mirror of
https://github.com/pgvector/pgvector.git
synced 2026-07-22 20:15:46 +08:00
Compare commits
5 Commits
datum
...
hnsw-build
| Author | SHA1 | Date | |
|---|---|---|---|
|
|
5934aeb30d | ||
|
|
115c2bd91d | ||
|
|
97ac01773d | ||
|
|
8d7abb6590 | ||
|
|
b247b688a8 |
@@ -1,7 +1,6 @@
|
|||||||
## 0.5.1 (unreleased)
|
## 0.5.1 (unreleased)
|
||||||
|
|
||||||
- Improved performance of index scans for IVFFlat after updates and deletes
|
- Improved performance of index scans for IVFFlat after updates and deletes
|
||||||
- Fixed locking for index scans for HNSW
|
|
||||||
|
|
||||||
## 0.5.0 (2023-08-28)
|
## 0.5.0 (2023-08-28)
|
||||||
|
|
||||||
|
|||||||
@@ -57,7 +57,7 @@
|
|||||||
/* PROGRESS_CREATEIDX_SUBPHASE_INITIALIZE is 1 */
|
/* PROGRESS_CREATEIDX_SUBPHASE_INITIALIZE is 1 */
|
||||||
#define PROGRESS_HNSW_PHASE_LOAD 2
|
#define PROGRESS_HNSW_PHASE_LOAD 2
|
||||||
|
|
||||||
#define HNSW_ELEMENT_TUPLE_SIZE(_datum) MAXALIGN(offsetof(HnswElementTupleData, value) + VARSIZE_ANY(_datum))
|
#define HNSW_ELEMENT_TUPLE_SIZE(_dim) MAXALIGN(offsetof(HnswElementTupleData, vec) + VECTOR_SIZE(_dim))
|
||||||
#define HNSW_NEIGHBOR_TUPLE_SIZE(level, m) MAXALIGN(offsetof(HnswNeighborTupleData, indextids) + ((level) + 2) * (m) * sizeof(ItemPointerData))
|
#define HNSW_NEIGHBOR_TUPLE_SIZE(level, m) MAXALIGN(offsetof(HnswNeighborTupleData, indextids) + ((level) + 2) * (m) * sizeof(ItemPointerData))
|
||||||
|
|
||||||
#define HnswPageGetOpaque(page) ((HnswPageOpaque) PageGetSpecialPointer(page))
|
#define HnswPageGetOpaque(page) ((HnswPageOpaque) PageGetSpecialPointer(page))
|
||||||
@@ -96,13 +96,12 @@ typedef struct HnswElementData
|
|||||||
List *heaptids;
|
List *heaptids;
|
||||||
uint8 level;
|
uint8 level;
|
||||||
uint8 deleted;
|
uint8 deleted;
|
||||||
bool loaded;
|
|
||||||
HnswNeighborArray *neighbors;
|
HnswNeighborArray *neighbors;
|
||||||
BlockNumber blkno;
|
BlockNumber blkno;
|
||||||
OffsetNumber offno;
|
OffsetNumber offno;
|
||||||
OffsetNumber neighborOffno;
|
OffsetNumber neighborOffno;
|
||||||
BlockNumber neighborPage;
|
BlockNumber neighborPage;
|
||||||
Datum value;
|
Vector *vec;
|
||||||
} HnswElementData;
|
} HnswElementData;
|
||||||
|
|
||||||
typedef HnswElementData * HnswElement;
|
typedef HnswElementData * HnswElement;
|
||||||
@@ -111,11 +110,13 @@ typedef struct HnswCandidate
|
|||||||
{
|
{
|
||||||
HnswElement element;
|
HnswElement element;
|
||||||
float distance;
|
float distance;
|
||||||
|
bool closer;
|
||||||
} HnswCandidate;
|
} HnswCandidate;
|
||||||
|
|
||||||
typedef struct HnswNeighborArray
|
typedef struct HnswNeighborArray
|
||||||
{
|
{
|
||||||
int length;
|
int length;
|
||||||
|
bool closerSet;
|
||||||
HnswCandidate *items;
|
HnswCandidate *items;
|
||||||
} HnswNeighborArray;
|
} HnswNeighborArray;
|
||||||
|
|
||||||
@@ -201,7 +202,7 @@ typedef struct HnswElementTupleData
|
|||||||
ItemPointerData heaptids[HNSW_HEAPTIDS];
|
ItemPointerData heaptids[HNSW_HEAPTIDS];
|
||||||
ItemPointerData neighbortid;
|
ItemPointerData neighbortid;
|
||||||
uint16 unused2;
|
uint16 unused2;
|
||||||
char value[FLEXIBLE_ARRAY_MEMBER];
|
Vector vec;
|
||||||
} HnswElementTupleData;
|
} HnswElementTupleData;
|
||||||
|
|
||||||
typedef HnswElementTupleData * HnswElementTuple;
|
typedef HnswElementTupleData * HnswElementTuple;
|
||||||
|
|||||||
@@ -8,7 +8,6 @@
|
|||||||
#include "lib/pairingheap.h"
|
#include "lib/pairingheap.h"
|
||||||
#include "nodes/pg_list.h"
|
#include "nodes/pg_list.h"
|
||||||
#include "storage/bufmgr.h"
|
#include "storage/bufmgr.h"
|
||||||
#include "utils/datum.h"
|
|
||||||
#include "utils/memutils.h"
|
#include "utils/memutils.h"
|
||||||
|
|
||||||
#if PG_VERSION_NUM >= 140000
|
#if PG_VERSION_NUM >= 140000
|
||||||
@@ -82,7 +81,6 @@ HnswBuildAppendPage(Relation index, Buffer *buf, Page *page, GenericXLogState **
|
|||||||
HnswPageGetOpaque(*page)->nextblkno = BufferGetBlockNumber(newbuf);
|
HnswPageGetOpaque(*page)->nextblkno = BufferGetBlockNumber(newbuf);
|
||||||
|
|
||||||
/* Commit */
|
/* Commit */
|
||||||
MarkBufferDirty(*buf);
|
|
||||||
GenericXLogFinish(*state);
|
GenericXLogFinish(*state);
|
||||||
UnlockReleaseBuffer(*buf);
|
UnlockReleaseBuffer(*buf);
|
||||||
|
|
||||||
@@ -107,6 +105,8 @@ CreateElementPages(HnswBuildState * buildstate)
|
|||||||
{
|
{
|
||||||
