is associated with a canonicalized key. For constant-containing
3738 : : * clauses, only the non-constant EM is used for lookup; see comments in
3739 : : * fill_ec_derives_key().
3740 : : */
3741 : : static void
3742 : 0 : ec_add_clause_to_derives_hash(EquivalenceClass *ec, RestrictInfo *rinfo)
3743 : : {
3744 : : ECDerivesKey key;
3745 : : ECDerivesEntry *entry;
3746 : : bool found;
3747 : :
3748 : : /*
3749 : : * Constants are always placed on the RHS; see
3750 : : * generate_base_implied_equalities_const().
3751 : : */
3752 : : Assert(!rinfo->left_em->em_is_const);
3753 : :
3754 : : /*
3755 : : * Clauses containing a constant are never considered redundant, so
3756 : : * parent_ec is not set.
3757 : : */
3758 : : Assert(!rinfo->parent_ec || !rinfo->right_em->em_is_const);
3759 : :
3760 : : /*
3761 : : * See fill_ec_derives_key() for details: we use a canonicalized key to
3762 : : * avoid storing both EM orderings. For constant EMs, only the
3763 : : * non-constant EM is included in the key.
3764 : : */
3765 : 0 : fill_ec_derives_key(&key,
3766 : : rinfo->left_em,
3767 [ # # ]: 0 : rinfo->right_em->em_is_const ? NULL : rinfo->right_em,
3768 : : rinfo->parent_ec);
3769 : 0 : entry = derives_insert(ec->ec_derives_hash, key, &found);
3770 : : Assert(!found);
3771 : 0 : entry->rinfo = rinfo;
3772 : 0 : }
3773 : :
3774 : : /*
3775 : : * ec_clear_derived_clauses
3776 : : * Reset ec_derives_list and ec_derives_hash.
3777 : : *
3778 : : * We destroy the hash table explicitly, since it may consume significant
3779 : : * space. The list holds the same set of entries and can become equally large
3780 : : * when thousands of partitions are involved, so we free it as well -- even
3781 : : * though we do not typically free lists.
3782 : : */
3783 : : static void
3784 : 33 : ec_clear_derived_clauses(EquivalenceClass *ec)
3785 : : {
3786 : 33 : list_free(ec->ec_derives_list);
3787 : 33 : ec->ec_derives_list = NIL;
3788 : :
3789 [ - + ]: 33 : if (ec->ec_derives_hash)
3790 : : {
3791 : 0 : derives_destroy(ec->ec_derives_hash);
3792 : 0 : ec->ec_derives_hash = NULL;
3793 : : }
3794 : 33 : }
3795 : :
3796 : : /*
3797 : : * ec_search_clause_for_ems
3798 : : * Search for an existing RestrictInfo that equates the given pair
3799 : : * of EquivalenceMembers, either from ec_sources or ec_derives.
3800 : : *
3801 : : * Returns a clause with matching operands in either given order or commuted
3802 : : * order. We used to require matching operator OIDs, but dropped that since any
3803 : : * semantically different operator here would indicate a broken operator family.
3804 : : *
3805 : : * Returns NULL if no matching clause is found.
3806 : : */
3807 : : static RestrictInfo *
3808 : 415274 : ec_search_clause_for_ems(PlannerInfo *root, EquivalenceClass *ec,
3809 : : EquivalenceMember *leftem, EquivalenceMember *rightem,
3810 : : EquivalenceClass *parent_ec)
3811 : : {
3812 : : /* Check original source clauses */
3813 [ + - + + : 1315656 : foreach_node(RestrictInfo, rinfo, ec->ec_sources)
+ + ]
3814 : : {
3815 [ + + ]: 487690 : if (rinfo->left_em == leftem &&
3816 [ + + ]: 231552 : rinfo->right_em == rightem &&
3817 [ + + ]: 205002 : rinfo->parent_ec == parent_ec)
3818 : 1291 : return rinfo;
3819 [ + + ]: 487600 : if (rinfo->left_em == rightem &&
3820 [ + + ]: 210259 : rinfo->right_em == leftem &&
3821 [ + + ]: 187463 : rinfo->parent_ec == parent_ec)
3822 : 1201 : return rinfo;
3823 : : }
3824 : :
3825 : : /* Not found in ec_sources; search derived clauses */
3826 : 413983 : return ec_search_derived_clause_for_ems(root, ec, leftem, rightem,
3827 : : parent_ec);
3828 : : }
3829 : :
3830 : : /*
3831 : : * ec_search_derived_clause_for_ems
3832 : : * Search for an existing derived clause between two EquivalenceMembers.
