From c81db872fde53a9dda33a5c80be7114d79d91606 Mon Sep 17 00:00:00 2001
From: Patric Mueller <bhaak@gmx.net>
Date: Sat, 12 Jan 2019 21:36:08 +0100
Subject: [PATCH] add file for the isaac64 random number generator

This is the version from the Comprehensive C Archive Network, licensed
under the CC0 "No Rights Reserved" Creative Common License.
http://ccodearchive.net/info/isaac.html
---
 include/isaac64.h | 131 ++++++++++++++++++++++++
 src/isaac64.c     | 255 ++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 386 insertions(+)
 create mode 100644 include/isaac64.h
 create mode 100644 src/isaac64.c

diff --git a/include/isaac64.h b/include/isaac64.h
new file mode 100644
index 000000000..5c22253bd
--- /dev/null
+++ b/include/isaac64.h
@@ -0,0 +1,131 @@
+/* CC0 (Public domain) - see LICENSE file for details */
+#if !defined(_isaac64_H)
+# define _isaac64_H (1)
+# include <stdint.h>
+
+
+
+typedef struct isaac64_ctx isaac64_ctx;
+
+
+
+#define ISAAC64_SZ_LOG      (8)
+#define ISAAC64_SZ          (1<<ISAAC64_SZ_LOG)
+#define ISAAC64_SEED_SZ_MAX (ISAAC64_SZ<<3)
+
+
+
+/*ISAAC is the most advanced of a series of pseudo-random number generators
+   designed by Robert J. Jenkins Jr. in 1996.
+  http://www.burtleburtle.net/bob/rand/isaac.html
+  This is the 64-bit version.
+  To quote:
+    ISAAC-64 generates a different sequence than ISAAC, but it uses the same
+     principles.
+    It uses 64-bit arithmetic.
+    It generates a 64-bit result every 19 instructions.
+    All cycles are at least 2**72 values, and the average cycle length is
+     2**16583.*/
+struct isaac64_ctx{
+  unsigned n;
+  uint64_t r[ISAAC64_SZ];
+  uint64_t m[ISAAC64_SZ];
+  uint64_t a;
+  uint64_t b;
+  uint64_t c;
+};
+
+
+/**
+ * isaac64_init - Initialize an instance of the ISAAC64 random number generator.
+ * @_ctx:   The ISAAC64 instance to initialize.
+ * @_seed:  The specified seed bytes.
+ *          This may be NULL if _nseed is less than or equal to zero.
+ * @_nseed: The number of bytes to use for the seed.
+ *          If this is greater than ISAAC64_SEED_SZ_MAX, the extra bytes are
+ *           ignored.
+ */
+void isaac64_init(isaac64_ctx *_ctx,const unsigned char *_seed,int _nseed);
+
+/**
+ * isaac64_reseed - Mix a new batch of entropy into the current state.
+ * To reset ISAAC64 to a known state, call isaac64_init() again instead.
+ * @_ctx:   The instance to reseed.
+ * @_seed:  The specified seed bytes.
+ *          This may be NULL if _nseed is zero.
+ * @_nseed: The number of bytes to use for the seed.
+ *          If this is greater than ISAAC64_SEED_SZ_MAX, the extra bytes are
+ *           ignored.
+ */
+void isaac64_reseed(isaac64_ctx *_ctx,const unsigned char *_seed,int _nseed);
+/**
+ * isaac64_next_uint64 - Return the next random 64-bit value.
+ * @_ctx: The ISAAC64 instance to generate the value with.
+ */
+uint64_t isaac64_next_uint64(isaac64_ctx *_ctx);
+/**
+ * isaac64_next_uint - Uniform random integer less than the given value.
+ * @_ctx: The ISAAC64 instance to generate the value with.
+ * @_n:   The upper bound on the range of numbers returned (not inclusive).
+ *        This must be greater than zero and less than 2**64.
+ *        To return integers in the full range 0...2**64-1, use
+ *         isaac64_next_uint64() instead.
+ * Return: An integer uniformly distributed between 0 and _n-1 (inclusive).
+ */
+uint64_t isaac64_next_uint(isaac64_ctx *_ctx,uint64_t _n);
+/**
+ * isaac64_next_float - Uniform random float in the range [0,1).
+ * @_ctx: The ISAAC64 instance to generate the value with.
+ * Returns a high-quality float uniformly distributed between 0 (inclusive)
+ *  and 1 (exclusive).
+ * All of the float's mantissa bits are random, e.g., the least significant bit
+ *  may still be non-zero even if the value is less than 0.5, and any
+ *  representable float in the range [0,1) has a chance to be returned, though
+ *  values very close to zero become increasingly unlikely.