Relation index = buildstate->index;
|
Relation index = buildstate->index;
|
||||||
ForkNumber forkNum = buildstate->forkNum;
|
ForkNumber forkNum = buildstate->forkNum;
|
||||||
|
int dimensions = buildstate->dimensions;
|
||||||
|
Size etupSize;
|
||||||
Size maxSize;
|
Size maxSize;
|
||||||
HnswElementTuple etup;
|
HnswElementTuple etup;
|
||||||
HnswNeighborTuple ntup;
|
HnswNeighborTuple ntup;
|
||||||
@@ -118,9 +118,10 @@ CreateElementPages(HnswBuildState * buildstate)
|
|||||||
|
|
||||||
/* Calculate sizes */
|
/* Calculate sizes */
|
||||||
maxSize = BLCKSZ - MAXALIGN(SizeOfPageHeaderData) - MAXALIGN(sizeof(HnswPageOpaqueData));
|
maxSize = BLCKSZ - MAXALIGN(SizeOfPageHeaderData) - MAXALIGN(sizeof(HnswPageOpaqueData));
|
||||||
|
etupSize = HNSW_ELEMENT_TUPLE_SIZE(dimensions);
|
||||||
|
|
||||||
/* Allocate once */
|
/* Allocate once */
|
||||||
etup = palloc0(maxSize);
|
etup = palloc0(etupSize);
|
||||||
ntup = palloc0(maxSize);
|
ntup = palloc0(maxSize);
|
||||||
|
|
||||||
/* Prepare first page */
|
/* Prepare first page */
|
||||||
@@ -132,14 +133,12 @@ CreateElementPages(HnswBuildState * buildstate)
|
|||||||
foreach(lc, buildstate->elements)
|
foreach(lc, buildstate->elements)
|
||||||
{
|
{
|
||||||
HnswElement element = lfirst(lc);
|
HnswElement element = lfirst(lc);
|
||||||
Size etupSize;
|
|
||||||
Size ntupSize;
|
Size ntupSize;
|
||||||
Size combinedSize;
|
Size combinedSize;
|
||||||
|
|
||||||
HnswSetElementTuple(etup, element);
|
HnswSetElementTuple(etup, element);
|
||||||
|
|
||||||
/* Calculate sizes */
|
/* Calculate sizes */
|
||||||
etupSize = HNSW_ELEMENT_TUPLE_SIZE(element->value);
|
|
||||||
ntupSize = HNSW_NEIGHBOR_TUPLE_SIZE(element->level, buildstate->m);
|
ntupSize = HNSW_NEIGHBOR_TUPLE_SIZE(element->level, buildstate->m);
|
||||||
combinedSize = etupSize + ntupSize + sizeof(ItemIdData);
|
combinedSize = etupSize + ntupSize + sizeof(ItemIdData);
|
||||||
|
|
||||||
@@ -179,7 +178,6 @@ CreateElementPages(HnswBuildState * buildstate)
|
|||||||
insertPage = BufferGetBlockNumber(buf);
|
insertPage = BufferGetBlockNumber(buf);
|
||||||
|
|
||||||
/* Commit */
|
/* Commit */
|
||||||
MarkBufferDirty(buf);
|
|
||||||
GenericXLogFinish(state);
|
GenericXLogFinish(state);
|
||||||
UnlockReleaseBuffer(buf);
|
UnlockReleaseBuffer(buf);
|
||||||
|
|
||||||
@@ -227,7 +225,6 @@ CreateNeighborPages(HnswBuildState * buildstate)
|
|||||||
elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
|
elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
|
||||||
|
|
||||||
/* Commit */
|
/* Commit */
|
||||||
MarkBufferDirty(buf);
|
|
||||||
GenericXLogFinish(state);
|
GenericXLogFinish(state);
|
||||||
UnlockReleaseBuffer(buf);
|
UnlockReleaseBuffer(buf);
|
||||||
}
|
}
|
||||||
@@ -276,15 +273,18 @@ InsertTuple(Relation index, Datum *values, HnswElement element, HnswBuildState *
|
|||||||
int m = buildstate->m;
|
int m = buildstate->m;
|
||||||
|
|
||||||
/* Detoast once for all calls */
|
/* Detoast once for all calls */
|
||||||
element->value = PointerGetDatum(PG_DETOAST_DATUM(values[0]));
|
Datum value = PointerGetDatum(PG_DETOAST_DATUM(values[0]));
|
||||||
|
|
||||||
/* Normalize if needed */
|
/* Normalize if needed */
|
||||||
if (buildstate->normprocinfo != NULL)
|
if (buildstate->normprocinfo != NULL)
|
||||||
{
|
{
|
||||||
if (!HnswNormValue(buildstate->normprocinfo, collation, &element->value, buildstate->normvec))
|
if (!HnswNormValue(buildstate->normprocinfo, collation, &value, buildstate->normvec))
|
||||||
return false;
|
return false;
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/* Copy value to element so accessible outside of memory context */
|
||||||
|
memcpy(element->vec, DatumGetVector(value), VECTOR_SIZE(buildstate->dimensions));
|
||||||
|
|
||||||
/* Insert element in graph */
|
/* Insert element in graph */
|
||||||
HnswInsertElement(element, entryPoint, NULL, procinfo, collation, m, efConstruction, false);
|
HnswInsertElement(element, entryPoint, NULL, procinfo, collation, m, efConstruction, false);
|
||||||
|
|
||||||
@@ -360,6 +360,7 @@ BuildCallback(Relation index, CALLBACK_ITEM_POINTER, Datum *values,
|
|||||||
|
|
||||||
/* Allocate necessary memory outside of memory context */
|
/* Allocate necessary memory outside of memory context */
|
||||||
