Prevent synchronous scan during GIN index build, because GIN is optimized

for inserting tuples in increasing TID order.  It's not clear whether this
fully explains Ivan Sergio Borgonovo's complaint, but simple testing
confirms that a scan that doesn't start at block 0 can slow GIN build by
a factor of three or four.

Backpatch to 8.3.  Sync scan didn't exist before that.

diff --git a/src/backend/access/gin/gininsert.c b/src/backend/access/gin/gininsert.c
index a20687f70327bdbe348f27bae083c0c798fe52b9..31af4a82f1d2887cf5eca9a658638c80ef73e017 100644 (file)
--- a/src/backend/access/gin/gininsert.c
+++ b/src/backend/access/gin/gininsert.c
@@ -8,7 +8,7 @@
  * Portions Copyright (c) 1994, Regents of the University of California
  *
  * IDENTIFICATION
- *                     $PostgreSQL: pgsql/src/backend/access/gin/gininsert.c,v 1.11 2008/01/01 19:45:46 momjian Exp $
+ *                     $PostgreSQL: pgsql/src/backend/access/gin/gininsert.c,v 1.11.2.1 2008/11/13 17:42:18 tgl Exp $
  *-------------------------------------------------------------------------
  */
 
@@ -326,8 +326,11 @@ ginbuild(PG_FUNCTION_ARGS)
        buildstate.accum.ginstate = &buildstate.ginstate;
        ginInitBA(&buildstate.accum);
 
-       /* do the heap scan */
-       reltuples = IndexBuildHeapScan(heap, index, indexInfo,
+       /*
+        * Do the heap scan.  We disallow sync scan here because dataPlaceToPage
+        * prefers to receive tuples in TID order.
+        */
+       reltuples = IndexBuildHeapScan(heap, index, indexInfo, false,
                                                                   ginBuildCallback, (void *) &buildstate);
 
        oldCtx = MemoryContextSwitchTo(buildstate.tmpCtx);

diff --git a/src/backend/access/gist/gist.c b/src/backend/access/gist/gist.c
index 91b8fe4d055f0ef1a38d685aee648390f612a6be..26f3ddbf8534cb75379100f64df223ad2844264a 100644 (file)
--- a/src/backend/access/gist/gist.c
+++ b/src/backend/access/gist/gist.c
@@ -8,7 +8,7 @@
  * Portions Copyright (c) 1994, Regents of the University of California
  *
  * IDENTIFICATION
- *       $PostgreSQL: pgsql/src/backend/access/gist/gist.c,v 1.149 2008/01/01 19:45:46 momjian Exp $
+ *       $PostgreSQL: pgsql/src/backend/access/gist/gist.c,v 1.149.2.1 2008/11/13 17:42:18 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -147,7 +147,7 @@ gistbuild(PG_FUNCTION_ARGS)
        buildstate.tmpCtx = createTempGistContext();
 
        /* do the heap scan */
-       reltuples = IndexBuildHeapScan(heap, index, indexInfo,
+       reltuples = IndexBuildHeapScan(heap, index, indexInfo, true,
                                                                   gistbuildCallback, (void *) &buildstate);
 
        /* okay, all heap tuples are indexed */

diff --git a/src/backend/access/hash/hash.c b/src/backend/access/hash/hash.c
index f6c4d5705d4284a6393d3ebcb81ba09be88b66bd..4b58f46b13642739021535d88934680ddcd2e456 100644 (file)
--- a/src/backend/access/hash/hash.c
+++ b/src/backend/access/hash/hash.c
@@ -8,7 +8,7 @@
  *
  *
  * IDENTIFICATION
- *       $PostgreSQL: pgsql/src/backend/access/hash/hash.c,v 1.98 2008/01/01 19:45:46 momjian Exp $
+ *       $PostgreSQL: pgsql/src/backend/access/hash/hash.c,v 1.98.2.1 2008/11/13 17:42:18 tgl Exp $
  *
  * NOTES
  *       This file contains only the public interface routines.
@@ -66,7 +66,7 @@ hashbuild(PG_FUNCTION_ARGS)
        buildstate.indtuples = 0;
 
        /* do the heap scan */
-       reltuples = IndexBuildHeapScan(heap, index, indexInfo,
+       reltuples = IndexBuildHeapScan(heap, index, indexInfo, true,
                                                                   hashbuildCallback, (void *) &buildstate);
 
