Header And Logo

PostgreSQL
| The world's most advanced open source database.

Data Structures | Defines | Typedefs | Functions

btreefuncs.c File Reference

#include "postgres.h"
#include "access/nbtree.h"
#include "catalog/namespace.h"
#include "funcapi.h"
#include "miscadmin.h"
#include "utils/builtins.h"
#include "utils/rel.h"
Include dependency graph for btreefuncs.c:

Go to the source code of this file.

Data Structures

struct  BTPageStat
struct  user_args

Defines

#define IS_INDEX(r)   ((r)->rd_rel->relkind == RELKIND_INDEX)
#define IS_BTREE(r)   ((r)->rd_rel->relam == BTREE_AM_OID)
#define CHECK_PAGE_OFFSET_RANGE(pg, offnum)
#define CHECK_RELATION_BLOCK_RANGE(rel, blkno)

Typedefs

typedef struct BTPageStat BTPageStat

Functions

Datum bt_metap (PG_FUNCTION_ARGS)
Datum bt_page_items (PG_FUNCTION_ARGS)
Datum bt_page_stats (PG_FUNCTION_ARGS)
 PG_FUNCTION_INFO_V1 (bt_metap)
 PG_FUNCTION_INFO_V1 (bt_page_items)
 PG_FUNCTION_INFO_V1 (bt_page_stats)
static void GetBTPageStatistics (BlockNumber blkno, Buffer buffer, BTPageStat *stat)

Define Documentation

#define CHECK_PAGE_OFFSET_RANGE (   pg,
  offnum 
)
Value:
{ \
        if ( !(FirstOffsetNumber <= (offnum) && \
                        (offnum) <= PageGetMaxOffsetNumber(pg)) ) \
             elog(ERROR, "page offset number out of range"); }

Definition at line 49 of file btreefuncs.c.

#define CHECK_RELATION_BLOCK_RANGE (   rel,
  blkno 
)
Value:
{ \
        if ( RelationGetNumberOfBlocks(rel) <= (BlockNumber) (blkno) ) \
             elog(ERROR, "block number out of range"); }

Definition at line 55 of file btreefuncs.c.

Referenced by bt_page_items(), and bt_page_stats().

#define IS_BTREE (   r  )     ((r)->rd_rel->relam == BTREE_AM_OID)

Definition at line 47 of file btreefuncs.c.

Referenced by bt_metap(), bt_page_items(), and bt_page_stats().

#define IS_INDEX (   r  )     ((r)->rd_rel->relkind == RELKIND_INDEX)

Definition at line 46 of file btreefuncs.c.

Referenced by bt_metap(), bt_page_items(), and bt_page_stats().


Typedef Documentation

typedef struct BTPageStat BTPageStat

Function Documentation

Datum bt_metap ( PG_FUNCTION_ARGS   ) 

Definition at line 433 of file btreefuncs.c.

References AccessShareLock, BTMetaPageData::btm_fastlevel, BTMetaPageData::btm_fastroot, BTMetaPageData::btm_level, BTMetaPageData::btm_magic, BTMetaPageData::btm_root, BTMetaPageData::btm_version, BTPageGetMeta, BUFFER_LOCK_SHARE, BufferGetPage, BuildTupleFromCStrings(), elog, ereport, errcode(), errmsg(), ERROR, get_call_result_type(), HeapTupleGetDatum, IS_BTREE, IS_INDEX, LockBuffer(), makeRangeVarFromNameList(), NULL, user_args::page, palloc(), PG_GETARG_TEXT_P, PG_RETURN_DATUM, ReadBuffer(), relation_close(), RELATION_IS_OTHER_TEMP, relation_openrv(), RelationGetRelationName, snprintf(), superuser(), textToQualifiedNameList(), TupleDescGetAttInMetadata(), TYPEFUNC_COMPOSITE, UnlockReleaseBuffer(), and values.

