postgresql/src/include/access/gin_private.h
Alexander Korotkov 4b754d6c16 Avoid full scan of GIN indexes when possible
The strategy of GIN index scan is driven by opclass-specific extract_query
method.  This method that needed search mode is GIN_SEARCH_MODE_ALL.  This
mode means that matching tuple may contain none of extracted entries.  Simple
example is '!term' tsquery, which doesn't need any term to exist in matching
tsvector.

In order to handle such scan key GIN calculates virtual entry, which contains
all TIDs of all entries of attribute.  In fact this is full scan of index
attribute.  And typically this is very slow, but allows to handle some queries
correctly in GIN.  However, current algorithm calculate such virtual entry for
each GIN_SEARCH_MODE_ALL scan key even if they are multiple for the same
attribute.  This is clearly not optimal.

This commit improves the situation by introduction of "exclude only" scan keys.
Such scan keys are not capable to return set of matching TIDs.  Instead, they
are capable only to filter TIDs produced by normal scan keys.  Therefore,
each attribute should contain at least one normal scan key, while rest of them
may be "exclude only" if search mode is GIN_SEARCH_MODE_ALL.

The same optimization might be applied to the whole scan, not per-attribute.
But that leads to NULL values elimination problem.  There is trade-off between
multiple possible ways to do this.  We probably want to do this later using
some cost-based decision algorithm.

Discussion: https://postgr.es/m/CAOBaU_YGP5-BEt5Cc0%3DzMve92vocPzD%2BXiZgiZs1kjY0cj%3DXBg%40mail.gmail.com
Author: Nikita Glukhov, Alexander Korotkov, Tom Lane, Julien Rouhaud
Reviewed-by: Julien Rouhaud, Tomas Vondra, Tom Lane
2020-01-18 01:11:39 +03:00

496 lines
17 KiB
C

/*--------------------------------------------------------------------------
* gin_private.h
* header file for postgres inverted index access method implementation.
*
* Copyright (c) 2006-2020, PostgreSQL Global Development Group
*
* src/include/access/gin_private.h
*--------------------------------------------------------------------------
*/
#ifndef GIN_PRIVATE_H
#define GIN_PRIVATE_H
#include "access/amapi.h"
#include "access/gin.h"
#include "access/ginblock.h"
#include "access/itup.h"
#include "catalog/pg_am_d.h"
#include "fmgr.h"
#include "lib/rbtree.h"
#include "storage/bufmgr.h"
/*
* Storage type for GIN's reloptions
*/
typedef struct GinOptions
{
int32 vl_len_; /* varlena header (do not touch directly!) */
bool useFastUpdate; /* use fast updates? */
int pendingListCleanupSize; /* maximum size of pending list */
} GinOptions;
#define GIN_DEFAULT_USE_FASTUPDATE true
#define GinGetUseFastUpdate(relation) \
(AssertMacro(relation->rd_rel->relkind == RELKIND_INDEX && \
relation->rd_rel->relam == GIN_AM_OID), \
(relation)->rd_options ? \
((GinOptions *) (relation)->rd_options)->useFastUpdate : GIN_DEFAULT_USE_FASTUPDATE)
#define GinGetPendingListCleanupSize(relation) \
(AssertMacro(relation->rd_rel->relkind == RELKIND_INDEX && \
relation->rd_rel->relam == GIN_AM_OID), \
(relation)->rd_options && \
((GinOptions *) (relation)->rd_options)->pendingListCleanupSize != -1 ? \
((GinOptions *) (relation)->rd_options)->pendingListCleanupSize : \
gin_pending_list_limit)
/* Macros for buffer lock/unlock operations */
#define GIN_UNLOCK BUFFER_LOCK_UNLOCK
#define GIN_SHARE BUFFER_LOCK_SHARE
#define GIN_EXCLUSIVE BUFFER_LOCK_EXCLUSIVE
/*
* GinState: working data structure describing the index being worked on
*/
typedef struct GinState
{
Relation index;
bool oneCol; /* true if single-column index */
/*
* origTupdesc is the nominal tuple descriptor of the index, ie, the i'th
* attribute shows the key type (not the input data type!) of the i'th
* index column. In a single-column index this describes the actual leaf
* index tuples. In a multi-column index, the actual leaf tuples contain
* a smallint column number followed by a key datum of the appropriate
* type for that column. We set up tupdesc[i] to describe the actual
* rowtype of the index tuples for the i'th column, ie, (int2, keytype).
