1
0
mirror of https://github.com/postgres/postgres.git synced 2025-06-26 12:21:12 +03:00

Minor performance improvement: avoid unnecessary creation/unioning of

bitmaps for multiple indexscans.  Instead just let each indexscan add
TIDs directly into the BitmapOr node's result bitmap.
This commit is contained in:
Tom Lane
2005-04-20 15:48:36 +00:00
parent de4fbfadc5
commit 9d64632144
4 changed files with 65 additions and 17 deletions

View File

@ -8,7 +8,7 @@
*
*
* IDENTIFICATION
* $PostgreSQL: pgsql/src/backend/executor/nodeBitmapOr.c,v 1.1 2005/04/19 22:35:12 tgl Exp $
* $PostgreSQL: pgsql/src/backend/executor/nodeBitmapOr.c,v 1.2 2005/04/20 15:48:36 tgl Exp $
*
*-------------------------------------------------------------------------
*/
@ -31,6 +31,7 @@
#include "executor/execdebug.h"
#include "executor/instrument.h"
#include "executor/nodeBitmapOr.h"
#include "miscadmin.h"
/* ----------------------------------------------------------------
@ -46,6 +47,7 @@ ExecInitBitmapOr(BitmapOr *node, EState *estate)
PlanState **bitmapplanstates;
int nplans;
int i;
ListCell *l;
Plan *initNode;
CXT1_printf("ExecInitBitmapOr: context is %d\n", CurrentMemoryContext);
@ -78,10 +80,12 @@ ExecInitBitmapOr(BitmapOr *node, EState *estate)
* call ExecInitNode on each of the plans to be executed and save the
* results into the array "bitmapplanstates".
*/
for (i = 0; i < nplans; i++)
i = 0;
foreach(l, node->bitmapplans)
{
initNode = (Plan *) list_nth(node->bitmapplans, i);
initNode = (Plan *) lfirst(l);
bitmapplanstates[i] = ExecInitNode(initNode, estate);
i++;
}
return bitmaporstate;
@ -128,17 +132,41 @@ MultiExecBitmapOr(BitmapOrState *node)
PlanState *subnode = bitmapplans[i];
TIDBitmap *subresult;
subresult = (TIDBitmap *) MultiExecProcNode(subnode);
/*
* We can special-case BitmapIndexScan children to avoid an
* explicit tbm_union step for each child: just pass down the
* current result bitmap and let the child OR directly into it.
*/
if (IsA(subnode, BitmapIndexScanState))
{
if (result == NULL) /* first subplan */
{
/* XXX should we use less than work_mem for this? */
result = tbm_create(work_mem * 1024L);
}
if (!subresult || !IsA(subresult, TIDBitmap))
elog(ERROR, "unrecognized result from subplan");
((BitmapIndexScanState *) subnode)->biss_result = result;
if (result == NULL)
result = subresult; /* first subplan */
subresult = (TIDBitmap *) MultiExecProcNode(subnode);
if (subresult != result)
elog(ERROR, "unrecognized result from subplan");
}
else
{
tbm_union(result, subresult);
tbm_free(subresult);
/* standard implementation */
subresult = (TIDBitmap *) MultiExecProcNode(subnode);
if (!subresult || !IsA(subresult, TIDBitmap))
elog(ERROR, "unrecognized result from subplan");
if (result == NULL)
result = subresult; /* first subplan */
else
{
tbm_union(result, subresult);
tbm_free(subresult);
}
}
}