mirror of
https://github.com/pgvector/pgvector.git
synced 2026-07-22 03:57:34 +08:00
Compare commits
1 Commits
fanalyzer
...
non-mvcc-s
| Author | SHA1 | Date | |
|---|---|---|---|
|
|
cb91de3332 |
4
.github/workflows/build.yml
vendored
4
.github/workflows/build.yml
vendored
@@ -39,10 +39,6 @@ jobs:
|
||||
sudo apt-get update
|
||||
sudo apt-get install libipc-run-perl
|
||||
- run: make prove_installcheck
|
||||
- if: ${{ matrix.os == 'ubuntu-22.04' }}
|
||||
run: make clean && make
|
||||
env:
|
||||
PG_CFLAGS: -Werror -fanalyzer
|
||||
mac:
|
||||
runs-on: macos-latest
|
||||
if: ${{ !startsWith(github.ref_name, 'windows') }}
|
||||
|
||||
@@ -1,6 +1,7 @@
|
||||
## 0.5.1 (unreleased)
|
||||
|
||||
- Improved performance of index scans for IVFFlat after updates and deletes
|
||||
- Fixed locking for index scans for HNSW
|
||||
|
||||
## 0.5.0 (2023-08-28)
|
||||
|
||||
|
||||
@@ -258,7 +258,7 @@ typedef struct HnswVacuumState
|
||||
/* Methods */
|
||||
int HnswGetM(Relation index);
|
||||
int HnswGetEfConstruction(Relation index);
|
||||
FmgrInfo *HnswOptionalProcInfo(Relation index, uint16 procnum);
|
||||
FmgrInfo *HnswOptionalProcInfo(Relation rel, uint16 procnum);
|
||||
bool HnswNormValue(FmgrInfo *procinfo, Oid collation, Datum *value, Vector * result);
|
||||
void HnswCommitBuffer(Buffer buf, GenericXLogState *state);
|
||||
Buffer HnswNewBuffer(Relation index, ForkNumber forkNum);
|
||||
|
||||
@@ -81,6 +81,7 @@ HnswBuildAppendPage(Relation index, Buffer *buf, Page *page, GenericXLogState **
|
||||
HnswPageGetOpaque(*page)->nextblkno = BufferGetBlockNumber(newbuf);
|
||||
|
||||
/* Commit */
|
||||
MarkBufferDirty(*buf);
|
||||
GenericXLogFinish(*state);
|
||||
UnlockReleaseBuffer(*buf);
|
||||
|
||||
@@ -178,6 +179,7 @@ CreateElementPages(HnswBuildState * buildstate)
|
||||
insertPage = BufferGetBlockNumber(buf);
|
||||
|
||||
/* Commit */
|
||||
MarkBufferDirty(buf);
|
||||
GenericXLogFinish(state);
|
||||
UnlockReleaseBuffer(buf);
|
||||
|
||||
@@ -225,6 +227,7 @@ CreateNeighborPages(HnswBuildState * buildstate)
|
||||
elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
|
||||
|
||||
/* Commit */
|
||||
MarkBufferDirty(buf);
|
||||
GenericXLogFinish(state);
|
||||
UnlockReleaseBuffer(buf);
|
||||
}
|
||||
|
||||
@@ -202,6 +202,8 @@ WriteNewElementPages(Relation index, HnswElement e, int m, BlockNumber insertPag
|
||||
HnswInsertAppendPage(index, &newbuf, &newpage, state, page);
|
||||
|
||||
/* Commit */
|
||||
MarkBufferDirty(newbuf);
|
||||
MarkBufferDirty(buf);
|
||||
GenericXLogFinish(state);
|
||||
|
||||
/* Unlock previous buffer */
|
||||
@@ -268,6 +270,9 @@ WriteNewElementPages(Relation index, HnswElement e, int m, BlockNumber insertPag
|
||||
}
|
||||
|
||||
/* Commit */
|
||||
MarkBufferDirty(buf);
|
||||
if (nbuf != buf)
|
||||
MarkBufferDirty(nbuf);
|
||||
GenericXLogFinish(state);
|
||||
UnlockReleaseBuffer(buf);
|
||||
if (nbuf != buf)
|
||||
@@ -386,6 +391,7 @@ HnswUpdateNeighborPages(Relation index, FmgrInfo *procinfo, Oid collation, HnswE
|
||||
elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
|
||||
|
||||
/* Commit */
|
||||
MarkBufferDirty(buf);
|
||||
GenericXLogFinish(state);
|
||||
}
|
||||
else
|
||||
@@ -439,6 +445,7 @@ HnswAddDuplicate(Relation index, HnswElement element, HnswElement dup)
|
||||
elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
|
||||
|
||||
/* Commit */
|
||||
MarkBufferDirty(buf);
|
||||
GenericXLogFinish(state);
|
||||
UnlockReleaseBuffer(buf);
|
||||
|
||||
|
||||