* Note that in any case, leaf tuples contain more data than is known to
* the TupleDesc; see access/gin/README for details.
*/
TupleDesc origTupdesc;
TupleDesc tupdesc[INDEX_MAX_KEYS];
/*
* Per-index-column opclass support functions
*/
FmgrInfo compareFn[INDEX_MAX_KEYS];
FmgrInfo extractValueFn[INDEX_MAX_KEYS];
FmgrInfo extractQueryFn[INDEX_MAX_KEYS];
FmgrInfo consistentFn[INDEX_MAX_KEYS];
FmgrInfo triConsistentFn[INDEX_MAX_KEYS];
FmgrInfo comparePartialFn[INDEX_MAX_KEYS]; /* optional method */
/* canPartialMatch[i] is true if comparePartialFn[i] is valid */
bool canPartialMatch[INDEX_MAX_KEYS];
/* Collations to pass to the support functions */
Oid supportCollation[INDEX_MAX_KEYS];
} GinState;
/* ginutil.c */
extern bytea *ginoptions(Datum reloptions, bool validate);
extern void initGinState(GinState *state, Relation index);
extern Buffer GinNewBuffer(Relation index);
extern void GinInitBuffer(Buffer b, uint32 f);
extern void GinInitPage(Page page, uint32 f, Size pageSize);
extern void GinInitMetabuffer(Buffer b);
extern int ginCompareEntries(GinState *ginstate, OffsetNumber attnum,
Datum a, GinNullCategory categorya,
Datum b, GinNullCategory categoryb);
extern int ginCompareAttEntries(GinState *ginstate,
OffsetNumber attnuma, Datum a, GinNullCategory categorya,
OffsetNumber attnumb, Datum b, GinNullCategory categoryb);
extern Datum *ginExtractEntries(GinState *ginstate, OffsetNumber attnum,
Datum value, bool isNull,
int32 *nentries, GinNullCategory **categories);
extern OffsetNumber gintuple_get_attrnum(GinState *ginstate, IndexTuple tuple);
extern Datum gintuple_get_key(GinState *ginstate, IndexTuple tuple,
GinNullCategory *category);
/* gininsert.c */
extern IndexBuildResult *ginbuild(Relation heap, Relation index,
struct IndexInfo *indexInfo);
extern void ginbuildempty(Relation index);
extern bool gininsert(Relation index, Datum *values, bool *isnull,
ItemPointer ht_ctid, Relation heapRel,
IndexUniqueCheck checkUnique,
struct IndexInfo *indexInfo);
extern void ginEntryInsert(GinState *ginstate,
OffsetNumber attnum, Datum key, GinNullCategory category,
ItemPointerData *items, uint32 nitem,
GinStatsData *buildStats);
/* ginbtree.c */
typedef struct GinBtreeStack
{
BlockNumber blkno;
Buffer buffer;
OffsetNumber off;
ItemPointerData iptr;
/* predictNumber contains predicted number of pages on current level */
uint32 predictNumber;
struct GinBtreeStack *parent;
} GinBtreeStack;
typedef struct GinBtreeData *GinBtree;
/* Return codes for GinBtreeData.beginPlaceToPage method */
typedef enum
{
GPTP_NO_WORK,
GPTP_INSERT,
GPTP_SPLIT
} GinPlaceToPageRC;
typedef struct GinBtreeData
{
/* search methods */
BlockNumber (*findChildPage) (GinBtree, GinBtreeStack *);
BlockNumber (*getLeftMostChild) (GinBtree, Page);
bool (*isMoveRight) (GinBtree, Page);
bool (*findItem) (GinBtree, GinBtreeStack *);
/* insert methods */
OffsetNumber (*findChildPtr) (GinBtree, Page, BlockNumber, OffsetNumber);
GinPlaceToPageRC (*beginPlaceToPage) (GinBtree, Buffer, GinBtreeStack *, void *, BlockNumber, void **, Page *, Page *);
void (*execPlaceToPage) (GinBtree, Buffer, GinBtreeStack *, void *, BlockNumber, void *);
void *(*prepareDownlink) (GinBtree, Buffer);
void (*fillRoot) (GinBtree, Page, BlockNumber, Page, BlockNumber, Page);
bool isData;
Relation index;
BlockNumber rootBlkno;
GinState *ginstate; /* not valid in a data scan */
bool fullScan;
