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 64099cd1e50cf10f2832cc723e0fdab2af424d60..6e0a194a69452938eb5872f54c0f91aabc7c68ab 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 1.15 2008/09/30 10:52:10 heikki Exp $
+ *                     $PostgreSQL: pgsql/src/backend/access/gin/gininsert.c,v 1.16 2008/11/13 17:42:09 tgl Exp $
  *-------------------------------------------------------------------------
  */
 
@@ -340,8 +340,11 @@ 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);
 
        /* dump remaining entries to the index */

diff --git a/src/backend/access/gist/gist.c b/src/backend/access/gist/gist.c
index e1f068ee7bcd7abdb40bd9b3375f94af94096390..1dff6d2f2ae9a6ddda9605d9e100e1d6293d67e1 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 1.153 2008/11/03 20:47:48 tgl Exp $
+ *       $PostgreSQL: pgsql/src/backend/access/gist/gist.c,v 1.154 2008/11/13 17:42:09 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -152,7 +152,7 @@ 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 a24350367fc1a4f868d63191e7cdc7919cfd54bc..86d4df5e4dff8b09ce93f937ad1951cf86c16cb1 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 1.106 2008/10/17 23:50:57 tgl Exp $
+ *       $PostgreSQL: pgsql/src/backend/access/hash/hash.c,v 1.107 2008/11/13 17:42:10 tgl Exp $
  *
  * NOTES
  *       This file contains only the public interface routines.
@@ -93,7 +93,7 @@ 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);
 
        if (buildstate.spool)

diff --git a/src/backend/access/nbtree/nbtree.c b/src/backend/access/nbtree/nbtree.c
index f47a8e5f3648ae22c449c0ffbfa52b2befe7b171..bd486afabb7269c35baa2590f0f75544cb827db5 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 1.164 2008/10/31 15:04:59 heikki Exp $
+ *       $PostgreSQL: pgsql/src/backend/access/nbtree/nbtree.c,v 1.165 2008/11/13 17:42:10 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -122,7 +122,7 @@ 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 7c0c313e91b520c082fcae4e562f8cdaf1622ef5..0ad1e04a748ae15a60b0aa67eb951c346a5c6455 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 1.307 2008/11/02 01:45:27 tgl Exp $
+ *       $PostgreSQL: pgsql/src/backend/catalog/index.c,v 1.308 2008/11/13 17:42:10 tgl Exp $
  *
  *
  * INTERFACE ROUTINES
@@ -1513,6 +1513,7 @@ double
 IndexBuildHeapScan(Relation heapRelation,
                                   Relation indexRelation,
                                   IndexInfo *indexInfo,
+                                  bool allow_sync,
                                   IndexBuildCallback callback,
                                   void *callback_state)
 {
@@ -1575,7 +1576,12 @@ IndexBuildHeapScan(Relation heapRelation,
                OldestXmin = GetOldestXmin(heapRelation->rd_rel->relisshared, true);
        }
 
-       scan = heap_beginscan(heapRelation, snapshot, 0, NULL);
+       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..939c47cd916e2bfa0849dbb7bb87e83fb7c5d5e0 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 1.76 2008/11/13 17:42:10 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);