@@ -113,6 +113,12 @@ hnswbeginscan(Relation index, int nkeys, int norderbys)
|
||||
|
||||
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;
|
||||
}
|
||||
|
||||
@@ -163,17 +169,8 @@ hnswgettuple(IndexScanDesc scan, ScanDirection dir)
|
||||
/* Get scan value */
|
||||
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);
|
||||
|
||||
/* Release shared lock */
|
||||
UnlockPage(scan->indexRelation, HNSW_SCAN_LOCK, ShareLock);
|
||||
|
||||
so->first = false;
|
||||
}
|
||||
|
||||
@@ -226,6 +223,9 @@ hnswendscan(IndexScanDesc scan)
|
||||
{
|
||||
HnswScanOpaque so = (HnswScanOpaque) scan->opaque;
|
||||
|
||||
/* Release shared lock */
|
||||
UnlockPage(scan->indexRelation, HNSW_SCAN_LOCK, ShareLock);
|
||||
|
||||
MemoryContextDelete(so->tmpCtx);
|
||||
|
||||
pfree(so);
|
||||
|
||||
@@ -38,12 +38,12 @@ HnswGetEfConstruction(Relation index)
|
||||
* Get proc
|
||||
*/
|
||||
FmgrInfo *
|
||||
HnswOptionalProcInfo(Relation index, uint16 procnum)
|
||||
HnswOptionalProcInfo(Relation rel, uint16 procnum)
|
||||
{
|
||||
if (!OidIsValid(index_getprocid(index, 1, procnum)))
|
||||
if (!OidIsValid(index_getprocid(rel, 1, procnum)))
|
||||
return NULL;
|
||||
|
||||
return index_getprocinfo(index, 1, procnum);
|
||||
return index_getprocinfo(rel, 1, procnum);
|
||||
}
|
||||
|
||||
/*
|
||||
@@ -117,6 +117,7 @@ HnswInitRegisterPage(Relation index, Buffer *buf, Page *page, GenericXLogState *
|
||||
void
|
||||
HnswCommitBuffer(Buffer buf, GenericXLogState *state)
|
||||
{
|
||||
MarkBufferDirty(buf);
|
||||
GenericXLogFinish(state);
|
||||
UnlockReleaseBuffer(buf);
|
||||
}
|
||||
@@ -142,17 +143,6 @@ HnswInitNeighbors(HnswElement element, int m)
|
||||
}
|
||||
}
|
||||
|
||||
/*
|
||||
* Free neighbors
|
||||
*/
|
||||
static void
|
||||
HnswFreeNeighbors(HnswElement element)
|
||||
{
|
||||
for (int lc = 0; lc <= element->level; lc++)
|
||||
pfree(element->neighbors[lc].items);
|
||||
pfree(element->neighbors);
|
||||
}
|
||||
|
||||
/*
|
||||
* Allocate an element
|
||||
*/
|
||||
@@ -184,8 +174,10 @@ HnswInitElement(ItemPointer heaptid, int m, double ml, int maxLevel)
|
||||
void
|
||||
HnswFreeElement(HnswElement element)
|
||||
{
|
||||
HnswFreeNeighbors(element);
|
||||
list_free_deep(element->heaptids);
|
||||
for (int lc = 0; lc <= element->level; lc++)
|
||||
pfree(element->neighbors[lc].items);
|
||||
pfree(element->neighbors);
|
||||
pfree(element->vec);
|
||||
pfree(element);
|
||||
}
|
||||
|
||||
@@ -128,7 +128,10 @@ RemoveHeapTids(HnswVacuumState * vacuumstate)
|
||||
blkno = HnswPageGetOpaque(page)->nextblkno;
|
||||
|
||||
if (updated)
|
||||
{
|
||||
MarkBufferDirty(buf);
|
||||
GenericXLogFinish(state);
|
||||
}
|
||||
else
|
||||
GenericXLogAbort(state);
|
||||
|
||||
@@ -226,6 +229,7 @@ RepairGraphElement(HnswVacuumState * vacuumstate, HnswElement element, HnswEleme
|
||||
elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
|
||||
|
||||
/* Commit */
|
||||
MarkBufferDirty(buf);
|
||||
GenericXLogFinish(state);
|
||||
UnlockReleaseBuffer(buf);
|
||||
|
||||
@@ -543,6 +547,9 @@ MarkDeleted(HnswVacuumState * vacuumstate)
|
||||
elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
|
||||
|
||||
/* Commit */
|
||||
MarkBufferDirty(buf);
|
||||
if (nbuf != buf)
|
||||
MarkBufferDirty(nbuf);
|
||||
GenericXLogFinish(state);
|
||||
if (nbuf != buf)
|
||||
UnlockReleaseBuffer(nbuf);
|
||||
|
||||
@@ -506,30 +506,29 @@ CreateListPages(Relation index, VectorArray centers, int dimensions,
|
||||
Buffer buf;
|
||||
Page page;
|
||||
GenericXLogState *state;
