public inbox for [email protected]help / color / mirror / Atom feed
[PATCH] Remove self-joins. 2+ messages / 2 participants [nested] [flat]
* [PATCH] Remove self-joins. @ 2021-01-11 04:01 Andrey Lepikhov <[email protected]> 0 siblings, 0 replies; 2+ messages in thread From: Andrey Lepikhov @ 2021-01-11 04:01 UTC (permalink / raw) Remove inner joins of a relation to itself if can be proven that such join can be replaced with a scan. We can build the required proofs of uniqueness using the existing innerrel_is_unique machinery. We can remove a self-join when for each outer row, if: 1. At most one inner row matches the join clauses. 2. If the join target list contains any inner vars then the inner row is (physically) same row as the outer one. In this patch we use Rowley's [1] approach to identify a self-join: 1. Collect all mergejoinable join quals looks like a.x = b.x 2. Collect all another join quals. 3. Check innerrel_is_unique() for the qual list from (1). If it returns true, then outer row matches only the same row from the inner relation. Proved, that this join is self-join and can be replaced by a scan. Some regression tests changed due to self-join removal logic. [1] https://www.postgresql.org/message-id/raw/CAApHDvpggnFMC4yP-jUO7PKN%3DfXeErW5bOxisvJ0HvkHQEY%3DWw%40... --- src/backend/optimizer/plan/analyzejoins.c | 1186 +++++++++++++++++++++ src/backend/optimizer/plan/planmain.c | 5 + src/backend/optimizer/util/relnode.c | 26 +- src/backend/utils/misc/guc.c | 10 + src/include/optimizer/pathnode.h | 4 + src/include/optimizer/planmain.h | 2 + src/test/regress/expected/equivclass.out | 32 + src/test/regress/expected/join.out | 331 ++++++ src/test/regress/expected/sysviews.out | 3 +- src/test/regress/sql/equivclass.sql | 16 + src/test/regress/sql/join.sql | 166 +++ 11 files changed, 1765 insertions(+), 16 deletions(-) diff --git a/src/backend/optimizer/plan/analyzejoins.c b/src/backend/optimizer/plan/analyzejoins.c index 90460a69bd..d631e95f89 100644 --- a/src/backend/optimizer/plan/analyzejoins.c +++ b/src/backend/optimizer/plan/analyzejoins.c @@ -22,6 +22,7 @@ */ #include "postgres.h" +#include "catalog/pg_class.h" #include "nodes/nodeFuncs.h" #include "optimizer/clauses.h" #include "optimizer/joininfo.h" @@ -29,8 +30,12 @@ #include "optimizer/pathnode.h" #include "optimizer/paths.h" #include "optimizer/planmain.h" +#include "optimizer/restrictinfo.h" #include "optimizer/tlist.h" #include "utils/lsyscache.h" +#include "utils/memutils.h" + +bool enable_self_join_removal; /* local functions */ static bool join_is_removable(PlannerInfo *root, SpecialJoinInfo *sjinfo); @@ -47,6 +52,7 @@ static bool is_innerrel_unique_for(PlannerInfo *root, RelOptInfo *innerrel, JoinType jointype, List *restrictlist); +static void change_rinfo(RestrictInfo* rinfo, Index from, Index to); /* @@ -1118,3 +1124,1183 @@ is_innerrel_unique_for(PlannerInfo *root, /* Let rel_is_distinct_for() do the hard work */ return rel_is_distinct_for(root, innerrel, clause_list); } + +typedef struct +{ + Index oldRelid; + Index newRelid; +} ChangeVarnoContext; + + +static bool +change_varno_walker(Node *node, ChangeVarnoContext *context) +{ + if (node == NULL) + return false; + + if (IsA(node, Var)) + { + Var* var = (Var*)node; + if (var->varno == context->oldRelid) + { + var->varno = context->newRelid; + var->varnosyn = context->newRelid; + var->location = -1; + } + else if (var->varno == context->newRelid) + var->location = -1; + + return false; + } + if (IsA(node, RestrictInfo)) + { + change_rinfo((RestrictInfo*)node, context->oldRelid, context->newRelid); + return false; + } + return expression_tree_walker(node, change_varno_walker, context); +} + +/* + * For all Vars in the expression that have varno = oldRelid, set + * varno = newRelid. + */ +static void +change_varno(Expr *expr, Index oldRelid, Index newRelid) +{ + ChangeVarnoContext context; + + context.oldRelid = oldRelid; + context.newRelid = newRelid; + change_varno_walker((Node *) expr, &context); +} + +/* + * Substitute newId for oldId in relids. + */ +static void +change_relid(Relids *relids, Index oldId, Index newId) +{ + if (bms_is_member(oldId, *relids)) + *relids = bms_add_member(bms_del_member(bms_copy(*relids), oldId), newId); +} + +static void +change_rinfo(RestrictInfo* rinfo, Index from, Index to) +{ + bool is_req_equal = + (rinfo->required_relids == rinfo->clause_relids) ? true : false; + + change_varno(rinfo->clause, from, to); + change_varno(rinfo->orclause, from, to); + change_relid(&rinfo->clause_relids, from, to); + if (is_req_equal) + rinfo->required_relids = rinfo->clause_relids; + else + change_relid(&rinfo->required_relids, from, to); + change_relid(&rinfo->left_relids, from, to); + change_relid(&rinfo->right_relids, from, to); + change_relid(&rinfo->outer_relids, from, to); + change_relid(&rinfo->nullable_relids, from, to); +} + +/* + * Update EC members to point to the remaining relation instead of the removed + * one, removing duplicates. + */ +static void +update_ec_members(EquivalenceClass *ec, Index toRemove, Index toKeep) +{ + int counter; + + for (counter = 0; counter < list_length(ec->ec_members); ) + { + ListCell *cell = list_nth_cell(ec->ec_members, counter); + EquivalenceMember *em = lfirst(cell); + int counter1; + + if (!bms_is_member(toRemove, em->em_relids)) + { + counter++; + continue; + } + + change_relid(&em->em_relids, toRemove, toKeep); + /* We only process inner joins */ + change_varno(em->em_expr, toRemove, toKeep); + + /* + * After we switched the equivalence member to the remaining relation, + * check that it is not the same as the existing member, and if it + * is, delete it. + */ + for (counter1 = 0; counter1 < list_length(ec->ec_members); counter1++) + { + EquivalenceMember *other; + + if (counter1 == counter) + continue; + + other = castNode(EquivalenceMember, list_nth(ec->ec_members, counter1)); + + if (equal(other->em_expr, em->em_expr)) + break; + } + + if (counter1 < list_length(ec->ec_members)) + ec->ec_members = list_delete_cell(ec->ec_members, cell); + else + counter++; + } +} + +/* + * Update EC sources to point to the remaining relation instead of the + * removed one. + */ +static void +update_ec_sources(List **sources, Index toRemove, Index toKeep) +{ + int counter; + int cc=0; + + for (counter = 0; counter < list_length(*sources); ) + { + ListCell *cell = list_nth_cell(*sources, counter); + RestrictInfo *rinfo = castNode(RestrictInfo, lfirst(cell)); + int counter1; + + if (!bms_is_member(toRemove, rinfo->required_relids)) + { + counter++; + continue; + } + + change_varno(rinfo->clause, toRemove, toKeep); + + /* + * After switching the clause to the remaining relation, check it for + * redundancy with existing ones. We don't have to check for + * redundancy with derived clauses, because we've just deleted them. + */ + for (counter1 = 0; counter1 < list_length(*sources); counter1++) + { + RestrictInfo *other; + + if (counter1 == counter) + continue; + + other = castNode(RestrictInfo, list_nth(*sources, counter1)); + if (equal(rinfo->clause, other->clause)) + break; + } + + if (counter1 < list_length(*sources)) + { + *sources = list_delete_cell(*sources, cell); + cc++; + } + else + { + counter++; + + /* We will keep this RestrictInfo, correct its relids. */ + change_rinfo(rinfo, toRemove, toKeep); + } + } +} + +/* + * Scratch space for the unique self join removal code. + */ +typedef struct +{ + PlannerInfo *root; + + /* Temporary array for relation ids. */ + Index *relids; + + /* Array of row marks indexed by relid. */ + PlanRowMark **row_marks; + + /* Bitmapset for join relids that is used to avoid reallocation. */ + Relids joinrelids; +} UsjScratch; + +/* + * Remove a relation after we have proven that it participates only in an + * unneeded unique self join. + * + * The joinclauses list is destructively changed. + */ +static void +remove_self_join_rel(UsjScratch *scratch, Relids joinrelids, List *joinclauses, + RelOptInfo *toKeep, RelOptInfo *toRemove) +{ + PlannerInfo *root = scratch->root; + ListCell *cell; + int i; + List *joininfos = NIL; + + /* + * Include all eclass mentions of removed relation into the eclass mentions + * of kept relation. + */ + toKeep->eclass_indexes = bms_add_members(toRemove->eclass_indexes, + toKeep->eclass_indexes); + + /* + * Transfer join and restriction clauses from the removed relation to the + * remaining one. We change the Vars of the clause to point to the + * remaining relation instead of the removed one. The clauses that require + * a subset of joinrelids become restriction clauses of the remaining + * relation, and others remain join clauses. We append them to + * baserestrictinfo and joininfo respectively, trying not to introduce + * duplicates. + * + * We also have to process the 'joinclauses' list here, because it + * contains EC-derived join clauses which must become filter clauses. It + * is not enough to just correct the ECs, because the EC-derived + * restrictions are generated before join removal (see + * generate_base_implied_equalities). + */ + + /* Replace removed relation in joininfo */ + foreach(cell, toKeep->joininfo) + { + RestrictInfo *rinfo = lfirst_node(RestrictInfo, cell); + + change_rinfo(rinfo, toRemove->relid, toKeep->relid); + + /* + * If this clause is a mergejoinable equality clause that compares a + * variable to itself, i.e., has the form of "X=X", replace it with + * null test. + */ + if (rinfo->mergeopfamilies && IsA(rinfo->clause, OpExpr)) + { + Expr *leftOp; + Expr *rightOp; + + leftOp = (Expr *) get_leftop(rinfo->clause); + rightOp = (Expr *) get_rightop(rinfo->clause); + + if (leftOp != NULL && equal(leftOp, rightOp)) + { + NullTest *nullTest = makeNode(NullTest); + nullTest->arg = leftOp; + nullTest->nulltesttype = IS_NOT_NULL; + nullTest->argisrow = false; + nullTest->location = -1; + rinfo->clause = (Expr *) nullTest; + } + } + } + + /* Replace removed relation in joinclauses. + * It is similar with code above but also adds substituted null-tests to + * baserestrict list of kept relation. + * joinclauses contains all restrictions, include general expressions like + * f(a'.x1..a'.xN, a''.x1..a''.xN) op g(a'.x1..a'.xN, a''.x1..a''.xN) + * that couldn't fall into the EC. + */ + foreach(cell, joinclauses) + { + RestrictInfo *rinfo = lfirst_node(RestrictInfo, cell); + + change_rinfo(rinfo, toRemove->relid, toKeep->relid); + + /* + * If this clause is a mergejoinable equality clause that compares a + * variable to itself, i.e., has the form of "X=X", replace it with + * null test. + */ + if (rinfo->mergeopfamilies && IsA(rinfo->clause, OpExpr)) + { + Expr *leftOp; + Expr *rightOp; + ListCell *otherCell; + List **target = &toKeep->baserestrictinfo; + + leftOp = (Expr *) get_leftop(rinfo->clause); + rightOp = (Expr *) get_rightop(rinfo->clause); + + if (leftOp != NULL && equal(leftOp, rightOp)) + { + NullTest *nullTest = makeNode(NullTest); + nullTest->arg = leftOp; + nullTest->nulltesttype = IS_NOT_NULL; + nullTest->argisrow = false; + nullTest->location = -1; + rinfo->clause = (Expr *) nullTest; + } + + foreach(otherCell, *target) + { + RestrictInfo *other = lfirst_node(RestrictInfo, otherCell); + + if (other == rinfo || + (rinfo->parent_ec != NULL + && other->parent_ec == rinfo->parent_ec) + || equal(rinfo->clause, other->clause)) + { + break; + } + } + + if (otherCell != NULL) + continue; + + /* + * XXX: In the case of clauses: + * toKeep.x = toRemove.y and toKeep.y = toRemove.x + * we pass into baserestrictinfo both as: + * toKeep.x = toKeep.y and toKeep.y = toKeep.x + * that is may be not fine, but correct. + */ + toKeep->baserestrictinfo = lappend(toKeep->baserestrictinfo, rinfo); + } + } + + /* Tranfer removed relation baserestrictinfo clauses to kept relation */ + foreach(cell, toRemove->baserestrictinfo) + { + ListCell *otherCell; + RestrictInfo *rinfo = lfirst_node(RestrictInfo, cell); + List **target = &toKeep->baserestrictinfo; + + /* + * Replace the references to the removed relation with references to + * the remaining one. We won't necessarily add this clause, because + * it may be already present in the joininfo or baserestrictinfo. + * Still, we have to switch it to point to the remaining relation. + * This is important for join clauses that reference both relations, + * because they are included in both joininfos. + */ + change_rinfo(rinfo, toRemove->relid, toKeep->relid); + + /* + * Don't add the clause if it is already present in the list, or + * derived from the same equivalence class, or is the same as another + * clause. + */ + foreach (otherCell, *target) + { + RestrictInfo *other = lfirst_node(RestrictInfo, otherCell); + + if (other == rinfo || + (rinfo->parent_ec != NULL + && other->parent_ec == rinfo->parent_ec) + || equal(rinfo->clause, other->clause)) + { + break; + } + } + + /* We can't have an EC-derived clause that joins to some third relation */ + if (otherCell != NULL) + continue; + + /* + * During replacing relids in joininfo some restrictions could change + * semantic from join to baserestrict info, for example: + * a1.x = a2.y => a1.x = a1.y + * If we already have restriction in the baserestrictinfo list: + * a1.x = a1.y, here we will remove duplicates. + */ + target = &toKeep->joininfo; + foreach (otherCell, *target) + { + RestrictInfo *other = lfirst_node(RestrictInfo, otherCell); + + if (equal(rinfo->clause, other->clause)) + break; + } + + if (otherCell != NULL) + /* Duplicate found */ + continue; + + *target = lappend(*target, rinfo); + } + + /* Transfer removed relation join-clauses to kept relation. + * It is similar with code above but also substituted redundant quals with + * null-tests. + */ + foreach(cell, toRemove->joininfo) + { + ListCell *otherCell; + RestrictInfo *rinfo = lfirst_node(RestrictInfo, cell); + List **target = &toKeep->joininfo; + + /* + * Replace the references to the removed relation with references to + * the remaining one. We won't necessarily add this clause, because + * it may be already present in the joininfo or baserestrictinfo. + * Still, we have to switch it to point to the remaining relation. + * This is important for join clauses that reference both relations, + * because they are included in both joininfos. + */ + change_rinfo(rinfo, toRemove->relid, toKeep->relid); + + /* + * Don't add the clause if it is already present in the list, or + * derived from the same equivalence class, or is the same as another + * clause. + */ + foreach (otherCell, *target) + { + RestrictInfo *other = lfirst_node(RestrictInfo, otherCell); + + if (other == rinfo || + (rinfo->parent_ec != NULL + && other->parent_ec == rinfo->parent_ec) + || equal(rinfo->clause, other->clause)) + { + break; + } + } + + if (otherCell != NULL) + continue; + + /* + * If this clause is a mergejoinable equality clause that compares a + * variable to itself, i.e., has the form of "X=X", replace it with + * null test. + */ + if (rinfo->mergeopfamilies && IsA(rinfo->clause, OpExpr)) + { + Expr *leftOp, *rightOp; + + Assert(IsA(rinfo->clause, OpExpr)); + + leftOp = (Expr *) get_leftop(rinfo->clause); + rightOp = (Expr *) get_rightop(rinfo->clause); + + if (leftOp != NULL && equal(leftOp, rightOp)) + { + NullTest *nullTest = makeNode(NullTest); + nullTest->arg = leftOp; + nullTest->nulltesttype = IS_NOT_NULL; + nullTest->argisrow = false; + nullTest->location = -1; + rinfo->clause = (Expr*)nullTest; + } + } + + *target = lappend(*target, rinfo); + toKeep->baserestrict_min_security = Min(toKeep->baserestrict_min_security, rinfo->security_level); + } + + /* + * Now the state of the joininfo list of the toKeep relation is changed: + * some of clauses can be removed, some need to move into baserestrictinfo. + * To do this use the same technique as the remove_rel_from_query() routine. + */ + joininfos = list_copy(toKeep->joininfo); + foreach(cell, joininfos) + { + RestrictInfo *rinfo = (RestrictInfo *) lfirst(cell); + + remove_join_clause_from_rels(root, rinfo, rinfo->required_relids); + + if (RINFO_IS_PUSHED_DOWN(rinfo, joinrelids)) + { + /* Recheck that qual doesn't actually reference the target rel */ + Assert(!bms_is_member(toRemove->relid, rinfo->clause_relids)); + + /* + * The required_relids probably aren't shared with anything else, + * but let's copy them just to be sure. + */ + rinfo->required_relids = bms_copy(rinfo->required_relids); + rinfo->required_relids = bms_del_member(rinfo->required_relids, + toRemove->relid); + distribute_restrictinfo_to_rels(root, rinfo); + } + } + + /* + * Transfer the targetlist and attr_needed flags. + */ + Assert(toRemove->reltarget->sortgrouprefs == 0); + + foreach (cell, toRemove->reltarget->exprs) + { + Expr *node = lfirst(cell); + change_varno(node, toRemove->relid, toKeep->relid); + if (!list_member(toKeep->reltarget->exprs, node)) + toKeep->reltarget->exprs = lappend(toKeep->reltarget->exprs, node); + } + + for (i = toKeep->min_attr; i <= toKeep->max_attr; i++) + { + int attno = i - toKeep->min_attr; + toKeep->attr_needed[attno] = bms_add_members(toKeep->attr_needed[attno], + toRemove->attr_needed[attno]); + } + + /* + * If the removed relation has a row mark, transfer it to the remaining + * one. + * + * If both rels have row marks, just keep the one corresponding to the + * remaining relation, because we verified earlier that they have the same + * strength. + * + * Also make sure that the scratch->row_marks cache is up to date, because + * we are going to use it for further join removals. + */ + if (scratch->row_marks[toRemove->relid]) + { + PlanRowMark **markToRemove = &scratch->row_marks[toRemove->relid]; + PlanRowMark **markToKeep = &scratch->row_marks[toKeep->relid]; + + if (*markToKeep) + { + Assert((*markToKeep)->markType == (*markToRemove)->markType); + + root->rowMarks = list_delete_ptr(root->rowMarks, *markToKeep); + *markToKeep = NULL; + } + else + { + *markToKeep = *markToRemove; + *markToRemove = NULL; + + /* Shouldn't have inheritance children here. */ + Assert((*markToKeep)->rti == (*markToKeep)->prti); + + (*markToKeep)->rti = toKeep->relid; + (*markToKeep)->prti = toKeep->relid; + } + } + + /* + * Likewise replace references in SpecialJoinInfo data structures. + * + * This is relevant in case the join we're deleting is nested inside some + * special joins: the upper joins' relid sets have to be adjusted. + */ + foreach (cell, root->join_info_list) + { + SpecialJoinInfo *sjinfo = (SpecialJoinInfo *) lfirst(cell); + + change_relid(&sjinfo->min_lefthand, toRemove->relid, toKeep->relid); + change_relid(&sjinfo->min_righthand, toRemove->relid, toKeep->relid); + change_relid(&sjinfo->syn_lefthand, toRemove->relid, toKeep->relid); + change_relid(&sjinfo->syn_righthand, toRemove->relid, toKeep->relid); + change_varno((Expr*)sjinfo->semi_rhs_exprs, toRemove->relid, toKeep->relid); + } + + /* + * Likewise update references in PlaceHolderVar data structures. + */ + foreach(cell, root->placeholder_list) + { + PlaceHolderInfo *phinfo = (PlaceHolderInfo *) lfirst(cell); + + /* + * We are at an inner join of two base relations. A placeholder can't + * be needed here or evaluated here. + */ + Assert(!bms_is_subset(phinfo->ph_eval_at, joinrelids)); + Assert(!bms_is_subset(phinfo->ph_needed, joinrelids)); + + change_relid(&phinfo->ph_eval_at, toRemove->relid, toKeep->relid); + change_relid(&phinfo->ph_needed, toRemove->relid, toKeep->relid); + change_relid(&phinfo->ph_lateral, toRemove->relid, toKeep->relid); + change_relid(&phinfo->ph_var->phrels, toRemove->relid, toKeep->relid); + } + + /* + * Update the equivalence classes that reference the removed relations. + */ + foreach(cell, root->eq_classes) + { + EquivalenceClass *ec = lfirst(cell); + + if (!bms_is_member(toRemove->relid, ec->ec_relids)) + { + /* + * This EC doesn't reference the removed relation, nothing to be + * done for it. + */ + continue; + } + + /* + * Update the EC members to reference the remaining relation instead + * of the removed one. + */ + update_ec_members(ec, toRemove->relid, toKeep->relid); + change_relid(&ec->ec_relids, toRemove->relid, toKeep->relid); + + /* + * We will now update source and derived clauses of the EC. + * + * Restriction clauses for base relations are already distributed to + * the respective baserestrictinfo lists (see + * generate_implied_equalities). The above code has already processed + * this list, and updated these clauses to reference the remaining + * relation, so we can skip them here based on their relids. + * + * Likewise, we have already processed the join clauses that join the + * removed relation to the remaining one. + * + * Finally, there are join clauses that join the removed relation to + * some third relation. We can't just delete the source clauses and + * regenerate them from the EC, because the corresponding equality + * operators might be missing (see the handling of ec_broken). + * Therefore, we will update the references in the source clauses. + * + * Derived clauses can be generated again, so it is simpler to just + * delete them. + */ + list_free(ec->ec_derives); + ec->ec_derives = NULL; + update_ec_sources(&ec->ec_sources, toRemove->relid, toKeep->relid); + } + + /* + * Mark the rel as "dead" to show it is no longer part of the join tree. + * (Removing it from the baserel array altogether seems too risky.) + */ + toRemove->reloptkind = RELOPT_DEADREL; + + /* + * Update references to the removed relation from other baserels. + */ + for (i = 1; i < root->simple_rel_array_size; i++) + { + RelOptInfo *otherrel = root->simple_rel_array[i]; + int attroff; + + /* no point in processing target rel itself */ + if (i == toRemove->relid) + continue; + + /* there may be empty slots corresponding to non-baserel RTEs */ + if (otherrel == NULL) + continue; + + Assert(otherrel->relid == i); /* sanity check on array */ + + /* Update attr_needed arrays. */ + for (attroff = otherrel->max_attr - otherrel->min_attr; + attroff >= 0; attroff--) + { + change_relid(&otherrel->attr_needed[attroff], toRemove->relid, + toKeep->relid); + } + + /* Update lateral references. */ + change_varno((Expr*)otherrel->lateral_vars, toRemove->relid, toKeep->relid); + } + + /* + * Change varno in some special cases with non-trivial RangeTblEntry + */ + foreach(cell, root->parse->rtable) + { + RangeTblEntry *rte = lfirst(cell); + + switch(rte->rtekind) + { + case RTE_FUNCTION: + change_varno((Expr*)rte->functions, toRemove->relid, toKeep->relid); + break; + case RTE_TABLEFUNC: + change_varno((Expr*)rte->tablefunc, toRemove->relid, toKeep->relid); + break; + case RTE_VALUES: + change_varno((Expr*)rte->values_lists, toRemove->relid, toKeep->relid); + break; + default: + /* no op */ + break; + } + } +} + +/* + * split_selfjoin_quals + * Processes 'joinquals' building two lists, one with a list of quals + * where the columns/exprs on either side of the join match and another + * list containing the remaining quals. + * + * 'joinquals' must only contain quals for a RTE_RELATION being joined to + * itself. + */ +static void +split_selfjoin_quals(PlannerInfo *root, List *joinquals, List **selfjoinquals, + List **otherjoinquals) +{ + ListCell *lc; + List *sjoinquals = NIL; + List *rjoinquals = NIL; + + foreach(lc, joinquals) + { + RestrictInfo *rinfo = (RestrictInfo *) lfirst(lc); + OpExpr *expr; + Expr *leftexpr; + Expr *rightexpr; + + if (bms_num_members(rinfo->clause_relids) != 2 || + bms_num_members(rinfo->left_relids) != 1 || + bms_num_members(rinfo->right_relids) != 1) + { + rjoinquals = lappend(rjoinquals, rinfo); + continue; + } + + expr = (OpExpr *) rinfo->clause; + + if (!IsA(expr, OpExpr) || list_length(expr->args) != 2) + { + rjoinquals = lappend(rjoinquals, rinfo); + continue; + } + + leftexpr = (Expr *) copyObject(get_leftop(rinfo->clause)); + rightexpr = (Expr *) copyObject(get_rightop(rinfo->clause)); + + /* Can't match of the exprs are not of the same type */ + if (leftexpr->type != rightexpr->type) + { + rjoinquals = lappend(rjoinquals, rinfo); + continue; + } + + change_varno(rightexpr, + bms_next_member(rinfo->right_relids, -1), + bms_next_member(rinfo->left_relids, -1)); + + if (equal(leftexpr, rightexpr)) + sjoinquals = lappend(sjoinquals, rinfo); + else + rjoinquals = lappend(rjoinquals, rinfo); + } + + *selfjoinquals = sjoinquals; + *otherjoinquals = rjoinquals; +} + +static bool +tlist_contains_rel_exprs(PlannerInfo *root, Relids relids, RelOptInfo *rel) +{ + ListCell *vars; + + foreach(vars, rel->reltarget->exprs) + { + Var *var = (Var *) lfirst(vars); + RelOptInfo *baserel; + int ndx; + + /* + * Ignore PlaceHolderVars in the input tlists; we'll make our own + * decisions about whether to copy them. + */ + if (IsA(var, PlaceHolderVar)) + return true; + + /* + * Otherwise, anything in a baserel or joinrel targetlist ought to be + * a Var. (More general cases can only appear in appendrel child + * rels, which will never be seen here.) + */ + if (!IsA(var, Var)) + elog(ERROR, "unexpected node type in rel targetlist: %d", + (int) nodeTag(var)); + + /* Get the Var's original base rel */ + baserel = find_base_rel(root, var->varno); + + /* Is it still needed above this joinrel? */ + ndx = var->varattno - baserel->min_attr; + if (bms_nonempty_difference(baserel->attr_needed[ndx], relids)) + return true; + } + return false; +} + +/* + * Find and remove unique self joins in a group of base relations that have + * the same Oid. + * + * Returns a set of relids that were removed. + */ +static Relids +remove_self_joins_one_group(UsjScratch *scratch, Index *relids, int n) +{ + PlannerInfo *root = scratch->root; + Relids joinrelids = scratch->joinrelids; + Relids result = NULL; + int i, o; + + if (n < 2) + return NULL; + + for (o = 0; o < n; o++) + { + RelOptInfo *outer = root->simple_rel_array[relids[o]]; + + for (i = o + 1; i < n; i++) + { + RelOptInfo *inner = root->simple_rel_array[relids[i]]; + List *restrictlist; + List *selfjoinquals; + List *otherjoinquals; + ListCell *lc; + bool jinfo_check = true; + + /* A sanity check: the relations have the same Oid. */ + Assert(root->simple_rte_array[relids[i]]->relid + == root->simple_rte_array[relids[o]]->relid); + + /* + * It is impossible to optimize two relations if they belong to + * different rules of order restriction. Otherwise planner can't + * be able to find any variants of correct query plan. + */ + foreach(lc, root->join_info_list) + { + SpecialJoinInfo *info = (SpecialJoinInfo *) lfirst(lc); + + if (bms_is_member(relids[i], info->syn_lefthand) && + !bms_is_member(relids[o], info->syn_lefthand)) + jinfo_check = false; + + if (bms_is_member(relids[i], info->syn_righthand) && + !bms_is_member(relids[o], info->syn_righthand)) + jinfo_check = false; + + if (bms_is_member(relids[o], info->syn_lefthand) && + !bms_is_member(relids[i], info->syn_lefthand)) + jinfo_check = false; + + if (bms_is_member(relids[o], info->syn_righthand) && + !bms_is_member(relids[i], info->syn_righthand)) + jinfo_check = false; + + if (!jinfo_check) + break; + } + + if (!jinfo_check) + continue; + + /* Reuse joinrelids bitset to avoid reallocation. */ + joinrelids = bms_del_members(joinrelids, joinrelids); + + /* + * We only deal with base rels here, so their relids bitset + * contains only one member -- their relid. + */ + joinrelids = bms_add_member(joinrelids, relids[o]); + joinrelids = bms_add_member(joinrelids, relids[i]); + + /* Is it a unique self join? */ + restrictlist = build_joinrel_restrictlist(root, joinrelids, outer, + inner); + + /* + * Process restrictlist to seperate out the self join quals from + * the other quals. e.g x = x goes to selfjoinquals and a = b to + * otherjoinquals. + */ + split_selfjoin_quals(root, restrictlist, &selfjoinquals, + &otherjoinquals); + + if (list_length(selfjoinquals) == 0) + { + /* + * Have a chance to remove join if target list contains vars from + * the only one relation. + */ + if (list_length(otherjoinquals) == 0) + { + /* Can't determine uniqueness without any quals. */ + continue; + + } + else if (!tlist_contains_rel_exprs(root, joinrelids, inner)) + { + /* + * TODO: + * In this case, we only have a chance in the case of a + * foreign key reference. + */ + continue; + } + else + /* + * The target list contains vars from both inner and outer + * relations. + */ + continue; + } + /* + * Determine if the inner table can duplicate outer rows. We must + * bypass the unique rel cache here since we're possibly using a + * subset of join quals. We can use 'force_cache' = true when all + * join quals are selfjoin quals. Otherwise we could end up + * putting false negatives in the cache. + */ + else if (!innerrel_is_unique(root, joinrelids, outer->relids, + inner, JOIN_INNER, selfjoinquals, + list_length(otherjoinquals) == 0)) + continue; + + /* + * We can't remove the join if the relations have row marks of + * different strength (e.g. one is locked FOR UPDATE and another + * just has ROW_MARK_REFERENCE for EvalPlanQual rechecking). + */ + if (scratch->row_marks[relids[i]] && scratch->row_marks[relids[o]] + && scratch->row_marks[relids[i]]->markType + != scratch->row_marks[relids[o]]->markType) + { + continue; + } + + /* + * Be safe to do not remove table participated in complicated PH + */ + foreach(lc, root->placeholder_list) + { + PlaceHolderInfo *phinfo = (PlaceHolderInfo *) lfirst(lc); + + /* there isn't any other place to eval PHV */ + if (bms_is_subset(phinfo->ph_eval_at, joinrelids) || + bms_is_subset(phinfo->ph_needed, joinrelids)) + break; + } + + if (lc) + continue; + + /* + * We can remove either relation, so remove the outer one, to + * simplify this loop. + */ + remove_self_join_rel(scratch, joinrelids, restrictlist, inner, outer); + result = bms_add_member(result, relids[o]); + + /* + * Replace varno in root targetlist and HAVING clause. + */ + change_varno((Expr *) root->processed_tlist, relids[o], relids[i]); + change_varno((Expr *) root->parse->havingQual, relids[o], relids[i]); + + /* We removed the outer relation, try the next one. */ + break; + } + } + + scratch->joinrelids = joinrelids; + return result; +} + +/* + * A qsort comparator to sort the relids by the relation Oid. + */ +static int +compare_rte(const Index *left, const Index *right, PlannerInfo *root) +{ + Oid l = root->simple_rte_array[*left]->relid; + Oid r = root->simple_rte_array[*right]->relid; + + return l < r ? 1 : (l == r ? 0 : -1); +} + +/* + * Find and remove unique self joins on a particular level of the join tree. + * + * We sort the relations by Oid and then examine each group with the same Oid. + * If we removed any relation, remove it from joinlist as well. + */ +static Relids +remove_self_joins_one_level(UsjScratch *scratch, List *joinlist, + Relids relidsToRemove) +{ + ListCell *lc; + Oid groupOid; + int groupStart; + int i; + int n = 0; + Index *relid_ascending = scratch->relids; + PlannerInfo *root = scratch->root; + + /* + * Collect the ids of base relations at this level of the join tree. + */ + foreach (lc, joinlist) + { + RangeTblEntry *rte; + RelOptInfo *rel; + RangeTblRef *ref = (RangeTblRef *) lfirst(lc); + if (!IsA(ref, RangeTblRef)) + continue; + + rte = root->simple_rte_array[ref->rtindex]; + rel = root->simple_rel_array[ref->rtindex]; + + /* We only care about base relations from which we select something. */ + if (rte->rtekind != RTE_RELATION || rte->relkind != RELKIND_RELATION + || rel == NULL) + { + continue; + } + + relid_ascending[n++] = ref->rtindex; + + /* + * Limit the number of joins we process to control the quadratic + * behavior. + */ + if (n > join_collapse_limit) + break; + } + + if (n < 2) + return relidsToRemove; + + /* + * Find and process the groups of relations that have same Oid. + */ + qsort_arg(relid_ascending, n, sizeof(*relid_ascending), + (qsort_arg_comparator) compare_rte, root); + groupOid = root->simple_rte_array[relid_ascending[0]]->relid; + groupStart = 0; + for (i = 1; i < n; i++) + { + RangeTblEntry *rte = root->simple_rte_array[relid_ascending[i]]; + Assert(rte->relid != InvalidOid); + if (rte->relid != groupOid) + { + relidsToRemove = bms_add_members(relidsToRemove, + remove_self_joins_one_group(scratch, &relid_ascending[groupStart], + i - groupStart)); + groupOid = rte->relid; + groupStart = i; + } + } + Assert(groupOid != InvalidOid); + Assert(groupStart < n); + relidsToRemove = bms_add_members(relidsToRemove, + remove_self_joins_one_group(scratch, &relid_ascending[groupStart], + n - groupStart)); + + return relidsToRemove; +} + +/* + * Find and remove unique self joins on a single level of a join tree, and + * recurse to handle deeper levels. + */ +static Relids +remove_self_joins_recurse(UsjScratch *scratch, List *joinlist, + Relids relidsToRemove) +{ + ListCell *lc; + + foreach (lc, joinlist) + { + switch (((Node *) lfirst(lc))->type) + { + case T_List: + relidsToRemove = remove_self_joins_recurse( + scratch, + (List *) lfirst(lc), + relidsToRemove); + break; + case T_RangeTblRef: + break; + default: + Assert(false); + } + } + return remove_self_joins_one_level(scratch, joinlist, relidsToRemove); +} + +/* + * Find and remove useless self joins. + * + * We search for joins where the same relation is joined to itself on all + * columns of some unique index. If this condition holds, then, for + * each outer row, only one inner row matches, and it is the same row + * of the same relation. This allows us to remove the join and replace + * it with a scan that combines WHERE clauses from both sides. The join + * clauses themselves assume the form of X = X and can be replaced with + * NOT NULL clauses. + * + * For the sake of simplicity, we don't apply this optimization to special + * joins. Here is a list of what we could do in some particular cases: + * 'a a1 semi join a a2': is reduced to inner by reduce_unique_semijoins, + * and then removed normally. + * 'a a1 anti join a a2': could simplify to a scan with 'outer quals AND + * (IS NULL on join columns OR NOT inner quals)'. + * 'a a1 left join a a2': could simplify to a scan like inner, but without + * NOT NULL conditions on join columns. + * 'a a1 left join (a a2 join b)': can't simplify this, because join to b + * can both remove rows and introduce duplicates. + * + * To search for removable joins, we order all the relations on their Oid, + * go over each set with the same Oid, and consider each pair of relations + * in this set. We check that both relation are made unique by the same + * unique index with the same clauses. + * + * To remove the join, we mark one of the participating relation as + * dead, and rewrite all references to it to point to the remaining + * relation. This includes modifying RestrictInfos, EquivalenceClasses and + * EquivalenceMembers. We also have to modify the row marks. The join clauses + * of the removed relation become either restriction or join clauses, based on + * whether they reference any relations not participating in the removed join. + * + * 'targetlist' is the top-level targetlist of query. If it has any references + * to the removed relations, we update them to point to the remaining ones. + */ +List * +remove_useless_self_joins(PlannerInfo *root, List *joinlist) +{ + ListCell *lc; + UsjScratch scratch; + Relids relidsToRemove = NULL; + int relid = -1; + + if (!enable_self_join_removal) + return joinlist; + + scratch.root = root; + scratch.relids = palloc(root->simple_rel_array_size * sizeof(Index)); + scratch.row_marks = palloc0(root->simple_rel_array_size * sizeof(PlanRowMark *)); + scratch.joinrelids = NULL; + + /* Collect row marks. */ + foreach (lc, root->rowMarks) + { + PlanRowMark *rowMark = (PlanRowMark *) lfirst(lc); + + /* Can't have more than one row mark for a relation. */ + Assert(scratch.row_marks[rowMark->rti] == NULL); + + scratch.row_marks[rowMark->rti] = rowMark; + } + + /* Finally, remove the joins. */ + relidsToRemove = remove_self_joins_recurse(&scratch, + joinlist, + relidsToRemove); + while ((relid = bms_next_member(relidsToRemove, relid)) >= 0) + { + int nremoved = 0; + joinlist = remove_rel_from_joinlist(joinlist, relid, &nremoved); + } + + pfree(scratch.relids); + pfree(scratch.row_marks); + return joinlist; +} diff --git a/src/backend/optimizer/plan/planmain.c b/src/backend/optimizer/plan/planmain.c index e1a13e20c5..444ebdb63c 100644 --- a/src/backend/optimizer/plan/planmain.c +++ b/src/backend/optimizer/plan/planmain.c @@ -226,6 +226,11 @@ query_planner(PlannerInfo *root, */ reduce_unique_semijoins(root); + /* + * Remove self joins on a unique column. + */ + joinlist = remove_useless_self_joins(root, joinlist); + /* * Now distribute "placeholders" to base rels as needed. This has to be * done after join removal because removal could change whether a diff --git a/src/backend/optimizer/util/relnode.c b/src/backend/optimizer/util/relnode.c index 731ff708b9..4f831bb17c 100644 --- a/src/backend/optimizer/util/relnode.c +++ b/src/backend/optimizer/util/relnode.c @@ -40,14 +40,10 @@ typedef struct JoinHashEntry static void build_joinrel_tlist(PlannerInfo *root, RelOptInfo *joinrel, RelOptInfo *input_rel); -static List *build_joinrel_restrictlist(PlannerInfo *root, - RelOptInfo *joinrel, - RelOptInfo *outer_rel, - RelOptInfo *inner_rel); static void build_joinrel_joinlist(RelOptInfo *joinrel, RelOptInfo *outer_rel, RelOptInfo *inner_rel); -static List *subbuild_joinrel_restrictlist(RelOptInfo *joinrel, +static List *subbuild_joinrel_restrictlist(Relids joinrelids, List *joininfo_list, List *new_restrictlist); static List *subbuild_joinrel_joinlist(RelOptInfo *joinrel, @@ -598,7 +594,7 @@ build_join_rel(PlannerInfo *root, */ if (restrictlist_ptr) *restrictlist_ptr = build_joinrel_restrictlist(root, - joinrel, + joinrel->relids, outer_rel, inner_rel); return joinrel; @@ -703,7 +699,7 @@ build_join_rel(PlannerInfo *root, * caller might or might not need the restrictlist, but I need it anyway * for set_joinrel_size_estimates().) */ - restrictlist = build_joinrel_restrictlist(root, joinrel, + restrictlist = build_joinrel_restrictlist(root, joinrel->relids, outer_rel, inner_rel); if (restrictlist_ptr) *restrictlist_ptr = restrictlist; @@ -1037,7 +1033,7 @@ build_joinrel_tlist(PlannerInfo *root, RelOptInfo *joinrel, * the various joinlist entries ultimately refer to RestrictInfos * pushed into them by distribute_restrictinfo_to_rels(). * - * 'joinrel' is a join relation node + * 'joinrelids' is a join relation id set * 'outer_rel' and 'inner_rel' are a pair of relations that can be joined * to form joinrel. * @@ -1050,9 +1046,9 @@ build_joinrel_tlist(PlannerInfo *root, RelOptInfo *joinrel, * RestrictInfo nodes are no longer context-dependent. Instead, just include * the original nodes in the lists made for the join relation. */ -static List * +List * build_joinrel_restrictlist(PlannerInfo *root, - RelOptInfo *joinrel, + Relids joinrelids, RelOptInfo *outer_rel, RelOptInfo *inner_rel) { @@ -1063,8 +1059,8 @@ build_joinrel_restrictlist(PlannerInfo *root, * eliminating any duplicates (important since we will see many of the * same clauses arriving from both input relations). */ - result = subbuild_joinrel_restrictlist(joinrel, outer_rel->joininfo, NIL); - result = subbuild_joinrel_restrictlist(joinrel, inner_rel->joininfo, result); + result = subbuild_joinrel_restrictlist(joinrelids, outer_rel->joininfo, NIL); + result = subbuild_joinrel_restrictlist(joinrelids, inner_rel->joininfo, result); /* * Add on any clauses derived from EquivalenceClasses. These cannot be @@ -1073,7 +1069,7 @@ build_joinrel_restrictlist(PlannerInfo *root, */ result = list_concat(result, generate_join_implied_equalities(root, - joinrel->relids, + joinrelids, outer_rel->relids, inner_rel)); @@ -1099,7 +1095,7 @@ build_joinrel_joinlist(RelOptInfo *joinrel, } static List * -subbuild_joinrel_restrictlist(RelOptInfo *joinrel, +subbuild_joinrel_restrictlist(Relids joinrelids, List *joininfo_list, List *new_restrictlist) { @@ -1109,7 +1105,7 @@ subbuild_joinrel_restrictlist(RelOptInfo *joinrel, { RestrictInfo *rinfo = (RestrictInfo *) lfirst(l); - if (bms_is_subset(rinfo->required_relids, joinrel->relids)) + if (bms_is_subset(rinfo->required_relids, joinrelids)) { /* * This clause becomes a restriction clause for the joinrel, since diff --git a/src/backend/utils/misc/guc.c b/src/backend/utils/misc/guc.c index 17579eeaca..6baabd7273 100644 --- a/src/backend/utils/misc/guc.c +++ b/src/backend/utils/misc/guc.c @@ -1111,6 +1111,16 @@ static struct config_bool ConfigureNamesBool[] = true, NULL, NULL, NULL }, + { + {"enable_self_join_removal", PGC_USERSET, QUERY_TUNING_METHOD, + gettext_noop("Enable removal of unique self-joins."), + NULL, + GUC_EXPLAIN + }, + &enable_self_join_removal, + true, + NULL, NULL, NULL + }, { {"geqo", PGC_USERSET, QUERY_TUNING_GEQO, gettext_noop("Enables genetic query optimization."), diff --git a/src/include/optimizer/pathnode.h b/src/include/optimizer/pathnode.h index 23dec14cbd..4b7aedfe3e 100644 --- a/src/include/optimizer/pathnode.h +++ b/src/include/optimizer/pathnode.h @@ -295,6 +295,10 @@ extern RelOptInfo *build_join_rel(PlannerInfo *root, RelOptInfo *inner_rel, SpecialJoinInfo *sjinfo, List **restrictlist_ptr); +extern List *build_joinrel_restrictlist(PlannerInfo *root, + Relids joinrelids, + RelOptInfo *outer_rel, + RelOptInfo *inner_rel); extern Relids min_join_parameterization(PlannerInfo *root, Relids joinrelids, RelOptInfo *outer_rel, diff --git a/src/include/optimizer/planmain.h b/src/include/optimizer/planmain.h index 777655210b..b237005408 100644 --- a/src/include/optimizer/planmain.h +++ b/src/include/optimizer/planmain.h @@ -20,6 +20,7 @@ /* GUC parameters */ #define DEFAULT_CURSOR_TUPLE_FRACTION 0.1 extern double cursor_tuple_fraction; +extern bool enable_self_join_removal; /* query_planner callback to compute query_pathkeys */ typedef void (*query_pathkeys_callback) (PlannerInfo *root, void *extra); @@ -106,6 +107,7 @@ extern bool query_is_distinct_for(Query *query, List *colnos, List *opids); extern bool innerrel_is_unique(PlannerInfo *root, Relids joinrelids, Relids outerrelids, RelOptInfo *innerrel, JoinType jointype, List *restrictlist, bool force_cache); +extern List * remove_useless_self_joins(PlannerInfo *root, List *jointree); /* * prototypes for plan/setrefs.c diff --git a/src/test/regress/expected/equivclass.out b/src/test/regress/expected/equivclass.out index 126f7047fe..de71441052 100644 --- a/src/test/regress/expected/equivclass.out +++ b/src/test/regress/expected/equivclass.out @@ -430,6 +430,38 @@ explain (costs off) Filter: ((unique1 IS NOT NULL) AND (unique2 IS NOT NULL)) (2 rows) +-- Test that broken ECs are processed correctly during self join removal. +-- Disable merge joins so that we don't get an error about missing commutator. +-- Test both orientations of the join clause, because only one of them breaks +-- the EC. +set enable_mergejoin to off; +explain (costs off) + select * from ec0 m join ec0 n on m.ff = n.ff + join ec1 p on m.ff + n.ff = p.f1; + QUERY PLAN +---------------------------------------- + Nested Loop + Join Filter: ((n.ff + n.ff) = p.f1) + -> Seq Scan on ec1 p + -> Materialize + -> Seq Scan on ec0 n + Filter: (ff IS NOT NULL) +(6 rows) + +explain (costs off) + select * from ec0 m join ec0 n on m.ff = n.ff + join ec1 p on p.f1::int8 = (m.ff + n.ff)::int8alias1; + QUERY PLAN +--------------------------------------------------------------- + Nested Loop + Join Filter: ((p.f1)::bigint = ((n.ff + n.ff))::int8alias1) + -> Seq Scan on ec1 p + -> Materialize + -> Seq Scan on ec0 n + Filter: (ff IS NOT NULL) +(6 rows) + +reset enable_mergejoin; -- this could be converted, but isn't at present explain (costs off) select * from tenk1 where unique1 = unique1 or unique2 = unique2; diff --git a/src/test/regress/expected/join.out b/src/test/regress/expected/join.out index 81b42c601b..576c08b4b2 100644 --- a/src/test/regress/expected/join.out +++ b/src/test/regress/expected/join.out @@ -4774,6 +4774,337 @@ select * from ----+----+----+---- (0 rows) +-- +-- test that semi- or inner self-joins on a unique column are removed +-- +-- enable only nestloop to get more predictable plans +set enable_hashjoin to off; +set enable_mergejoin to off; +create table sj (a int unique, b int); +insert into sj values (1, null), (null, 2), (2, 1); +analyze sj; +select p.* from sj p, sj q where q.a = p.a and q.b = q.a - 1; + a | b +---+--- + 2 | 1 +(1 row) + +explain (costs off) +select p.* from sj p, sj q where q.a = p.a and q.b = q.a - 1; + QUERY PLAN +----------------------------------------------- + Seq Scan on sj q + Filter: ((a IS NOT NULL) AND (b = (a - 1))) +(2 rows) + +explain (costs off) +select * from sj p +where exists (select * from sj q + where q.a = p.a and q.b < 10); + QUERY PLAN +------------------------------------------ + Seq Scan on sj q + Filter: ((a IS NOT NULL) AND (b < 10)) +(2 rows) + +-- Double self-join removal. +-- Use a condition on "b + 1", not on "b", for the second join, so that +-- the equivalence class is different from the first one, and we can +-- test the non-ec code path. +explain (costs off) +select * from sj t1 join sj t2 on t1.a = t2.a and t1.b = t2.b + join sj t3 on t2.a = t3.a and t2.b + 1 = t3.b + 1; + QUERY PLAN +--------------------------------------------------------------------------- + Seq Scan on sj t3 + Filter: ((a IS NOT NULL) AND (b IS NOT NULL) AND ((b + 1) IS NOT NULL)) +(2 rows) + +-- subselect that references the removed relation +explain (costs off) +select t1.a, (select a from sj where a = t2.a and a = t1.a) +from sj t1, sj t2 +where t1.a = t2.a; + QUERY PLAN +------------------------------------------ + Seq Scan on sj t2 + Filter: (a IS NOT NULL) + SubPlan 1 + -> Result + One-Time Filter: (t2.a = t2.a) + -> Seq Scan on sj + Filter: (a = t2.a) +(7 rows) + +-- self-join under outer join +explain (costs off) +select * from sj x join sj y on x.a = y.a +left join int8_tbl z on x.a = z.q1; + QUERY PLAN +------------------------------------ + Nested Loop Left Join + Join Filter: (y.a = z.q1) + -> Seq Scan on sj y + Filter: (a IS NOT NULL) + -> Materialize + -> Seq Scan on int8_tbl z +(6 rows) + +explain (costs off) +select * from sj x join sj y on x.a = y.a +left join int8_tbl z on y.a = z.q1; + QUERY PLAN +------------------------------------ + Nested Loop Left Join + Join Filter: (y.a = z.q1) + -> Seq Scan on sj y + Filter: (a IS NOT NULL) + -> Materialize + -> Seq Scan on int8_tbl z +(6 rows) + +-- Test that placeholders are updated correctly after join removal +explain (costs off) +select * from (values (1)) x +left join (select coalesce(y.q1, 1) from int8_tbl y + right join sj j1 inner join sj j2 on j1.a = j2.a + on true) z +on true; + QUERY PLAN +------------------------------------------ + Nested Loop Left Join + -> Result + -> Nested Loop Left Join + -> Seq Scan on sj j2 + Filter: (a IS NOT NULL) + -> Materialize + -> Seq Scan on int8_tbl y +(7 rows) + +-- Test that OR predicated are updated correctly after join removal +CREATE TABLE tab_with_flag ( id INT PRIMARY KEY, is_flag SMALLINT); +CREATE INDEX idx_test_is_flag ON tab_with_flag (is_flag); +explain (costs off) +SELECT COUNT(*) FROM tab_with_flag WHERE (is_flag IS NULL OR is_flag = 0) AND id IN (SELECT id FROM tab_with_flag WHERE id IN (2, 3)); + QUERY PLAN +---------------------------------------------------------------------------------- + Aggregate + -> Bitmap Heap Scan on tab_with_flag + Recheck Cond: ((id = ANY ('{2,3}'::integer[])) AND (id IS NOT NULL)) + Filter: ((is_flag IS NULL) OR (is_flag = 0)) + -> Bitmap Index Scan on tab_with_flag_pkey + Index Cond: ((id = ANY ('{2,3}'::integer[])) AND (id IS NOT NULL)) +(6 rows) + +DROP TABLE tab_with_flag; +-- HAVING clause +explain (costs off) +select p.b from sj p join sj q on p.a = q.a group by p.b having sum(p.a) = 1; + QUERY PLAN +--------------------------------- + HashAggregate + Group Key: q.b + Filter: (sum(q.a) = 1) + -> Seq Scan on sj q + Filter: (a IS NOT NULL) +(5 rows) + +-- update lateral references and range table entry reference +explain (verbose, costs off) +select 1 from (select x.* from sj x, sj y where x.a = y.a) q, + lateral generate_series(1, q.a) gs(i); + QUERY PLAN +------------------------------------------------------ + Nested Loop + Output: 1 + -> Seq Scan on public.sj y + Output: y.a, y.b + Filter: (y.a IS NOT NULL) + -> Function Scan on pg_catalog.generate_series gs + Output: gs.i + Function Call: generate_series(1, y.a) +(8 rows) + +explain (verbose, costs off) +select 1 from (select y.