element = HnswInitElement(tid, buildstate->m, buildstate->ml, buildstate->maxLevel);
|
element = HnswInitElement(tid, buildstate->m, buildstate->ml, buildstate->maxLevel);
|
||||||
|
element->vec = palloc(VECTOR_SIZE(buildstate->dimensions));
|
||||||
|
|
||||||
/* Use memory context since detoast can allocate */
|
/* Use memory context since detoast can allocate */
|
||||||
oldCtx = MemoryContextSwitchTo(buildstate->tmpCtx);
|
oldCtx = MemoryContextSwitchTo(buildstate->tmpCtx);
|
||||||
@@ -367,8 +368,9 @@ BuildCallback(Relation index, CALLBACK_ITEM_POINTER, Datum *values,
|
|||||||
/* Insert tuple */
|
/* Insert tuple */
|
||||||
inserted = InsertTuple(index, values, element, buildstate, &dup);
|
inserted = InsertTuple(index, values, element, buildstate, &dup);
|
||||||
|
|
||||||
/* Switch memory context */
|
/* Reset memory context */
|
||||||
MemoryContextSwitchTo(oldCtx);
|
MemoryContextSwitchTo(oldCtx);
|
||||||
|
MemoryContextReset(buildstate->tmpCtx);
|
||||||
|
|
||||||
/* Add outside memory context */
|
/* Add outside memory context */
|
||||||
if (dup != NULL)
|
if (dup != NULL)
|
||||||
@@ -376,16 +378,9 @@ BuildCallback(Relation index, CALLBACK_ITEM_POINTER, Datum *values,
|
|||||||
|
|
||||||
/* Add to buildstate or free */
|
/* Add to buildstate or free */
|
||||||
if (inserted)
|
if (inserted)
|
||||||
{
|
|
||||||
element->value = datumCopy(element->value, false, -1);
|
|
||||||
element->loaded = true;
|
|
||||||
buildstate->elements = lappend(buildstate->elements, element);
|
buildstate->elements = lappend(buildstate->elements, element);
|
||||||
}
|
|
||||||
else
|
else
|
||||||
HnswFreeElement(element);
|
HnswFreeElement(element);
|
||||||
|
|
||||||
/* Reset memory context */
|
|
||||||
MemoryContextReset(buildstate->tmpCtx);
|
|
||||||
}
|
}
|
||||||
|
|
||||||
/*
|
/*
|
||||||
|
|||||||
@@ -123,6 +123,7 @@ WriteNewElementPages(Relation index, HnswElement e, int m, BlockNumber insertPag
|
|||||||
Size minCombinedSize;
|
Size minCombinedSize;
|
||||||
HnswElementTuple etup;
|
HnswElementTuple etup;
|
||||||
BlockNumber currentPage = insertPage;
|
BlockNumber currentPage = insertPage;
|
||||||
|
int dimensions = e->vec->dim;
|
||||||
HnswNeighborTuple ntup;
|
HnswNeighborTuple ntup;
|
||||||
Buffer nbuf;
|
Buffer nbuf;
|
||||||
Page npage;
|
Page npage;
|
||||||
@@ -131,7 +132,7 @@ WriteNewElementPages(Relation index, HnswElement e, int m, BlockNumber insertPag
|
|||||||
BlockNumber newInsertPage = InvalidBlockNumber;
|
BlockNumber newInsertPage = InvalidBlockNumber;
|
||||||
|
|
||||||
/* Calculate sizes */
|
/* Calculate sizes */
|
||||||
etupSize = HNSW_ELEMENT_TUPLE_SIZE(e->value);
|
etupSize = HNSW_ELEMENT_TUPLE_SIZE(dimensions);
|
||||||
ntupSize = HNSW_NEIGHBOR_TUPLE_SIZE(e->level, m);
|
ntupSize = HNSW_NEIGHBOR_TUPLE_SIZE(e->level, m);
|
||||||
combinedSize = etupSize + ntupSize + sizeof(ItemIdData);
|
combinedSize = etupSize + ntupSize + sizeof(ItemIdData);
|
||||||
maxSize = BLCKSZ - MAXALIGN(SizeOfPageHeaderData) - MAXALIGN(sizeof(HnswPageOpaqueData));
|
maxSize = BLCKSZ - MAXALIGN(SizeOfPageHeaderData) - MAXALIGN(sizeof(HnswPageOpaqueData));
|
||||||
@@ -201,8 +202,6 @@ WriteNewElementPages(Relation index, HnswElement e, int m, BlockNumber insertPag
|
|||||||
HnswInsertAppendPage(index, &newbuf, &newpage, state, page);
|
HnswInsertAppendPage(index, &newbuf, &newpage, state, page);
|
||||||
|
|
||||||
/* Commit */
|
/* Commit */
|
||||||
MarkBufferDirty(newbuf);
|
|
||||||
MarkBufferDirty(buf);
|
|
||||||
GenericXLogFinish(state);
|
GenericXLogFinish(state);
|
||||||
|
|
||||||
/* Unlock previous buffer */
|
/* Unlock previous buffer */
|
||||||
@@ -269,9 +268,6 @@ WriteNewElementPages(Relation index, HnswElement e, int m, BlockNumber insertPag
|
|||||||
}
|
}
|
||||||
|
|
||||||
/* Commit */
|
/* Commit */
|
||||||
MarkBufferDirty(buf);
|
|
||||||
if (nbuf != buf)
|
|
||||||
MarkBufferDirty(nbuf);
|
|
||||||
GenericXLogFinish(state);
|
GenericXLogFinish(state);
|
||||||
UnlockReleaseBuffer(buf);
|
UnlockReleaseBuffer(buf);
|
||||||
if (nbuf != buf)
|
if (nbuf != buf)
|
||||||
@@ -390,7 +386,6 @@ HnswUpdateNeighborPages(Relation index, FmgrInfo *procinfo, Oid collation, HnswE
|
|||||||
elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
|
elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
|
||||||
|
|
||||||
/* Commit */
|
/* Commit */