|
||||
Size listSize;
|
||||
OffsetNumber offno;
|
||||
Size itemsz;
|
||||
IvfflatList list;
|
||||
|
||||
listSize = MAXALIGN(IVFFLAT_LIST_SIZE(dimensions));
|
||||
list = palloc(listSize);
|
||||
itemsz = MAXALIGN(IVFFLAT_LIST_SIZE(dimensions));
|
||||
list = palloc(itemsz);
|
||||
|
||||
buf = IvfflatNewBuffer(index, forkNum);
|
||||
IvfflatInitRegisterPage(index, &buf, &page, &state);
|
||||
|
||||
for (int i = 0; i < lists; i++)
|
||||
{
|
||||
OffsetNumber offno;
|
||||
|
||||
/* Load list */
|
||||
list->startPage = InvalidBlockNumber;
|
||||
list->insertPage = InvalidBlockNumber;
|
||||
memcpy(&list->center, VectorArrayGet(centers, i), VECTOR_SIZE(dimensions));
|
||||
|
||||
/* Ensure free space */
|
||||
if (PageGetFreeSpace(page) < listSize)
|
||||
if (PageGetFreeSpace(page) < itemsz)
|
||||
IvfflatAppendPage(index, &buf, &page, &state, forkNum);
|
||||
|
||||
/* Add the item */
|
||||
offno = PageAddItem(page, (Item) list, listSize, InvalidOffsetNumber, false, false);
|
||||
offno = PageAddItem(page, (Item) list, itemsz, InvalidOffsetNumber, false, false);
|
||||
if (offno == InvalidOffsetNumber)
|
||||
elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
|
||||
|
||||
|
||||
@@ -40,6 +40,9 @@
|
||||
#define IVFFLAT_METAPAGE_BLKNO 0
|
||||
#define IVFFLAT_HEAD_BLKNO 1 /* first list page */
|
||||
|
||||
/* Must correspond to page numbers since page lock is used */
|
||||
#define IVFFLAT_SCAN_LOCK 0
|
||||
|
||||
/* IVFFlat parameters */
|
||||
#define IVFFLAT_DEFAULT_LISTS 100
|
||||
#define IVFFLAT_MIN_LISTS 1
|
||||
@@ -246,6 +249,7 @@ typedef struct IvfflatScanOpaqueData
|
||||
int probes;
|
||||
int dimensions;
|
||||
bool first;
|
||||
bool hasLock;
|
||||
Buffer buf;
|
||||
ItemPointerData heaptid;
|
||||
|
||||
@@ -277,7 +281,7 @@ VectorArray VectorArrayInit(int maxlen, int dimensions);
|
||||
void VectorArrayFree(VectorArray arr);
|
||||
void PrintVectorArray(char *msg, VectorArray arr);
|
||||
void IvfflatKmeans(Relation index, VectorArray samples, VectorArray centers);
|
||||
FmgrInfo *IvfflatOptionalProcInfo(Relation index, uint16 procnum);
|
||||
FmgrInfo *IvfflatOptionalProcInfo(Relation rel, uint16 procnum);
|
||||
bool IvfflatNormValue(FmgrInfo *procinfo, Oid collation, Datum *value, Vector * result);
|
||||
int IvfflatGetLists(Relation index);
|
||||
void IvfflatGetMetaPageInfo(Relation index, int *lists, int *dimensions);
|
||||
|
||||
@@ -11,37 +11,36 @@
|
||||
* Find the list that minimizes the distance function
|
||||
*/
|
||||
static void
|
||||
FindInsertPage(Relation index, Datum *values, BlockNumber *insertPage, ListInfo * listInfo)
|
||||
FindInsertPage(Relation rel, Datum *values, BlockNumber *insertPage, ListInfo * listInfo)
|
||||
{
|
||||
Buffer cbuf;
|
||||
Page cpage;
|
||||
IvfflatList list;
|
||||
double distance;
|
||||
double minDistance = DBL_MAX;
|
||||
BlockNumber nextblkno = IVFFLAT_HEAD_BLKNO;
|
||||
FmgrInfo *procinfo;
|
||||
Oid collation;
|
||||
OffsetNumber offno;
|
||||
OffsetNumber maxoffno;
|
||||
|
||||
/* Avoid compiler warning */
|
||||
listInfo->blkno = nextblkno;
|
||||
listInfo->offno = FirstOffsetNumber;
|
||||
|
||||
procinfo = index_getprocinfo(index, 1, IVFFLAT_DISTANCE_PROC);
|
||||
collation = index->rd_indcollation[0];
|
||||
procinfo = index_getprocinfo(rel, 1, IVFFLAT_DISTANCE_PROC);
|
||||
collation = rel->rd_indcollation[0];
|
||||
|
||||
/* Search all list pages */
|
||||
while (BlockNumberIsValid(nextblkno))
|
||||
{
|
||||
Buffer cbuf;
|
||||
Page cpage;
|
||||