{
    text       *relname = PG_GETARG_TEXT_P(0);
    Datum       result;
    Relation    rel;
    RangeVar   *relrv;
    BTMetaPageData *metad;
    TupleDesc   tupleDesc;
    int         j;
    char       *values[6];
    Buffer      buffer;
    Page        page;
    HeapTuple   tuple;

    if (!superuser())
        ereport(ERROR,
                (errcode(ERRCODE_INSUFFICIENT_PRIVILEGE),
                 (errmsg("must be superuser to use pageinspect functions"))));

    relrv = makeRangeVarFromNameList(textToQualifiedNameList(relname));
    rel = relation_openrv(relrv, AccessShareLock);

    if (!IS_INDEX(rel) || !IS_BTREE(rel))
        elog(ERROR, "relation \"%s\" is not a btree index",
             RelationGetRelationName(rel));

    /*
     * Reject attempts to read non-local temporary relations; we would be
     * likely to get wrong data since we have no visibility into the owning
     * session's local buffers.
     */
    if (RELATION_IS_OTHER_TEMP(rel))
        ereport(ERROR,
                (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
                 errmsg("cannot access temporary tables of other sessions")));

    buffer = ReadBuffer(rel, 0);
    LockBuffer(buffer, BUFFER_LOCK_SHARE);

    page = BufferGetPage(buffer);
    metad = BTPageGetMeta(page);

    /* Build a tuple descriptor for our result type */
    if (get_call_result_type(fcinfo, NULL, &tupleDesc) != TYPEFUNC_COMPOSITE)
        elog(ERROR, "return type must be a row type");

    j = 0;
    values[j] = palloc(32);
    snprintf(values[j++], 32, "%d", metad->btm_magic);
    values[j] = palloc(32);
    snprintf(values[j++], 32, "%d", metad->btm_version);
    values[j] = palloc(32);
    snprintf(values[j++], 32, "%d", metad->btm_root);
    values[j] = palloc(32);
    snprintf(values[j++], 32, "%d", metad->btm_level);
    values[j] = palloc(32);
    snprintf(values[j++], 32, "%d", metad->btm_fastroot);
    values[j] = palloc(32);
    snprintf(values[j++], 32, "%d", metad->btm_fastlevel);

    tuple = BuildTupleFromCStrings(TupleDescGetAttInMetadata(tupleDesc),
                                   values);

    result = HeapTupleGetDatum(tuple);

    UnlockReleaseBuffer(buffer);
    relation_close(rel, AccessShareLock);

    PG_RETURN_DATUM(result);
}

Datum bt_page_items ( PG_FUNCTION_ARGS   ) 

Definition at line 276 of file btreefuncs.c.

References AccessShareLock, FuncCallContext::attinmeta, BlockIdGetBlockNumber, BUFFER_LOCK_SHARE, BufferGetPage, BuildTupleFromCStrings(), FuncCallContext::call_cntr, CHECK_RELATION_BLOCK_RANGE, elog, ereport, errcode(), errmsg(), ERROR, get_call_result_type(), HeapTupleGetDatum, IndexInfoFindDataOffset, IndexTupleHasNulls, IndexTupleHasVarwidths, IndexTupleSize, ItemPointerData::ip_blkid, ItemPointerData::ip_posid, IS_BTREE, IS_INDEX, ItemIdIsValid, LockBuffer(), makeRangeVarFromNameList(), FuncCallContext::max_calls, MemoryContextSwitchTo(), FuncCallContext::multi_call_memory_ctx, NOTICE, NULL, user_args::offset, P_ISDELETED, user_args::page, PageGetItem, PageGetItemId, PageGetMaxOffsetNumber, PageGetSpecialPointer, palloc(), palloc0(), pfree(), PG_GETARG_TEXT_P, PG_GETARG_UINT32, ReadBuffer(), relation_close(), RELATION_IS_OTHER_TEMP, relation_openrv(), RelationGetRelationName, snprintf(), SRF_FIRSTCALL_INIT, SRF_IS_FIRSTCALL, SRF_PERCALL_SETUP, SRF_RETURN_DONE, SRF_RETURN_NEXT, superuser(), IndexTupleData::t_info, IndexTupleData::t_tid, textToQualifiedNameList(), TupleDescGetAttInMetadata(), TYPEFUNC_COMPOSITE, UnlockReleaseBuffer(), FuncCallContext::user_fctx, and values.

