From: Marko Kreen <markokr@gmail.com>
Date: Sat, 11 Aug 2007 07:01:13 +0000 (+0000)
Subject: comment out the recv() avoidance logic
X-Git-Tag: pgbouncer_1_1~60
X-Git-Url: https://granicus.if.org/sourcecode?a=commitdiff_plain;h=cb4d8ca5e09d4be20348d32f6540e6c9fcad4098;p=pgbouncer

comment out the recv() avoidance logic
---

diff --git a/src/sbuf.c b/src/sbuf.c
index 40eb41a..057b5bf 100644
--- a/src/sbuf.c
+++ b/src/sbuf.c
@@ -175,25 +175,23 @@ void sbuf_pause(SBuf *sbuf)
 /* resume from pause, start waiting for data */
 void sbuf_continue(SBuf *sbuf)
 {
-	bool do_recv = true;
+	bool do_recv = DO_RECV;
 	AssertActive(sbuf);
 
 	sbuf_wait_for_data(sbuf);
 
 	/*
-	 * FIXME: is it safe?
+	 * It's tempting to try to avoid the recv() but that would
+	 * only work if no code wants to see full packet.
 	 *
-	 * if avail > SMALL_PKT in buffer, skip the recv().
+	 * This is not true in ServerParameter case.
 	 */
-	if (sbuf->recv_pos - sbuf->pkt_pos >= SMALL_PKT)
-		do_recv = false;
-
 	/*
-	 * There may be some data already received,
-	 * but not certain, so avoid SKIP_RECV.
-	 * Anyway, it affects only client sockets.
+	 * if (sbuf->recv_pos - sbuf->pkt_pos >= SMALL_PKT)
+	 *	do_recv = false;
 	 */
-	sbuf_main_loop(sbuf, DO_RECV);
+
+	sbuf_main_loop(sbuf, do_recv);
 }
 
 /*