OffsetNumber maxoffno;
|
||||
|
||||
cbuf = ReadBuffer(index, nextblkno);
|
||||
cbuf = ReadBuffer(rel, nextblkno);
|
||||
LockBuffer(cbuf, BUFFER_LOCK_SHARE);
|
||||
cpage = BufferGetPage(cbuf);
|
||||
maxoffno = PageGetMaxOffsetNumber(cpage);
|
||||
|
||||
for (OffsetNumber offno = FirstOffsetNumber; offno <= maxoffno; offno = OffsetNumberNext(offno))
|
||||
for (offno = FirstOffsetNumber; offno <= maxoffno; offno = OffsetNumberNext(offno))
|
||||
{
|
||||
IvfflatList list;
|
||||
double distance;
|
||||
|
||||
list = (IvfflatList) PageGetItem(cpage, PageGetItemId(cpage, offno));
|
||||
distance = DatumGetFloat8(FunctionCall2Coll(procinfo, collation, values[0], PointerGetDatum(&list->center)));
|
||||
|
||||
@@ -64,7 +63,7 @@ FindInsertPage(Relation index, Datum *values, BlockNumber *insertPage, ListInfo
|
||||
* Insert a tuple into the index
|
||||
*/
|
||||
static void
|
||||
InsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heap_tid, Relation heapRel)
|
||||
InsertTuple(Relation rel, Datum *values, bool *isnull, ItemPointer heap_tid, Relation heapRel)
|
||||
{
|
||||
IndexTuple itup;
|
||||
Datum value;
|
||||
@@ -81,20 +80,20 @@ InsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heap_tid, R
|
||||
value = PointerGetDatum(PG_DETOAST_DATUM(values[0]));
|
||||
|
||||
/* Normalize if needed */
|
||||
normprocinfo = IvfflatOptionalProcInfo(index, IVFFLAT_NORM_PROC);
|
||||
normprocinfo = IvfflatOptionalProcInfo(rel, IVFFLAT_NORM_PROC);
|
||||
if (normprocinfo != NULL)
|
||||
{
|
||||
if (!IvfflatNormValue(normprocinfo, index->rd_indcollation[0], &value, NULL))
|
||||
if (!IvfflatNormValue(normprocinfo, rel->rd_indcollation[0], &value, NULL))
|
||||
return;
|
||||
}
|
||||
|
||||
/* Find the insert page - sets the page and list info */
|
||||
FindInsertPage(index, values, &insertPage, &listInfo);
|
||||
FindInsertPage(rel, values, &insertPage, &listInfo);
|
||||
Assert(BlockNumberIsValid(insertPage));
|
||||
originalInsertPage = insertPage;
|
||||
|
||||
/* Form tuple */
|
||||
itup = index_form_tuple(RelationGetDescr(index), &value, isnull);
|
||||
itup = index_form_tuple(RelationGetDescr(rel), &value, isnull);
|
||||
itup->t_tid = *heap_tid;
|
||||
|
||||
/* Get tuple size */
|
||||
@@ -104,10 +103,10 @@ InsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heap_tid, R
|
||||
/* Find a page to insert the item */
|
||||
for (;;)
|
||||
{
|
||||
buf = ReadBuffer(index, insertPage);
|
||||
buf = ReadBuffer(rel, insertPage);
|
||||
LockBuffer(buf, BUFFER_LOCK_EXCLUSIVE);
|
||||
|
||||
state = GenericXLogStart(index);
|
||||
state = GenericXLogStart(rel);
|
||||
page = GenericXLogRegisterBuffer(state, buf, 0);
|
||||
|
||||
if (PageGetFreeSpace(page) >= itemsz)
|
||||
@@ -127,9 +126,9 @@ InsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heap_tid, R
|
||||
Page newpage;
|
||||
|
||||
/* Add a new page */
|
||||
LockRelationForExtension(index, ExclusiveLock);
|
||||
newbuf = IvfflatNewBuffer(index, MAIN_FORKNUM);
|
||||
UnlockRelationForExtension(index, ExclusiveLock);
|
||||
LockRelationForExtension(rel, ExclusiveLock);
|
||||
newbuf = IvfflatNewBuffer(rel, MAIN_FORKNUM);
|
||||
UnlockRelationForExtension(rel, ExclusiveLock);
|
||||
|
||||
/* Init new page */
|
||||
newpage = GenericXLogRegisterBuffer(state, newbuf, GENERIC_XLOG_FULL_IMAGE);
|
||||
@@ -142,13 +141,15 @@ InsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heap_tid, R
|
||||
IvfflatPageGetOpaque(page)->nextblkno = insertPage;
|
||||
|
||||
/* Commit */
|
||||
MarkBufferDirty(newbuf);