|
||||||
MarkBufferDirty(buf);
|
|
||||||
GenericXLogFinish(state);
|
GenericXLogFinish(state);
|
||||||
}
|
}
|
||||||
else
|
else
|
||||||
@@ -410,7 +405,7 @@ HnswAddDuplicate(Relation index, HnswElement element, HnswElement dup)
|
|||||||
Buffer buf;
|
Buffer buf;
|
||||||
Page page;
|
Page page;
|
||||||
GenericXLogState *state;
|
GenericXLogState *state;
|
||||||
Size etupSize = HNSW_ELEMENT_TUPLE_SIZE(dup->value);
|
Size etupSize = HNSW_ELEMENT_TUPLE_SIZE(dup->vec->dim);
|
||||||
HnswElementTuple etup;
|
HnswElementTuple etup;
|
||||||
int i;
|
int i;
|
||||||
|
|
||||||
@@ -444,7 +439,6 @@ HnswAddDuplicate(Relation index, HnswElement element, HnswElement dup)
|
|||||||
elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
|
elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
|
||||||
|
|
||||||
/* Commit */
|
/* Commit */
|
||||||
MarkBufferDirty(buf);
|
|
||||||
GenericXLogFinish(state);
|
GenericXLogFinish(state);
|
||||||
UnlockReleaseBuffer(buf);
|
UnlockReleaseBuffer(buf);
|
||||||
|
|
||||||
@@ -521,7 +515,7 @@ HnswInsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heap_ti
|
|||||||
|
|
||||||
/* Create an element */
|
/* Create an element */
|
||||||
element = HnswInitElement(heap_tid, m, HnswGetMl(m), HnswGetMaxLevel(m));
|
element = HnswInitElement(heap_tid, m, HnswGetMl(m), HnswGetMaxLevel(m));
|
||||||
element->value = value;
|
element->vec = DatumGetVector(value);
|
||||||
|
|
||||||
/* Prevent concurrent inserts when likely updating entry point */
|
/* Prevent concurrent inserts when likely updating entry point */
|
||||||
if (entryPoint == NULL || element->level > entryPoint->level)
|
if (entryPoint == NULL || element->level > entryPoint->level)
|
||||||
|
|||||||
@@ -113,12 +113,6 @@ hnswbeginscan(Relation index, int nkeys, int norderbys)
|
|||||||
|
|
||||||
scan->opaque = so;
|
scan->opaque = so;
|
||||||
|
|
||||||
/*
|
|
||||||
* Get a shared lock. This allows vacuum to ensure no in-flight scans
|
|
||||||
* before marking tuples as deleted.
|
|
||||||
*/
|
|
||||||
LockPage(scan->indexRelation, HNSW_SCAN_LOCK, ShareLock);
|
|
||||||
|
|
||||||
return scan;
|
return scan;
|
||||||
}
|
}
|
||||||
|
|
||||||
@@ -169,8 +163,17 @@ hnswgettuple(IndexScanDesc scan, ScanDirection dir)
|
|||||||
/* Get scan value */
|
/* Get scan value */
|
||||||
value = GetScanValue(scan);
|
value = GetScanValue(scan);
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Get a shared lock. This allows vacuum to ensure no in-flight scans
|
||||||
|
* before marking tuples as deleted.
|
||||||
|
*/
|
||||||
|
LockPage(scan->indexRelation, HNSW_SCAN_LOCK, ShareLock);
|
||||||
|
|
||||||
so->w = GetScanItems(scan, value);
|
so->w = GetScanItems(scan, value);
|
||||||
|
|
||||||
|
/* Release shared lock */
|
||||||
|
UnlockPage(scan->indexRelation, HNSW_SCAN_LOCK, ShareLock);
|
||||||
|
|
||||||
so->first = false;
|
so->first = false;
|
||||||
}
|
}
|
||||||
|
|
||||||
@@ -223,9 +226,6 @@ hnswendscan(IndexScanDesc scan)
|
|||||||
{
|
{
|
||||||
HnswScanOpaque so = (HnswScanOpaque) scan->opaque;
|
HnswScanOpaque so = (HnswScanOpaque) scan->opaque;
|
||||||
|
|
||||||
/* Release shared lock */
|
|
||||||
UnlockPage(scan->indexRelation, HNSW_SCAN_LOCK, ShareLock);
|
|
||||||
|
|
||||||
MemoryContextDelete(so->tmpCtx);
|
MemoryContextDelete(so->tmpCtx);
|
||||||
|
|
||||||
pfree(so);
|
pfree(so);
|
||||||
|
|||||||
@@ -4,7 +4,6 @@
|
|||||||
|
|
||||||
#include "hnsw.h"
|
#include "hnsw.h"
|
||||||
#include "storage/bufmgr.h"
|
#include "storage/bufmgr.h"
|
||||||
#include "utils/datum.h"
|
|
||||||
#include "vector.h"
|
#include "vector.h"
|
||||||
|
|
||||||
/*
|
/*
|
||||||
@@ -118,7 +117,6 @@ HnswInitRegisterPage(Relation index, Buffer *buf, Page *page, GenericXLogState *
|
|||||||
void
|
void
|
||||||
HnswCommitBuffer(Buffer buf, GenericXLogState *state)
|
HnswCommitBuffer(Buffer buf, GenericXLogState *state)
|
||||||
{
|
{
|
||||||
MarkBufferDirty(buf);
|
|
||||||
GenericXLogFinish(state);
|
GenericXLogFinish(state);
|
||||||
UnlockReleaseBuffer(buf);
|
UnlockReleaseBuffer(buf);
|
||||||
}
|
}
|
||||||
@@ -141,6 +139,7 @@ HnswInitNeighbors(HnswElement element, int m)
|
|||||||
a = &element->neighbors[lc];
|
a = &element->neighbors[lc];
|
||||||
a->length = 0;
|
a->length = 0;
|
||||||
a->items = palloc(sizeof(HnswCandidate) * lm);