bool isBuild;
/* Search key for Entry tree */
OffsetNumber entryAttnum;
Datum entryKey;
GinNullCategory entryCategory;
/* Search key for data tree (posting tree) */
ItemPointerData itemptr;
} GinBtreeData;
/* This represents a tuple to be inserted to entry tree. */
typedef struct
{
IndexTuple entry; /* tuple to insert */
bool isDelete; /* delete old tuple at same offset? */
} GinBtreeEntryInsertData;
/*
* This represents an itempointer, or many itempointers, to be inserted to
* a data (posting tree) leaf page
*/
typedef struct
{
ItemPointerData *items;
uint32 nitem;
uint32 curitem;
} GinBtreeDataLeafInsertData;
/*
* For internal data (posting tree) pages, the insertion payload is a
* PostingItem
*/
extern GinBtreeStack *ginFindLeafPage(GinBtree btree, bool searchMode,
bool rootConflictCheck, Snapshot snapshot);
extern Buffer ginStepRight(Buffer buffer, Relation index, int lockmode);
extern void freeGinBtreeStack(GinBtreeStack *stack);
extern void ginInsertValue(GinBtree btree, GinBtreeStack *stack,
void *insertdata, GinStatsData *buildStats);
/* ginentrypage.c */
extern IndexTuple GinFormTuple(GinState *ginstate,
OffsetNumber attnum, Datum key, GinNullCategory category,
Pointer data, Size dataSize, int nipd, bool errorTooBig);
extern void ginPrepareEntryScan(GinBtree btree, OffsetNumber attnum,
Datum key, GinNullCategory category,
GinState *ginstate);
extern void ginEntryFillRoot(GinBtree btree, Page root, BlockNumber lblkno, Page lpage, BlockNumber rblkno, Page rpage);
extern ItemPointer ginReadTuple(GinState *ginstate, OffsetNumber attnum,
IndexTuple itup, int *nitems);
/* gindatapage.c */
extern ItemPointer GinDataLeafPageGetItems(Page page, int *nitems, ItemPointerData advancePast);
extern int GinDataLeafPageGetItemsToTbm(Page page, TIDBitmap *tbm);
extern BlockNumber createPostingTree(Relation index,
ItemPointerData *items, uint32 nitems,
GinStatsData *buildStats, Buffer entrybuffer);
extern void GinDataPageAddPostingItem(Page page, PostingItem *data, OffsetNumber offset);
extern void GinPageDeletePostingItem(Page page, OffsetNumber offset);
extern void ginInsertItemPointers(Relation index, BlockNumber rootBlkno,
ItemPointerData *items, uint32 nitem,
GinStatsData *buildStats);
extern GinBtreeStack *ginScanBeginPostingTree(GinBtree btree, Relation index, BlockNumber rootBlkno, Snapshot snapshot);
extern void ginDataFillRoot(GinBtree btree, Page root, BlockNumber lblkno, Page lpage, BlockNumber rblkno, Page rpage);
/*
* This is declared in ginvacuum.c, but is passed between ginVacuumItemPointers
* and ginVacuumPostingTreeLeaf and as an opaque struct, so we need a forward
* declaration for it.
*/
typedef struct GinVacuumState GinVacuumState;
extern void ginVacuumPostingTreeLeaf(Relation rel, Buffer buf, GinVacuumState *gvs);
/* ginscan.c */
/*
* GinScanKeyData describes a single GIN index qualifier expression.
*
* From each qual expression, we extract one or more specific index search
* conditions, which are represented by GinScanEntryData. It's quite
* possible for identical search conditions to be requested by more than
* one qual expression, in which case we merge such conditions to have just
* one unique GinScanEntry --- this is particularly important for efficiency
* when dealing with full-index-scan entries. So there can be multiple
* GinScanKeyData.scanEntry pointers to the same GinScanEntryData.