|
||||
MarkBufferDirty(buf);
|
||||
GenericXLogFinish(state);
|
||||
|
||||
/* Unlock previous buffer */
|
||||
UnlockReleaseBuffer(buf);
|
||||
|
||||
/* Prepare new buffer */
|
||||
state = GenericXLogStart(index);
|
||||
state = GenericXLogStart(rel);
|
||||
buf = newbuf;
|
||||
page = GenericXLogRegisterBuffer(state, buf, 0);
|
||||
break;
|
||||
@@ -157,13 +158,13 @@ InsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heap_tid, R
|
||||
|
||||
/* Add to next offset */
|
||||
if (PageAddItem(page, (Item) itup, itemsz, InvalidOffsetNumber, false, false) == InvalidOffsetNumber)
|
||||
elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
|
||||
elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(rel));
|
||||
|
||||
IvfflatCommitBuffer(buf, state);
|
||||
|
||||
/* Update the insert page */
|
||||
if (insertPage != originalInsertPage)
|
||||
IvfflatUpdateList(index, listInfo, insertPage, originalInsertPage, InvalidBlockNumber, MAIN_FORKNUM);
|
||||
IvfflatUpdateList(rel, listInfo, insertPage, originalInsertPage, InvalidBlockNumber, MAIN_FORKNUM);
|
||||
}
|
||||
|
||||
/*
|
||||
|
||||
@@ -17,6 +17,10 @@ InitCenters(Relation index, VectorArray samples, VectorArray centers, float *low
|
||||
FmgrInfo *procinfo;
|
||||
Oid collation;
|
||||
int64 j;
|
||||
double distance;
|
||||
double sum;
|
||||
double choice;
|
||||
Vector *vec;
|
||||
float *weight = palloc(samples->length * sizeof(float));
|
||||
int numCenters = centers->maxlen;
|
||||
int numSamples = samples->length;
|
||||
@@ -29,21 +33,17 @@ InitCenters(Relation index, VectorArray samples, VectorArray centers, float *low
|
||||
centers->length++;
|
||||
|
||||
for (j = 0; j < numSamples; j++)
|
||||
weight[j] = FLT_MAX;
|
||||
weight[j] = DBL_MAX;
|
||||
|
||||
for (int i = 0; i < numCenters; i++)
|
||||
{
|
||||
double sum;
|
||||
double choice;
|
||||
|
||||
CHECK_FOR_INTERRUPTS();
|
||||
|
||||
sum = 0.0;
|
||||
|
||||
for (j = 0; j < numSamples; j++)
|
||||
{
|
||||
Vector *vec = VectorArrayGet(samples, j);
|
||||
double distance;
|
||||
vec = VectorArrayGet(samples, j);
|
||||
|
||||
/* Only need to compute distance for new center */
|
||||
/* TODO Use triangle inequality to reduce distance calculations */
|
||||
@@ -112,6 +112,7 @@ CompareVectors(const void *a, const void *b)
|
||||
static void
|
||||
QuickCenters(Relation index, VectorArray samples, VectorArray centers)
|
||||
{
|
||||
Vector *vec;
|
||||
int dimensions = centers->dim;
|
||||
Oid collation = index->rd_indcollation[0];
|
||||
FmgrInfo *normprocinfo = IvfflatOptionalProcInfo(index, IVFFLAT_KMEANS_NORM_PROC);
|
||||
@@ -122,7 +123,7 @@ QuickCenters(Relation index, VectorArray samples, VectorArray centers)
|
||||
qsort(samples->items, samples->length, VECTOR_SIZE(samples->dim), CompareVectors);
|
||||
for (int i = 0; i < samples->length; i++)
|
||||
{
|
||||
Vector *vec = VectorArrayGet(samples, i);
|
||||
vec = VectorArrayGet(samples, i);
|
||||
|
||||
if (i == 0 || CompareVectors(vec, VectorArrayGet(samples, i - 1)) != 0)
|
||||
{
|
||||
@@ -135,7 +136,7 @@ QuickCenters(Relation index, VectorArray samples, VectorArray centers)
|
||||
/* Fill remaining with random data */
|
||||
while (centers->length < centers->maxlen)
|
||||
{
|
||||
Vector *vec = VectorArrayGet(centers, centers->length);
|
||||
vec = VectorArrayGet(centers, centers->length);
|
||||
|
||||
SET_VARSIZE(vec, VECTOR_SIZE(dimensions));
|
||||
vec->dim = dimensions;
|
||||
@@ -167,6 +168,7 @@ ElkanKmeans(Relation index, VectorArray samples, VectorArray centers)