|
a->items = palloc(sizeof(HnswCandidate) * lm);
|
||||||
|
a->closerSet = false;
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
@@ -188,8 +187,7 @@ HnswFreeElement(HnswElement element)
|
|||||||
{
|
{
|
||||||
HnswFreeNeighbors(element);
|
HnswFreeNeighbors(element);
|
||||||
list_free_deep(element->heaptids);
|
list_free_deep(element->heaptids);
|
||||||
if (element->loaded)
|
pfree(element->vec);
|
||||||
pfree(DatumGetPointer(element->value));
|
|
||||||
pfree(element);
|
pfree(element);
|
||||||
}
|
}
|
||||||
|
|
||||||
@@ -216,7 +214,7 @@ HnswInitElementFromBlock(BlockNumber blkno, OffsetNumber offno)
|
|||||||
element->blkno = blkno;
|
element->blkno = blkno;
|
||||||
element->offno = offno;
|
element->offno = offno;
|
||||||
element->neighbors = NULL;
|
element->neighbors = NULL;
|
||||||
element->loaded = false;
|
element->vec = NULL;
|
||||||
return element;
|
return element;
|
||||||
}
|
}
|
||||||
|
|
||||||
@@ -326,7 +324,7 @@ HnswSetElementTuple(HnswElementTuple etup, HnswElement element)
|
|||||||
else
|
else
|
||||||
ItemPointerSetInvalid(&etup->heaptids[i]);
|
ItemPointerSetInvalid(&etup->heaptids[i]);
|
||||||
}
|
}
|
||||||
memcpy(&etup->value, DatumGetPointer(element->value), VARSIZE_ANY(element->value));
|
memcpy(&etup->vec, element->vec, VECTOR_SIZE(element->vec->dim));
|
||||||
}
|
}
|
||||||
|
|
||||||
/*
|
/*
|
||||||
@@ -449,10 +447,8 @@ HnswLoadElementFromTuple(HnswElement element, HnswElementTuple etup, bool loadHe
|
|||||||
|
|
||||||
if (loadVec)
|
if (loadVec)
|
||||||
{
|
{
|
||||||
Datum value = PointerGetDatum(&etup->value);
|
element->vec = palloc(VECTOR_SIZE(etup->vec.dim));
|
||||||
|
memcpy(element->vec, &etup->vec, VECTOR_SIZE(etup->vec.dim));
|
||||||
element->value = datumCopy(value, false, -1);
|
|
||||||
element->loaded = true;
|
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
@@ -480,7 +476,7 @@ HnswLoadElement(HnswElement element, float *distance, Datum *q, Relation index,
|
|||||||
|
|
||||||
/* Calculate distance */
|
/* Calculate distance */
|
||||||
if (distance != NULL)
|
if (distance != NULL)
|
||||||
*distance = (float) DatumGetFloat8(FunctionCall2Coll(procinfo, collation, *q, PointerGetDatum(&etup->value)));
|
*distance = (float) DatumGetFloat8(FunctionCall2Coll(procinfo, collation, *q, PointerGetDatum(&etup->vec)));
|
||||||
|
|
||||||
UnlockReleaseBuffer(buf);
|
UnlockReleaseBuffer(buf);
|
||||||
}
|
}
|
||||||
@@ -491,7 +487,7 @@ HnswLoadElement(HnswElement element, float *distance, Datum *q, Relation index,
|
|||||||
static float
|
static float
|
||||||
GetCandidateDistance(HnswCandidate * hc, Datum q, FmgrInfo *procinfo, Oid collation)
|
GetCandidateDistance(HnswCandidate * hc, Datum q, FmgrInfo *procinfo, Oid collation)
|
||||||
{
|
{
|
||||||
return DatumGetFloat8(FunctionCall2Coll(procinfo, collation, q, hc->element->value));
|
return DatumGetFloat8(FunctionCall2Coll(procinfo, collation, q, PointerGetDatum(hc->element->vec)));
|
||||||
}
|
}
|
||||||
|
|
||||||
/*
|
/*
|
||||||
@@ -726,7 +722,7 @@ HnswGetDistance(HnswElement a, HnswElement b, int lc, FmgrInfo *procinfo, Oid co
|
|||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
return DatumGetFloat8(FunctionCall2Coll(procinfo, collation, a->value, b->value));
|
return DatumGetFloat8(FunctionCall2Coll(procinfo, collation, PointerGetDatum(a->vec), PointerGetDatum(b->vec)));
|
||||||
}
|
}
|
||||||
|
|
||||||
/*
|
/*
|
||||||
@@ -753,11 +749,13 @@ CheckElementCloser(HnswCandidate * e, List *r, int lc, FmgrInfo *procinfo, Oid c
|
|||||||
* Algorithm 4 from paper
|
* Algorithm 4 from paper
|
||||||
*/
|
*/
|
||||||
static List *
|
static List *
|
||||||
SelectNeighbors(List *c, int m, int lc, FmgrInfo *procinfo, Oid collation, HnswCandidate * *pruned)
|
SelectNeighbors(List *c, int m, int lc, FmgrInfo *procinfo, Oid collation, HnswElement e2, HnswCandidate * newCandidate, HnswCandidate * *pruned)
|
||||||
{
|
{
|
||||||
List *r = NIL;
|
List *r = NIL;
|
||||||
List *w = list_copy(c);
|
List *w = list_copy(c);
|
||||||
pairingheap *wd;
|
pairingheap *wd;
|
||||||
|
bool mustCalculate = !e2->neighbors[lc].closerSet;
|
||||||
|
bool foundNew = false;
|
||||||
|
|
||||||
if (list_length(w) <= m)
|
if (list_length(w) <= m)
|
||||||
return w;
|
return w;
|
||||||
@@ -768,18 +766,40 @@ SelectNeighbors(List *c, int m, int lc, FmgrInfo *procinfo, Oid collation, HnswC