* from sj x, sj y where x.a = y.a) q, + lateral generate_series(1, q.a) gs(i); + QUERY PLAN +------------------------------------------------------ + Nested Loop + Output: 1 + -> Seq Scan on public.sj y + Output: y.a, y.b + Filter: (y.a IS NOT NULL) + -> Function Scan on pg_catalog.generate_series gs + Output: gs.i + Function Call: generate_series(1, y.a) +(8 rows) + +-- Test that a non-EC-derived join clause is processed correctly. Use an +-- outer join so that we can't form an EC. +explain (costs off) select * from sj p join sj q on p.a = q.a + left join sj r on p.a + q.a = r.a; + QUERY PLAN +------------------------------------ + Nested Loop Left Join + Join Filter: ((q.a + q.a) = r.a) + -> Seq Scan on sj q + Filter: (a IS NOT NULL) + -> Materialize + -> Seq Scan on sj r +(6 rows) + +-- FIXME this constant false filter doesn't look good. Should we merge +-- equivalence classes? +explain (costs off) +select * from sj p, sj q where p.a = q.a and p.b = 1 and q.b = 2; + QUERY PLAN +----------------------------------------------------- + Seq Scan on sj q + Filter: ((a IS NOT NULL) AND (b = 2) AND (b = 1)) +(2 rows) + +-- Check that attr_needed is updated correctly after self-join removal. In this +-- test, the join of j1 with j2 is removed. k1.b is required at either j1 or j2. +-- If this info is lost, join targetlist for (k1, k2) will not contain k1.b. +-- Use index scan for k1 so that we don't get 'b' from physical tlist used for +-- seqscan. Also disable reordering of joins because this test depends on a +-- particular join tree. +create table sk (a int, b int); +create index on sk(a); +set join_collapse_limit to 1; +set enable_seqscan to off; +explain (costs off) select 1 from + (sk k1 join sk k2 on k1.a = k2.a) + join (sj j1 join sj j2 on j1.a = j2.a) on j1.b = k1.b; + QUERY PLAN +----------------------------------------------------- + Nested Loop + Join Filter: (k1.b = j2.b) + -> Nested Loop + -> Index Scan using sk_a_idx on sk k1 + -> Index Only Scan using sk_a_idx on sk k2 + Index Cond: (a = k1.a) + -> Materialize + -> Index Scan using sj_a_key on sj j2 + Index Cond: (a IS NOT NULL) +(9 rows) + +explain (costs off) select 1 from + (sk k1 join sk k2 on k1.a = k2.a) + join (sj j1 join sj j2 on j1.a = j2.a) on j2.b = k1.b; + QUERY PLAN +----------------------------------------------------- + Nested Loop + Join Filter: (k1.b = j2.b) + -> Nested Loop + -> Index Scan using sk_a_idx on sk k1 + -> Index Only Scan using sk_a_idx on sk k2 + Index Cond: (a = k1.a) + -> Materialize + -> Index Scan using sj_a_key on sj j2 + Index Cond: (a IS NOT NULL) +(9 rows) + +reset join_collapse_limit; +reset enable_seqscan; +-- Check that clauses from the join filter list is not lost on the self-join removal +CREATE TABLE emp1 ( id SERIAL PRIMARY KEY NOT NULL, code int); +explain (verbose, costs off) +SELECT * FROM emp1 e1, emp1 e2 WHERE e1.id = e2.id AND e2.code <> e1.code; + QUERY PLAN +---------------------------------------------------------- + Seq Scan on public.emp1 e2 + Output: e2.id, e2.code, e2.id, e2.code + Filter: ((e2.id IS NOT NULL) AND (e2.code <> e2.code)) +(3 rows) + +-- We can remove the join even if we find the join can't duplicate rows and +-- the base quals of each side are different. In the following case we end up +-- moving quals over to s1 to make it so it can't match any rows. +create table sl(a int, b int); +create unique index on sl(a, b); +vacuum analyze sl; +-- Both sides are unique, but base quals are different +explain (costs off) +select * from sl t1, sl t2 where t1.a = t2.a and t1.b = 1 and t2.b = 2; + QUERY PLAN +----------------------------------------------------- + Seq Scan on sl t2 + Filter: ((a IS NOT NULL) AND (b = 2) AND (b = 1)) +(2 rows) + +-- +---- Only one side is unqiue +--select * from sl t1, sl t2 where t1.a = t2.a and t1.b = 1; +--select * from sl t1, sl t2 where t1.a = t2.a and t2.b = 1; +-- +---- Several uniques indexes match, and we select a different one +---- for each side, so the join is not removed +--create table sm(a int unique, b int unique, c int unique); +--explain (costs off) +--select * from sm m, sm n where m.a = n.b and m.c = n.c; +--explain (costs off) +--select * from sm m, sm n where m.a = n.c and m.b = n.b; +--explain (costs off) +--select * from sm m, sm n where m.c = n.b and m.a = n.a; +-- Check optimization disabling if it will violate special join conditions. +-- Two identical joined relations satisfies self join removal conditions but +-- stay in different special join infos. +CREATE TABLE sj_t1 (id serial, a int); +CREATE TABLE sj_t2 (id serial, a int); +CREATE TABLE sj_t3 (id serial, a int); +CREATE TABLE sj_t4 (id serial, a int); +CREATE UNIQUE INDEX ON sj_t3 USING btree (a,id); +CREATE UNIQUE INDEX ON sj_t2 USING btree (id); +EXPLAIN (COSTS OFF) +SELECT * FROM sj_t1 +JOIN ( + SELECT sj_t2.id AS id FROM sj_t2 + WHERE EXISTS + ( + SELECT TRUE FROM sj_t3,sj_t4 WHERE sj_t3.a = 1 AND sj_t3.id = sj_t2.id + ) + ) t2t3t4 +ON sj_t1.id = t2t3t4.id +JOIN ( + SELECT sj_t2.id AS id FROM sj_t2 + WHERE EXISTS + ( + SELECT TRUE FROM sj_t3,sj_t4 WHERE sj_t3.a = 1 AND sj_t3.id = sj_t2.id + ) + ) _t2t3t4 +ON sj_t1.id = _t2t3t4.id; + QUERY PLAN +------------------------------------------------------------------------------------- + Nested Loop + Join Filter: (sj_t3.id = sj_t1.id) + -> Nested Loop + Join Filter: (sj_t3.id = sj_t2.id) + -> Nested Loop Semi Join + -> Nested Loop + -> HashAggregate + Group Key: sj_t3.id + -> Nested Loop + -> Seq Scan on sj_t4 + -> Materialize + -> Bitmap Heap Scan on sj_t3 + Recheck Cond: (a = 1) + -> Bitmap Index Scan on sj_t3_a_id_idx + Index Cond: (a = 1) + -> Index Only Scan using sj_t2_id_idx on sj_t2 sj_t2_1 + Index Cond: (id = sj_t3.id) + -> Nested Loop + -> Index Only Scan using sj_t3_a_id_idx on sj_t3 sj_t3_1 + Index Cond: ((a = 1) AND (id = sj_t3.id)) + -> Seq Scan on sj_t4 sj_t4_1 + -> Index Only Scan using sj_t2_id_idx on sj_t2 + Index Cond: (id = sj_t2_1.id) + -> Seq Scan on sj_t1 +(24 rows) + +reset enable_hashjoin; +reset enable_mergejoin; -- -- Test hints given on incorrect column references are useful -- diff --git a/src/test/regress/expected/sysviews.out b/src/test/regress/expected/sysviews.out index 81bdacf59d..9847097700 100644 --- a/src/test/regress/expected/sysviews.out +++ b/src/test/regress/expected/sysviews.out @@ -103,10 +103,11 @@ select name, setting from pg_settings where name like 'enable%'; enable_partition_pruning | on enable_partitionwise_aggregate | off enable_partitionwise_join | off + enable_self_join_removal | on enable_seqscan | on enable_sort | on enable_tidscan | on -(18 rows) +(19 rows) -- Test that the pg_timezone_names and pg_timezone_abbrevs views are -- more-or-less working. We can't test their contents in any great detail diff --git a/src/test/regress/sql/equivclass.sql b/src/test/regress/sql/equivclass.sql index 247b0a3105..77dd964ebf 100644 --- a/src/test/regress/sql/equivclass.sql +++ b/src/test/regress/sql/equivclass.sql @@ -259,6 +259,22 @@ drop user regress_user_ectest; explain (costs off) select * from tenk1 where unique1 = unique1 and unique2 = unique2; +-- Test that broken ECs are processed correctly during self join removal. +-- Disable merge joins so that we don't get an error about missing commutator. +-- Test both orientations of the join clause, because only one of them breaks +-- the EC. +set enable_mergejoin to off; + +explain (costs off) + select * from ec0 m join ec0 n on m.ff = n.ff + join ec1 p on m.ff + n.ff = p.f1; + +explain (costs off) + select * from ec0 m join ec0 n on m.ff = n.ff + join ec1 p on p.f1::int8 = (m.ff + n.ff)::int8alias1; + +reset enable_mergejoin; + -- this could be converted, but isn't at present explain (costs off) select * from tenk1 where unique1 = unique1 or unique2 = unique2; diff --git a/src/test/regress/sql/join.sql b/src/test/regress/sql/join.sql index 9887fe0c0b..7168900bc8 100644 --- a/src/test/regress/sql/join.sql +++ b/src/test/regress/sql/join.sql @@ -1705,6 +1705,172 @@ select * from select * from int8_tbl x join (int4_tbl x cross join int4_tbl y(ff)) j on q1 = f1; -- ok +-- +-- test that semi- or inner self-joins on a unique column are removed +-- + +-- enable only nestloop to get more predictable plans +set enable_hashjoin to off; +set enable_mergejoin to off; + +create table sj (a int unique, b int); +insert into sj values (1, null), (null, 2), (2, 1); +analyze sj; + +select p.* from sj p, sj q where q.a = p.a and q.b = q.a - 1; + +explain (costs off) +select p.* from sj p, sj q where q.a = p.a and q.b = q.a - 1; + +explain (costs off) +select * from sj p +where exists (select * from sj q + where q.a = p.a and q.b < 10); + +-- Double self-join removal. +-- Use a condition on "b + 1", not on "b", for the second join, so that +-- the equivalence class is different from the first one, and we can +-- test the non-ec code path. +explain (costs off) +select * from sj t1 join sj t2 on t1.a = t2.a and t1.b = t2.b + join sj t3 on t2.a = t3.a and t2.b + 1 = t3.b + 1; + +-- subselect that references the removed relation +explain (costs off) +select t1.a, (select a from sj where a = t2.a and a = t1.a) +from sj t1, sj t2 +where t1.a = t2.a; + +-- self-join under outer join +explain (costs off) +select * from sj x join sj y on x.a = y.a +left join int8_tbl z on x.a = z.q1; + +explain (costs off) +select * from sj x join sj y on x.a = y.a +left join int8_tbl z on y.a = z.q1; + +-- Test that placeholders are updated correctly after join removal +explain (costs off) +select * from (values (1)) x +left join (select coalesce(y.q1, 1) from int8_tbl y + right join sj j1 inner join sj j2 on j1.a = j2.a + on true) z +on true; + +-- Test that OR predicated are updated correctly after join removal +CREATE TABLE tab_with_flag ( id INT PRIMARY KEY, is_flag SMALLINT); +CREATE INDEX idx_test_is_flag ON tab_with_flag (is_flag); +explain (costs off) +SELECT COUNT(*) FROM tab_with_flag WHERE (is_flag IS NULL OR is_flag = 0) AND id IN (SELECT id FROM tab_with_flag WHERE id IN (2, 3)); +DROP TABLE tab_with_flag; + +-- HAVING clause +explain (costs off) +select p.b from sj p join sj q on p.a = q.a group by p.b having sum(p.a) = 1; + +-- update lateral references and range table entry reference +explain (verbose, costs off) +select 1 from (select x.* from sj x, sj y where x.a = y.a) q, + lateral generate_series(1, q.a) gs(i); + +explain (verbose, costs off) +select 1 from (select y.* from sj x, sj y where x.a = y.a) q, + lateral generate_series(1, q.a) gs(i); + +-- Test that a non-EC-derived join clause is processed correctly. Use an +-- outer join so that we can't form an EC. +explain (costs off) select * from sj p join sj q on p.a = q.a + left join sj r on p.a + q.a = r.a; + +-- FIXME this constant false filter doesn't look good. Should we merge +-- equivalence classes? +explain (costs off) +select * from sj p, sj q where p.a = q.a and p.b = 1 and q.b = 2; + +-- Check that attr_needed is updated correctly after self-join removal. In this +-- test, the join of j1 with j2 is removed. k1.b is required at either j1 or j2. +-- If this info is lost, join targetlist for (k1, k2) will not contain k1.b. +-- Use index scan for k1 so that we don't get 'b' from physical tlist used for +-- seqscan. Also disable reordering of joins because this test depends on a +-- particular join tree. +create table sk (a int, b int); +create index on sk(a); +set join_collapse_limit to 1; +set enable_seqscan to off; +explain (costs off) select 1 from + (sk k1 join sk k2 on k1.a = k2.a) + join (sj j1 join sj j2 on j1.a = j2.a) on j1.b = k1.b; +explain (costs off) select 1 from + (sk k1 join sk k2 on k1.a = k2.a) + join (sj j1 join sj j2 on j1.a = j2.a) on j2.b = k1.b; +reset join_collapse_limit; +reset enable_seqscan; + +-- Check that clauses from the join filter list is not lost on the self-join removal +CREATE TABLE emp1 ( id SERIAL PRIMARY KEY NOT NULL, code int); +explain (verbose, costs off) +SELECT * FROM emp1 e1, emp1 e2 WHERE e1.id = e2.id AND e2.code <> e1.code; + +-- We can remove the join even if we find the join can't duplicate rows and +-- the base quals of each side are different. In the following case we end up +-- moving quals over to s1 to make it so it can't match any rows. +create table sl(a int, b int); +create unique index on sl(a, b); +vacuum analyze sl; + +-- Both sides are unique, but base quals are different +explain (costs off) +select * from sl t1, sl t2 where t1.a = t2.a and t1.b = 1 and t2.b = 2; + +-- +---- Only one side is unqiue +--select * from sl t1, sl t2 where t1.a = t2.a and t1.b = 1; +--select * from sl t1, sl t2 where t1.a = t2.a and t2.b = 1; +-- +---- Several uniques indexes match, and we select a different one +---- for each side, so the join is not removed +--create table sm(a int unique, b int unique, c int unique); +--explain (costs off) +--select * from sm m, sm n where m.a = n.b and m.c = n.c; +--explain (costs off) +--select * from sm m, sm n where m.a = n.c and m.b = n.b; +--explain (costs off) +--select * from sm m, sm n where m.c = n.b and m.a = n.a; + +-- Check optimization disabling if it will violate special join conditions. +-- Two identical joined relations satisfies self join removal conditions but +-- stay in different special join infos. +CREATE TABLE sj_t1 (id serial, a int); +CREATE TABLE sj_t2 (id serial, a int); +CREATE TABLE sj_t3 (id serial, a int); +CREATE TABLE sj_t4 (id serial, a int); + +CREATE UNIQUE INDEX ON sj_t3 USING btree (a,id); +CREATE UNIQUE INDEX ON sj_t2 USING btree (id); + +EXPLAIN (COSTS OFF) +SELECT * FROM sj_t1 +JOIN ( + SELECT sj_t2.id AS id FROM sj_t2 + WHERE EXISTS + ( + SELECT TRUE FROM sj_t3,sj_t4 WHERE sj_t3.a = 1 AND sj_t3.id = sj_t2.id + ) + ) t2t3t4 +ON sj_t1.id = t2t3t4.id +JOIN ( + SELECT sj_t2.id AS id FROM sj_t2 + WHERE EXISTS + ( + SELECT TRUE FROM sj_t3,sj_t4 WHERE sj_t3.a = 1 AND sj_t3.id = sj_t2.id + ) + ) _t2t3t4 +ON sj_t1.id = _t2t3t4.id; + +reset enable_hashjoin; +reset enable_mergejoin; + -- -- Test hints given on incorrect column references are useful -- -- 2.25.1 --------------1F3B3A4B8A220BEE26665206-- ^ permalink raw reply [nested|flat] 2+ messages in thread
* Re: SLRUs in the main buffer pool - Page Header definitions @ 2023-01-10 19:05 Bagga, Rishu <[email protected]> 0 siblings, 0 replies; 2+ messages in thread From: Bagga, Rishu @ 2023-01-10 19:05 UTC (permalink / raw) To: vignesh C <[email protected]>; pgsql-hackers; +Cc: Heikki Linnakangas <[email protected]>; Thomas Munro <[email protected]>; Debnath, Shawn <[email protected]> >On 1/3/23, 5:06 AM, "vignesh C" <[email protected]> wrote: >>On Fri, 16 Dec 2022 at 04:47, Bagga, Rishu <[email protected]> >>wrote: >>Rebased and updated a new patch addressing the critical section issue in >>RecordNewMultliXact.In GetNewMultiXactId, we now make our ReadBuffer >>calls before starting the critical section, but while holding the >>MultiXactGenLock, so we always fetch the correct buffers. We store >>them in an array that is accessed later in RecordNewMultiXact. >>This way we can keep the existing functionality of only holding the >>MultiXactGenLock while reading in buffers, but can let go when we are >>writing, to preserve the existing concurrency paradigm. >>Let me know your thoughts on this approach. > The patch does not apply on top of HEAD as in [1], please post a >rebased patch: >=== Applying patches on top of PostgreSQL commit ID >92957ed98c5c565362ce665266132a7f08f6b0c0 === >=== applying patch ./slru_to_buffer_cache_with_page_headers_v3.patch ... >patching file src/include/catalog/catversion.h >Hunk #1 FAILED at 57. >1 out of 1 hunk FAILED -- saving rejects to file >src/include/catalog/catversion.h.rej >[1] - http://cfbot.cputube.org/patch _41_3514.log >Regards, >Vignesh Hi all, Rebased the patch, and fixed a bug I introduced in the previous patch in TrimCLOG. We ran a quick set of pgbench tests and observed no regressions. Here are the numbers: 3 trials, with scale 10,000, 350 clients, 350 threads, over 30 minutes: Median TPS: Control

Trial 1: 58331.0 Trial 2: 57191.0 Trial 3: 57101.3 Average of Medians: 57541.1 SLRUs to BufferCache + Page Headers: Trial 1: 62605.0 Trial 2: 62891.2 Trial 3: 59906.3 Average of Medians: 61800.8 Machine Specs: Driver Instance: m5d.metal Architecture x86_64 CPUs: 96 RAM: 384 GiB OS: Amazon Linux 2 Server Instance: r5dn.metal Architecture x86_64 CPUs: 64 RAM: 500GiB OS: Amazon Linux 2 Looking forward to your feedback on this. Sincerely, Rishu Bagga, Amazon Web Services (AWS)  Attachments: [application/octet-stream] slru_to_buffer_cache_with_page_headers_v4.patch (348.2K, ../../[email protected]/2-slru_to_buffer_cache_with_page_headers_v4.patch) download | inline diff: From 5d66fad1fac30ee6869be234e5d0b9c5e6bfebb0 Mon Sep 17 00:00:00 2001 From: Rishu Bagga <[email protected]> Date: Thu, 15 Sep 2022 00:55:25 +0000 Subject: [PATCH] heikki s2b patch + page headers + upgrade --- contrib/amcheck/verify_nbtree.c | 2 +- contrib/bloom/blinsert.c | 6 +- contrib/pg_prewarm/autoprewarm.c | 2 +- contrib/pg_prewarm/pg_prewarm.c | 4 +- contrib/pg_visibility/pg_visibility.c | 5 +- src/backend/access/gist/gistbuild.c | 11 +- src/backend/access/hash/hashpage.c | 2 +- src/backend/access/heap/heapam_handler.c | 32 +- src/backend/access/heap/rewriteheap.c | 6 +- src/backend/access/heap/visibilitymap.c | 41 +- src/backend/access/nbtree/nbtree.c | 6 +- src/backend/access/nbtree/nbtsort.c | 8 +- src/backend/access/spgist/spginsert.c | 17 +- src/backend/access/table/tableam.c | 4 +- src/backend/access/transam/clog.c | 227 +-- src/backend/access/transam/commit_ts.c | 199 +-- src/backend/access/transam/multixact.c | 508 +++---- src/backend/access/transam/slru.c | 1484 ++----------------- src/backend/access/transam/subtrans.c | 114 +- src/backend/access/transam/xact.c | 2 + src/backend/access/transam/xlog.c | 15 +- src/backend/access/transam/xlogprefetcher.c | 28 +- src/backend/access/transam/xlogutils.c | 17 +- src/backend/catalog/catalog.c | 53 +- src/backend/catalog/index.c | 4 +- src/backend/catalog/storage.c | 205 ++- src/backend/commands/async.c | 97 +- src/backend/commands/dbcommands.c | 8 +- src/backend/commands/sequence.c | 8 +- src/backend/commands/tablecmds.c | 19 +- src/backend/storage/buffer/buf_init.c | 17 +- src/backend/storage/buffer/bufmgr.c | 379 ++--- src/backend/storage/buffer/localbuf.c | 25 +- src/backend/storage/freespace/freespace.c | 41 +- src/backend/storage/ipc/ipci.c | 4 - src/backend/storage/lmgr/predicate.c | 72 +- src/backend/storage/page/bufpage.c | 23 +- src/backend/storage/smgr/md.c | 366 ++--- src/backend/storage/smgr/smgr.c | 399 +++-- src/backend/storage/sync/sync.c | 20 +- src/backend/utils/cache/inval.c | 17 +- src/backend/utils/cache/relcache.c | 27 +- src/backend/utils/mmgr/mcxt.c | 13 +- src/bin/pg_upgrade/file.c | 175 ++- src/bin/pg_upgrade/function.c | 66 + src/bin/pg_upgrade/pg_upgrade.c | 76 +- src/bin/pg_upgrade/pg_upgrade.h | 19 +- src/common/relpath.c | 31 +- src/include/access/clog.h | 6 - src/include/access/commit_ts.h | 3 - src/include/access/multixact.h | 3 - src/include/access/slru.h | 181 +-- src/include/access/slrudefs.h | 19 + src/include/access/slrulist.h | 30 + src/include/access/subtrans.h | 3 - src/include/catalog/storage.h | 11 +- src/include/common/relpath.h | 15 +- src/include/storage/buf_internals.h | 11 +- src/include/storage/bufmgr.h | 38 +- src/include/storage/bufpage.h | 17 + src/include/storage/md.h | 28 +- src/include/storage/relfilelocator.h | 29 +- src/include/storage/smgr.h | 99 +- src/include/utils/inval.h | 2 +- src/include/utils/rel.h | 33 +- 65 files changed, 2173 insertions(+), 3259 deletions(-) create mode 100644 src/include/access/slrudefs.h create mode 100644 src/include/access/slrulist.h diff --git a/contrib/amcheck/verify_nbtree.c b/contrib/amcheck/verify_nbtree.c index 257cff671bd..2ab837c3062 100644 --- a/contrib/amcheck/verify_nbtree.c +++ b/contrib/amcheck/verify_nbtree.c @@ -324,7 +324,7 @@ bt_index_check_internal(Oid indrelid, bool parentcheck, bool heapallindexed, bool heapkeyspace, allequalimage; - if (!smgrexists(RelationGetSmgr(indrel), MAIN_FORKNUM)) + if (!smgrexists(RelationGetSmgr(indrel, MAIN_FORKNUM))) ereport(ERROR, (errcode(ERRCODE_INDEX_CORRUPTED), errmsg("index \"%s\" lacks a main relation fork", diff --git a/contrib/bloom/blinsert.c b/contrib/bloom/blinsert.c index f81442efb3f..49b952f3c3b 100644 --- a/contrib/bloom/blinsert.c +++ b/contrib/bloom/blinsert.c @@ -177,9 +177,9 @@ blbuildempty(Relation index) * this even when wal_level=minimal. */ PageSetChecksumInplace(metapage, BLOOM_METAPAGE_BLKNO); - smgrwrite(RelationGetSmgr(index), INIT_FORKNUM, BLOOM_METAPAGE_BLKNO, + smgrwrite(RelationGetSmgr(index, INIT_FORKNUM), BLOOM_METAPAGE_BLKNO, (char *) metapage, true); - log_newpage(&(RelationGetSmgr(index))->smgr_rlocator.locator, INIT_FORKNUM, + log_newpage(&index->rd_locator, INIT_FORKNUM, BLOOM_METAPAGE_BLKNO, metapage, true); /* @@ -187,7 +187,7 @@ blbuildempty(Relation index) * write did not go through shared_buffers and therefore a concurrent * checkpoint may have moved the redo pointer past our xlog record. */ - smgrimmedsync(RelationGetSmgr(index), INIT_FORKNUM); + smgrimmedsync(RelationGetSmgr(index, INIT_FORKNUM)); } /* diff --git a/contrib/pg_prewarm/autoprewarm.c b/contrib/pg_prewarm/autoprewarm.c index 93835449c0e..74cc8df6481 100644 --- a/contrib/pg_prewarm/autoprewarm.c +++ b/contrib/pg_prewarm/autoprewarm.c @@ -535,7 +535,7 @@ autoprewarm_database_main(Datum main_arg) */ if (blk->forknum > InvalidForkNumber && blk->forknum <= MAX_FORKNUM && - smgrexists(RelationGetSmgr(rel), blk->forknum)) + smgrexists(RelationGetSmgr(rel, blk->forknum))) nblocks = RelationGetNumberOfBlocksInFork(rel, blk->forknum); else nblocks = 0; diff --git a/contrib/pg_prewarm/pg_prewarm.c b/contrib/pg_prewarm/pg_prewarm.c index 54209924aed..479bbafb745 100644 --- a/contrib/pg_prewarm/pg_prewarm.c +++ b/contrib/pg_prewarm/pg_prewarm.c @@ -109,7 +109,7 @@ pg_prewarm(PG_FUNCTION_ARGS) aclcheck_error(aclresult, get_relkind_objtype(rel->rd_rel->relkind), get_rel_name(relOid)); /* Check that the fork exists. */ - if (!smgrexists(RelationGetSmgr(rel), forkNumber)) + if (!smgrexists(RelationGetSmgr(rel, forkNumber))) ereport(ERROR, (errcode(ERRCODE_INVALID_PARAMETER_VALUE), errmsg("fork \"%s\" does not exist for this relation", @@ -177,7 +177,7 @@ pg_prewarm(PG_FUNCTION_ARGS) for (block = first_block; block <= last_block; ++block) { CHECK_FOR_INTERRUPTS(); - smgrread(RelationGetSmgr(rel), forkNumber, block, blockbuffer.data); + smgrread(RelationGetSmgr(rel, forkNumber), block, blockbuffer.data); ++blocks_done; } } diff --git a/contrib/pg_visibility/pg_visibility.c b/contrib/pg_visibility/pg_visibility.c index 2a4acfd1eee..bb7cd8ec43d 100644 --- a/contrib/pg_visibility/pg_visibility.c +++ b/contrib/pg_visibility/pg_visibility.c @@ -386,13 +386,14 @@ pg_truncate_visibility_map(PG_FUNCTION_ARGS) check_relation_relkind(rel); /* Forcibly reset cached file size */ - RelationGetSmgr(rel)->smgr_cached_nblocks[VISIBILITYMAP_FORKNUM] = InvalidBlockNumber; + RelationGetSmgr(rel, VISIBILITYMAP_FORKNUM)->smgr_cached_nblocks = InvalidBlockNumber; block = visibilitymap_prepare_truncate(rel, 0); if (BlockNumberIsValid(block)) { fork = VISIBILITYMAP_FORKNUM; - smgrtruncate(RelationGetSmgr(rel), &fork, 1, &block); + DropRelationBuffers(rel->rd_locator, rel->rd_backend, &fork, 1, &block); + smgrtruncate_multi(rel->rd_locator, rel->rd_backend, &fork, 1, &block); } if (RelationNeedsWAL(rel)) diff --git a/src/backend/access/gist/gistbuild.c b/src/backend/access/gist/gistbuild.c index d21a308d41d..5ee6c26f084 100644 --- a/src/backend/access/gist/gistbuild.c +++ b/src/backend/access/gist/gistbuild.c @@ -416,7 +416,7 @@ gist_indexsortbuild(GISTBuildState *state) * replaced with the real root page at the end. */ page = palloc0(BLCKSZ); - smgrextend(RelationGetSmgr(state->indexrel), MAIN_FORKNUM, GIST_ROOT_BLKNO, + smgrextend(RelationGetSmgr(state->indexrel, MAIN_FORKNUM), GIST_ROOT_BLKNO, page, true); state->pages_allocated++; state->pages_written++; @@ -460,7 +460,7 @@ gist_indexsortbuild(GISTBuildState *state) /* Write out the root */ PageSetLSN(levelstate->pages[0], GistBuildLSN); PageSetChecksumInplace(levelstate->pages[0], GIST_ROOT_BLKNO); - smgrwrite(RelationGetSmgr(state->indexrel), MAIN_FORKNUM, GIST_ROOT_BLKNO, + smgrwrite(RelationGetSmgr(state->indexrel, MAIN_FORKNUM), GIST_ROOT_BLKNO, levelstate->pages[0], true); if (RelationNeedsWAL(state->indexrel)) log_newpage(&state->indexrel->rd_locator, MAIN_FORKNUM, GIST_ROOT_BLKNO, @@ -479,7 +479,7 @@ gist_indexsortbuild(GISTBuildState *state) * still not be on disk when the crash occurs. */ if (RelationNeedsWAL(state->indexrel)) - smgrimmedsync(RelationGetSmgr(state->indexrel), MAIN_FORKNUM); + smgrimmedsync(RelationGetSmgr(state->indexrel, MAIN_FORKNUM)); } /* @@ -657,7 +657,7 @@ gist_indexsortbuild_flush_ready_pages(GISTBuildState *state) PageSetLSN(page, GistBuildLSN); PageSetChecksumInplace(page, blkno); - smgrextend(RelationGetSmgr(state->indexrel), MAIN_FORKNUM, blkno, page, + smgrextend(RelationGetSmgr(state->indexrel, MAIN_FORKNUM), blkno, page, true); state->pages_written++; @@ -943,8 +943,7 @@ gistBuildCallback(Relation index, */ if ((buildstate->buildMode == GIST_BUFFERING_AUTO && buildstate->indtuples % BUFFERING_MODE_SWITCH_CHECK_STEP == 0 && - effective_cache_size < smgrnblocks(RelationGetSmgr(index), - MAIN_FORKNUM)) || + effective_cache_size < smgrnblocks(RelationGetSmgr(index, MAIN_FORKNUM))) || (buildstate->buildMode == GIST_BUFFERING_STATS && buildstate->indtuples >= BUFFERING_MODE_TUPLE_SIZE_STATS_TARGET)) { diff --git a/src/backend/access/hash/hashpage.c b/src/backend/access/hash/hashpage.c index 2d8fdec98ed..166bf81e028 100644 --- a/src/backend/access/hash/hashpage.c +++ b/src/backend/access/hash/hashpage.c @@ -1030,7 +1030,7 @@ _hash_alloc_buckets(Relation rel, BlockNumber firstblock, uint32 nblocks) true); PageSetChecksumInplace(page, lastblock); - smgrextend(RelationGetSmgr(rel), MAIN_FORKNUM, lastblock, zerobuf.data, + smgrextend(RelationGetSmgr(rel, MAIN_FORKNUM), lastblock, zerobuf.data, false); return true; diff --git a/src/backend/access/heap/heapam_handler.c b/src/backend/access/heap/heapam_handler.c index c4b1916d36e..8af225502cc 100644 --- a/src/backend/access/heap/heapam_handler.c +++ b/src/backend/access/heap/heapam_handler.c @@ -572,7 +572,7 @@ heapam_relation_set_new_filelocator(Relation rel, TransactionId *freezeXid, MultiXactId *minmulti) { - SMgrRelation srel; + SMgrFileHandle sfile; /* * Initialize to the minimum XID that could put tuples in the table. We @@ -591,7 +591,7 @@ heapam_relation_set_new_filelocator(Relation rel, */ *minmulti = GetOldestMultiXactId(); - srel = RelationCreateStorage(*newrlocator, persistence, true); + sfile = RelationCreateStorage(*newrlocator, persistence, true); /* * If required, set up an init fork for an unlogged table so that it can @@ -604,15 +604,18 @@ heapam_relation_set_new_filelocator(Relation rel, */ if (persistence == RELPERSISTENCE_UNLOGGED) { + SMgrFileHandle sfile_init; + Assert(rel->rd_rel->relkind == RELKIND_RELATION || rel->rd_rel->relkind == RELKIND_MATVIEW || rel->rd_rel->relkind == RELKIND_TOASTVALUE); - smgrcreate(srel, INIT_FORKNUM, false); + sfile_init = smgropen(*newrlocator, InvalidBackendId, INIT_FORKNUM); + smgrcreate(sfile_init, false); log_smgrcreate(newrlocator, INIT_FORKNUM); - smgrimmedsync(srel, INIT_FORKNUM); + smgrimmedsync(sfile); } - smgrclose(srel); + smgrclose(sfile); } static void @@ -624,9 +627,7 @@ heapam_relation_nontransactional_truncate(Relation rel) static void heapam_relation_copy_data(Relation rel, const RelFileLocator *newrlocator) { - SMgrRelation dstrel; - - dstrel = smgropen(*newrlocator, rel->rd_backend); + SMgrFileHandle dstmain; /* * Since we copy the file directly without looking at the shared buffers, @@ -646,16 +647,21 @@ heapam_relation_copy_data(Relation rel, const RelFileLocator *newrlocator) RelationCreateStorage(*newrlocator, rel->rd_rel->relpersistence, true); /* copy main fork */ - RelationCopyStorage(RelationGetSmgr(rel), dstrel, MAIN_FORKNUM, + dstmain = smgropen(*newrlocator, rel->rd_backend, MAIN_FORKNUM); + RelationCopyStorage(RelationGetSmgr(rel, MAIN_FORKNUM), dstmain, rel->rd_rel->relpersistence); /* copy those extra forks that exist */ for (ForkNumber forkNum = MAIN_FORKNUM + 1; forkNum <= MAX_FORKNUM; forkNum++) { - if (smgrexists(RelationGetSmgr(rel), forkNum)) + SMgrFileHandle src_fork = RelationGetSmgr(rel, forkNum); + + if (smgrexists(src_fork)) { - smgrcreate(dstrel, forkNum, false); + SMgrFileHandle dst_fork = smgropen(*newrlocator, rel->rd_backend, forkNum); + + smgrcreate(dst_fork, false); /* * WAL log creation if the relation is persistent, or this is the @@ -665,7 +671,7 @@ heapam_relation_copy_data(Relation rel, const RelFileLocator *newrlocator) (rel->rd_rel->relpersistence == RELPERSISTENCE_UNLOGGED && forkNum == INIT_FORKNUM)) log_smgrcreate(newrlocator, forkNum); - RelationCopyStorage(RelationGetSmgr(rel), dstrel, forkNum, + RelationCopyStorage(RelationGetSmgr(rel, forkNum), dst_fork, rel->rd_rel->relpersistence); } } @@ -673,7 +679,7 @@ heapam_relation_copy_data(Relation rel, const RelFileLocator *newrlocator) /* drop old relation, and close new one */ RelationDropStorage(rel); - smgrclose(dstrel); + smgrclose(dstmain); } static void diff --git a/src/backend/access/heap/rewriteheap.c b/src/backend/access/heap/rewriteheap.c index 8993c1ed5a9..d991e7765f3 100644 --- a/src/backend/access/heap/rewriteheap.c +++ b/src/backend/access/heap/rewriteheap.c @@ -325,7 +325,7 @@ end_heap_rewrite(RewriteState state) PageSetChecksumInplace(state->rs_buffer, state->rs_blockno); - smgrextend(RelationGetSmgr(state->rs_new_rel), MAIN_FORKNUM, + smgrextend(RelationGetSmgr(state->rs_new_rel, MAIN_FORKNUM), state->rs_blockno, (char *) state->rs_buffer, true); } @@ -337,7 +337,7 @@ end_heap_rewrite(RewriteState state) * wrote before the checkpoint. */ if (RelationNeedsWAL(state->rs_new_rel)) - smgrimmedsync(RelationGetSmgr(state->rs_new_rel), MAIN_FORKNUM); + smgrimmedsync(RelationGetSmgr(state->rs_new_rel, MAIN_FORKNUM)); logical_end_heap_rewrite(state); @@ -691,7 +691,7 @@ raw_heap_insert(RewriteState state, HeapTuple tup) */ PageSetChecksumInplace(page, state->rs_blockno); - smgrextend(RelationGetSmgr(state->rs_new_rel), MAIN_FORKNUM, + smgrextend(RelationGetSmgr(state->rs_new_rel, MAIN_FORKNUM), state->rs_blockno, (char *) page, true); state->rs_blockno++; diff --git a/src/backend/access/heap/visibilitymap.c b/src/backend/access/heap/visibilitymap.c index 1d1ca423a98..b03a0dc5d40 100644 --- a/src/backend/access/heap/visibilitymap.c +++ b/src/backend/access/heap/visibilitymap.c @@ -465,7 +465,7 @@ visibilitymap_prepare_truncate(Relation rel, BlockNumber nheapblocks) * If no visibility map has been created yet for this relation, there's * nothing to truncate. */ - if (!smgrexists(RelationGetSmgr(rel), VISIBILITYMAP_FORKNUM)) + if (!smgrexists(RelationGetSmgr(rel, VISIBILITYMAP_FORKNUM))) return InvalidBlockNumber; /* @@ -532,7 +532,7 @@ visibilitymap_prepare_truncate(Relation rel, BlockNumber nheapblocks) else newnblocks = truncBlock; - if (smgrnblocks(RelationGetSmgr(rel), VISIBILITYMAP_FORKNUM) <= newnblocks) + if (smgrnblocks(RelationGetSmgr(rel, VISIBILITYMAP_FORKNUM)) <= newnblocks) { /* nothing to do, the file was already smaller than requested size */ return InvalidBlockNumber; @@ -551,29 +551,29 @@ static Buffer vm_readbuf(Relation rel, BlockNumber blkno, bool extend) { Buffer buf; - SMgrRelation reln; + SMgrFileHandle vm_sfile; /* * Caution: re-using this smgr pointer could fail if the relcache entry * gets closed. It's safe as long as we only do smgr-level operations * between here and the last use of the pointer. */ - reln = RelationGetSmgr(rel); + vm_sfile = RelationGetSmgr(rel, VISIBILITYMAP_FORKNUM); /* * If we haven't cached the size of the visibility map fork yet, check it * first. */ - if (reln->smgr_cached_nblocks[VISIBILITYMAP_FORKNUM] == InvalidBlockNumber) + if (vm_sfile->smgr_cached_nblocks == InvalidBlockNumber) { - if (smgrexists(reln, VISIBILITYMAP_FORKNUM)) - smgrnblocks(reln, VISIBILITYMAP_FORKNUM); + if (smgrexists(vm_sfile)) + smgrnblocks(vm_sfile); else - reln->smgr_cached_nblocks[VISIBILITYMAP_FORKNUM] = 0; + vm_sfile->smgr_cached_nblocks = 0; } /* Handle requests beyond EOF */ - if (blkno >= reln->smgr_cached_nblocks[VISIBILITYMAP_FORKNUM]) + if (blkno >= vm_sfile->smgr_cached_nblocks) { if (extend) vm_extend(rel, blkno + 1); @@ -600,8 +600,7 @@ vm_readbuf(Relation rel, BlockNumber blkno, bool extend) * long as it doesn't depend on the page header having correct contents. * Current usage is safe because PageGetContents() does not require that. */ - buf = ReadBufferExtended(rel, VISIBILITYMAP_FORKNUM, blkno, - RBM_ZERO_ON_ERROR, NULL); + buf = ReadBufferExtended(rel, VISIBILITYMAP_FORKNUM, blkno, RBM_ZERO_ON_ERROR, NULL); if (PageIsNew(BufferGetPage(buf))) { LockBuffer(buf, BUFFER_LOCK_EXCLUSIVE); @@ -621,7 +620,7 @@ vm_extend(Relation rel, BlockNumber vm_nblocks) { BlockNumber vm_nblocks_now; PGAlignedBlock pg; - SMgrRelation reln; + SMgrFileHandle vm_sfile; PageInit((Page) pg.data, BLCKSZ, 0); @@ -642,27 +641,27 @@ vm_extend(Relation rel, BlockNumber vm_nblocks) * gets closed. It's safe as long as we only do smgr-level operations * between here and the last use of the pointer. */ - reln = RelationGetSmgr(rel); + vm_sfile = RelationGetSmgr(rel, VISIBILITYMAP_FORKNUM); /* * Create the file first if it doesn't exist. If smgr_vm_nblocks is * positive then it must exist, no need for an smgrexists call. */ - if ((reln->smgr_cached_nblocks[VISIBILITYMAP_FORKNUM] == 0 || - reln->smgr_cached_nblocks[VISIBILITYMAP_FORKNUM] == InvalidBlockNumber) && - !smgrexists(reln, VISIBILITYMAP_FORKNUM)) - smgrcreate(reln, VISIBILITYMAP_FORKNUM, false); + if ((vm_sfile->smgr_cached_nblocks == 0 || + vm_sfile->smgr_cached_nblocks == InvalidBlockNumber) && + !smgrexists(vm_sfile)) + smgrcreate(vm_sfile, false); /* Invalidate cache so that smgrnblocks() asks the kernel. */ - reln->smgr_cached_nblocks[VISIBILITYMAP_FORKNUM] = InvalidBlockNumber; - vm_nblocks_now = smgrnblocks(reln, VISIBILITYMAP_FORKNUM); + vm_sfile->smgr_cached_nblocks = InvalidBlockNumber; + vm_nblocks_now = smgrnblocks(vm_sfile); /* Now extend the file */ while (vm_nblocks_now < vm_nblocks) { PageSetChecksumInplace((Page) pg.data, vm_nblocks_now); - smgrextend(reln, VISIBILITYMAP_FORKNUM, vm_nblocks_now, pg.data, false); + smgrextend(vm_sfile, vm_nblocks_now, pg.data, false); vm_nblocks_now++; } @@ -673,7 +672,7 @@ vm_extend(Relation rel, BlockNumber vm_nblocks) * to keep checking for creation or extension of the file, which happens * infrequently. */ - CacheInvalidateSmgr(reln->smgr_rlocator); + CacheInvalidateSmgr(rel->rd_locator, rel->rd_backend); UnlockRelationForExtension(rel, ExclusiveLock); } diff --git a/src/backend/access/nbtree/nbtree.c b/src/backend/access/nbtree/nbtree.c index 1cc88da032d..7d9968d241a 100644 --- a/src/backend/access/nbtree/nbtree.c +++ b/src/backend/access/nbtree/nbtree.c @@ -164,9 +164,9 @@ btbuildempty(Relation index) * this even when wal_level=minimal. */ PageSetChecksumInplace(metapage, BTREE_METAPAGE); - smgrwrite(RelationGetSmgr(index), INIT_FORKNUM, BTREE_METAPAGE, + smgrwrite(RelationGetSmgr(index, INIT_FORKNUM), BTREE_METAPAGE, (char *) metapage, true); - log_newpage(&RelationGetSmgr(index)->smgr_rlocator.locator, INIT_FORKNUM, + log_newpage(&index->rd_locator, INIT_FORKNUM, BTREE_METAPAGE, metapage, true); /* @@ -174,7 +174,7 @@ btbuildempty(Relation index) * write did not go through shared_buffers and therefore a concurrent * checkpoint may have moved the redo pointer past our xlog record. */ - smgrimmedsync(RelationGetSmgr(index), INIT_FORKNUM); + smgrimmedsync(RelationGetSmgr(index, INIT_FORKNUM)); } /* diff --git a/src/backend/access/nbtree/nbtsort.c b/src/backend/access/nbtree/nbtsort.c index 67b7b1710c6..216c26cad17 100644 --- a/src/backend/access/nbtree/nbtsort.c +++ b/src/backend/access/nbtree/nbtsort.c @@ -662,7 +662,7 @@ _bt_blwritepage(BTWriteState *wstate, Page page, BlockNumber blkno) if (!wstate->btws_zeropage) wstate->btws_zeropage = (Page) palloc0(BLCKSZ); /* don't set checksum for all-zero page */ - smgrextend(RelationGetSmgr(wstate->index), MAIN_FORKNUM, + smgrextend(RelationGetSmgr(wstate->index, MAIN_FORKNUM), wstate->btws_pages_written++, (char *) wstate->btws_zeropage, true); @@ -677,14 +677,14 @@ _bt_blwritepage(BTWriteState *wstate, Page page, BlockNumber blkno) if (blkno == wstate->btws_pages_written) { /* extending the file... */ - smgrextend(RelationGetSmgr(wstate->index), MAIN_FORKNUM, blkno, + smgrextend(RelationGetSmgr(wstate->index, MAIN_FORKNUM), blkno, (char *) page, true); wstate->btws_pages_written++; } else { /* overwriting a block we zero-filled before */ - smgrwrite(RelationGetSmgr(wstate->index), MAIN_FORKNUM, blkno, + smgrwrite(RelationGetSmgr(wstate->index, MAIN_FORKNUM), blkno, (char *) page, true); } @@ -1431,7 +1431,7 @@ _bt_load(BTWriteState *wstate, BTSpool *btspool, BTSpool *btspool2) * still not be on disk when the crash occurs. */ if (wstate->btws_use_wal) - smgrimmedsync(RelationGetSmgr(wstate->index), MAIN_FORKNUM); + smgrimmedsync(RelationGetSmgr(wstate->index, MAIN_FORKNUM)); } /* diff --git a/src/backend/access/spgist/spginsert.c b/src/backend/access/spgist/spginsert.c index 4ed067f0d93..7ff26de9790 100644 --- a/src/backend/access/spgist/spginsert.c +++ b/src/backend/access/spgist/spginsert.c @@ -155,8 +155,11 @@ spgbuild(Relation heap, Relation index, IndexInfo *indexInfo) void spgbuildempty(Relation index) { + SMgrFileHandle sfile; Page page; + sfile = RelationGetSmgr(index, INIT_FORKNUM); + /* Construct metapage. */ page = (Page) palloc(BLCKSZ); SpGistInitMetapage(page); @@ -169,27 +172,27 @@ spgbuildempty(Relation index) * replayed. */ PageSetChecksumInplace(page, SPGIST_METAPAGE_BLKNO); - smgrwrite(RelationGetSmgr(index), INIT_FORKNUM, SPGIST_METAPAGE_BLKNO, + smgrwrite(sfile, SPGIST_METAPAGE_BLKNO, (char *) page, true); - log_newpage(&(RelationGetSmgr(index))->smgr_rlocator.locator, INIT_FORKNUM, + log_newpage(&index->rd_locator, INIT_FORKNUM, SPGIST_METAPAGE_BLKNO, page, true); /* Likewise for the root page. */ SpGistInitPage(page, SPGIST_LEAF); PageSetChecksumInplace(page, SPGIST_ROOT_BLKNO); - smgrwrite(RelationGetSmgr(index), INIT_FORKNUM, SPGIST_ROOT_BLKNO, + smgrwrite(sfile, SPGIST_ROOT_BLKNO, (char *) page, true); - log_newpage(&(RelationGetSmgr(index))->smgr_rlocator.locator, INIT_FORKNUM, + log_newpage(&index->rd_locator, INIT_FORKNUM, SPGIST_ROOT_BLKNO, page, true); /* Likewise for the null-tuples root page. */ SpGistInitPage(page, SPGIST_LEAF | SPGIST_NULLS); PageSetChecksumInplace(page, SPGIST_NULL_BLKNO); - smgrwrite(RelationGetSmgr(index), INIT_FORKNUM, SPGIST_NULL_BLKNO, + smgrwrite(sfile, SPGIST_NULL_BLKNO, (char *) page, true); - log_newpage(&(RelationGetSmgr(index))->smgr_rlocator.locator, INIT_FORKNUM, + log_newpage(&index->rd_locator, INIT_FORKNUM, SPGIST_NULL_BLKNO, page, true); /* @@ -197,7 +200,7 @@ spgbuildempty(Relation index) * writes did not go through shared buffers and therefore a concurrent * checkpoint may have moved the redo pointer past our xlog record. */ - smgrimmedsync(RelationGetSmgr(index), INIT_FORKNUM); + smgrimmedsync(sfile); } /* diff --git a/src/backend/access/table/tableam.c b/src/backend/access/table/tableam.c index ef0d34fceee..9ebf174b189 100644 --- a/src/backend/access/table/tableam.c +++ b/src/backend/access/table/tableam.c @@ -631,10 +631,10 @@ table_block_relation_size(Relation rel, ForkNumber forkNumber) if (forkNumber == InvalidForkNumber) { for (int i = 0; i < MAX_FORKNUM; i++) - nblocks += smgrnblocks(RelationGetSmgr(rel), i); + nblocks += smgrnblocks(RelationGetSmgr(rel, i)); } else - nblocks = smgrnblocks(RelationGetSmgr(rel), forkNumber); + nblocks = smgrnblocks(RelationGetSmgr(rel, forkNumber)); return nblocks * BLCKSZ; } diff --git a/src/backend/access/transam/clog.c b/src/backend/access/transam/clog.c index 4a431d58767..c08f010cf15 100644 --- a/src/backend/access/transam/clog.c +++ b/src/backend/access/transam/clog.c @@ -41,6 +41,8 @@ #include "miscadmin.h" #include "pg_trace.h" #include "pgstat.h" +#include "storage/bufmgr.h" +#include "storage/buf_internals.h" #include "storage/proc.h" #include "storage/sync.h" @@ -59,7 +61,7 @@ /* We need two bits per xact, so four xacts fit in a byte */ #define CLOG_BITS_PER_XACT 2 #define CLOG_XACTS_PER_BYTE 4 -#define CLOG_XACTS_PER_PAGE (BLCKSZ * CLOG_XACTS_PER_BYTE) +#define CLOG_XACTS_PER_PAGE ((BLCKSZ - SizeOfPageHeaderData) * CLOG_XACTS_PER_BYTE) #define CLOG_XACT_BITMASK ((1 << CLOG_BITS_PER_XACT) - 1) #define TransactionIdToPage(xid) ((xid) / (TransactionId) CLOG_XACTS_PER_PAGE) @@ -81,17 +83,10 @@ */ #define THRESHOLD_SUBTRANS_CLOG_OPT 5 -/* - * Link to shared-memory data structures for CLOG control - */ -static SlruCtlData XactCtlData; - -#define XactCtl (&XactCtlData) - -static int ZeroCLOGPage(int pageno, bool writeXlog); +static Buffer ZeroCLOGPage(int pageno, bool writeXlog); static bool CLOGPagePrecedes(int page1, int page2); -static void WriteZeroPageXlogRec(int pageno); +static XLogRecPtr WriteZeroPageXlogRec(int pageno); static void WriteTruncateXlogRec(int pageno, TransactionId oldestXact, Oid oldestXactDb); static void TransactionIdSetPageStatus(TransactionId xid, int nsubxids, @@ -99,7 +94,7 @@ static void TransactionIdSetPageStatus(TransactionId xid, int nsubxids, XLogRecPtr lsn, int pageno, bool all_xact_same_page); static void TransactionIdSetStatusBit(TransactionId xid, XidStatus status, - XLogRecPtr lsn, int slotno); + XLogRecPtr lsn, Buffer buffer); static void set_status_by_pages(int nsubxids, TransactionId *subxids, XidStatus status, XLogRecPtr lsn); static bool TransactionGroupUpdateXidStatus(TransactionId xid, @@ -339,13 +334,12 @@ TransactionIdSetPageStatusInternal(TransactionId xid, int nsubxids, TransactionId *subxids, XidStatus status, XLogRecPtr lsn, int pageno) { - int slotno; + Buffer buffer; int i; Assert(status == TRANSACTION_STATUS_COMMITTED || status == TRANSACTION_STATUS_ABORTED || (status == TRANSACTION_STATUS_SUB_COMMITTED && !TransactionIdIsValid(xid))); - Assert(LWLockHeldByMeInMode(XactSLRULock, LW_EXCLUSIVE)); /* * If we're doing an async commit (ie, lsn is valid), then we must wait @@ -356,7 +350,8 @@ TransactionIdSetPageStatusInternal(TransactionId xid, int nsubxids, * write-busy, since we don't care if the update reaches disk sooner than * we think. */ - slotno = SimpleLruReadPage(XactCtl, pageno, XLogRecPtrIsInvalid(lsn), xid); + buffer = ReadSlruBuffer(SLRU_CLOG_ID, pageno, RBM_NORMAL); + LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE); /* * Set the main transaction id, if any. @@ -374,25 +369,27 @@ TransactionIdSetPageStatusInternal(TransactionId xid, int nsubxids, { for (i = 0; i < nsubxids; i++) { - Assert(XactCtl->shared->page_number[slotno] == TransactionIdToPage(subxids[i])); + Assert(pageno == TransactionIdToPage(subxids[i])); TransactionIdSetStatusBit(subxids[i], TRANSACTION_STATUS_SUB_COMMITTED, - lsn, slotno); + lsn, buffer); } } /* ... then the main transaction */ - TransactionIdSetStatusBit(xid, status, lsn, slotno); + TransactionIdSetStatusBit(xid, status, lsn, buffer); } /* Set the subtransactions */ for (i = 0; i < nsubxids; i++) { - Assert(XactCtl->shared->page_number[slotno] == TransactionIdToPage(subxids[i])); - TransactionIdSetStatusBit(subxids[i], status, lsn, slotno); + Assert(pageno == TransactionIdToPage(subxids[i])); + TransactionIdSetStatusBit(subxids[i], status, lsn, buffer); } - XactCtl->shared->page_dirty[slotno] = true; + + MarkBufferDirty(buffer); + UnlockReleaseBuffer(buffer); } /* @@ -566,7 +563,7 @@ TransactionGroupUpdateXidStatus(TransactionId xid, XidStatus status, * Must be called with XactSLRULock held */ static void -TransactionIdSetStatusBit(TransactionId xid, XidStatus status, XLogRecPtr lsn, int slotno) +TransactionIdSetStatusBit(TransactionId xid, XidStatus status, XLogRecPtr lsn, Buffer buffer) { int byteno = TransactionIdToByte(xid); int bshift = TransactionIdToBIndex(xid) * CLOG_BITS_PER_XACT; @@ -574,7 +571,10 @@ TransactionIdSetStatusBit(TransactionId xid, XidStatus status, XLogRecPtr lsn, i char byteval; char curval; - byteptr = XactCtl->shared->page_buffer[slotno] + byteno; + Assert(LWLockHeldByMeInMode(BufferDescriptorGetContentLock(GetBufferDescriptor(buffer - 1)), + LW_EXCLUSIVE)); + + byteptr = PageGetContents(BufferGetPage(buffer)) + byteno; curval = (*byteptr >> bshift) & CLOG_XACT_BITMASK; /* @@ -602,8 +602,9 @@ TransactionIdSetStatusBit(TransactionId xid, XidStatus status, XLogRecPtr lsn, i byteval |= (status << bshift); *byteptr = byteval; + /* - * Update the group LSN if the transaction completion LSN is higher. + * Update the buffer LSN if the transaction completion LSN is higher. * * Note: lsn will be invalid when supplied during InRecovery processing, * so we don't need to do anything special to avoid LSN updates during @@ -612,10 +613,8 @@ TransactionIdSetStatusBit(TransactionId xid, XidStatus status, XLogRecPtr lsn, i */ if (!XLogRecPtrIsInvalid(lsn)) { - int lsnindex = GetLSNIndex(slotno, xid); - - if (XactCtl->shared->group_lsn[lsnindex] < lsn) - XactCtl->shared->group_lsn[lsnindex] = lsn; + if (PageGetLSN(BufferGetPage(buffer)) < lsn) + PageSetLSN(BufferGetPage(buffer), lsn); } } @@ -640,67 +639,21 @@ TransactionIdGetStatus(TransactionId xid, XLogRecPtr *lsn) int pageno = TransactionIdToPage(xid); int byteno = TransactionIdToByte(xid); int bshift = TransactionIdToBIndex(xid) * CLOG_BITS_PER_XACT; - int slotno; - int lsnindex; char *byteptr; XidStatus status; + Buffer buffer; - /* lock is acquired by SimpleLruReadPage_ReadOnly */ - - slotno = SimpleLruReadPage_ReadOnly(XactCtl, pageno, xid); - byteptr = XactCtl->shared->page_buffer[slotno] + byteno; + buffer = ReadSlruBuffer(SLRU_CLOG_ID, pageno, RBM_NORMAL); + byteptr = PageGetContents(BufferGetPage(buffer)) + byteno; status = (*byteptr >> bshift) & CLOG_XACT_BITMASK; + *lsn = PageGetLSN(BufferGetPage(buffer)); - lsnindex = GetLSNIndex(slotno, xid); - *lsn = XactCtl->shared->group_lsn[lsnindex]; - - LWLockRelease(XactSLRULock); + ReleaseBuffer(buffer); return status; } -/* - * Number of shared CLOG buffers. - * - * On larger multi-processor systems, it is possible to have many CLOG page - * requests in flight at one time which could lead to disk access for CLOG - * page if the required page is not found in memory. Testing revealed that we - * can get the best performance by having 128 CLOG buffers, more than that it - * doesn't improve performance. - * - * Unconditionally keeping the number of CLOG buffers to 128 did not seem like - * a good idea, because it would increase the minimum amount of shared memory - * required to start, which could be a problem for people running very small - * configurations. The following formula seems to represent a reasonable - * compromise: people with very low values for shared_buffers will get fewer - * CLOG buffers as well, and everyone else will get 128. - */ -Size -CLOGShmemBuffers(void) -{ - return Min(128, Max(4, NBuffers / 512)); -} - -/* - * Initialization of shared memory for CLOG - */ -Size -CLOGShmemSize(void) -{ - return SimpleLruShmemSize(CLOGShmemBuffers(), CLOG_LSNS_PER_PAGE); -} - -void -CLOGShmemInit(void) -{ - XactCtl->PagePrecedes = CLOGPagePrecedes; - SimpleLruInit(XactCtl, "Xact", CLOGShmemBuffers(), CLOG_LSNS_PER_PAGE, - XactSLRULock, "pg_xact", LWTRANCHE_XACT_BUFFER, - SYNC_HANDLER_CLOG); - SlruPagePrecedesUnitTests(XactCtl, CLOG_XACTS_PER_PAGE); -} - /* * This func must be called ONCE on system install. It creates * the initial CLOG segment. (The CLOG directory is assumed to @@ -710,18 +663,15 @@ CLOGShmemInit(void) void BootStrapCLOG(void) { - int slotno; - - LWLockAcquire(XactSLRULock, LW_EXCLUSIVE); + Buffer buffer; /* Create and zero the first page of the commit log */ - slotno = ZeroCLOGPage(0, false); + buffer = ZeroCLOGPage(0, false); /* Make sure it's written out */ - SimpleLruWritePage(XactCtl, slotno); - Assert(!XactCtl->shared->page_dirty[slotno]); + FlushOneBuffer(buffer); - LWLockRelease(XactSLRULock); + UnlockReleaseBuffer(buffer); } /* @@ -733,17 +683,29 @@ BootStrapCLOG(void) * * Control lock must be held at entry, and will be held at exit. */ -static int +static Buffer ZeroCLOGPage(int pageno, bool writeXlog) { - int slotno; + Buffer buffer; + Page page; + XLogRecPtr lsn; + + buffer = ZeroSlruBuffer(SLRU_CLOG_ID, pageno); + + page = BufferGetPage(buffer); - slotno = SimpleLruZeroPage(XactCtl, pageno); + PageInitSLRU(page, BLCKSZ, 0); + lsn = 0; if (writeXlog) - WriteZeroPageXlogRec(pageno); + { + lsn = WriteZeroPageXlogRec(pageno); + } + + PageSetHeaderDataNonRel(page, pageno, lsn, BLCKSZ, PG_METAPAGE_LAYOUT_VERSION); + MarkBufferDirty(buffer); - return slotno; + return buffer; } /* @@ -753,17 +715,6 @@ ZeroCLOGPage(int pageno, bool writeXlog) void StartupCLOG(void) { - TransactionId xid = XidFromFullTransactionId(ShmemVariableCache->nextXid); - int pageno = TransactionIdToPage(xid); - - LWLockAcquire(XactSLRULock, LW_EXCLUSIVE); - - /* - * Initialize our idea of the latest page number. - */ - XactCtl->shared->latest_page_number = pageno; - - LWLockRelease(XactSLRULock); } /* @@ -775,8 +726,6 @@ TrimCLOG(void) TransactionId xid = XidFromFullTransactionId(ShmemVariableCache->nextXid); int pageno = TransactionIdToPage(xid); - LWLockAcquire(XactSLRULock, LW_EXCLUSIVE); - /* * Zero out the remainder of the current clog page. Under normal * circumstances it should be zeroes already, but it seems at least @@ -793,40 +742,25 @@ TrimCLOG(void) { int byteno = TransactionIdToByte(xid); int bshift = TransactionIdToBIndex(xid) * CLOG_BITS_PER_XACT; - int slotno; char *byteptr; + Buffer buffer; - slotno = SimpleLruReadPage(XactCtl, pageno, false, xid); - byteptr = XactCtl->shared->page_buffer[slotno] + byteno; + buffer = ReadSlruBuffer(SLRU_CLOG_ID, pageno, RBM_TRIM); + LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE); + byteptr = PageGetContents(BufferGetPage(buffer)) + byteno; /* Zero so-far-unused positions in the current byte */ *byteptr &= (1 << bshift) - 1; /* Zero the rest of the page */ - MemSet(byteptr + 1, 0, BLCKSZ - byteno - 1); + MemSet(byteptr + 1, 0, BLCKSZ - byteno - MAXALIGN(SizeOfPageHeaderData) - 1); - XactCtl->shared->page_dirty[slotno] = true; - } - LWLockRelease(XactSLRULock); -} + MarkBufferDirty(buffer); -/* - * Perform a checkpoint --- either during shutdown, or on-the-fly - */ -void -CheckPointCLOG(void) -{ - /* - * Write dirty CLOG pages to disk. This may result in sync requests - * queued for later handling by ProcessSyncRequests(), as part of the - * checkpoint. - */ - TRACE_POSTGRESQL_CLOG_CHECKPOINT_START(true); - SimpleLruWriteAll(XactCtl, true); - TRACE_POSTGRESQL_CLOG_CHECKPOINT_DONE(true); + UnlockReleaseBuffer(buffer); + } } - /* * Make sure that CLOG has room for a newly-allocated XID. * @@ -850,12 +784,8 @@ ExtendCLOG(TransactionId newestXact) pageno = TransactionIdToPage(newestXact); - LWLockAcquire(XactSLRULock, LW_EXCLUSIVE); - /* Zero the page and make an XLOG entry about it */ - ZeroCLOGPage(pageno, true); - - LWLockRelease(XactSLRULock); + UnlockReleaseBuffer(ZeroCLOGPage(pageno, true)); } @@ -886,7 +816,8 @@ TruncateCLOG(TransactionId oldestXact, Oid oldestxid_datoid) cutoffPage = TransactionIdToPage(oldestXact); /* Check to see if there's any files that could be removed */ - if (!SlruScanDirectory(XactCtl, SlruScanDirCbReportPresence, &cutoffPage)) + if (!SlruScanDirectory(SLRU_CLOG_ID, CLOGPagePrecedes, + SlruScanDirCbReportPresence, &cutoffPage)) return; /* nothing to remove */ /* @@ -907,7 +838,7 @@ TruncateCLOG(TransactionId oldestXact, Oid oldestxid_datoid) WriteTruncateXlogRec(cutoffPage, oldestXact, oldestxid_datoid); /* Now we can remove the old CLOG segment(s) */ - SimpleLruTruncate(XactCtl, cutoffPage); + SimpleLruTruncate(SLRU_CLOG_ID, CLOGPagePrecedes, cutoffPage); } @@ -948,12 +879,15 @@ CLOGPagePrecedes(int page1, int page2) /* * Write a ZEROPAGE xlog record */ -static void +static XLogRecPtr WriteZeroPageXlogRec(int pageno) { + XLogRecPtr lsn; XLogBeginInsert(); XLogRegisterData((char *) (&pageno), sizeof(int)); - (void) XLogInsert(RM_CLOG_ID, CLOG_ZEROPAGE); + lsn = XLogInsert(RM_CLOG_ID, CLOG_ZEROPAGE); + + return lsn; } /* @@ -992,17 +926,13 @@ clog_redo(XLogReaderState *record) if (info == CLOG_ZEROPAGE) { int pageno; - int slotno; + Buffer buffer; memcpy(&pageno, XLogRecGetData(record), sizeof(int)); - LWLockAcquire(XactSLRULock, LW_EXCLUSIVE); - - slotno = ZeroCLOGPage(pageno, false); - SimpleLruWritePage(XactCtl, slotno); - Assert(!XactCtl->shared->page_dirty[slotno]); - - LWLockRelease(XactSLRULock); + buffer = ZeroCLOGPage(pageno, false); + FlushOneBuffer(buffer); + UnlockReleaseBuffer(buffer); } else if (info == CLOG_TRUNCATE) { @@ -1012,17 +942,8 @@ clog_redo(XLogReaderState *record) AdvanceOldestClogXid(xlrec.oldestXact); - SimpleLruTruncate(XactCtl, xlrec.pageno); + SimpleLruTruncate(SLRU_CLOG_ID, CLOGPagePrecedes, xlrec.pageno); } else elog(PANIC, "clog_redo: unknown op code %u", info); } - -/* - * Entrypoint for sync.c to sync clog files. - */ -int -clogsyncfiletag(const FileTag *ftag, char *path) -{ - return SlruSyncFileTag(XactCtl, ftag, path); -} diff --git a/src/backend/access/transam/commit_ts.c b/src/backend/access/transam/commit_ts.c index b897fabc702..495cfaed03f 100644 --- a/src/backend/access/transam/commit_ts.c +++ b/src/backend/access/transam/commit_ts.c @@ -63,19 +63,14 @@ typedef struct CommitTimestampEntry sizeof(RepOriginId)) #define COMMIT_TS_XACTS_PER_PAGE \ - (BLCKSZ / SizeOfCommitTimestampEntry) + ((BLCKSZ - SizeOfPageHeaderData) / SizeOfCommitTimestampEntry) #define TransactionIdToCTsPage(xid) \ ((xid) / (TransactionId) COMMIT_TS_XACTS_PER_PAGE) #define TransactionIdToCTsEntry(xid) \ ((xid) % (TransactionId) COMMIT_TS_XACTS_PER_PAGE) -/* - * Link to shared-memory data structures for CommitTs control - */ -static SlruCtlData CommitTsCtlData; -#define CommitTsCtl (&CommitTsCtlData) /* * We keep a cache of the last value set in shared memory. @@ -107,11 +102,12 @@ static void SetXidCommitTsInPage(TransactionId xid, int nsubxids, static void TransactionIdSetCommitTs(TransactionId xid, TimestampTz ts, RepOriginId nodeid, int slotno); static void error_commit_ts_disabled(void); -static int ZeroCommitTsPage(int pageno, bool writeXlog); +static Buffer ZeroCommitTsPage(int pageno, bool writeXlog); static bool CommitTsPagePrecedes(int page1, int page2); static void ActivateCommitTs(void); static void DeactivateCommitTs(void); -static void WriteZeroPageXlogRec(int pageno); +static XLogRecPtr WriteZeroPageXlogRec(int pageno); + static void WriteTruncateXlogRec(int pageno, TransactionId oldestXid); /* @@ -216,32 +212,31 @@ SetXidCommitTsInPage(TransactionId xid, int nsubxids, TransactionId *subxids, TimestampTz ts, RepOriginId nodeid, int pageno) { - int slotno; int i; + Buffer buffer; - LWLockAcquire(CommitTsSLRULock, LW_EXCLUSIVE); + buffer = ReadSlruBuffer(SLRU_COMMIT_TS_ID, pageno, RBM_NORMAL); + LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE); - slotno = SimpleLruReadPage(CommitTsCtl, pageno, true, xid); - - TransactionIdSetCommitTs(xid, ts, nodeid, slotno); + TransactionIdSetCommitTs(xid, ts, nodeid, buffer); for (i = 0; i < nsubxids; i++) - TransactionIdSetCommitTs(subxids[i], ts, nodeid, slotno); + TransactionIdSetCommitTs(subxids[i], ts, nodeid, buffer); - CommitTsCtl->shared->page_dirty[slotno] = true; + MarkBufferDirty(buffer); - LWLockRelease(CommitTsSLRULock); + UnlockReleaseBuffer(buffer); } /* * Sets the commit timestamp of a single transaction. - * - * Must be called with CommitTsSLRULock held */ static void TransactionIdSetCommitTs(TransactionId xid, TimestampTz ts, - RepOriginId nodeid, int slotno) + RepOriginId nodeid, Buffer buffer) { int entryno = TransactionIdToCTsEntry(xid); + int pageno = TransactionIdToCTsPage(xid); + CommitTimestampEntry entry; Assert(TransactionIdIsNormal(xid)); @@ -249,9 +244,12 @@ TransactionIdSetCommitTs(TransactionId xid, TimestampTz ts, entry.time = ts; entry.nodeid = nodeid; - memcpy(CommitTsCtl->shared->page_buffer[slotno] + + Assert(xid == pageno * COMMIT_TS_XACTS_PER_PAGE + entryno); + + memcpy(PageGetContents(BufferGetPage(buffer)) + \ SizeOfCommitTimestampEntry * entryno, &entry, SizeOfCommitTimestampEntry); + } /* @@ -268,10 +266,10 @@ TransactionIdGetCommitTsData(TransactionId xid, TimestampTz *ts, { int pageno = TransactionIdToCTsPage(xid); int entryno = TransactionIdToCTsEntry(xid); - int slotno; CommitTimestampEntry entry; TransactionId oldestCommitTsXid; TransactionId newestCommitTsXid; + Buffer buffer; if (!TransactionIdIsValid(xid)) ereport(ERROR, @@ -325,18 +323,19 @@ TransactionIdGetCommitTsData(TransactionId xid, TimestampTz *ts, return false; } - /* lock is acquired by SimpleLruReadPage_ReadOnly */ - slotno = SimpleLruReadPage_ReadOnly(CommitTsCtl, pageno, xid); + buffer = ReadSlruBuffer(SLRU_COMMIT_TS_ID, pageno, RBM_NORMAL); + LockBuffer(buffer, BUFFER_LOCK_SHARE); + memcpy(&entry, - CommitTsCtl->shared->page_buffer[slotno] + - SizeOfCommitTimestampEntry * entryno, - SizeOfCommitTimestampEntry); + PageGetContents(BufferGetPage(buffer)) + \ + SizeOfCommitTimestampEntry * entryno, + SizeOfCommitTimestampEntry); *ts = entry.time; if (nodeid) *nodeid = entry.nodeid; - LWLockRelease(CommitTsSLRULock); + UnlockReleaseBuffer(buffer); return *ts != 0; } @@ -487,27 +486,13 @@ pg_xact_commit_timestamp_origin(PG_FUNCTION_ARGS) PG_RETURN_DATUM(HeapTupleGetDatum(htup)); } -/* - * Number of shared CommitTS buffers. - * - * We use a very similar logic as for the number of CLOG buffers (except we - * scale up twice as fast with shared buffers, and the maximum is twice as - * high); see comments in CLOGShmemBuffers. - */ -Size -CommitTsShmemBuffers(void) -{ - return Min(256, Max(4, NBuffers / 256)); -} - /* * Shared memory sizing for CommitTs */ Size CommitTsShmemSize(void) { - return SimpleLruShmemSize(CommitTsShmemBuffers(), 0) + - sizeof(CommitTimestampShared); + return sizeof(CommitTimestampShared); } /* @@ -519,12 +504,7 @@ CommitTsShmemInit(void) { bool found; - CommitTsCtl->PagePrecedes = CommitTsPagePrecedes; - SimpleLruInit(CommitTsCtl, "CommitTs", CommitTsShmemBuffers(), 0, - CommitTsSLRULock, "pg_commit_ts", - LWTRANCHE_COMMITTS_BUFFER, - SYNC_HANDLER_COMMIT_TS); - SlruPagePrecedesUnitTests(CommitTsCtl, COMMIT_TS_XACTS_PER_PAGE); + SlruPagePrecedesUnitTests(CommitTsPagePrecedes, COMMIT_TS_XACTS_PER_PAGE); commitTsShared = ShmemInitStruct("CommitTs shared", sizeof(CommitTimestampShared), @@ -568,17 +548,26 @@ BootStrapCommitTs(void) * * Control lock must be held at entry, and will be held at exit. */ -static int +static Buffer ZeroCommitTsPage(int pageno, bool writeXlog) { - int slotno; - - slotno = SimpleLruZeroPage(CommitTsCtl, pageno); - + Buffer buffer; + Page page; + XLogRecPtr lsn; + + buffer = ZeroSlruBuffer(SLRU_COMMIT_TS_ID, pageno); + + page = BufferGetPage(buffer); + PageInitSLRU(page, BLCKSZ, 0); + if (writeXlog) - WriteZeroPageXlogRec(pageno); + { + lsn = WriteZeroPageXlogRec(pageno); + PageSetHeaderDataNonRel(page, pageno, lsn, BLCKSZ, PG_METAPAGE_LAYOUT_VERSION); + } + MarkBufferDirty(buffer); - return slotno; + return buffer; } /* @@ -676,13 +665,6 @@ ActivateCommitTs(void) xid = XidFromFullTransactionId(ShmemVariableCache->nextXid); pageno = TransactionIdToCTsPage(xid); - /* - * Re-Initialize our idea of the latest page number. - */ - LWLockAcquire(CommitTsSLRULock, LW_EXCLUSIVE); - CommitTsCtl->shared->latest_page_number = pageno; - LWLockRelease(CommitTsSLRULock); - /* * If CommitTs is enabled, but it wasn't in the previous server run, we * need to set the oldest and newest values to the next Xid; that way, we @@ -705,15 +687,19 @@ ActivateCommitTs(void) LWLockRelease(CommitTsLock); /* Create the current segment file, if necessary */ - if (!SimpleLruDoesPhysicalPageExist(CommitTsCtl, pageno)) + if (!SimpleLruDoesPhysicalPageExist(SLRU_COMMIT_TS_ID, pageno)) { - int slotno; + Buffer buffer; + Page page; - LWLockAcquire(CommitTsSLRULock, LW_EXCLUSIVE); - slotno = ZeroCommitTsPage(pageno, false); - SimpleLruWritePage(CommitTsCtl, slotno); - Assert(!CommitTsCtl->shared->page_dirty[slotno]); - LWLockRelease(CommitTsSLRULock); + buffer = ZeroSlruBuffer(SLRU_COMMIT_TS_ID, pageno); + + page = BufferGetPage(buffer); + PageInitSLRU(page, BLCKSZ, 0); + + MarkBufferDirty(buffer); + FlushOneBuffer(buffer); + UnlockReleaseBuffer(buffer); } /* Change the activation status in shared memory. */ @@ -762,23 +748,9 @@ DeactivateCommitTs(void) * be overwritten anyway when we wrap around, but it seems better to be * tidy.) */ - LWLockAcquire(CommitTsSLRULock, LW_EXCLUSIVE); - (void) SlruScanDirectory(CommitTsCtl, SlruScanDirCbDeleteAll, NULL); - LWLockRelease(CommitTsSLRULock); -} - -/* - * Perform a checkpoint --- either during shutdown, or on-the-fly - */ -void -CheckPointCommitTs(void) -{ - /* - * Write dirty CommitTs pages to disk. This may result in sync requests - * queued for later handling by ProcessSyncRequests(), as part of the - * checkpoint. - */ - SimpleLruWriteAll(CommitTsCtl, true); + (void) SlruScanDirectory(SLRU_COMMIT_TS_ID, + CommitTsPagePrecedes, + SlruScanDirCbDeleteAll, NULL); } /* @@ -816,12 +788,8 @@ ExtendCommitTs(TransactionId newestXact) pageno = TransactionIdToCTsPage(newestXact); - LWLockAcquire(CommitTsSLRULock, LW_EXCLUSIVE); - /* Zero the page and make an XLOG entry about it */ - ZeroCommitTsPage(pageno, !InRecovery); - - LWLockRelease(CommitTsSLRULock); + UnlockReleaseBuffer(ZeroCommitTsPage(pageno, !InRecovery)); } /* @@ -842,7 +810,9 @@ TruncateCommitTs(TransactionId oldestXact) cutoffPage = TransactionIdToCTsPage(oldestXact); /* Check to see if there's any files that could be removed */ - if (!SlruScanDirectory(CommitTsCtl, SlruScanDirCbReportPresence, + if (!SlruScanDirectory(SLRU_COMMIT_TS_ID, + CommitTsPagePrecedes, + SlruScanDirCbReportPresence, &cutoffPage)) return; /* nothing to remove */ @@ -850,7 +820,7 @@ TruncateCommitTs(TransactionId oldestXact) WriteTruncateXlogRec(cutoffPage, oldestXact); /* Now we can remove the old CommitTs segment(s) */ - SimpleLruTruncate(CommitTsCtl, cutoffPage); + SimpleLruTruncate(SLRU_COMMIT_TS_ID, CommitTsPagePrecedes, cutoffPage); } /* @@ -936,12 +906,16 @@ CommitTsPagePrecedes(int page1, int page2) /* * Write a ZEROPAGE xlog record */ -static void +static XLogRecPtr WriteZeroPageXlogRec(int pageno) { + XLogRecPtr lsn; + XLogBeginInsert(); XLogRegisterData((char *) (&pageno), sizeof(int)); - (void) XLogInsert(RM_COMMIT_TS_ID, COMMIT_TS_ZEROPAGE); + lsn = XLogInsert(RM_COMMIT_TS_ID, COMMIT_TS_ZEROPAGE); + + return lsn; } /* @@ -974,17 +948,19 @@ commit_ts_redo(XLogReaderState *record) if (info == COMMIT_TS_ZEROPAGE) { int pageno; - int slotno; + Buffer buffer; + Page page; memcpy(&pageno, XLogRecGetData(record), sizeof(int)); - LWLockAcquire(CommitTsSLRULock, LW_EXCLUSIVE); - - slotno = ZeroCommitTsPage(pageno, false); - SimpleLruWritePage(CommitTsCtl, slotno); - Assert(!CommitTsCtl->shared->page_dirty[slotno]); - - LWLockRelease(CommitTsSLRULock); + buffer = ZeroSlruBuffer(SLRU_COMMIT_TS_ID, pageno); + + page = BufferGetPage(buffer); + PageInitSLRU(page, BLCKSZ, 0); + + MarkBufferDirty(buffer); + FlushOneBuffer(buffer); + UnlockReleaseBuffer(buffer); } else if (info == COMMIT_TS_TRUNCATE) { @@ -992,23 +968,8 @@ commit_ts_redo(XLogReaderState *record) AdvanceOldestCommitTsXid(trunc->oldestXid); - /* - * During XLOG replay, latest_page_number isn't set up yet; insert a - * suitable value to bypass the sanity test in SimpleLruTruncate. - */ - CommitTsCtl->shared->latest_page_number = trunc->pageno; - - SimpleLruTruncate(CommitTsCtl, trunc->pageno); + SimpleLruTruncate(SLRU_COMMIT_TS_ID, CommitTsPagePrecedes, trunc->pageno); } else elog(PANIC, "commit_ts_redo: unknown op code %u", info); } - -/* - * Entrypoint for sync.c to sync commit_ts files. - */ -int -committssyncfiletag(const FileTag *ftag, char *path) -{ - return SlruSyncFileTag(CommitTsCtl, ftag, path); -} diff --git a/src/backend/access/transam/multixact.c b/src/backend/access/transam/multixact.c index e75e1fdf740..74b8d434fb4 100644 --- a/src/backend/access/transam/multixact.c +++ b/src/backend/access/transam/multixact.c @@ -106,7 +106,7 @@ */ /* We need four bytes per offset */ -#define MULTIXACT_OFFSETS_PER_PAGE (BLCKSZ / sizeof(MultiXactOffset)) +#define MULTIXACT_OFFSETS_PER_PAGE ((BLCKSZ - SizeOfPageHeaderData)/ sizeof(MultiXactOffset)) #define MultiXactIdToOffsetPage(xid) \ ((xid) / (MultiXactOffset) MULTIXACT_OFFSETS_PER_PAGE) @@ -138,7 +138,7 @@ /* size in bytes of a complete group */ #define MULTIXACT_MEMBERGROUP_SIZE \ (sizeof(TransactionId) * MULTIXACT_MEMBERS_PER_MEMBERGROUP + MULTIXACT_FLAGBYTES_PER_GROUP) -#define MULTIXACT_MEMBERGROUPS_PER_PAGE (BLCKSZ / MULTIXACT_MEMBERGROUP_SIZE) +#define MULTIXACT_MEMBERGROUPS_PER_PAGE ((BLCKSZ - SizeOfPageHeaderData) / MULTIXACT_MEMBERGROUP_SIZE) #define MULTIXACT_MEMBERS_PER_PAGE \ (MULTIXACT_MEMBERGROUPS_PER_PAGE * MULTIXACT_MEMBERS_PER_MEMBERGROUP) @@ -161,9 +161,9 @@ /* Location (byte offset within page) of flag word for a given member */ #define MXOffsetToFlagsOffset(xid) \ - ((((xid) / (TransactionId) MULTIXACT_MEMBERS_PER_MEMBERGROUP) % \ + (((((xid) / (TransactionId) MULTIXACT_MEMBERS_PER_MEMBERGROUP) % \ (TransactionId) MULTIXACT_MEMBERGROUPS_PER_PAGE) * \ - (TransactionId) MULTIXACT_MEMBERGROUP_SIZE) + (TransactionId) MULTIXACT_MEMBERGROUP_SIZE)) #define MXOffsetToFlagsBitShift(xid) \ (((xid) % (TransactionId) MULTIXACT_MEMBERS_PER_MEMBERGROUP) * \ MXACT_MEMBER_BITS_PER_XACT) @@ -181,15 +181,6 @@ #define PreviousMultiXactId(xid) \ ((xid) == FirstMultiXactId ? MaxMultiXactId : (xid) - 1) -/* - * Links to shared-memory data structures for MultiXact control - */ -static SlruCtlData MultiXactOffsetCtlData; -static SlruCtlData MultiXactMemberCtlData; - -#define MultiXactOffsetCtl (&MultiXactOffsetCtlData) -#define MultiXactMemberCtl (&MultiXactMemberCtlData) - /* * MultiXact state shared across all backends. All this state is protected * by MultiXactGenLock. (We also use MultiXactOffsetSLRULock and @@ -341,8 +332,8 @@ static MemoryContext MXactContext = NULL; /* internal MultiXactId management */ static void MultiXactIdSetOldestVisible(void); static void RecordNewMultiXact(MultiXactId multi, MultiXactOffset offset, - int nmembers, MultiXactMember *members); -static MultiXactId GetNewMultiXactId(int nmembers, MultiXactOffset *offset); + int nmembers, MultiXactMember *members, Buffer * offset_buf_ptr, Buffer * member_bufs); +static MultiXactId GetNewMultiXactId(int nmembers, MultiXactOffset *offset, Buffer * offset_buf, Buffer ** member_bufs); /* MultiXact cache management */ static int mxactMemberComparator(const void *arg1, const void *arg2); @@ -354,19 +345,18 @@ static void mXactCachePut(MultiXactId multi, int nmembers, static char *mxstatus_to_string(MultiXactStatus status); /* management of SLRU infrastructure */ -static int ZeroMultiXactOffsetPage(int pageno, bool writeXlog); -static int ZeroMultiXactMemberPage(int pageno, bool writeXlog); +static Buffer ZeroMultiXactOffsetPage(int pageno, bool writeXlog); +static Buffer ZeroMultiXactMemberPage(int pageno, bool writeXlog); static bool MultiXactOffsetPagePrecedes(int page1, int page2); -static bool MultiXactMemberPagePrecedes(int page1, int page2); static bool MultiXactOffsetPrecedes(MultiXactOffset offset1, MultiXactOffset offset2); -static void ExtendMultiXactOffset(MultiXactId multi); -static void ExtendMultiXactMember(MultiXactOffset offset, int nmembers); +static void ExtendMultiXactOffset(MultiXactId multi, Buffer * buffer); +static void ExtendMultiXactMember(MultiXactOffset offset, int nmembers, Buffer ** member_buffers); static bool MultiXactOffsetWouldWrap(MultiXactOffset boundary, MultiXactOffset start, uint32 distance); static bool SetOffsetVacuumLimit(bool is_startup); static bool find_multixact_start(MultiXactId multi, MultiXactOffset *result); -static void WriteMZeroPageXlogRec(int pageno, uint8 info); +static XLogRecPtr WriteMZeroPageXlogRec(int pageno, uint8 info); static void WriteMTruncateXlogRec(Oid oldestMultiDB, MultiXactId startTruncOff, MultiXactId endTruncOff, @@ -767,6 +757,9 @@ ReadMultiXactIdRange(MultiXactId *oldest, MultiXactId *next) MultiXactId MultiXactIdCreateFromMembers(int nmembers, MultiXactMember *members) { + Buffer * member_bufs; + Buffer offset_buff; + MultiXactId multi; MultiXactOffset offset; xl_multixact_create xlrec; @@ -820,7 +813,8 @@ MultiXactIdCreateFromMembers(int nmembers, MultiXactMember *members) * in vacuum. During vacuum, in particular, it would be unacceptable to * keep OldestMulti set, in case it runs for long. */ - multi = GetNewMultiXactId(nmembers, &offset); + + multi = GetNewMultiXactId(nmembers, &offset, &offset_buff, &member_bufs); /* Make an XLOG entry describing the new MXID. */ xlrec.mid = multi; @@ -840,8 +834,8 @@ MultiXactIdCreateFromMembers(int nmembers, MultiXactMember *members) (void) XLogInsert(RM_MULTIXACT_ID, XLOG_MULTIXACT_CREATE_ID); /* Now enter the information into the OFFSETs and MEMBERs logs */ - RecordNewMultiXact(multi, offset, nmembers, members); - + RecordNewMultiXact(multi, offset, nmembers, members, &offset_buff, member_bufs); + /* Done with critical section */ END_CRIT_SECTION(); @@ -862,40 +856,38 @@ MultiXactIdCreateFromMembers(int nmembers, MultiXactMember *members) */ static void RecordNewMultiXact(MultiXactId multi, MultiXactOffset offset, - int nmembers, MultiXactMember *members) + int nmembers, MultiXactMember *members, Buffer * offset_buf_ptr, Buffer * member_bufs) { int pageno; int prev_pageno; + int min_pageno; int entryno; - int slotno; MultiXactOffset *offptr; int i; - - LWLockAcquire(MultiXactOffsetSLRULock, LW_EXCLUSIVE); + Buffer buffer; + Buffer offset_buf; pageno = MultiXactIdToOffsetPage(multi); entryno = MultiXactIdToOffsetEntry(multi); - /* - * Note: we pass the MultiXactId to SimpleLruReadPage as the "transaction" - * to complain about if there's any I/O error. This is kinda bogus, but - * since the errors will always give the full pathname, it should be clear - * enough that a MultiXactId is really involved. Perhaps someday we'll - * take the trouble to generalize the slru.c error reporting code. - */ - slotno = SimpleLruReadPage(MultiXactOffsetCtl, pageno, true, multi); - offptr = (MultiXactOffset *) MultiXactOffsetCtl->shared->page_buffer[slotno]; - offptr += entryno; + if (offset_buf_ptr) + offset_buf = *offset_buf_ptr; + else + offset_buf = ReadSlruBuffer(SLRU_MULTIXACT_OFFSET_ID, pageno, RBM_NORMAL); - *offptr = offset; + LockBuffer(offset_buf, BUFFER_LOCK_EXCLUSIVE); + - MultiXactOffsetCtl->shared->page_dirty[slotno] = true; + offptr = (MultiXactOffset *) PageGetContents(BufferGetPage(offset_buf)); + offptr += entryno; - /* Exchange our lock */ - LWLockRelease(MultiXactOffsetSLRULock); + *offptr = offset; - LWLockAcquire(MultiXactMemberSLRULock, LW_EXCLUSIVE); + MarkBufferDirty(offset_buf); + UnlockReleaseBuffer(offset_buf); + buffer = InvalidBuffer; + min_pageno = MXOffsetToMemberPage(offset); prev_pageno = -1; for (i = 0; i < nmembers; i++, offset++) @@ -916,27 +908,35 @@ RecordNewMultiXact(MultiXactId multi, MultiXactOffset offset, if (pageno != prev_pageno) { - slotno = SimpleLruReadPage(MultiXactMemberCtl, pageno, true, multi); + if (BufferIsValid(buffer)) + UnlockReleaseBuffer(buffer); + + if (member_bufs) + buffer = member_bufs[pageno - min_pageno]; + else + buffer = ReadSlruBuffer(SLRU_MULTIXACT_MEMBER_ID, pageno, RBM_NORMAL); + + LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE); prev_pageno = pageno; } - memberptr = (TransactionId *) - (MultiXactMemberCtl->shared->page_buffer[slotno] + memberoff); + memberptr = (TransactionId *) (PageGetContents(BufferGetPage(buffer)) + memberoff); *memberptr = members[i].xid; - flagsptr = (uint32 *) - (MultiXactMemberCtl->shared->page_buffer[slotno] + flagsoff); + flagsptr = (uint32 *) (PageGetContents(BufferGetPage(buffer)) + flagsoff); flagsval = *flagsptr; flagsval &= ~(((1 << MXACT_MEMBER_BITS_PER_XACT) - 1) << bshift); flagsval |= (members[i].status << bshift); *flagsptr = flagsval; - MultiXactMemberCtl->shared->page_dirty[slotno] = true; + MarkBufferDirty(buffer); } - LWLockRelease(MultiXactMemberSLRULock); + UnlockReleaseBuffer(buffer); + if (member_bufs != NULL) + pfree(member_bufs); } /* @@ -955,8 +955,11 @@ RecordNewMultiXact(MultiXactId multi, MultiXactOffset offset, * caller must end the critical section after writing SLRU data. */ static MultiXactId -GetNewMultiXactId(int nmembers, MultiXactOffset *offset) +GetNewMultiXactId(int nmembers, MultiXactOffset *offset, Buffer * offset_buf, Buffer ** member_bufs) { + int min_pageno; + int max_pageno; + MultiXactId result; MultiXactOffset nextOffset; @@ -1074,7 +1077,7 @@ GetNewMultiXactId(int nmembers, MultiXactOffset *offset) } /* Make sure there is room for the MXID in the file. */ - ExtendMultiXactOffset(result); + ExtendMultiXactOffset(result, offset_buf); /* * Reserve the members space, similarly to above. Also, be careful not to @@ -1162,7 +1165,12 @@ GetNewMultiXactId(int nmembers, MultiXactOffset *offset) MultiXactState->offsetStopLimit - nextOffset + nmembers), errhint("Execute a database-wide VACUUM in that database with reduced vacuum_multixact_freeze_min_age and vacuum_multixact_freeze_table_age settings."))); - ExtendMultiXactMember(nextOffset, nmembers); + min_pageno = MXOffsetToMemberPage(nextOffset); + max_pageno = MXOffsetToMemberPage(nextOffset + nmembers - 1); + + *member_bufs = (Buffer *) palloc(sizeof(Buffer) * (max_pageno - min_pageno + 1)); + + ExtendMultiXactMember(nextOffset, nmembers, member_bufs); /* * Critical section from here until caller has written the data into the @@ -1228,7 +1236,6 @@ GetMultiXactIdMembers(MultiXactId multi, MultiXactMember **members, int pageno; int prev_pageno; int entryno; - int slotno; MultiXactOffset *offptr; MultiXactOffset offset; int length; @@ -1239,6 +1246,7 @@ GetMultiXactIdMembers(MultiXactId multi, MultiXactMember **members, MultiXactId tmpMXact; MultiXactOffset nextOffset; MultiXactMember *ptr; + Buffer buffer; debug_elog3(DEBUG2, "GetMembers: asked for %u", multi); @@ -1342,13 +1350,12 @@ GetMultiXactIdMembers(MultiXactId multi, MultiXactMember **members, * time on every multixact creation. */ retry: - LWLockAcquire(MultiXactOffsetSLRULock, LW_EXCLUSIVE); - pageno = MultiXactIdToOffsetPage(multi); entryno = MultiXactIdToOffsetEntry(multi); - slotno = SimpleLruReadPage(MultiXactOffsetCtl, pageno, true, multi); - offptr = (MultiXactOffset *) MultiXactOffsetCtl->shared->page_buffer[slotno]; + buffer = ReadSlruBuffer(SLRU_MULTIXACT_OFFSET_ID, pageno, RBM_NORMAL); + LockBuffer(buffer, BUFFER_LOCK_SHARE); + offptr = (MultiXactOffset *) (PageGetContents(BufferGetPage(buffer))); offptr += entryno; offset = *offptr; @@ -1379,16 +1386,20 @@ retry: entryno = MultiXactIdToOffsetEntry(tmpMXact); if (pageno != prev_pageno) - slotno = SimpleLruReadPage(MultiXactOffsetCtl, pageno, true, tmpMXact); + { + UnlockReleaseBuffer(buffer); + buffer = ReadSlruBuffer(SLRU_MULTIXACT_OFFSET_ID, pageno, RBM_NORMAL); + LockBuffer(buffer, BUFFER_LOCK_SHARE); + } - offptr = (MultiXactOffset *) MultiXactOffsetCtl->shared->page_buffer[slotno]; + offptr = (MultiXactOffset *) (PageGetContents(BufferGetPage(buffer))); offptr += entryno; nextMXOffset = *offptr; if (nextMXOffset == 0) { /* Corner case 2: next multixact is still being filled in */ - LWLockRelease(MultiXactOffsetSLRULock); + UnlockReleaseBuffer(buffer); CHECK_FOR_INTERRUPTS(); pg_usleep(1000L); goto retry; @@ -1396,14 +1407,11 @@ retry: length = nextMXOffset - offset; } - - LWLockRelease(MultiXactOffsetSLRULock); + UnlockReleaseBuffer(buffer); + buffer = InvalidBuffer; ptr = (MultiXactMember *) palloc(length * sizeof(MultiXactMember)); - /* Now get the members themselves. */ - LWLockAcquire(MultiXactMemberSLRULock, LW_EXCLUSIVE); - truelength = 0; prev_pageno = -1; for (i = 0; i < length; i++, offset++) @@ -1419,12 +1427,14 @@ retry: if (pageno != prev_pageno) { - slotno = SimpleLruReadPage(MultiXactMemberCtl, pageno, true, multi); + if (BufferIsValid(buffer)) + UnlockReleaseBuffer(buffer); + buffer = ReadSlruBuffer(SLRU_MULTIXACT_MEMBER_ID, pageno, RBM_NORMAL); + LockBuffer(buffer, BUFFER_LOCK_SHARE); prev_pageno = pageno; } - xactptr = (TransactionId *) - (MultiXactMemberCtl->shared->page_buffer[slotno] + memberoff); + xactptr = (TransactionId *) (PageGetContents((BufferGetPage(buffer)) + memberoff)); if (!TransactionIdIsValid(*xactptr)) { @@ -1435,14 +1445,13 @@ retry: flagsoff = MXOffsetToFlagsOffset(offset); bshift = MXOffsetToFlagsBitShift(offset); - flagsptr = (uint32 *) (MultiXactMemberCtl->shared->page_buffer[slotno] + flagsoff); + flagsptr = (uint32 *) (PageGetContents(BufferGetPage(buffer)) + flagsoff); ptr[truelength].xid = *xactptr; ptr[truelength].status = (*flagsptr >> bshift) & MXACT_MEMBER_XACT_BITMASK; truelength++; } - - LWLockRelease(MultiXactMemberSLRULock); + UnlockReleaseBuffer(buffer); /* A multixid with zero members should not happen */ Assert(truelength > 0); @@ -1834,8 +1843,6 @@ MultiXactShmemSize(void) mul_size(sizeof(MultiXactId) * 2, MaxOldestSlot)) size = SHARED_MULTIXACT_STATE_SIZE; - size = add_size(size, SimpleLruShmemSize(NUM_MULTIXACTOFFSET_BUFFERS, 0)); - size = add_size(size, SimpleLruShmemSize(NUM_MULTIXACTMEMBER_BUFFERS, 0)); return size; } @@ -1847,22 +1854,6 @@ MultiXactShmemInit(void) debug_elog2(DEBUG2, "Shared Memory Init for MultiXact"); - MultiXactOffsetCtl->PagePrecedes = MultiXactOffsetPagePrecedes; - MultiXactMemberCtl->PagePrecedes = MultiXactMemberPagePrecedes; - - SimpleLruInit(MultiXactOffsetCtl, - "MultiXactOffset", NUM_MULTIXACTOFFSET_BUFFERS, 0, - MultiXactOffsetSLRULock, "pg_multixact/offsets", - LWTRANCHE_MULTIXACTOFFSET_BUFFER, - SYNC_HANDLER_MULTIXACT_OFFSET); - SlruPagePrecedesUnitTests(MultiXactOffsetCtl, MULTIXACT_OFFSETS_PER_PAGE); - SimpleLruInit(MultiXactMemberCtl, - "MultiXactMember", NUM_MULTIXACTMEMBER_BUFFERS, 0, - MultiXactMemberSLRULock, "pg_multixact/members", - LWTRANCHE_MULTIXACTMEMBER_BUFFER, - SYNC_HANDLER_MULTIXACT_MEMBER); - /* doesn't call SimpleLruTruncate() or meet criteria for unit tests */ - /* Initialize our shared state struct */ MultiXactState = ShmemInitStruct("Shared MultiXact State", SHARED_MULTIXACT_STATE_SIZE, @@ -1893,29 +1884,17 @@ MultiXactShmemInit(void) void BootStrapMultiXact(void) { - int slotno; - - LWLockAcquire(MultiXactOffsetSLRULock, LW_EXCLUSIVE); + Buffer buffer; /* Create and zero the first page of the offsets log */ - slotno = ZeroMultiXactOffsetPage(0, false); - - /* Make sure it's written out */ - SimpleLruWritePage(MultiXactOffsetCtl, slotno); - Assert(!MultiXactOffsetCtl->shared->page_dirty[slotno]); - - LWLockRelease(MultiXactOffsetSLRULock); - - LWLockAcquire(MultiXactMemberSLRULock, LW_EXCLUSIVE); + buffer = ZeroMultiXactOffsetPage(0, false); + FlushOneBuffer(buffer); + UnlockReleaseBuffer(buffer); /* Create and zero the first page of the members log */ - slotno = ZeroMultiXactMemberPage(0, false); - - /* Make sure it's written out */ - SimpleLruWritePage(MultiXactMemberCtl, slotno); - Assert(!MultiXactMemberCtl->shared->page_dirty[slotno]); - - LWLockRelease(MultiXactMemberSLRULock); + buffer = ZeroMultiXactMemberPage(0, false); + FlushOneBuffer(buffer); + UnlockReleaseBuffer(buffer); } /* @@ -1927,33 +1906,54 @@ BootStrapMultiXact(void) * * Control lock must be held at entry, and will be held at exit. */ -static int +static Buffer ZeroMultiXactOffsetPage(int pageno, bool writeXlog) { - int slotno; + Buffer buffer; + Page page; + XLogRecPtr recptr; - slotno = SimpleLruZeroPage(MultiXactOffsetCtl, pageno); + buffer = ZeroSlruBuffer(SLRU_MULTIXACT_OFFSET_ID, pageno); + page = BufferGetPage(buffer); + PageInitSLRU(page, BLCKSZ, 0); + recptr = 0; + if (writeXlog) - WriteMZeroPageXlogRec(pageno, XLOG_MULTIXACT_ZERO_OFF_PAGE); + { + recptr = WriteMZeroPageXlogRec(pageno, XLOG_MULTIXACT_ZERO_OFF_PAGE); + } + + PageSetHeaderDataNonRel(page, pageno, recptr, BLCKSZ, PG_METAPAGE_LAYOUT_VERSION); + MarkBufferDirty(buffer); - return slotno; + return buffer; } /* * Ditto, for MultiXactMember */ -static int +static Buffer ZeroMultiXactMemberPage(int pageno, bool writeXlog) { - int slotno; - - slotno = SimpleLruZeroPage(MultiXactMemberCtl, pageno); - + Buffer buffer; + Page page; + XLogRecPtr recptr; + + buffer = ZeroSlruBuffer(SLRU_MULTIXACT_MEMBER_ID, pageno); + page = BufferGetPage(buffer); + PageInitSLRU(page, BLCKSZ, 0); + + recptr = 0; if (writeXlog) - WriteMZeroPageXlogRec(pageno, XLOG_MULTIXACT_ZERO_MEM_PAGE); - - return slotno; + { + recptr = WriteMZeroPageXlogRec(pageno, XLOG_MULTIXACT_ZERO_MEM_PAGE); + } + + PageSetHeaderDataNonRel(page, pageno, recptr, BLCKSZ, PG_METAPAGE_LAYOUT_VERSION); + MarkBufferDirty(buffer); + + return buffer; } /* @@ -1978,22 +1978,14 @@ MaybeExtendOffsetSlru(void) pageno = MultiXactIdToOffsetPage(MultiXactState->nextMXact); - LWLockAcquire(MultiXactOffsetSLRULock, LW_EXCLUSIVE); - - if (!SimpleLruDoesPhysicalPageExist(MultiXactOffsetCtl, pageno)) + if (!SimpleLruDoesPhysicalPageExist(SLRU_MULTIXACT_OFFSET_ID, pageno)) { - int slotno; + Buffer buffer; - /* - * Fortunately for us, SimpleLruWritePage is already prepared to deal - * with creating a new segment file even if the page we're writing is - * not the first in it, so this is enough. - */ - slotno = ZeroMultiXactOffsetPage(pageno, false); - SimpleLruWritePage(MultiXactOffsetCtl, slotno); + buffer = ZeroMultiXactOffsetPage(pageno, false); + FlushOneBuffer(buffer); + UnlockReleaseBuffer(buffer); } - - LWLockRelease(MultiXactOffsetSLRULock); } /* @@ -2007,21 +1999,6 @@ MaybeExtendOffsetSlru(void) void StartupMultiXact(void) { - MultiXactId multi = MultiXactState->nextMXact; - MultiXactOffset offset = MultiXactState->nextOffset; - int pageno; - - /* - * Initialize offset's idea of the latest page number. - */ - pageno = MultiXactIdToOffsetPage(multi); - MultiXactOffsetCtl->shared->latest_page_number = pageno; - - /* - * Initialize member's idea of the latest page number. - */ - pageno = MXOffsetToMemberPage(offset); - MultiXactMemberCtl->shared->latest_page_number = pageno; } /* @@ -2045,14 +2022,8 @@ TrimMultiXact(void) oldestMXactDB = MultiXactState->oldestMultiXactDB; LWLockRelease(MultiXactGenLock); - /* Clean up offsets state */ - LWLockAcquire(MultiXactOffsetSLRULock, LW_EXCLUSIVE); - - /* - * (Re-)Initialize our idea of the latest page number for offsets. - */ pageno = MultiXactIdToOffsetPage(nextMXact); - MultiXactOffsetCtl->shared->latest_page_number = pageno; + /* * Zero out the remainder of the current offsets page. See notes in @@ -2065,46 +2036,39 @@ TrimMultiXact(void) entryno = MultiXactIdToOffsetEntry(nextMXact); if (entryno != 0) { - int slotno; MultiXactOffset *offptr; + Buffer buffer; - slotno = SimpleLruReadPage(MultiXactOffsetCtl, pageno, true, nextMXact); - offptr = (MultiXactOffset *) MultiXactOffsetCtl->shared->page_buffer[slotno]; + buffer = ReadSlruBuffer(SLRU_MULTIXACT_OFFSET_ID, pageno, RBM_TRIM); + LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE); + offptr = (MultiXactOffset *) (PageGetContents(BufferGetPage(buffer))); offptr += entryno; - MemSet(offptr, 0, BLCKSZ - (entryno * sizeof(MultiXactOffset))); + MemSet(offptr, 0, BLCKSZ - MAXALIGN(SizeOfPageHeaderData) - (entryno * sizeof(MultiXactOffset))); - MultiXactOffsetCtl->shared->page_dirty[slotno] = true; + MarkBufferDirty(buffer); + UnlockReleaseBuffer(buffer); } - LWLockRelease(MultiXactOffsetSLRULock); - - /* And the same for members */ - LWLockAcquire(MultiXactMemberSLRULock, LW_EXCLUSIVE); - - /* - * (Re-)Initialize our idea of the latest page number for members. - */ - pageno = MXOffsetToMemberPage(offset); - MultiXactMemberCtl->shared->latest_page_number = pageno; - /* * Zero out the remainder of the current members page. See notes in * TrimCLOG() for motivation. */ + + pageno = MXOffsetToMemberPage(offset); flagsoff = MXOffsetToFlagsOffset(offset); if (flagsoff != 0) { - int slotno; TransactionId *xidptr; int memberoff; + Buffer buffer; memberoff = MXOffsetToMemberOffset(offset); - slotno = SimpleLruReadPage(MultiXactMemberCtl, pageno, true, offset); - xidptr = (TransactionId *) - (MultiXactMemberCtl->shared->page_buffer[slotno] + memberoff); + buffer = ReadSlruBuffer(SLRU_MULTIXACT_MEMBER_ID, pageno, RBM_TRIM); + LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE); + xidptr = (TransactionId *) (PageGetContents(BufferGetPage(buffer)) + memberoff); - MemSet(xidptr, 0, BLCKSZ - memberoff); + MemSet(xidptr, 0, BLCKSZ - memberoff - MAXALIGN(SizeOfPageHeaderData)); /* * Note: we don't need to zero out the flag bits in the remaining @@ -2112,11 +2076,10 @@ TrimMultiXact(void) * writing. */ - MultiXactMemberCtl->shared->page_dirty[slotno] = true; + MarkBufferDirty(buffer); + UnlockReleaseBuffer(buffer); } - LWLockRelease(MultiXactMemberSLRULock); - /* signal that we're officially up */ LWLockAcquire(MultiXactGenLock, LW_EXCLUSIVE); MultiXactState->finishedStartup = true; @@ -2148,25 +2111,6 @@ MultiXactGetCheckptMulti(bool is_shutdown, *nextMulti, *nextMultiOffset, *oldestMulti, *oldestMultiDB); } -/* - * Perform a checkpoint --- either during shutdown, or on-the-fly - */ -void -CheckPointMultiXact(void) -{ - TRACE_POSTGRESQL_MULTIXACT_CHECKPOINT_START(true); - - /* - * Write dirty MultiXact pages to disk. This may result in sync requests - * queued for later handling by ProcessSyncRequests(), as part of the - * checkpoint. - */ - SimpleLruWriteAll(MultiXactOffsetCtl, true); - SimpleLruWriteAll(MultiXactMemberCtl, true); - - TRACE_POSTGRESQL_MULTIXACT_CHECKPOINT_DONE(true); -} - /* * Set the next-to-be-assigned MultiXactId and offset * @@ -2401,26 +2345,30 @@ MultiXactAdvanceOldest(MultiXactId oldestMulti, Oid oldestMultiDB) * room in shared memory. */ static void -ExtendMultiXactOffset(MultiXactId multi) +ExtendMultiXactOffset(MultiXactId multi, Buffer * buffer) { int pageno; + /* - * No work except at first MultiXactId of a page. But beware: just after - * wraparound, the first MultiXactId of page zero is FirstMultiXactId. + * Make a ReadBuffer call for the page we need beforehand so that we don't need + * to malloc later. + * If we're at the first MultiXactId of a page, make sure we also zero the page */ + + pageno = MultiXactIdToOffsetPage(multi); if (MultiXactIdToOffsetEntry(multi) != 0 && multi != FirstMultiXactId) + { + /* make a read buffer call to enlarge the resource owner */ + *buffer = ReadSlruBuffer(SLRU_MULTIXACT_OFFSET_ID, pageno, RBM_NORMAL); return; - - pageno = MultiXactIdToOffsetPage(multi); - - LWLockAcquire(MultiXactOffsetSLRULock, LW_EXCLUSIVE); - - /* Zero the page and make an XLOG entry about it */ - ZeroMultiXactOffsetPage(pageno, true); - - LWLockRelease(MultiXactOffsetSLRULock); + } else + { + /* Zero the page and make an XLOG entry about it */ + *buffer = ZeroMultiXactOffsetPage(pageno, true); + LockBuffer(*buffer, BUFFER_LOCK_UNLOCK); /* release lock but don't unpin */ + } } /* @@ -2431,7 +2379,7 @@ ExtendMultiXactOffset(MultiXactId multi) * same comments apply. */ static void -ExtendMultiXactMember(MultiXactOffset offset, int nmembers) +ExtendMultiXactMember(MultiXactOffset offset, int nmembers, Buffer ** buffers) { /* * It's possible that the members span more than one page of the members @@ -2439,10 +2387,17 @@ ExtendMultiXactMember(MultiXactOffset offset, int nmembers) * optimal if the members span several pages, but that seems unusual * enough to not worry much about. */ + int min_pageno; + + min_pageno = MXOffsetToMemberPage(offset); while (nmembers > 0) { + Buffer buf; + int flagsoff; int flagsbit; + int pageno; + uint32 difference; /* @@ -2450,20 +2405,24 @@ ExtendMultiXactMember(MultiXactOffset offset, int nmembers) */ flagsoff = MXOffsetToFlagsOffset(offset); flagsbit = MXOffsetToFlagsBitShift(offset); + pageno = MXOffsetToMemberPage(offset); + + + if (flagsoff == 0 && flagsbit == 0) { - int pageno; - - pageno = MXOffsetToMemberPage(offset); - - LWLockAcquire(MultiXactMemberSLRULock, LW_EXCLUSIVE); - /* Zero the page and make an XLOG entry about it */ - ZeroMultiXactMemberPage(pageno, true); - - LWLockRelease(MultiXactMemberSLRULock); + buf = ZeroMultiXactMemberPage(pageno, true); + LockBuffer(buf, BUFFER_LOCK_UNLOCK); + } else + { + /* do a read buffer call to allocate space beforehand */ + buf = ReadSlruBuffer(SLRU_MULTIXACT_MEMBER_ID, pageno, RBM_NORMAL); } + if (buffers) + (*buffers)[pageno - min_pageno] = buf; + /* * Compute the number of items till end of current page. Careful: if * addition of unsigned ints wraps around, we're at the last page of @@ -2737,8 +2696,8 @@ find_multixact_start(MultiXactId multi, MultiXactOffset *result) MultiXactOffset offset; int pageno; int entryno; - int slotno; MultiXactOffset *offptr; + Buffer buffer; Assert(MultiXactState->finishedStartup); @@ -2746,20 +2705,19 @@ find_multixact_start(MultiXactId multi, MultiXactOffset *result) entryno = MultiXactIdToOffsetEntry(multi); /* - * Write out dirty data, so PhysicalPageExists can work correctly. + * Cope with missing/bogus oldest MultiXact in inconsistent states (see + * commit 068cfadf9). */ - SimpleLruWriteAll(MultiXactOffsetCtl, true); - SimpleLruWriteAll(MultiXactMemberCtl, true); - - if (!SimpleLruDoesPhysicalPageExist(MultiXactOffsetCtl, pageno)) + if (!ProbeSlruBuffer(SLRU_MULTIXACT_OFFSET_ID, pageno) && + !SimpleLruDoesPhysicalPageExist(SLRU_MULTIXACT_OFFSET_ID, pageno)) return false; - /* lock is acquired by SimpleLruReadPage_ReadOnly */ - slotno = SimpleLruReadPage_ReadOnly(MultiXactOffsetCtl, pageno, multi); - offptr = (MultiXactOffset *) MultiXactOffsetCtl->shared->page_buffer[slotno]; + buffer = ReadSlruBuffer(SLRU_MULTIXACT_OFFSET_ID, pageno, RBM_NORMAL); + LockBuffer(buffer, BUFFER_LOCK_SHARE); + offptr = (MultiXactOffset *) (PageGetContents(BufferGetPage(buffer))); offptr += entryno; offset = *offptr; - LWLockRelease(MultiXactOffsetSLRULock); + UnlockReleaseBuffer(buffer); *result = offset; return true; @@ -2862,12 +2820,13 @@ typedef struct mxtruncinfo * This callback determines the earliest existing page number. */ static bool -SlruScanDirCbFindEarliest(SlruCtl ctl, char *filename, int segpage, void *data) +SlruScanDirCbFindEarliest(int slru_id, SlruPagePrecedesFunction PagePrecedes, + char *filename, int segpage, void *data) { mxtruncinfo *trunc = (mxtruncinfo *) data; if (trunc->earliestExistingPage == -1 || - ctl->PagePrecedes(segpage, trunc->earliestExistingPage)) + PagePrecedes(segpage, trunc->earliestExistingPage)) { trunc->earliestExistingPage = segpage; } @@ -2899,7 +2858,7 @@ PerformMembersTruncation(MultiXactOffset oldestOffset, MultiXactOffset newOldest while (segment != endsegment) { elog(DEBUG2, "truncating multixact members segment %x", segment); - SlruDeleteSegment(MultiXactMemberCtl, segment); + SlruDeleteSegment(SLRU_MULTIXACT_MEMBER_ID, segment); /* move to next segment, handling wraparound correctly */ if (segment == maxsegment) @@ -2922,7 +2881,8 @@ PerformOffsetsTruncation(MultiXactId oldestMulti, MultiXactId newOldestMulti) * didn't subtract one, we'd trigger SimpleLruTruncate's wraparound * detection. */ - SimpleLruTruncate(MultiXactOffsetCtl, + SimpleLruTruncate(SLRU_MULTIXACT_OFFSET_ID, + MultiXactOffsetPagePrecedes, MultiXactIdToOffsetPage(PreviousMultiXactId(newOldestMulti))); } @@ -2996,7 +2956,9 @@ TruncateMultiXact(MultiXactId newOldestMulti, Oid newOldestMultiDB) * been truncated away, and we crashed before updating oldestMulti. */ trunc.earliestExistingPage = -1; - SlruScanDirectory(MultiXactOffsetCtl, SlruScanDirCbFindEarliest, &trunc); + SlruScanDirectory(SLRU_MULTIXACT_OFFSET_ID, + MultiXactOffsetPagePrecedes, + SlruScanDirCbFindEarliest, &trunc); earliest = trunc.earliestExistingPage * MULTIXACT_OFFSETS_PER_PAGE; if (earliest < FirstMultiXactId) earliest = FirstMultiXactId; @@ -3128,24 +3090,6 @@ MultiXactOffsetPagePrecedes(int page1, int page2) multi2 + MULTIXACT_OFFSETS_PER_PAGE - 1)); } -/* - * Decide whether a MultiXactMember page number is "older" for truncation - * purposes. There is no "invalid offset number" so use the numbers verbatim. - */ -static bool -MultiXactMemberPagePrecedes(int page1, int page2) -{ - MultiXactOffset offset1; - MultiXactOffset offset2; - - offset1 = ((MultiXactOffset) page1) * MULTIXACT_MEMBERS_PER_PAGE; - offset2 = ((MultiXactOffset) page2) * MULTIXACT_MEMBERS_PER_PAGE; - - return (MultiXactOffsetPrecedes(offset1, offset2) && - MultiXactOffsetPrecedes(offset1, - offset2 + MULTIXACT_MEMBERS_PER_PAGE - 1)); -} - /* * Decide which of two MultiXactIds is earlier. * @@ -3190,12 +3134,16 @@ MultiXactOffsetPrecedes(MultiXactOffset offset1, MultiXactOffset offset2) * Write an xlog record reflecting the zeroing of either a MEMBERs or * OFFSETs page (info shows which) */ -static void +static XLogRecPtr WriteMZeroPageXlogRec(int pageno, uint8 info) { + XLogRecPtr recptr; + XLogBeginInsert(); XLogRegisterData((char *) (&pageno), sizeof(int)); - (void) XLogInsert(RM_MULTIXACT_ID, info); + recptr = XLogInsert(RM_MULTIXACT_ID, info); + + return recptr; } /* @@ -3240,32 +3188,18 @@ multixact_redo(XLogReaderState *record) if (info == XLOG_MULTIXACT_ZERO_OFF_PAGE) { int pageno; - int slotno; memcpy(&pageno, XLogRecGetData(record), sizeof(int)); - LWLockAcquire(MultiXactOffsetSLRULock, LW_EXCLUSIVE); - - slotno = ZeroMultiXactOffsetPage(pageno, false); - SimpleLruWritePage(MultiXactOffsetCtl, slotno); - Assert(!MultiXactOffsetCtl->shared->page_dirty[slotno]); - - LWLockRelease(MultiXactOffsetSLRULock); + UnlockReleaseBuffer(ZeroMultiXactOffsetPage(pageno, false)); } else if (info == XLOG_MULTIXACT_ZERO_MEM_PAGE) { int pageno; - int slotno; memcpy(&pageno, XLogRecGetData(record), sizeof(int)); - LWLockAcquire(MultiXactMemberSLRULock, LW_EXCLUSIVE); - - slotno = ZeroMultiXactMemberPage(pageno, false); - SimpleLruWritePage(MultiXactMemberCtl, slotno); - Assert(!MultiXactMemberCtl->shared->page_dirty[slotno]); - - LWLockRelease(MultiXactMemberSLRULock); + UnlockReleaseBuffer(ZeroMultiXactMemberPage(pageno, false)); } else if (info == XLOG_MULTIXACT_CREATE_ID) { @@ -3276,7 +3210,7 @@ multixact_redo(XLogReaderState *record) /* Store the data back into the SLRU files */ RecordNewMultiXact(xlrec->mid, xlrec->moff, xlrec->nmembers, - xlrec->members); + xlrec->members, NULL, NULL); /* Make sure nextMXact/nextOffset are beyond what this record has */ MultiXactAdvanceNextMXact(xlrec->mid + 1, @@ -3299,7 +3233,6 @@ multixact_redo(XLogReaderState *record) else if (info == XLOG_MULTIXACT_TRUNCATE_ID) { xl_multixact_truncate xlrec; - int pageno; memcpy(&xlrec, XLogRecGetData(record), SizeOfMultiXactTruncate); @@ -3325,13 +3258,6 @@ multixact_redo(XLogReaderState *record) PerformMembersTruncation(xlrec.startTruncMemb, xlrec.endTruncMemb); - /* - * During XLOG replay, latest_page_number isn't necessarily set up - * yet; insert a suitable value to bypass the sanity test in - * SimpleLruTruncate. - */ - pageno = MultiXactIdToOffsetPage(xlrec.endTruncOff); - MultiXactOffsetCtl->shared->latest_page_number = pageno; PerformOffsetsTruncation(xlrec.startTruncOff, xlrec.endTruncOff); LWLockRelease(MultiXactTruncationLock); @@ -3401,21 +3327,3 @@ pg_get_multixact_members(PG_FUNCTION_ARGS) SRF_RETURN_DONE(funccxt); } - -/* - * Entrypoint for sync.c to sync offsets files. - */ -int -multixactoffsetssyncfiletag(const FileTag *ftag, char *path) -{ - return SlruSyncFileTag(MultiXactOffsetCtl, ftag, path); -} - -/* - * Entrypoint for sync.c to sync members files. - */ -int -multixactmemberssyncfiletag(const FileTag *ftag, char *path) -{ - return SlruSyncFileTag(MultiXactMemberCtl, ftag, path); -} diff --git a/src/backend/access/transam/slru.c b/src/backend/access/transam/slru.c index 5ab86238a92..00f178c4293 100644 --- a/src/backend/access/transam/slru.c +++ b/src/backend/access/transam/slru.c @@ -1,41 +1,9 @@ /*------------------------------------------------------------------------- * * slru.c - * Simple LRU buffering for transaction status logfiles + * Simple buffering for transaction status logfiles * - * We use a simple least-recently-used scheme to manage a pool of page - * buffers. Under ordinary circumstances we expect that write - * traffic will occur mostly to the latest page (and to the just-prior - * page, soon after a page transition). Read traffic will probably touch - * a larger span of pages, but in any case a fairly small number of page - * buffers should be sufficient. So, we just search the buffers using plain - * linear search; there's no need for a hashtable or anything fancy. - * The management algorithm is straight LRU except that we will never swap - * out the latest page (since we know it's going to be hit again eventually). - * - * We use a control LWLock to protect the shared data structures, plus - * per-buffer LWLocks that synchronize I/O for each buffer. The control lock - * must be held to examine or modify any shared state. A process that is - * reading in or writing out a page buffer does not hold the control lock, - * only the per-buffer lock for the buffer it is working on. - * - * "Holding the control lock" means exclusive lock in all cases except for - * SimpleLruReadPage_ReadOnly(); see comments for SlruRecentlyUsed() for - * the implications of that. - * - * When initiating I/O on a buffer, we acquire the per-buffer lock exclusively - * before releasing the control lock. The per-buffer lock is released after - * completing the I/O, re-acquiring the control lock, and updating the shared - * state. (Deadlock is not possible here, because we never try to initiate - * I/O when someone else is already doing I/O on the same buffer.) - * To wait for I/O to complete, release the control lock, acquire the - * per-buffer lock in shared mode, immediately release the per-buffer lock, - * reacquire the control lock, and then recheck state (since arbitrary things - * could have happened while we didn't have the lock). - * - * As with the regular buffer manager, it is possible for another process - * to re-dirty a page that is currently being written out. This is handled - * by re-setting the page's page_dirty flag. + * XXX write me * * * Portions Copyright (c) 1996-2023, PostgreSQL Global Development Group @@ -60,562 +28,34 @@ #include "storage/fd.h" #include "storage/shmem.h" -#define SlruFileName(ctl, path, seg) \ - snprintf(path, MAXPGPATH, "%s/%04X", (ctl)->Dir, seg) - /* - * During SimpleLruWriteAll(), we will usually not need to write more than one - * or two physical files, but we may need to write several pages per file. We - * can consolidate the I/O requests by leaving files open until control returns - * to SimpleLruWriteAll(). This data structure remembers which files are open. + * SLRU ID to path mapping */ -#define MAX_WRITEALL_BUFFERS 16 +#define PG_SLRU(symname,name,path,synchronize) \ + path, -typedef struct SlruWriteAllData +static char *slru_dirs[] = { - int num_files; /* # files actually open */ - int fd[MAX_WRITEALL_BUFFERS]; /* their FD's */ - int segno[MAX_WRITEALL_BUFFERS]; /* their log seg#s */ -} SlruWriteAllData; - -typedef struct SlruWriteAllData *SlruWriteAll; +#include "access/slrulist.h" +}; /* - * Populate a file tag describing a segment file. We only use the segment - * number, since we can derive everything else we need by having separate - * sync handler functions for clog, multixact etc. + * We'll maintain a little cache of recently seen buffers, to try to avoid the + * buffer mapping table on repeat access (ie the busy end of the CLOG). One + * entry per SLRU. */ -#define INIT_SLRUFILETAG(a,xx_handler,xx_segno) \ -( \ - memset(&(a), 0, sizeof(FileTag)), \ - (a).handler = (xx_handler), \ - (a).segno = (xx_segno) \ -) - -/* - * Macro to mark a buffer slot "most recently used". Note multiple evaluation - * of arguments! - * - * The reason for the if-test is that there are often many consecutive - * accesses to the same page (particularly the latest page). By suppressing - * useless increments of cur_lru_count, we reduce the probability that old - * pages' counts will "wrap around" and make them appear recently used. - * - * We allow this code to be executed concurrently by multiple processes within - * SimpleLruReadPage_ReadOnly(). As long as int reads and writes are atomic, - * this should not cause any completely-bogus values to enter the computation. - * However, it is possible for either cur_lru_count or individual - * page_lru_count entries to be "reset" to lower values than they should have, - * in case a process is delayed while it executes this macro. With care in - * SlruSelectLRUPage(), this does little harm, and in any case the absolute - * worst possible consequence is a nonoptimal choice of page to evict. The - * gain from allowing concurrent reads of SLRU pages seems worth it. - */ -#define SlruRecentlyUsed(shared, slotno) \ - do { \ - int new_lru_count = (shared)->cur_lru_count; \ - if (new_lru_count != (shared)->page_lru_count[slotno]) { \ - (shared)->cur_lru_count = ++new_lru_count; \ - (shared)->page_lru_count[slotno] = new_lru_count; \ - } \ - } while (0) - -/* Saved info for SlruReportIOError */ -typedef enum -{ - SLRU_OPEN_FAILED, - SLRU_SEEK_FAILED, - SLRU_READ_FAILED, - SLRU_WRITE_FAILED, - SLRU_FSYNC_FAILED, - SLRU_CLOSE_FAILED -} SlruErrorCause; +struct SlruRecentBuffer { + int pageno; + Buffer recent_buffer; +}; -static SlruErrorCause slru_errcause; -static int slru_errno; +static struct SlruRecentBuffer slru_recent_buffers[SLRU_NEXT_ID]; - -static void SimpleLruZeroLSNs(SlruCtl ctl, int slotno); -static void SimpleLruWaitIO(SlruCtl ctl, int slotno); -static void SlruInternalWritePage(SlruCtl ctl, int slotno, SlruWriteAll fdata); -static bool SlruPhysicalReadPage(SlruCtl ctl, int pageno, int slotno); -static bool SlruPhysicalWritePage(SlruCtl ctl, int pageno, int slotno, - SlruWriteAll fdata); -static void SlruReportIOError(SlruCtl ctl, int pageno, TransactionId xid); -static int SlruSelectLRUPage(SlruCtl ctl, int pageno); - -static bool SlruScanDirCbDeleteCutoff(SlruCtl ctl, char *filename, +static bool SlruScanDirCbDeleteCutoff(int slru_id, + SlruPagePrecedesFunction PagePrecedes, + char *filename, int segpage, void *data); -static void SlruInternalDeleteSegment(SlruCtl ctl, int segno); - -/* - * Initialization of shared memory - */ - -Size -SimpleLruShmemSize(int nslots, int nlsns) -{ - Size sz; - - /* we assume nslots isn't so large as to risk overflow */ - sz = MAXALIGN(sizeof(SlruSharedData)); - sz += MAXALIGN(nslots * sizeof(char *)); /* page_buffer[] */ - sz += MAXALIGN(nslots * sizeof(SlruPageStatus)); /* page_status[] */ - sz += MAXALIGN(nslots * sizeof(bool)); /* page_dirty[] */ - sz += MAXALIGN(nslots * sizeof(int)); /* page_number[] */ - sz += MAXALIGN(nslots * sizeof(int)); /* page_lru_count[] */ - sz += MAXALIGN(nslots * sizeof(LWLockPadded)); /* buffer_locks[] */ - - if (nlsns > 0) - sz += MAXALIGN(nslots * nlsns * sizeof(XLogRecPtr)); /* group_lsn[] */ - - return BUFFERALIGN(sz) + BLCKSZ * nslots; -} - -/* - * Initialize, or attach to, a simple LRU cache in shared memory. - * - * ctl: address of local (unshared) control