Fast ALTER TABLE ADD COLUMN with a non-NULL default

Currently adding a column to a table with a non-NULL default results in
a rewrite of the table. For large tables this can be both expensive and
disruptive. This patch removes the need for the rewrite as long as the
default value is not volatile. The default expression is evaluated at
the time of the ALTER TABLE and the result stored in a new column
(attmissingval) in pg_attribute, and a new column (atthasmissing) is set
to true. Any existing row when fetched will be supplied with the
attmissingval. New rows will have the supplied value or the default and
so will never need the attmissingval.

Any time the table is rewritten all the atthasmissing and attmissingval
settings for the attributes are cleared, as they are no longer needed.

The most visible code change from this is in heap_attisnull, which
acquires a third TupleDesc argument, allowing it to detect a missing
value if there is one. In many cases where it is known that there will
not be any (e.g.  catalog relations) NULL can be passed for this
argument.

Andrew Dunstan, heavily modified from an original patch from Serge
Rielau.
Reviewed by Tom Lane, Andres Freund, Tomas Vondra and David Rowley.

Discussion: https://postgr.es/m/31e2e921-7002-4c27-59f5-51f08404c858@2ndQuadrant.com

1 files changed, 70 insertions, 12 deletions

diff --git a/src/backend/utils/cache/relcache.c b/src/backend/utils/cache/relcache.c
index de502f9bc9e..48f92dc430d 100644
--- a/src/backend/utils/cache/relcache.c
+++ b/src/backend/utils/cache/relcache.c
@@ -36,6 +36,7 @@
 #include "access/nbtree.h"
 #include "access/reloptions.h"
 #include "access/sysattr.h"
+#include "access/tupdesc_details.h"
 #include "access/xact.h"
 #include "access/xlog.h"
 #include "catalog/catalog.h"
@@ -79,6 +80,7 @@
 #include "storage/smgr.h"
 #include "utils/array.h"
 #include "utils/builtins.h"
+#include "utils/datum.h"
 #include "utils/fmgroids.h"
 #include "utils/inval.h"
 #include "utils/lsyscache.h"
@@ -495,6 +497,7 @@ RelationBuildTupleDesc(Relation relation)
 	int			need;
 	TupleConstr *constr;
 	AttrDefault *attrdef = NULL;
+	AttrMissing *attrmiss = NULL;
 	int			ndef = 0;
 
 	/* copy some fields from pg_class row to rd_att */
@@ -540,15 +543,17 @@ RelationBuildTupleDesc(Relation relation)
 	while (HeapTupleIsValid(pg_attribute_tuple = systable_getnext(pg_attribute_scan)))
 	{
 		Form_pg_attribute attp;
+		int			attnum;
 
 		attp = (Form_pg_attribute) GETSTRUCT(pg_attribute_tuple);
 
-		if (attp->attnum <= 0 ||
-			attp->attnum > relation->rd_rel->relnatts)
+		attnum = attp->attnum;
+		if (attnum <= 0 || attnum > relation->rd_rel->relnatts)
 			elog(ERROR, "invalid attribute number %d for %s",
 				 attp->attnum, RelationGetRelationName(relation));
 
-		memcpy(TupleDescAttr(relation->rd_att, attp->attnum - 1),
+
+		memcpy(TupleDescAttr(relation->rd_att, attnum - 1),
 			   attp,
 			   ATTRIBUTE_FIXED_PART_SIZE);
 
@@ -556,6 +561,7 @@ RelationBuildTupleDesc(Relation relation)
 		if (attp->attnotnull)
 			constr->has_not_null = true;
 
