In gnuradio-core/src/lib/general/random.h, RNADOM_MAX is defined
unconditionally:
Recently NetBSD-current has added RANDOM_MAX, for essentially the same
reason, to parallel with RAND_MAX. (random(3)'s definition is 0 to
2**31-1, so this shouldn’t be different).
So, is it ok to commit this patch?
I would also update the comment (it’s not a bug that Solaris RAND_MAX is
32767 if rand(3) on Solaris returns from 0 to 215-1 - standards don’t
require 231-1 that I know of)?
$NetBSD: patch-ag,v 1.3 2008/03/15 15:09:54 joerg Exp $
— gnuradio-core/src/lib/general/random.h.orig 2008-03-15
15:34:46.000000000 +0
100
+++ gnuradio-core/src/lib/general/random.h
@@ -24,8 +24,9 @@
#define RANDOM_H
// we use this because some systems (solaris) define RAND_MAX as 32767
+#ifndef RANDOM_MAX
static const int RANDOM_MAX = 2147483647;
+#endif
#include <stdlib.h>