+ * To generate cheaper float values that do not have these properties, use
+ *   ldexpf((float)isaac64_next_uint64(_ctx),-64);
+ */
+float isaac64_next_float(isaac64_ctx *_ctx);
+/**
+ * isaac64_next_signed_float - Uniform random float in the range (-1,1).
+ * @_ctx: The ISAAC64 instance to generate the value with.
+ * Returns a high-quality float uniformly distributed between -1 and 1
+ *  (exclusive).
+ * All of the float's mantissa bits are random, e.g., the least significant bit
+ *  may still be non-zero even if the magnitude is less than 0.5, and any
+ *  representable float in the range (-1,1) has a chance to be returned, though
+ *  values very close to zero become increasingly unlikely.
+ * To generate cheaper float values that do not have these properties, use
+ *   ldexpf((float)isaac64_next_uint64(_ctx),-63)-1;
+ *  though this returns values in the range [-1,1).
+ */
+float isaac64_next_signed_float(isaac64_ctx *_ctx);
+/**
+ * isaac64_next_double - Uniform random double in the range [0,1).
+ * @_ctx: The ISAAC64 instance to generate the value with.
+ * Returns a high-quality double uniformly distributed between 0 (inclusive)
+ *  and 1 (exclusive).
+ * All of the double's mantissa bits are random, e.g., the least significant
+ *  bit may still be non-zero even if the value is less than 0.5, and any
+ *  representable double in the range [0,1) has a chance to be returned, though
+ *  values very close to zero become increasingly unlikely.
+ * To generate cheaper double values that do not have these properties, use
+ *   ldexp((double)isaac64_next_uint64(_ctx),-64);
+ */
+double isaac64_next_double(isaac64_ctx *_ctx);
+/**
+ * isaac64_next_signed_double - Uniform random double in the range (-1,1).
+ * @_ctx: The ISAAC64 instance to generate the value with.
+ * Returns a high-quality double uniformly distributed between -1 and 1
+ *  (exclusive).
+ * All of the double's mantissa bits are random, e.g., the least significant
+ *  bit may still be non-zero even if the value is less than 0.5, and any
+ *  representable double in the range (-1,1) has a chance to be returned,
+ *  though values very close to zero become increasingly unlikely.
+ * To generate cheaper double values that do not have these properties, use
+ *   ldexp((double)isaac64_next_uint64(_ctx),-63)-1;
+ *  though this returns values in the range [-1,1).
+ */
+double isaac64_next_signed_double(isaac64_ctx *_ctx);
+
+#endif
diff --git a/src/isaac64.c b/src/isaac64.c
new file mode 100644
index 000000000..3a392fc9b
--- /dev/null
+++ b/src/isaac64.c
@@ -0,0 +1,255 @@
+/*Written by Timothy B. Terriberry (tterribe@xiph.org) 1999-2009
+  CC0 (Public domain) - see LICENSE file for details
+  Based on the public domain ISAAC implementation by Robert J. Jenkins Jr.*/
+#include <float.h>
+#include <math.h>
+#include <string.h>
+#include <ccan/ilog/ilog.h>
+#include "isaac64.h"
+
+
+#define ISAAC64_MASK ((uint64_t)0xFFFFFFFFFFFFFFFFULL)
+
+/* Extract ISAAC64_SZ_LOG bits (starting at bit 3). */
+static inline uint32_t lower_bits(uint64_t x)
+{
+	return (x & ((ISAAC64_SZ-1) << 3)) >>3;
+}
+
+/* Extract next ISAAC64_SZ_LOG bits (starting at bit ISAAC64_SZ_LOG+2). */
+static inline uint32_t upper_bits(uint32_t y)
+{
+	return (y >> (ISAAC64_SZ_LOG+3)) & (ISAAC64_SZ-1);
