[src/trunk]: src/sys/sys Add an additional usage note about using __predict_*...

[thorpej <thorpej%NetBSD.org@localhost>]

details:   https://anonhg.NetBSD.org/src/rev/a226ea1a14fe
branches:  trunk
changeset: 485927:a226ea1a14fe
user:      thorpej <thorpej%NetBSD.org@localhost>
date:      Mon May 08 22:41:38 2000 +0000

description:
Add an additional usage note about using __predict_*() only in hot spots,
as suggested by Chris Demetriou.

diffstat:

 sys/sys/cdefs.h |  8 +++++++-
 1 files changed, 7 insertions(+), 1 deletions(-)

diffs (22 lines):

diff -r 1acbc84e0387 -r a226ea1a14fe sys/sys/cdefs.h
--- a/sys/sys/cdefs.h   Mon May 08 21:58:40 2000 +0000
+++ b/sys/sys/cdefs.h   Mon May 08 22:41:38 2000 +0000
@@ -1,4 +1,4 @@
-/*     $NetBSD: cdefs.h,v 1.33 2000/05/08 18:36:00 thorpej Exp $       */
+/*     $NetBSD: cdefs.h,v 1.34 2000/05/08 22:41:38 thorpej Exp $       */
 
 /*
  * Copyright (c) 1991, 1993
@@ -215,6 +215,12 @@
  *     * Other than that, if you don't know the liklyhood of a test
  *       succeeding from empirical or other `hard' evidence, don't
  *       make predictions.
+ *
+ *     * These are meant to be used in places that are run `a lot'.
+ *       It is wasteful to make predictions in code that is run
+ *       seldomly (e.g. at subsystem initialization time) as the
+ *       basic block reordering that this affects can often generate
+ *       larger code.
  */
 #if __GNUC_PREREQ__(2, 96)
 #define        __predict_true(exp)     __builtin_expect(((exp) != 0), 1)