3833 : : *
3834 : : * If the number of derived clauses exceeds a threshold, switch to hash table
3835 : : * lookup; otherwise, scan ec_derives_list linearly.
3836 : : *
3837 : : * Clauses involving constants are looked up by passing the non-constant EM
3838 : : * as leftem and setting rightem to NULL. In that case, we expect to find a
3839 : : * clause with a constant on the RHS.
3840 : : *
3841 : : * While searching the list, we compare each given EM with both sides of each
3842 : : * clause. But for hash table lookups, we construct a canonicalized key and
3843 : : * perform a single lookup.
3844 : : */
3845 : : static RestrictInfo *
3846 : 413988 : ec_search_derived_clause_for_ems(PlannerInfo *root, EquivalenceClass *ec,
3847 : : EquivalenceMember *leftem,
3848 : : EquivalenceMember *rightem,
3849 : : EquivalenceClass *parent_ec)
3850 : : {
3851 : : /* Switch to using hash lookup when list grows "too long". */
3852 [ + - - + ]: 827976 : if (!ec->ec_derives_hash &&
3853 : 413988 : list_length(ec->ec_derives_list) >= EC_DERIVES_HASH_THRESHOLD)
3854 : 0 : ec_build_derives_hash(root, ec);
3855 : :
3856 : : /* Perform hash table lookup if available */
3857 [ - + ]: 413988 : if (ec->ec_derives_hash)
3858 : : {
3859 : : ECDerivesKey key;
3860 : : RestrictInfo *rinfo;
3861 : : ECDerivesEntry *entry;
3862 : :
3863 : 0 : fill_ec_derives_key(&key, leftem, rightem, parent_ec);
3864 : 0 : entry = derives_lookup(ec->ec_derives_hash, key);
3865 [ # # ]: 0 : if (entry)
3866 : : {
3867 : 0 : rinfo = entry->rinfo;
3868 : : Assert(rinfo);
3869 : : Assert(rightem || rinfo->right_em->em_is_const);
3870 : 0 : return rinfo;
3871 : : }
3872 : : }
3873 : : else
3874 : : {
3875 : : /* Fallback to linear search over ec_derives_list */
3876 [ + + + + : 591326 : foreach_node(RestrictInfo, rinfo, ec->ec_derives_list)
+ + ]
3877 : : {
3878 : : /* Handle special case: lookup by non-const EM alone */
3879 [ + + ]: 452186 : if (!rightem &&
3880 [ + - ]: 5 : rinfo->left_em == leftem)
3881 : : {
3882 : : Assert(rinfo->right_em->em_is_const);
3883 : 344418 : return rinfo;
3884 : : }
3885 [ + + ]: 452181 : if (rinfo->left_em == leftem &&
3886 [ + + ]: 185400 : rinfo->right_em == rightem &&
3887 [ + + ]: 168044 : rinfo->parent_ec == parent_ec)
3888 : 168034 : return rinfo;
3889 [ + + ]: 284147 : if (rinfo->left_em == rightem &&
3890 [ + + ]: 186505 : rinfo->right_em == leftem &&
3891 [ + - ]: 176379 : rinfo->parent_ec == parent_ec)
3892 : 176379 : return rinfo;
3893 : : }
3894 : : }
3895 : :
3896 : 69570 : return NULL;
3897 : : }
|