*
* In each GinScanKeyData, nentries is the true number of entries, while
* nuserentries is the number that extractQueryFn returned (which is what
* we report to consistentFn). The "user" entries must come first.
*/
typedef struct GinScanKeyData *GinScanKey;
typedef struct GinScanEntryData *GinScanEntry;
typedef struct GinScanKeyData
{
/* Real number of entries in scanEntry[] (always > 0) */
uint32 nentries;
/* Number of entries that extractQueryFn and consistentFn know about */
uint32 nuserentries;
/* array of GinScanEntry pointers, one per extracted search condition */
GinScanEntry *scanEntry;
/*
* At least one of the entries in requiredEntries must be present for a
* tuple to match the overall qual.
*
* additionalEntries contains entries that are needed by the consistent
* function to decide if an item matches, but are not sufficient to
* satisfy the qual without entries from requiredEntries.
*/
GinScanEntry *requiredEntries;
int nrequired;
GinScanEntry *additionalEntries;
int nadditional;
/* array of check flags, reported to consistentFn */
GinTernaryValue *entryRes;
bool (*boolConsistentFn) (GinScanKey key);
GinTernaryValue (*triConsistentFn) (GinScanKey key);
FmgrInfo *consistentFmgrInfo;
FmgrInfo *triConsistentFmgrInfo;
Oid collation;
/* other data needed for calling consistentFn */
Datum query;
/* NB: these three arrays have only nuserentries elements! */
Datum *queryValues;
GinNullCategory *queryCategories;
Pointer *extra_data;
StrategyNumber strategy;
int32 searchMode;
OffsetNumber attnum;
/*
* An excludeOnly scan key is not able to enumerate all matching tuples.
* That is, to be semantically correct on its own, it would need to have a
* GIN_CAT_EMPTY_QUERY scanEntry, but it doesn't. Such a key can still be
* used to filter tuples returned by other scan keys, so we will get the
* right answers as long as there's at least one non-excludeOnly scan key
* for each index attribute considered by the search. For efficiency
* reasons we don't want to have unnecessary GIN_CAT_EMPTY_QUERY entries,
* so we will convert an excludeOnly scan key to non-excludeOnly (by
* adding a GIN_CAT_EMPTY_QUERY scanEntry) only if there are no other
* non-excludeOnly scan keys.
*/
bool excludeOnly;
/*
* Match status data. curItem is the TID most recently tested (could be a
* lossy-page pointer). curItemMatches is true if it passes the
* consistentFn test; if so, recheckCurItem is the recheck flag.
* isFinished means that all the input entry streams are finished, so this
* key cannot succeed for any later TIDs.
*/
ItemPointerData curItem;
bool curItemMatches;
bool recheckCurItem;
bool isFinished;
} GinScanKeyData;
typedef struct GinScanEntryData
{
/* query key and other information from extractQueryFn */
Datum queryKey;
GinNullCategory queryCategory;
bool isPartialMatch;
Pointer extra_data;
StrategyNumber strategy;
int32 searchMode;
OffsetNumber attnum;
/* Current page in posting tree */
Buffer buffer;
/* current ItemPointer to heap */
ItemPointerData curItem;
/* for a partial-match or full-scan query, we accumulate all TIDs here */
TIDBitmap *matchBitmap;
TBMIterator *matchIterator;
TBMIterateResult *matchResult;
/* used for Posting list and one page in Posting tree */
ItemPointerData *list;
int nlist;
OffsetNumber offset;
bool isFinished;
bool reduceResult;
uint32 predictNumberResult;
GinBtreeData btree;
} GinScanEntryData;
typedef struct GinScanOpaqueData
{
MemoryContext tempCtx;
GinState ginstate;
GinScanKey keys; /* one per scan qualifier expr */
uint32 nkeys;
GinScanEntry *entries; /* one per index search condition */
uint32 totalentries;