structure. - * name: name of SLRU. (This is user-visible, pick with care!) - * nslots: number of page slots to use. - * nlsns: number of LSN groups per page (set to zero if not relevant). - * ctllock: LWLock to use to control access to the shared control structure. - * subdir: PGDATA-relative subdirectory that will contain the files. - * tranche_id: LWLock tranche ID to use for the SLRU's per-buffer LWLocks. - * sync_handler: which set of functions to use to handle sync requests - */ -void -SimpleLruInit(SlruCtl ctl, const char *name, int nslots, int nlsns, - LWLock *ctllock, const char *subdir, int tranche_id, - SyncRequestHandler sync_handler) -{ - SlruShared shared; - bool found; - - shared = (SlruShared) ShmemInitStruct(name, - SimpleLruShmemSize(nslots, nlsns), - &found); - - if (!IsUnderPostmaster) - { - /* Initialize locks and shared memory area */ - char *ptr; - Size offset; - int slotno; - - Assert(!found); - - memset(shared, 0, sizeof(SlruSharedData)); - - shared->ControlLock = ctllock; - - shared->num_slots = nslots; - shared->lsn_groups_per_page = nlsns; - - shared->cur_lru_count = 0; - - /* shared->latest_page_number will be set later */ - - shared->slru_stats_idx = pgstat_get_slru_index(name); - - ptr = (char *) shared; - offset = MAXALIGN(sizeof(SlruSharedData)); - shared->page_buffer = (char **) (ptr + offset); - offset += MAXALIGN(nslots * sizeof(char *)); - shared->page_status = (SlruPageStatus *) (ptr + offset); - offset += MAXALIGN(nslots * sizeof(SlruPageStatus)); - shared->page_dirty = (bool *) (ptr + offset); - offset += MAXALIGN(nslots * sizeof(bool)); - shared->page_number = (int *) (ptr + offset); - offset += MAXALIGN(nslots * sizeof(int)); - shared->page_lru_count = (int *) (ptr + offset); - offset += MAXALIGN(nslots * sizeof(int)); - - /* Initialize LWLocks */ - shared->buffer_locks = (LWLockPadded *) (ptr + offset); - offset += MAXALIGN(nslots * sizeof(LWLockPadded)); - - if (nlsns > 0) - { - shared->group_lsn = (XLogRecPtr *) (ptr + offset); - offset += MAXALIGN(nslots * nlsns * sizeof(XLogRecPtr)); - } - - ptr += BUFFERALIGN(offset); - for (slotno = 0; slotno < nslots; slotno++) - { - LWLockInitialize(&shared->buffer_locks[slotno].lock, - tranche_id); - - shared->page_buffer[slotno] = ptr; - shared->page_status[slotno] = SLRU_PAGE_EMPTY; - shared->page_dirty[slotno] = false; - shared->page_lru_count[slotno] = 0; - ptr += BLCKSZ; - } - - /* Should fit to estimated shmem size */ - Assert(ptr - (char *) shared <= SimpleLruShmemSize(nslots, nlsns)); - } - else - Assert(found); - - /* - * Initialize the unshared control struct, including directory path. We - * assume caller set PagePrecedes. - */ - ctl->shared = shared; - ctl->sync_handler = sync_handler; - strlcpy(ctl->Dir, subdir, sizeof(ctl->Dir)); -} - -/* - * Initialize (or reinitialize) a page to zeroes. - * - * The page is not actually written, just set up in shared memory. - * The slot number of the new page is returned. - * - * Control lock must be held at entry, and will be held at exit. - */ -int -SimpleLruZeroPage(SlruCtl ctl, int pageno) -{ - SlruShared shared = ctl->shared; - int slotno; - - /* Find a suitable buffer slot for the page */ - slotno = SlruSelectLRUPage(ctl, pageno); - Assert(shared->page_status[slotno] == SLRU_PAGE_EMPTY || - (shared->page_status[slotno] == SLRU_PAGE_VALID && - !shared->page_dirty[slotno]) || - shared->page_number[slotno] == pageno); - - /* Mark the slot as containing this page */ - shared->page_number[slotno] = pageno; - shared->page_status[slotno] = SLRU_PAGE_VALID; - shared->page_dirty[slotno] = true; - SlruRecentlyUsed(shared, slotno); - - /* Set the buffer to zeroes */ - MemSet(shared->page_buffer[slotno], 0, BLCKSZ); - - /* Set the LSNs for this new page to zero */ - SimpleLruZeroLSNs(ctl, slotno); - - /* Assume this page is now the latest active page */ - shared->latest_page_number = pageno; - - /* update the stats counter of zeroed pages */ - pgstat_count_slru_page_zeroed(shared->slru_stats_idx); - - return slotno; -} - -/* - * Zero all the LSNs we store for this slru page. - * - * This should be called each time we create a new page, and each time we read - * in a page from disk into an existing buffer. (Such an old page cannot - * have any interesting LSNs, since we'd have flushed them before writing - * the page in the first place.) - * - * This assumes that InvalidXLogRecPtr is bitwise-all-0. - */ -static void -SimpleLruZeroLSNs(SlruCtl ctl, int slotno) -{ - SlruShared shared = ctl->shared; - - if (shared->lsn_groups_per_page > 0) - MemSet(&shared->group_lsn[slotno * shared->lsn_groups_per_page], 0, - shared->lsn_groups_per_page * sizeof(XLogRecPtr)); -} - -/* - * Wait for any active I/O on a page slot to finish. (This does not - * guarantee that new I/O hasn't been started before we return, though. - * In fact the slot might not even contain the same page anymore.) - * - * Control lock must be held at entry, and will be held at exit. - */ -static void -SimpleLruWaitIO(SlruCtl ctl, int slotno) -{ - SlruShared shared = ctl->shared; - - /* See notes at top of file */ - LWLockRelease(shared->ControlLock); - LWLockAcquire(&shared->buffer_locks[slotno].lock, LW_SHARED); - LWLockRelease(&shared->buffer_locks[slotno].lock); - LWLockAcquire(shared->ControlLock, LW_EXCLUSIVE); - - /* - * If the slot is still in an io-in-progress state, then either someone - * already started a new I/O on the slot, or a previous I/O failed and - * neglected to reset the page state. That shouldn't happen, really, but - * it seems worth a few extra cycles to check and recover from it. We can - * cheaply test for failure by seeing if the buffer lock is still held (we - * assume that transaction abort would release the lock). - */ - if (shared->page_status[slotno] == SLRU_PAGE_READ_IN_PROGRESS || - shared->page_status[slotno] == SLRU_PAGE_WRITE_IN_PROGRESS) - { - if (LWLockConditionalAcquire(&shared->buffer_locks[slotno].lock, LW_SHARED)) - { - /* indeed, the I/O must have failed */ - if (shared->page_status[slotno] == SLRU_PAGE_READ_IN_PROGRESS) - shared->page_status[slotno] = SLRU_PAGE_EMPTY; - else /* write_in_progress */ - { - shared->page_status[slotno] = SLRU_PAGE_VALID; - shared->page_dirty[slotno] = true; - } - LWLockRelease(&shared->buffer_locks[slotno].lock); - } - } -} - -/* - * Find a page in a shared buffer, reading it in if necessary. - * The page number must correspond to an already-initialized page. - * - * If write_ok is true then it is OK to return a page that is in - * WRITE_IN_PROGRESS state; it is the caller's responsibility to be sure - * that modification of the page is safe. If write_ok is false then we - * will not return the page until it is not undergoing active I/O. - * - * The passed-in xid is used only for error reporting, and may be - * InvalidTransactionId if no specific xid is associated with the action. - * - * Return value is the shared-buffer slot number now holding the page. - * The buffer's LRU access info is updated. - * - * Control lock must be held at entry, and will be held at exit. - */ -int -SimpleLruReadPage(SlruCtl ctl, int pageno, bool write_ok, - TransactionId xid) -{ - SlruShared shared = ctl->shared; - - /* Outer loop handles restart if we must wait for someone else's I/O */ - for (;;) - { - int slotno; - bool ok; - - /* See if page already is in memory; if not, pick victim slot */ - slotno = SlruSelectLRUPage(ctl, pageno); - - /* Did we find the page in memory? */ - if (shared->page_number[slotno] == pageno && - shared->page_status[slotno] != SLRU_PAGE_EMPTY) - { - /* - * If page is still being read in, we must wait for I/O. Likewise - * if the page is being written and the caller said that's not OK. - */ - if (shared->page_status[slotno] == SLRU_PAGE_READ_IN_PROGRESS || - (shared->page_status[slotno] == SLRU_PAGE_WRITE_IN_PROGRESS && - !write_ok)) - { - SimpleLruWaitIO(ctl, slotno); - /* Now we must recheck state from the top */ - continue; - } - /* Otherwise, it's ready to use */ - SlruRecentlyUsed(shared, slotno); - - /* update the stats counter of pages found in the SLRU */ - pgstat_count_slru_page_hit(shared->slru_stats_idx); - - return slotno; - } - - /* We found no match; assert we selected a freeable slot */ - Assert(shared->page_status[slotno] == SLRU_PAGE_EMPTY || - (shared->page_status[slotno] == SLRU_PAGE_VALID && - !shared->page_dirty[slotno])); - - /* Mark the slot read-busy */ - shared->page_number[slotno] = pageno; - shared->page_status[slotno] = SLRU_PAGE_READ_IN_PROGRESS; - shared->page_dirty[slotno] = false; - - /* Acquire per-buffer lock (cannot deadlock, see notes at top) */ - LWLockAcquire(&shared->buffer_locks[slotno].lock, LW_EXCLUSIVE); - - /* Release control lock while doing I/O */ - LWLockRelease(shared->ControlLock); - - /* Do the read */ - ok = SlruPhysicalReadPage(ctl, pageno, slotno); - - /* Set the LSNs for this newly read-in page to zero */ - SimpleLruZeroLSNs(ctl, slotno); - - /* Re-acquire control lock and update page state */ - LWLockAcquire(shared->ControlLock, LW_EXCLUSIVE); - - Assert(shared->page_number[slotno] == pageno && - shared->page_status[slotno] == SLRU_PAGE_READ_IN_PROGRESS && - !shared->page_dirty[slotno]); - - shared->page_status[slotno] = ok ? SLRU_PAGE_VALID : SLRU_PAGE_EMPTY; - - LWLockRelease(&shared->buffer_locks[slotno].lock); - - /* Now it's okay to ereport if we failed */ - if (!ok) - SlruReportIOError(ctl, pageno, xid); - - SlruRecentlyUsed(shared, slotno); - - /* update the stats counter of pages not found in SLRU */ - pgstat_count_slru_page_read(shared->slru_stats_idx); - - return slotno; - } -} - -/* - * Find a page in a shared buffer, reading it in if necessary. - * The page number must correspond to an already-initialized page. - * The caller must intend only read-only access to the page. - * - * The passed-in xid is used only for error reporting, and may be - * InvalidTransactionId if no specific xid is associated with the action. - * - * Return value is the shared-buffer slot number now holding the page. - * The buffer's LRU access info is updated. - * - * Control lock must NOT be held at entry, but will be held at exit. - * It is unspecified whether the lock will be shared or exclusive. - */ -int -SimpleLruReadPage_ReadOnly(SlruCtl ctl, int pageno, TransactionId xid) -{ - SlruShared shared = ctl->shared; - int slotno; - - /* Try to find the page while holding only shared lock */ - LWLockAcquire(shared->ControlLock, LW_SHARED); - - /* See if page is already in a buffer */ - for (slotno = 0; slotno < shared->num_slots; slotno++) - { - if (shared->page_number[slotno] == pageno && - shared->page_status[slotno] != SLRU_PAGE_EMPTY && - shared->page_status[slotno] != SLRU_PAGE_READ_IN_PROGRESS) - { - /* See comments for SlruRecentlyUsed macro */ - SlruRecentlyUsed(shared, slotno); - - /* update the stats counter of pages found in the SLRU */ - pgstat_count_slru_page_hit(shared->slru_stats_idx); - - return slotno; - } - } - - /* No luck, so switch to normal exclusive lock and do regular read */ - LWLockRelease(shared->ControlLock); - LWLockAcquire(shared->ControlLock, LW_EXCLUSIVE); - - return SimpleLruReadPage(ctl, pageno, true, xid); -} - -/* - * Write a page from a shared buffer, if necessary. - * Does nothing if the specified slot is not dirty. - * - * NOTE: only one write attempt is made here. Hence, it is possible that - * the page is still dirty at exit (if someone else re-dirtied it during - * the write). However, we *do* attempt a fresh write even if the page - * is already being written; this is for checkpoints. - * - * Control lock must be held at entry, and will be held at exit. - */ -static void -SlruInternalWritePage(SlruCtl ctl, int slotno, SlruWriteAll fdata) -{ - SlruShared shared = ctl->shared; - int pageno = shared->page_number[slotno]; - bool ok; - - /* If a write is in progress, wait for it to finish */ - while (shared->page_status[slotno] == SLRU_PAGE_WRITE_IN_PROGRESS && - shared->page_number[slotno] == pageno) - { - SimpleLruWaitIO(ctl, slotno); - } - - /* - * Do nothing if page is not dirty, or if buffer no longer contains the - * same page we were called for. - */ - if (!shared->page_dirty[slotno] || - shared->page_status[slotno] != SLRU_PAGE_VALID || - shared->page_number[slotno] != pageno) - return; - - /* - * Mark the slot write-busy, and clear the dirtybit. After this point, a - * transaction status update on this page will mark it dirty again. - */ - shared->page_status[slotno] = SLRU_PAGE_WRITE_IN_PROGRESS; - shared->page_dirty[slotno] = false; - - /* Acquire per-buffer lock (cannot deadlock, see notes at top) */ - LWLockAcquire(&shared->buffer_locks[slotno].lock, LW_EXCLUSIVE); - - /* Release control lock while doing I/O */ - LWLockRelease(shared->ControlLock); - - /* Do the write */ - ok = SlruPhysicalWritePage(ctl, pageno, slotno, fdata); - - /* If we failed, and we're in a flush, better close the files */ - if (!ok && fdata) - { - int i; - - for (i = 0; i < fdata->num_files; i++) - CloseTransientFile(fdata->fd[i]); - } - - /* Re-acquire control lock and update page state */ - LWLockAcquire(shared->ControlLock, LW_EXCLUSIVE); - - Assert(shared->page_number[slotno] == pageno && - shared->page_status[slotno] == SLRU_PAGE_WRITE_IN_PROGRESS); - - /* If we failed to write, mark the page dirty again */ - if (!ok) - shared->page_dirty[slotno] = true; - - shared->page_status[slotno] = SLRU_PAGE_VALID; - - LWLockRelease(&shared->buffer_locks[slotno].lock); - - /* Now it's okay to ereport if we failed */ - if (!ok) - SlruReportIOError(ctl, pageno, InvalidTransactionId); - - /* If part of a checkpoint, count this as a buffer written. */ - if (fdata) - CheckpointStats.ckpt_bufs_written++; -} - -/* - * Wrapper of SlruInternalWritePage, for external callers. - * fdata is always passed a NULL here. - */ -void -SimpleLruWritePage(SlruCtl ctl, int slotno) -{ - SlruInternalWritePage(ctl, slotno, NULL); -} +static void SlruInternalDeleteSegment(int slru_id, int segno); /* * Return whether the given page exists on disk. @@ -624,592 +64,24 @@ SimpleLruWritePage(SlruCtl ctl, int slotno) * large enough to contain the given page. */ bool -SimpleLruDoesPhysicalPageExist(SlruCtl ctl, int pageno) +SimpleLruDoesPhysicalPageExist(int slru_id, int pageno) { - int segno = pageno / SLRU_PAGES_PER_SEGMENT; - int rpageno = pageno % SLRU_PAGES_PER_SEGMENT; - int offset = rpageno * BLCKSZ; - char path[MAXPGPATH]; - int fd; - bool result; - off_t endpos; - - /* update the stats counter of checked pages */ - pgstat_count_slru_page_exists(ctl->shared->slru_stats_idx); - - SlruFileName(ctl, path, segno); - - fd = OpenTransientFile(path, O_RDONLY | PG_BINARY); - if (fd < 0) - { - /* expected: file doesn't exist */ - if (errno == ENOENT) - return false; - - /* report error normally */ - slru_errcause = SLRU_OPEN_FAILED; - slru_errno = errno; - SlruReportIOError(ctl, pageno, 0); - } - - if ((endpos = lseek(fd, 0, SEEK_END)) < 0) - { - slru_errcause = SLRU_SEEK_FAILED; - slru_errno = errno; - SlruReportIOError(ctl, pageno, 0); - } - - result = endpos >= (off_t) (offset + BLCKSZ); - - if (CloseTransientFile(fd) != 0) - { - slru_errcause = SLRU_CLOSE_FAILED; - slru_errno = errno; - return false; - } - - return result; -} - -/* - * Physical read of a (previously existing) page into a buffer slot - * - * On failure, we cannot just ereport(ERROR) since caller has put state in - * shared memory that must be undone. So, we return false and save enough - * info in static variables to let SlruReportIOError make the report. - * - * For now, assume it's not worth keeping a file pointer open across - * read/write operations. We could cache one virtual file pointer ... - */ -static bool -SlruPhysicalReadPage(SlruCtl ctl, int pageno, int slotno) -{ - SlruShared shared = ctl->shared; - int segno = pageno / SLRU_PAGES_PER_SEGMENT; - int rpageno = pageno % SLRU_PAGES_PER_SEGMENT; - off_t offset = rpageno * BLCKSZ; - char path[MAXPGPATH]; - int fd; - - SlruFileName(ctl, path, segno); - - /* - * In a crash-and-restart situation, it's possible for us to receive - * commands to set the commit status of transactions whose bits are in - * already-truncated segments of the commit log (see notes in - * SlruPhysicalWritePage). Hence, if we are InRecovery, allow the case - * where the file doesn't exist, and return zeroes instead. - */ - fd = OpenTransientFile(path, O_RDONLY | PG_BINARY); - if (fd < 0) - { - if (errno != ENOENT || !InRecovery) - { - slru_errcause = SLRU_OPEN_FAILED; - slru_errno = errno; - return false; - } - - ereport(LOG, - (errmsg("file \"%s\" doesn't exist, reading as zeroes", - path))); - MemSet(shared->page_buffer[slotno], 0, BLCKSZ); - return true; - } - - errno = 0; - pgstat_report_wait_start(WAIT_EVENT_SLRU_READ); - if (pg_pread(fd, shared->page_buffer[slotno], BLCKSZ, offset) != BLCKSZ) - { - pgstat_report_wait_end(); - slru_errcause = SLRU_READ_FAILED; - slru_errno = errno; - CloseTransientFile(fd); - return false; - } - pgstat_report_wait_end(); - - if (CloseTransientFile(fd) != 0) - { - slru_errcause = SLRU_CLOSE_FAILED; - slru_errno = errno; - return false; - } - - return true; -} - -/* - * Physical write of a page from a buffer slot - * - * On failure, we cannot just ereport(ERROR) since caller has put state in - * shared memory that must be undone. So, we return false and save enough - * info in static variables to let SlruReportIOError make the report. - * - * For now, assume it's not worth keeping a file pointer open across - * independent read/write operations. We do batch operations during - * SimpleLruWriteAll, though. - * - * fdata is NULL for a standalone write, pointer to open-file info during - * SimpleLruWriteAll. - */ -static bool -SlruPhysicalWritePage(SlruCtl ctl, int pageno, int slotno, SlruWriteAll fdata) -{ - SlruShared shared = ctl->shared; int segno = pageno / SLRU_PAGES_PER_SEGMENT; int rpageno = pageno % SLRU_PAGES_PER_SEGMENT; off_t offset = rpageno * BLCKSZ; - char path[MAXPGPATH]; - int fd = -1; - - /* update the stats counter of written pages */ - pgstat_count_slru_page_written(shared->slru_stats_idx); - - /* - * Honor the write-WAL-before-data rule, if appropriate, so that we do not - * write out data before associated WAL records. This is the same action - * performed during FlushBuffer() in the main buffer manager. - */ - if (shared->group_lsn != NULL) - { - /* - * We must determine the largest async-commit LSN for the page. This - * is a bit tedious, but since this entire function is a slow path - * anyway, it seems better to do this here than to maintain a per-page - * LSN variable (which'd need an extra comparison in the - * transaction-commit path). - */ - XLogRecPtr max_lsn; - int lsnindex, - lsnoff; - - lsnindex = slotno * shared->lsn_groups_per_page; - max_lsn = shared->group_lsn[lsnindex++]; - for (lsnoff = 1; lsnoff < shared->lsn_groups_per_page; lsnoff++) - { - XLogRecPtr this_lsn = shared->group_lsn[lsnindex++]; - - if (max_lsn < this_lsn) - max_lsn = this_lsn; - } - - if (!XLogRecPtrIsInvalid(max_lsn)) - { - /* - * As noted above, elog(ERROR) is not acceptable here, so if - * XLogFlush were to fail, we must PANIC. This isn't much of a - * restriction because XLogFlush is just about all critical - * section anyway, but let's make sure. - */ - START_CRIT_SECTION(); - XLogFlush(max_lsn); - END_CRIT_SECTION(); - } - } - - /* - * During a WriteAll, we may already have the desired file open. - */ - if (fdata) - { - int i; - - for (i = 0; i < fdata->num_files; i++) - { - if (fdata->segno[i] == segno) - { - fd = fdata->fd[i]; - break; - } - } - } - - if (fd < 0) - { - /* - * If the file doesn't already exist, we should create it. It is - * possible for this to need to happen when writing a page that's not - * first in its segment; we assume the OS can cope with that. (Note: - * it might seem that it'd be okay to create files only when - * SimpleLruZeroPage is called for the first page of a segment. - * However, if after a crash and restart the REDO logic elects to - * replay the log from a checkpoint before the latest one, then it's - * possible that we will get commands to set transaction status of - * transactions that have already been truncated from the commit log. - * Easiest way to deal with that is to accept references to - * nonexistent files here and in SlruPhysicalReadPage.) - * - * Note: it is possible for more than one backend to be executing this - * code simultaneously for different pages of the same file. Hence, - * don't use O_EXCL or O_TRUNC or anything like that. - */ - SlruFileName(ctl, path, segno); - fd = OpenTransientFile(path, O_RDWR | O_CREAT | PG_BINARY); - if (fd < 0) - { - slru_errcause = SLRU_OPEN_FAILED; - slru_errno = errno; - return false; - } - - if (fdata) - { - if (fdata->num_files < MAX_WRITEALL_BUFFERS) - { - fdata->fd[fdata->num_files] = fd; - fdata->segno[fdata->num_files] = segno; - fdata->num_files++; - } - else - { - /* - * In the unlikely event that we exceed MAX_FLUSH_BUFFERS, - * fall back to treating it as a standalone write. - */ - fdata = NULL; - } - } - } - - errno = 0; - pgstat_report_wait_start(WAIT_EVENT_SLRU_WRITE); - if (pg_pwrite(fd, shared->page_buffer[slotno], BLCKSZ, offset) != BLCKSZ) - { - pgstat_report_wait_end(); - /* if write didn't set errno, assume problem is no disk space */ - if (errno == 0) - errno = ENOSPC; - slru_errcause = SLRU_WRITE_FAILED; - slru_errno = errno; - if (!fdata) - CloseTransientFile(fd); - return false; - } - pgstat_report_wait_end(); - - /* Queue up a sync request for the checkpointer. */ - if (ctl->sync_handler != SYNC_HANDLER_NONE) - { - FileTag tag; - - INIT_SLRUFILETAG(tag, ctl->sync_handler, segno); - if (!RegisterSyncRequest(&tag, SYNC_REQUEST, false)) - { - /* No space to enqueue sync request. Do it synchronously. */ - pgstat_report_wait_start(WAIT_EVENT_SLRU_SYNC); - if (pg_fsync(fd) != 0) - { - pgstat_report_wait_end(); - slru_errcause = SLRU_FSYNC_FAILED; - slru_errno = errno; - CloseTransientFile(fd); - return false; - } - pgstat_report_wait_end(); - } - } - - /* Close file, unless part of flush request. */ - if (!fdata) - { - if (CloseTransientFile(fd) != 0) - { - slru_errcause = SLRU_CLOSE_FAILED; - slru_errno = errno; - return false; - } - } - - return true; -} - -/* - * Issue the error message after failure of SlruPhysicalReadPage or - * SlruPhysicalWritePage. Call this after cleaning up shared-memory state. - */ -static void -SlruReportIOError(SlruCtl ctl, int pageno, TransactionId xid) -{ - int segno = pageno / SLRU_PAGES_PER_SEGMENT; - int rpageno = pageno % SLRU_PAGES_PER_SEGMENT; - int offset = rpageno * BLCKSZ; - char path[MAXPGPATH]; - - SlruFileName(ctl, path, segno); - errno = slru_errno; - switch (slru_errcause) - { - case SLRU_OPEN_FAILED: - ereport(ERROR, - (errcode_for_file_access(), - errmsg("could not access status of transaction %u", xid), - errdetail("Could not open file \"%s\": %m.", path))); - break; - case SLRU_SEEK_FAILED: - ereport(ERROR, - (errcode_for_file_access(), - errmsg("could not access status of transaction %u", xid), - errdetail("Could not seek in file \"%s\" to offset %d: %m.", - path, offset))); - break; - case SLRU_READ_FAILED: - if (errno) - ereport(ERROR, - (errcode_for_file_access(), - errmsg("could not access status of transaction %u", xid), - errdetail("Could not read from file \"%s\" at offset %d: %m.", - path, offset))); - else - ereport(ERROR, - (errmsg("could not access status of transaction %u", xid), - errdetail("Could not read from file \"%s\" at offset %d: read too few bytes.", path, offset))); - break; - case SLRU_WRITE_FAILED: - if (errno) - ereport(ERROR, - (errcode_for_file_access(), - errmsg("could not access status of transaction %u", xid), - errdetail("Could not write to file \"%s\" at offset %d: %m.", - path, offset))); - else - ereport(ERROR, - (errmsg("could not access status of transaction %u", xid), - errdetail("Could not write to file \"%s\" at offset %d: wrote too few bytes.", - path, offset))); - break; - case SLRU_FSYNC_FAILED: - ereport(data_sync_elevel(ERROR), - (errcode_for_file_access(), - errmsg("could not access status of transaction %u", xid), - errdetail("Could not fsync file \"%s\": %m.", - path))); - break; - case SLRU_CLOSE_FAILED: - ereport(ERROR, - (errcode_for_file_access(), - errmsg("could not access status of transaction %u", xid), - errdetail("Could not close file \"%s\": %m.", - path))); - break; - default: - /* can't get here, we trust */ - elog(ERROR, "unrecognized SimpleLru error cause: %d", - (int) slru_errcause); - break; - } -} - -/* - * Select the slot to re-use when we need a free slot. - * - * The target page number is passed because we need to consider the - * possibility that some other process reads in the target page while - * we are doing I/O to free a slot. Hence, check or recheck to see if - * any slot already holds the target page, and return that slot if so. - * Thus, the returned slot is *either* a slot already holding the pageno - * (could be any state except EMPTY), *or* a freeable slot (state EMPTY - * or CLEAN). - * - * Control lock must be held at entry, and will be held at exit. - */ -static int -SlruSelectLRUPage(SlruCtl ctl, int pageno) -{ - SlruShared shared = ctl->shared; - - /* Outer loop handles restart after I/O */ - for (;;) - { - int slotno; - int cur_count; - int bestvalidslot = 0; /* keep compiler quiet */ - int best_valid_delta = -1; - int best_valid_page_number = 0; /* keep compiler quiet */ - int bestinvalidslot = 0; /* keep compiler quiet */ - int best_invalid_delta = -1; - int best_invalid_page_number = 0; /* keep compiler quiet */ - - /* See if page already has a buffer assigned */ - for (slotno = 0; slotno < shared->num_slots; slotno++) - { - if (shared->page_number[slotno] == pageno && - shared->page_status[slotno] != SLRU_PAGE_EMPTY) - return slotno; - } - - /* - * If we find any EMPTY slot, just select that one. Else choose a - * victim page to replace. We normally take the least recently used - * valid page, but we will never take the slot containing - * latest_page_number, even if it appears least recently used. We - * will select a slot that is already I/O busy only if there is no - * other choice: a read-busy slot will not be least recently used once - * the read finishes, and waiting for an I/O on a write-busy slot is - * inferior to just picking some other slot. Testing shows the slot - * we pick instead will often be clean, allowing us to begin a read at - * once. - * - * Normally the page_lru_count values will all be different and so - * there will be a well-defined LRU page. But since we allow - * concurrent execution of SlruRecentlyUsed() within - * SimpleLruReadPage_ReadOnly(), it is possible that multiple pages - * acquire the same lru_count values. In that case we break ties by - * choosing the furthest-back page. - * - * Notice that this next line forcibly advances cur_lru_count to a - * value that is certainly beyond any value that will be in the - * page_lru_count array after the loop finishes. This ensures that - * the next execution of SlruRecentlyUsed will mark the page newly - * used, even if it's for a page that has the current counter value. - * That gets us back on the path to having good data when there are - * multiple pages with the same lru_count. - */ - cur_count = (shared->cur_lru_count)++; - for (slotno = 0; slotno < shared->num_slots; slotno++) - { - int this_delta; - int this_page_number; - - if (shared->page_status[slotno] == SLRU_PAGE_EMPTY) - return slotno; - this_delta = cur_count - shared->page_lru_count[slotno]; - if (this_delta < 0) - { - /* - * Clean up in case shared updates have caused cur_count - * increments to get "lost". We back off the page counts, - * rather than trying to increase cur_count, to avoid any - * question of infinite loops or failure in the presence of - * wrapped-around counts. - */ - shared->page_lru_count[slotno] = cur_count; - this_delta = 0; - } - this_page_number = shared->page_number[slotno]; - if (this_page_number == shared->latest_page_number) - continue; - if (shared->page_status[slotno] == SLRU_PAGE_VALID) - { - if (this_delta > best_valid_delta || - (this_delta == best_valid_delta && - ctl->PagePrecedes(this_page_number, - best_valid_page_number))) - { - bestvalidslot = slotno; - best_valid_delta = this_delta; - best_valid_page_number = this_page_number; - } - } - else - { - if (this_delta > best_invalid_delta || - (this_delta == best_invalid_delta && - ctl->PagePrecedes(this_page_number, - best_invalid_page_number))) - { - bestinvalidslot = slotno; - best_invalid_delta = this_delta; - best_invalid_page_number = this_page_number; - } - } - } - - /* - * If all pages (except possibly the latest one) are I/O busy, we'll - * have to wait for an I/O to complete and then retry. In that - * unhappy case, we choose to wait for the I/O on the least recently - * used slot, on the assumption that it was likely initiated first of - * all the I/Os in progress and may therefore finish first. - */ - if (best_valid_delta < 0) - { - SimpleLruWaitIO(ctl, bestinvalidslot); - continue; - } - - /* - * If the selected page is clean, we're set. - */ - if (!shared->page_dirty[bestvalidslot]) - return bestvalidslot; - - /* - * Write the page. - */ - SlruInternalWritePage(ctl, bestvalidslot, NULL); - - /* - * Now loop back and try again. This is the easiest way of dealing - * with corner cases such as the victim page being re-dirtied while we - * wrote it. - */ - } -} - -/* - * Write dirty pages to disk during checkpoint or database shutdown. Flushing - * is deferred until the next call to ProcessSyncRequests(), though we do fsync - * the containing directory here to make sure that newly created directory - * entries are on disk. - */ -void -SimpleLruWriteAll(SlruCtl ctl, bool allow_redirtied) -{ - SlruShared shared = ctl->shared; - SlruWriteAllData fdata; - int slotno; - int pageno = 0; - int i; - bool ok; - - /* update the stats counter of flushes */ - pgstat_count_slru_flush(shared->slru_stats_idx); - - /* - * Find and write dirty pages - */ - fdata.num_files = 0; + off_t size; + RelFileLocator rlocator = SlruRelFileLocator(slru_id, segno); + SMgrFileHandle sfile = smgropen(rlocator, InvalidBackendId, MAIN_FORKNUM); - LWLockAcquire(shared->ControlLock, LW_EXCLUSIVE); - - for (slotno = 0; slotno < shared->num_slots; slotno++) - { - SlruInternalWritePage(ctl, slotno, &fdata); - - /* - * In some places (e.g. checkpoints), we cannot assert that the slot - * is clean now, since another process might have re-dirtied it - * already. That's okay. - */ - Assert(allow_redirtied || - shared->page_status[slotno] == SLRU_PAGE_EMPTY || - (shared->page_status[slotno] == SLRU_PAGE_VALID && - !shared->page_dirty[slotno])); - } - - LWLockRelease(shared->ControlLock); + /* update the stats counter of checked pages */ + pgstat_count_slru_page_exists(slru_id); - /* - * Now close any files that were open - */ - ok = true; - for (i = 0; i < fdata.num_files; i++) - { - if (CloseTransientFile(fdata.fd[i]) != 0) - { - slru_errcause = SLRU_CLOSE_FAILED; - slru_errno = errno; - pageno = fdata.segno[i] * SLRU_PAGES_PER_SEGMENT; - ok = false; - } - } - if (!ok) - SlruReportIOError(ctl, pageno, InvalidTransactionId); + if (smgrexists(sfile)) + size = smgrnblocks(sfile); + else + size = 0; - /* Ensure that directory entries for new files are on disk. */ - if (ctl->sync_handler != SYNC_HANDLER_NONE) - fsync_fname(ctl->Dir, true); + return size >= offset + BLCKSZ; } /* @@ -1224,75 +96,14 @@ SimpleLruWriteAll(SlruCtl ctl, bool allow_redirtied) * after it has accrued freshly-written data. */ void -SimpleLruTruncate(SlruCtl ctl, int cutoffPage) +SimpleLruTruncate(int slru_id, SlruPagePrecedesFunction PagePrecedes, int cutoffPage) { - SlruShared shared = ctl->shared; - int slotno; - /* update the stats counter of truncates */ - pgstat_count_slru_truncate(shared->slru_stats_idx); - - /* - * Scan shared memory and remove any pages preceding the cutoff page, to - * ensure we won't rewrite them later. (Since this is normally called in - * or just after a checkpoint, any dirty pages should have been flushed - * already ... we're just being extra careful here.) - */ - LWLockAcquire(shared->ControlLock, LW_EXCLUSIVE); - -restart: - - /* - * While we are holding the lock, make an important safety check: the - * current endpoint page must not be eligible for removal. - */ - if (ctl->PagePrecedes(shared->latest_page_number, cutoffPage)) - { - LWLockRelease(shared->ControlLock); - ereport(LOG, - (errmsg("could not truncate directory \"%s\": apparent wraparound", - ctl->Dir))); - return; - } - - for (slotno = 0; slotno < shared->num_slots; slotno++) - { - if (shared->page_status[slotno] == SLRU_PAGE_EMPTY) - continue; - if (!ctl->PagePrecedes(shared->page_number[slotno], cutoffPage)) - continue; - - /* - * If page is clean, just change state to EMPTY (expected case). - */ - if (shared->page_status[slotno] == SLRU_PAGE_VALID && - !shared->page_dirty[slotno]) - { - shared->page_status[slotno] = SLRU_PAGE_EMPTY; - continue; - } - - /* - * Hmm, we have (or may have) I/O operations acting on the page, so - * we've got to wait for them to finish and then start again. This is - * the same logic as in SlruSelectLRUPage. (XXX if page is dirty, - * wouldn't it be OK to just discard it without writing it? - * SlruMayDeleteSegment() uses a stricter qualification, so we might - * not delete this page in the end; even if we don't delete it, we - * won't have cause to read its data again. For now, keep the logic - * the same as it was.) - */ - if (shared->page_status[slotno] == SLRU_PAGE_VALID) - SlruInternalWritePage(ctl, slotno, NULL); - else - SimpleLruWaitIO(ctl, slotno); - goto restart; - } - - LWLockRelease(shared->ControlLock); + pgstat_count_slru_truncate(slru_id); /* Now we can remove the old segment(s) */ - (void) SlruScanDirectory(ctl, SlruScanDirCbDeleteCutoff, &cutoffPage); + (void) SlruScanDirectory(slru_id, PagePrecedes, SlruScanDirCbDeleteCutoff, + &cutoffPage); } /* @@ -1302,77 +113,22 @@ restart: * they either can't yet contain anything, or have already been cleaned out. */ static void -SlruInternalDeleteSegment(SlruCtl ctl, int segno) +SlruInternalDeleteSegment(int slru_id, int segno) { - char path[MAXPGPATH]; - - /* Forget any fsync requests queued for this segment. */ - if (ctl->sync_handler != SYNC_HANDLER_NONE) - { - FileTag tag; - - INIT_SLRUFILETAG(tag, ctl->sync_handler, segno); - RegisterSyncRequest(&tag, SYNC_FORGET_REQUEST, true); - } + RelFileLocator rlocator = SlruRelFileLocator(slru_id, segno); + SMgrFileHandle sfile = smgropen(rlocator, InvalidBackendId, MAIN_FORKNUM); /* Unlink the file. */ - SlruFileName(ctl, path, segno); - ereport(DEBUG2, (errmsg_internal("removing file \"%s\"", path))); - unlink(path); + smgrunlink(sfile, false); } /* * Delete an individual SLRU segment, identified by the segment number. */ void -SlruDeleteSegment(SlruCtl ctl, int segno) +SlruDeleteSegment(int slru_id, int segno) { - SlruShared shared = ctl->shared; - int slotno; - bool did_write; - - /* Clean out any possibly existing references to the segment. */ - LWLockAcquire(shared->ControlLock, LW_EXCLUSIVE); -restart: - did_write = false; - for (slotno = 0; slotno < shared->num_slots; slotno++) - { - int pagesegno = shared->page_number[slotno] / SLRU_PAGES_PER_SEGMENT; - - if (shared->page_status[slotno] == SLRU_PAGE_EMPTY) - continue; - - /* not the segment we're looking for */ - if (pagesegno != segno) - continue; - - /* If page is clean, just change state to EMPTY (expected case). */ - if (shared->page_status[slotno] == SLRU_PAGE_VALID && - !shared->page_dirty[slotno]) - { - shared->page_status[slotno] = SLRU_PAGE_EMPTY; - continue; - } - - /* Same logic as SimpleLruTruncate() */ - if (shared->page_status[slotno] == SLRU_PAGE_VALID) - SlruInternalWritePage(ctl, slotno, NULL); - else - SimpleLruWaitIO(ctl, slotno); - - did_write = true; - } - - /* - * Be extra careful and re-check. The IO functions release the control - * lock, so new pages could have been read in. - */ - if (did_write) - goto restart; - - SlruInternalDeleteSegment(ctl, segno); - - LWLockRelease(shared->ControlLock); + SlruInternalDeleteSegment(slru_id, segno); } /* @@ -1389,19 +145,21 @@ restart: * first>=cutoff && last>=cutoff: no; every page of this segment is too young */ static bool -SlruMayDeleteSegment(SlruCtl ctl, int segpage, int cutoffPage) +SlruMayDeleteSegment(SlruPagePrecedesFunction PagePrecedes, + int segpage, int cutoffPage) { int seg_last_page = segpage + SLRU_PAGES_PER_SEGMENT - 1; Assert(segpage % SLRU_PAGES_PER_SEGMENT == 0); - return (ctl->PagePrecedes(segpage, cutoffPage) && - ctl->PagePrecedes(seg_last_page, cutoffPage)); + return (PagePrecedes(segpage, cutoffPage) && + PagePrecedes(seg_last_page, cutoffPage)); } #ifdef USE_ASSERT_CHECKING static void -SlruPagePrecedesTestOffset(SlruCtl ctl, int per_page, uint32 offset) +SlruPagePrecedesTestOffset(SlruPagePrecedesFunction PagePrecedes, + int per_page, uint32 offset) { TransactionId lhs, rhs; @@ -1426,19 +184,19 @@ SlruPagePrecedesTestOffset(SlruCtl ctl, int per_page, uint32 offset) Assert(!TransactionIdPrecedes(rhs, lhs + 1)); Assert(!TransactionIdFollowsOrEquals(lhs, rhs)); Assert(!TransactionIdFollowsOrEquals(rhs, lhs)); - Assert(!ctl->PagePrecedes(lhs / per_page, lhs / per_page)); - Assert(!ctl->PagePrecedes(lhs / per_page, rhs / per_page)); - Assert(!ctl->PagePrecedes(rhs / per_page, lhs / per_page)); - Assert(!ctl->PagePrecedes((lhs - per_page) / per_page, rhs / per_page)); - Assert(ctl->PagePrecedes(rhs / per_page, (lhs - 3 * per_page) / per_page)); - Assert(ctl->PagePrecedes(rhs / per_page, (lhs - 2 * per_page) / per_page)); - Assert(ctl->PagePrecedes(rhs / per_page, (lhs - 1 * per_page) / per_page) + Assert(!PagePrecedes(lhs / per_page, lhs / per_page)); + Assert(!PagePrecedes(lhs / per_page, rhs / per_page)); + Assert(!PagePrecedes(rhs / per_page, lhs / per_page)); + Assert(!PagePrecedes((lhs - per_page) / per_page, rhs / per_page)); + Assert(PagePrecedes(rhs / per_page, (lhs - 3 * per_page) / per_page)); + Assert(PagePrecedes(rhs / per_page, (lhs - 2 * per_page) / per_page)); + Assert(PagePrecedes(rhs / per_page, (lhs - 1 * per_page) / per_page) || (1U << 31) % per_page != 0); /* See CommitTsPagePrecedes() */ - Assert(ctl->PagePrecedes((lhs + 1 * per_page) / per_page, rhs / per_page) + Assert(PagePrecedes((lhs + 1 * per_page) / per_page, rhs / per_page) || (1U << 31) % per_page != 0); - Assert(ctl->PagePrecedes((lhs + 2 * per_page) / per_page, rhs / per_page)); - Assert(ctl->PagePrecedes((lhs + 3 * per_page) / per_page, rhs / per_page)); - Assert(!ctl->PagePrecedes(rhs / per_page, (lhs + per_page) / per_page)); + Assert(PagePrecedes((lhs + 2 * per_page) / per_page, rhs / per_page)); + Assert(PagePrecedes((lhs + 3 * per_page) / per_page, rhs / per_page)); + Assert(!PagePrecedes(rhs / per_page, (lhs + per_page) / per_page)); /* * GetNewTransactionId() has assigned the last XID it can safely use, and @@ -1451,7 +209,7 @@ SlruPagePrecedesTestOffset(SlruCtl ctl, int per_page, uint32 offset) oldestXact = newestXact + 1; oldestXact -= 1U << 31; oldestPage = oldestXact / per_page; - Assert(!SlruMayDeleteSegment(ctl, + Assert(!SlruMayDeleteSegment(PagePrecedes, (newestPage - newestPage % SLRU_PAGES_PER_SEGMENT), oldestPage)); @@ -1467,7 +225,7 @@ SlruPagePrecedesTestOffset(SlruCtl ctl, int per_page, uint32 offset) oldestXact = newestXact + 1; oldestXact -= 1U << 31; oldestPage = oldestXact / per_page; - Assert(!SlruMayDeleteSegment(ctl, + Assert(!SlruMayDeleteSegment(PagePrecedes, (newestPage - newestPage % SLRU_PAGES_PER_SEGMENT), oldestPage)); @@ -1483,12 +241,12 @@ SlruPagePrecedesTestOffset(SlruCtl ctl, int per_page, uint32 offset) * do not apply to them.) */ void -SlruPagePrecedesUnitTests(SlruCtl ctl, int per_page) +SlruPagePrecedesUnitTests(SlruPagePrecedesFunction PagePrecedes, int per_page) { /* Test first, middle and last entries of a page. */ - SlruPagePrecedesTestOffset(ctl, per_page, 0); - SlruPagePrecedesTestOffset(ctl, per_page, per_page / 2); - SlruPagePrecedesTestOffset(ctl, per_page, per_page - 1); + SlruPagePrecedesTestOffset(PagePrecedes, per_page, 0); + SlruPagePrecedesTestOffset(PagePrecedes, per_page, per_page / 2); + SlruPagePrecedesTestOffset(PagePrecedes, per_page, per_page - 1); } #endif @@ -1498,11 +256,12 @@ SlruPagePrecedesUnitTests(SlruCtl ctl, int per_page) * one containing the page passed as "data". */ bool -SlruScanDirCbReportPresence(SlruCtl ctl, char *filename, int segpage, void *data) +SlruScanDirCbReportPresence(int slru_id, SlruPagePrecedesFunction PagePrecedes, + char *filename, int segpage, void *data) { int cutoffPage = *(int *) data; - if (SlruMayDeleteSegment(ctl, segpage, cutoffPage)) + if (SlruMayDeleteSegment(PagePrecedes, segpage, cutoffPage)) return true; /* found one; don't iterate any more */ return false; /* keep going */ @@ -1513,12 +272,15 @@ SlruScanDirCbReportPresence(SlruCtl ctl, char *filename, int segpage, void *data * This callback deletes segments prior to the one passed in as "data". */ static bool -SlruScanDirCbDeleteCutoff(SlruCtl ctl, char *filename, int segpage, void *data) +SlruScanDirCbDeleteCutoff(int slru_id, SlruPagePrecedesFunction PagePrecedes, + char *filename, int segpage, void *data) { int cutoffPage = *(int *) data; - if (SlruMayDeleteSegment(ctl, segpage, cutoffPage)) - SlruInternalDeleteSegment(ctl, segpage / SLRU_PAGES_PER_SEGMENT); + if (SlruMayDeleteSegment(PagePrecedes, segpage, cutoffPage)) + { + SlruDeleteSegment(slru_id, segpage / SLRU_PAGES_PER_SEGMENT); + } return false; /* keep going */ } @@ -1528,9 +290,10 @@ SlruScanDirCbDeleteCutoff(SlruCtl ctl, char *filename, int segpage, void *data) * This callback deletes all segments. */ bool -SlruScanDirCbDeleteAll(SlruCtl ctl, char *filename, int segpage, void *data) +SlruScanDirCbDeleteAll(int slru_id, SlruPagePrecedesFunction PagePrecedes, + char *filename, int segpage, void *data) { - SlruInternalDeleteSegment(ctl, segpage / SLRU_PAGES_PER_SEGMENT); + SlruInternalDeleteSegment(slru_id, segpage / SLRU_PAGES_PER_SEGMENT); return false; /* keep going */ } @@ -1551,16 +314,20 @@ SlruScanDirCbDeleteAll(SlruCtl ctl, char *filename, int segpage, void *data) * Note that no locking is applied. */ bool -SlruScanDirectory(SlruCtl ctl, SlruScanCallback callback, void *data) +SlruScanDirectory(int slru_id, SlruPagePrecedesFunction PagePrecedes, + SlruScanCallback callback, void *data) { bool retval = false; DIR *cldir; struct dirent *clde; int segno; int segpage; + const char *path; + + path = slru_dirs[slru_id]; - cldir = AllocateDir(ctl->Dir); - while ((clde = ReadDir(cldir, ctl->Dir)) != NULL) + cldir = AllocateDir(path); + while ((clde = ReadDir(cldir, path)) != NULL) { size_t len; @@ -1573,8 +340,8 @@ SlruScanDirectory(SlruCtl ctl, SlruScanCallback callback, void *data) segpage = segno * SLRU_PAGES_PER_SEGMENT; elog(DEBUG2, "SlruScanDirectory invoking callback on %s/%s", - ctl->Dir, clde->d_name); - retval = callback(ctl, clde->d_name, segpage, data); + path, clde->d_name); + retval = callback(slru_id, PagePrecedes, clde->d_name, segpage, data); if (retval) break; } @@ -1585,29 +352,74 @@ SlruScanDirectory(SlruCtl ctl, SlruScanCallback callback, void *data) } /* - * Individual SLRUs (clog, ...) have to provide a sync.c handler function so - * that they can provide the correct "SlruCtl" (otherwise we don't know how to - * build the path), but they just forward to this common implementation that - * performs the fsync. + * Read a buffer. Buffer is pinned on return. + */ +Buffer +ReadSlruBuffer(int slru_id, int pageno, ReadBufferMode mode) +{ + int segno = pageno / SLRU_PAGES_PER_SEGMENT; + int rpageno = pageno % SLRU_PAGES_PER_SEGMENT; + RelFileLocator rlocator = SlruRelFileLocator(slru_id, segno); + Buffer buffer; + bool hit; + + /* Try to avoid doing a buffer mapping table lookup for repeated access. */ + buffer = slru_recent_buffers[slru_id].recent_buffer; + if (slru_recent_buffers[slru_id].pageno == pageno && + BufferIsValid(buffer) && + ReadRecentBuffer(rlocator, MAIN_FORKNUM, pageno, buffer)) + { + pgstat_count_slru_page_hit(slru_id); + return buffer; + } + + /* Regular lookup. */ + buffer = ReadBufferWithoutRelcacheWithHit(rlocator, MAIN_FORKNUM, rpageno, + mode, NULL, true, &hit); + + /* Remember where this page is for next time. */ + slru_recent_buffers[slru_id].pageno = pageno; + slru_recent_buffers[slru_id].recent_buffer = buffer; + + if (hit) + pgstat_count_slru_page_hit(slru_id); + + return buffer; +} + +/* + * Zero-initialize a buffer. Buffer is pinned and exclusively locked on return. */ -int -SlruSyncFileTag(SlruCtl ctl, const FileTag *ftag, char *path) +Buffer +ZeroSlruBuffer(int slru_id, int pageno) { - int fd; - int save_errno; - int result; + int segno = pageno / SLRU_PAGES_PER_SEGMENT; + int rpageno = pageno % SLRU_PAGES_PER_SEGMENT; + RelFileLocator rlocator = SlruRelFileLocator(slru_id, segno); + Buffer buffer; + SMgrFileHandle sfile; - SlruFileName(ctl, path, ftag->segno); + sfile = smgropen(rlocator, InvalidBackendId, MAIN_FORKNUM); + if (!smgrexists(sfile)) + smgrcreate(sfile, false); + + buffer = ReadBufferWithoutRelcache(rlocator, MAIN_FORKNUM, rpageno, RBM_ZERO_AND_LOCK, NULL, true); - fd = OpenTransientFile(path, O_RDWR | PG_BINARY); - if (fd < 0) - return -1; + /* Remember where this page is for next time. */ + slru_recent_buffers[slru_id].pageno = pageno; + slru_recent_buffers[slru_id].recent_buffer = buffer; - result = pg_fsync(fd); - save_errno = errno; + pgstat_count_slru_page_zeroed(slru_id); - CloseTransientFile(fd); + return buffer; +} + +bool +ProbeSlruBuffer(int slru_id, int pageno) +{ + int segno = pageno / SLRU_PAGES_PER_SEGMENT; + int rpageno = pageno % SLRU_PAGES_PER_SEGMENT; + RelFileLocator rlocator = SlruRelFileLocator(slru_id, segno); - errno = save_errno; - return result; + return BufferProbe(rlocator, MAIN_FORKNUM, rpageno); } diff --git a/src/backend/access/transam/subtrans.c b/src/backend/access/transam/subtrans.c index 62bb610167c..33ede8138d6 100644 --- a/src/backend/access/transam/subtrans.c +++ b/src/backend/access/transam/subtrans.c @@ -32,6 +32,7 @@ #include "access/subtrans.h" #include "access/transam.h" #include "pg_trace.h" +#include "storage/bufmgr.h" #include "utils/snapmgr.h" @@ -49,21 +50,13 @@ */ /* We need four bytes per xact */ -#define SUBTRANS_XACTS_PER_PAGE (BLCKSZ / sizeof(TransactionId)) +#define SUBTRANS_XACTS_PER_PAGE ((BLCKSZ - SizeOfPageHeaderData) / sizeof(TransactionId)) #define TransactionIdToPage(xid) ((xid) / (TransactionId) SUBTRANS_XACTS_PER_PAGE) #define TransactionIdToEntry(xid) ((xid) % (TransactionId) SUBTRANS_XACTS_PER_PAGE) -/* - * Link to shared-memory data structures for SUBTRANS control - */ -static SlruCtlData SubTransCtlData; - -#define SubTransCtl (&SubTransCtlData) - - -static int ZeroSUBTRANSPage(int pageno); +static Buffer ZeroSUBTRANSPage(int pageno); static bool SubTransPagePrecedes(int page1, int page2); @@ -75,16 +68,15 @@ SubTransSetParent(TransactionId xid, TransactionId parent) { int pageno = TransactionIdToPage(xid); int entryno = TransactionIdToEntry(xid); - int slotno; TransactionId *ptr; + Buffer buffer; Assert(TransactionIdIsValid(parent)); Assert(TransactionIdFollows(xid, parent)); - LWLockAcquire(SubtransSLRULock, LW_EXCLUSIVE); - - slotno = SimpleLruReadPage(SubTransCtl, pageno, true, xid); - ptr = (TransactionId *) SubTransCtl->shared->page_buffer[slotno]; + buffer = ReadSlruBuffer(SLRU_SUBTRANS_ID, pageno, RBM_NORMAL); + LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE); + ptr = (TransactionId *) PageGetContents(BufferGetPage(buffer)); ptr += entryno; /* @@ -96,10 +88,10 @@ SubTransSetParent(TransactionId xid, TransactionId parent) { Assert(*ptr == InvalidTransactionId); *ptr = parent; - SubTransCtl->shared->page_dirty[slotno] = true; + MarkBufferDirty(buffer); } - LWLockRelease(SubtransSLRULock); + UnlockReleaseBuffer(buffer); } /* @@ -110,9 +102,9 @@ SubTransGetParent(TransactionId xid) { int pageno = TransactionIdToPage(xid); int entryno = TransactionIdToEntry(xid); - int slotno; TransactionId *ptr; TransactionId parent; + Buffer buffer; /* Can't ask about stuff that might not be around anymore */ Assert(TransactionIdFollowsOrEquals(xid, TransactionXmin)); @@ -121,15 +113,14 @@ SubTransGetParent(TransactionId xid) if (!TransactionIdIsNormal(xid)) return InvalidTransactionId; - /* lock is acquired by SimpleLruReadPage_ReadOnly */ + buffer = ReadSlruBuffer(SLRU_SUBTRANS_ID, pageno, RBM_NORMAL); - slotno = SimpleLruReadPage_ReadOnly(SubTransCtl, pageno, xid); - ptr = (TransactionId *) SubTransCtl->shared->page_buffer[slotno]; + ptr = (TransactionId *) PageGetContents(BufferGetPage(buffer)); ptr += entryno; parent = *ptr; - LWLockRelease(SubtransSLRULock); + ReleaseBuffer(buffer); return parent; } @@ -177,26 +168,6 @@ SubTransGetTopmostTransaction(TransactionId xid) return previousXid; } - -/* - * Initialization of shared memory for SUBTRANS - */ -Size -SUBTRANSShmemSize(void) -{ - return SimpleLruShmemSize(NUM_SUBTRANS_BUFFERS, 0); -} - -void -SUBTRANSShmemInit(void) -{ - SubTransCtl->PagePrecedes = SubTransPagePrecedes; - SimpleLruInit(SubTransCtl, "Subtrans", NUM_SUBTRANS_BUFFERS, 0, - SubtransSLRULock, "pg_subtrans", - LWTRANCHE_SUBTRANS_BUFFER, SYNC_HANDLER_NONE); - SlruPagePrecedesUnitTests(SubTransCtl, SUBTRANS_XACTS_PER_PAGE); -} - /* * This func must be called ONCE on system install. It creates * the initial SUBTRANS segment. (The SUBTRANS directory is assumed to @@ -210,18 +181,16 @@ SUBTRANSShmemInit(void) void BootStrapSUBTRANS(void) { - int slotno; + Buffer buffer; - LWLockAcquire(SubtransSLRULock, LW_EXCLUSIVE); + SlruPagePrecedesUnitTests(SubTransPagePrecedes, SUBTRANS_XACTS_PER_PAGE); /* Create and zero the first page of the subtrans log */ - slotno = ZeroSUBTRANSPage(0); + buffer = ZeroSUBTRANSPage(0); /* Make sure it's written out */ - SimpleLruWritePage(SubTransCtl, slotno); - Assert(!SubTransCtl->shared->page_dirty[slotno]); - - LWLockRelease(SubtransSLRULock); + FlushOneBuffer(buffer); + UnlockReleaseBuffer(buffer); } /* @@ -232,10 +201,19 @@ BootStrapSUBTRANS(void) * * Control lock must be held at entry, and will be held at exit. */ -static int +static Buffer ZeroSUBTRANSPage(int pageno) { - return SimpleLruZeroPage(SubTransCtl, pageno); + Buffer buffer; + Page page; + + buffer = ZeroSlruBuffer(SLRU_SUBTRANS_ID, pageno); + page = BufferGetPage(buffer); + PageInitSLRU(page, BLCKSZ, 0); + + MarkBufferDirty(buffer); + + return buffer; } /* @@ -258,7 +236,6 @@ StartupSUBTRANS(TransactionId oldestActiveXID) * Whenever we advance into a new page, ExtendSUBTRANS will likewise zero * the new page without regard to whatever was previously on disk. */ - LWLockAcquire(SubtransSLRULock, LW_EXCLUSIVE); startPage = TransactionIdToPage(oldestActiveXID); nextXid = ShmemVariableCache->nextXid; @@ -266,36 +243,15 @@ StartupSUBTRANS(TransactionId oldestActiveXID) while (startPage != endPage) { - (void) ZeroSUBTRANSPage(startPage); + UnlockReleaseBuffer(ZeroSUBTRANSPage(startPage)); startPage++; /* must account for wraparound */ if (startPage > TransactionIdToPage(MaxTransactionId)) startPage = 0; } - (void) ZeroSUBTRANSPage(startPage); - - LWLockRelease(SubtransSLRULock); -} - -/* - * Perform a checkpoint --- either during shutdown, or on-the-fly - */ -void -CheckPointSUBTRANS(void) -{ - /* - * Write dirty SUBTRANS pages to disk - * - * This is not actually necessary from a correctness point of view. We do - * it merely to improve the odds that writing of dirty pages is done by - * the checkpoint process and not by backends. - */ - TRACE_POSTGRESQL_SUBTRANS_CHECKPOINT_START(true); - SimpleLruWriteAll(SubTransCtl, true); - TRACE_POSTGRESQL_SUBTRANS_CHECKPOINT_DONE(true); + UnlockReleaseBuffer(ZeroSUBTRANSPage(startPage)); } - /* * Make sure that SUBTRANS has room for a newly-allocated XID. * @@ -319,12 +275,8 @@ ExtendSUBTRANS(TransactionId newestXact) pageno = TransactionIdToPage(newestXact); - LWLockAcquire(SubtransSLRULock, LW_EXCLUSIVE); - /* Zero the page */ - ZeroSUBTRANSPage(pageno); - - LWLockRelease(SubtransSLRULock); + UnlockReleaseBuffer(ZeroSUBTRANSPage(pageno)); } @@ -350,7 +302,7 @@ TruncateSUBTRANS(TransactionId oldestXact) TransactionIdRetreat(oldestXact); cutoffPage = TransactionIdToPage(oldestXact); - SimpleLruTruncate(SubTransCtl, cutoffPage); + SimpleLruTruncate(SLRU_SUBTRANS_ID, SubTransPagePrecedes, cutoffPage); } diff --git a/src/backend/access/transam/xact.c b/src/backend/access/transam/xact.c index d85e3139082..1f179ba2329 100644 --- a/src/backend/access/transam/xact.c +++ b/src/backend/access/transam/xact.c @@ -68,6 +68,7 @@ #include "utils/inval.h" #include "utils/memutils.h" #include "utils/relmapper.h" +#include "utils/resowner_private.h" #include "utils/snapmgr.h" #include "utils/timeout.h" #include "utils/timestamp.h" @@ -1397,6 +1398,7 @@ RecordTransactionCommit(void) * are delaying the checkpoint a bit fuzzy, but it doesn't matter. */ Assert((MyProc->delayChkptFlags & DELAY_CHKPT_START) == 0); + START_CRIT_SECTION(); MyProc->delayChkptFlags |= DELAY_CHKPT_START; diff --git a/src/backend/access/transam/xlog.c b/src/backend/access/transam/xlog.c index 0070d56b0b0..7b92124d7fd 100644 --- a/src/backend/access/transam/xlog.c +++ b/src/backend/access/transam/xlog.c @@ -4636,6 +4636,7 @@ BootStrapXLOG(void) uint64 sysidentifier; struct timeval tv; pg_crc32c crc; + ResourceOwner resowner; /* allow ordinary WAL segment creation, like StartupXLOG() would */ SetInstallXLogFileSegmentActive(); @@ -4775,10 +4776,14 @@ BootStrapXLOG(void) WriteControlFile(); /* Bootstrap the commit log, too */ + resowner = ResourceOwnerCreate(NULL, "bootstrap resowner"); + CurrentResourceOwner = resowner; BootStrapCLOG(); BootStrapCommitTs(); BootStrapSUBTRANS(); BootStrapMultiXact(); + CurrentResourceOwner = NULL; + ResourceOwnerDelete(resowner); pfree(buffer); @@ -4787,6 +4792,8 @@ BootStrapXLOG(void) * otherwise never run the checks and GUC related initializations therein. */ ReadControlFile(); + + smgrcloseall(); } static char * @@ -6995,15 +7002,11 @@ CheckPointGuts(XLogRecPtr checkPointRedo, int flags) CheckPointSnapBuild(); CheckPointLogicalRewriteHeap(); CheckPointReplicationOrigin(); + CheckPointPredicate(); - /* Write out all dirty data in SLRUs and the main buffer pool */ + /* Write out all dirty data in the buffer pool */ TRACE_POSTGRESQL_BUFFER_CHECKPOINT_START(flags); CheckpointStats.ckpt_write_t = GetCurrentTimestamp(); - CheckPointCLOG(); - CheckPointCommitTs(); - CheckPointSUBTRANS(); - CheckPointMultiXact(); - CheckPointPredicate(); CheckPointBuffers(flags); /* Perform all queued up fsyncs */ diff --git a/src/backend/access/transam/xlogprefetcher.c b/src/backend/access/transam/xlogprefetcher.c index 046e40d143a..50674cd8b53 100644 --- a/src/backend/access/transam/xlogprefetcher.c +++ b/src/backend/access/transam/xlogprefetcher.c @@ -652,7 +652,7 @@ XLogPrefetcherNextBlock(uintptr_t pgsr_private, XLogRecPtr *lsn) { int block_id = prefetcher->next_block_id++; DecodedBkpBlock *block = &record->blocks[block_id]; - SMgrRelation reln; + SMgrFileHandle sfile; PrefetchBufferResult result; if (!block->in_use) @@ -722,7 +722,7 @@ XLogPrefetcherNextBlock(uintptr_t pgsr_private, XLogRecPtr *lsn) * same relation (with some scheme to handle invalidations * safely), but for now we'll call smgropen() every time. */ - reln = smgropen(block->rlocator, InvalidBackendId); + sfile = smgropen(block->rlocator, InvalidBackendId, block->forknum); /* * If the relation file doesn't exist on disk, for example because @@ -731,14 +731,14 @@ XLogPrefetcherNextBlock(uintptr_t pgsr_private, XLogRecPtr *lsn) * further prefetching in the relation until this record is * replayed. */ - if (!smgrexists(reln, MAIN_FORKNUM)) + if (!smgrexists(sfile)) { #ifdef XLOGPREFETCHER_DEBUG_LEVEL elog(XLOGPREFETCHER_DEBUG_LEVEL, "suppressing all prefetch in relation %u/%u/%u until %X/%X is replayed, because the relation does not exist on disk", - reln->smgr_rlocator.locator.spcOid, - reln->smgr_rlocator.locator.dbOid, - reln->smgr_rlocator.locator.relNumber, + sfile->smgr_rlocator.locator.spcOid, + sfile->smgr_rlocator.locator.dbOid, + sfile->smgr_rlocator.locator.relNumber, LSN_FORMAT_ARGS(record->lsn)); #endif XLogPrefetcherAddFilter(prefetcher, block->rlocator, 0, @@ -752,14 +752,14 @@ XLogPrefetcherNextBlock(uintptr_t pgsr_private, XLogRecPtr *lsn) * block yet, suppress prefetching of this block and higher until * this record is replayed. */ - if (block->blkno >= smgrnblocks(reln, block->forknum)) + if (block->blkno >= smgrnblocks(sfile)) { #ifdef XLOGPREFETCHER_DEBUG_LEVEL elog(XLOGPREFETCHER_DEBUG_LEVEL, "suppressing prefetch in relation %u/%u/%u from block %u until %X/%X is replayed, because the relation is too small", - reln->smgr_rlocator.locator.spcOid, - reln->smgr_rlocator.locator.dbOid, - reln->smgr_rlocator.locator.relNumber, + sfile->smgr_rlocator.locator.spcOid, + sfile->smgr_rlocator.locator.dbOid, + sfile->smgr_rlocator.locator.relNumber, block->blkno, LSN_FORMAT_ARGS(record->lsn)); #endif @@ -770,7 +770,7 @@ XLogPrefetcherNextBlock(uintptr_t pgsr_private, XLogRecPtr *lsn) } /* Try to initiate prefetching. */ - result = PrefetchSharedBuffer(reln, block->forknum, block->blkno); + result = PrefetchSharedBuffer(sfile, block->blkno); if (BufferIsValid(result.recent_buffer)) { /* Cache hit, nothing to do. */ @@ -796,9 +796,9 @@ XLogPrefetcherNextBlock(uintptr_t pgsr_private, XLogRecPtr *lsn) */ elog(ERROR, "could not prefetch relation %u/%u/%u block %u", - reln->smgr_rlocator.locator.spcOid, - reln->smgr_rlocator.locator.dbOid, - reln->smgr_rlocator.locator.relNumber, + sfile->smgr_locator.locator.spcOid, + sfile->smgr_locator.locator.dbOid, + sfile->smgr_locator.locator.relNumber, block->blkno); } } diff --git a/src/backend/access/transam/xlogutils.c b/src/backend/access/transam/xlogutils.c index 87ca9dfef16..93a01ad844a 100644 --- a/src/backend/access/transam/xlogutils.c +++ b/src/backend/access/transam/xlogutils.c @@ -477,7 +477,7 @@ XLogReadBufferExtended(RelFileLocator rlocator, ForkNumber forknum, { BlockNumber lastblock; Buffer buffer; - SMgrRelation smgr; + SMgrFileHandle sfile; Assert(blkno != P_NEW); @@ -491,7 +491,7 @@ XLogReadBufferExtended(RelFileLocator rlocator, ForkNumber forknum, } /* Open the relation at smgr level */ - smgr = smgropen(rlocator, InvalidBackendId); + sfile = smgropen(rlocator, InvalidBackendId, forknum); /* * Create the target file if it doesn't already exist. This lets us cope @@ -501,9 +501,9 @@ XLogReadBufferExtended(RelFileLocator rlocator, ForkNumber forknum, * filesystem loses an inode during a crash. Better to write the data * until we are actually told to delete the file.) */ - smgrcreate(smgr, forknum, true); + smgrcreate(sfile, true); - lastblock = smgrnblocks(smgr, forknum); + lastblock = smgrnblocks(sfile); if (blkno < lastblock) { @@ -631,7 +631,7 @@ CreateFakeRelcacheEntry(RelFileLocator rlocator) rel->rd_lockInfo.lockRelId.dbId = rlocator.dbOid; rel->rd_lockInfo.lockRelId.relId = rlocator.relNumber; - rel->rd_smgr = NULL; + MemSet(rel->rd_smgr, 0, sizeof(rel->rd_smgr)); return rel; } @@ -643,8 +643,11 @@ void FreeFakeRelcacheEntry(Relation fakerel) { /* make sure the fakerel is not referenced by the SmgrRelation anymore */ - if (fakerel->rd_smgr != NULL) - smgrclearowner(&fakerel->rd_smgr, fakerel->rd_smgr); + for (int i = 0; i <= MAX_FORKNUM; i++) + { + if (fakerel->rd_smgr[i] != NULL) + smgrclearowner(&fakerel->rd_smgr[i], fakerel->rd_smgr[i]); + } pfree(fakerel); } diff --git a/src/backend/catalog/catalog.c b/src/backend/catalog/catalog.c index 1bf6c5633cd..6b01158615b 100644 --- a/src/backend/catalog/catalog.c +++ b/src/backend/catalog/catalog.c @@ -501,10 +501,10 @@ GetNewOidWithIndex(Relation relation, Oid indexId, AttrNumber oidcolumn) RelFileNumber GetNewRelFileNumber(Oid reltablespace, Relation pg_class, char relpersistence) { - RelFileLocatorBackend rlocator; - char *rpath; - bool collides; + RelFileLocator rlocator; BackendId backend; + SMgrFileHandle sfile; + bool collides; /* * If we ever get here during pg_upgrade, there's something wrong; all @@ -513,6 +513,11 @@ GetNewRelFileNumber(Oid reltablespace, Relation pg_class, char relpersistence) */ Assert(!IsBinaryUpgrade); + /* + * The relpath will vary based on the backend ID, so we must initialize + * that properly here to make sure that any collisions based on filename + * are properly detected. + */ switch (relpersistence) { case RELPERSISTENCE_TEMP: @@ -528,53 +533,29 @@ GetNewRelFileNumber(Oid reltablespace, Relation pg_class, char relpersistence) } /* This logic should match RelationInitPhysicalAddr */ - rlocator.locator.spcOid = reltablespace ? reltablespace : MyDatabaseTableSpace; - rlocator.locator.dbOid = - (rlocator.locator.spcOid == GLOBALTABLESPACE_OID) ? + rlocator.spcOid = reltablespace ? reltablespace : MyDatabaseTableSpace; + rlocator.dbOid = + (rlocator.spcOid == GLOBALTABLESPACE_OID) ? InvalidOid : MyDatabaseId; - /* - * The relpath will vary based on the backend ID, so we must initialize - * that properly here to make sure that any collisions based on filename - * are properly detected. - */ - rlocator.backend = backend; - do { CHECK_FOR_INTERRUPTS(); /* Generate the OID */ if (pg_class) - rlocator.locator.relNumber = GetNewOidWithIndex(pg_class, ClassOidIndexId, + rlocator.relNumber = GetNewOidWithIndex(pg_class, ClassOidIndexId, Anum_pg_class_oid); else - rlocator.locator.relNumber = GetNewObjectId(); + rlocator.relNumber = GetNewObjectId(); /* Check for existing file of same name */ - rpath = relpath(rlocator, MAIN_FORKNUM); - - if (access(rpath, F_OK) == 0) - { - /* definite collision */ - collides = true; - } - else - { - /* - * Here we have a little bit of a dilemma: if errno is something - * other than ENOENT, should we declare a collision and loop? In - * practice it seems best to go ahead regardless of the errno. If - * there is a colliding file we will get an smgr failure when we - * attempt to create the new relation file. - */ - collides = false; - } - - pfree(rpath); + sfile = smgropen(rlocator, backend, MAIN_FORKNUM); + collides = smgrexists(sfile); + smgrclose(sfile); } while (collides); - return rlocator.locator.relNumber; + return rlocator.relNumber; } /* diff --git a/src/backend/catalog/index.c b/src/backend/catalog/index.c index e6579f29793..683210067b2 100644 --- a/src/backend/catalog/index.c +++ b/src/backend/catalog/index.c @@ -3029,9 +3029,9 @@ index_build(Relation heapRelation, * relfilenumber won't change, and nothing needs to be done here. */ if (indexRelation->rd_rel->relpersistence == RELPERSISTENCE_UNLOGGED && - !smgrexists(RelationGetSmgr(indexRelation), INIT_FORKNUM)) + !smgrexists(RelationGetSmgr(indexRelation, INIT_FORKNUM))) { - smgrcreate(RelationGetSmgr(indexRelation), INIT_FORKNUM, false); + smgrcreate(RelationGetSmgr(indexRelation, INIT_FORKNUM), false); indexRelation->rd_indam->ambuildempty(indexRelation); } diff --git a/src/backend/catalog/storage.c b/src/backend/catalog/storage.c index 039ada94c4f..7e45a3d1ec8 100644 --- a/src/backend/catalog/storage.c +++ b/src/backend/catalog/storage.c @@ -116,11 +116,11 @@ AddPendingSync(const RelFileLocator *rlocator) * that does not want the storage to be destroyed in case of an abort may * pass register_delete = false. */ -SMgrRelation +SMgrFileHandle RelationCreateStorage(RelFileLocator rlocator, char relpersistence, bool register_delete) { - SMgrRelation srel; + SMgrFileHandle sfile; BackendId backend; bool needs_wal; @@ -145,11 +145,11 @@ RelationCreateStorage(RelFileLocator rlocator, char relpersistence, return NULL; /* placate compiler */ } - srel = smgropen(rlocator, backend); - smgrcreate(srel, MAIN_FORKNUM, false); + sfile = smgropen(rlocator, backend, MAIN_FORKNUM); + smgrcreate(sfile, false); if (needs_wal) - log_smgrcreate(&srel->smgr_rlocator.locator, MAIN_FORKNUM); + log_smgrcreate(&rlocator, MAIN_FORKNUM); /* * Add the relation to the list of stuff to delete at abort, if we are @@ -175,7 +175,7 @@ RelationCreateStorage(RelFileLocator rlocator, char relpersistence, AddPendingSync(&rlocator); } - return srel; + return sfile; } /* @@ -292,16 +292,18 @@ RelationTruncate(Relation rel, BlockNumber nblocks) ForkNumber forks[MAX_FORKNUM]; BlockNumber blocks[MAX_FORKNUM]; int nforks = 0; - SMgrRelation reln; /* * Make sure smgr_targblock etc aren't pointing somewhere past new end. * (Note: don't rely on this reln pointer below this loop.) */ - reln = RelationGetSmgr(rel); - reln->smgr_targblock = InvalidBlockNumber; - for (int i = 0; i <= MAX_FORKNUM; ++i) - reln->smgr_cached_nblocks[i] = InvalidBlockNumber; + for (int i = 0; i <= MAX_FORKNUM; i++) + { + SMgrFileHandle sfile = RelationGetSmgr(rel, i); + + sfile->smgr_targblock = InvalidBlockNumber; + sfile->smgr_cached_nblocks = InvalidBlockNumber; + } /* Prepare for truncation of MAIN fork of the relation */ forks[nforks] = MAIN_FORKNUM; @@ -309,7 +311,7 @@ RelationTruncate(Relation rel, BlockNumber nblocks) nforks++; /* Prepare for truncation of the FSM if it exists */ - fsm = smgrexists(RelationGetSmgr(rel), FSM_FORKNUM); + fsm = smgrexists(RelationGetSmgr(rel, FSM_FORKNUM)); if (fsm) { blocks[nforks] = FreeSpaceMapPrepareTruncateRel(rel, nblocks); @@ -322,7 +324,7 @@ RelationTruncate(Relation rel, BlockNumber nblocks) } /* Prepare for truncation of the visibility map too if it exists */ - vm = smgrexists(RelationGetSmgr(rel), VISIBILITYMAP_FORKNUM); + vm = smgrexists(RelationGetSmgr(rel, VISIBILITYMAP_FORKNUM)); if (vm) { blocks[nforks] = visibilitymap_prepare_truncate(rel, nblocks); @@ -390,11 +392,12 @@ RelationTruncate(Relation rel, BlockNumber nblocks) } /* - * This will first remove any buffers from the buffer pool that should no + * First remove any buffers from the buffer pool that should no * longer exist after truncation is complete, and then truncate the * corresponding files on disk. */ - smgrtruncate(RelationGetSmgr(rel), forks, nforks, blocks); + DropRelationBuffers(rel->rd_locator, rel->rd_backend, forks, nforks, blocks); + smgrtruncate_multi(rel->rd_locator, rel->rd_backend, forks, nforks, blocks); /* We've done all the critical work, so checkpoints are OK now. */ MyProc->delayChkptFlags &= ~DELAY_CHKPT_COMPLETE; @@ -428,7 +431,7 @@ RelationPreTruncate(Relation rel) return; pending = hash_search(pendingSyncHash, - &(RelationGetSmgr(rel)->smgr_rlocator.locator), + &rel->rd_locator, HASH_FIND, NULL); if (pending) pending->is_truncated = true; @@ -444,12 +447,12 @@ RelationPreTruncate(Relation rel) * Also note that this is frequently called via locutions such as * RelationCopyStorage(RelationGetSmgr(rel), ...); * That's safe only because we perform only smgr and WAL operations here. - * If we invoked anything else, a relcache flush could cause our SMgrRelation + * If we invoked anything else, a relcache flush could cause our SMgrFileHandle * argument to become a dangling pointer. */ void -RelationCopyStorage(SMgrRelation src, SMgrRelation dst, - ForkNumber forkNum, char relpersistence) +RelationCopyStorage(SMgrFileHandle src, SMgrFileHandle dst, + char relpersistence) { PGAlignedBlock buf; Page page; @@ -466,7 +469,7 @@ RelationCopyStorage(SMgrRelation src, SMgrRelation dst, * it needs to be synced to disk. */ copying_initfork = relpersistence == RELPERSISTENCE_UNLOGGED && - forkNum == INIT_FORKNUM; + src->smgr_locator.forknum == INIT_FORKNUM; /* * We need to log the copied data in WAL iff WAL archiving/streaming is @@ -477,14 +480,14 @@ RelationCopyStorage(SMgrRelation src, SMgrRelation dst, use_wal = XLogIsNeeded() && (relpersistence == RELPERSISTENCE_PERMANENT || copying_initfork); - nblocks = smgrnblocks(src, forkNum); + nblocks = smgrnblocks(src); for (blkno = 0; blkno < nblocks; blkno++) { /* If we got a cancel signal during the copy of the data, quit */ CHECK_FOR_INTERRUPTS(); - smgrread(src, forkNum, blkno, buf.data); + smgrread(src, blkno, buf.data); if (!PageIsVerifiedExtended(page, blkno, PIV_LOG_WARNING | PIV_REPORT_STAT)) @@ -496,9 +499,7 @@ RelationCopyStorage(SMgrRelation src, SMgrRelation dst, * (errcontext callbacks shouldn't be risking any such thing, but * people have been known to forget that rule.) */ - char *relpath = relpathbackend(src->smgr_rlocator.locator, - src->smgr_rlocator.backend, - forkNum); + char *relpath = smgrfilepath(src->smgr_locator); ereport(ERROR, (errcode(ERRCODE_DATA_CORRUPTED), @@ -512,7 +513,7 @@ RelationCopyStorage(SMgrRelation src, SMgrRelation dst, * space. */ if (use_wal) - log_newpage(&dst->smgr_rlocator.locator, forkNum, blkno, page, false); + log_newpage(&dst->smgr_locator.locator, dst->smgr_locator.forknum, blkno, page, false); PageSetChecksumInplace(page, blkno); @@ -521,7 +522,7 @@ RelationCopyStorage(SMgrRelation src, SMgrRelation dst, * need for smgr to schedule an fsync for this write; we'll do it * ourselves below. */ - smgrextend(dst, forkNum, blkno, buf.data, true); + smgrextend(dst, blkno, buf.data, true); } /* @@ -534,7 +535,7 @@ RelationCopyStorage(SMgrRelation src, SMgrRelation dst, * they might still not be on disk when the crash occurs. */ if (use_wal || copying_initfork) - smgrimmedsync(dst, forkNum); + smgrimmedsync(dst); } /* @@ -653,9 +654,9 @@ smgrDoPendingDeletes(bool isCommit) PendingRelDelete *pending; PendingRelDelete *prev; PendingRelDelete *next; - int nrels = 0, - maxrels = 0; - SMgrRelation *srels = NULL; + int nlocators = 0, + maxlocators = 0; + RelFileLocatorBackend *locators = NULL; prev = NULL; for (pending = pendingDeletes; pending != NULL; pending = next) @@ -676,23 +677,21 @@ smgrDoPendingDeletes(bool isCommit) /* do deletion if called for */ if (pending->atCommit == isCommit) { - SMgrRelation srel; - - srel = smgropen(pending->rlocator, pending->backend); + RelFileLocatorBackend rlocator = { pending->rlocator, pending->backend }; /* allocate the initial array, or extend it, if needed */ - if (maxrels == 0) + if (maxlocators == 0) { - maxrels = 8; - srels = palloc(sizeof(SMgrRelation) * maxrels); + maxlocators = 8; + locators = palloc(sizeof(RelFileLocatorBackend) * maxlocators); } - else if (maxrels <= nrels) + else if (maxlocators <= nlocators) { - maxrels *= 2; - srels = repalloc(srels, sizeof(SMgrRelation) * maxrels); + maxlocators *= 2; + locators = repalloc(locators, sizeof(RelFileLocatorBackend) * maxlocators); } - srels[nrels++] = srel; + locators[nlocators++] = rlocator; } /* must explicitly free the list entry */ pfree(pending); @@ -700,15 +699,58 @@ smgrDoPendingDeletes(bool isCommit) } } - if (nrels > 0) + if (nlocators > 0) { - smgrdounlinkall(srels, nrels, false); + ForkNumber forks[MAX_FORKNUM + 1]; + + for (int i = 0; i <= MAX_FORKNUM; i++) + forks[i] = i; + + /* + * Get rid of any remaining buffers for the relations. bufmgr will just + * drop them without bothering to write the contents. + */ + DropRelationsAllBuffers(locators, nlocators); - for (int i = 0; i < nrels; i++) - smgrclose(srels[i]); + for (int i = 0; i < nlocators; i++) + smgrunlink_multi(locators[i].locator, locators[i].backend, forks, MAX_FORKNUM + 1, false); + pfree(locators); + } +} + +/* + * DropRelationFiles -- drop files of all given relations + */ +void +DropRelationFiles(RelFileLocator *delrels, int ndelrels, bool isRedo) +{ + RelFileLocatorBackend *locators; + int i; + ForkNumber all_forks[MAX_FORKNUM + 1]; - pfree(srels); + locators = palloc(sizeof(RelFileLocatorBackend) * ndelrels); + for (i = 0; i < ndelrels; i++) + { + if (isRedo) + { + for (int fork = 0; fork <= MAX_FORKNUM; fork++) + XLogDropRelation(delrels[i], fork); + } + locators[i].locator = delrels[i]; + locators[i].backend = InvalidBackendId; } + + /* + * Get rid of any remaining buffers for the relations. bufmgr will just + * drop them without bothering to write the contents. + */ + DropRelationsAllBuffers(locators, ndelrels); + + for (int fork = 0; fork <= MAX_FORKNUM; fork++) + all_forks[fork] = fork; + + for (i = 0; i < ndelrels; i++) + smgrunlink_multi(locators[i].locator, locators[i].backend, all_forks, MAX_FORKNUM + 1, true); } /* @@ -718,9 +760,9 @@ void smgrDoPendingSyncs(bool isCommit, bool isParallelWorker) { PendingRelDelete *pending; - int nrels = 0, - maxrels = 0; - SMgrRelation *srels = NULL; + int nlocators = 0, + maxlocators = 0; + RelFileLocator *locators = NULL; HASH_SEQ_STATUS scan; PendingRelSync *pendingsync; @@ -757,9 +799,6 @@ smgrDoPendingSyncs(bool isCommit, bool isParallelWorker) ForkNumber fork; BlockNumber nblocks[MAX_FORKNUM + 1]; BlockNumber total_blocks = 0; - SMgrRelation srel; - - srel = smgropen(pendingsync->rlocator, InvalidBackendId); /* * We emit newpage WAL records for smaller relations. @@ -773,9 +812,12 @@ smgrDoPendingSyncs(bool isCommit, bool isParallelWorker) { for (fork = 0; fork <= MAX_FORKNUM; fork++) { - if (smgrexists(srel, fork)) + SMgrFileHandle sfile; + + sfile = smgropen(pendingsync->rlocator, InvalidBackendId, fork); + if (smgrexists(sfile)) { - BlockNumber n = smgrnblocks(srel, fork); + BlockNumber n = smgrnblocks(sfile); /* we shouldn't come here for unlogged relations */ Assert(fork != INIT_FORKNUM); @@ -803,18 +845,19 @@ smgrDoPendingSyncs(bool isCommit, bool isParallelWorker) total_blocks * BLCKSZ / 1024 >= wal_skip_threshold) { /* allocate the initial array, or extend it, if needed */ - if (maxrels == 0) + if (maxlocators == 0) { - maxrels = 8; - srels = palloc(sizeof(SMgrRelation) * maxrels); + maxlocators = 8; + locators = palloc(sizeof(RelFileLocatorBackend) * maxlocators); } - else if (maxrels <= nrels) + else if (maxlocators <= nlocators) { - maxrels *= 2; - srels = repalloc(srels, sizeof(SMgrRelation) * maxrels); + maxlocators *= 2; + locators = repalloc(locators, sizeof(RelFileLocatorBackend) * maxlocators); } - srels[nrels++] = srel; + locators[nlocators] = pendingsync->rlocator; + nlocators++; } else { @@ -833,7 +876,7 @@ smgrDoPendingSyncs(bool isCommit, bool isParallelWorker) * page including any unused space. ReadBufferExtended() * counts some pgstat events; unfortunately, we discard them. */ - rel = CreateFakeRelcacheEntry(srel->smgr_rlocator.locator); + rel = CreateFakeRelcacheEntry(pendingsync->rlocator); log_newpage_range(rel, fork, 0, n, false); FreeFakeRelcacheEntry(rel); } @@ -842,11 +885,20 @@ smgrDoPendingSyncs(bool isCommit, bool isParallelWorker) pendingSyncHash = NULL; - if (nrels > 0) + FlushRelationsAllBuffers(locators, nlocators); + + for (int i = 0; i < nlocators; i++) { - smgrdosyncall(srels, nrels); - pfree(srels); + for (int fork = 0; fork <= MAX_FORKNUM; fork++) + { + SMgrFileHandle sfile = smgropen(locators[i], InvalidBackendId, fork); + + if (smgrexists(sfile)) + smgrimmedsync(sfile); + } } + if (locators != NULL) + pfree(locators); } /* @@ -966,22 +1018,22 @@ smgr_redo(XLogReaderState *record) if (info == XLOG_SMGR_CREATE) { xl_smgr_create *xlrec = (xl_smgr_create *) XLogRecGetData(record); - SMgrRelation reln; + SMgrFileHandle sfile; - reln = smgropen(xlrec->rlocator, InvalidBackendId); - smgrcreate(reln, xlrec->forkNum, true); + sfile = smgropen(xlrec->rlocator, InvalidBackendId, xlrec->forkNum); + smgrcreate(sfile, true); } else if (info == XLOG_SMGR_TRUNCATE) { xl_smgr_truncate *xlrec = (xl_smgr_truncate *) XLogRecGetData(record); - SMgrRelation reln; + SMgrFileHandle sfile; Relation rel; ForkNumber forks[MAX_FORKNUM]; BlockNumber blocks[MAX_FORKNUM]; int nforks = 0; bool need_fsm_vacuum = false; - reln = smgropen(xlrec->rlocator, InvalidBackendId); + sfile = smgropen(xlrec->rlocator, InvalidBackendId, MAIN_FORKNUM); /* * Forcibly create relation if it doesn't exist (which suggests that @@ -989,7 +1041,7 @@ smgr_redo(XLogReaderState *record) * XLogReadBufferForRedo, we prefer to recreate the rel and replay the * log as best we can until the drop is seen. */ - smgrcreate(reln, MAIN_FORKNUM, true); + smgrcreate(sfile, true); /* * Before we perform the truncation, update minimum recovery point to @@ -1022,8 +1074,10 @@ smgr_redo(XLogReaderState *record) /* Prepare for truncation of FSM and VM too */ rel = CreateFakeRelcacheEntry(xlrec->rlocator); + DropRelationBuffers(xlrec->rlocator, InvalidBackendId, forks, nforks, blocks); + if ((xlrec->flags & SMGR_TRUNCATE_FSM) != 0 && - smgrexists(reln, FSM_FORKNUM)) + smgrexists(smgropen(xlrec->rlocator, InvalidBackendId, FSM_FORKNUM))) { blocks[nforks] = FreeSpaceMapPrepareTruncateRel(rel, xlrec->blkno); if (BlockNumberIsValid(blocks[nforks])) @@ -1034,7 +1088,7 @@ smgr_redo(XLogReaderState *record) } } if ((xlrec->flags & SMGR_TRUNCATE_VM) != 0 && - smgrexists(reln, VISIBILITYMAP_FORKNUM)) + smgrexists(smgropen(xlrec->rlocator, InvalidBackendId, VISIBILITYMAP_FORKNUM))) { blocks[nforks] = visibilitymap_prepare_truncate(rel, xlrec->blkno); if (BlockNumberIsValid(blocks[nforks])) @@ -1046,7 +1100,10 @@ smgr_redo(XLogReaderState *record) /* Do the real work to truncate relation forks */ if (nforks > 0) - smgrtruncate(reln, forks, nforks, blocks); + { + DropRelationBuffers(xlrec->rlocator, InvalidBackendId, forks, nforks, blocks); + smgrtruncate_multi(xlrec->rlocator, InvalidBackendId, forks, nforks, blocks); + } /* * Update upper-level FSM pages to account for the truncation. This is diff --git a/src/backend/commands/async.c b/src/backend/commands/async.c index ef909cf4e08..523a378752c 100644 --- a/src/backend/commands/async.c +++ b/src/backend/commands/async.c @@ -141,6 +141,7 @@ #include "libpq/libpq.h" #include "libpq/pqformat.h" #include "miscadmin.h" +#include "storage/bufmgr.h" #include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/proc.h" @@ -163,7 +164,7 @@ * than that, so changes in that data structure won't affect user-visible * restrictions. */ -#define NOTIFY_PAYLOAD_MAX_LENGTH (BLCKSZ - NAMEDATALEN - 128) +#define NOTIFY_PAYLOAD_MAX_LENGTH (BLCKSZ - NAMEDATALEN - SizeOfPageHeaderData - 128) /* * Struct representing an entry in the global notify queue @@ -213,7 +214,7 @@ typedef struct QueuePosition ((x).page == (y).page && (x).offset == (y).offset) #define QUEUE_POS_IS_ZERO(x) \ - ((x).page == 0 && (x).offset == 0) + ((x).page == 0 && (x).offset == MAXALIGN(SizeOfPageHeaderData)) /* choose logically smaller QueuePosition */ #define QUEUE_POS_MIN(x,y) \ @@ -305,12 +306,6 @@ static AsyncQueueControl *asyncQueueControl; #define QUEUE_NEXT_LISTENER(i) (asyncQueueControl->backend[i].nextListener) #define QUEUE_BACKEND_POS(i) (asyncQueueControl->backend[i].pos) -/* - * The SLRU buffer area through which we access the notification queue - */ -static SlruCtlData NotifyCtlData; - -#define NotifyCtl (&NotifyCtlData) #define QUEUE_PAGESIZE BLCKSZ #define QUEUE_FULL_WARN_INTERVAL 5000 /* warn at most once every 5s */ @@ -521,8 +516,6 @@ AsyncShmemSize(void) size = mul_size(MaxBackends + 1, sizeof(QueueBackendStatus)); size = add_size(size, offsetof(AsyncQueueControl, backend)); - size = add_size(size, SimpleLruShmemSize(NUM_NOTIFY_BUFFERS, 0)); - return size; } @@ -550,8 +543,8 @@ AsyncShmemInit(void) if (!found) { /* First time through, so initialize it */ - SET_QUEUE_POS(QUEUE_HEAD, 0, 0); - SET_QUEUE_POS(QUEUE_TAIL, 0, 0); + SET_QUEUE_POS(QUEUE_HEAD, 0, MAXALIGN(SizeOfPageHeaderData)); + SET_QUEUE_POS(QUEUE_TAIL, 0, MAXALIGN(SizeOfPageHeaderData)); QUEUE_STOP_PAGE = 0; QUEUE_FIRST_LISTENER = InvalidBackendId; asyncQueueControl->lastQueueFillWarn = 0; @@ -561,24 +554,17 @@ AsyncShmemInit(void) QUEUE_BACKEND_PID(i) = InvalidPid; QUEUE_BACKEND_DBOID(i) = InvalidOid; QUEUE_NEXT_LISTENER(i) = InvalidBackendId; - SET_QUEUE_POS(QUEUE_BACKEND_POS(i), 0, 0); + SET_QUEUE_POS(QUEUE_BACKEND_POS(i), 0, MAXALIGN(SizeOfPageHeaderData)); } } - /* - * Set up SLRU management of the pg_notify data. - */ - NotifyCtl->PagePrecedes = asyncQueuePagePrecedes; - SimpleLruInit(NotifyCtl, "Notify", NUM_NOTIFY_BUFFERS, 0, - NotifySLRULock, "pg_notify", LWTRANCHE_NOTIFY_BUFFER, - SYNC_HANDLER_NONE); - if (!found) { /* * During start or reboot, clean out the pg_notify directory. */ - (void) SlruScanDirectory(NotifyCtl, SlruScanDirCbDeleteAll, NULL); + (void) SlruScanDirectory(SLRU_NOTIFY_ID, asyncQueuePagePrecedes, + SlruScanDirCbDeleteAll, NULL); } } @@ -1345,19 +1331,19 @@ asyncQueueAdvance(volatile QueuePosition *position, int entryLength) * written or read. */ offset += entryLength; - Assert(offset <= QUEUE_PAGESIZE); + Assert(offset <= QUEUE_PAGESIZE - MAXALIGN(SizeOfPageHeaderData)); /* * In a second step check if another entry can possibly be written to the * page. If so, stay here, we have reached the next position. If not, then * we need to move on to the next page. */ - if (offset + QUEUEALIGN(AsyncQueueEntryEmptySize) > QUEUE_PAGESIZE) + if (offset + QUEUEALIGN(AsyncQueueEntryEmptySize) > QUEUE_PAGESIZE - MAXALIGN(SizeOfPageHeaderData)) { pageno++; if (pageno > QUEUE_MAX_PAGE) pageno = 0; /* wrap around */ - offset = 0; + offset = MAXALIGN(SizeOfPageHeaderData); /* start at SizeOfPageHeaderData */ pageJump = true; } @@ -1411,10 +1397,7 @@ asyncQueueAddEntries(ListCell *nextNotify) QueuePosition queue_head; int pageno; int offset; - int slotno; - - /* We hold both NotifyQueueLock and NotifySLRULock during this operation */ - LWLockAcquire(NotifySLRULock, LW_EXCLUSIVE); + Buffer buffer; /* * We work with a local copy of QUEUE_HEAD, which we write back to shared @@ -1439,13 +1422,20 @@ asyncQueueAddEntries(ListCell *nextNotify) */ pageno = QUEUE_POS_PAGE(queue_head); if (QUEUE_POS_IS_ZERO(queue_head)) - slotno = SimpleLruZeroPage(NotifyCtl, pageno); + { + buffer = ZeroSlruBuffer(SLRU_NOTIFY_ID, pageno); + PageSetHeaderDataNonRel(BufferGetPage(buffer), pageno, InvalidXLogRecPtr, BLCKSZ, PG_METAPAGE_LAYOUT_VERSION); + + } else - slotno = SimpleLruReadPage(NotifyCtl, pageno, true, - InvalidTransactionId); + { + buffer = ReadSlruBuffer(SLRU_NOTIFY_ID, pageno, RBM_NORMAL); + LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE); + } + /* Note we mark the page dirty before writing in it */ - NotifyCtl->shared->page_dirty[slotno] = true; + MarkBufferDirty(buffer); while (nextNotify != NULL) { @@ -1457,7 +1447,7 @@ asyncQueueAddEntries(ListCell *nextNotify) offset = QUEUE_POS_OFFSET(queue_head); /* Check whether the entry really fits on the current page */ - if (offset + qe.length <= QUEUE_PAGESIZE) + if (offset + qe.length <= QUEUE_PAGESIZE - MAXALIGN(SizeOfPageHeaderData)) { /* OK, so advance nextNotify past this item */ nextNotify = lnext(pendingNotifies->events, nextNotify); @@ -1469,17 +1459,18 @@ asyncQueueAddEntries(ListCell *nextNotify) * only check dboid and since it won't match any reader's database * OID, they will ignore this entry and move on. */ - qe.length = QUEUE_PAGESIZE - offset; + qe.length = QUEUE_PAGESIZE - MAXALIGN(SizeOfPageHeaderData) - offset; qe.dboid = InvalidOid; qe.data[0] = '\0'; /* empty channel */ qe.data[1] = '\0'; /* empty payload */ } /* Now copy qe into the shared buffer page */ - memcpy(NotifyCtl->shared->page_buffer[slotno] + offset, + memcpy(PageGetContents(BufferGetPage(buffer)) + offset, &qe, qe.length); + /* Advance queue_head appropriately, and detect if page is full */ if (asyncQueueAdvance(&(queue_head), qe.length)) { @@ -1491,7 +1482,10 @@ asyncQueueAddEntries(ListCell *nextNotify) * asyncQueueIsFull() ensured that there is room to create this * page without overrunning the queue. */ - slotno = SimpleLruZeroPage(NotifyCtl, QUEUE_POS_PAGE(queue_head)); + UnlockReleaseBuffer(buffer); + buffer = ZeroSlruBuffer(SLRU_NOTIFY_ID, + QUEUE_POS_PAGE(queue_head)); + MarkBufferDirty(buffer); /* * If the new page address is a multiple of QUEUE_CLEANUP_DELAY, @@ -1505,12 +1499,11 @@ asyncQueueAddEntries(ListCell *nextNotify) break; } } + UnlockReleaseBuffer(buffer); /* Success, so update the global QUEUE_HEAD */ QUEUE_HEAD = queue_head; - LWLockRelease(NotifySLRULock); - return nextNotify; } @@ -1983,17 +1976,16 @@ asyncQueueReadAllNotifications(void) { int curpage = QUEUE_POS_PAGE(pos); int curoffset = QUEUE_POS_OFFSET(pos); - int slotno; int copysize; + Buffer buffer; /* - * We copy the data from SLRU into a local buffer, so as to avoid - * holding the NotifySLRULock while we are examining the entries - * and possibly transmitting them to our frontend. Copy only the - * part of the page we will actually inspect. + * We copy the data into a local buffer, so as to avoid holding a + * buffer pin while we are examining the entries and possibly + * transmitting them to our frontend. Copy only the part of the + * page we will actually inspect. */ - slotno = SimpleLruReadPage_ReadOnly(NotifyCtl, curpage, - InvalidTransactionId); + buffer = ReadSlruBuffer(SLRU_NOTIFY_ID, curpage, RBM_NORMAL); if (curpage == QUEUE_POS_PAGE(head)) { /* we only want to read as far as head */ @@ -2004,13 +1996,12 @@ asyncQueueReadAllNotifications(void) else { /* fetch all the rest of the page */ - copysize = QUEUE_PAGESIZE - curoffset; + copysize = QUEUE_PAGESIZE - MAXALIGN(SizeOfPageHeaderData) - curoffset; } - memcpy(page_buffer.buf + curoffset, - NotifyCtl->shared->page_buffer[slotno] + curoffset, + memcpy(PageGetContents(page_buffer.buf) + curoffset, + PageGetContents(BufferGetPage(buffer)) + curoffset, copysize); - /* Release lock that we got from SimpleLruReadPage_ReadOnly() */ - LWLockRelease(NotifySLRULock); + ReleaseBuffer(buffer); /* * Process messages up to the stop position, end of page, or an @@ -2078,7 +2069,7 @@ asyncQueueProcessPageEntries(volatile QueuePosition *current, if (QUEUE_POS_EQUAL(thisentry, stop)) break; - qe = (AsyncQueueEntry *) (page_buffer + QUEUE_POS_OFFSET(thisentry)); + qe = (AsyncQueueEntry *) (PageGetContents(page_buffer) + QUEUE_POS_OFFSET(thisentry)); /* * Advance *current over this message, possibly to the next page. As @@ -2207,7 +2198,7 @@ asyncQueueAdvanceTail(void) * SimpleLruTruncate() will ask for NotifySLRULock but will also * release the lock again. */ - SimpleLruTruncate(NotifyCtl, newtailpage); + SimpleLruTruncate(SLRU_NOTIFY_ID, asyncQueuePagePrecedes, newtailpage); /* * Update QUEUE_STOP_PAGE. This changes asyncQueueIsFull()'s verdict diff --git a/src/backend/commands/dbcommands.c b/src/backend/commands/dbcommands.c index 518ffca09a0..0bfe148e036 100644 --- a/src/backend/commands/dbcommands.c +++ b/src/backend/commands/dbcommands.c @@ -259,7 +259,7 @@ ScanSourceDatabasePgClass(Oid tbid, Oid dbid, char *srcpath) List *rlocatorlist = NIL; LockRelId relid; Snapshot snapshot; - SMgrRelation smgr; + SMgrFileHandle sfile; BufferAccessStrategy bstrategy; /* Get pg_class relfilenumber. */ @@ -276,9 +276,9 @@ ScanSourceDatabasePgClass(Oid tbid, Oid dbid, char *srcpath) rlocator.dbOid = dbid; rlocator.relNumber = relfilenumber; - smgr = smgropen(rlocator, InvalidBackendId); - nblocks = smgrnblocks(smgr, MAIN_FORKNUM); - smgrclose(smgr); + sfile = smgropen(rlocator, InvalidBackendId, MAIN_FORKNUM); + nblocks = smgrnblocks(sfile); + smgrclose(sfile); /* Use a buffer access strategy since this is a bulk read operation. */ bstrategy = GetAccessStrategy(BAS_BULKREAD); diff --git a/src/backend/commands/sequence.c b/src/backend/commands/sequence.c index bfe279cddf1..460e2bb1822 100644 --- a/src/backend/commands/sequence.c +++ b/src/backend/commands/sequence.c @@ -353,14 +353,14 @@ fill_seq_with_data(Relation rel, HeapTuple tuple) if (rel->rd_rel->relpersistence == RELPERSISTENCE_UNLOGGED) { - SMgrRelation srel; + SMgrFileHandle sfile; - srel = smgropen(rel->rd_locator, InvalidBackendId); - smgrcreate(srel, INIT_FORKNUM, false); + sfile = smgropen(rel->rd_locator, InvalidBackendId, INIT_FORKNUM); + smgrcreate(sfile, false); log_smgrcreate(&rel->rd_locator, INIT_FORKNUM); fill_seq_fork_with_data(rel, tuple, INIT_FORKNUM); FlushRelationBuffers(rel); - smgrclose(srel); + smgrclose(sfile); } } diff --git a/src/backend/commands/tablecmds.c b/src/backend/commands/tablecmds.c index 1db3bd9e2e8..7d53fea6a81 100644 --- a/src/backend/commands/tablecmds.c +++ b/src/backend/commands/tablecmds.c @@ -14634,9 +14634,7 @@ AlterTableMoveAll(AlterTableMoveAllStmt *stmt) static void index_copy_data(Relation rel, RelFileLocator newrlocator) { - SMgrRelation dstrel; - - dstrel = smgropen(newrlocator, rel->rd_backend); + SMgrFileHandle dstmain; /* * Since we copy the file directly without looking at the shared buffers, @@ -14656,16 +14654,20 @@ index_copy_data(Relation rel, RelFileLocator newrlocator) RelationCreateStorage(newrlocator, rel->rd_rel->relpersistence, true); /* copy main fork */ - RelationCopyStorage(RelationGetSmgr(rel), dstrel, MAIN_FORKNUM, + dstmain = smgropen(newrlocator, rel->rd_backend, MAIN_FORKNUM); + RelationCopyStorage(RelationGetSmgr(rel, MAIN_FORKNUM), dstmain, rel->rd_rel->relpersistence); /* copy those extra forks that exist */ for (ForkNumber forkNum = MAIN_FORKNUM + 1; forkNum <= MAX_FORKNUM; forkNum++) { - if (smgrexists(RelationGetSmgr(rel), forkNum)) + if (smgrexists(RelationGetSmgr(rel, forkNum))) { - smgrcreate(dstrel, forkNum, false); + SMgrFileHandle src_fork = RelationGetSmgr(rel, forkNum); + SMgrFileHandle dst_fork = smgropen(newrlocator, rel->rd_backend, forkNum); + + smgrcreate(dst_fork, false); /* * WAL log creation if the relation is persistent, or this is the @@ -14675,14 +14677,15 @@ index_copy_data(Relation rel, RelFileLocator newrlocator) (rel->rd_rel->relpersistence == RELPERSISTENCE_UNLOGGED && forkNum == INIT_FORKNUM)) log_smgrcreate(&newrlocator, forkNum); - RelationCopyStorage(RelationGetSmgr(rel), dstrel, forkNum, + RelationCopyStorage(src_fork, dst_fork, rel->rd_rel->relpersistence); + smgrclose(dst_fork); } } /* drop old relation, and close new one */ RelationDropStorage(rel); - smgrclose(dstrel); + smgrclose(dstmain); } /* diff --git a/src/backend/storage/buffer/buf_init.c b/src/backend/storage/buffer/buf_init.c index 20946c47cb4..9746838bf9f 100644 --- a/src/backend/storage/buffer/buf_init.c +++ b/src/backend/storage/buffer/buf_init.c @@ -20,6 +20,7 @@ BufferDescPadded *BufferDescriptors; char *BufferBlocks; +XLogRecPtr *BufferExternalLSNs; ConditionVariableMinimallyPadded *BufferIOCVArray; WritebackContext BackendWritebackContext; CkptSortItem *CkptBufferIds; @@ -69,9 +70,11 @@ InitBufferPool(void) { bool foundBufs, foundDescs, + foundLSNs, foundIOCV, foundBufCkpt; + /* Align descriptors to a cacheline boundary. */ BufferDescriptors = (BufferDescPadded *) ShmemInitStruct("Buffer Descriptors", @@ -88,6 +91,11 @@ InitBufferPool(void) NBuffers * sizeof(ConditionVariableMinimallyPadded), &foundIOCV); + BufferExternalLSNs = (XLogRecPtr *) + ShmemInitStruct("Buffer External LSNs", + NBuffers * sizeof(XLogRecPtr), + &foundLSNs); + /* * The array used to sort to-be-checkpointed buffer ids is located in * shared memory, to avoid having to allocate significant amounts of @@ -99,10 +107,10 @@ InitBufferPool(void) ShmemInitStruct("Checkpoint BufferIds", NBuffers * sizeof(CkptSortItem), &foundBufCkpt); - if (foundDescs || foundBufs || foundIOCV || foundBufCkpt) + if (foundDescs || foundBufs || foundIOCV || foundBufCkpt || foundLSNs) { /* should find all of these, or none of them */ - Assert(foundDescs && foundBufs && foundIOCV && foundBufCkpt); + Assert(foundDescs && foundBufs && foundIOCV && foundBufCkpt && foundLSNs); /* note: this path is only taken in EXEC_BACKEND case */ } else @@ -133,6 +141,8 @@ InitBufferPool(void) LWTRANCHE_BUFFER_CONTENT); ConditionVariableInit(BufferDescriptorGetIOCV(buf)); + + BufferExternalLSNs[i] = InvalidXLogRecPtr; } /* Correct last entry of linked list */ @@ -166,6 +176,9 @@ BufferShmemSize(void) /* size of data pages */ size = add_size(size, mul_size(NBuffers, BLCKSZ)); + /* size of external LSNs */ + size = add_size(size, mul_size(NBuffers, sizeof(XLogRecPtr))); + /* size of stuff controlled by freelist.c */ size = add_size(size, StrategyShmemSize()); diff --git a/src/backend/storage/buffer/bufmgr.c b/src/backend/storage/buffer/bufmgr.c index 3fb38a25cfa..5cefc4098ee 100644 --- a/src/backend/storage/buffer/bufmgr.c +++ b/src/backend/storage/buffer/bufmgr.c @@ -57,10 +57,17 @@ #include "utils/resowner_private.h" #include "utils/timestamp.h" +/* + * XXX Ideally we'd switch to standard pages for SLRU data, but in the + * meantime we need some way to identify buffers that hold raw data (no + * invasive LSN, no checksums). + */ /* Note: these two macros only work on shared buffers, not local ones! */ #define BufHdrGetBlock(bufHdr) ((Block) (BufferBlocks + ((Size) (bufHdr)->buf_id) * BLCKSZ)) -#define BufferGetLSN(bufHdr) (PageGetLSN(BufHdrGetBlock(bufHdr))) + +#define BufferGetLSN(bufHdr) \ + PageGetLSN(BufHdrGetBlock(bufHdr)) /* Note: this macro only works on local buffers, not shared ones! */ #define LocalBufHdrGetBlock(bufHdr) \ @@ -117,19 +124,6 @@ typedef struct CkptTsStatus int index; } CkptTsStatus; -/* - * Type for array used to sort SMgrRelations - * - * FlushRelationsAllBuffers shares the same comparator function with - * DropRelationsAllBuffers. Pointer to this struct and RelFileLocator must be - * compatible. - */ -typedef struct SMgrSortArray -{ - RelFileLocator rlocator; /* This must be the first member */ - SMgrRelation srel; -} SMgrSortArray; - /* GUC variables */ bool zero_damaged_pages = false; int bgwriter_lru_maxpages = 100; @@ -459,8 +453,8 @@ ForgetPrivateRefCountEntry(PrivateRefCountEntry *ref) ) -static Buffer ReadBuffer_common(SMgrRelation smgr, char relpersistence, - ForkNumber forkNum, BlockNumber blockNum, +static Buffer ReadBuffer_common(SMgrFileHandle sfile, char relpersistence, + BlockNumber blockNum, ReadBufferMode mode, BufferAccessStrategy strategy, bool *hit); static bool PinBuffer(BufferDesc *buf, BufferAccessStrategy strategy); @@ -476,20 +470,19 @@ static void TerminateBufferIO(BufferDesc *buf, bool clear_dirty, uint32 set_flag_bits); static void shared_buffer_write_error_callback(void *arg); static void local_buffer_write_error_callback(void *arg); -static BufferDesc *BufferAlloc(SMgrRelation smgr, +static BufferDesc *BufferAlloc(SMgrFileHandle smgr, char relpersistence, - ForkNumber forkNum, BlockNumber blockNum, BufferAccessStrategy strategy, bool *foundPtr); -static void FlushBuffer(BufferDesc *buf, SMgrRelation reln); +static void FlushBuffer(BufferDesc *buf, SMgrFileHandle sfile); static void FindAndDropRelationBuffers(RelFileLocator rlocator, - ForkNumber forkNum, + ForkNumber forknum, BlockNumber nForkBlock, BlockNumber firstDelBlock); static void RelationCopyStorageUsingBuffer(RelFileLocator srclocator, RelFileLocator dstlocator, - ForkNumber forkNum, bool permanent); + ForkNumber forknum, bool permanent); static void AtProcExit_Buffers(int code, Datum arg); static void CheckForBufferLeaks(void); static int rlocator_comparator(const void *p1, const void *p2); @@ -502,9 +495,7 @@ static int ts_ckpt_progress_comparator(Datum a, Datum b, void *arg); * Implementation of PrefetchBuffer() for shared buffers. */ PrefetchBufferResult -PrefetchSharedBuffer(SMgrRelation smgr_reln, - ForkNumber forkNum, - BlockNumber blockNum) +PrefetchSharedBuffer(SMgrFileHandle sfile, BlockNumber blockNum) { PrefetchBufferResult result = {InvalidBuffer, false}; BufferTag newTag; /* identity of requested block */ @@ -515,8 +506,8 @@ PrefetchSharedBuffer(SMgrRelation smgr_reln, Assert(BlockNumberIsValid(blockNum)); /* create a tag so we can lookup the buffer */ - InitBufferTag(&newTag, &smgr_reln->smgr_rlocator.locator, - forkNum, blockNum); + InitBufferTag(&newTag, &sfile->smgr_locator.locator, + sfile->smgr_locator.forknum, blockNum); /* determine its hash code and partition lock ID */ newHash = BufTableHashCode(&newTag); @@ -535,7 +526,7 @@ PrefetchSharedBuffer(SMgrRelation smgr_reln, * Try to initiate an asynchronous read. This returns false in * recovery if the relation file doesn't exist. */ - if (smgrprefetch(smgr_reln, forkNum, blockNum)) + if (smgrprefetch(sfile, blockNum)) result.initiated_io = true; #endif /* USE_PREFETCH */ } @@ -589,7 +580,7 @@ PrefetchSharedBuffer(SMgrRelation smgr_reln, * recovery, an error is raised). */ PrefetchBufferResult -PrefetchBuffer(Relation reln, ForkNumber forkNum, BlockNumber blockNum) +PrefetchBuffer(Relation reln, ForkNumber forknum, BlockNumber blockNum) { Assert(RelationIsValid(reln)); Assert(BlockNumberIsValid(blockNum)); @@ -603,12 +594,12 @@ PrefetchBuffer(Relation reln, ForkNumber forkNum, BlockNumber blockNum) errmsg("cannot access temporary tables of other sessions"))); /* pass it off to localbuf.c */ - return PrefetchLocalBuffer(RelationGetSmgr(reln), forkNum, blockNum); + return PrefetchLocalBuffer(RelationGetSmgr(reln, forknum), blockNum); } else { /* pass it to the shared buffer version */ - return PrefetchSharedBuffer(RelationGetSmgr(reln), forkNum, blockNum); + return PrefetchSharedBuffer(RelationGetSmgr(reln, forknum), blockNum); } } @@ -620,7 +611,7 @@ PrefetchBuffer(Relation reln, ForkNumber forkNum, BlockNumber blockNum) * tag. In that case, the buffer is pinned and the usage count is bumped. */ bool -ReadRecentBuffer(RelFileLocator rlocator, ForkNumber forkNum, BlockNumber blockNum, +ReadRecentBuffer(RelFileLocator rlocator, ForkNumber forknum, BlockNumber blockNum, Buffer recent_buffer) { BufferDesc *bufHdr; @@ -632,7 +623,7 @@ ReadRecentBuffer(RelFileLocator rlocator, ForkNumber forkNum, BlockNumber blockN ResourceOwnerEnlargeBuffers(CurrentResourceOwner); ReservePrivateRefCountEntry(); - InitBufferTag(&tag, &rlocator, forkNum, blockNum); + InitBufferTag(&tag, &rlocator, forknum, blockNum); if (BufferIsLocal(recent_buffer)) { @@ -756,7 +747,7 @@ ReadBuffer(Relation reln, BlockNumber blockNum) * See buffer/README for details. */ Buffer -ReadBufferExtended(Relation reln, ForkNumber forkNum, BlockNumber blockNum, +ReadBufferExtended(Relation reln, ForkNumber forknum, BlockNumber blockNum, ReadBufferMode mode, BufferAccessStrategy strategy) { bool hit; @@ -777,8 +768,9 @@ ReadBufferExtended(Relation reln, ForkNumber forkNum, BlockNumber blockNum, * miss. */ pgstat_count_buffer_read(reln); - buf = ReadBuffer_common(RelationGetSmgr(reln), reln->rd_rel->relpersistence, - forkNum, blockNum, mode, strategy, &hit); + buf = ReadBuffer_common(RelationGetSmgr(reln, forknum), + reln->rd_rel->relpersistence, + blockNum, mode, strategy, &hit); if (hit) pgstat_count_buffer_hit(reln); return buf; @@ -796,19 +788,31 @@ ReadBufferExtended(Relation reln, ForkNumber forkNum, BlockNumber blockNum, * BackendId). */ Buffer -ReadBufferWithoutRelcache(RelFileLocator rlocator, ForkNumber forkNum, +ReadBufferWithoutRelcache(RelFileLocator rlocator, ForkNumber forknum, BlockNumber blockNum, ReadBufferMode mode, BufferAccessStrategy strategy, bool permanent) { bool hit; - SMgrRelation smgr = smgropen(rlocator, InvalidBackendId); + SMgrFileHandle sfile = smgropen(rlocator, InvalidBackendId, forknum); - return ReadBuffer_common(smgr, permanent ? RELPERSISTENCE_PERMANENT : - RELPERSISTENCE_UNLOGGED, forkNum, blockNum, + return ReadBuffer_common(sfile, permanent ? RELPERSISTENCE_PERMANENT : + RELPERSISTENCE_UNLOGGED, blockNum, mode, strategy, &hit); } +Buffer +ReadBufferWithoutRelcacheWithHit(RelFileLocator rlocator, ForkNumber forknum, + BlockNumber blockNum, ReadBufferMode mode, + BufferAccessStrategy strategy, bool permanent, bool *hit) +{ + SMgrFileHandle sfile = smgropen(rlocator, InvalidBackendId, forknum); + + return ReadBuffer_common(sfile, permanent ? RELPERSISTENCE_PERMANENT : + RELPERSISTENCE_UNLOGGED, blockNum, + mode, strategy, hit); +} + /* * ReadBuffer_common -- common logic for all ReadBuffer variants @@ -816,7 +820,7 @@ ReadBufferWithoutRelcache(RelFileLocator rlocator, ForkNumber forkNum, * *hit is set to true if the request was satisfied from shared buffer cache. */ static Buffer -ReadBuffer_common(SMgrRelation smgr, char relpersistence, ForkNumber forkNum, +ReadBuffer_common(SMgrFileHandle sfile, char relpersistence, BlockNumber blockNum, ReadBufferMode mode, BufferAccessStrategy strategy, bool *hit) { @@ -824,7 +828,7 @@ ReadBuffer_common(SMgrRelation smgr, char relpersistence, ForkNumber forkNum, Block bufBlock; bool found; bool isExtend; - bool isLocalBuf = SmgrIsTemp(smgr); + bool isLocalBuf = SmgrIsTemp(sfile); *hit = false; @@ -833,29 +837,30 @@ ReadBuffer_common(SMgrRelation smgr, char relpersistence, ForkNumber forkNum, isExtend = (blockNum == P_NEW); - TRACE_POSTGRESQL_BUFFER_READ_START(forkNum, blockNum, - smgr->smgr_rlocator.locator.spcOid, - smgr->smgr_rlocator.locator.dbOid, - smgr->smgr_rlocator.locator.relNumber, - smgr->smgr_rlocator.backend, + TRACE_POSTGRESQL_BUFFER_READ_START(smgr->smgr_locator.forknum, + blockNum, + smgr->smgr_locator.locator.spcOid, + smgr->smgr_locator.locator.dbOid, + smgr->smgr_locator.locator.relNumber, + smgr->smgr_locator.backend, isExtend); /* Substitute proper block number if caller asked for P_NEW */ if (isExtend) { - blockNum = smgrnblocks(smgr, forkNum); + blockNum = smgrnblocks(sfile); /* Fail if relation is already at maximum possible length */ if (blockNum == P_NEW) ereport(ERROR, (errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED), errmsg("cannot extend relation %s beyond %u blocks", - relpath(smgr->smgr_rlocator, forkNum), + smgrfilepath(sfile->smgr_locator), P_NEW))); } if (isLocalBuf) { - bufHdr = LocalBufferAlloc(smgr, forkNum, blockNum, &found); + bufHdr = LocalBufferAlloc(sfile, blockNum, &found); if (found) pgBufferUsage.local_blks_hit++; else if (isExtend) @@ -870,7 +875,7 @@ ReadBuffer_common(SMgrRelation smgr, char relpersistence, ForkNumber forkNum, * lookup the buffer. IO_IN_PROGRESS is set if the requested block is * not currently in memory. */ - bufHdr = BufferAlloc(smgr, relpersistence, forkNum, blockNum, + bufHdr = BufferAlloc(sfile, relpersistence, blockNum, strategy, &found); if (found) pgBufferUsage.shared_blks_hit++; @@ -895,11 +900,12 @@ ReadBuffer_common(SMgrRelation smgr, char relpersistence, ForkNumber forkNum, if (VacuumCostActive) VacuumCostBalance += VacuumCostPageHit; - TRACE_POSTGRESQL_BUFFER_READ_DONE(forkNum, blockNum, - smgr->smgr_rlocator.locator.spcOid, - smgr->smgr_rlocator.locator.dbOid, - smgr->smgr_rlocator.locator.relNumber, - smgr->smgr_rlocator.backend, + TRACE_POSTGRESQL_BUFFER_READ_DONE(sfile->smgr_forknum, + blockNum, + sfile->smgr_rlocator.locator.spcOid, + sfile->smgr_rlocator.locator.dbOid, + sfile->smgr_rlocator.locator.relNumber, + sfile->smgr_rlocator.backend, isExtend, found); @@ -936,7 +942,7 @@ ReadBuffer_common(SMgrRelation smgr, char relpersistence, ForkNumber forkNum, if (!PageIsNew((Page) bufBlock)) ereport(ERROR, (errmsg("unexpected data beyond EOF in block %u of relation %s", - blockNum, relpath(smgr->smgr_rlocator, forkNum)), + blockNum, smgrfilepath(sfile->smgr_locator)), errhint("This has been seen to occur with buggy kernels; consider updating your system."))); /* @@ -993,7 +999,7 @@ ReadBuffer_common(SMgrRelation smgr, char relpersistence, ForkNumber forkNum, /* new buffers are zero-filled */ MemSet((char *) bufBlock, 0, BLCKSZ); /* don't set checksum for all-zero page */ - smgrextend(smgr, forkNum, blockNum, (char *) bufBlock, false); + smgrextend(sfile, blockNum, (char *) bufBlock, false); /* * NB: we're *not* doing a ScheduleBufferTagForWriteback here; @@ -1018,7 +1024,7 @@ ReadBuffer_common(SMgrRelation smgr, char relpersistence, ForkNumber forkNum, if (track_io_timing) INSTR_TIME_SET_CURRENT(io_start); - smgrread(smgr, forkNum, blockNum, (char *) bufBlock); + smgrread(sfile, blockNum, (char *) bufBlock); if (track_io_timing) { @@ -1029,8 +1035,8 @@ ReadBuffer_common(SMgrRelation smgr, char relpersistence, ForkNumber forkNum, } /* check for garbage data */ - if (!PageIsVerifiedExtended((Page) bufBlock, blockNum, - PIV_LOG_WARNING | PIV_REPORT_STAT)) + if ((!