From b191695afcba438ae8c5d1c3b4d5939c76d22a4f Mon Sep 17 00:00:00 2001 From: Melanie Plageman Date: Tue, 2 Dec 2025 16:16:22 -0500 Subject: [PATCH v24 07/16] Eliminate XLOG_HEAP2_VISIBLE from vacuum phase I prune/freeze Vacuum no longer emits a separate WAL record for each page set all-visible or all-frozen during phase I. Instead, visibility map updates are now included in the XLOG_HEAP2_PRUNE_VACUUM_SCAN record that is already emitted for pruning and freezing. Previously, heap_page_prune_and_freeze() determined whether a page was all-visible, but the corresponding VM bits were only set later in lazy_scan_prune(). Now the VM is updated immediately in heap_page_prune_and_freeze(), at the same time as the heap modifications. This change applies only to vacuum phase I, not to pruning performed during normal page access. NOTE: This commit is the main commit and all review-only commits preceding it will be squashed into it. Author: Melanie Plageman Reviewed-by: Andres Freund Reviewed-by: Robert Haas Reviewed-by: Kirill Reshke Reviewed-by: Chao Li Discussion: https://postgr.es/m/flat/CAAKRu_ZMw6Npd_qm2KM%2BFwQ3cMOMx1Dh3VMhp8-V7SOLxdK9-g%40mail.gmail.com --- src/backend/access/heap/pruneheap.c | 255 ++++++++++++++++------------ 1 file changed, 144 insertions(+), 111 deletions(-) diff --git a/src/backend/access/heap/pruneheap.c b/src/backend/access/heap/pruneheap.c index 96dc902ec12..489b8487599 100644 --- a/src/backend/access/heap/pruneheap.c +++ b/src/backend/access/heap/pruneheap.c @@ -194,6 +194,12 @@ static void page_verify_redirects(Page page); static bool heap_page_will_freeze(Relation relation, Buffer buffer, bool did_tuple_hint_fpi, bool do_prune, bool do_hint_prune, PruneState *prstate); +static TransactionId get_conflict_xid(bool do_prune, bool do_freeze, bool do_set_vm, + uint8 new_vmbits, + TransactionId latest_xid_removed, + TransactionId frz_conflict_horizon, + TransactionId visibility_cutoff_xid, + bool blk_already_av); static bool heap_page_will_set_vm(Relation relation, BlockNumber heap_blk, Buffer heap_buf, @@ -782,6 +788,64 @@ heap_page_will_freeze(Relation relation, Buffer buffer, return do_freeze; } +/* + * Calculate the conflict horizon for the whole XLOG_HEAP2_PRUNE_VACUUM_SCAN + * or XLOG_HEAP2_PRUNE_ON_ACCESS record. + */ +static TransactionId +get_conflict_xid(bool do_prune, bool do_freeze, bool do_set_vm, uint8 new_vmbits, + TransactionId latest_xid_removed, TransactionId frz_conflict_horizon, + TransactionId visibility_cutoff_xid, bool blk_already_av) +{ + TransactionId conflict_xid; + + /* + * We can omit the snapshot conflict horizon if we are not pruning or + * freezing any tuples and are setting an already all-visible page + * all-frozen in the VM. In this case, all of the tuples on the page must + * already be visible to all MVCC snapshots on the standby. + */ + if (!do_prune && !do_freeze && + do_set_vm && blk_already_av && (new_vmbits & VISIBILITYMAP_ALL_FROZEN)) + return InvalidTransactionId; + + /* + * The snapshotConflictHorizon for the whole record should be the most + * conservative of all the horizons calculated for any of the possible + * modifications. If this record will prune tuples, any transactions on + * the standby older than the youngest xmax of the most recently removed + * tuple this record will prune will conflict. If this record will freeze + * tuples, any transactions on the standby with xids older than the + * youngest tuple this record will freeze will conflict. + */ + conflict_xid = InvalidTransactionId; + + /* + * If we are updating the VM, the conflict horizon is almost always the + * visibility cutoff XID. + * + * Separately, if we are freezing any tuples, as an optimization, we can + * use the visibility_cutoff_xid as the conflict horizon if the page will + * be all-frozen. This is true even if there are LP_DEAD line pointers + * because we ignored those when maintaining the visibility_cutoff_xid. + * This will have been calculated earlier as the frz_conflict_horizon when + * we determined we would freeze. + */ + if (do_set_vm) + conflict_xid = visibility_cutoff_xid; + else if (do_freeze) + conflict_xid = frz_conflict_horizon; + + /* + * If we are removing tuples with a younger xmax than our so far + * calculated conflict_xid, we must use this as our horizon. + */ + if (TransactionIdFollows(latest_xid_removed, conflict_xid)) + conflict_xid = latest_xid_removed; + + return conflict_xid; +} + /* * Decide whether to set the visibility map bits (all-visible and all-frozen) * for heap_blk using