|
||||
Oid collation;
|
||||
Vector *vec;
|
||||
Vector *newCenter;
|
||||
int iteration;
|
||||
int64 j;
|
||||
int64 k;
|
||||
int dimensions = centers->dim;
|
||||
@@ -180,6 +182,14 @@ ElkanKmeans(Relation index, VectorArray samples, VectorArray centers)
|
||||
float *s;
|
||||
float *halfcdist;
|
||||
float *newcdist;
|
||||
int changes;
|
||||
double minDistance;
|
||||
int closestCenter;
|
||||
double distance;
|
||||
bool rj;
|
||||
bool rjreset;
|
||||
double dxcx;
|
||||
double dxc;
|
||||
|
||||
/* Calculate allocation sizes */
|
||||
Size samplesSize = VECTOR_ARRAY_SIZE(samples->maxlen, samples->dim);
|
||||
@@ -237,14 +247,14 @@ ElkanKmeans(Relation index, VectorArray samples, VectorArray centers)
|
||||
/* Assign each x to its closest initial center c(x) = argmin d(x,c) */
|
||||
for (j = 0; j < numSamples; j++)
|
||||
{
|
||||
float minDistance = FLT_MAX;
|
||||
int closestCenter = 0;
|
||||
minDistance = DBL_MAX;
|
||||
closestCenter = 0;
|
||||
|
||||
/* Find closest center */
|
||||
for (k = 0; k < numCenters; k++)
|
||||
{
|
||||
/* TODO Use Lemma 1 in k-means++ initialization */
|
||||
float distance = lowerBound[j * numCenters + k];
|
||||
distance = lowerBound[j * numCenters + k];
|
||||
|
||||
if (distance < minDistance)
|
||||
{
|
||||
@@ -258,14 +268,13 @@ ElkanKmeans(Relation index, VectorArray samples, VectorArray centers)
|
||||
}
|
||||
|
||||
/* Give 500 iterations to converge */
|
||||
for (int iteration = 0; iteration < 500; iteration++)
|
||||
for (iteration = 0; iteration < 500; iteration++)
|
||||
{
|
||||
int changes = 0;
|
||||
bool rjreset;
|
||||
|
||||
/* Can take a while, so ensure we can interrupt */
|
||||
CHECK_FOR_INTERRUPTS();
|
||||
|
||||
changes = 0;
|
||||
|
||||
/* Step 1: For all centers, compute distance */
|
||||
for (j = 0; j < numCenters; j++)
|
||||
{
|
||||
@@ -273,8 +282,7 @@ ElkanKmeans(Relation index, VectorArray samples, VectorArray centers)
|
||||
|
||||
for (k = j + 1; k < numCenters; k++)
|
||||
{
|
||||
float distance = 0.5 * DatumGetFloat8(FunctionCall2Coll(procinfo, collation, PointerGetDatum(vec), PointerGetDatum(VectorArrayGet(centers, k))));
|
||||
|
||||
distance = 0.5 * DatumGetFloat8(FunctionCall2Coll(procinfo, collation, PointerGetDatum(vec), PointerGetDatum(VectorArrayGet(centers, k))));
|
||||
halfcdist[j * numCenters + k] = distance;
|
||||
halfcdist[k * numCenters + j] = distance;
|
||||
}
|
||||
@@ -283,12 +291,10 @@ ElkanKmeans(Relation index, VectorArray samples, VectorArray centers)
|
||||
/* For all centers c, compute s(c) */
|
||||
for (j = 0; j < numCenters; j++)
|
||||
{
|
||||
float minDistance = FLT_MAX;
|
||||
minDistance = DBL_MAX;
|
||||
|
||||
for (k = 0; k < numCenters; k++)
|
||||
{
|
||||
float distance;
|
||||
|
||||
if (j == k)
|
||||
continue;
|
||||
|
||||
@@ -304,8 +310,6 @@ ElkanKmeans(Relation index, VectorArray samples, VectorArray centers)
|
||||
|
||||
for (j = 0; j < numSamples; j++)
|
||||
{
|
||||
bool rj;
|
||||
|
||||
/* Step 2: Identify all points x such that u(x) <= s(c(x)) */
|
||||
if (upperBound[j] <= s[closestCenters[j]])
|
||||
continue;
|
||||
@@ -314,8 +318,6 @@ ElkanKmeans(Relation index, VectorArray samples, VectorArray centers)
|
||||
|
||||
for (k = 0; k < numCenters; k++)
|
||||
{
|
||||
float dxcx;
|
||||
|
||||
/* Step 3: For all remaining points x and centers c */
|
||||
if (k == closestCenters[j])
|
||||
continue;
|
||||
@@ -345,7 +347,7 @@ ElkanKmeans(Relation index, VectorArray samples, VectorArray centers)
|
||||
/* Step 3b */