(mode == RBM_TRIM)) && (!PageIsVerifiedExtended((Page) bufBlock, blockNum, + PIV_LOG_WARNING | PIV_REPORT_STAT))) { if (mode == RBM_ZERO_ON_ERROR || zero_damaged_pages) { @@ -1038,7 +1044,7 @@ ReadBuffer_common(SMgrRelation smgr, char relpersistence, ForkNumber forkNum, (errcode(ERRCODE_DATA_CORRUPTED), errmsg("invalid page in block %u of relation %s; zeroing out page", blockNum, - relpath(smgr->smgr_rlocator, forkNum)))); + smgrfilepath(sfile->smgr_locator)))); MemSet((char *) bufBlock, 0, BLCKSZ); } else @@ -1046,7 +1052,7 @@ ReadBuffer_common(SMgrRelation smgr, char relpersistence, ForkNumber forkNum, (errcode(ERRCODE_DATA_CORRUPTED), errmsg("invalid page in block %u of relation %s", blockNum, - relpath(smgr->smgr_rlocator, forkNum)))); + smgrfilepath(sfile->smgr_locator)))); } } } @@ -1085,11 +1091,12 @@ ReadBuffer_common(SMgrRelation smgr, char relpersistence, ForkNumber forkNum, if (VacuumCostActive) VacuumCostBalance += VacuumCostPageMiss; - TRACE_POSTGRESQL_BUFFER_READ_DONE(forkNum, blockNum, - smgr->smgr_rlocator.locator.spcOid, - smgr->smgr_rlocator.locator.dbOid, - smgr->smgr_rlocator.locator.relNumber, - smgr->smgr_rlocator.backend, + TRACE_POSTGRESQL_BUFFER_READ_DONE(sfile->smgr_locator.forknum, + blockNum, + sfile->smgr_locator.locator.spcOid, + sfile->smgr_locator.locator.dbOid, + sfile->smgr_locator.locator.relNumber, + sfile->smgr_locator.backend, isExtend, found); @@ -1116,7 +1123,7 @@ ReadBuffer_common(SMgrRelation smgr, char relpersistence, ForkNumber forkNum, * No locks are held either at entry or exit. */ static BufferDesc * -BufferAlloc(SMgrRelation smgr, char relpersistence, ForkNumber forkNum, +BufferAlloc(SMgrFileHandle sfile, char relpersistence, BlockNumber blockNum, BufferAccessStrategy strategy, bool *foundPtr) @@ -1134,7 +1141,7 @@ BufferAlloc(SMgrRelation smgr, char relpersistence, ForkNumber forkNum, uint32 buf_state; /* create a tag so we can lookup the buffer */ - InitBufferTag(&newTag, &smgr->smgr_rlocator.locator, forkNum, blockNum); + InitBufferTag(&newTag, &sfile->smgr_locator.locator, sfile->smgr_locator.forknum, blockNum); /* determine its hash code and partition lock ID */ newHash = BufTableHashCode(&newTag); @@ -1264,10 +1271,11 @@ BufferAlloc(SMgrRelation smgr, char relpersistence, ForkNumber forkNum, } /* OK, do the I/O */ - TRACE_POSTGRESQL_BUFFER_WRITE_DIRTY_START(forkNum, blockNum, - smgr->smgr_rlocator.locator.spcOid, - smgr->smgr_rlocator.locator.dbOid, - smgr->smgr_rlocator.locator.relNumber); + TRACE_POSTGRESQL_BUFFER_WRITE_DIRTY_START(sfile->smgr_locator.forknum, + blockNum, + sfile->smgr_locator.locator.spcOid, + sfile->smgr_locator.locator.dbOid, + sfile->smgr_locator.locator.relNumber); FlushBuffer(buf, NULL); LWLockRelease(BufferDescriptorGetContentLock(buf)); @@ -1275,10 +1283,11 @@ BufferAlloc(SMgrRelation smgr, char relpersistence, ForkNumber forkNum, ScheduleBufferTagForWriteback(&BackendWritebackContext, &buf->tag); - TRACE_POSTGRESQL_BUFFER_WRITE_DIRTY_DONE(forkNum, blockNum, - smgr->smgr_rlocator.locator.spcOid, - smgr->smgr_rlocator.locator.dbOid, - smgr->smgr_rlocator.locator.relNumber); + TRACE_POSTGRESQL_BUFFER_WRITE_DIRTY_DONE(sfile->smgr_locator.forknum, + blockNum, + sfile->smgr_locator.locator.spcOid, + sfile->smgr_locator.locator.dbOid, + sfile->smgr_locator.locator.relNumber); } else { @@ -1434,7 +1443,7 @@ BufferAlloc(SMgrRelation smgr, char relpersistence, ForkNumber forkNum, buf_state &= ~(BM_VALID | BM_DIRTY | BM_JUST_DIRTIED | BM_CHECKPOINT_NEEDED | BM_IO_ERROR | BM_PERMANENT | BUF_USAGECOUNT_MASK); - if (relpersistence == RELPERSISTENCE_PERMANENT || forkNum == INIT_FORKNUM) + if (relpersistence == RELPERSISTENCE_PERMANENT || sfile->smgr_locator.forknum == INIT_FORKNUM) buf_state |= BM_TAG_VALID | BM_PERMANENT | BUF_USAGECOUNT_ONE; else buf_state |= BM_TAG_VALID | BUF_USAGECOUNT_ONE; @@ -1647,7 +1656,7 @@ ReleaseAndReadBuffer(Buffer buffer, Relation relation, BlockNumber blockNum) { - ForkNumber forkNum = MAIN_FORKNUM; + ForkNumber forknum = MAIN_FORKNUM; BufferDesc *bufHdr; if (BufferIsValid(buffer)) @@ -1658,7 +1667,7 @@ ReleaseAndReadBuffer(Buffer buffer, bufHdr = GetLocalBufferDescriptor(-buffer - 1); if (bufHdr->tag.blockNum == blockNum && BufTagMatchesRelFileLocator(&bufHdr->tag, &relation->rd_locator) && - BufTagGetForkNum(&bufHdr->tag) == forkNum) + BufTagGetForkNum(&bufHdr->tag) == forknum) return buffer; ResourceOwnerForgetBuffer(CurrentResourceOwner, buffer); LocalRefCount[-buffer - 1]--; @@ -1669,7 +1678,7 @@ ReleaseAndReadBuffer(Buffer buffer, /* we have pin, so it's ok to examine tag without spinlock */ if (bufHdr->tag.blockNum == blockNum && BufTagMatchesRelFileLocator(&bufHdr->tag, &relation->rd_locator) && - BufTagGetForkNum(&bufHdr->tag) == forkNum) + BufTagGetForkNum(&bufHdr->tag) == forknum) return buffer; UnpinBuffer(bufHdr); } @@ -2820,7 +2829,7 @@ BufferGetTag(Buffer buffer, RelFileLocator *rlocator, ForkNumber *forknum, * as the second parameter. If not, pass NULL. */ static void -FlushBuffer(BufferDesc *buf, SMgrRelation reln) +FlushBuffer(BufferDesc *buf, SMgrFileHandle sfile) { XLogRecPtr recptr; ErrorContextCallback errcallback; @@ -2845,14 +2854,14 @@ FlushBuffer(BufferDesc *buf, SMgrRelation reln) error_context_stack = &errcallback; /* Find smgr relation for buffer */ - if (reln == NULL) - reln = smgropen(BufTagGetRelFileLocator(&buf->tag), InvalidBackendId); + if (sfile == NULL) + sfile = smgropen(BufTagGetRelFileLocator(&buf->tag), InvalidBackendId, BufTagGetForkNum(&buf->tag)); - TRACE_POSTGRESQL_BUFFER_FLUSH_START(BufTagGetForkNum(&buf->tag), + TRACE_POSTGRESQL_BUFFER_FLUSH_START(sfile->smgr_locator.forknum, buf->tag.blockNum, - reln->smgr_rlocator.locator.spcOid, - reln->smgr_rlocator.locator.dbOid, - reln->smgr_rlocator.locator.relNumber); + sfile->smgr_locator.locator.spcOid, + sfile->smgr_locator.locator.dbOid, + sfile->smgr_locator.locator.relNumber); buf_state = LockBufHdr(buf); @@ -2906,8 +2915,7 @@ FlushBuffer(BufferDesc *buf, SMgrRelation reln) /* * bufToWrite is either the shared buffer or a copy, as appropriate. */ - smgrwrite(reln, - BufTagGetForkNum(&buf->tag), + smgrwrite(sfile, buf->tag.blockNum, bufToWrite, false); @@ -2928,11 +2936,11 @@ FlushBuffer(BufferDesc *buf, SMgrRelation reln) */ TerminateBufferIO(buf, true, 0); - TRACE_POSTGRESQL_BUFFER_FLUSH_DONE(BufTagGetForkNum(&buf->tag), + TRACE_POSTGRESQL_BUFFER_FLUSH_DONE(sfile->smgr_locator.forknum, buf->tag.blockNum, - reln->smgr_rlocator.locator.spcOid, - reln->smgr_rlocator.locator.dbOid, - reln->smgr_rlocator.locator.relNumber); + sfile->smgr_locator.locator.spcOid, + sfile->smgr_locator.locator.dbOid, + sfile->smgr_locator.locator.relNumber); /* Pop the error context stack */ error_context_stack = errcallback.previous; @@ -2947,7 +2955,7 @@ FlushBuffer(BufferDesc *buf, SMgrRelation reln) * it might not be. */ BlockNumber -RelationGetNumberOfBlocksInFork(Relation relation, ForkNumber forkNum) +RelationGetNumberOfBlocksInFork(Relation relation, ForkNumber forknum) { if (RELKIND_HAS_TABLE_AM(relation->rd_rel->relkind)) { @@ -2959,13 +2967,13 @@ RelationGetNumberOfBlocksInFork(Relation relation, ForkNumber forkNum) */ uint64 szbytes; - szbytes = table_relation_size(relation, forkNum); + szbytes = table_relation_size(relation, forknum); return (szbytes + (BLCKSZ - 1)) / BLCKSZ; } else if (RELKIND_HAS_STORAGE(relation->rd_rel->relkind)) { - return smgrnblocks(RelationGetSmgr(relation), forkNum); + return smgrnblocks(RelationGetSmgr(relation, forknum)); } else Assert(false); @@ -3028,6 +3036,7 @@ BufferGetLSNAtomic(Buffer buffer) buf_state = LockBufHdr(bufHdr); lsn = PageGetLSN(page); + UnlockBufHdr(bufHdr, buf_state); return lsn; @@ -3055,26 +3064,20 @@ BufferGetLSNAtomic(Buffer buffer) * -------------------------------------------------------------------- */ void -DropRelationBuffers(SMgrRelation smgr_reln, ForkNumber *forkNum, +DropRelationBuffers(RelFileLocator rlocator, BackendId backend, ForkNumber *forknum, int nforks, BlockNumber *firstDelBlock) { int i; int j; - RelFileLocatorBackend rlocator; BlockNumber nForkBlock[MAX_FORKNUM]; uint64 nBlocksToInvalidate = 0; - rlocator = smgr_reln->smgr_rlocator; - /* If it's a local relation, it's localbuf.c's problem. */ - if (RelFileLocatorBackendIsTemp(rlocator)) + if (backend == MyBackendId) { - if (rlocator.backend == MyBackendId) - { - for (j = 0; j < nforks; j++) - DropRelationLocalBuffers(rlocator.locator, forkNum[j], - firstDelBlock[j]); - } + for (j = 0; j < nforks; j++) + DropRelationLocalBuffers(rlocator, forknum[j], + firstDelBlock[j]); return; } @@ -3103,7 +3106,10 @@ DropRelationBuffers(SMgrRelation smgr_reln, ForkNumber *forkNum, for (i = 0; i < nforks; i++) { /* Get the number of blocks for a relation's fork */ - nForkBlock[i] = smgrnblocks_cached(smgr_reln, forkNum[i]); + SMgrFileHandle sfile; + + sfile = smgropen(rlocator, backend, forknum[i]); + nForkBlock[i] = smgrnblocks_cached(sfile); if (nForkBlock[i] == InvalidBlockNumber) { @@ -3123,7 +3129,7 @@ DropRelationBuffers(SMgrRelation smgr_reln, ForkNumber *forkNum, nBlocksToInvalidate < BUF_DROP_FULL_SCAN_THRESHOLD) { for (j = 0; j < nforks; j++) - FindAndDropRelationBuffers(rlocator.locator, forkNum[j], + FindAndDropRelationBuffers(rlocator, forknum[j], nForkBlock[j], firstDelBlock[j]); return; } @@ -3146,18 +3152,18 @@ DropRelationBuffers(SMgrRelation smgr_reln, ForkNumber *forkNum, * false positives are safe because we'll recheck after getting the * buffer lock. * - * We could check forkNum and blockNum as well as the rlocator, but + * We could check forknum and blockNum as well as the rlocator, but * the incremental win from doing so seems small. */ - if (!BufTagMatchesRelFileLocator(&bufHdr->tag, &rlocator.locator)) + if (!BufTagMatchesRelFileLocator(&bufHdr->tag, &rlocator)) continue; buf_state = LockBufHdr(bufHdr); for (j = 0; j < nforks; j++) { - if (BufTagMatchesRelFileLocator(&bufHdr->tag, &rlocator.locator) && - BufTagGetForkNum(&bufHdr->tag) == forkNum[j] && + if (BufTagMatchesRelFileLocator(&bufHdr->tag, &rlocator) && + BufTagGetForkNum(&bufHdr->tag) == forknum[j] && bufHdr->tag.blockNum >= firstDelBlock[j]) { InvalidateBuffer(bufHdr); /* releases spinlock */ @@ -3178,11 +3184,10 @@ DropRelationBuffers(SMgrRelation smgr_reln, ForkNumber *forkNum, * -------------------------------------------------------------------- */ void -DropRelationsAllBuffers(SMgrRelation *smgr_reln, int nlocators) +DropRelationsAllBuffers(RelFileLocatorBackend *rlocators, int nlocators) { int i; int n = 0; - SMgrRelation *rels; BlockNumber (*block)[MAX_FORKNUM + 1]; uint64 nBlocksToInvalidate = 0; RelFileLocator *locators; @@ -3192,18 +3197,18 @@ DropRelationsAllBuffers(SMgrRelation *smgr_reln, int nlocators) if (nlocators == 0) return; - rels = palloc(sizeof(SMgrRelation) * nlocators); /* non-local relations */ + locators = palloc(sizeof(RelFileLocator) * nlocators); /* non-local relations */ /* If it's a local relation, it's localbuf.c's problem. */ for (i = 0; i < nlocators; i++) { - if (RelFileLocatorBackendIsTemp(smgr_reln[i]->smgr_rlocator)) + if (rlocators[i].backend != InvalidBackendId) { - if (smgr_reln[i]->smgr_rlocator.backend == MyBackendId) - DropRelationAllLocalBuffers(smgr_reln[i]->smgr_rlocator.locator); + if (rlocators[i].backend == MyBackendId) + DropRelationAllLocalBuffers(rlocators[i].locator); } else - rels[n++] = smgr_reln[i]; + locators[n++] = rlocators[i].locator; } /* @@ -3212,7 +3217,7 @@ DropRelationsAllBuffers(SMgrRelation *smgr_reln, int nlocators) */ if (n == 0) { - pfree(rels); + pfree(locators); return; } @@ -3232,12 +3237,13 @@ DropRelationsAllBuffers(SMgrRelation *smgr_reln, int nlocators) for (int j = 0; j <= MAX_FORKNUM; j++) { /* Get the number of blocks for a relation's fork. */ - block[i][j] = smgrnblocks_cached(rels[i], j); + SMgrFileHandle sfile = smgropen(locators[i], InvalidBackendId, j); + block[i][j] = smgrnblocks_cached(sfile); /* We need to only consider the relation forks that exists. */ if (block[i][j] == InvalidBlockNumber) { - if (!smgrexists(rels[i], j)) + if (!smgrexists(sfile)) continue; cached = false; break; @@ -3263,20 +3269,17 @@ DropRelationsAllBuffers(SMgrRelation *smgr_reln, int nlocators) continue; /* drop all the buffers for a particular relation fork */ - FindAndDropRelationBuffers(rels[i]->smgr_rlocator.locator, + FindAndDropRelationBuffers(locators[i], j, block[i][j], 0); } } pfree(block); - pfree(rels); + pfree(locators); return; } pfree(block); - locators = palloc(sizeof(RelFileLocator) * n); /* non-local relations */ - for (i = 0; i < n; i++) - locators[i] = rels[i]->smgr_rlocator.locator; /* * For low number of relations to drop just use a simple walk through, to @@ -3336,7 +3339,6 @@ DropRelationsAllBuffers(SMgrRelation *smgr_reln, int nlocators) } pfree(locators); - pfree(rels); } /* --------------------------------------------------------------------- @@ -3349,7 +3351,7 @@ DropRelationsAllBuffers(SMgrRelation *smgr_reln, int nlocators) * -------------------------------------------------------------------- */ static void -FindAndDropRelationBuffers(RelFileLocator rlocator, ForkNumber forkNum, +FindAndDropRelationBuffers(RelFileLocator rlocator, ForkNumber forknum, BlockNumber nForkBlock, BlockNumber firstDelBlock) { @@ -3365,7 +3367,7 @@ FindAndDropRelationBuffers(RelFileLocator rlocator, ForkNumber forkNum, uint32 buf_state; /* create a tag so we can lookup the buffer */ - InitBufferTag(&bufTag, &rlocator, forkNum, curBlock); + InitBufferTag(&bufTag, &rlocator, forknum, curBlock); /* determine its hash code and partition lock ID */ bufHash = BufTableHashCode(&bufTag); @@ -3390,7 +3392,7 @@ FindAndDropRelationBuffers(RelFileLocator rlocator, ForkNumber forkNum, buf_state = LockBufHdr(bufHdr); if (BufTagMatchesRelFileLocator(&bufHdr->tag, &rlocator) && - BufTagGetForkNum(&bufHdr->tag) == forkNum && + BufTagGetForkNum(&bufHdr->tag) == forknum && bufHdr->tag.blockNum >= firstDelBlock) InvalidateBuffer(bufHdr); /* releases spinlock */ else @@ -3545,8 +3547,7 @@ FlushRelationBuffers(Relation rel) PageSetChecksumInplace(localpage, bufHdr->tag.blockNum); - smgrwrite(RelationGetSmgr(rel), - BufTagGetForkNum(&bufHdr->tag), + smgrwrite(RelationGetSmgr(rel, BufTagGetForkNum(&bufHdr->tag)), bufHdr->tag.blockNum, localpage, false); @@ -3586,7 +3587,7 @@ FlushRelationBuffers(Relation rel) { PinBuffer_Locked(bufHdr); LWLockAcquire(BufferDescriptorGetContentLock(bufHdr), LW_SHARED); - FlushBuffer(bufHdr, RelationGetSmgr(rel)); + FlushBuffer(bufHdr, RelationGetSmgr(rel, bufHdr->tag.forkNum)); LWLockRelease(BufferDescriptorGetContentLock(bufHdr)); UnpinBuffer(bufHdr); } @@ -3605,25 +3606,20 @@ FlushRelationBuffers(Relation rel) * -------------------------------------------------------------------- */ void -FlushRelationsAllBuffers(SMgrRelation *smgrs, int nrels) +FlushRelationsAllBuffers(RelFileLocator *rels, int nrels) { int i; - SMgrSortArray *srels; + RelFileLocator *locators = NULL; bool use_bsearch; if (nrels == 0) return; /* fill-in array for qsort */ - srels = palloc(sizeof(SMgrSortArray) * nrels); + locators = palloc(sizeof(RelFileLocator) * nrels); for (i = 0; i < nrels; i++) - { - Assert(!RelFileLocatorBackendIsTemp(smgrs[i]->smgr_rlocator)); - - srels[i].rlocator = smgrs[i]->smgr_rlocator.locator; - srels[i].srel = smgrs[i]; - } + locators[i] = rels[i]; /* * Save the bsearch overhead for low number of relations to sync. See @@ -3631,16 +3627,16 @@ FlushRelationsAllBuffers(SMgrRelation *smgrs, int nrels) */ use_bsearch = nrels > RELS_BSEARCH_THRESHOLD; - /* sort the list of SMgrRelations if necessary */ + /* sort the list of locators if necessary */ if (use_bsearch) - pg_qsort(srels, nrels, sizeof(SMgrSortArray), rlocator_comparator); + pg_qsort(locators, nrels, sizeof(RelFileLocator), rlocator_comparator); /* Make sure we can handle the pin inside the loop */ ResourceOwnerEnlargeBuffers(CurrentResourceOwner); for (i = 0; i < NBuffers; i++) { - SMgrSortArray *srelent = NULL; + RelFileLocator *found = NULL; BufferDesc *bufHdr = GetBufferDescriptor(i); uint32 buf_state; @@ -3655,9 +3651,9 @@ FlushRelationsAllBuffers(SMgrRelation *smgrs, int nrels) for (j = 0; j < nrels; j++) { - if (BufTagMatchesRelFileLocator(&bufHdr->tag, &srels[j].rlocator)) + if (BufTagMatchesRelFileLocator(&bufHdr->tag, &locators[j])) { - srelent = &srels[j]; + found = &locators[j]; break; } } @@ -3667,24 +3663,26 @@ FlushRelationsAllBuffers(SMgrRelation *smgrs, int nrels) RelFileLocator rlocator; rlocator = BufTagGetRelFileLocator(&bufHdr->tag); - srelent = bsearch((const void *) &(rlocator), - srels, nrels, sizeof(SMgrSortArray), + found = bsearch((const void *) &(rlocator), + locators, nrels, sizeof(RelFileLocator), rlocator_comparator); } /* buffer doesn't belong to any of the given relfilelocators; skip it */ - if (srelent == NULL) + if (found == NULL) continue; + /* FIXME: cache SMgrFileHandles for the rels, and pass to FlushBuffer */ + ReservePrivateRefCountEntry(); buf_state = LockBufHdr(bufHdr); - if (BufTagMatchesRelFileLocator(&bufHdr->tag, &srelent->rlocator) && + if (BufTagMatchesRelFileLocator(&bufHdr->tag, found) && (buf_state & (BM_VALID | BM_DIRTY)) == (BM_VALID | BM_DIRTY)) { PinBuffer_Locked(bufHdr); LWLockAcquire(BufferDescriptorGetContentLock(bufHdr), LW_SHARED); - FlushBuffer(bufHdr, srelent->srel); + FlushBuffer(bufHdr, NULL); LWLockRelease(BufferDescriptorGetContentLock(bufHdr)); UnpinBuffer(bufHdr); } @@ -3692,7 +3690,7 @@ FlushRelationsAllBuffers(SMgrRelation *smgrs, int nrels) UnlockBufHdr(bufHdr, buf_state); } - pfree(srels); + pfree(locators); } /* --------------------------------------------------------------------- @@ -3708,7 +3706,7 @@ FlushRelationsAllBuffers(SMgrRelation *smgrs, int nrels) static void RelationCopyStorageUsingBuffer(RelFileLocator srclocator, RelFileLocator dstlocator, - ForkNumber forkNum, bool permanent) + ForkNumber forknum, bool permanent) { Buffer srcBuf; Buffer dstBuf; @@ -3726,11 +3724,11 @@ RelationCopyStorageUsingBuffer(RelFileLocator srclocator, * can skip it when copying any fork of an unlogged relation other than * the init fork. */ - use_wal = XLogIsNeeded() && (permanent || forkNum == INIT_FORKNUM); + use_wal = XLogIsNeeded() && (permanent || forknum == INIT_FORKNUM); /* Get number of blocks in the source relation. */ - nblocks = smgrnblocks(smgropen(srclocator, InvalidBackendId), - forkNum); + nblocks = smgrnblocks(smgropen(srclocator, InvalidBackendId, forknum)); + /* Nothing to copy; just return. */ if (nblocks == 0) @@ -3741,7 +3739,7 @@ RelationCopyStorageUsingBuffer(RelFileLocator srclocator, * relation before starting to copy block by block. */ memset(buf.data, 0, BLCKSZ); - smgrextend(smgropen(dstlocator, InvalidBackendId), forkNum, nblocks - 1, + smgrextend(smgropen(dstlocator, InvalidBackendId, forknum), nblocks - 1, buf.data, true); /* This is a bulk operation, so use buffer access strategies. */ @@ -3754,14 +3752,14 @@ RelationCopyStorageUsingBuffer(RelFileLocator srclocator, CHECK_FOR_INTERRUPTS(); /* Read block from source relation. */ - srcBuf = ReadBufferWithoutRelcache(srclocator, forkNum, blkno, + srcBuf = ReadBufferWithoutRelcache(srclocator, forknum, blkno, RBM_NORMAL, bstrategy_src, permanent); LockBuffer(srcBuf, BUFFER_LOCK_SHARE); srcPage = BufferGetPage(srcBuf); /* Use P_NEW to extend the destination relation. */ - dstBuf = ReadBufferWithoutRelcache(dstlocator, forkNum, blkno, + dstBuf = ReadBufferWithoutRelcache(dstlocator, forknum, blkno, RBM_NORMAL, bstrategy_dst, permanent); LockBuffer(dstBuf, BUFFER_LOCK_EXCLUSIVE); @@ -3799,7 +3797,6 @@ void CreateAndCopyRelationData(RelFileLocator src_rlocator, RelFileLocator dst_rlocator, bool permanent) { - RelFileLocatorBackend rlocator; char relpersistence; /* Set the relpersistence. */ @@ -3819,34 +3816,30 @@ CreateAndCopyRelationData(RelFileLocator src_rlocator, permanent); /* copy those extra forks that exist */ - for (ForkNumber forkNum = MAIN_FORKNUM + 1; - forkNum <= MAX_FORKNUM; forkNum++) + for (ForkNumber forknum = MAIN_FORKNUM + 1; + forknum <= MAX_FORKNUM; forknum++) { - if (smgrexists(smgropen(src_rlocator, InvalidBackendId), forkNum)) + if (smgrexists(smgropen(src_rlocator, InvalidBackendId, forknum))) { - smgrcreate(smgropen(dst_rlocator, InvalidBackendId), forkNum, false); + smgrcreate(smgropen(dst_rlocator, InvalidBackendId, forknum), false); /* * WAL log creation if the relation is persistent, or this is the * init fork of an unlogged relation. */ - if (permanent || forkNum == INIT_FORKNUM) - log_smgrcreate(&dst_rlocator, forkNum); + if (permanent || forknum == INIT_FORKNUM) + log_smgrcreate(&dst_rlocator, forknum); /* Copy a fork's data, block by block. */ - RelationCopyStorageUsingBuffer(src_rlocator, dst_rlocator, forkNum, + RelationCopyStorageUsingBuffer(src_rlocator, dst_rlocator, forknum, permanent); } } - /* close source and destination smgr if exists. */ - rlocator.backend = InvalidBackendId; - rlocator.locator = src_rlocator; - smgrcloserellocator(rlocator); + smgrcloserellocator(src_rlocator, InvalidBackendId); - rlocator.locator = dst_rlocator; - smgrcloserellocator(rlocator); + smgrcloserellocator(dst_rlocator, InvalidBackendId); } /* --------------------------------------------------------------------- @@ -4967,7 +4960,7 @@ IssuePendingWritebacks(WritebackContext *context) { PendingWriteback *cur; PendingWriteback *next; - SMgrRelation reln; + SMgrFileHandle sfile; int ahead; BufferTag tag; RelFileLocator currlocator; @@ -5007,8 +5000,8 @@ IssuePendingWritebacks(WritebackContext *context) i += ahead; /* and finally tell the kernel to write the data to storage */ - reln = smgropen(currlocator, InvalidBackendId); - smgrwriteback(reln, BufTagGetForkNum(&tag), tag.blockNum, nblocks); + sfile = smgropen(currlocator, InvalidBackendId, BufTagGetForkNum(&tag)); + smgrwriteback(sfile, tag.blockNum, nblocks); } context->nr_pending = 0; @@ -5030,3 +5023,29 @@ TestForOldSnapshot_impl(Snapshot snapshot, Relation relation) (errcode(ERRCODE_SNAPSHOT_TOO_OLD), errmsg("snapshot too old"))); } + +/* + * Check if a buffer tag is currently mapped. + * + * XXX Dubious semantics; needed only for multixact's handling for + * inconsistent states. + */ +bool +BufferProbe(RelFileLocator rlocator, ForkNumber forknum, BlockNumber blockNum) +{ + BufferTag tag; + uint32 hash; + LWLock *partitionLock; + int buf_id; + + InitBufferTag(&tag, &rlocator, forknum, blockNum); + + hash = BufTableHashCode(&tag); + partitionLock = BufMappingPartitionLock(hash); + + LWLockAcquire(partitionLock, LW_SHARED); + buf_id = BufTableLookup(&tag, hash); + LWLockRelease(partitionLock); + + return buf_id >= 0; +} diff --git a/src/backend/storage/buffer/localbuf.c b/src/backend/storage/buffer/localbuf.c index b2720df6eaa..d268458c4b5 100644 --- a/src/backend/storage/buffer/localbuf.c +++ b/src/backend/storage/buffer/localbuf.c @@ -61,14 +61,13 @@ static Block GetLocalBufferStorage(void); * No-op if prefetching isn't compiled in. */ PrefetchBufferResult -PrefetchLocalBuffer(SMgrRelation smgr, ForkNumber forkNum, - BlockNumber blockNum) +PrefetchLocalBuffer(SMgrFileHandle sfile, BlockNumber blockNum) { PrefetchBufferResult result = {InvalidBuffer, false}; BufferTag newTag; /* identity of requested block */ LocalBufferLookupEnt *hresult; - InitBufferTag(&newTag, &smgr->smgr_rlocator.locator, forkNum, blockNum); + InitBufferTag(&newTag, &sfile->smgr_locator.locator, sfile->smgr_locator.forknum, blockNum); /* Initialize local buffers if first request in this session */ if (LocalBufHash == NULL) @@ -87,7 +86,7 @@ PrefetchLocalBuffer(SMgrRelation smgr, ForkNumber forkNum, { #ifdef USE_PREFETCH /* Not in buffers, so initiate prefetch */ - smgrprefetch(smgr, forkNum, blockNum); + smgrprefetch(sfile, blockNum); result.initiated_io = true; #endif /* USE_PREFETCH */ } @@ -106,8 +105,7 @@ PrefetchLocalBuffer(SMgrRelation smgr, ForkNumber forkNum, * (hence, usage_count is always advanced). */ BufferDesc * -LocalBufferAlloc(SMgrRelation smgr, ForkNumber forkNum, BlockNumber blockNum, - bool *foundPtr) +LocalBufferAlloc(SMgrFileHandle sfile, BlockNumber blockNum, bool *foundPtr) { BufferTag newTag; /* identity of requested block */ LocalBufferLookupEnt *hresult; @@ -117,7 +115,7 @@ LocalBufferAlloc(SMgrRelation smgr, ForkNumber forkNum, BlockNumber blockNum, bool found; uint32 buf_state; - InitBufferTag(&newTag, &smgr->smgr_rlocator.locator, forkNum, blockNum); + InitBufferTag(&newTag, &sfile->smgr_locator.locator, sfile->smgr_locator.forknum, blockNum); /* Initialize local buffers if first request in this session */ if (LocalBufHash == NULL) @@ -134,7 +132,7 @@ LocalBufferAlloc(SMgrRelation smgr, ForkNumber forkNum, BlockNumber blockNum, Assert(BufferTagsEqual(&bufHdr->tag, &newTag)); #ifdef LBDEBUG fprintf(stderr, "LB ALLOC (%u,%d,%d) %d\n", - smgr->smgr_rlocator.locator.relNumber, forkNum, blockNum, -b - 1); + sfile->smgr_locator.locator.relNumber, sfile->smgr_locator.forknum, blockNum, -b - 1); #endif buf_state = pg_atomic_read_u32(&bufHdr->state); @@ -162,7 +160,7 @@ LocalBufferAlloc(SMgrRelation smgr, ForkNumber forkNum, BlockNumber blockNum, #ifdef LBDEBUG fprintf(stderr, "LB ALLOC (%u,%d,%d) %d\n", - smgr->smgr_rlocator.locator.relNumber, forkNum, blockNum, + sfile->smgr_locator.locator.relNumber, sfile->smgr_locator.forknum, blockNum, -nextFreeLocalBuf - 1); #endif @@ -211,17 +209,16 @@ LocalBufferAlloc(SMgrRelation smgr, ForkNumber forkNum, BlockNumber blockNum, */ if (buf_state & BM_DIRTY) { - SMgrRelation oreln; + SMgrFileHandle ofile; Page localpage = (char *) LocalBufHdrGetBlock(bufHdr); - /* Find smgr relation for buffer */ - oreln = smgropen(BufTagGetRelFileLocator(&bufHdr->tag), MyBackendId); + /* Find smgr file handle for buffer */ + ofile = smgropen(BufTagGetRelFileLocator(&bufHdr->tag), MyBackendId, BufTagGetForkNum(&bufHdr->tag)); PageSetChecksumInplace(localpage, bufHdr->tag.blockNum); /* And write... */ - smgrwrite(oreln, - BufTagGetForkNum(&bufHdr->tag), + smgrwrite(ofile, bufHdr->tag.blockNum, localpage, false); diff --git a/src/backend/storage/freespace/freespace.c b/src/backend/storage/freespace/freespace.c index 3e9693b293b..a825a0b5af3 100644 --- a/src/backend/storage/freespace/freespace.c +++ b/src/backend/storage/freespace/freespace.c @@ -271,7 +271,7 @@ FreeSpaceMapPrepareTruncateRel(Relation rel, BlockNumber nblocks) * If no FSM has been created yet for this relation, there's nothing to * truncate. */ - if (!smgrexists(RelationGetSmgr(rel), FSM_FORKNUM)) + if (!smgrexists(RelationGetSmgr(rel, FSM_FORKNUM))) return InvalidBlockNumber; /* Get the location in the FSM of the first removed heap block */ @@ -317,7 +317,7 @@ FreeSpaceMapPrepareTruncateRel(Relation rel, BlockNumber nblocks) else { new_nfsmblocks = fsm_logical_to_physical(first_removed_address); - if (smgrnblocks(RelationGetSmgr(rel), FSM_FORKNUM) <= new_nfsmblocks) + if (smgrnblocks(RelationGetSmgr(rel, FSM_FORKNUM)) <= new_nfsmblocks) return InvalidBlockNumber; /* nothing to do; the FSM was already * smaller */ } @@ -532,14 +532,14 @@ fsm_readbuf(Relation rel, FSMAddress addr, bool extend) { BlockNumber blkno = fsm_logical_to_physical(addr); Buffer buf; - SMgrRelation reln; + SMgrFileHandle fsm_file; /* * Caution: re-using this smgr pointer could fail if the relcache entry * gets closed. It's safe as long as we only do smgr-level operations * between here and the last use of the pointer. */ - reln = RelationGetSmgr(rel); + fsm_file = RelationGetSmgr(rel, FSM_FORKNUM); /* * If we haven't cached the size of the FSM yet, check it first. Also @@ -547,19 +547,19 @@ fsm_readbuf(Relation rel, FSMAddress addr, bool extend) * value might be stale. (We send smgr inval messages on truncation, but * not on extension.) */ - if (reln->smgr_cached_nblocks[FSM_FORKNUM] == InvalidBlockNumber || - blkno >= reln->smgr_cached_nblocks[FSM_FORKNUM]) + if (fsm_file->smgr_cached_nblocks == InvalidBlockNumber || + blkno >= fsm_file->smgr_cached_nblocks) { /* Invalidate the cache so smgrnblocks asks the kernel. */ - reln->smgr_cached_nblocks[FSM_FORKNUM] = InvalidBlockNumber; - if (smgrexists(reln, FSM_FORKNUM)) - smgrnblocks(reln, FSM_FORKNUM); + fsm_file->smgr_cached_nblocks = InvalidBlockNumber; + if (smgrexists(fsm_file)) + smgrnblocks(fsm_file); else - reln->smgr_cached_nblocks[FSM_FORKNUM] = 0; + fsm_file->smgr_cached_nblocks = 0; } /* Handle requests beyond EOF */ - if (blkno >= reln->smgr_cached_nblocks[FSM_FORKNUM]) + if (blkno >= fsm_file->smgr_cached_nblocks) { if (extend) fsm_extend(rel, blkno + 1); @@ -609,7 +609,7 @@ fsm_extend(Relation rel, BlockNumber fsm_nblocks) { BlockNumber fsm_nblocks_now; PGAlignedBlock pg; - SMgrRelation reln; + SMgrFileHandle fsm_file; PageInit((Page) pg.data, BLCKSZ, 0); @@ -630,29 +630,28 @@ fsm_extend(Relation rel, BlockNumber fsm_nblocks) * gets closed. It's safe as long as we only do smgr-level operations * between here and the last use of the pointer. */ - reln = RelationGetSmgr(rel); + fsm_file = RelationGetSmgr(rel, FSM_FORKNUM); /* * Create the FSM file first if it doesn't exist. If * smgr_cached_nblocks[FSM_FORKNUM] is positive then it must exist, no * need for an smgrexists call. */ - if ((reln->smgr_cached_nblocks[FSM_FORKNUM] == 0 || - reln->smgr_cached_nblocks[FSM_FORKNUM] == InvalidBlockNumber) && - !smgrexists(reln, FSM_FORKNUM)) - smgrcreate(reln, FSM_FORKNUM, false); + if ((fsm_file->smgr_cached_nblocks == 0 || + fsm_file->smgr_cached_nblocks == InvalidBlockNumber) && + !smgrexists(fsm_file)) + smgrcreate(fsm_file, false); /* Invalidate cache so that smgrnblocks() asks the kernel. */ - reln->smgr_cached_nblocks[FSM_FORKNUM] = InvalidBlockNumber; - fsm_nblocks_now = smgrnblocks(reln, FSM_FORKNUM); + fsm_file->smgr_cached_nblocks = InvalidBlockNumber; + fsm_nblocks_now = smgrnblocks(fsm_file); /* Extend as needed. */ while (fsm_nblocks_now < fsm_nblocks) { PageSetChecksumInplace((Page) pg.data, fsm_nblocks_now); - smgrextend(reln, FSM_FORKNUM, fsm_nblocks_now, - pg.data, false); + smgrextend(fsm_file, fsm_nblocks_now, pg.data, false); fsm_nblocks_now++; } diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c index 8f1ded7338f..8601e1c0dfb 100644 --- a/src/backend/storage/ipc/ipci.c +++ b/src/backend/storage/ipc/ipci.c @@ -117,9 +117,7 @@ CalculateShmemSize(int *num_semaphores) size = add_size(size, XLogPrefetchShmemSize()); size = add_size(size, XLOGShmemSize()); size = add_size(size, XLogRecoveryShmemSize()); - size = add_size(size, CLOGShmemSize()); size = add_size(size, CommitTsShmemSize()); - size = add_size(size, SUBTRANSShmemSize()); size = add_size(size, TwoPhaseShmemSize()); size = add_size(size, BackgroundWorkerShmemSize()); size = add_size(size, MultiXactShmemSize()); @@ -241,9 +239,7 @@ CreateSharedMemoryAndSemaphores(void) XLOGShmemInit(); XLogPrefetchShmemInit(); XLogRecoveryShmemInit(); - CLOGShmemInit(); CommitTsShmemInit(); - SUBTRANSShmemInit(); MultiXactShmemInit(); InitBufferPool(); diff --git a/src/backend/storage/lmgr/predicate.c b/src/backend/storage/lmgr/predicate.c index 327adef5d3c..22b06a29bc1 100644 --- a/src/backend/storage/lmgr/predicate.c +++ b/src/backend/storage/lmgr/predicate.c @@ -313,17 +313,9 @@ ((targethash) ^ ((uint32) PointerGetDatum((predicatelocktag)->myXact)) \ << LOG2_NUM_PREDICATELOCK_PARTITIONS) - -/* - * The SLRU buffer area through which we access the old xids. - */ -static SlruCtlData SerialSlruCtlData; - -#define SerialSlruCtl (&SerialSlruCtlData) - #define SERIAL_PAGESIZE BLCKSZ #define SERIAL_ENTRYSIZE sizeof(SerCommitSeqNo) -#define SERIAL_ENTRIESPERPAGE (SERIAL_PAGESIZE / SERIAL_ENTRYSIZE) +#define SERIAL_ENTRIESPERPAGE ((SERIAL_PAGESIZE - SizeOfPageHeaderData) / SERIAL_ENTRYSIZE) /* * Set maximum pages based on the number needed to track all transactions. @@ -332,12 +324,13 @@ static SlruCtlData SerialSlruCtlData; #define SerialNextPage(page) (((page) >= SERIAL_MAX_PAGE) ? 0 : (page) + 1) -#define SerialValue(slotno, xid) (*((SerCommitSeqNo *) \ - (SerialSlruCtl->shared->page_buffer[slotno] + \ +#define SerialValue(buffer, xid) (*((SerCommitSeqNo *) \ + (PageGetContents(BufferGetPage(buffer)) + \ ((((uint32) (xid)) % SERIAL_ENTRIESPERPAGE) * SERIAL_ENTRYSIZE)))) #define SerialPage(xid) (((uint32) (xid)) / SERIAL_ENTRIESPERPAGE) + typedef struct SerialControlData { int headPage; /* newest initialized page */ @@ -849,10 +842,14 @@ SerialPagePrecedesLogicallyUnitTests(void) * requires burning ~2B XIDs in single-user mode, a negligible * possibility. Moreover, if it does happen, the consequence would be * mild, namely a new transaction failing in SimpleLruReadPage(). + * + * NOTE: After adding page headers, the defect affects two pages. + * We now assert correct treatment of its second to prior page. + * */ headPage = oldestPage; targetPage = newestPage; - Assert(SerialPagePrecedesLogically(headPage, targetPage - 1)); + Assert(SerialPagePrecedesLogically(headPage, targetPage - 2)); #if 0 Assert(SerialPagePrecedesLogically(headPage, targetPage)); #endif @@ -867,17 +864,10 @@ SerialInit(void) { bool found; - /* - * Set up SLRU management of the pg_serial data. - */ - SerialSlruCtl->PagePrecedes = SerialPagePrecedesLogically; - SimpleLruInit(SerialSlruCtl, "Serial", - NUM_SERIAL_BUFFERS, 0, SerialSLRULock, "pg_serial", - LWTRANCHE_SERIAL_BUFFER, SYNC_HANDLER_NONE); #ifdef USE_ASSERT_CHECKING SerialPagePrecedesLogicallyUnitTests(); #endif - SlruPagePrecedesUnitTests(SerialSlruCtl, SERIAL_ENTRIESPERPAGE); + SlruPagePrecedesUnitTests(SerialPagePrecedesLogically, SERIAL_ENTRIESPERPAGE); /* * Create or attach to the SerialControl structure. @@ -907,9 +897,9 @@ SerialAdd(TransactionId xid, SerCommitSeqNo minConflictCommitSeqNo) { TransactionId tailXid; int targetPage; - int slotno; int firstZeroPage; bool isNewPage; + Buffer buffer; Assert(TransactionIdIsValid(xid)); @@ -954,16 +944,23 @@ SerialAdd(TransactionId xid, SerCommitSeqNo minConflictCommitSeqNo) /* Initialize intervening pages. */ while (firstZeroPage != targetPage) { - (void) SimpleLruZeroPage(SerialSlruCtl, firstZeroPage); + buffer = ZeroSlruBuffer(SLRU_SERIAL_ID, firstZeroPage); + PageSetHeaderDataNonRel(BufferGetPage(buffer), firstZeroPage, InvalidXLogRecPtr, BLCKSZ, PG_METAPAGE_LAYOUT_VERSION); + MarkBufferDirty(buffer); + UnlockReleaseBuffer(buffer); firstZeroPage = SerialNextPage(firstZeroPage); } - slotno = SimpleLruZeroPage(SerialSlruCtl, targetPage); + buffer = ZeroSlruBuffer(SLRU_SERIAL_ID, targetPage); } else - slotno = SimpleLruReadPage(SerialSlruCtl, targetPage, true, xid); + { + buffer = ReadSlruBuffer(SLRU_SERIAL_ID, targetPage, RBM_NORMAL); + LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE); + } - SerialValue(slotno, xid) = minConflictCommitSeqNo; - SerialSlruCtl->shared->page_dirty[slotno] = true; + SerialValue(buffer, xid) = minConflictCommitSeqNo; + MarkBufferDirty(buffer); + UnlockReleaseBuffer(buffer); LWLockRelease(SerialSLRULock); } @@ -979,7 +976,7 @@ SerialGetMinConflictCommitSeqNo(TransactionId xid) TransactionId headXid; TransactionId tailXid; SerCommitSeqNo val; - int slotno; + Buffer buffer; Assert(TransactionIdIsValid(xid)); @@ -1001,9 +998,9 @@ SerialGetMinConflictCommitSeqNo(TransactionId xid) * The following function must be called without holding SerialSLRULock, * but will return with that lock held, which must then be released. */ - slotno = SimpleLruReadPage_ReadOnly(SerialSlruCtl, - SerialPage(xid), xid); - val = SerialValue(slotno, xid); + buffer = ReadSlruBuffer(SLRU_SERIAL_ID, SerialPage(xid), RBM_NORMAL); + val = SerialValue(buffer, xid); + ReleaseBuffer(buffer); LWLockRelease(SerialSLRULock); return val; } @@ -1122,19 +1119,7 @@ CheckPointPredicate(void) LWLockRelease(SerialSLRULock); /* Truncate away pages that are no longer required */ - SimpleLruTruncate(SerialSlruCtl, tailPage); - - /* - * Write dirty SLRU pages to disk - * - * This is not actually necessary from a correctness point of view. We do - * it merely as a debugging aid. - * - * We're doing this after the truncation to avoid writing pages right - * before deleting the file in which they sit, which would be completely - * pointless. - */ - SimpleLruWriteAll(SerialSlruCtl, true); + SimpleLruTruncate(SLRU_SERIAL_ID, SerialPagePrecedesLogically, tailPage); } /*------------------------------------------------------------------------*/ @@ -1396,7 +1381,6 @@ PredicateLockShmemSize(void) /* Shared memory structures for SLRU tracking of old committed xids. */ size = add_size(size, sizeof(SerialControlData)); - size = add_size(size, SimpleLruShmemSize(NUM_SERIAL_BUFFERS, 0)); return size; } diff --git a/src/backend/storage/page/bufpage.c b/src/backend/storage/page/bufpage.c index 92994f8f395..1a810754c5c 100644 --- a/src/backend/storage/page/bufpage.c +++ b/src/backend/storage/page/bufpage.c @@ -59,6 +59,27 @@ PageInit(Page page, Size pageSize, Size specialSize) /* p->pd_prune_xid = InvalidTransactionId; done by above MemSet */ } +void +PageInitSLRU(Page page, Size pageSize, Size specialSize) +{ + PageHeader p = (PageHeader) page; + + specialSize = MAXALIGN(specialSize); + + Assert(pageSize == BLCKSZ); + Assert(pageSize > specialSize + SizeOfPageHeaderData); + + /* Make sure all fields of page are zero, as well as unused space */ + MemSet(p, 0, pageSize); + + p->pd_flags = 0; + p->pd_lower = SizeOfPageHeaderData; + p->pd_upper = pageSize - specialSize; + p->pd_special = pageSize - specialSize; + PageSetPageSizeAndVersion(page, pageSize, PG_METAPAGE_LAYOUT_VERSION); + /* p->pd_prune_xid = InvalidTransactionId; done by above MemSet */ +} + /* * PageIsVerifiedExtended @@ -103,7 +124,7 @@ PageIsVerifiedExtended(Page page, BlockNumber blkno, int flags) if (DataChecksumsEnabled()) { checksum = pg_checksum_page((char *) page, blkno); - + if (checksum != p->pd_checksum) checksum_failure = true; } diff --git a/src/backend/storage/smgr/md.c b/src/backend/storage/smgr/md.c index 60c9905eff9..55ebd45b7d1 100644 --- a/src/backend/storage/smgr/md.c +++ b/src/backend/storage/smgr/md.c @@ -66,10 +66,10 @@ * out to an unlinked old copy of a segment file that will eventually * disappear. * - * File descriptors are stored in the per-fork md_seg_fds arrays inside - * SMgrRelation. The length of these arrays is stored in md_num_open_segs. - * Note that a fork's md_num_open_segs having a specific value does not - * necessarily mean the relation doesn't have additional segments; we may + * File descriptors are stored in the md_seg_fds array inside + * SMgrFileData. The length of the array is stored in md_num_open_segs. + * Note that md_num_open_segs having a specific value does not + * necessarily mean the file doesn't have additional segments; we may * just not have opened the next segment yet. (We could not have "all * segments are in the array" as an invariant anyway, since another backend * could extend the relation while we aren't looking.) We do not have @@ -121,26 +121,18 @@ static MemoryContext MdCxt; /* context for all MdfdVec objects */ /* local routines */ -static void mdunlinkfork(RelFileLocatorBackend rlocator, ForkNumber forknum, - bool isRedo); -static MdfdVec *mdopenfork(SMgrRelation reln, ForkNumber forknum, int behavior); -static void register_dirty_segment(SMgrRelation reln, ForkNumber forknum, - MdfdVec *seg); -static void register_unlink_segment(RelFileLocatorBackend rlocator, ForkNumber forknum, - BlockNumber segno); -static void register_forget_request(RelFileLocatorBackend rlocator, ForkNumber forknum, - BlockNumber segno); -static void _fdvec_resize(SMgrRelation reln, - ForkNumber forknum, - int nseg); -static char *_mdfd_segpath(SMgrRelation reln, ForkNumber forknum, - BlockNumber segno); -static MdfdVec *_mdfd_openseg(SMgrRelation reln, ForkNumber forknum, +static void mdunlinkfork(SMgrFileLocator slocator, bool isRedo); +static MdfdVec *mdopenfork(SMgrFileHandle sfile, int behavior); +static void register_dirty_segment(SMgrFileHandle sfile, MdfdVec *seg); +static void register_unlink_segment(SMgrFileLocator slocator, BlockNumber segno); +static void register_forget_request(SMgrFileLocator slocator, BlockNumber segno); +static void _fdvec_resize(SMgrFileHandle sfile, int nseg); +static char *_mdfd_segpath(SMgrFileHandle sfile, BlockNumber segno); +static MdfdVec *_mdfd_openseg(SMgrFileHandle sfile, BlockNumber segno, int oflags); -static MdfdVec *_mdfd_getseg(SMgrRelation reln, ForkNumber forknum, +static MdfdVec *_mdfd_getseg(SMgrFileHandle sfile, BlockNumber blkno, bool skipFsync, int behavior); -static BlockNumber _mdnblocks(SMgrRelation reln, ForkNumber forknum, - MdfdVec *seg); +static BlockNumber _mdnblocks(SMgrFileHandle sfile, MdfdVec *seg); /* @@ -160,7 +152,7 @@ mdinit(void) * Note: this will return true for lingering files, with pending deletions */ bool -mdexists(SMgrRelation reln, ForkNumber forknum) +mdexists(SMgrFileHandle sfile) { /* * Close it first, to ensure that we notice if the fork has been unlinked @@ -168,9 +160,9 @@ mdexists(SMgrRelation reln, ForkNumber forknum) * which already closes relations when dropping them. */ if (!InRecovery) - mdclose(reln, forknum); + mdclose(sfile); - return (mdopenfork(reln, forknum, EXTENSION_RETURN_NULL) != NULL); + return (mdopenfork(sfile, EXTENSION_RETURN_NULL) != NULL); } /* @@ -179,16 +171,16 @@ mdexists(SMgrRelation reln, ForkNumber forknum) * If isRedo is true, it's okay for the relation to exist already. */ void -mdcreate(SMgrRelation reln, ForkNumber forknum, bool isRedo) +mdcreate(SMgrFileHandle sfile, bool isRedo) { MdfdVec *mdfd; char *path; File fd; - if (isRedo && reln->md_num_open_segs[forknum] > 0) + if (isRedo && sfile->md_num_open_segs > 0) return; /* created and opened already... */ - Assert(reln->md_num_open_segs[forknum] == 0); + Assert(sfile->md_num_open_segs == 0); /* * We may be using the target table space for the first time in this @@ -199,11 +191,14 @@ mdcreate(SMgrRelation reln, ForkNumber forknum, bool isRedo) * should be here and not in commands/tablespace.c? But that would imply * importing a lot of stuff that smgr.c oughtn't know, either. */ - TablespaceCreateDbspace(reln->smgr_rlocator.locator.spcOid, - reln->smgr_rlocator.locator.dbOid, - isRedo); + if (sfile->smgr_locator.locator.spcOid != SLRU_SPC_OID) + { + TablespaceCreateDbspace(sfile->smgr_locator.locator.spcOid, + sfile->smgr_locator.locator.dbOid, + isRedo); + } - path = relpath(reln->smgr_rlocator, forknum); + path = smgrfilepath(sfile->smgr_locator); fd = PathNameOpenFile(path, O_RDWR | O_CREAT | O_EXCL | PG_BINARY); @@ -225,8 +220,8 @@ mdcreate(SMgrRelation reln, ForkNumber forknum, bool isRedo) pfree(path); - _fdvec_resize(reln, forknum, 1); - mdfd = &reln->md_seg_fds[forknum][0]; + _fdvec_resize(sfile, 1); + mdfd = &sfile->md_seg_fds[0]; mdfd->mdfd_vfd = fd; mdfd->mdfd_segno = 0; } @@ -293,16 +288,9 @@ mdcreate(SMgrRelation reln, ForkNumber forknum, bool isRedo) * we are usually not in a transaction anymore when this is called. */ void -mdunlink(RelFileLocatorBackend rlocator, ForkNumber forknum, bool isRedo) +mdunlink(SMgrFileLocator slocator, bool isRedo) { - /* Now do the per-fork work */ - if (forknum == InvalidForkNumber) - { - for (forknum = 0; forknum <= MAX_FORKNUM; forknum++) - mdunlinkfork(rlocator, forknum, isRedo); - } - else - mdunlinkfork(rlocator, forknum, isRedo); + mdunlinkfork(slocator, isRedo); } /* @@ -330,29 +318,29 @@ do_truncate(const char *path) } static void -mdunlinkfork(RelFileLocatorBackend rlocator, ForkNumber forknum, bool isRedo) +mdunlinkfork(SMgrFileLocator slocator, bool isRedo) { char *path; int ret; int save_errno; - path = relpath(rlocator, forknum); + path = smgrfilepath(slocator); /* * Truncate and then unlink the first segment, or just register a request * to unlink it later, as described in the comments for mdunlink(). */ - if (isRedo || IsBinaryUpgrade || forknum != MAIN_FORKNUM || - RelFileLocatorBackendIsTemp(rlocator)) + if (isRedo || IsBinaryUpgrade || slocator.forknum != MAIN_FORKNUM || + SMgrFileLocatorIsTemp(slocator)) { - if (!RelFileLocatorBackendIsTemp(rlocator)) + if (!SMgrFileLocatorIsTemp(slocator)) { /* Prevent other backends' fds from holding on to the disk space */ ret = do_truncate(path); /* Forget any pending sync requests for the first segment */ save_errno = errno; - register_forget_request(rlocator, forknum, 0 /* first seg */ ); + register_forget_request(slocator, 0 /* first seg */ ); errno = save_errno; } else @@ -379,7 +367,7 @@ mdunlinkfork(RelFileLocatorBackend rlocator, ForkNumber forknum, bool isRedo) /* Register request to unlink first segment later */ save_errno = errno; - register_unlink_segment(rlocator, forknum, 0 /* first seg */ ); + register_unlink_segment(slocator, 0 /* first seg */ ); errno = save_errno; } @@ -404,7 +392,7 @@ mdunlinkfork(RelFileLocatorBackend rlocator, ForkNumber forknum, bool isRedo) { sprintf(segpath, "%s.