information from PruneState and blk_known_av. Some @@ -969,7 +1033,6 @@ heap_page_prune_and_freeze(PruneFreezeParams *params, Buffer vmbuffer = params->vmbuffer; Page page = BufferGetPage(buffer); BlockNumber blockno = BufferGetBlockNumber(buffer); - TransactionId vm_conflict_horizon = InvalidTransactionId; PruneState prstate; bool do_freeze; bool do_prune; @@ -977,6 +1040,9 @@ heap_page_prune_and_freeze(PruneFreezeParams *params, bool do_set_vm; bool did_tuple_hint_fpi; int64 fpi_before = pgWalUsage.wal_fpi; + TransactionId conflict_xid = InvalidTransactionId; + uint8 new_vmbits = 0; + uint8 old_vmbits = 0; /* Initialize prstate */ prune_freeze_setup(params, @@ -1038,6 +1104,36 @@ heap_page_prune_and_freeze(PruneFreezeParams *params, prstate.all_visible = prstate.all_frozen = false; Assert(!prstate.all_frozen || prstate.all_visible); + Assert(!prstate.all_visible || (prstate.lpdead_items == 0)); + + /* + * Decide whether to set the page-level PD_ALL_VISIBLE bit and the VM bits + * based on information from the VM and the all_visible/all_frozen flags. + * + * While it is valid for PD_ALL_VISIBLE to be set when the corresponding + * VM bit is clear, we strongly prefer to keep them in sync. + * + * Accordingly, we also allow updating only the VM when PD_ALL_VISIBLE has + * already been set. Setting only the VM is most common when setting an + * already all-visible page all-frozen. + */ + do_set_vm = heap_page_will_set_vm(params->relation, + blockno, buffer, vmbuffer, params->blk_known_av, + &prstate, &new_vmbits); + + /* + * new_vmbits should be 0 regardless of whether or not the page is + * all-visible if we do not intend to set the VM. + */ + Assert(do_set_vm || new_vmbits == 0); + + conflict_xid = get_conflict_xid(do_prune, do_freeze, do_set_vm, new_vmbits, + prstate.latest_xid_removed, prstate.frz_conflict_horizon, + prstate.visibility_cutoff_xid, params->blk_known_av); + + /* Lock vmbuffer before entering a critical section */ + if (do_set_vm) + LockBuffer(vmbuffer, BUFFER_LOCK_EXCLUSIVE); /* Any error while applying the changes is critical */ START_CRIT_SECTION(); @@ -1059,14 +1155,17 @@ heap_page_prune_and_freeze(PruneFreezeParams *params, /* * If that's all we had to do to the page, this is a non-WAL-logged - * hint. If we are going to freeze or prune the page, we will mark - * the buffer dirty below. + * hint. If we are going to freeze or prune the page or set + * PD_ALL_VISIBLE, we will mark the buffer dirty below. + * + * Setting PD_ALL_VISIBLE is fully WAL-logged because it is forbidden + * for the VM to be set and PD_ALL_VISIBLE to be clear. */ - if (!do_freeze && !do_prune) + if (!do_freeze && !do_prune && !do_set_vm) MarkBufferDirtyHint(buffer, true); } - if (do_prune || do_freeze) + if (do_prune || do_freeze || do_set_vm) { /* Apply the planned item changes and repair page fragmentation. */ if (do_prune) @@ -1080,6 +1179,15 @@ heap_page_prune_and_freeze(PruneFreezeParams *params, if (do_freeze) heap_freeze_prepared_tuples(buffer, prstate.frozen, prstate.nfrozen); + if (do_set_vm) + { + PageSetAllVisible(page); + old_vmbits = visibilitymap_set_vmbits(blockno, + vmbuffer, new_vmbits, + params->relation->rd_locator); + Assert(old_vmbits != new_vmbits); + } + MarkBufferDirty(buffer); /* @@ -1087,29 +1195,12 @@ heap_page_prune_and_freeze(PruneFreezeParams *params, */ if (RelationNeedsWAL(params->relation)) { - /* - * The snapshotConflictHorizon for the whole record should be the - * most conservative of all the horizons calculated for any of the - * possible modifications. If this record will prune tuples, any - * transactions on the standby older than the youngest xmax of the - * most recently removed tuple this record will prune will - * conflict. If this record will freeze tuples, any transactions - * on the standby with xids older than the youngest tuple this - * record will freeze will conflict. - */ - TransactionId conflict_xid; - - if (TransactionIdFollows(prstate.frz_conflict_horizon, - prstate.latest_xid_removed)) - conflict_xid = prstate.frz_conflict_horizon; - else - conflict_xid = prstate.latest_xid_removed; - log_heap_prune_and_freeze(params->relation, buffer, - InvalidBuffer, /* vmbuffer */ - 0, /* vmflags */ + do_set_vm ? vmbuffer : InvalidBuffer, + do_set_vm ? new_vmbits : 0, conflict_xid, - true, params->reason, + true, /* cleanup lock */ + params->reason, prstate.frozen, prstate.nfrozen, prstate.redirected, prstate.nredirected, prstate.nowdead, prstate.ndead, @@ -1119,46 +1210,8 @@ heap_page_prune_and_freeze(PruneFreezeParams *params, END_CRIT_SECTION(); - /* Copy information back for caller */ - presult->ndeleted = prstate.ndeleted; - presult->nnewlpdead = prstate.ndead; - presult->nfrozen = prstate.nfrozen; - presult->live_tuples = prstate.live_tuples; - presult->recently_dead_tuples = prstate.recently_dead_tuples; - presult->hastup = prstate.hastup; - - presult->lpdead_items = prstate.lpdead_items; - /* the presult->deadoffsets array was already filled in */ - - if (prstate.attempt_freeze) - { - if (presult->nfrozen > 0) - { - *new_relfrozen_xid = prstate.pagefrz.FreezePageRelfrozenXid; - *new_relmin_mxid = prstate.pagefrz.FreezePageRelminMxid; - } - else - { - *new_relfrozen_xid = prstate.pagefrz.NoFreezePageRelfrozenXid; - *new_relmin_mxid = prstate.pagefrz.NoFreezePageRelminMxid; - } - } - - presult->new_vmbits = 0; - presult->old_vmbits = 0; - - /* - * If updating the visibility map, the conflict horizon for that record - * must be the newest xmin on the page. However, if the page is - * completely frozen, there can be no conflict and the vm_conflict_horizon - * should remain InvalidTransactionId. This includes the case that we - * just froze all the tuples; the prune-freeze record included the - * conflict XID already so we don't need to again. - */ - if (prstate.all_frozen) - vm_conflict_horizon = InvalidTransactionId; - else - vm_conflict_horizon = prstate.visibility_cutoff_xid; + if (do_set_vm) + LockBuffer(vmbuffer, BUFFER_LOCK_UNLOCK); /* * During its second pass over the heap, VACUUM calls @@ -1173,7 +1226,8 @@ heap_page_prune_and_freeze(PruneFreezeParams *params, TransactionId debug_cutoff; bool debug_all_frozen; - Assert(presult->lpdead_items == 0); + Assert(prstate.lpdead_items == 0); + Assert(prstate.cutoffs); Assert(heap_page_is_all_visible(params->relation, buffer, prstate.cutoffs->OldestXmin, @@ -1183,56 +1237,35 @@ heap_page_prune_and_freeze(PruneFreezeParams *params, Assert(prstate.all_frozen == debug_all_frozen); Assert(!TransactionIdIsValid(debug_cutoff) || - debug_cutoff == vm_conflict_horizon); + debug_cutoff == prstate.visibility_cutoff_xid); } #endif - Assert(!prstate.all_frozen || prstate.all_visible); - Assert(!prstate.all_visible || (prstate.lpdead_items == 0)); - - /* - * Decide whether to set the page-level PD_ALL_VISIBLE bit and the VM bits - * based on information from the VM and the all_visible/all_frozen flags. - * - * While it is valid for PD_ALL_VISIBLE to be set when the corresponding - * VM bit is clear, we strongly prefer to keep them in sync. - * - * Accordingly, we also allow updating only the VM when PD_ALL_VISIBLE has - * already been set. Setting only the VM is most common when setting an - * already all-visible page all-frozen. - */ - do_set_vm = heap_page_will_set_vm(params->relation, - blockno, - buffer, - vmbuffer, - params->blk_known_av, - &prstate, - &presult->new_vmbits); + /* Copy information back for caller */ + presult->ndeleted = prstate.ndeleted; + presult->nnewlpdead = prstate.ndead; + presult->nfrozen = prstate.nfrozen; + presult->live_tuples = prstate.live_tuples; + presult->recently_dead_tuples = prstate.recently_dead_tuples; + presult->hastup = prstate.hastup; + presult->new_vmbits = new_vmbits; + presult->old_vmbits = old_vmbits; - /* - * new_vmbits should be 0 regardless of whether or not the page is - * all-visible if we do not intend to set the VM. - */ - Assert(do_set_vm || presult->new_vmbits == 0); + presult->lpdead_items = prstate.lpdead_items; + /* the presult->deadoffsets array was already filled in */ - if (do_set_vm) + if (prstate.attempt_freeze) { - /* - * It should never be the case that the visibility map page is set - * while the page-level bit is clear, but the reverse is allowed (if - * checksums are not enabled). - * - * The heap page is added to the WAL chain even if it wasn't modified, - * so we still need to mark it dirty. The only scenario where it isn't - * modified in phase I is when the VM was truncated or removed, which - * isn't worth optimizing for. - */ - PageSetAllVisible(page); - MarkBufferDirty(buffer); - presult->old_vmbits = visibilitymap_set(params->relation, blockno, buffer, - InvalidXLogRecPtr, - vmbuffer, vm_conflict_horizon, - presult->new_vmbits); + if (presult->nfrozen > 0) + { + *new_relfrozen_xid = prstate.pagefrz.FreezePageRelfrozenXid; + *new_relmin_mxid = prstate.pagefrz.FreezePageRelminMxid; + } + else + { + *new_relfrozen_xid = prstate.pagefrz.NoFreezePageRelfrozenXid; + *new_relmin_mxid = prstate.pagefrz.NoFreezePageRelminMxid; + } } } -- 2.43.0