From: Andres Freund <andres@anarazel.de>
Date: Thu, 28 Feb 2019 21:55:27 +0000 (-0800)
Subject: Don't force materializing when copying a buffer tuple table slot.
X-Git-Tag: REL_12_BETA1~646
X-Git-Url: https://granicus.if.org/sourcecode?a=commitdiff_plain;h=8f0577386e68c3b1c83a07b95756e5ee3f4ae73f;p=postgresql

Don't force materializing when copying a buffer tuple table slot.

After 5408e233f0667478 it's not necessary to force materializing the
target slot when copying from one buffer slot to another. Previously
that was required because the HeapTupleData portion of the source slot
wasn't guaranteed to stay valid long enough, but now we can simply
copy that part into the destination slot's tupdata.

Author: Andres Freund
---

diff --git a/src/backend/executor/execTuples.c b/src/backend/executor/execTuples.c
index 121649f343..e194867049 100644
--- a/src/backend/executor/execTuples.c
+++ b/src/backend/executor/execTuples.c
@@ -747,10 +747,12 @@ tts_buffer_heap_copyslot(TupleTableSlot *dstslot, TupleTableSlot *srcslot)
 
 	/*
 	 * If the source slot is of a different kind, or is a buffer slot that has
-	 * been materialized, make a new copy of the tuple.
+	 * been materialized / is virtual, make a new copy of the tuple. Otherwise
+	 * make a new reference to the in-buffer tuple.
 	 */
 	if (dstslot->tts_ops != srcslot->tts_ops ||
-		TTS_SHOULDFREE(srcslot))
+		TTS_SHOULDFREE(srcslot) ||
+		!bsrcslot->base.tuple)
 	{
 		MemoryContext oldContext;
 
@@ -763,18 +765,19 @@ tts_buffer_heap_copyslot(TupleTableSlot *dstslot, TupleTableSlot *srcslot)
 	}
 	else
 	{
-		if (!bsrcslot->base.tuple)
-			tts_buffer_heap_materialize(srcslot);
+		Assert(BufferIsValid(bsrcslot->buffer));
 
 		tts_buffer_heap_store_tuple(dstslot, bsrcslot->base.tuple,
 									bsrcslot->buffer, false);
 
 		/*
-		 * Need to materialize because the HeapTupleData portion of the tuple
-		 * might be in a foreign memory context. That's annoying, but until
-		 * that's moved into the slot, unavoidable.
+		 * The HeapTupleData portion of the source tuple might be shorter
+		 * lived than the destination slot. Therefore copy the HeapTuple into
+		 * our slot's tupdata, which is guaranteed to live long enough (but
+		 * will still point into the buffer).
 		 */
-		tts_buffer_heap_materialize(dstslot);
+		memcpy(&bdstslot->base.tupdata, bdstslot->base.tuple, sizeof(HeapTupleData));
+		bdstslot->base.tuple = &bdstslot->base.tupdata;
 	}
 }