%u", path, segno); - if (!RelFileLocatorBackendIsTemp(rlocator)) + if (!SMgrFileLocatorIsTemp(slocator)) { /* * Prevent other backends' fds from holding on to the disk @@ -417,7 +405,7 @@ mdunlinkfork(RelFileLocatorBackend rlocator, ForkNumber forknum, bool isRedo) * Forget any pending sync requests for this segment before we * try to unlink. */ - register_forget_request(rlocator, forknum, segno); + register_forget_request(slocator, segno); } if (unlink(segpath) < 0) @@ -446,7 +434,7 @@ mdunlinkfork(RelFileLocatorBackend rlocator, ForkNumber forknum, bool isRedo) * causes intervening file space to become filled with zeroes. */ void -mdextend(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, +mdextend(SMgrFileHandle sfile, BlockNumber blocknum, char *buffer, bool skipFsync) { off_t seekpos; @@ -455,7 +443,7 @@ mdextend(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, /* This assert is too expensive to have on normally ... */ #ifdef CHECK_WRITE_VS_EXTEND - Assert(blocknum >= mdnblocks(reln, forknum)); + Assert(blocknum >= mdnblocks(sfile)); #endif /* @@ -468,10 +456,10 @@ mdextend(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, ereport(ERROR, (errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED), errmsg("cannot extend file \"%s\" beyond %u blocks", - relpath(reln->smgr_rlocator, forknum), + smgrfilepath(sfile->smgr_locator), InvalidBlockNumber))); - v = _mdfd_getseg(reln, forknum, blocknum, skipFsync, EXTENSION_CREATE); + v = _mdfd_getseg(sfile, blocknum, skipFsync, EXTENSION_CREATE); seekpos = (off_t) BLCKSZ * (blocknum % ((BlockNumber) RELSEG_SIZE)); @@ -494,10 +482,10 @@ mdextend(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, errhint("Check free disk space."))); } - if (!skipFsync && !SmgrIsTemp(reln)) - register_dirty_segment(reln, forknum, v); + if (!skipFsync && !SmgrIsTemp(sfile)) + register_dirty_segment(sfile, v); - Assert(_mdnblocks(reln, forknum, v) <= ((BlockNumber) RELSEG_SIZE)); + Assert(_mdnblocks(sfile, v) <= ((BlockNumber) RELSEG_SIZE)); } /* @@ -511,17 +499,17 @@ mdextend(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, * invent one out of whole cloth. */ static MdfdVec * -mdopenfork(SMgrRelation reln, ForkNumber forknum, int behavior) +mdopenfork(SMgrFileHandle sfile, int behavior) { MdfdVec *mdfd; char *path; File fd; /* No work if already open */ - if (reln->md_num_open_segs[forknum] > 0) - return &reln->md_seg_fds[forknum][0]; + if (sfile->md_num_open_segs > 0) + return &sfile->md_seg_fds[0]; - path = relpath(reln->smgr_rlocator, forknum); + path = smgrfilepath(sfile->smgr_locator); fd = PathNameOpenFile(path, O_RDWR | PG_BINARY); @@ -540,12 +528,12 @@ mdopenfork(SMgrRelation reln, ForkNumber forknum, int behavior) pfree(path); - _fdvec_resize(reln, forknum, 1); - mdfd = &reln->md_seg_fds[forknum][0]; + _fdvec_resize(sfile, 1); + mdfd = &sfile->md_seg_fds[0]; mdfd->mdfd_vfd = fd; mdfd->mdfd_segno = 0; - Assert(_mdnblocks(reln, forknum, mdfd) <= ((BlockNumber) RELSEG_SIZE)); + Assert(_mdnblocks(sfile, mdfd) <= ((BlockNumber) RELSEG_SIZE)); return mdfd; } @@ -554,20 +542,19 @@ mdopenfork(SMgrRelation reln, ForkNumber forknum, int behavior) * mdopen() -- Initialize newly-opened relation. */ void -mdopen(SMgrRelation reln) +mdopen(SMgrFileHandle sfile) { /* mark it not open */ - for (int forknum = 0; forknum <= MAX_FORKNUM; forknum++) - reln->md_num_open_segs[forknum] = 0; + sfile->md_num_open_segs = 0; } /* * mdclose() -- Close the specified relation, if it isn't closed already. */ void -mdclose(SMgrRelation reln, ForkNumber forknum) +mdclose(SMgrFileHandle sfile) { - int nopensegs = reln->md_num_open_segs[forknum]; + int nopensegs = sfile->md_num_open_segs; /* No work if already closed */ if (nopensegs == 0) @@ -576,10 +563,10 @@ mdclose(SMgrRelation reln, ForkNumber forknum) /* close segments starting from the end */ while (nopensegs > 0) { - MdfdVec *v = &reln->md_seg_fds[forknum][nopensegs - 1]; + MdfdVec *v = &sfile->md_seg_fds[nopensegs - 1]; FileClose(v->mdfd_vfd); - _fdvec_resize(reln, forknum, nopensegs - 1); + _fdvec_resize(sfile, nopensegs - 1); nopensegs--; } } @@ -588,13 +575,13 @@ mdclose(SMgrRelation reln, ForkNumber forknum) * mdprefetch() -- Initiate asynchronous read of the specified block of a relation */ bool -mdprefetch(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum) +mdprefetch(SMgrFileHandle sfile, BlockNumber blocknum) { #ifdef USE_PREFETCH off_t seekpos; MdfdVec *v; - v = _mdfd_getseg(reln, forknum, blocknum, false, + v = _mdfd_getseg(sfile, blocknum, false, InRecovery ? EXTENSION_RETURN_NULL : EXTENSION_FAIL); if (v == NULL) return false; @@ -616,8 +603,7 @@ mdprefetch(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum) * considerably more efficient than doing so individually. */ void -mdwriteback(SMgrRelation reln, ForkNumber forknum, - BlockNumber blocknum, BlockNumber nblocks) +mdwriteback(SMgrFileHandle sfile, BlockNumber blocknum, BlockNumber nblocks) { /* * Issue flush requests in as few requests as possible; have to split at @@ -631,7 +617,7 @@ mdwriteback(SMgrRelation reln, ForkNumber forknum, int segnum_start, segnum_end; - v = _mdfd_getseg(reln, forknum, blocknum, true /* not used */ , + v = _mdfd_getseg(sfile, blocknum, true /* not used */ , EXTENSION_DONT_OPEN); /* @@ -668,20 +654,21 @@ mdwriteback(SMgrRelation reln, ForkNumber forknum, * mdread() -- Read the specified block from a relation. */ void -mdread(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, +mdread(SMgrFileHandle sfile, BlockNumber blocknum, char *buffer) { off_t seekpos; int nbytes; MdfdVec *v; - TRACE_POSTGRESQL_SMGR_MD_READ_START(forknum, blocknum, - reln->smgr_rlocator.locator.spcOid, - reln->smgr_rlocator.locator.dbOid, - reln->smgr_rlocator.locator.relNumber, - reln->smgr_rlocator.backend); + TRACE_POSTGRESQL_SMGR_MD_READ_START(sfile->smgr_forknum, + blocknum, + sfile->smgr_rlocator.locator.spcOid, + sfile->smgr_rlocator.locator.dbOid, + sfile->smgr_rlocator.locator.relNumber, + sfile->smgr_rlocator.backend); - v = _mdfd_getseg(reln, forknum, blocknum, false, + v = _mdfd_getseg(sfile, blocknum, false, EXTENSION_FAIL | EXTENSION_CREATE_RECOVERY); seekpos = (off_t) BLCKSZ * (blocknum % ((BlockNumber) RELSEG_SIZE)); @@ -690,11 +677,12 @@ mdread(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, nbytes = FileRead(v->mdfd_vfd, buffer, BLCKSZ, seekpos, WAIT_EVENT_DATA_FILE_READ); - TRACE_POSTGRESQL_SMGR_MD_READ_DONE(forknum, blocknum, - reln->smgr_rlocator.locator.spcOid, - reln->smgr_rlocator.locator.dbOid, - reln->smgr_rlocator.locator.relNumber, - reln->smgr_rlocator.backend, + TRACE_POSTGRESQL_SMGR_MD_READ_DONE(sfile->smgr_forknum, + blocknum, + sfile->smgr_rlocator.locator.spcOid, + sfile->smgr_rlocator.locator.dbOid, + sfile->smgr_rlocator.locator.relNumber, + sfile->smgr_rlocator.backend, nbytes, BLCKSZ); @@ -733,7 +721,7 @@ mdread(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, * use mdextend(). */ void -mdwrite(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, +mdwrite(SMgrFileHandle sfile, BlockNumber blocknum, char *buffer, bool skipFsync) { off_t seekpos; @@ -742,16 +730,17 @@ mdwrite(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, /* This assert is too expensive to have on normally ... */ #ifdef CHECK_WRITE_VS_EXTEND - Assert(blocknum < mdnblocks(reln, forknum)); + Assert(blocknum < mdnblocks(sfile)); #endif - TRACE_POSTGRESQL_SMGR_MD_WRITE_START(forknum, blocknum, - reln->smgr_rlocator.locator.spcOid, - reln->smgr_rlocator.locator.dbOid, - reln->smgr_rlocator.locator.relNumber, - reln->smgr_rlocator.backend); + TRACE_POSTGRESQL_SMGR_MD_WRITE_START(sfile->smgr_forknum, + blocknum, + sfile->smgr_rlocator.locator.spcOid, + sfile->smgr_rlocator.locator.dbOid, + sfile->smgr_rlocator.locator.relNumber, + sfile->smgr_rlocator.backend); - v = _mdfd_getseg(reln, forknum, blocknum, skipFsync, + v = _mdfd_getseg(sfile, blocknum, skipFsync, EXTENSION_FAIL | EXTENSION_CREATE_RECOVERY); seekpos = (off_t) BLCKSZ * (blocknum % ((BlockNumber) RELSEG_SIZE)); @@ -760,11 +749,12 @@ mdwrite(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, nbytes = FileWrite(v->mdfd_vfd, buffer, BLCKSZ, seekpos, WAIT_EVENT_DATA_FILE_WRITE); - TRACE_POSTGRESQL_SMGR_MD_WRITE_DONE(forknum, blocknum, - reln->smgr_rlocator.locator.spcOid, - reln->smgr_rlocator.locator.dbOid, - reln->smgr_rlocator.locator.relNumber, - reln->smgr_rlocator.backend, + TRACE_POSTGRESQL_SMGR_MD_WRITE_DONE(sfile->smgr_forknum, + blocknum, + sfile->smgr_rlocator.locator.spcOid, + sfile->smgr_rlocator.locator.dbOid, + sfile->smgr_rlocator.locator.relNumber, + sfile->smgr_rlocator.backend, nbytes, BLCKSZ); @@ -785,8 +775,8 @@ mdwrite(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, errhint("Check free disk space."))); } - if (!skipFsync && !SmgrIsTemp(reln)) - register_dirty_segment(reln, forknum, v); + if (!skipFsync && !SmgrIsTemp(sfile)) + register_dirty_segment(sfile, v); } /* @@ -798,16 +788,16 @@ mdwrite(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, * are present in the array. */ BlockNumber -mdnblocks(SMgrRelation reln, ForkNumber forknum) +mdnblocks(SMgrFileHandle sfile) { MdfdVec *v; BlockNumber nblocks; BlockNumber segno; - mdopenfork(reln, forknum, EXTENSION_FAIL); + mdopenfork(sfile, EXTENSION_FAIL); /* mdopen has opened the first segment */ - Assert(reln->md_num_open_segs[forknum] > 0); + Assert(sfile->md_num_open_segs > 0); /* * Start from the last open segments, to avoid redundant seeks. We have @@ -822,12 +812,12 @@ mdnblocks(SMgrRelation reln, ForkNumber forknum) * that's OK because the checkpointer never needs to compute relation * size.) */ - segno = reln->md_num_open_segs[forknum] - 1; - v = &reln->md_seg_fds[forknum][segno]; + segno = sfile->md_num_open_segs - 1; + v = &sfile->md_seg_fds[segno]; for (;;) { - nblocks = _mdnblocks(reln, forknum, v); + nblocks = _mdnblocks(sfile, v); if (nblocks > ((BlockNumber) RELSEG_SIZE)) elog(FATAL, "segment too big"); if (nblocks < ((BlockNumber) RELSEG_SIZE)) @@ -845,7 +835,7 @@ mdnblocks(SMgrRelation reln, ForkNumber forknum) * undermines _mdfd_getseg's attempts to notice and report an error * upon access to a missing segment. */ - v = _mdfd_openseg(reln, forknum, segno, 0); + v = _mdfd_openseg(sfile, segno, 0); if (v == NULL) return segno * ((BlockNumber) RELSEG_SIZE); } @@ -855,7 +845,7 @@ mdnblocks(SMgrRelation reln, ForkNumber forknum) * mdtruncate() -- Truncate relation to specified number of blocks. */ void -mdtruncate(SMgrRelation reln, ForkNumber forknum, BlockNumber nblocks) +mdtruncate(SMgrFileHandle sfile, BlockNumber nblocks) { BlockNumber curnblk; BlockNumber priorblocks; @@ -865,7 +855,7 @@ mdtruncate(SMgrRelation reln, ForkNumber forknum, BlockNumber nblocks) * NOTE: mdnblocks makes sure we have opened all active segments, so that * truncation loop will get them all! */ - curnblk = mdnblocks(reln, forknum); + curnblk = mdnblocks(sfile); if (nblocks > curnblk) { /* Bogus request ... but no complaint if InRecovery */ @@ -873,7 +863,7 @@ mdtruncate(SMgrRelation reln, ForkNumber forknum, BlockNumber nblocks) return; ereport(ERROR, (errmsg("could not truncate file \"%s\" to %u blocks: it's only %u blocks now", - relpath(reln->smgr_rlocator, forknum), + smgrfilepath(sfile->smgr_locator), nblocks, curnblk))); } if (nblocks == curnblk) @@ -883,14 +873,14 @@ mdtruncate(SMgrRelation reln, ForkNumber forknum, BlockNumber nblocks) * Truncate segments, starting at the last one. Starting at the end makes * managing the memory for the fd array easier, should there be errors. */ - curopensegs = reln->md_num_open_segs[forknum]; + curopensegs = sfile->md_num_open_segs; while (curopensegs > 0) { MdfdVec *v; priorblocks = (curopensegs - 1) * RELSEG_SIZE; - v = &reln->md_seg_fds[forknum][curopensegs - 1]; + v = &sfile->md_seg_fds[curopensegs - 1]; if (priorblocks > nblocks) { @@ -904,14 +894,14 @@ mdtruncate(SMgrRelation reln, ForkNumber forknum, BlockNumber nblocks) errmsg("could not truncate file \"%s\": %m", FilePathName(v->mdfd_vfd)))); - if (!SmgrIsTemp(reln)) - register_dirty_segment(reln, forknum, v); + if (!SmgrIsTemp(sfile)) + register_dirty_segment(sfile, v); /* we never drop the 1st segment */ - Assert(v != &reln->md_seg_fds[forknum][0]); + Assert(v != &sfile->md_seg_fds[0]); FileClose(v->mdfd_vfd); - _fdvec_resize(reln, forknum, curopensegs - 1); + _fdvec_resize(sfile, curopensegs - 1); } else if (priorblocks + ((BlockNumber) RELSEG_SIZE) > nblocks) { @@ -930,8 +920,8 @@ mdtruncate(SMgrRelation reln, ForkNumber forknum, BlockNumber nblocks) errmsg("could not truncate file \"%s\" to %u blocks: %m", FilePathName(v->mdfd_vfd), nblocks))); - if (!SmgrIsTemp(reln)) - register_dirty_segment(reln, forknum, v); + if (!SmgrIsTemp(sfile)) + register_dirty_segment(sfile, v); } else { @@ -957,7 +947,7 @@ mdtruncate(SMgrRelation reln, ForkNumber forknum, BlockNumber nblocks) * segment may survive recovery, reintroducing unwanted data into the table. */ void -mdimmedsync(SMgrRelation reln, ForkNumber forknum) +mdimmedsync(SMgrFileHandle sfile) { int segno; int min_inactive_seg; @@ -966,9 +956,9 @@ mdimmedsync(SMgrRelation reln, ForkNumber forknum) * NOTE: mdnblocks makes sure we have opened all active segments, so that * fsync loop will get them all! */ - mdnblocks(reln, forknum); + mdnblocks(sfile); - min_inactive_seg = segno = reln->md_num_open_segs[forknum]; + min_inactive_seg = segno = sfile->md_num_open_segs; /* * Temporarily open inactive segments, then close them after sync. There @@ -976,12 +966,12 @@ mdimmedsync(SMgrRelation reln, ForkNumber forknum) * is harmless. We don't bother to clean them up and take a risk of * further trouble. The next mdclose() will soon close them. */ - while (_mdfd_openseg(reln, forknum, segno, 0) != NULL) + while (_mdfd_openseg(sfile, segno, 0) != NULL) segno++; while (segno > 0) { - MdfdVec *v = &reln->md_seg_fds[forknum][segno - 1]; + MdfdVec *v = &sfile->md_seg_fds[segno - 1]; if (FileSync(v->mdfd_vfd, WAIT_EVENT_DATA_FILE_IMMEDIATE_SYNC) < 0) ereport(data_sync_elevel(ERROR), @@ -993,7 +983,7 @@ mdimmedsync(SMgrRelation reln, ForkNumber forknum) if (segno > min_inactive_seg) { FileClose(v->mdfd_vfd); - _fdvec_resize(reln, forknum, segno - 1); + _fdvec_resize(sfile, segno - 1); } segno--; @@ -1010,14 +1000,14 @@ mdimmedsync(SMgrRelation reln, ForkNumber forknum) * enough to be a performance problem). */ static void -register_dirty_segment(SMgrRelation reln, ForkNumber forknum, MdfdVec *seg) +register_dirty_segment(SMgrFileHandle sfile, MdfdVec *seg) { FileTag tag; - INIT_MD_FILETAG(tag, reln->smgr_rlocator.locator, forknum, seg->mdfd_segno); + INIT_MD_FILETAG(tag, sfile->smgr_locator.locator, sfile->smgr_locator.forknum, seg->mdfd_segno); /* Temp relations should never be fsync'd */ - Assert(!SmgrIsTemp(reln)); + Assert(!SmgrIsTemp(sfile)); if (!RegisterSyncRequest(&tag, SYNC_REQUEST, false /* retryOnError */ )) { @@ -1036,15 +1026,14 @@ register_dirty_segment(SMgrRelation reln, ForkNumber forknum, MdfdVec *seg) * register_unlink_segment() -- Schedule a file to be deleted after next checkpoint */ static void -register_unlink_segment(RelFileLocatorBackend rlocator, ForkNumber forknum, - BlockNumber segno) +register_unlink_segment(SMgrFileLocator slocator, BlockNumber segno) { FileTag tag; - INIT_MD_FILETAG(tag, rlocator.locator, forknum, segno); + INIT_MD_FILETAG(tag, slocator.locator, slocator.forknum, segno); /* Should never be used with temp relations */ - Assert(!RelFileLocatorBackendIsTemp(rlocator)); + Assert(!SMgrFileLocatorIsTemp(slocator)); RegisterSyncRequest(&tag, SYNC_UNLINK_REQUEST, true /* retryOnError */ ); } @@ -1053,12 +1042,11 @@ register_unlink_segment(RelFileLocatorBackend rlocator, ForkNumber forknum, * register_forget_request() -- forget any fsyncs for a relation fork's segment */ static void -register_forget_request(RelFileLocatorBackend rlocator, ForkNumber forknum, - BlockNumber segno) +register_forget_request(SMgrFileLocator slocator, BlockNumber segno) { FileTag tag; - INIT_MD_FILETAG(tag, rlocator.locator, forknum, segno); + INIT_MD_FILETAG(tag, slocator.locator, slocator.forknum, segno); RegisterSyncRequest(&tag, SYNC_FORGET_REQUEST, true /* retryOnError */ ); } @@ -1081,57 +1069,25 @@ ForgetDatabaseSyncRequests(Oid dbid) RegisterSyncRequest(&tag, SYNC_FILTER_REQUEST, true /* retryOnError */ ); } -/* - * DropRelationFiles -- drop files of all given relations - */ -void -DropRelationFiles(RelFileLocator *delrels, int ndelrels, bool isRedo) -{ - SMgrRelation *srels; - int i; - - srels = palloc(sizeof(SMgrRelation) * ndelrels); - for (i = 0; i < ndelrels; i++) - { - SMgrRelation srel = smgropen(delrels[i], InvalidBackendId); - - if (isRedo) - { - ForkNumber fork; - - for (fork = 0; fork <= MAX_FORKNUM; fork++) - XLogDropRelation(delrels[i], fork); - } - srels[i] = srel; - } - - smgrdounlinkall(srels, ndelrels, isRedo); - - for (i = 0; i < ndelrels; i++) - smgrclose(srels[i]); - pfree(srels); -} - /* * _fdvec_resize() -- Resize the fork's open segments array */ static void -_fdvec_resize(SMgrRelation reln, - ForkNumber forknum, +_fdvec_resize(SMgrFileHandle sfile, int nseg) { if (nseg == 0) { - if (reln->md_num_open_segs[forknum] > 0) + if (sfile->md_num_open_segs > 0) { - pfree(reln->md_seg_fds[forknum]); - reln->md_seg_fds[forknum] = NULL; + pfree(sfile->md_seg_fds); + sfile->md_seg_fds = NULL; } } - else if (reln->md_num_open_segs[forknum] == 0) + else if (sfile->md_num_open_segs == 0) { - reln->md_seg_fds[forknum] = + sfile->md_seg_fds = MemoryContextAlloc(MdCxt, sizeof(MdfdVec) * nseg); } else @@ -1142,12 +1098,12 @@ _fdvec_resize(SMgrRelation reln, * FileClose(), and the memory context internally will sometimes avoid * doing an actual reallocation. */ - reln->md_seg_fds[forknum] = - repalloc(reln->md_seg_fds[forknum], + sfile->md_seg_fds = + repalloc(sfile->md_seg_fds, sizeof(MdfdVec) * nseg); } - reln->md_num_open_segs[forknum] = nseg; + sfile->md_num_open_segs = nseg; } /* @@ -1155,12 +1111,12 @@ _fdvec_resize(SMgrRelation reln, * returned string is palloc'd. */ static char * -_mdfd_segpath(SMgrRelation reln, ForkNumber forknum, BlockNumber segno) +_mdfd_segpath(SMgrFileHandle sfile, BlockNumber segno) { char *path, *fullpath; - path = relpath(reln->smgr_rlocator, forknum); + path = smgrfilepath(sfile->smgr_locator); if (segno > 0) { @@ -1178,14 +1134,14 @@ _mdfd_segpath(SMgrRelation reln, ForkNumber forknum, BlockNumber segno) * and make a MdfdVec object for it. Returns NULL on failure. */ static MdfdVec * -_mdfd_openseg(SMgrRelation reln, ForkNumber forknum, BlockNumber segno, +_mdfd_openseg(SMgrFileHandle sfile, BlockNumber segno, int oflags) { MdfdVec *v; File fd; char *fullpath; - fullpath = _mdfd_segpath(reln, forknum, segno); + fullpath = _mdfd_segpath(sfile, segno); /* open the file */ fd = PathNameOpenFile(fullpath, O_RDWR | PG_BINARY | oflags); @@ -1199,16 +1155,16 @@ _mdfd_openseg(SMgrRelation reln, ForkNumber forknum, BlockNumber segno, * Segments are always opened in order from lowest to highest, so we must * be adding a new one at the end. */ - Assert(segno == reln->md_num_open_segs[forknum]); + Assert(segno == sfile->md_num_open_segs); - _fdvec_resize(reln, forknum, segno + 1); + _fdvec_resize(sfile, segno + 1); /* fill the entry */ - v = &reln->md_seg_fds[forknum][segno]; + v = &sfile->md_seg_fds[segno]; v->mdfd_vfd = fd; v->mdfd_segno = segno; - Assert(_mdnblocks(reln, forknum, v) <= ((BlockNumber) RELSEG_SIZE)); + Assert(_mdnblocks(sfile, v) <= ((BlockNumber) RELSEG_SIZE)); /* all done */ return v; @@ -1223,7 +1179,7 @@ _mdfd_openseg(SMgrRelation reln, ForkNumber forknum, BlockNumber segno, * EXTENSION_CREATE case. */ static MdfdVec * -_mdfd_getseg(SMgrRelation reln, ForkNumber forknum, BlockNumber blkno, +_mdfd_getseg(SMgrFileHandle sfile, BlockNumber blkno, bool skipFsync, int behavior) { MdfdVec *v; @@ -1238,9 +1194,9 @@ _mdfd_getseg(SMgrRelation reln, ForkNumber forknum, BlockNumber blkno, targetseg = blkno / ((BlockNumber) RELSEG_SIZE); /* if an existing and opened segment, we're done */ - if (targetseg < reln->md_num_open_segs[forknum]) + if (targetseg < sfile->md_num_open_segs) { - v = &reln->md_seg_fds[forknum][targetseg]; + v = &sfile->md_seg_fds[targetseg]; return v; } @@ -1255,19 +1211,19 @@ _mdfd_getseg(SMgrRelation reln, ForkNumber forknum, BlockNumber blkno, * 'behavior'). Start with either the last opened, or the first segment if * none was opened before. */ - if (reln->md_num_open_segs[forknum] > 0) - v = &reln->md_seg_fds[forknum][reln->md_num_open_segs[forknum] - 1]; + if (sfile->md_num_open_segs > 0) + v = &sfile->md_seg_fds[sfile->md_num_open_segs - 1]; else { - v = mdopenfork(reln, forknum, behavior); + v = mdopenfork(sfile, behavior); if (!v) return NULL; /* if behavior & EXTENSION_RETURN_NULL */ } - for (nextsegno = reln->md_num_open_segs[forknum]; + for (nextsegno = sfile->md_num_open_segs; nextsegno <= targetseg; nextsegno++) { - BlockNumber nblocks = _mdnblocks(reln, forknum, v); + BlockNumber nblocks = _mdnblocks(sfile, v); int flags = 0; Assert(nextsegno == v->mdfd_segno + 1); @@ -1296,7 +1252,7 @@ _mdfd_getseg(SMgrRelation reln, ForkNumber forknum, BlockNumber blkno, { char *zerobuf = palloc0(BLCKSZ); - mdextend(reln, forknum, + mdextend(sfile, nextsegno * ((BlockNumber) RELSEG_SIZE) - 1, zerobuf, skipFsync); pfree(zerobuf); @@ -1327,11 +1283,11 @@ _mdfd_getseg(SMgrRelation reln, ForkNumber forknum, BlockNumber blkno, ereport(ERROR, (errcode_for_file_access(), errmsg("could not open file \"%s\" (target block %u): previous segment is only %u blocks", - _mdfd_segpath(reln, forknum, nextsegno), + _mdfd_segpath(sfile, nextsegno), blkno, nblocks))); } - v = _mdfd_openseg(reln, forknum, nextsegno, flags); + v = _mdfd_openseg(sfile, nextsegno, flags); if (v == NULL) { @@ -1341,7 +1297,7 @@ _mdfd_getseg(SMgrRelation reln, ForkNumber forknum, BlockNumber blkno, ereport(ERROR, (errcode_for_file_access(), errmsg("could not open file \"%s\" (target block %u): %m", - _mdfd_segpath(reln, forknum, nextsegno), + _mdfd_segpath(sfile, nextsegno), blkno))); } } @@ -1353,7 +1309,7 @@ _mdfd_getseg(SMgrRelation reln, ForkNumber forknum, BlockNumber blkno, * Get number of blocks present in a single disk file */ static BlockNumber -_mdnblocks(SMgrRelation reln, ForkNumber forknum, MdfdVec *seg) +_mdnblocks(SMgrFileHandle sfile, MdfdVec *seg) { off_t len; @@ -1376,16 +1332,16 @@ _mdnblocks(SMgrRelation reln, ForkNumber forknum, MdfdVec *seg) int mdsyncfiletag(const FileTag *ftag, char *path) { - SMgrRelation reln = smgropen(ftag->rlocator, InvalidBackendId); + SMgrFileHandle sfile = smgropen(ftag->rlocator, InvalidBackendId, ftag->forknum); File file; bool need_to_close; int result, save_errno; /* See if we already have the file open, or need to open it. */ - if (ftag->segno < reln->md_num_open_segs[ftag->forknum]) + if (ftag->segno < sfile->md_num_open_segs) { - file = reln->md_seg_fds[ftag->forknum][ftag->segno].mdfd_vfd; + file = sfile->md_seg_fds[ftag->segno].mdfd_vfd; strlcpy(path, FilePathName(file), MAXPGPATH); need_to_close = false; } @@ -1393,7 +1349,7 @@ mdsyncfiletag(const FileTag *ftag, char *path) { char *p; - p = _mdfd_segpath(reln, ftag->forknum, ftag->segno); + p = _mdfd_segpath(sfile, ftag->segno); strlcpy(path, p, MAXPGPATH); pfree(p); diff --git a/src/backend/storage/smgr/smgr.c b/src/backend/storage/smgr/smgr.c index 80eb6311e74..8329d7b983e 100644 --- a/src/backend/storage/smgr/smgr.c +++ b/src/backend/storage/smgr/smgr.c @@ -41,27 +41,24 @@ typedef struct f_smgr { void (*smgr_init) (void); /* may be NULL */ void (*smgr_shutdown) (void); /* may be NULL */ - void (*smgr_open) (SMgrRelation reln); - void (*smgr_close) (SMgrRelation reln, ForkNumber forknum); - void (*smgr_create) (SMgrRelation reln, ForkNumber forknum, - bool isRedo); - bool (*smgr_exists) (SMgrRelation reln, ForkNumber forknum); - void (*smgr_unlink) (RelFileLocatorBackend rlocator, ForkNumber forknum, - bool isRedo); - void (*smgr_extend) (SMgrRelation reln, ForkNumber forknum, + void (*smgr_open) (SMgrFileHandle sfile); + void (*smgr_close) (SMgrFileHandle sfile); + void (*smgr_create) (SMgrFileHandle sfile, bool isRedo); + bool (*smgr_exists) (SMgrFileHandle sfile); + void (*smgr_unlink) (SMgrFileLocator slocator, bool isRedo); + void (*smgr_extend) (SMgrFileHandle sfile, BlockNumber blocknum, char *buffer, bool skipFsync); - bool (*smgr_prefetch) (SMgrRelation reln, ForkNumber forknum, + bool (*smgr_prefetch) (SMgrFileHandle sfile, BlockNumber blocknum); - void (*smgr_read) (SMgrRelation reln, ForkNumber forknum, + void (*smgr_read) (SMgrFileHandle sfile, BlockNumber blocknum, char *buffer); - void (*smgr_write) (SMgrRelation reln, ForkNumber forknum, + void (*smgr_write) (SMgrFileHandle sfile, BlockNumber blocknum, char *buffer, bool skipFsync); - void (*smgr_writeback) (SMgrRelation reln, ForkNumber forknum, + void (*smgr_writeback) (SMgrFileHandle sfile, BlockNumber blocknum, BlockNumber nblocks); - BlockNumber (*smgr_nblocks) (SMgrRelation reln, ForkNumber forknum); - void (*smgr_truncate) (SMgrRelation reln, ForkNumber forknum, - BlockNumber nblocks); - void (*smgr_immedsync) (SMgrRelation reln, ForkNumber forknum); + BlockNumber (*smgr_nblocks) (SMgrFileHandle sfile); + void (*smgr_truncate) (SMgrFileHandle sfile, BlockNumber nblocks); + void (*smgr_immedsync) (SMgrFileHandle sfile); } f_smgr; static const f_smgr smgrsw[] = { @@ -88,12 +85,12 @@ static const f_smgr smgrsw[] = { static const int NSmgr = lengthof(smgrsw); /* - * Each backend has a hashtable that stores all extant SMgrRelation objects. - * In addition, "unowned" SMgrRelation objects are chained together in a list. + * Each backend has a hashtable that stores all extant SMgrFileData objects. + * In addition, "unowned" SMgrFile objects are chained together in a list. */ -static HTAB *SMgrRelationHash = NULL; +static HTAB *SMgrFileHash = NULL; -static dlist_head unowned_relns; +static dlist_head unowned_sfiles; /* local function prototypes */ static void smgrshutdown(int code, Datum arg); @@ -142,50 +139,50 @@ smgrshutdown(int code, Datum arg) * * This does not attempt to actually open the underlying file. */ -SMgrRelation -smgropen(RelFileLocator rlocator, BackendId backend) +SMgrFileHandle +smgropen(RelFileLocator rlocator, BackendId backend, ForkNumber forkNum) { - RelFileLocatorBackend brlocator; - SMgrRelation reln; + SMgrFileLocator slocator; + SMgrFileHandle sfile; bool found; - if (SMgrRelationHash == NULL) + if (SMgrFileHash == NULL) { /* First time through: initialize the hash table */ HASHCTL ctl; - ctl.keysize = sizeof(RelFileLocatorBackend); - ctl.entrysize = sizeof(SMgrRelationData); - SMgrRelationHash = hash_create("smgr relation table", 400, + ctl.keysize = sizeof(SMgrFileLocator); + ctl.entrysize = sizeof(SMgrFileData); + SMgrFileHash = hash_create("smgr relation table", 400, &ctl, HASH_ELEM | HASH_BLOBS); - dlist_init(&unowned_relns); + dlist_init(&unowned_sfiles); } /* Look up or create an entry */ - brlocator.locator = rlocator; - brlocator.backend = backend; - reln = (SMgrRelation) hash_search(SMgrRelationHash, - (void *) &brlocator, - HASH_ENTER, &found); + slocator.locator = rlocator; + slocator.backend = backend; + slocator.forknum = forkNum; + sfile = (SMgrFileHandle) hash_search(SMgrFileHash, + (void *) &slocator, + HASH_ENTER, &found); /* Initialize it if not present before */ if (!found) { /* hash_search already filled in the lookup key */ - reln->smgr_owner = NULL; - reln->smgr_targblock = InvalidBlockNumber; - for (int i = 0; i <= MAX_FORKNUM; ++i) - reln->smgr_cached_nblocks[i] = InvalidBlockNumber; - reln->smgr_which = 0; /* we only have md.c at present */ + sfile->smgr_owner = NULL; + sfile->smgr_targblock = InvalidBlockNumber; + sfile->smgr_cached_nblocks = InvalidBlockNumber; + sfile->smgr_which = 0; /* we only have md.c at present */ /* implementation-specific initialization */ - smgrsw[reln->smgr_which].smgr_open(reln); + smgrsw[sfile->smgr_which].smgr_open(sfile); /* it has no owner yet */ - dlist_push_tail(&unowned_relns, &reln->node); + dlist_push_tail(&unowned_sfiles, &sfile->node); } - return reln; + return sfile; } /* @@ -195,7 +192,7 @@ smgropen(RelFileLocator rlocator, BackendId backend) * the only such owners exist in the relcache. */ void -smgrsetowner(SMgrRelation *owner, SMgrRelation reln) +smgrsetowner(SMgrFileHandle *owner, SMgrFileHandle sfile) { /* We don't support "disowning" an SMgrRelation here, use smgrclearowner */ Assert(owner != NULL); @@ -206,68 +203,66 @@ smgrsetowner(SMgrRelation *owner, SMgrRelation reln) * depending on the order of processing. It's ok to close the old * relcache entry early in that case.) * - * If there isn't an old owner, then the reln should be in the unowned + * If there isn't an old owner, then the sfile should be in the unowned * list, and we need to remove it. */ - if (reln->smgr_owner) - *(reln->smgr_owner) = NULL; + if (sfile->smgr_owner) + *(sfile->smgr_owner) = NULL; else - dlist_delete(&reln->node); + dlist_delete(&sfile->node); /* Now establish the ownership relationship. */ - reln->smgr_owner = owner; - *owner = reln; + sfile->smgr_owner = owner; + *owner = sfile; } /* - * smgrclearowner() -- Remove long-lived reference to an SMgrRelation object + * smgrclearowner() -- Remove long-lived reference to an SMgrFileHandle object * if one exists */ void -smgrclearowner(SMgrRelation *owner, SMgrRelation reln) +smgrclearowner(SMgrFileHandle *owner, SMgrFileHandle sfile) { /* Do nothing if the SMgrRelation object is not owned by the owner */ - if (reln->smgr_owner != owner) + if (sfile->smgr_owner != owner) return; /* unset the owner's reference */ *owner = NULL; /* unset our reference to the owner */ - reln->smgr_owner = NULL; + sfile->smgr_owner = NULL; /* add to list of unowned relations */ - dlist_push_tail(&unowned_relns, &reln->node); + dlist_push_tail(&unowned_sfiles, &sfile->node); } /* - * smgrexists() -- Does the underlying file for a fork exist? + * smgrexists() -- Does the underlying file exist? */ bool -smgrexists(SMgrRelation reln, ForkNumber forknum) +smgrexists(SMgrFileHandle sfile) { - return smgrsw[reln->smgr_which].smgr_exists(reln, forknum); + return smgrsw[sfile->smgr_which].smgr_exists(sfile); } /* - * smgrclose() -- Close and delete an SMgrRelation object. + * smgrclose() -- Close and delete an SMgrFile object. */ void -smgrclose(SMgrRelation reln) +smgrclose(SMgrFileHandle sfile) { - SMgrRelation *owner; - ForkNumber forknum; + SMgrFileHandle *owner; - for (forknum = 0; forknum <= MAX_FORKNUM; forknum++) - smgrsw[reln->smgr_which].smgr_close(reln, forknum); + smgrsw[sfile->smgr_which].smgr_close(sfile); - owner = reln->smgr_owner; + owner = sfile->smgr_owner; if (!owner) - dlist_delete(&reln->node); + dlist_delete(&sfile->node); - if (hash_search(SMgrRelationHash, - (void *) &(reln->smgr_rlocator), + if (hash_search(SMgrFileHash, + (void *) &(sfile->smgr_locator), HASH_REMOVE, NULL) == NULL) elog(ERROR, "SMgrRelation hashtable corrupted"); @@ -284,14 +279,11 @@ smgrclose(SMgrRelation reln) * * The object remains valid. */ -void -smgrrelease(SMgrRelation reln) +static void +smgrrelease(SMgrFileHandle sfile) { - for (ForkNumber forknum = 0; forknum <= MAX_FORKNUM; forknum++) - { - smgrsw[reln->smgr_which].smgr_close(reln, forknum); - reln->smgr_cached_nblocks[forknum] = InvalidBlockNumber; - } + smgrsw[sfile->smgr_which].smgr_close(sfile); + sfile->smgr_cached_nblocks = InvalidBlockNumber; } /* @@ -303,16 +295,16 @@ void smgrreleaseall(void) { HASH_SEQ_STATUS status; - SMgrRelation reln; + SMgrFileHandle sfile; /* Nothing to do if hashtable not set up */ - if (SMgrRelationHash == NULL) + if (SMgrFileHash == NULL) return; - hash_seq_init(&status, SMgrRelationHash); + hash_seq_init(&status, SMgrFileHash); - while ((reln = (SMgrRelation) hash_seq_search(&status)) != NULL) - smgrrelease(reln); + while ((sfile = (SMgrFileHandle) hash_seq_search(&status)) != NULL) + smgrrelease(sfile); } /* @@ -322,16 +314,16 @@ void smgrcloseall(void) { HASH_SEQ_STATUS status; - SMgrRelation reln; + SMgrFileHandle sfile; /* Nothing to do if hashtable not set up */ - if (SMgrRelationHash == NULL) + if (SMgrFileHash == NULL) return; - hash_seq_init(&status, SMgrRelationHash); + hash_seq_init(&status, SMgrFileHash); - while ((reln = (SMgrRelation) hash_seq_search(&status)) != NULL) - smgrclose(reln); + while ((sfile = (SMgrFileHandle) hash_seq_search(&status)) != NULL) + smgrclose(sfile); } /* @@ -343,111 +335,112 @@ smgrcloseall(void) * such entry exists already. */ void -smgrcloserellocator(RelFileLocatorBackend rlocator) +smgrcloserellocator(RelFileLocator rlocator, BackendId backend) { - SMgrRelation reln; + SMgrFileHandle sfile; /* Nothing to do if hashtable not set up */ - if (SMgrRelationHash == NULL) + if (SMgrFileHash == NULL) return; - reln = (SMgrRelation) hash_search(SMgrRelationHash, - (void *) &rlocator, - HASH_FIND, NULL); - if (reln != NULL) - smgrclose(reln); + for (int i = 0; i <= MAX_FORKNUM; i++) + { + SMgrFileLocator slocator = { rlocator, backend, i }; + + sfile = (SMgrFileHandle) hash_search(SMgrFileHash, + (void *) &slocator, + HASH_FIND, NULL); + if (sfile != NULL) + smgrclose(sfile); + } } /* - * smgrcreate() -- Create a new relation. + * smgrcreate() -- Create a new file. * - * Given an already-created (but presumably unused) SMgrRelation, + * Given an already-created (but presumably unused) SMgrFileHandle, * cause the underlying disk file or other storage for the fork * to be created. */ void -smgrcreate(SMgrRelation reln, ForkNumber forknum, bool isRedo) +smgrcreate(SMgrFileHandle sfile, bool isRedo) { - smgrsw[reln->smgr_which].smgr_create(reln, forknum, isRedo); + smgrsw[sfile->smgr_which].smgr_create(sfile, isRedo); } /* - * smgrdosyncall() -- Immediately sync all forks of all given relations + * smgrunlink_multi() -- Immediately unlink given forks of given relation * - * All forks of all given relations are synced out to the store. + * The given forks of the relation are removed from the store. This + * should not be used during transactional operations, since it can't be + * undone. + * + * This handles multiple forks in one call, because the cache invalidation + * happens at relation granularity. If we had an smgrunlink() function + * to unlink just a single smgr file, and the caller wanted to delete + * multiple forks of a single relation, each call would send a new + * cache invalidation event, which would be wasteful. * - * This is equivalent to FlushRelationBuffers() for each smgr relation, - * then calling smgrimmedsync() for all forks of each relation, but it's - * significantly quicker so should be preferred when possible. + * If isRedo is true, it is okay for the underlying file(s) to be gone + * already. */ void -smgrdosyncall(SMgrRelation *rels, int nrels) +smgrunlink_multi(RelFileLocator rlocator, BackendId backend, ForkNumber *forks, int nforks, bool isRedo) { + int which; int i = 0; - ForkNumber forknum; - if (nrels == 0) - return; + which = 0; /* we only have md.c at present */ - FlushRelationsAllBuffers(rels, nrels); + /* Close the forks at smgr level */ + smgrcloserellocator(rlocator, backend); /* - * Sync the physical file(s). + * Send a shared-inval message to force other backends to close any + * dangling smgr references they may have for these rels. We should do + * this before starting the actual unlinking, in case we fail partway + * through that step. Note that the sinval messages will eventually come + * back to this backend, too, and thereby provide a backstop that we + * closed our own smgr rel. */ - for (i = 0; i < nrels; i++) - { - int which = rels[i]->smgr_which; + CacheInvalidateSmgr(rlocator, backend); - for (forknum = 0; forknum <= MAX_FORKNUM; forknum++) - { - if (smgrsw[which].smgr_exists(rels[i], forknum)) - smgrsw[which].smgr_immedsync(rels[i], forknum); - } + /* + * Delete the physical file(s). + * + * Note: smgr_unlink must treat deletion failure as a WARNING, not an + * ERROR, because we've already decided to commit or abort the current + * xact. + */ + for (i = 0; i < nforks; i++) + { + SMgrFileLocator slocator = { rlocator, backend, forks[i] }; + smgrsw[which].smgr_unlink(slocator, isRedo); } } /* - * smgrdounlinkall() -- Immediately unlink all forks of all given relations - * - * All forks of all given relations are removed from the store. This - * should not be used during transactional operations, since it can't be - * undone. + * smgrdounlink() -- Immediately unlink a file * * If isRedo is true, it is okay for the underlying file(s) to be gone * already. + * + * To remove a relation transactionally, see RelationDropStorage() instead. + * This will cause cache invalidation of all forks of the relation, not just + * this one. */ void -smgrdounlinkall(SMgrRelation *rels, int nrels, bool isRedo) +smgrunlink(SMgrFileHandle sfile, bool isRedo) { - int i = 0; - RelFileLocatorBackend *rlocators; - ForkNumber forknum; - - if (nrels == 0) - return; - - /* - * Get rid of any remaining buffers for the relations. bufmgr will just - * drop them without bothering to write the contents. - */ - DropRelationsAllBuffers(rels, nrels); - - /* - * create an array which contains all relations to be dropped, and close - * each relation's forks at the smgr level while at it - */ - rlocators = palloc(sizeof(RelFileLocatorBackend) * nrels); - for (i = 0; i < nrels; i++) - { - RelFileLocatorBackend rlocator = rels[i]->smgr_rlocator; - int which = rels[i]->smgr_which; + SMgrFileLocator locator; + int which; - rlocators[i] = rlocator; + /* remember before closing */ + which = sfile->smgr_which; + locator = sfile->smgr_locator; - /* Close the forks at smgr level */ - for (forknum = 0; forknum <= MAX_FORKNUM; forknum++) - smgrsw[which].smgr_close(rels[i], forknum); - } + /* Close the file at smgr level */ + smgrclose(sfile); /* * Send a shared-inval message to force other backends to close any @@ -457,8 +450,7 @@ smgrdounlinkall(SMgrRelation *rels, int nrels, bool isRedo) * back to this backend, too, and thereby provide a backstop that we * closed our own smgr rel. */ - for (i = 0; i < nrels; i++) - CacheInvalidateSmgr(rlocators[i]); + CacheInvalidateSmgr(locator.locator, locator.backend); /* * Delete the physical file(s). @@ -467,16 +459,7 @@ smgrdounlinkall(SMgrRelation *rels, int nrels, bool isRedo) * ERROR, because we've already decided to commit or abort the current * xact. */ - - for (i = 0; i < nrels; i++) - { - int which = rels[i]->smgr_which; - - for (forknum = 0; forknum <= MAX_FORKNUM; forknum++) - smgrsw[which].smgr_unlink(rlocators[i], forknum, isRedo); - } - - pfree(rlocators); + smgrsw[which].smgr_unlink(locator, isRedo); } @@ -490,21 +473,21 @@ smgrdounlinkall(SMgrRelation *rels, int nrels, bool isRedo) * causes intervening file space to become filled with zeroes. */ void -smgrextend(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, +smgrextend(SMgrFileHandle sfile, BlockNumber blocknum, char *buffer, bool skipFsync) { - smgrsw[reln->smgr_which].smgr_extend(reln, forknum, blocknum, - buffer, skipFsync); + smgrsw[sfile->smgr_which].smgr_extend(sfile, blocknum, + buffer, skipFsync); /* * Normally we expect this to increase nblocks by one, but if the cached * value isn't as expected, just invalidate it so the next call asks the * kernel. */ - if (reln->smgr_cached_nblocks[forknum] == blocknum) - reln->smgr_cached_nblocks[forknum] = blocknum + 1; + if (sfile->smgr_cached_nblocks == blocknum) + sfile->smgr_cached_nblocks = blocknum + 1; else - reln->smgr_cached_nblocks[forknum] = InvalidBlockNumber; + sfile->smgr_cached_nblocks = InvalidBlockNumber; } /* @@ -515,13 +498,13 @@ smgrextend(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, * record). */ bool -smgrprefetch(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum) +smgrprefetch(SMgrFileHandle sfile, BlockNumber blocknum) { - return smgrsw[reln->smgr_which].smgr_prefetch(reln, forknum, blocknum); + return smgrsw[sfile->smgr_which].smgr_prefetch(sfile, blocknum); } /* - * smgrread() -- read a particular block from a relation into the supplied + * smgrread() -- read a particular block from a file into the supplied * buffer. * * This routine is called from the buffer manager in order to @@ -529,10 +512,9 @@ smgrprefetch(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum) * return pages in the format that POSTGRES expects. */ void -smgrread(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, - char *buffer) +smgrread(SMgrFileHandle sfile, BlockNumber blocknum, char *buffer) { - smgrsw[reln->smgr_which].smgr_read(reln, forknum, blocknum, buffer); + smgrsw[sfile->smgr_which].smgr_read(sfile, blocknum, buffer); } /* @@ -551,11 +533,11 @@ smgrread(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, * do not require fsync. */ void -smgrwrite(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, +smgrwrite(SMgrFileHandle sfile, BlockNumber blocknum, char *buffer, bool skipFsync) { - smgrsw[reln->smgr_which].smgr_write(reln, forknum, blocknum, - buffer, skipFsync); + smgrsw[sfile->smgr_which].smgr_write(sfile, blocknum, + buffer, skipFsync); } @@ -564,11 +546,11 @@ smgrwrite(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, * blocks. */ void -smgrwriteback(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, +smgrwriteback(SMgrFileHandle sfile, BlockNumber blocknum, BlockNumber nblocks) { - smgrsw[reln->smgr_which].smgr_writeback(reln, forknum, blocknum, - nblocks); + smgrsw[sfile->smgr_which].smgr_writeback(sfile, blocknum, + nblocks); } /* @@ -576,18 +558,18 @@ smgrwriteback(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, * supplied relation. */ BlockNumber -smgrnblocks(SMgrRelation reln, ForkNumber forknum) +smgrnblocks(SMgrFileHandle sfile) { BlockNumber result; /* Check and return if we get the cached value for the number of blocks. */ - result = smgrnblocks_cached(reln, forknum); + result = smgrnblocks_cached(sfile); if (result != InvalidBlockNumber) return result; - result = smgrsw[reln->smgr_which].smgr_nblocks(reln, forknum); + result = smgrsw[sfile->smgr_which].smgr_nblocks(sfile); - reln->smgr_cached_nblocks[forknum] = result; + sfile->smgr_cached_nblocks = result; return result; } @@ -600,38 +582,41 @@ smgrnblocks(SMgrRelation reln, ForkNumber forknum) * fork size is not cached. */ BlockNumber -smgrnblocks_cached(SMgrRelation reln, ForkNumber forknum) +smgrnblocks_cached(SMgrFileHandle sfile) { /* * For now, we only use cached values in recovery due to lack of a shared * invalidation mechanism for changes in file size. */ - if (InRecovery && reln->smgr_cached_nblocks[forknum] != InvalidBlockNumber) - return reln->smgr_cached_nblocks[forknum]; + if (InRecovery && sfile->smgr_cached_nblocks != InvalidBlockNumber) + return sfile->smgr_cached_nblocks; return InvalidBlockNumber; } /* - * smgrtruncate() -- Truncate the given forks of supplied relation to - * each specified numbers of blocks + * smgrtruncate_multi() -- Truncate the given forks of supplied relation to + * each specified numbers of blocks * * The truncation is done immediately, so this can't be rolled back. * * The caller must hold AccessExclusiveLock on the relation, to ensure that * other backends receive the smgr invalidation event that this function sends * before they access any forks of the relation again. + * + * Like smgrunlink_multi(), this handles multiple forks in one call because + * the cache invalidation happens at relation granularity. + * + * NB: The caller is responsible for dropping buffers! Before v16, this + * function did it. */ void -smgrtruncate(SMgrRelation reln, ForkNumber *forknum, int nforks, BlockNumber *nblocks) +smgrtruncate_multi(RelFileLocator rlocator, BackendId backend, ForkNumber *forks, + int nforks, BlockNumber *nblocks) { int i; - /* - * Get rid of any buffers for the about-to-be-deleted blocks. bufmgr will - * just drop them without bothering to write the contents. - */ - DropRelationBuffers(reln, forknum, nforks, nblocks); + Assert(nforks < MAX_FORKNUM + 1); /* * Send a shared-inval message to force other backends to close any smgr @@ -643,15 +628,19 @@ smgrtruncate(SMgrRelation reln, ForkNumber *forknum, int nforks, BlockNumber *nb * is a performance-critical path.) As in the unlink code, we want to be * sure the message is sent before we start changing things on-disk. */ - CacheInvalidateSmgr(reln->smgr_rlocator); + CacheInvalidateSmgr(rlocator, backend); - /* Do the truncation */ + /* Do the truncations */ for (i = 0; i < nforks; i++) { + SMgrFileHandle sfile; + + sfile = smgropen(rlocator, backend, forks[i]); + /* Make the cached size is invalid if we encounter an error. */ - reln->smgr_cached_nblocks[forknum[i]] = InvalidBlockNumber; + sfile->smgr_cached_nblocks = InvalidBlockNumber; - smgrsw[reln->smgr_which].smgr_truncate(reln, forknum[i], nblocks[i]); + smgrsw[sfile->smgr_which].smgr_truncate(sfile, nblocks[i]); /* * We might as well update the local smgr_cached_nblocks values. The @@ -660,7 +649,7 @@ smgrtruncate(SMgrRelation reln, ForkNumber *forknum, int nforks, BlockNumber *nb * smgr_vm_nblocks, and these ones too at the next command boundary. * But these ensure they aren't outright wrong until then. */ - reln->smgr_cached_nblocks[forknum[i]] = nblocks[i]; + sfile->smgr_cached_nblocks = nblocks[i]; } } @@ -688,9 +677,9 @@ smgrtruncate(SMgrRelation reln, ForkNumber *forknum, int nforks, BlockNumber *nb * otherwise the sync is not very meaningful. */ void -smgrimmedsync(SMgrRelation reln, ForkNumber forknum) +smgrimmedsync(SMgrFileHandle sfile) { - smgrsw[reln->smgr_which].smgr_immedsync(reln, forknum); + smgrsw[sfile->smgr_which].smgr_immedsync(sfile); } /* @@ -714,14 +703,14 @@ AtEOXact_SMgr(void) * Zap all unowned SMgrRelations. We rely on smgrclose() to remove each * one from the list. */ - dlist_foreach_modify(iter, &unowned_relns) + dlist_foreach_modify(iter, &unowned_sfiles) { - SMgrRelation rel = dlist_container(SMgrRelationData, node, - iter.cur); + SMgrFileHandle sfile = dlist_container(SMgrFileData, node, + iter.cur); - Assert(rel->smgr_owner == NULL); + Assert(sfile->smgr_owner == NULL); - smgrclose(rel); + smgrclose(sfile); } } diff --git a/src/backend/storage/sync/sync.c b/src/backend/storage/sync/sync.c index 7135150b167..2ed70576482 100644 --- a/src/backend/storage/sync/sync.c +++ b/src/backend/storage/sync/sync.c @@ -18,9 +18,7 @@ #include <fcntl.h> #include <sys/file.h> -#include "access/commit_ts.h" -#include "access/clog.h" -#include "access/multixact.h" +#include "access/slru.h" #include "access/xlog.h" #include "access/xlogutils.h" #include "commands/tablespace.h" @@ -106,22 +104,6 @@ static const SyncOps syncsw[] = { .sync_unlinkfiletag = mdunlinkfiletag, .sync_filetagmatches = mdfiletagmatches }, - /* pg_xact */ - [SYNC_HANDLER_CLOG] = { - .sync_syncfiletag = clogsyncfiletag - }, - /* pg_commit_ts */ - [SYNC_HANDLER_COMMIT_TS] = { - .sync_syncfiletag = committssyncfiletag - }, - /* pg_multixact/offsets */ - [SYNC_HANDLER_MULTIXACT_OFFSET] = { - .sync_syncfiletag = multixactoffsetssyncfiletag - }, - /* pg_multixact/members */ - [SYNC_HANDLER_MULTIXACT_MEMBER] = { - .sync_syncfiletag = multixactmemberssyncfiletag - } }; /* diff --git a/src/backend/utils/cache/inval.c b/src/backend/utils/cache/inval.c index 0008826f67c..7a1d8c012c2 100644 --- a/src/backend/utils/cache/inval.c +++ b/src/backend/utils/cache/inval.c @@ -661,11 +661,12 @@ LocalExecuteInvalidationMessage(SharedInvalidationMessage *msg) * We could have smgr entries for relations of other databases, so no * short-circuit test is possible here. */ - RelFileLocatorBackend rlocator; + RelFileLocator rlocator; + BackendId backend; - rlocator.locator = msg->sm.rlocator; - rlocator.backend = (msg->sm.backend_hi << 16) | (int) msg->sm.backend_lo; - smgrcloserellocator(rlocator); + rlocator = msg->sm.rlocator; + backend = (msg->sm.backend_hi << 16) | (int) msg->sm.backend_lo; + smgrcloserellocator(rlocator, backend); } else if (msg->id == SHAREDINVALRELMAP_ID) { @@ -1459,14 +1460,14 @@ CacheInvalidateRelcacheByRelid(Oid relid) * Thus, the maximum possible backend ID is 2^23-1. */ void -CacheInvalidateSmgr(RelFileLocatorBackend rlocator) +CacheInvalidateSmgr(RelFileLocator rlocator, BackendId backend) { SharedInvalidationMessage msg; msg.sm.id = SHAREDINVALSMGR_ID; - msg.sm.backend_hi = rlocator.backend >> 16; - msg.sm.backend_lo = rlocator.backend & 0xffff; - msg.sm.rlocator = rlocator.locator; + msg.sm.backend_hi = backend >> 16; + msg.sm.backend_lo = backend & 0xffff; + msg.sm.rlocator = rlocator; /* check AddCatcacheInvalidationMessage() for an explanation */ VALGRIND_MAKE_MEM_DEFINED(&msg, sizeof(msg)); diff --git a/src/backend/utils/cache/relcache.c b/src/backend/utils/cache/relcache.c index d171cfcf2fc..a193f515c66 100644 --- a/src/backend/utils/cache/relcache.c +++ b/src/backend/utils/cache/relcache.c @@ -419,7 +419,7 @@ AllocateRelationDesc(Form_pg_class relp) relation = (Relation) palloc0(sizeof(RelationData)); /* make sure relation is marked as having no open file yet */ - relation->rd_smgr = NULL; + MemSet(relation->rd_smgr, 0, sizeof(relation->rd_smgr)); /* * Copy the relation tuple form @@ -1248,7 +1248,7 @@ retry: RelationInitPhysicalAddr(relation); /* make sure relation is marked as having no open file yet */ - relation->rd_smgr = NULL; + MemSet(relation->rd_smgr, 0, sizeof(relation->rd_smgr)); /* * now we can free the memory allocated for pg_class_tuple @@ -1877,7 +1877,7 @@ formrdesc(const char *relationName, Oid relationReltype, relation = (Relation) palloc0(sizeof(RelationData)); /* make sure relation is marked as having no open file yet */ - relation->rd_smgr = NULL; + MemSet(relation->rd_smgr, 0, sizeof(relation->rd_smgr)); /* * initialize reference count: 1 because it is nailed in cache @@ -2701,7 +2701,8 @@ RelationClearRelation(Relation relation, bool rebuild) } /* rd_smgr must not be swapped, due to back-links from smgr level */ - SWAPFIELD(SMgrRelation, rd_smgr); + for (int i = 0; i <= MAX_FORKNUM; i++) + SWAPFIELD(SMgrFileHandle, rd_smgr[i]); /* rd_refcnt must be preserved */ SWAPFIELD(int, rd_refcnt); /* isnailed shouldn't change */ @@ -3532,7 +3533,7 @@ RelationBuildLocalRelation(const char *relname, rel = (Relation) palloc0(sizeof(RelationData)); /* make sure relation is marked as having no open file yet */ - rel->rd_smgr = NULL; + MemSet(rel->rd_smgr, 0, sizeof(rel->rd_smgr)); /* mark it nailed if appropriate */ rel->rd_isnailed = nailit; @@ -3764,7 +3765,6 @@ RelationSetNewRelfilenumber(Relation relation, char persistence) */ if (IsBinaryUpgrade) { - SMgrRelation srel; /* * During a binary upgrade, we use this code path to ensure that @@ -3781,9 +3781,14 @@ RelationSetNewRelfilenumber(Relation relation, char persistence) * fails at this stage, the new cluster will need to be recreated * anyway. */ - srel = smgropen(relation->rd_locator, relation->rd_backend); - smgrdounlinkall(&srel, 1, false); - smgrclose(srel); + ForkNumber forks[MAX_FORKNUM + 1]; + for (int i = 0; i <= MAX_FORKNUM; i ++) + { + smgropen(relation->rd_locator, relation->rd_backend, i); + forks[i] = i; + } + + smgrunlink_multi(relation->rd_locator, relation->rd_backend, forks, MAX_FORKNUM + 1, false); } else { @@ -3811,7 +3816,7 @@ RelationSetNewRelfilenumber(Relation relation, char persistence) else if (RELKIND_HAS_STORAGE(relation->rd_rel->relkind)) { /* handle these directly, at least for now */ - SMgrRelation srel; + SMgrFileHandle srel; srel = RelationCreateStorage(newrlocator, persistence, true); smgrclose(srel); @@ -6298,7 +6303,7 @@ load_relcache_init_file(bool shared) /* * Reset transient-state fields in the relcache entry */ - rel->rd_smgr = NULL; + MemSet(rel->rd_smgr, 0, sizeof(rel->rd_smgr)); if (rel->rd_isnailed) rel->rd_refcnt = 1; else diff --git a/src/backend/utils/mmgr/mcxt.c b/src/backend/utils/mmgr/mcxt.c index 8b6591abfb2..7e2635b124d 100644 --- a/src/backend/utils/mmgr/mcxt.c +++ b/src/backend/utils/mmgr/mcxt.c @@ -162,9 +162,20 @@ static void MemoryContextStatsPrint(MemoryContext context, void *passthru, * You should not do memory allocations within a critical section, because * an out-of-memory error will be escalated to a PANIC. To enforce that * rule, the allocation functions Assert that. + * + * FIXME: bypass this for the critical section in RecordTransactionCommit() + * for now. It does a lot of things that can allocate: + * - calls TransactionIdCommitTree, which pins buffers, which requires + * space in the ResourceOwner for the pin (ResourceOwnerEnlargeBuffers()) + * - same for TransactionTreeSetCommitTsData() call. + * - reading a page can require flushing other pages, which in turn + * can call CompactCheckpointerRequestQueue(), which allocates + * - reading a page calls smgropen(), which allocates the SMgrFile entry + * if it's not open already */ #define AssertNotInCriticalSection(context) \ - Assert(CritSectionCount == 0 || (context)->allowInCritSection) + Assert(CritSectionCount == 0 || (context)->allowInCritSection || \ + (MyProc != NULL && (MyProc->delayChkptFlags & DELAY_CHKPT_START != 0))) /* * Call the given function in the MemoryContextMethods for the memory context diff --git a/src/bin/pg_upgrade/file.c b/src/bin/pg_upgrade/file.c index ed874507ff4..27f240470ba 100644 --- a/src/bin/pg_upgrade/file.c +++ b/src/bin/pg_upgrade/file.c @@ -20,6 +20,7 @@ #endif #include "access/visibilitymapdefs.h" +#include "access/slrudefs.h" #include "common/file_perm.h" #include "pg_upgrade.h" #include "storage/bufpage.h" @@ -139,7 +140,6 @@ copyFile(const char *src, const char *dst, #endif /* WIN32 */ } - /* * linkFile() * @@ -316,6 +316,179 @@ rewriteVisibilityMap(const char *fromfile, const char *tofile, close(src_fd); } + +//create new file and initialize to 256K +static int +init_new_slru_file(FILE * fp) +{ + int fd; + int res; + + fd = fileno(fp); + res = ftruncate(fd, BLCKSZ * SLRU_PAGES_PER_SEGMENT); + return res; +} + +int +copy_to_new_format(const char *old_subdir, const char *new_subdir, int element_size) +{ + DIR *dr; + struct dirent *de; + + size_t read_items; + size_t n_items; + size_t to_read_items; + size_t