--- /dev/null
+++ Libc/Libc-594.9.4/gen/FreeBSD/rand48.h.patch
@@ -0,0 +1,67 @@
+--- rand48.h.orig	2003-05-20 15:21:02.000000000 -0700
++++ rand48.h	2005-11-03 14:06:17.000000000 -0800
+@@ -19,8 +19,6 @@
+ #include <math.h>
+ #include <stdlib.h>
+ 
+-void		_dorand48(unsigned short[3]);
+-
+ #define	RAND48_SEED_0	(0x330e)
+ #define	RAND48_SEED_1	(0xabcd)
+ #define	RAND48_SEED_2	(0x1234)
+@@ -29,4 +27,55 @@
+ #define	RAND48_MULT_2	(0x0005)
+ #define	RAND48_ADD	(0x000b)
+ 
++typedef unsigned long long uint48;
++
++extern uint48 _rand48_seed;
++extern uint48 _rand48_mult;
++extern uint48 _rand48_add;
++
++#define TOUINT48(x,y,z) \
++	((uint48)(x) + (((uint48)(y)) << 16) + (((uint48)(z)) << 32))
++
++#define RAND48_SEED	TOUINT48(RAND48_SEED_0, RAND48_SEED_1, RAND48_SEED_2)
++#define RAND48_MULT	TOUINT48(RAND48_MULT_0, RAND48_MULT_1, RAND48_MULT_2)
++
++#define LOADRAND48(l,x) \
++	(l) = TOUINT48((x)[0], (x)[1], (x)[2])
++
++#define STORERAND48(l,x) \
++	(x)[0] = (unsigned short)(l); \
++	(x)[1] = (unsigned short)((l) >> 16); \
++	(x)[2] = (unsigned short)((l) >> 32)
++
++#define _DORAND48(l) \
++	(l) = (l) * _rand48_mult + _rand48_add
++
++#define DORAND48(l,x) \
++	LOADRAND48(l, x); \
++	_DORAND48(l); \
++	STORERAND48(l, x)
++
++#include "fpmath.h"
++
++/*
++ * Optimization for speed: avoid int-to-double conversion.  Assume doubles
++ * are IEEE-754 and insert the bits directly.  To normalize, the (1 << 52)
++ * is the hidden bit, which the first set bit is shifted to.
++ */
++#define ERAND48_BEGIN \
++	union { \
++		union IEEEd2bits ieee; \
++		unsigned long long l; \
++	} u; \
++	int s
++
++#define ERAND48_END(x) \
++	u.l = ((x) & 0xffffffffffffULL); \
++	if (u.l == 0) \
++		return 0.0; \
++	u.l <<= 5; \
++	for(s = 0; !(u.l & (1LL << 52)); s++, u.l <<= 1) {} \
++	u.ieee.bits.exp = 1022 - s; \
++	return u.ieee.d
++
+ #endif /* _RAND48_H_ */