|
|||||||
{
|
{
|
||||||
/* Assumes w is already ordered desc */
|
/* Assumes w is already ordered desc */
|
||||||
HnswCandidate *e = llast(w);
|
HnswCandidate *e = llast(w);
|
||||||
bool closer;
|
|
||||||
|
|
||||||
w = list_delete_last(w);
|
w = list_delete_last(w);
|
||||||
|
|
||||||
closer = CheckElementCloser(e, r, lc, procinfo, collation);
|
/* Use previous state of r and wd to skip work when possible */
|
||||||
|
if (mustCalculate)
|
||||||
|
e->closer = CheckElementCloser(e, r, lc, procinfo, collation);
|
||||||
|
else if (foundNew)
|
||||||
|
{
|
||||||
|
/* If new or current candidate is not closer, no change in state */
|
||||||
|
if (newCandidate->closer && e->closer)
|
||||||
|
{
|
||||||
|
/* Only need to compare with new candidate */
|
||||||
|
float distance = HnswGetDistance(e->element, newCandidate->element, lc, procinfo, collation);
|
||||||
|
|
||||||
if (closer)
|
e->closer = e->distance < distance;
|
||||||
|
|
||||||
|
if (!e->closer)
|
||||||
|
mustCalculate = true;
|
||||||
|
}
|
||||||
|
}
|
||||||
|
else if (e == newCandidate)
|
||||||
|
{
|
||||||
|
e->closer = CheckElementCloser(e, r, lc, procinfo, collation);
|
||||||
|
foundNew = true;
|
||||||
|
}
|
||||||
|
|
||||||
|
if (e->closer)
|
||||||
r = lappend(r, e);
|
r = lappend(r, e);
|
||||||
else
|
else
|
||||||
pairingheap_add(wd, &(CreatePairingHeapNode(e)->ph_node));
|
pairingheap_add(wd, &(CreatePairingHeapNode(e)->ph_node));
|
||||||
}
|
}
|
||||||
|
|
||||||
|
e2->neighbors[lc].closerSet = true;
|
||||||
|
|
||||||
/* Keep pruned connections */
|
/* Keep pruned connections */
|
||||||
while (!pairingheap_is_empty(wd) && list_length(r) < m)
|
while (!pairingheap_is_empty(wd) && list_length(r) < m)
|
||||||
r = lappend(r, ((HnswPairingHeapNode *) pairingheap_remove_first(wd))->inner);
|
r = lappend(r, ((HnswPairingHeapNode *) pairingheap_remove_first(wd))->inner);
|
||||||
@@ -809,7 +829,7 @@ HnswFindDuplicate(HnswElement e)
|
|||||||
HnswCandidate *neighbor = &neighbors->items[i];
|
HnswCandidate *neighbor = &neighbors->items[i];
|
||||||
|
|
||||||
/* Exit early since ordered by distance */
|
/* Exit early since ordered by distance */
|
||||||
if (!datumIsEqual(e->value, neighbor->element->value, false, -1))
|
if (vector_cmp_internal(e->vec, neighbor->element->vec) != 0)
|
||||||
break;
|
break;
|
||||||
|
|
||||||
/* Check for space */
|
/* Check for space */
|
||||||
@@ -884,13 +904,13 @@ HnswUpdateConnection(HnswElement element, HnswCandidate * hc, int m, int lc, int
|
|||||||
/* Load elements on insert */
|
/* Load elements on insert */
|
||||||
if (index != NULL)
|
if (index != NULL)
|
||||||
{
|
{
|
||||||
Datum q = hc->element->value;
|
Datum q = PointerGetDatum(hc->element->vec);
|
||||||
|
|
||||||
for (int i = 0; i < currentNeighbors->length; i++)
|
for (int i = 0; i < currentNeighbors->length; i++)
|
||||||
{
|
{
|
||||||
HnswCandidate *hc3 = ¤tNeighbors->items[i];
|
HnswCandidate *hc3 = ¤tNeighbors->items[i];
|
||||||
|
|
||||||
if (!hc3->element->loaded)
|
if (hc3->element->vec == NULL)
|
||||||
HnswLoadElement(hc3->element, &hc3->distance, &q, index, procinfo, collation, true);
|
HnswLoadElement(hc3->element, &hc3->distance, &q, index, procinfo, collation, true);
|
||||||
else
|
else
|
||||||
hc3->distance = GetCandidateDistance(hc3, q, procinfo, collation);
|
hc3->distance = GetCandidateDistance(hc3, q, procinfo, collation);
|
||||||
@@ -914,7 +934,7 @@ HnswUpdateConnection(HnswElement element, HnswCandidate * hc, int m, int lc, int
|
|||||||
c = lappend(c, &hc2);
|
c = lappend(c, &hc2);
|
||||||
list_sort(c, CompareCandidateDistances);
|
list_sort(c, CompareCandidateDistances);
|
||||||
|
|
||||||
SelectNeighbors(c, m, lc, procinfo, collation, &pruned);
|
SelectNeighbors(c, m, lc, procinfo, collation, hc->element, &hc2, &pruned);
|
||||||
|
|
||||||
/* Should not happen */
|
/* Should not happen */
|
||||||
if (pruned == NULL)
|
if (pruned == NULL)
|
||||||
@@ -972,7 +992,7 @@ HnswInsertElement(HnswElement element, HnswElement entryPoint, Relation index, F
|
|||||||
List *w;
|
List *w;
|
||||||
int level = element->level;
|
int level = element->level;
|
||||||
int entryLevel;
|
int entryLevel;
|
||||||
Datum q = element->value;
|
Datum q = PointerGetDatum(element->vec);
|
||||||