write_items; + size_t pages_written; + size_t write_offset; + + char * read_file; + char * write_file; + + int res; + int write_file_segno; + int total_read_files; + + FILE * old_fd; + FILE * new_fp; + +#define MAXBUFSIZE (BLCKSZ - SizeOfPageHeaderData) + + char old_path[MAXPGPATH]; + char new_path[MAXPGPATH]; + PGAlignedBlock pg_buf; + BlockNumber blkno; //track aboslute block number + + + struct dirent** all_dirents; + + + snprintf(old_path, sizeof(old_path), "%s/%s", old_cluster.pgdata, old_subdir); + snprintf(new_path, sizeof(new_path), "%s/%s", new_cluster.pgdata, new_subdir); + + /* + * calculate the number of items that can fit + * inside BLCKSZ - SizeOfPageHeaderData + */ + + n_items = (size_t) (MAXBUFSIZE / element_size); + + /* number of items to read in at a time */ + to_read_items = n_items; + + //bufsize = (size_t) (n_items * element_size); + write_file_segno = 0; /* track which segment we are writing into */ + write_offset = (size_t) (SizeOfPageHeaderData); + + dr = opendir((char *)old_path); + if (dr == NULL) + return -1; + + write_file = psprintf("%s/%04X", new_path, write_file_segno); + + new_fp = fopen(write_file, "wb"); + + if (new_fp == NULL) + return -1; + + res = init_new_slru_file(new_fp); + if (res < 0) + return -1; + + if (res < 0) + return -1; + + pages_written = 0; + blkno = 0; /* absolute blkno */ + + all_dirents = get_sorted_hex_files(old_path, &total_read_files); + + write_items = 0; + for (int i = 0; i < total_read_files; i++) + { + de = all_dirents[i]; + + if ((!strcmp((char *) de->d_name, ".")) || !strcmp((char *) de->d_name, "..")) + { + continue; + } + + read_file = psprintf("%s/%s", (char *) old_path, (char *) de->d_name); + + old_fd = fopen(read_file, "rb"); + read_items = 0; + + do + { + memset(pg_buf.data, 0, BLCKSZ); + read_items = fread(pg_buf.data + write_offset, element_size, to_read_items, old_fd); /* how many items we read */ + + if (to_read_items == n_items) /* not finishing up a partial read */ + { + + ((PageHeader) pg_buf.data)->pd_lower = SizeOfPageHeaderData; + ((PageHeader) pg_buf.data)->pd_upper = BLCKSZ; + ((PageHeader) pg_buf.data)->pd_special = BLCKSZ; + + ((PageHeader) pg_buf.data)->pd_checksum = + pg_checksum_page(pg_buf.data, blkno); + + fwrite(pg_buf.data, SizeOfPageHeaderData, 1, new_fp); /* write page header data */ + } + + + write_items += fwrite(pg_buf.data + write_offset, element_size, read_items, new_fp); /* increment how many items written */ + + if (write_items % n_items == 0 && write_items > 0 && errno == 0) /* finished writing into new page. */ + { + if (write_items == n_items * SLRU_PAGES_PER_SEGMENT) /* end of segment */ + { + fclose(new_fp); + pages_written = 0; + write_file_segno ++; + write_items = 0; + write_file = psprintf("%s/%04X", new_path, write_file_segno); + new_fp = fopen((char *)write_file, "wb"); + res = init_new_slru_file(new_fp); + if (res < 0 || new_fp == NULL) + { + fclose(new_fp); + cleanup_dirents(all_dirents, total_read_files); + + } + } else { /* end of page but not segment */ + //fflush(new_fp); + pages_written ++; + } + + blkno++; + + to_read_items = n_items; /* to_read_items = n */ + res = fseek(new_fp, (pages_written * BLCKSZ), SEEK_SET); + if (res < 0) + { + fclose(new_fp); + cleanup_dirents(all_dirents, total_read_files); + return -1; + } + } else { + /*end of segment in source dir*/ + if (read_items < to_read_items) + { + to_read_items = n_items - read_items; /* read remaining items */ + } + } + } while (!feof(old_fd)); /* until end of file */ + fclose(old_fd); + } + fflush(new_fp); + + /* free memory malloc'd by scandir while sorting */ + cleanup_dirents(all_dirents, total_read_files); + + return 0; +} + void check_file_clone(void) { diff --git a/src/bin/pg_upgrade/function.c b/src/bin/pg_upgrade/function.c index 9ad75446eec..39b36a1e0ed 100644 --- a/src/bin/pg_upgrade/function.c +++ b/src/bin/pg_upgrade/function.c @@ -7,6 +7,7 @@ * src/bin/pg_upgrade/function.c */ + #include "postgres_fe.h" #include "access/transam.h" @@ -42,6 +43,43 @@ library_name_compare(const void *p1, const void *p2) ((const LibraryInfo *) p2)->dbnum; } +/* + * qsort comparator for hex filenames + */ +static int +file_name_compare(const struct dirent ** de_1, const struct dirent ** de_2) +{ + int n1; + int n2; + + + char * fname_1; + char * fname_2; + + fname_1 = (char *) (*de_1)->d_name; + fname_2 = (char *) (*de_2)->d_name; + + if ((strcmp(fname_1, ".")) || strcmp(fname_1, "..")) + { + n1 = (int) strtol(fname_1, NULL, 16); + } else { + n1 = -1; + } + + if ((strcmp(fname_2, ".")) || strcmp(fname_2, "..")) + { + n2 = (int) strtol(fname_2, NULL, 16); + } else { + n2 = -1; + } + + if (n1 == n2) + { + return 1; //arbitrarily select the first input + } else { + return n1 - n2; + } +} /* * get_loadable_libraries() @@ -109,6 +147,34 @@ get_loadable_libraries(void) } +/* + * get_sorted_hex_files() + * given the filepath of a directory, + * return array of child dirents with hex filenames e.g '000A' + * in sorted order + */ +struct dirent** +get_sorted_hex_files(char * dr, int * size) +{ + struct dirent **entry_list; + + *size = scandir(dr, &entry_list, NULL, file_name_compare); + if (*size < 0) + { + return NULL; //error + } + return entry_list; +} + +void +cleanup_dirents(struct dirent ** all_dirents, int total_read_files) +{ + for (int i = 0; i < total_read_files; i++) + { + free((struct dirent **) all_dirents[i]); + } + free((struct dirent **) all_dirents); +} /* * check_loadable_libraries() * diff --git a/src/bin/pg_upgrade/pg_upgrade.c b/src/bin/pg_upgrade/pg_upgrade.c index e5597d31052..8c264cc98dd 100644 --- a/src/bin/pg_upgrade/pg_upgrade.c +++ b/src/bin/pg_upgrade/pg_upgrade.c @@ -38,27 +38,35 @@ #include "postgres_fe.h" +#include <dirent.h> #include <time.h> +#include <unistd.h> #ifdef HAVE_LANGINFO_H #include <langinfo.h> #endif +#include "access/slrudefs.h" #include "catalog/pg_class_d.h" #include "common/file_perm.h" #include "common/logging.h" #include "common/restricted_token.h" #include "fe_utils/string_utils.h" #include "pg_upgrade.h" +#include "storage/bufpage.h" static void prepare_new_cluster(void); static void prepare_new_globals(void); static void create_new_objects(void); static void copy_xact_xlog_xid(void); -static void set_frozenxids(bool minmxid_only); +static void set_frozenxids(bool minmxid_only);; static void make_outputdirs(char *pgdata); static void setup(char *argv0, bool *live_check); + +#define MAXBUFSIZE (BLCKSZ - SizeOfPageHeaderData) +#define SLRU_PAGES_PER_SEGMENT 32 + ClusterInfo old_cluster, new_cluster; OSInfo os_info; @@ -573,11 +581,36 @@ copy_xact_xlog_xid(void) * Copy old commit logs to new data dir. pg_clog has been renamed to * pg_xact in post-10 clusters. */ - copy_subdir_files(GET_MAJOR_VERSION(old_cluster.major_version) <= 906 ? - "pg_clog" : "pg_xact", - GET_MAJOR_VERSION(new_cluster.major_version) <= 906 ? - "pg_clog" : "pg_xact"); + + if (old_cluster.controldata.cat_ver <= CLOG_FORMATCHANGE_CAT_VER && + new_cluster.controldata.cat_ver >= CLOG_FORMATCHANGE_CAT_VER) + { + int ret; + ret = copy_to_new_format(GET_MAJOR_VERSION(old_cluster.major_version) <= 906 ? + "pg_clog" : "pg_xact", + "pg_xact", 1); + + if (ret < 0) + { + pg_fatal("could not reformat clog files"); + } + } + if (old_cluster.controldata.cat_ver >= CLOG_FORMATCHANGE_CAT_VER && + new_cluster.controldata.cat_ver >= CLOG_FORMATCHANGE_CAT_VER) + { + copy_subdir_files("pg_xact", "pg_xact"); + } + + if (old_cluster.controldata.cat_ver <= CLOG_FORMATCHANGE_CAT_VER && + new_cluster.controldata.cat_ver <= CLOG_FORMATCHANGE_CAT_VER) + { + copy_subdir_files(GET_MAJOR_VERSION(old_cluster.major_version) <= 906 ? + "pg_clog" : "pg_xact", + GET_MAJOR_VERSION(new_cluster.major_version) <= 906 ? + "pg_clog" : "pg_xact"); + } + prep_status("Setting oldest XID for new cluster"); exec_prog(UTILITY_LOG_FILE, NULL, true, true, "\"%s/pg_resetwal\" -f -u %u \"%s\"", @@ -633,23 +666,44 @@ copy_xact_xlog_xid(void) } else if (new_cluster.controldata.cat_ver >= MULTIXACT_FORMATCHANGE_CAT_VER) { + copy_to_new_format("pg_multixact/offsets", "pg_multixact/offsets", MULTIXACT_MEMBER_ENTRY_SIZE); + copy_to_new_format("pg_multixact/members", "pg_multixact/members", MULTIXACT_OFFSET_ENTRY_SIZE); + + + prep_status("Setting next multixact ID and offset for new cluster"); + /* + * we preserve all files and contents, so we must preserve both "next" + * counters here and the oldest multi present on system. + */ + exec_prog(UTILITY_LOG_FILE, NULL, true, true, + "\"%s/pg_resetwal\" -O %u -m %u,%u \"%s\"", + new_cluster.bindir, + old_cluster.controldata.chkpnt_nxtmxoff, + old_cluster.controldata.chkpnt_nxtmulti, + old_cluster.controldata.chkpnt_oldstMulti, + new_cluster.pgdata); + check_ok(); + + } + /*else if (new_cluster.controldata.cat_ver >= MULTIXACT_FORMATCHANGE_CAT_VER) + { + * * Remove offsets/0000 file created by initdb that no longer matches * the new multi-xid value. "members" starts at zero so no need to * remove it. - */ + * remove_new_subdir("pg_multixact/offsets", false); prep_status("Setting oldest multixact ID in new cluster"); - - /* + * * We don't preserve files in this case, but it's important that the * oldest multi is set to the latest value used by the old system, so * that multixact.c returns the empty set for multis that might be * present on disk. We set next multi to the value following that; it * might end up wrapped around (i.e. 0) if the old cluster had * next=MaxMultiXactId, but multixact.c can cope with that just fine. - */ + * exec_prog(UTILITY_LOG_FILE, NULL, true, true, "\"%s/pg_resetwal\" -m %u,%u \"%s\"", new_cluster.bindir, @@ -657,7 +711,8 @@ copy_xact_xlog_xid(void) old_cluster.controldata.chkpnt_nxtmulti, new_cluster.pgdata); check_ok(); - } + + }*/ /* now reset the wal archives in the new cluster */ prep_status("Resetting WAL archives"); @@ -669,7 +724,6 @@ copy_xact_xlog_xid(void) check_ok(); } - /* * set_frozenxids() * diff --git a/src/bin/pg_upgrade/pg_upgrade.h b/src/bin/pg_upgrade/pg_upgrade.h index 5f2a116f23e..19e16edcc55 100644 --- a/src/bin/pg_upgrade/pg_upgrade.h +++ b/src/bin/pg_upgrade/pg_upgrade.h @@ -7,6 +7,7 @@ #include <unistd.h> #include <assert.h> +#include <dirent.h> #include <sys/stat.h> #include <sys/time.h> @@ -113,7 +114,16 @@ extern char *output_files[]; * version to this value. pg_upgrade behavior depends on whether old and new * server versions are both newer than this, or only the new one is. */ -#define MULTIXACT_FORMATCHANGE_CAT_VER 201301231 + +/* + * page header format change + */ +#define MULTIXACT_FORMATCHANGE_CAT_VER 202209141 + +/* + * page header format change + */ +#define CLOG_FORMATCHANGE_CAT_VER 202210141 /* * large object chunk size added to pg_controldata, @@ -122,7 +132,7 @@ extern char *output_files[]; #define LARGE_OBJECT_SIZE_PG_CONTROL_VER 942 /* - * change in JSONB format during 9.4 beta + * addition of page header */ #define JSONB_FORMAT_CHANGE_CAT_VER 201409291 @@ -378,6 +388,9 @@ void rewriteVisibilityMap(const char *fromfile, const char *tofile, void check_file_clone(void); void check_hard_link(void); + +int copy_to_new_format(const char *old_subdir, const char *new_subdir, int element_size); + /* fopen_priv() is no longer different from fopen() */ #define fopen_priv(path, mode) fopen(path, mode) @@ -385,6 +398,8 @@ void check_hard_link(void); void get_loadable_libraries(void); void check_loadable_libraries(void); +void cleanup_dirents(struct dirent ** all_dirents, int total_read_files); +struct dirent** get_sorted_hex_files(char * dr, int * size); /* info.c */ diff --git a/src/common/relpath.c b/src/common/relpath.c index 87de5f6c960..4715d06d287 100644 --- a/src/common/relpath.c +++ b/src/common/relpath.c @@ -22,6 +22,16 @@ #include "common/relpath.h" #include "storage/backendid.h" +/* + * SLRU ID to path mapping + */ +#define PG_SLRU(symname,name,path,synchronize) \ + path, + +static char *slru_dirs[] = +{ +#include "access/slrulist.h" +}; /* * Lookup table of fork name by fork number. @@ -129,7 +139,7 @@ GetDatabasePath(Oid dbOid, Oid spcOid) } /* - * GetRelationPath - construct path to a relation's file + * GetSMgrFilePath - construct path to a relation's file * * Result is a palloc'd string. * @@ -138,12 +148,27 @@ GetDatabasePath(Oid dbOid, Oid spcOid) * the trouble considering BackendId is just int anyway. */ char * -GetRelationPath(Oid dbOid, Oid spcOid, RelFileNumber relNumber, +GetSMgrFilePath(Oid dbOid, Oid spcOid, RelFileNumber relNumber, int backendId, ForkNumber forkNumber) { char *path; - if (spcOid == GLOBALTABLESPACE_OID) + if (spcOid == SLRU_SPC_OID) + { + if (dbOid >= lengthof(slru_dirs) || forkNumber != 0 || backendId != InvalidBackendId) + { +#ifndef FRONTEND + ereport(ERROR, + (errcode(ERRCODE_INVALID_PARAMETER_VALUE), + errmsg("invalid SLRU file locator %u/%u/%u/%u/%u", + spcOid, dbOid, relNumber, backendId, forkNumber))); +#else + return NULL; +#endif + } + path = psprintf("%s/%04X", slru_dirs[dbOid], relNumber); + } + else if (spcOid == GLOBALTABLESPACE_OID) { /* Shared system relations live in {datadir}/global */ Assert(dbOid == 0); diff --git a/src/include/access/clog.h b/src/include/access/clog.h index d99444f073f..aacf10ca522 100644 --- a/src/include/access/clog.h +++ b/src/include/access/clog.h @@ -40,18 +40,12 @@ extern void TransactionIdSetTreeStatus(TransactionId xid, int nsubxids, TransactionId *subxids, XidStatus status, XLogRecPtr lsn); extern XidStatus TransactionIdGetStatus(TransactionId xid, XLogRecPtr *lsn); -extern Size CLOGShmemBuffers(void); -extern Size CLOGShmemSize(void); -extern void CLOGShmemInit(void); extern void BootStrapCLOG(void); extern void StartupCLOG(void); extern void TrimCLOG(void); -extern void CheckPointCLOG(void); extern void ExtendCLOG(TransactionId newestXact); extern void TruncateCLOG(TransactionId oldestXact, Oid oldestxid_datoid); -extern int clogsyncfiletag(const FileTag *ftag, char *path); - /* XLOG stuff */ #define CLOG_ZEROPAGE 0x00 #define CLOG_TRUNCATE 0x10 diff --git a/src/include/access/commit_ts.h b/src/include/access/commit_ts.h index 5087cdce51e..605dc75b3b0 100644 --- a/src/include/access/commit_ts.h +++ b/src/include/access/commit_ts.h @@ -27,7 +27,6 @@ extern bool TransactionIdGetCommitTsData(TransactionId xid, extern TransactionId GetLatestCommitTsData(TimestampTz *ts, RepOriginId *nodeid); -extern Size CommitTsShmemBuffers(void); extern Size CommitTsShmemSize(void); extern void CommitTsShmemInit(void); extern void BootStrapCommitTs(void); @@ -41,8 +40,6 @@ extern void SetCommitTsLimit(TransactionId oldestXact, TransactionId newestXact); extern void AdvanceOldestCommitTsXid(TransactionId oldestXact); -extern int committssyncfiletag(const FileTag *ftag, char *path); - /* XLOG stuff */ #define COMMIT_TS_ZEROPAGE 0x00 #define COMMIT_TS_TRUNCATE 0x10 diff --git a/src/include/access/multixact.h b/src/include/access/multixact.h index 246f757f6ab..5848e4072ba 100644 --- a/src/include/access/multixact.h +++ b/src/include/access/multixact.h @@ -118,9 +118,6 @@ extern bool MultiXactIdPrecedes(MultiXactId multi1, MultiXactId multi2); extern bool MultiXactIdPrecedesOrEquals(MultiXactId multi1, MultiXactId multi2); -extern int multixactoffsetssyncfiletag(const FileTag *ftag, char *path); -extern int multixactmemberssyncfiletag(const FileTag *ftag, char *path); - extern void AtEOXact_MultiXact(void); extern void AtPrepare_MultiXact(void); extern void PostPrepare_MultiXact(TransactionId xid); diff --git a/src/include/access/slru.h b/src/include/access/slru.h index a8a424d92da..4b3e2e4a34d 100644 --- a/src/include/access/slru.h +++ b/src/include/access/slru.h @@ -1,7 +1,7 @@ /*------------------------------------------------------------------------- * * slru.h - * Simple LRU buffering for transaction status logfiles + * Buffering for transaction status logfiles * * Portions Copyright (c) 1996-2023, PostgreSQL Global Development Group * Portions Copyright (c) 1994, Regents of the University of California @@ -14,9 +14,36 @@ #define SLRU_H #include "access/xlogdefs.h" +#include "catalog/pg_tablespace_d.h" +#include "storage/buf.h" +#include "storage/bufmgr.h" #include "storage/lwlock.h" +#include "storage/smgr.h" #include "storage/sync.h" +/* Pseudo database ID used for SLRU data. */ +#define SLRU_SPC_ID 9 + +/* Pseudo database IDs used by each cache. */ +#define PG_SLRU(symname,name,path, synchronize) \ + symname, + +typedef enum SlruIds +{ +#include "access/slrulist.h" + SLRU_NEXT_ID +} SlruIds; +#undef PG_SLRU + +typedef bool (*SlruPagePrecedesFunction) (int, int); + +static inline RelFileLocator +SlruRelFileLocator(uint32 slru_db_id, uint32 segment_id) +{ + RelFileLocator rlocator = {SLRU_SPC_ID, slru_db_id, segment_id}; + return rlocator; +} + /* * Define SLRU segment size. A page is the same BLCKSZ as is used everywhere @@ -33,142 +60,40 @@ */ #define SLRU_PAGES_PER_SEGMENT 32 -/* - * Page status codes. Note that these do not include the "dirty" bit. - * page_dirty can be true only in the VALID or WRITE_IN_PROGRESS states; - * in the latter case it implies that the page has been re-dirtied since - * the write started. - */ -typedef enum -{ - SLRU_PAGE_EMPTY, /* buffer is not in use */ - SLRU_PAGE_READ_IN_PROGRESS, /* page is being read in */ - SLRU_PAGE_VALID, /* page is valid and not being written */ - SLRU_PAGE_WRITE_IN_PROGRESS /* page is being written out */ -} SlruPageStatus; - -/* - * Shared-memory state - */ -typedef struct SlruSharedData -{ - LWLock *ControlLock; - - /* Number of buffers managed by this SLRU structure */ - int num_slots; - - /* - * Arrays holding info for each buffer slot. Page number is undefined - * when status is EMPTY, as is page_lru_count. - */ - char **page_buffer; - SlruPageStatus *page_status; - bool *page_dirty; - int *page_number; - int *page_lru_count; - LWLockPadded *buffer_locks; - - /* - * Optional array of WAL flush LSNs associated with entries in the SLRU - * pages. If not zero/NULL, we must flush WAL before writing pages (true - * for pg_xact, false for multixact, pg_subtrans, pg_notify). group_lsn[] - * has lsn_groups_per_page entries per buffer slot, each containing the - * highest LSN known for a contiguous group of SLRU entries on that slot's - * page. - */ - XLogRecPtr *group_lsn; - int lsn_groups_per_page; - - /*---------- - * We mark a page "most recently used" by setting - * page_lru_count[slotno] = ++cur_lru_count; - * The oldest page is therefore the one with the highest value of - * cur_lru_count - page_lru_count[slotno] - * The counts will eventually wrap around, but this calculation still - * works as long as no page's age exceeds INT_MAX counts. - *---------- - */ - int cur_lru_count; - - /* - * latest_page_number is the page number of the current end of the log; - * this is not critical data, since we use it only to avoid swapping out - * the latest page. - */ - int latest_page_number; - - /* SLRU's index for statistics purposes (might not be unique) */ - int slru_stats_idx; -} SlruSharedData; - -typedef SlruSharedData *SlruShared; - -/* - * SlruCtlData is an unshared structure that points to the active information - * in shared memory. - */ -typedef struct SlruCtlData -{ - SlruShared shared; - - /* - * Which sync handler function to use when handing sync requests over to - * the checkpointer. SYNC_HANDLER_NONE to disable fsync (eg pg_notify). - */ - SyncRequestHandler sync_handler; - - /* - * Decide whether a page is "older" for truncation and as a hint for - * evicting pages in LRU order. Return true if every entry of the first - * argument is older than every entry of the second argument. Note that - * !PagePrecedes(a,b) && !PagePrecedes(b,a) need not imply a==b; it also - * arises when some entries are older and some are not. For SLRUs using - * SimpleLruTruncate(), this must use modular arithmetic. (For others, - * the behavior of this callback has no functional implications.) Use - * SlruPagePrecedesUnitTests() in SLRUs meeting its criteria. - */ - bool (*PagePrecedes) (int, int); - - /* - * Dir is set during SimpleLruInit and does not change thereafter. Since - * it's always the same, it doesn't need to be in shared memory. - */ - char Dir[64]; -} SlruCtlData; - -typedef SlruCtlData *SlruCtl; - - -extern Size SimpleLruShmemSize(int nslots, int nlsns); -extern void SimpleLruInit(SlruCtl ctl, const char *name, int nslots, int nlsns, - LWLock *ctllock, const char *subdir, int tranche_id, - SyncRequestHandler sync_handler); -extern int SimpleLruZeroPage(SlruCtl ctl, int pageno); -extern int SimpleLruReadPage(SlruCtl ctl, int pageno, bool write_ok, - TransactionId xid); -extern int SimpleLruReadPage_ReadOnly(SlruCtl ctl, int pageno, - TransactionId xid); -extern void SimpleLruWritePage(SlruCtl ctl, int slotno); -extern void SimpleLruWriteAll(SlruCtl ctl, bool allow_redirtied); #ifdef USE_ASSERT_CHECKING -extern void SlruPagePrecedesUnitTests(SlruCtl ctl, int per_page); +extern void SlruPagePrecedesUnitTests(SlruPagePrecedesFunction PagePrecedes, + int per_page); #else #define SlruPagePrecedesUnitTests(ctl, per_page) do {} while (0) #endif -extern void SimpleLruTruncate(SlruCtl ctl, int cutoffPage); -extern bool SimpleLruDoesPhysicalPageExist(SlruCtl ctl, int pageno); +extern void SimpleLruTruncate(int slru_id, SlruPagePrecedesFunction PagePrecedes, + int cutoffPage); +extern bool SimpleLruDoesPhysicalPageExist(int slru_id, int pageno); -typedef bool (*SlruScanCallback) (SlruCtl ctl, char *filename, int segpage, +typedef bool (*SlruScanCallback) (int slru_id, + SlruPagePrecedesFunction PagePrecedes, + char *filename, int segpage, void *data); -extern bool SlruScanDirectory(SlruCtl ctl, SlruScanCallback callback, void *data); -extern void SlruDeleteSegment(SlruCtl ctl, int segno); - -extern int SlruSyncFileTag(SlruCtl ctl, const FileTag *ftag, char *path); +extern bool SlruScanDirectory(int slru_id, SlruPagePrecedesFunction PagePrecedes, + SlruScanCallback callback, void *data); +extern void SlruDeleteSegment(int slru_id, int segno); /* SlruScanDirectory public callbacks */ -extern bool SlruScanDirCbReportPresence(SlruCtl ctl, char *filename, +extern bool SlruScanDirCbReportPresence(int slru_id, + SlruPagePrecedesFunction PagePrecedes, + char *filename, int segpage, void *data); -extern bool SlruScanDirCbDeleteAll(SlruCtl ctl, char *filename, int segpage, +extern bool SlruScanDirCbDeleteAll(int slru_id, SlruPagePrecedesFunction PagePrecedes, + char *filename, int segpage, void *data); +/* Buffer access */ +extern Buffer ReadSlruBuffer(int slru_id, int pageno, ReadBufferMode mode); +extern Buffer ZeroSlruBuffer(int slru_id, int pageno); +extern bool ProbeSlruBuffer(int slru_id, int pageno); + +/* Interfaces use by stats view */ +extern Oid SlruRelIdByName(const char *name); +extern const char *SlruName(int slru_id); + #endif /* SLRU_H */ diff --git a/src/include/access/slrudefs.h b/src/include/access/slrudefs.h new file mode 100644 index 00000000000..49cd78d923d --- /dev/null +++ b/src/include/access/slrudefs.h @@ -0,0 +1,19 @@ +/*------------------------------------------------------------------------- + * + * slrudefs.h + * macros for accessing contents of "slru" pages + * + * + * Copyright (c) 2021-2022, PostgreSQL Global Development Group + * + * src/include/access/slrudefs.h + * + *-------------------------------------------------------------------------- + */ + +#define SLRU_PAGES_PER_SEGMENT 32 + +#define MULTIXACT_MEMBER_ENTRY_SIZE 20 + +#define MULTIXACT_OFFSET_ENTRY_SIZE 8 + diff --git a/src/include/access/slrulist.h b/src/include/access/slrulist.h new file mode 100644 index 00000000000..c1289a1326c --- /dev/null +++ b/src/include/access/slrulist.h @@ -0,0 +1,30 @@ +/*--------------------------------------------------------------------------- + * slrulist.h + * + * The SLRU list is kept in its own source file for possible + * use by automatic tools. The exact representation of a rmgr is determined + * by the PG_SLRU macro, which is not defined in this file; it can be + * defined by the caller for special purposes. + * + * Portions Copyright (c) 1996-2022, PostgreSQL Global Development Group + * Portions Copyright (c) 1994, Regents of the University of California + * + * src/include/access/slrulist.h + *--------------------------------------------------------------------------- + */ + +/* there is deliberately not an #ifndef SLRULIST_H here */ + +/* + * List of SLRU entries. Note that order of entries defines the + * numerical values of each SLRU's ID, which is used in in-memory structus. + */ + +/* symbol name, textual name, path, synchronize */ +PG_SLRU(SLRU_CLOG_ID, "Xact", "pg_xact", true) +PG_SLRU(SLRU_SUBTRANS_ID, "Subtrans", "pg_subtrans", false) +PG_SLRU(SLRU_MULTIXACT_OFFSET_ID, "MultiXactOffset", "pg_multixact/offsets", true) +PG_SLRU(SLRU_MULTIXACT_MEMBER_ID, "MultiXactMember", "pg_multixact/members", true) +PG_SLRU(SLRU_COMMIT_TS_ID, "CommitTs", "pg_commit_ts", true) +PG_SLRU(SLRU_SERIAL_ID, "Serial", "pg_serial", false) +PG_SLRU(SLRU_NOTIFY_ID, "Notify", "pg_notify", false) diff --git a/src/include/access/subtrans.h b/src/include/access/subtrans.h index 46a473c77f5..14e3bf720fe 100644 --- a/src/include/access/subtrans.h +++ b/src/include/access/subtrans.h @@ -18,11 +18,8 @@ extern void SubTransSetParent(TransactionId xid, TransactionId parent); extern TransactionId SubTransGetParent(TransactionId xid); extern TransactionId SubTransGetTopmostTransaction(TransactionId xid); -extern Size SUBTRANSShmemSize(void); -extern void SUBTRANSShmemInit(void); extern void BootStrapSUBTRANS(void); extern void StartupSUBTRANS(TransactionId oldestActiveXID); -extern void CheckPointSUBTRANS(void); extern void ExtendSUBTRANS(TransactionId newestXact); extern void TruncateSUBTRANS(TransactionId oldestXact); diff --git a/src/include/catalog/storage.h b/src/include/catalog/storage.h index 45a3c7835cb..c96f448225b 100644 --- a/src/include/catalog/storage.h +++ b/src/include/catalog/storage.h @@ -22,19 +22,20 @@ /* GUC variables */ extern PGDLLIMPORT int wal_skip_threshold; -extern SMgrRelation RelationCreateStorage(RelFileLocator rlocator, - char relpersistence, - bool register_delete); +extern SMgrFileHandle RelationCreateStorage(RelFileLocator rlocator, + char relpersistence, + bool register_delete); extern void RelationDropStorage(Relation rel); extern void RelationPreserveStorage(RelFileLocator rlocator, bool atCommit); extern void RelationPreTruncate(Relation rel); extern void RelationTruncate(Relation rel, BlockNumber nblocks); -extern void RelationCopyStorage(SMgrRelation src, SMgrRelation dst, - ForkNumber forkNum, char relpersistence); +extern void RelationCopyStorage(SMgrFileHandle src, SMgrFileHandle dst, + char relpersistence); extern bool RelFileLocatorSkippingWAL(RelFileLocator rlocator); extern Size EstimatePendingSyncsSpace(void); extern void SerializePendingSyncs(Size maxSize, char *startAddress); extern void RestorePendingSyncs(char *startAddress); +extern void DropRelationFiles(RelFileLocator *delrels, int ndelrels, bool isRedo); /* * These functions used to be in storage/smgr/smgr.c, which explains the diff --git a/src/include/common/relpath.h b/src/include/common/relpath.h index 511c21682ec..7f132864592 100644 --- a/src/include/common/relpath.h +++ b/src/include/common/relpath.h @@ -63,6 +63,9 @@ typedef enum ForkNumber #define FORKNAMECHARS 4 /* max chars for a fork name */ +/* Pseudo tablespace ID used for SLRUs. */ +#define SLRU_SPC_OID 9 + extern PGDLLIMPORT const char *const forkNames[]; extern ForkNumber forkname_to_number(const char *forkName); @@ -73,25 +76,25 @@ extern int forkname_chars(const char *str, ForkNumber *fork); */ extern char *GetDatabasePath(Oid dbOid, Oid spcOid); -extern char *GetRelationPath(Oid dbOid, Oid spcOid, RelFileNumber relNumber, +extern char *GetSMgrFilePath(Oid dbOid, Oid spcOid, RelFileNumber relNumber, int backendId, ForkNumber forkNumber); /* * Wrapper macros for GetRelationPath. Beware of multiple - * evaluation of the RelFileLocator or RelFileLocatorBackend argument! + * evaluation of the RelFileLocator or SMgrFileLocator argument! */ /* First argument is a RelFileLocator */ #define relpathbackend(rlocator, backend, forknum) \ - GetRelationPath((rlocator).dbOid, (rlocator).spcOid, (rlocator).relNumber, \ + GetSMgrFilePath((rlocator).dbOid, (rlocator).spcOid, (rlocator).relNumber, \ backend, forknum) /* First argument is a RelFileLocator */ #define relpathperm(rlocator, forknum) \ relpathbackend(rlocator, InvalidBackendId, forknum) -/* First argument is a RelFileLocatorBackend */ -#define relpath(rlocator, forknum) \ - relpathbackend((rlocator).locator, (rlocator).backend, forknum) +/* First argument is a SMgrFileLocator */ +#define smgrfilepath(slocator) \ + GetSMgrFilePath((slocator).locator.dbOid, (slocator).locator.spcOid, (slocator).locator.relNumber, (slocator).backend, (slocator).forknum) #endif /* RELPATH_H */ diff --git a/src/include/storage/buf_internals.h b/src/include/storage/buf_internals.h index ed8aa2519c0..2c8cf856224 100644 --- a/src/include/storage/buf_internals.h +++ b/src/include/storage/buf_internals.h @@ -281,6 +281,11 @@ typedef union BufferDescPadded char pad[BUFFERDESC_PAD_TO_SIZE]; } BufferDescPadded; +#define BufferGetExternalLSN(bufHdr) \ + BufferExternalLSNs[(bufHdr)->buf_id] +#define BufferSetExternalLSN(bufHdr, lsn) \ + BufferExternalLSNs[(bufHdr)->buf_id] = (lsn) + /* * The PendingWriteback & WritebackContext structure are used to keep * information about pending flush requests to be issued to the OS. @@ -307,6 +312,7 @@ typedef struct WritebackContext /* in buf_init.c */ extern PGDLLIMPORT BufferDescPadded *BufferDescriptors; extern PGDLLIMPORT ConditionVariableMinimallyPadded *BufferIOCVArray; +extern PGDLLIMPORT XLogRecPtr *BufferExternalLSNs; extern PGDLLIMPORT WritebackContext BackendWritebackContext; /* in localbuf.c */ @@ -413,10 +419,9 @@ extern int BufTableInsert(BufferTag *tagPtr, uint32 hashcode, int buf_id); extern void BufTableDelete(BufferTag *tagPtr, uint32 hashcode); /* localbuf.c */ -extern PrefetchBufferResult PrefetchLocalBuffer(SMgrRelation smgr, - ForkNumber forkNum, +extern PrefetchBufferResult PrefetchLocalBuffer(SMgrFileHandle smgr, BlockNumber blockNum); -extern BufferDesc *LocalBufferAlloc(SMgrRelation smgr, ForkNumber forkNum, +extern BufferDesc *LocalBufferAlloc(SMgrFileHandle smgr, BlockNumber blockNum, bool *foundPtr); extern void MarkLocalBufferDirty(Buffer buffer); extern void DropRelationLocalBuffers(RelFileLocator rlocator, diff --git a/src/include/storage/bufmgr.h b/src/include/storage/bufmgr.h index 33eadbc1291..e6aed18f0fc 100644 --- a/src/include/storage/bufmgr.h +++ b/src/include/storage/bufmgr.h @@ -42,8 +42,11 @@ typedef enum RBM_ZERO_AND_CLEANUP_LOCK, /* Like RBM_ZERO_AND_LOCK, but locks the page * in "cleanup" mode */ RBM_ZERO_ON_ERROR, /* Read, but return an all-zeros page on error */ - RBM_NORMAL_NO_LOG /* Don't log page as invalid during WAL + RBM_NORMAL_NO_LOG, /* Don't log page as invalid during WAL * replay; otherwise same as RBM_NORMAL */ + + RBM_TRIM /*Read for TRIM functions in CLOG / MultiXact. + Don't validate checksum or zero. */ } ReadBufferMode; /* @@ -59,7 +62,7 @@ typedef struct PrefetchBufferResult struct WritebackContext; /* forward declared, to avoid including smgr.h here */ -struct SMgrRelationData; +struct SMgrFileData; /* in globals.c ... this duplicates miscadmin.h */ extern PGDLLIMPORT int NBuffers; @@ -110,8 +113,7 @@ extern PGDLLIMPORT int32 *LocalRefCount; /* * prototypes for functions in bufmgr.c */ -extern PrefetchBufferResult PrefetchSharedBuffer(struct SMgrRelationData *smgr_reln, - ForkNumber forkNum, +extern PrefetchBufferResult PrefetchSharedBuffer(struct SMgrFileData *smgr_file, BlockNumber blockNum); extern PrefetchBufferResult PrefetchBuffer(Relation reln, ForkNumber forkNum, BlockNumber blockNum); @@ -125,13 +127,19 @@ extern Buffer ReadBufferWithoutRelcache(RelFileLocator rlocator, ForkNumber forkNum, BlockNumber blockNum, ReadBufferMode mode, BufferAccessStrategy strategy, bool permanent); +extern Buffer ReadBufferWithoutRelcacheWithHit(RelFileLocator rlocator, + ForkNumber forkNum, BlockNumber blockNum, + ReadBufferMode mode, BufferAccessStrategy strategy, + bool permanent, bool *hit); extern void ReleaseBuffer(Buffer buffer); extern void UnlockReleaseBuffer(Buffer buffer); extern void MarkBufferDirty(Buffer buffer); extern void IncrBufferRefCount(Buffer buffer); extern Buffer ReleaseAndReadBuffer(Buffer buffer, Relation relation, BlockNumber blockNum); - +extern bool BufferProbe(RelFileLocator rlocator, ForkNumber forkNum, + BlockNumber blockNum); + extern void InitBufferPoolAccess(void); extern void AtEOXact_Buffers(bool isCommit); extern void PrintBufferLeakWarning(Buffer buffer); @@ -139,18 +147,26 @@ extern void CheckPointBuffers(int flags); extern BlockNumber BufferGetBlockNumber(Buffer buffer); extern BlockNumber RelationGetNumberOfBlocksInFork(Relation relation, ForkNumber forkNum); -extern void FlushOneBuffer(Buffer buffer); -extern void FlushRelationBuffers(Relation rel); -extern void FlushRelationsAllBuffers(struct SMgrRelationData **smgrs, int nrels); extern void CreateAndCopyRelationData(RelFileLocator src_rlocator, RelFileLocator dst_rlocator, bool permanent); + +extern void FlushOneBuffer(Buffer buffer); +extern void FlushRelationBuffers(Relation rel); +extern void FlushRelationsAllBuffers(RelFileLocator *locators, int nlocators); extern void FlushDatabaseBuffers(Oid dbid); -extern void DropRelationBuffers(struct SMgrRelationData *smgr_reln, + +extern void DropRelationBuffers(RelFileLocator rlocator, BackendId backend, ForkNumber *forkNum, int nforks, BlockNumber *firstDelBlock); -extern void DropRelationsAllBuffers(struct SMgrRelationData **smgr_reln, - int nlocators); + +typedef struct RelFileLocatorBackend +{ + RelFileLocator locator; + BackendId backend; +} RelFileLocatorBackend; + +extern void DropRelationsAllBuffers(RelFileLocatorBackend *locators, int nlocators); extern void DropDatabaseBuffers(Oid dbid); #define RelationGetNumberOfBlocks(reln) \ diff --git a/src/include/storage/bufpage.h b/src/include/storage/bufpage.h index 424ecba028f..c944e3dc928 100644 --- a/src/include/storage/bufpage.h +++ b/src/include/storage/bufpage.h @@ -201,6 +201,7 @@ typedef PageHeaderData *PageHeader; * handling pages. */ #define PG_PAGE_LAYOUT_VERSION 4 +#define PG_METAPAGE_LAYOUT_VERSION 1 #define PG_DATA_CHECKSUM_VERSION 1 /* ---------------------------------------------------------------- @@ -302,6 +303,20 @@ PageSetPageSizeAndVersion(Page page, Size size, uint8 version) ((PageHeader) page)->pd_pagesize_version = size | version; } +/* + * PageSetHeaderDataMinimal + * Sets the LSN, page size and version, and checksum + */ +#define PageSetHeaderDataNonRel(page, pageno, lsn, size, version) \ +( \ + PageSetLSN(page, lsn), \ + PageSetPageSizeAndVersion(page, size, version), \ + PageClearHasFreeLinePointers(page), \ + PageSetChecksumInplace(page, pageno) \ +) + + + /* ---------------- * page special data functions * ---------------- @@ -486,6 +501,8 @@ StaticAssertDecl(BLCKSZ == ((BLCKSZ / sizeof(size_t)) * sizeof(size_t)), "BLCKSZ has to be a multiple of sizeof(size_t)"); extern void PageInit(Page page, Size pageSize, Size specialSize); +extern void PageInitSLRU(Page page, Size pageSize, Size specialSize); + extern bool PageIsVerifiedExtended(Page page, BlockNumber blkno, int flags); extern OffsetNumber PageAddItemExtended(Page page, Item item, Size size, OffsetNumber offsetNumber, int flags); diff --git a/src/include/storage/md.h b/src/include/storage/md.h index bcada9ff221..a052945685c 100644 --- a/src/include/storage/md.h +++ b/src/include/storage/md.h @@ -21,28 +21,26 @@ /* md storage manager functionality */ extern void mdinit(void); -extern void mdopen(SMgrRelation reln); -extern void mdclose(SMgrRelation reln, ForkNumber forknum); -extern void mdcreate(SMgrRelation reln, ForkNumber forknum, bool isRedo); -extern bool mdexists(SMgrRelation reln, ForkNumber forknum); -extern void mdunlink(RelFileLocatorBackend rlocator, ForkNumber forknum, bool isRedo); -extern void mdextend(SMgrRelation reln, ForkNumber forknum, +extern void mdopen(SMgrFileHandle sfile); +extern void mdclose(SMgrFileHandle sfile); +extern void mdcreate(SMgrFileHandle sfile, bool isRedo); +extern bool mdexists(SMgrFileHandle sfile); +extern void mdunlink(SMgrFileLocator slocator, bool isRedo); +extern void mdextend(SMgrFileHandle sfile, BlockNumber blocknum, char *buffer, bool skipFsync); -extern bool mdprefetch(SMgrRelation reln, ForkNumber forknum, +extern bool mdprefetch(SMgrFileHandle sfile, BlockNumber blocknum); -extern void mdread(SMgrRelation reln, ForkNumber forknum, BlockNumber blocknum, +extern void mdread(SMgrFileHandle sfile, BlockNumber blocknum, char *buffer); -extern void mdwrite(SMgrRelation reln, ForkNumber forknum, +extern void mdwrite(SMgrFileHandle sfile, BlockNumber blocknum, char *buffer, bool skipFsync); -extern void mdwriteback(SMgrRelation reln, ForkNumber forknum, +extern void mdwriteback(SMgrFileHandle sfile, BlockNumber blocknum, BlockNumber nblocks); -extern BlockNumber mdnblocks(SMgrRelation reln, ForkNumber forknum); -extern void mdtruncate(SMgrRelation reln, ForkNumber forknum, - BlockNumber nblocks); -extern void mdimmedsync(SMgrRelation reln, ForkNumber forknum); +extern BlockNumber mdnblocks(SMgrFileHandle sfile); +extern void mdtruncate(SMgrFileHandle sfile, BlockNumber nblocks); +extern void mdimmedsync(SMgrFileHandle sfile); extern void ForgetDatabaseSyncRequests(Oid dbid); -extern void DropRelationFiles(RelFileLocator *delrels, int ndelrels, bool isRedo); /* md sync callbacks */ extern int mdsyncfiletag(const FileTag *ftag, char *path); diff --git a/src/include/storage/relfilelocator.h b/src/include/storage/relfilelocator.h index 61cf0169bd7..e2c22b8e6f9 100644 --- a/src/include/storage/relfilelocator.h +++ b/src/include/storage/relfilelocator.h @@ -53,6 +53,8 @@ * Note: various places use RelFileLocator in hashtable keys. Therefore, * there *must not* be any unused padding bytes in this struct. That * should be safe as long as all the fields are of type Oid. + * + * See also SMgrFileLocator in smgr.h. */ typedef struct RelFileLocator { @@ -62,38 +64,15 @@ typedef struct RelFileLocator } RelFileLocator; /* - * Augmenting a relfilelocator with the backend ID provides all the information - * we need to locate the physical storage. The backend ID is InvalidBackendId - * for regular relations (those accessible to more than one backend), or the - * owning backend's ID for backend-local relations. Backend-local relations - * are always transient and removed in case of a database crash; they are - * never WAL-logged or fsync'd. - */ -typedef struct RelFileLocatorBackend -{ - RelFileLocator locator; - BackendId backend; -} RelFileLocatorBackend; - -#define RelFileLocatorBackendIsTemp(rlocator) \ - ((rlocator).backend != InvalidBackendId) - -/* - * Note: RelFileLocatorEquals and RelFileLocatorBackendEquals compare relNumber + * Note: RelFileLocatorEquals compares relNumber * first since that is most likely to be different in two unequal * RelFileLocators. It is probably redundant to compare spcOid if the other * fields are found equal, but do it anyway to be sure. Likewise for checking - * the backend ID in RelFileLocatorBackendEquals. + * the backend ID in SMgrFileLocatorBackendEquals. */ #define RelFileLocatorEquals(locator1, locator2) \ ((locator1).relNumber == (locator2).relNumber && \ (locator1).dbOid == (locator2).dbOid && \ (locator1).spcOid == (locator2).spcOid) -#define RelFileLocatorBackendEquals(locator1, locator2) \ - ((locator1).locator.relNumber == (locator2).locator.relNumber && \ - (locator1).locator.dbOid == (locator2).locator.dbOid && \ - (locator1).backend == (locator2).backend && \ - (locator1).locator.spcOid == (locator2).locator.spcOid) - #endif /* RELFILELOCATOR_H */ diff --git a/src/include/storage/smgr.h b/src/include/storage/smgr.h index 56233c4d216..83c4a756ce5 100644 --- a/src/include/storage/smgr.h +++ b/src/include/storage/smgr.h @@ -18,31 +18,56 @@ #include "storage/block.h" #include "storage/relfilelocator.h" + /* - * smgr.c maintains a table of SMgrRelation objects, which are essentially - * cached file handles. An SMgrRelation is created (if not already present) + * SMgrFileLocator contains all the information needed to locate the physical + * storage of a relation fork, or some other file that is managed by the buffer + * manager. + * + * The backend ID is InvalidBackendId for regular relations (those accessible + * to more than one backend), or the owning backend's ID for backend-local + * relations. Backend-local relations are always transient and removed in + * case of a database crash; they are never WAL-logged or fsync'd. + */ +typedef struct SMgrFileLocator +{ + RelFileLocator locator; + BackendId backend; + ForkNumber forknum; +} SMgrFileLocator; + +#define SMgrFileLocatorIsTemp(slocator) \ + ((slocator).backend != InvalidBackendId) + +/* + * smgr.c maintains a table of SMgrFileData objects, which are essentially + * cached file handles. An SMgrFile is created (if not already present) * by smgropen(), and destroyed by smgrclose(). Note that neither of these * operations imply I/O, they just create or destroy a hashtable entry. * (But smgrclose() may release associated resources, such as OS-level file * descriptors.) * - * An SMgrRelation may have an "owner", which is just a pointer to it from - * somewhere else; smgr.c will clear this pointer if the SMgrRelation is + * An SMgrFile may have an "owner", which is just a pointer to it from + * somewhere else; smgr.c will clear this pointer if the SMgrFile is * closed. We use this to avoid dangling pointers from relcache to smgr * without having to make the smgr explicitly aware of relcache. There * can't be more than one "owner" pointer per SMgrRelation, but that's * all we need. * - * SMgrRelations that do not have an "owner" are considered to be transient, + * SMgrFiles that do not have an "owner" are considered to be transient, * and are deleted at end of transaction. + * + * A file that is represented by an SMgrFile can be managed by the buffer + * manager. Currently, it's only used for relation files, but could be used + * for SLRUs and other things in the future. */ -typedef struct SMgrRelationData +typedef struct SMgrFileData { - /* rlocator is the hashtable lookup key, so it must be first! */ - RelFileLocatorBackend smgr_rlocator; /* relation physical identifier */ + /* locator is the hashtable lookup key, so must be first! */ + SMgrFileLocator smgr_locator; /* file physical identifier */ /* pointer to owning pointer, or NULL if none */ - struct SMgrRelationData **smgr_owner; + struct SMgrFileData **smgr_owner; /* * The following fields are reset to InvalidBlockNumber upon a cache flush @@ -51,7 +76,7 @@ typedef struct SMgrRelationData * invalidation for fork extension. */ BlockNumber smgr_targblock; /* current insertion target block */ - BlockNumber smgr_cached_nblocks[MAX_FORKNUM + 1]; /* last known size */ + BlockNumber smgr_cached_nblocks; /* last known size */ /* additional public fields may someday exist here */ @@ -65,46 +90,46 @@ typedef struct SMgrRelationData * for md.c; per-fork arrays of the number of open segments * (md_num_open_segs) and the segments themselves (md_seg_fds). */ - int md_num_open_segs[MAX_FORKNUM + 1]; - struct _MdfdVec *md_seg_fds[MAX_FORKNUM + 1]; + int md_num_open_segs; + struct _MdfdVec *md_seg_fds; - /* if unowned, list link in list of all unowned SMgrRelations */ + /* if unowned, list link in list of all unowned SMgrFiles */ dlist_node node; -} SMgrRelationData; +} SMgrFileData; -typedef SMgrRelationData *SMgrRelation; +typedef SMgrFileData *SMgrFileHandle; #define SmgrIsTemp(smgr) \ - RelFileLocatorBackendIsTemp((smgr)->smgr_rlocator) + SMgrFileLocatorIsTemp((smgr)->smgr_locator) extern void smgrinit(void); -extern SMgrRelation smgropen(RelFileLocator rlocator, BackendId backend); -extern bool smgrexists(SMgrRelation reln, ForkNumber forknum); -extern void smgrsetowner(SMgrRelation *owner, SMgrRelation reln); -extern void smgrclearowner(SMgrRelation *owner, SMgrRelation reln); -extern void smgrclose(SMgrRelation reln); +extern SMgrFileHandle smgropen(RelFileLocator rlocator, BackendId backend, ForkNumber forkNum); +extern bool smgrexists(SMgrFileHandle sfile); +extern void smgrsetowner(SMgrFileHandle *owner, SMgrFileHandle sfile); +extern void smgrclearowner(SMgrFileHandle *owner, SMgrFileHandle sfile); +extern void smgrclose(SMgrFileHandle sfile); extern void smgrcloseall(void); -extern void smgrcloserellocator(RelFileLocatorBackend rlocator); -extern void smgrrelease(SMgrRelation reln); extern void smgrreleaseall(void); -extern void smgrcreate(SMgrRelation reln, ForkNumber forknum, bool isRedo); -extern void smgrdosyncall(SMgrRelation *rels, int nrels); -extern void smgrdounlinkall(SMgrRelation *rels, int nrels, bool isRedo); -extern void smgrextend(SMgrRelation reln, ForkNumber forknum, +extern void smgrcreate(SMgrFileHandle sfile, bool isRedo); +extern void smgrextend(SMgrFileHandle sfile, BlockNumber blocknum, char *buffer, bool skipFsync); -extern bool smgrprefetch(SMgrRelation reln, ForkNumber forknum, - BlockNumber blocknum); -extern void smgrread(SMgrRelation reln, ForkNumber forknum, +extern bool smgrprefetch(SMgrFileHandle sfile, BlockNumber blocknum); +extern void smgrread(SMgrFileHandle sfile, BlockNumber blocknum, char *buffer); -extern void smgrwrite(SMgrRelation reln, ForkNumber forknum, +extern void smgrwrite(SMgrFileHandle sfile, BlockNumber blocknum, char *buffer, bool skipFsync); -extern void smgrwriteback(SMgrRelation reln, ForkNumber forknum, +extern void smgrwriteback(SMgrFileHandle sfile, BlockNumber blocknum, BlockNumber nblocks); -extern BlockNumber smgrnblocks(SMgrRelation reln, ForkNumber forknum); -extern BlockNumber smgrnblocks_cached(SMgrRelation reln, ForkNumber forknum); -extern void smgrtruncate(SMgrRelation reln, ForkNumber *forknum, - int nforks, BlockNumber *nblocks); -extern void smgrimmedsync(SMgrRelation reln, ForkNumber forknum); +extern BlockNumber smgrnblocks(SMgrFileHandle sfile); +extern BlockNumber smgrnblocks_cached(SMgrFileHandle sfile); +extern void smgrimmedsync(SMgrFileHandle sfile); +extern void smgrunlink(SMgrFileHandle sfile, bool isRedo); + +extern void smgrtruncate_multi(RelFileLocator rlocator, BackendId backend, ForkNumber *forks, int nforks, BlockNumber *nblocks); +extern void smgrunlink_multi(RelFileLocator rlocator, BackendId backend, ForkNumber *forks, int nforks, bool isRedo); + +extern void smgrcloserellocator(RelFileLocator rlocator, BackendId backend); + extern void AtEOXact_SMgr(void); extern bool ProcessBarrierSmgrRelease(void); diff --git a/src/include/utils/inval.h b/src/include/utils/inval.h index 14b4eac0630..6c279422b2e 100644 --- a/src/include/utils/inval.h +++ b/src/include/utils/inval.h @@ -48,7 +48,7 @@ extern void CacheInvalidateRelcacheByTuple(HeapTuple classTuple); extern void CacheInvalidateRelcacheByRelid(Oid relid); -extern void CacheInvalidateSmgr(RelFileLocatorBackend rlocator); +extern void CacheInvalidateSmgr(RelFileLocator rlocator, BackendId backend); extern void CacheInvalidateRelmap(Oid databaseId); diff --git a/src/include/utils/rel.h b/src/include/utils/rel.h index af9785038d3..bd7b98dee99 100644 --- a/src/include/utils/rel.h +++ b/src/include/utils/rel.h @@ -54,7 +54,7 @@ typedef LockInfoData *LockInfo; typedef struct RelationData { RelFileLocator rd_locator; /* relation physical identifier */ - SMgrRelation rd_smgr; /* cached file handle, or NULL */ + SMgrFileHandle rd_smgr[MAX_FORKNUM + 1]; /* cached file handles, or NULLs */ int rd_refcnt; /* reference count */ BackendId rd_backend; /* owning backend id, if temporary relation */ bool rd_islocaltemp; /* rel is a temp rel of this session */ @@ -562,16 +562,17 @@ typedef struct ViewOptions * Note: since a relcache flush can cause the file handle to be closed again, * it's unwise to hold onto the pointer returned by this function for any * long period. Recommended practice is to just re-execute RelationGetSmgr - * each time you need to access the SMgrRelation. It's quite cheap in + * each time you need to access the SMgrFileHandle. It's quite cheap in * comparison to whatever an smgr function is going to do. */ -static inline SMgrRelation -RelationGetSmgr(Relation rel) +static inline SMgrFileHandle +RelationGetSmgr(Relation rel, ForkNumber forkNum) { - if (unlikely(rel->rd_smgr == NULL)) - smgrsetowner(&(rel->rd_smgr), smgropen(rel->rd_locator, rel->rd_backend)); - return rel->rd_smgr; + if (unlikely(rel->rd_smgr[forkNum] == NULL)) + smgrsetowner(&(rel->rd_smgr[forkNum]), smgropen(rel->rd_locator, rel->rd_backend, forkNum)); + return rel->rd_smgr[forkNum]; } +#endif /* * RelationCloseSmgr @@ -580,13 +581,15 @@ RelationGetSmgr(Relation rel) static inline void RelationCloseSmgr(Relation relation) { - if (relation->rd_smgr != NULL) - smgrclose(relation->rd_smgr); - - /* smgrclose should unhook from owner pointer */ - Assert(relation->rd_smgr == NULL); + for (int i = 0; i <= MAX_FORKNUM; i++) + { + if (relation->rd_smgr[i] != NULL) + { + smgrclose(relation->rd_smgr[i]); + Assert(relation->rd_smgr[i] == NULL); + } + } } -#endif /* !FRONTEND */ /* * RelationGetTargetBlock @@ -597,7 +600,7 @@ RelationCloseSmgr(Relation relation) * so there's no need to re-open the smgr handle if it's not currently open. */ #define RelationGetTargetBlock(relation) \ - ( (relation)->rd_smgr != NULL ? (relation)->rd_smgr->smgr_targblock : InvalidBlockNumber ) + ( (relation)->rd_smgr[MAIN_FORKNUM] != NULL ? (relation)->rd_smgr[MAIN_FORKNUM]->smgr_targblock : InvalidBlockNumber ) /* * RelationSetTargetBlock @@ -605,7 +608,7 @@ RelationCloseSmgr(Relation relation) */ #define RelationSetTargetBlock(relation, targblock) \ do { \ - RelationGetSmgr(relation)->smgr_targblock = (targblock); \ + RelationGetSmgr(relation, MAIN_FORKNUM)->smgr_targblock = (targblock); \ } while (0) /* -- 2.38.1 ^ permalink raw reply [nested|flat] 2+ messages in thread
end of thread, other threads:[~2023-01-10 19:05 UTC | newest] Thread overview: 2+ messages (download: mbox mbox.gz follow: Atom feed) -- links below jump to the message on this page -- 2021-01-11 04:01 [PATCH] Remove self-joins. Andrey Lepikhov <[email protected]> 2023-01-10 19:05 Re: SLRUs in the main buffer pool - Page Header definitions Bagga, Rishu <[email protected]>
This inbox is served by agora; see mirroring instructions for how to clone and mirror all data and code used for this inbox