+}
+
+static void isaac64_update(isaac64_ctx *_ctx){
+  uint64_t *m;
+  uint64_t *r;
+  uint64_t  a;
+  uint64_t  b;
+  uint64_t  x;
+  uint64_t  y;
+  int       i;
+  m=_ctx->m;
+  r=_ctx->r;
+  a=_ctx->a;
+  b=_ctx->b+(++_ctx->c);
+  for(i=0;i<ISAAC64_SZ/2;i++){
+    x=m[i];
+    a=~(a^a<<21)+m[i+ISAAC64_SZ/2];
+    m[i]=y=m[lower_bits(x)]+a+b;
+    r[i]=b=m[upper_bits(y)]+x;
+    x=m[++i];
+    a=(a^a>>5)+m[i+ISAAC64_SZ/2];
+    m[i]=y=m[lower_bits(x)]+a+b;
+    r[i]=b=m[upper_bits(y)]+x;
+    x=m[++i];
+    a=(a^a<<12)+m[i+ISAAC64_SZ/2];
+    m[i]=y=m[lower_bits(x)]+a+b;
+    r[i]=b=m[upper_bits(y)]+x;
+    x=m[++i];
+    a=(a^a>>33)+m[i+ISAAC64_SZ/2];
+    m[i]=y=m[lower_bits(x)]+a+b;
+    r[i]=b=m[upper_bits(y)]+x;
+  }
+  for(i=ISAAC64_SZ/2;i<ISAAC64_SZ;i++){
+    x=m[i];
+    a=~(a^a<<21)+m[i-ISAAC64_SZ/2];
+    m[i]=y=m[lower_bits(x)]+a+b;
+    r[i]=b=m[upper_bits(y)]+x;
+    x=m[++i];
+    a=(a^a>>5)+m[i-ISAAC64_SZ/2];
+    m[i]=y=m[lower_bits(x)]+a+b;
+    r[i]=b=m[upper_bits(y)]+x;
+    x=m[++i];
+    a=(a^a<<12)+m[i-ISAAC64_SZ/2];
+    m[i]=y=m[lower_bits(x)]+a+b;
+    r[i]=b=m[upper_bits(y)]+x;
+    x=m[++i];
+    a=(a^a>>33)+m[i-ISAAC64_SZ/2];
+    m[i]=y=m[lower_bits(x)]+a+b;
+    r[i]=b=m[upper_bits(y)]+x;
+  }
+  _ctx->b=b;
+  _ctx->a=a;
+  _ctx->n=ISAAC64_SZ;
+}
+
+static void isaac64_mix(uint64_t _x[8]){
+  static const unsigned char SHIFT[8]={9,9,23,15,14,20,17,14};
+  int i;
+  for(i=0;i<8;i++){
+    _x[i]-=_x[(i+4)&7];
+    _x[(i+5)&7]^=_x[(i+7)&7]>>SHIFT[i];
+    _x[(i+7)&7]+=_x[i];
+    i++;
+    _x[i]-=_x[(i+4)&7];
+    _x[(i+5)&7]^=_x[(i+7)&7]<<SHIFT[i];
+    _x[(i+7)&7]+=_x[i];
+  }
+}
+
+
+void isaac64_init(isaac64_ctx *_ctx,const unsigned char *_seed,int _nseed){
+  _ctx->a=_ctx->b=_ctx->c=0;
+  memset(_ctx->r,0,sizeof(_ctx->r));
+  isaac64_reseed(_ctx,_seed,_nseed);
+}
+
+void isaac64_reseed(isaac64_ctx *_ctx,const unsigned char *_seed,int _nseed){
+  uint64_t *m;
+  uint64_t *r;
+  uint64_t  x[8];
+  int       i;
+  int       j;
+  m=_ctx->m;
+  r=_ctx->r;
+  if(_nseed>ISAAC64_SEED_SZ_MAX)_nseed=ISAAC64_SEED_SZ_MAX;
+  for(i=0;i<_nseed>>3;i++){
+    r[i]^=(uint64_t)_seed[i<<3|7]<<56|(uint64_t)_seed[i<<3|6]<<48|
+     (uint64_t)_seed[i<<3|5]<<40|(uint64_t)_seed[i<<3|4]<<32|
+     (uint64_t)_seed[i<<3|3]<<24|(uint64_t)_seed[i<<3|2]<<16|
+     (uint64_t)_seed[i<<3|1]<<8|_seed[i<<3];
+  }
+  _nseed-=i<<3;
+  if(_nseed>0){
+    uint64_t ri;
+    ri=_seed[i<<3];
+    for(j=1;j<_nseed;j++)ri|=(uint64_t)_seed[i<<3|j]<<(j<<3);
+    r[i++]^=ri;
+  }
+  x[0]=x[1]=x[2]=x[3]=x[4]=x[5]=x[6]=x[7]=(uint64_t)0x9E3779B97F4A7C13ULL;
+  for(i=0;i<4;i++)isaac64_mix(x);
+  for(i=0;i<ISAAC64_SZ;i+=8){
+    for(j=0;j<8;j++)x[j]+=r[i+j];
+    isaac64_mix(x);
+    memcpy(m+i,x,sizeof(x));
+  }
+  for(i=0;i<ISAAC64_SZ;i+=8){
+    for(j=0;j<8;j++)x[j]+=m[i+j];
+    isaac64_mix(x);
+    memcpy(m+i,x,sizeof(x));
+  }
+  isaac64_update(_ctx);
+}
+
+uint64_t isaac64_next_uint64(isaac64_ctx *_ctx){
+  if(!_ctx->n)isaac64_update(_ctx);
+  return _ctx->r[--_ctx->n];
+}
+
+uint64_t isaac64_next_uint(isaac64_ctx *_ctx,uint64_t _n){
+  uint64_t r;
+  uint64_t v;
+  uint64_t d;
+  do{
+    r=isaac64_next_uint64(_ctx);
+    v=r%_n;
+    d=r-v;
+  }
+  while(((d+_n-1)&ISAAC64_MASK)<d);
+  return v;
+}
+
+/*Returns a uniform random float.