uint32 allocentries; /* allocated length of entries[] */
MemoryContext keyCtx; /* used to hold key and entry data */
bool isVoidRes; /* true if query is unsatisfiable */
} GinScanOpaqueData;
typedef GinScanOpaqueData *GinScanOpaque;
extern IndexScanDesc ginbeginscan(Relation rel, int nkeys, int norderbys);
extern void ginendscan(IndexScanDesc scan);
extern void ginrescan(IndexScanDesc scan, ScanKey key, int nscankeys,
ScanKey orderbys, int norderbys);
extern void ginNewScanKey(IndexScanDesc scan);
extern void ginFreeScanKeys(GinScanOpaque so);
/* ginget.c */
extern int64 gingetbitmap(IndexScanDesc scan, TIDBitmap *tbm);
/* ginlogic.c */
extern void ginInitConsistentFunction(GinState *ginstate, GinScanKey key);
/* ginvacuum.c */
extern IndexBulkDeleteResult *ginbulkdelete(IndexVacuumInfo *info,
IndexBulkDeleteResult *stats,
IndexBulkDeleteCallback callback,
void *callback_state);
extern IndexBulkDeleteResult *ginvacuumcleanup(IndexVacuumInfo *info,
IndexBulkDeleteResult *stats);
extern ItemPointer ginVacuumItemPointers(GinVacuumState *gvs,
ItemPointerData *items, int nitem, int *nremaining);
/* ginvalidate.c */
extern bool ginvalidate(Oid opclassoid);
/* ginbulk.c */
typedef struct GinEntryAccumulator
{
RBTNode rbtnode;
Datum key;
GinNullCategory category;
OffsetNumber attnum;
bool shouldSort;
ItemPointerData *list;
uint32 maxcount; /* allocated size of list[] */
uint32 count; /* current number of list[] entries */
} GinEntryAccumulator;
typedef struct
{
GinState *ginstate;
Size allocatedMemory;
GinEntryAccumulator *entryallocator;
uint32 eas_used;
RBTree *tree;
RBTreeIterator tree_walk;
} BuildAccumulator;
extern void ginInitBA(BuildAccumulator *accum);
extern void ginInsertBAEntries(BuildAccumulator *accum,
ItemPointer heapptr, OffsetNumber attnum,
Datum *entries, GinNullCategory *categories,
int32 nentries);
extern void ginBeginBAScan(BuildAccumulator *accum);
extern ItemPointerData *ginGetBAEntry(BuildAccumulator *accum,
OffsetNumber *attnum, Datum *key, GinNullCategory *category,
uint32 *n);
/* ginfast.c */
typedef struct GinTupleCollector
{
IndexTuple *tuples;
uint32 ntuples;
uint32 lentuples;
uint32 sumsize;
} GinTupleCollector;
extern void ginHeapTupleFastInsert(GinState *ginstate,
GinTupleCollector *collector);
extern void ginHeapTupleFastCollect(GinState *ginstate,
GinTupleCollector *collector,
OffsetNumber attnum, Datum value, bool isNull,
ItemPointer ht_ctid);
extern void ginInsertCleanup(GinState *ginstate, bool full_clean,
bool fill_fsm, bool forceCleanup, IndexBulkDeleteResult *stats);
/* ginpostinglist.c */
extern GinPostingList *ginCompressPostingList(const ItemPointer ipd, int nipd,
int maxsize, int *nwritten);
extern int ginPostingListDecodeAllSegmentsToTbm(GinPostingList *ptr, int totalsize, TIDBitmap *tbm);
extern ItemPointer ginPostingListDecodeAllSegments(GinPostingList *ptr, int len, int *ndecoded);
extern ItemPointer ginPostingListDecode(GinPostingList *ptr, int *ndecoded);
extern ItemPointer ginMergeItemPointers(ItemPointerData *a, uint32 na,
ItemPointerData *b, uint32 nb,
int *nmerged);
/*
* Merging the results of several gin scans compares item pointers a lot,
* so we want this to be inlined.
*/
static inline int
ginCompareItemPointers(ItemPointer a, ItemPointer b)
{
uint64 ia = (uint64) GinItemPointerGetBlockNumber(a) << 32 | GinItemPointerGetOffsetNumber(a);
uint64 ib = (uint64) GinItemPointerGetBlockNumber(b) << 32 | GinItemPointerGetOffsetNumber(b);
if (ia == ib)
return 0;
else if (ia > ib)
return 1;
else
return -1;
}
extern int ginTraverseLock(Buffer buffer, bool searchMode);
#endif /* GIN_PRIVATE_H */