HnswElement skipElement = existing ? element : NULL;
|
HnswElement skipElement = existing ? element : NULL;
|
||||||
|
|
||||||
/* No neighbors if no entry point */
|
/* No neighbors if no entry point */
|
||||||
@@ -1013,7 +1033,7 @@ HnswInsertElement(HnswElement element, HnswElement entryPoint, Relation index, F
|
|||||||
else
|
else
|
||||||
lw = w;
|
lw = w;
|
||||||
|
|
||||||
neighbors = SelectNeighbors(lw, lm, lc, procinfo, collation, NULL);
|
neighbors = SelectNeighbors(lw, lm, lc, procinfo, collation, element, NULL, NULL);
|
||||||
|
|
||||||
AddConnections(element, neighbors, lm, lc);
|
AddConnections(element, neighbors, lm, lc);
|
||||||
|
|
||||||
|
|||||||
@@ -93,7 +93,7 @@ RemoveHeapTids(HnswVacuumState * vacuumstate)
|
|||||||
|
|
||||||
if (itemUpdated)
|
if (itemUpdated)
|
||||||
{
|
{
|
||||||
Size etupSize = HNSW_ELEMENT_TUPLE_SIZE(PointerGetDatum(&etup->value));
|
Size etupSize = HNSW_ELEMENT_TUPLE_SIZE(etup->vec.dim);
|
||||||
|
|
||||||
/* Mark rest as invalid */
|
/* Mark rest as invalid */
|
||||||
for (int i = idx; i < HNSW_HEAPTIDS; i++)
|
for (int i = idx; i < HNSW_HEAPTIDS; i++)
|
||||||
@@ -128,10 +128,7 @@ RemoveHeapTids(HnswVacuumState * vacuumstate)
|
|||||||
blkno = HnswPageGetOpaque(page)->nextblkno;
|
blkno = HnswPageGetOpaque(page)->nextblkno;
|
||||||
|
|
||||||
if (updated)
|
if (updated)
|
||||||
{
|
|
||||||
MarkBufferDirty(buf);
|
|
||||||
GenericXLogFinish(state);
|
GenericXLogFinish(state);
|
||||||
}
|
|
||||||
else
|
else
|
||||||
GenericXLogAbort(state);
|
GenericXLogAbort(state);
|
||||||
|
|
||||||
@@ -229,7 +226,6 @@ RepairGraphElement(HnswVacuumState * vacuumstate, HnswElement element, HnswEleme
|
|||||||
elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
|
elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
|
||||||
|
|
||||||
/* Commit */
|
/* Commit */
|
||||||
MarkBufferDirty(buf);
|
|
||||||
GenericXLogFinish(state);
|
GenericXLogFinish(state);
|
||||||
UnlockReleaseBuffer(buf);
|
UnlockReleaseBuffer(buf);
|
||||||
|
|
||||||
@@ -485,7 +481,6 @@ MarkDeleted(HnswVacuumState * vacuumstate)
|
|||||||
HnswNeighborTuple ntup;
|
HnswNeighborTuple ntup;
|
||||||
Size etupSize;
|
Size etupSize;
|
||||||
Size ntupSize;
|
Size ntupSize;
|
||||||
Datum value;
|
|
||||||
Buffer nbuf;
|
Buffer nbuf;
|
||||||
Page npage;
|
Page npage;
|
||||||
BlockNumber neighborPage;
|
BlockNumber neighborPage;
|
||||||
@@ -509,11 +504,8 @@ MarkDeleted(HnswVacuumState * vacuumstate)
|
|||||||
if (ItemPointerIsValid(&etup->heaptids[0]))
|
if (ItemPointerIsValid(&etup->heaptids[0]))
|
||||||
continue;
|
continue;
|
||||||
|
|
||||||
/* Get datum */
|
|
||||||
value = PointerGetDatum(&etup->value);
|
|
||||||
|
|
||||||
/* Calculate sizes */
|
/* Calculate sizes */
|
||||||
etupSize = HNSW_ELEMENT_TUPLE_SIZE(value);
|
etupSize = HNSW_ELEMENT_TUPLE_SIZE(etup->vec.dim);
|
||||||
ntupSize = HNSW_NEIGHBOR_TUPLE_SIZE(etup->level, vacuumstate->m);
|
ntupSize = HNSW_NEIGHBOR_TUPLE_SIZE(etup->level, vacuumstate->m);
|
||||||
|
|
||||||
/* Get neighbor page */
|
/* Get neighbor page */
|
||||||
@@ -536,7 +528,7 @@ MarkDeleted(HnswVacuumState * vacuumstate)
|
|||||||
|
|
||||||
/* Overwrite element */
|
/* Overwrite element */
|
||||||
etup->deleted = 1;
|
etup->deleted = 1;
|
||||||
MemSet(&etup->value, 0, VARSIZE_ANY(value));
|
MemSet(&etup->vec.x, 0, etup->vec.dim * sizeof(float));
|
||||||
|
|
||||||
/* Overwrite neighbors */
|
/* Overwrite neighbors */
|
||||||
for (int i = 0; i < ntup->count; i++)
|
for (int i = 0; i < ntup->count; i++)
|
||||||
@@ -551,9 +543,6 @@ MarkDeleted(HnswVacuumState * vacuumstate)
|
|||||||
elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
|
elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
|
||||||
|
|
||||||
/* Commit */
|
/* Commit */
|
||||||
MarkBufferDirty(buf);
|
|
||||||
if (nbuf != buf)
|
|
||||||
MarkBufferDirty(nbuf);
|
|
||||||
GenericXLogFinish(state);
|
GenericXLogFinish(state);
|
||||||
if (nbuf != buf)
|
if (nbuf != buf)
|
||||||
UnlockReleaseBuffer(nbuf);
|
UnlockReleaseBuffer(nbuf);
|
||||||
|
|||||||
@@ -506,9 +506,11 @@ CreateListPages(Relation index, VectorArray centers, int dimensions,
|
|||||||
Buffer buf;
|
Buffer buf;
|
||||||
Page page;
|
Page page;
|
||||||
GenericXLogState *state;
|
GenericXLogState *state;
|
||||||
|
Size listSize;
|
||||||
IvfflatList list;