+  The expected value is within FLT_MIN (e.g., 1E-37) of 0.5.
+  _bits: An initial set of random bits.
+  _base: This should be -(the number of bits in _bits), up to -64.
+  Return: A float uniformly distributed between 0 (inclusive) and 1
+           (exclusive).
+          The average value was measured over 2**32 samples to be
+           0.499991407275206357.*/
+static float isaac64_float_bits(isaac64_ctx *_ctx,uint64_t _bits,int _base){
+  float ret;
+  int   nbits_needed;
+  while(!_bits){
+    if(_base+FLT_MANT_DIG<FLT_MIN_EXP)return 0;
+    _base-=64;
+    _bits=isaac64_next_uint64(_ctx);
+  }
+  nbits_needed=FLT_MANT_DIG-ilog64_nz(_bits);
+#if FLT_MANT_DIG>64
+  ret=ldexpf((float)_bits,_base);
+# if FLT_MANT_DIG>129
+  while(64-nbits_needed<0){
+# else
+  if(64-nbits_needed<0){
+# endif
+    _base-=64;
+    nbits_needed-=64;
+    ret+=ldexpf((float)isaac64_next_uint64(_ctx),_base);
+  }
+  _bits=isaac64_next_uint64(_ctx)>>(64-nbits_needed);
+  ret+=ldexpf((float)_bits,_base-nbits_needed);
+#else
+  if(nbits_needed>0){
+    _bits=_bits<<nbits_needed|isaac64_next_uint64(_ctx)>>(64-nbits_needed);
+  }
+# if FLT_MANT_DIG<64
+  else _bits>>=-nbits_needed;
+# endif
+  ret=ldexpf((float)_bits,_base-nbits_needed);
+#endif
+  return ret;
+}
+
+float isaac64_next_float(isaac64_ctx *_ctx){
+  return isaac64_float_bits(_ctx,0,0);
+}
+
+float isaac64_next_signed_float(isaac64_ctx *_ctx){
+  uint64_t bits;
+  bits=isaac64_next_uint64(_ctx);
+  return (1|-((int)bits&1))*isaac64_float_bits(_ctx,bits>>1,-63);
+}
+
+/*Returns a uniform random double.
+  _bits: An initial set of random bits.
+  _base: This should be -(the number of bits in _bits), up to -64.
+  Return: A double uniformly distributed between 0 (inclusive) and 1
+           (exclusive).
+          The average value was measured over 2**32 samples to be
+           0.499990992392019273.*/
+static double isaac64_double_bits(isaac64_ctx *_ctx,uint64_t _bits,int _base){
+  double ret;
+  int    nbits_needed;
+  while(!_bits){
+    if(_base+DBL_MANT_DIG<DBL_MIN_EXP)return 0;
+    _base-=64;
+    _bits=isaac64_next_uint64(_ctx);
+  }
+  nbits_needed=DBL_MANT_DIG-ilog64_nz(_bits);
+#if DBL_MANT_DIG>64
+  ret=ldexp((double)_bits,_base);
+# if DBL_MANT_DIG>129
+  while(64-nbits_needed<0){
+# else
+  if(64-nbits_needed<0){
+# endif
+    _base-=64;
+    nbits_needed-=64;
+    ret+=ldexp((double)isaac64_next_uint64(_ctx),_base);
+  }
+  _bits=isaac64_next_uint64(_ctx)>>(64-nbits_needed);
+  ret+=ldexp((double)_bits,_base-nbits_needed);
+#else
+  if(nbits_needed>0){
+    _bits=_bits<<nbits_needed|isaac64_next_uint64(_ctx)>>(64-nbits_needed);
+  }
+# if DBL_MANT_DIG<64
+  else _bits>>=-nbits_needed;
+# endif
+  ret=ldexp((double)_bits,_base-nbits_needed);
+#endif
+  return ret;
+}
+
+double isaac64_next_double(isaac64_ctx *_ctx){
+  return isaac64_double_bits(_ctx,0,0);
+}
+
+double isaac64_next_signed_double(isaac64_ctx *_ctx){
+  uint64_t bits;
+  bits=isaac64_next_uint64(_ctx);
+  return (1|-((int)bits&1))*isaac64_double_bits(_ctx,bits>>1,-63);
+}
-- 
2.40.0