{
    text       *relname = PG_GETARG_TEXT_P(0);
    uint32      blkno = PG_GETARG_UINT32(1);
    Datum       result;
    char       *values[6];
    HeapTuple   tuple;
    FuncCallContext *fctx;
    MemoryContext mctx;
    struct user_args *uargs;

    if (!superuser())
        ereport(ERROR,
                (errcode(ERRCODE_INSUFFICIENT_PRIVILEGE),
                 (errmsg("must be superuser to use pageinspect functions"))));

    if (SRF_IS_FIRSTCALL())
    {
        RangeVar   *relrv;
        Relation    rel;
        Buffer      buffer;
        BTPageOpaque opaque;
        TupleDesc   tupleDesc;

        fctx = SRF_FIRSTCALL_INIT();

        relrv = makeRangeVarFromNameList(textToQualifiedNameList(relname));
        rel = relation_openrv(relrv, AccessShareLock);

        if (!IS_INDEX(rel) || !IS_BTREE(rel))
            elog(ERROR, "relation \"%s\" is not a btree index",
                 RelationGetRelationName(rel));

        /*
         * Reject attempts to read non-local temporary relations; we would be
         * likely to get wrong data since we have no visibility into the
         * owning session's local buffers.
         */
        if (RELATION_IS_OTHER_TEMP(rel))
            ereport(ERROR,
                    (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
                errmsg("cannot access temporary tables of other sessions")));

        if (blkno == 0)
            elog(ERROR, "block 0 is a meta page");

        CHECK_RELATION_BLOCK_RANGE(rel, blkno);

        buffer = ReadBuffer(rel, blkno);
        LockBuffer(buffer, BUFFER_LOCK_SHARE);

        /*
         * We copy the page into local storage to avoid holding pin on the
         * buffer longer than we must, and possibly failing to release it at
         * all if the calling query doesn't fetch all rows.
         */
        mctx = MemoryContextSwitchTo(fctx->multi_call_memory_ctx);

        uargs = palloc(sizeof(struct user_args));

        uargs->page = palloc(BLCKSZ);
        memcpy(uargs->page, BufferGetPage(buffer), BLCKSZ);

        UnlockReleaseBuffer(buffer);
        relation_close(rel, AccessShareLock);

        uargs->offset = FirstOffsetNumber;

        opaque = (BTPageOpaque) PageGetSpecialPointer(uargs->page);

        if (P_ISDELETED(opaque))
            elog(NOTICE, "page is deleted");

        fctx->max_calls = PageGetMaxOffsetNumber(uargs->page);

        /* Build a tuple descriptor for our result type */
        if (get_call_result_type(fcinfo, NULL, &tupleDesc) != TYPEFUNC_COMPOSITE)
            elog(ERROR, "return type must be a row type");

        fctx->attinmeta = TupleDescGetAttInMetadata(tupleDesc);

        fctx->user_fctx = uargs;

        MemoryContextSwitchTo(mctx);
    }

    fctx = SRF_PERCALL_SETUP();
    uargs = fctx->user_fctx;

    if (fctx->call_cntr < fctx->max_calls)
    {
        ItemId      id;
        IndexTuple  itup;
        int         j;
        int         off;
        int         dlen;
        char       *dump;
        char       *ptr;

        id = PageGetItemId(uargs->page, uargs->offset);

        if (!ItemIdIsValid(id))
            elog(ERROR, "invalid ItemId");

        itup = (IndexTuple) PageGetItem(uargs->page, id);

        j = 0;
        values[j] = palloc(32);
        snprintf(values[j++], 32, "%d", uargs->offset);
        values[j] = palloc(32);
        snprintf(values[j++], 32, "(%u,%u)",
                 BlockIdGetBlockNumber(&(itup->t_tid.ip_blkid)),
                 itup->t_tid.ip_posid);
        values[j] = palloc(32);
        snprintf(values[j++], 32, "%d", (int) IndexTupleSize(itup));
        values[j] = palloc(32);
        snprintf(values[j++], 32, "%c", IndexTupleHasNulls(itup) ? 't' : 'f');
        values[j] = palloc(32);
        snprintf(values[j++], 32, "%c", IndexTupleHasVarwidths(itup) ? 't' : 'f');

        ptr = (char *) itup + IndexInfoFindDataOffset(itup->t_info);
        dlen = IndexTupleSize(itup) - IndexInfoFindDataOffset(itup->t_info);
        dump = palloc0(dlen * 3 + 1);
        values[j] = dump;
        for (off = 0; off < dlen; off++)
        {
            if (off > 0)
                *dump++ = ' ';
            sprintf(dump, "%02x", *(ptr + off) & 0xff);
            dump += 2;
        }

        tuple = BuildTupleFromCStrings(fctx->attinmeta, values);
        result = HeapTupleGetDatum(tuple);

        uargs->offset = uargs->offset + 1;

        SRF_RETURN_NEXT(fctx, result);
    }
    else
    {
        pfree(uargs->page);
        pfree(uargs);
        SRF_RETURN_DONE(fctx);
    }
}

Datum bt_page_stats ( PG_FUNCTION_ARGS   ) 

Definition at line 165 of file btreefuncs.c.