|
IvfflatList list;
|
||||||
|
|
||||||
list = palloc0(BLCKSZ);
|
listSize = MAXALIGN(IVFFLAT_LIST_SIZE(dimensions));
|
||||||
|
list = palloc(listSize);
|
||||||
|
|
||||||
buf = IvfflatNewBuffer(index, forkNum);
|
buf = IvfflatNewBuffer(index, forkNum);
|
||||||
IvfflatInitRegisterPage(index, &buf, &page, &state);
|
IvfflatInitRegisterPage(index, &buf, &page, &state);
|
||||||
@@ -516,13 +518,11 @@ CreateListPages(Relation index, VectorArray centers, int dimensions,
|
|||||||
for (int i = 0; i < lists; i++)
|
for (int i = 0; i < lists; i++)
|
||||||
{
|
{
|
||||||
OffsetNumber offno;
|
OffsetNumber offno;
|
||||||
Datum center = PointerGetDatum(VectorArrayGet(centers, i));
|
|
||||||
Size listSize = MAXALIGN(IVFFLAT_LIST_SIZE(center));
|
|
||||||
|
|
||||||
/* Load list */
|
/* Load list */
|
||||||
list->startPage = InvalidBlockNumber;
|
list->startPage = InvalidBlockNumber;
|
||||||
list->insertPage = InvalidBlockNumber;
|
list->insertPage = InvalidBlockNumber;
|
||||||
memcpy(&list->center, DatumGetPointer(center), VARSIZE_ANY(center));
|
memcpy(&list->center, VectorArrayGet(centers, i), VECTOR_SIZE(dimensions));
|
||||||
|
|
||||||
/* Ensure free space */
|
/* Ensure free space */
|
||||||
if (PageGetFreeSpace(page) < listSize)
|
if (PageGetFreeSpace(page) < listSize)
|
||||||
|
|||||||
@@ -52,7 +52,7 @@
|
|||||||
#define PROGRESS_IVFFLAT_PHASE_ASSIGN 3
|
#define PROGRESS_IVFFLAT_PHASE_ASSIGN 3
|
||||||
#define PROGRESS_IVFFLAT_PHASE_LOAD 4
|
#define PROGRESS_IVFFLAT_PHASE_LOAD 4
|
||||||
|
|
||||||
#define IVFFLAT_LIST_SIZE(_datum) (offsetof(IvfflatListData, center) + VARSIZE_ANY(_datum))
|
#define IVFFLAT_LIST_SIZE(_dim) (offsetof(IvfflatListData, center) + VECTOR_SIZE(_dim))
|
||||||
|
|
||||||
#define IvfflatPageGetOpaque(page) ((IvfflatPageOpaque) PageGetSpecialPointer(page))
|
#define IvfflatPageGetOpaque(page) ((IvfflatPageOpaque) PageGetSpecialPointer(page))
|
||||||
#define IvfflatPageGetMeta(page) ((IvfflatMetaPageData *) PageGetContents(page))
|
#define IvfflatPageGetMeta(page) ((IvfflatMetaPageData *) PageGetContents(page))
|
||||||
@@ -229,7 +229,7 @@ typedef struct IvfflatListData
|
|||||||
{
|
{
|
||||||
BlockNumber startPage;
|
BlockNumber startPage;
|
||||||
BlockNumber insertPage;
|
BlockNumber insertPage;
|
||||||
char center[FLEXIBLE_ARRAY_MEMBER];
|
Vector center;
|
||||||
} IvfflatListData;
|
} IvfflatListData;
|
||||||
|
|
||||||
typedef IvfflatListData * IvfflatList;
|
typedef IvfflatListData * IvfflatList;
|
||||||
|
|||||||
@@ -142,8 +142,6 @@ InsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heap_tid, R
|
|||||||
IvfflatPageGetOpaque(page)->nextblkno = insertPage;
|
IvfflatPageGetOpaque(page)->nextblkno = insertPage;
|
||||||
|
|
||||||
/* Commit */
|
/* Commit */
|
||||||
MarkBufferDirty(newbuf);
|
|
||||||
MarkBufferDirty(buf);
|
|
||||||
GenericXLogFinish(state);
|
GenericXLogFinish(state);
|
||||||
|
|
||||||
/* Unlock previous buffer */
|
/* Unlock previous buffer */
|
||||||
|
|||||||
@@ -136,7 +136,6 @@ IvfflatInitRegisterPage(Relation index, Buffer *buf, Page *page, GenericXLogStat
|
|||||||
void
|
void
|
||||||
IvfflatCommitBuffer(Buffer buf, GenericXLogState *state)
|
IvfflatCommitBuffer(Buffer buf, GenericXLogState *state)
|
||||||
{
|
{
|
||||||
MarkBufferDirty(buf);
|
|
||||||
GenericXLogFinish(state);
|
GenericXLogFinish(state);
|
||||||
UnlockReleaseBuffer(buf);
|
UnlockReleaseBuffer(buf);
|
||||||
}
|
}
|
||||||
@@ -160,8 +159,6 @@ IvfflatAppendPage(Relation index, Buffer *buf, Page *page, GenericXLogState **st
|
|||||||
IvfflatInitPage(newbuf, newpage);
|
IvfflatInitPage(newbuf, newpage);
|
||||||
|
|
||||||
/* Commit */
|
/* Commit */
|
||||||
MarkBufferDirty(*buf);
|
|
||||||
MarkBufferDirty(newbuf);
|
|
||||||
GenericXLogFinish(*state);
|
GenericXLogFinish(*state);
|
||||||
|
|
||||||
/* Unlock */
|
/* Unlock */
|
||||||
|
|||||||
@@ -107,7 +107,6 @@ ivfflatbulkdelete(IndexVacuumInfo *info, IndexBulkDeleteResult *stats,
|
|||||||
{
|
{
|
||||||
/* Delete tuples */
|
/* Delete tuples */
|
||||||
PageIndexMultiDelete(page, deletable, ndeletable);
|
PageIndexMultiDelete(page, deletable, ndeletable);
|
||||||
MarkBufferDirty(buf);
|
|
||||||
GenericXLogFinish(state);
|
GenericXLogFinish(state);
|
||||||
}
|
}
|
||||||
else
|
else
|
||||||
|
|||||||
Reference in New Issue
Block a user