+		/* If the column has a default, fill it into the attrdef array */
 		if (attp->atthasdef)
 		{
 			if (attrdef == NULL)
@@ -563,10 +569,63 @@ RelationBuildTupleDesc(Relation relation)
 					MemoryContextAllocZero(CacheMemoryContext,
 										   relation->rd_rel->relnatts *
 										   sizeof(AttrDefault));
-			attrdef[ndef].adnum = attp->attnum;
+			attrdef[ndef].adnum = attnum;
 			attrdef[ndef].adbin = NULL;
+
 			ndef++;
 		}
+
+		/* Likewise for a missing value */
+		if (attp->atthasmissing)
+		{
+			Datum		missingval;
+			bool		missingNull;
+
+			/* Do we have a missing value? */
+			missingval = heap_getattr(pg_attribute_tuple,
+									  Anum_pg_attribute_attmissingval,
+									  pg_attribute_desc->rd_att,
+									  &missingNull);
+			if (!missingNull)
+			{
+				/* Yes, fetch from the array */
+				MemoryContext oldcxt;
+				bool		is_null;
+				int			one = 1;
+				Datum		missval;
+
+				if (attrmiss == NULL)
+					attrmiss = (AttrMissing *)
+						MemoryContextAllocZero(CacheMemoryContext,
+											   relation->rd_rel->relnatts *
+											   sizeof(AttrMissing));
+
+				missval = array_get_element(missingval,
+											1,
+											&one,
+											-1,
+											attp->attlen,
+											attp->attbyval,
+											attp->attalign,
+											&is_null);
+				Assert(!is_null);
+				if (attp->attbyval)
+				{
+					/* for copy by val just copy the datum direct */
+					attrmiss[attnum - 1].ammissing = missval;
+				}
+				else
+				{
+					/* otherwise copy in the correct context */
+					oldcxt = MemoryContextSwitchTo(CacheMemoryContext);
+					attrmiss[attnum - 1].ammissing = datumCopy(missval,
+															   attp->attbyval,
+															   attp->attlen);
+					MemoryContextSwitchTo(oldcxt);
+				}
+				attrmiss[attnum - 1].ammissingPresent = true;
+			}
+		}
 		need--;
 		if (need == 0)
 			break;
@@ -607,7 +666,8 @@ RelationBuildTupleDesc(Relation relation)
 	/*
 	 * Set up constraint/default info
 	 */
-	if (constr->has_not_null || ndef > 0 || relation->rd_rel->relchecks)
+	if (constr->has_not_null || ndef > 0 ||
+		attrmiss || relation->rd_rel->relchecks)
 	{
 		relation->rd_att->constr = constr;
 
@@ -624,6 +684,8 @@ RelationBuildTupleDesc(Relation relation)
 		else
 			constr->num_defval = 0;
 
+		constr->missing = attrmiss;
+
 		if (relation->rd_rel->relchecks > 0)	/* CHECKs */
 		{
 			constr->num_check = relation->rd_rel->relchecks;
@@ -4063,10 +4125,6 @@ AttrDefaultFetch(Relation relation)
 
 	systable_endscan(adscan);
 	heap_close(adrel, AccessShareLock);
-
-	if (found != ndef)
-		elog(WARNING, "%d attrdef record(s) missing for rel %s",
-			 ndef - found, RelationGetRelationName(relation));
 }
 
 /*
@@ -4405,7 +4463,7 @@ RelationGetIndexList(Relation relation)
 		 */
 		if (!IndexIsValid(index) || !index->indisunique ||
 			!index->indimmediate ||
-			!heap_attisnull(htup, Anum_pg_index_indpred))
+			!heap_attisnull(htup, Anum_pg_index_indpred, NULL))
 			continue;
 
 		/* Check to see if is a usable btree index on OID */
@@ -4700,7 +4758,7 @@ RelationGetIndexExpressions(Relation relation)
 
 	/* Quick exit if there is nothing to do. */
 	if (relation->rd_indextuple == NULL ||
-		heap_attisnull(relation->rd_indextuple, Anum_pg_index_indexprs))
+		heap_attisnull(relation->rd_indextuple, Anum_pg_index_indexprs, NULL))
 		return NIL;
 
 	/*
@@ -4762,7 +4820,7 @@ RelationGetIndexPredicate(Relation relation)
 
 	/* Quick exit if there is nothing to do. */
 	if (relation->rd_indextuple == NULL ||
-		heap_attisnull(relation->rd_indextuple, Anum_pg_index_indpred))
+		heap_attisnull(relation->rd_indextuple, Anum_pg_index_indpred, NULL))
 		return NIL;
 
 	/*