References AccessShareLock, BTPageStat::avg_item_size, BTPageStat::blkno, BTPageStat::btpo, BTPageStat::btpo_flags, BTPageStat::btpo_next, BTPageStat::btpo_prev, BUFFER_LOCK_SHARE, BuildTupleFromCStrings(), CHECK_RELATION_BLOCK_RANGE, BTPageStat::dead_items, elog, ereport, errcode(), errmsg(), ERROR, BTPageStat::free_size, get_call_result_type(), GetBTPageStatistics(), HeapTupleGetDatum, IS_BTREE, IS_INDEX, BTPageStat::level, BTPageStat::live_items, LockBuffer(), makeRangeVarFromNameList(), NULL, BTPageStat::page_size, palloc(), PG_GETARG_TEXT_P, PG_GETARG_UINT32, PG_RETURN_DATUM, ReadBuffer(), relation_close(), RELATION_IS_OTHER_TEMP, relation_openrv(), RelationGetRelationName, snprintf(), superuser(), textToQualifiedNameList(), TupleDescGetAttInMetadata(), BTPageStat::type, TYPEFUNC_COMPOSITE, UnlockReleaseBuffer(), values, and BTPageStat::xact.

{
    text       *relname = PG_GETARG_TEXT_P(0);
    uint32      blkno = PG_GETARG_UINT32(1);
    Buffer      buffer;
    Relation    rel;
    RangeVar   *relrv;
    Datum       result;
    HeapTuple   tuple;
    TupleDesc   tupleDesc;
    int         j;
    char       *values[11];
    BTPageStat  stat;

    if (!superuser())
        ereport(ERROR,
                (errcode(ERRCODE_INSUFFICIENT_PRIVILEGE),
                 (errmsg("must be superuser to use pageinspect functions"))));

    relrv = makeRangeVarFromNameList(textToQualifiedNameList(relname));
    rel = relation_openrv(relrv, AccessShareLock);

    if (!IS_INDEX(rel) || !IS_BTREE(rel))
        elog(ERROR, "relation \"%s\" is not a btree index",
             RelationGetRelationName(rel));

    /*
     * Reject attempts to read non-local temporary relations; we would be
     * likely to get wrong data since we have no visibility into the owning
     * session's local buffers.
     */
    if (RELATION_IS_OTHER_TEMP(rel))
        ereport(ERROR,
                (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
                 errmsg("cannot access temporary tables of other sessions")));

    if (blkno == 0)
        elog(ERROR, "block 0 is a meta page");

    CHECK_RELATION_BLOCK_RANGE(rel, blkno);

    buffer = ReadBuffer(rel, blkno);
    LockBuffer(buffer, BUFFER_LOCK_SHARE);

    /* keep compiler quiet */
    stat.btpo_prev = stat.btpo_next = InvalidBlockNumber;
    stat.btpo_flags = stat.free_size = stat.avg_item_size = 0;

    GetBTPageStatistics(blkno, buffer, &stat);

    UnlockReleaseBuffer(buffer);
    relation_close(rel, AccessShareLock);

    /* Build a tuple descriptor for our result type */
    if (get_call_result_type(fcinfo, NULL, &tupleDesc) != TYPEFUNC_COMPOSITE)
        elog(ERROR, "return type must be a row type");

    j = 0;
    values[j] = palloc(32);
    snprintf(values[j++], 32, "%d", stat.blkno);
    values[j] = palloc(32);
    snprintf(values[j++], 32, "%c", stat.type);
    values[j] = palloc(32);
    snprintf(values[j++], 32, "%d", stat.live_items);
    values[j] = palloc(32);
    snprintf(values[j++], 32, "%d", stat.dead_items);
    values[j] = palloc(32);
    snprintf(values[j++], 32, "%d", stat.avg_item_size);
    values[j] = palloc(32);
    snprintf(values[j++], 32, "%d", stat.page_size);
    values[j] = palloc(32);
    snprintf(values[j++], 32, "%d", stat.free_size);
    values[j] = palloc(32);
    snprintf(values[j++], 32, "%d", stat.btpo_prev);
    values[j] = palloc(32);
    snprintf(values[j++], 32, "%d", stat.btpo_next);
    values[j] = palloc(32);
    if (stat.type == 'd')
        snprintf(values[j++], 32, "%d", stat.btpo.xact);
    else
        snprintf(values[j++], 32, "%d", stat.btpo.level);
    values[j] = palloc(32);
    snprintf(values[j++], 32, "%d", stat.btpo_flags);