|
||||
if (dxcx > lowerBound[j * numCenters + k] || dxcx > halfcdist[closestCenters[j] * numCenters + k])
|
||||
{
|
||||
float dxc = DatumGetFloat8(FunctionCall2Coll(procinfo, collation, PointerGetDatum(vec), PointerGetDatum(VectorArrayGet(centers, k))));
|
||||
dxc = DatumGetFloat8(FunctionCall2Coll(procinfo, collation, PointerGetDatum(vec), PointerGetDatum(VectorArrayGet(centers, k))));
|
||||
|
||||
/* d(x,c) calculated */
|
||||
lowerBound[j * numCenters + k] = dxc;
|
||||
@@ -359,6 +361,7 @@ ElkanKmeans(Relation index, VectorArray samples, VectorArray centers)
|
||||
|
||||
changes++;
|
||||
}
|
||||
|
||||
}
|
||||
}
|
||||
}
|
||||
@@ -375,8 +378,6 @@ ElkanKmeans(Relation index, VectorArray samples, VectorArray centers)
|
||||
|
||||
for (j = 0; j < numSamples; j++)
|
||||
{
|
||||
int closestCenter;
|
||||
|
||||
vec = VectorArrayGet(samples, j);
|
||||
closestCenter = closestCenters[j];
|
||||
|
||||
@@ -425,7 +426,7 @@ ElkanKmeans(Relation index, VectorArray samples, VectorArray centers)
|
||||
{
|
||||
for (k = 0; k < numCenters; k++)
|
||||
{
|
||||
float distance = lowerBound[j * numCenters + k] - newcdist[k];
|
||||
distance = lowerBound[j * numCenters + k] - newcdist[k];
|
||||
|
||||
if (distance < 0)
|
||||
distance = 0;
|
||||
@@ -441,7 +442,7 @@ ElkanKmeans(Relation index, VectorArray samples, VectorArray centers)
|
||||
|
||||
/* Step 7 */
|
||||
for (j = 0; j < numCenters; j++)
|
||||
VectorArraySet(centers, j, VectorArrayGet(newCenters, j));
|
||||
memcpy(VectorArrayGet(centers, j), VectorArrayGet(newCenters, j), VECTOR_SIZE(dimensions));
|
||||
|
||||
if (changes == 0 && iteration != 0)
|
||||
break;
|
||||
@@ -464,6 +465,9 @@ static void
|
||||
CheckCenters(Relation index, VectorArray centers)
|
||||
{
|
||||
FmgrInfo *normprocinfo;
|
||||
Oid collation;
|
||||
Vector *vec;
|
||||
double norm;
|
||||
|
||||
if (centers->length != centers->maxlen)
|
||||
elog(ERROR, "Not enough centers. Please report a bug.");
|
||||
@@ -471,7 +475,7 @@ CheckCenters(Relation index, VectorArray centers)
|
||||
/* Ensure no NaN or infinite values */
|
||||
for (int i = 0; i < centers->length; i++)
|
||||
{
|
||||
Vector *vec = VectorArrayGet(centers, i);
|
||||
vec = VectorArrayGet(centers, i);
|
||||
|
||||
for (int j = 0; j < vec->dim; j++)
|
||||
{
|
||||
@@ -497,12 +501,11 @@ CheckCenters(Relation index, VectorArray centers)
|
||||
normprocinfo = IvfflatOptionalProcInfo(index, IVFFLAT_NORM_PROC);
|
||||
if (normprocinfo != NULL)
|
||||
{
|
||||
Oid collation = index->rd_indcollation[0];
|
||||
collation = index->rd_indcollation[0];
|
||||
|
||||
for (int i = 0; i < centers->length; i++)
|
||||
{
|
||||
double norm = DatumGetFloat8(FunctionCall1Coll(normprocinfo, collation, PointerGetDatum(VectorArrayGet(centers, i))));
|
||||
|
||||
norm = DatumGetFloat8(FunctionCall1Coll(normprocinfo, collation, PointerGetDatum(VectorArrayGet(centers, i))));
|
||||
if (norm == 0)
|
||||
elog(ERROR, "Zero norm detected. Please report a bug.");
|
||||
}
|
||||
|
||||
@@ -9,6 +9,7 @@
|
||||
#include "miscadmin.h"
|
||||
#include "pgstat.h"
|
||||
#include "storage/bufmgr.h"
|
||||
#include "storage/lmgr.h"
|
||||
|
||||
/*
|
||||
* Compare list distances
|
||||
@@ -263,6 +264,7 @@ ivfflatbeginscan(Relation index, int nkeys, int norderbys)
|
||||
so = (IvfflatScanOpaque) palloc(offsetof(IvfflatScanOpaqueData, lists) + probes * sizeof(IvfflatScanList));
|
||||
so->buf = InvalidBuffer;
|
||||
so->first = true;
|
||||
so->hasLock = false;
|
||||
ItemPointerSetInvalid(&so->heaptid);