        /*

diff --git a/src/backend/access/nbtree/nbtree.c b/src/backend/access/nbtree/nbtree.c
index d547b910528cd01c4ba87eadf178c9dc12827cc6..8b30331be4b0b2ab6e8a01fec2f5606f6bb5dc3e 100644 (file)
--- a/src/backend/access/nbtree/nbtree.c
+++ b/src/backend/access/nbtree/nbtree.c
@@ -12,7 +12,7 @@
  * Portions Copyright (c) 1994, Regents of the University of California
  *
  * IDENTIFICATION
- *       $PostgreSQL: pgsql/src/backend/access/nbtree/nbtree.c,v 1.156.2.1 2008/04/16 23:59:51 tgl Exp $
+ *       $PostgreSQL: pgsql/src/backend/access/nbtree/nbtree.c,v 1.156.2.2 2008/11/13 17:42:18 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -117,7 +117,7 @@ btbuild(PG_FUNCTION_ARGS)
                buildstate.spool2 = _bt_spoolinit(index, false, true);
 
        /* do the heap scan */
-       reltuples = IndexBuildHeapScan(heap, index, indexInfo,
+       reltuples = IndexBuildHeapScan(heap, index, indexInfo, true,
                                                                   btbuildCallback, (void *) &buildstate);
 
        /* okay, all heap tuples are indexed */

diff --git a/src/backend/catalog/index.c b/src/backend/catalog/index.c
index 403e9ab5f25ed710b37e3f96c7bb4fae4a547e4f..85c368e4408b0bff0a32cd064d25d65daded6d7e 100644 (file)
--- a/src/backend/catalog/index.c
+++ b/src/backend/catalog/index.c
@@ -8,7 +8,7 @@
  *
  *
  * IDENTIFICATION
- *       $PostgreSQL: pgsql/src/backend/catalog/index.c,v 1.292.2.1 2008/08/10 19:02:46 tgl Exp $
+ *       $PostgreSQL: pgsql/src/backend/catalog/index.c,v 1.292.2.2 2008/11/13 17:42:18 tgl Exp $
  *
  *
  * INTERFACE ROUTINES
@@ -1450,6 +1450,7 @@ double
 IndexBuildHeapScan(Relation heapRelation,
                                   Relation indexRelation,
                                   IndexInfo *indexInfo,
+                                  bool allow_sync,
                                   IndexBuildCallback callback,
                                   void *callback_state)
 {
@@ -1512,10 +1513,12 @@ IndexBuildHeapScan(Relation heapRelation,
                OldestXmin = GetOldestXmin(heapRelation->rd_rel->relisshared, true);
        }
 
-       scan = heap_beginscan(heapRelation, /* relation */
-                                                 snapshot,             /* seeself */
-                                                 0,    /* number of keys */
-                                                 NULL);        /* scan key */
+       scan = heap_beginscan_strat(heapRelation,       /* relation */
+                                                               snapshot,               /* snapshot */
+                                                               0,                              /* number of keys */
+                                                               NULL,                   /* scan key */
+                                                               true,                   /* buffer access strategy OK */
+                                                               allow_sync);    /* syncscan OK? */
 
        reltuples = 0;
 

diff --git a/src/include/catalog/index.h b/src/include/catalog/index.h
index a5ab195c233d92eac2332e537a376bb1c70241fb..70c265232bf57ffbba9abffceffb688fabc09c09 100644 (file)
--- a/src/include/catalog/index.h
+++ b/src/include/catalog/index.h
@@ -7,7 +7,7 @@
  * Portions Copyright (c) 1996-2008, PostgreSQL Global Development Group
  * Portions Copyright (c) 1994, Regents of the University of California
  *
- * $PostgreSQL: pgsql/src/include/catalog/index.h,v 1.75 2008/01/01 19:45:56 momjian Exp $
+ * $PostgreSQL: pgsql/src/include/catalog/index.h,v 1.75.2.1 2008/11/13 17:42:19 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -63,6 +63,7 @@ extern void index_build(Relation heapRelation,
 extern double IndexBuildHeapScan(Relation heapRelation,
                                   Relation indexRelation,
                                   IndexInfo *indexInfo,
+                                  bool allow_sync,
                                   IndexBuildCallback callback,
                                   void *callback_state);