logical decoding: Remove TODO about unnecessary optimization.

anarazel · anarazel · commit 112b006fe79b · 2020-03-24T12:15:03.000-07:00
Measurements show, and intuition agrees, that there's currently no known cases where adding a fastpath to avoid allocating / ordering a heap for a single transaction is worthwhile. Author: Dilip Kumar Discussion: https://postgr.es/m/CAFiTN-sp701wvzvnLQJGk7JDqrFM8f--97-ihbwkU8qvn=p8nw@mail.gmail.com
diff --git a/src/backend/replication/logical/reorderbuffer.c b/src/backend/replication/logical/reorderbuffer.c
@@ -1037,11 +1037,6 @@ ReorderBufferIterTXNInit(ReorderBuffer *rb, ReorderBufferTXN *txn,
 			nr_txns++;
 	}
 
-	/*
-	 * TODO: Consider adding fastpath for the rather common nr_txns=1 case, no
-	 * need to allocate/build a heap then.
-	 */
-
 	/* allocate iteration state */
 	state = (ReorderBufferIterTXNState *)
 		MemoryContextAllocZero(rb->context,