|
||||
so->probes = probes;
|
||||
so->dimensions = dimensions;
|
||||
@@ -347,6 +349,13 @@ ivfflatgettuple(IndexScanDesc scan, ScanDirection dir)
|
||||
if (scan->orderByData == NULL)
|
||||
elog(ERROR, "cannot scan ivfflat index without order");
|
||||
|
||||
/* Get a shared lock for non-MVCC snapshots */
|
||||
if (!so->hasLock && !IsMVCCSnapshot(scan->xs_snapshot))
|
||||
{
|
||||
so->hasLock = true;
|
||||
LockPage(scan->indexRelation, IVFFLAT_SCAN_LOCK, ShareLock);
|
||||
}
|
||||
|
||||
if (scan->orderByData->sk_flags & SK_ISNULL)
|
||||
value = PointerGetDatum(InitVector(so->dimensions));
|
||||
else
|
||||
@@ -422,6 +431,10 @@ ivfflatendscan(IndexScanDesc scan)
|
||||
if (BufferIsValid(so->buf))
|
||||
ReleaseBuffer(so->buf);
|
||||
|
||||
/* Release lock */
|
||||
if (so->hasLock)
|
||||
UnlockPage(scan->indexRelation, IVFFLAT_SCAN_LOCK, ShareLock);
|
||||
|
||||
pairingheap_free(so->listQueue);
|
||||
tuplesort_end(so->sortstate);
|
||||
|
||||
|
||||
@@ -57,12 +57,12 @@ IvfflatGetLists(Relation index)
|
||||
* Get proc
|
||||
*/
|
||||
FmgrInfo *
|
||||
IvfflatOptionalProcInfo(Relation index, uint16 procnum)
|
||||
IvfflatOptionalProcInfo(Relation rel, uint16 procnum)
|
||||
{
|
||||
if (!OidIsValid(index_getprocid(index, 1, procnum)))
|
||||
if (!OidIsValid(index_getprocid(rel, 1, procnum)))
|
||||
return NULL;
|
||||
|
||||
return index_getprocinfo(index, 1, procnum);
|
||||
return index_getprocinfo(rel, 1, procnum);
|
||||
}
|
||||
|
||||
/*
|
||||
@@ -136,6 +136,7 @@ IvfflatInitRegisterPage(Relation index, Buffer *buf, Page *page, GenericXLogStat
|
||||
void
|
||||
IvfflatCommitBuffer(Buffer buf, GenericXLogState *state)
|
||||
{
|
||||
MarkBufferDirty(buf);
|
||||
GenericXLogFinish(state);
|
||||
UnlockReleaseBuffer(buf);
|
||||
}
|
||||
@@ -159,6 +160,8 @@ IvfflatAppendPage(Relation index, Buffer *buf, Page *page, GenericXLogState **st
|
||||
IvfflatInitPage(newbuf, newpage);
|
||||
|
||||
/* Commit */
|
||||
MarkBufferDirty(*buf);
|
||||
MarkBufferDirty(newbuf);
|
||||
GenericXLogFinish(*state);
|
||||
|
||||
/* Unlock */
|
||||
|
||||
@@ -3,6 +3,7 @@
|
||||
#include "commands/vacuum.h"
|
||||
#include "ivfflat.h"
|
||||
#include "storage/bufmgr.h"
|
||||
#include "storage/lmgr.h"
|
||||
|
||||
/*
|
||||
* Bulk delete tuples from the index
|
||||
@@ -65,14 +66,10 @@ ivfflatbulkdelete(IndexVacuumInfo *info, IndexBulkDeleteResult *stats,
|
||||
|
||||
vacuum_delay_point();
|
||||
|
||||
buf = ReadBufferExtended(index, MAIN_FORKNUM, searchPage, RBM_NORMAL, bas);
|
||||
/* Ensure no in-flight index scans for non-MVCC snapshots */
|
||||
LockPage(index, IVFFLAT_SCAN_LOCK, ExclusiveLock);
|
||||
|
||||
/*
|
||||
* ambulkdelete cannot delete entries from pages that are
|
||||
* pinned by other backends
|
||||
*
|
||||
* https://www.postgresql.org/docs/current/index-locking.html
|
||||
*/
|
||||
buf = ReadBufferExtended(index, MAIN_FORKNUM, searchPage, RBM_NORMAL, bas);
|
||||
LockBufferForCleanup(buf);
|
||||
|
||||
state = GenericXLogStart(index);
|
||||
@@ -107,12 +104,15 @@ ivfflatbulkdelete(IndexVacuumInfo *info, IndexBulkDeleteResult *stats,
|
||||
{
|
||||
/* Delete tuples */
|
||||
PageIndexMultiDelete(page, deletable, ndeletable);
|
||||
MarkBufferDirty(buf);
|
||||
GenericXLogFinish(state);
|
||||
}
|
||||
else
|
||||
GenericXLogAbort(state);
|
||||
|
||||
UnlockReleaseBuffer(buf);
|
||||
|
||||
UnlockPage(index, IVFFLAT_SCAN_LOCK, ExclusiveLock);
|
||||
}
|
||||
|
||||
/*
|
||||
|
||||
Reference in New Issue
Block a user