tidbitmap: Support shared iteration.

When a shared iterator is used, each call to tbm_shared_iterate()
returns a result that has not yet been returned to any process
attached to the shared iterator.  In other words, each cooperating
processes gets a disjoint subset of the full result set, but all
results are returned exactly once.

This is infrastructure for parallel bitmap heap scan.

Dilip Kumar.  The larger patch set of which this is a part has been
reviewed and tested by (at least) Andres Freund, Amit Khandekar,
Tushar Ahuja, Rafia Sabih, Haribabu Kommi, and Thomas Munro.

Discussion: http://postgr.es/m/CAFiTN-uc4=0WxRGfCzs-xfkMYcSEWUC-Fon6thkJGjkh9i=13A@mail.gmail.com

doc/src/sgml/monitoring.sgml

@@ -826,7 +826,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
 
       <tbody>
        <row>
-        <entry morerows="58"><literal>LWLock</></entry>
+        <entry morerows="59"><literal>LWLock</></entry>
         <entry><literal>ShmemIndexLock</></entry>
         <entry>Waiting to find or allocate space in shared memory.</entry>
        </row>
@@ -1081,6 +1081,10 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
          <entry><literal>parallel_query_dsa</></entry>
          <entry>Waiting for parallel query dynamic shared memory allocation lock.</entry>
         </row>
+        <row>
+         <entry><literal>tbm</></entry>
+         <entry>Waiting for TBM shared iterator lock.</entry>
+        </row>
         <row>
          <entry morerows="9"><literal>Lock</></entry>
          <entry><literal>relation</></entry>

src/backend/access/gin/ginget.c

@@ -120,7 +120,7 @@ collectMatchBitmap(GinBtreeData *btree, GinBtreeStack *stack,
 	Form_pg_attribute attr;
 
 	/* Initialize empty bitmap result */
-	scanEntry->matchBitmap = tbm_create(work_mem * 1024L);
+	scanEntry->matchBitmap = tbm_create(work_mem * 1024L, NULL);
 
 	/* Null query cannot partial-match anything */
 	if (scanEntry->isPartialMatch &&

src/backend/executor/nodeBitmapIndexscan.c

@@ -78,7 +78,7 @@ MultiExecBitmapIndexScan(BitmapIndexScanState *node)
 	else
 	{
 		/* XXX should we use less than work_mem for this? */
-		tbm = tbm_create(work_mem * 1024L);
+		tbm = tbm_create(work_mem * 1024L, NULL);
 	}
 
 	/*

src/backend/executor/nodeBitmapOr.c

@@ -129,7 +129,7 @@ MultiExecBitmapOr(BitmapOrState *node)
 			if (result == NULL) /* first subplan */
 			{
 				/* XXX should we use less than work_mem for this? */
-				result = tbm_create(work_mem * 1024L);
+				result = tbm_create(work_mem * 1024L, NULL);
 			}
 
 			((BitmapIndexScanState *) subnode)->biss_result = result;

src/backend/storage/lmgr/lwlock.c

@@ -510,6 +510,7 @@ RegisterLWLockTranches(void)
 						  "predicate_lock_manager");
 	LWLockRegisterTranche(LWTRANCHE_PARALLEL_QUERY_DSA,
 						  "parallel_query_dsa");
+	LWLockRegisterTranche(LWTRANCHE_TBM, "tbm");
 
 	/* Register named tranches. */
 	for (i = 0; i < NamedLWLockTrancheRequests; i++)

src/include/nodes/tidbitmap.h

@@ -23,6 +23,7 @@
 #define TIDBITMAP_H
 
 #include "storage/itemptr.h"
+#include "utils/dsa.h"
 
 
 /*
@@ -33,6 +34,7 @@ typedef struct TIDBitmap TIDBitmap;
 
 /* Likewise, TBMIterator is private */
 typedef struct TBMIterator TBMIterator;
+typedef struct TBMSharedIterator TBMSharedIterator;
 
 /* Result structure for tbm_iterate */
 typedef struct
@@ -46,8 +48,9 @@ typedef struct
 
 /* function prototypes in nodes/tidbitmap.c */
 
-extern TIDBitmap *tbm_create(long maxbytes);
+extern TIDBitmap *tbm_create(long maxbytes, dsa_area *dsa);
 extern void tbm_free(TIDBitmap *tbm);
+extern void tbm_free_shared_area(dsa_area *dsa, dsa_pointer dp);
 
 extern void tbm_add_tuples(TIDBitmap *tbm,
 			   const ItemPointer tids, int ntids,
@@ -60,7 +63,12 @@ extern void tbm_intersect(TIDBitmap *a, const TIDBitmap *b);
 extern bool tbm_is_empty(const TIDBitmap *tbm);
 
 extern TBMIterator *tbm_begin_iterate(TIDBitmap *tbm);
+extern dsa_pointer tbm_prepare_shared_iterate(TIDBitmap *tbm);
 extern TBMIterateResult *tbm_iterate(TBMIterator *iterator);
+extern TBMIterateResult *tbm_shared_iterate(TBMSharedIterator *iterator);
 extern void tbm_end_iterate(TBMIterator *iterator);
+extern void tbm_end_shared_iterate(TBMSharedIterator *iterator);
+extern TBMSharedIterator *tbm_attach_shared_iterate(dsa_area *dsa,
+						  dsa_pointer dp);
 
 #endif   /* TIDBITMAP_H */

src/include/storage/lwlock.h

@@ -212,6 +212,7 @@ typedef enum BuiltinTrancheIds
 	LWTRANCHE_LOCK_MANAGER,
 	LWTRANCHE_PREDICATE_LOCK_MANAGER,
 	LWTRANCHE_PARALLEL_QUERY_DSA,
+	LWTRANCHE_TBM,
 	LWTRANCHE_FIRST_USER_DEFINED
 }	BuiltinTrancheIds;