    tuple = BuildTupleFromCStrings(TupleDescGetAttInMetadata(tupleDesc),
                                   values);

    result = HeapTupleGetDatum(tuple);

    PG_RETURN_DATUM(result);
}

static void GetBTPageStatistics ( BlockNumber  blkno,
Buffer  buffer,
BTPageStat stat 
) [static]

Definition at line 94 of file btreefuncs.c.

References BTPageStat::avg_item_size, BTPageStat::blkno, BTPageOpaqueData::btpo, BTPageStat::btpo, BTPageOpaqueData::btpo_cycleid, BTPageStat::btpo_cycleid, BTPageOpaqueData::btpo_flags, BTPageStat::btpo_flags, BTPageOpaqueData::btpo_next, BTPageStat::btpo_next, BTPageOpaqueData::btpo_prev, BTPageStat::btpo_prev, BufferGetPage, BTPageStat::dead_items, FirstOffsetNumber, BTPageStat::free_size, IndexTupleSize, ItemIdIsDead, BTPageOpaqueData::level, BTPageStat::level, BTPageStat::live_items, BTPageStat::max_avail, P_IGNORE, P_ISDELETED, P_ISLEAF, P_ISROOT, BTPageStat::page_size, PageGetFreeSpace(), PageGetItem, PageGetItemId, PageGetMaxOffsetNumber, PageGetPageSize, PageGetSpecialPointer, PageHeaderData::pd_special, BTPageStat::type, BTPageOpaqueData::xact, and BTPageStat::xact.

Referenced by bt_page_stats().

{
    Page        page = BufferGetPage(buffer);
    PageHeader  phdr = (PageHeader) page;
    OffsetNumber maxoff = PageGetMaxOffsetNumber(page);
    BTPageOpaque opaque = (BTPageOpaque) PageGetSpecialPointer(page);
    int         item_size = 0;
    int         off;

    stat->blkno = blkno;

    stat->max_avail = BLCKSZ - (BLCKSZ - phdr->pd_special + SizeOfPageHeaderData);

    stat->dead_items = stat->live_items = 0;

    stat->page_size = PageGetPageSize(page);

    /* page type (flags) */
    if (P_ISDELETED(opaque))
    {
        stat->type = 'd';
        stat->btpo.xact = opaque->btpo.xact;
        return;
    }
    else if (P_IGNORE(opaque))
        stat->type = 'e';
    else if (P_ISLEAF(opaque))
        stat->type = 'l';
    else if (P_ISROOT(opaque))
        stat->type = 'r';
    else
        stat->type = 'i';

    /* btpage opaque data */
    stat->btpo_prev = opaque->btpo_prev;
    stat->btpo_next = opaque->btpo_next;
    stat->btpo.level = opaque->btpo.level;
    stat->btpo_flags = opaque->btpo_flags;
    stat->btpo_cycleid = opaque->btpo_cycleid;

    /* count live and dead tuples, and free space */
    for (off = FirstOffsetNumber; off <= maxoff; off++)
    {
        IndexTuple  itup;

        ItemId      id = PageGetItemId(page, off);

        itup = (IndexTuple) PageGetItem(page, id);

        item_size += IndexTupleSize(itup);

        if (!ItemIdIsDead(id))
            stat->live_items++;
        else
            stat->dead_items++;
    }
    stat->free_size = PageGetFreeSpace(page);

    if ((stat->live_items + stat->dead_items) > 0)
        stat->avg_item_size = item_size / (stat->live_items + stat->dead_items);
    else
        stat->avg_item_size = 0;
}

PG_FUNCTION_INFO_V1 ( bt_metap   ) 
PG_FUNCTION_INFO_V1 ( bt_page_stats   ) 
PG_FUNCTION_INFO_V1 ( bt_page_items   )