Reimplementation of Random using an LXM pseudo-random number generator

runtime/Makefile

@@ -27,7 +27,7 @@ BYTECODE_C_SOURCES := $(addsuffix .c, \
   lexing callback debugger weak finalise custom dynlink \
   platform fiber shared_heap addrmap \
   afl $(UNIX_OR_WIN32) bigarray main memprof domain sync \
-  skiplist lf_skiplist codefrag)
+  skiplist lf_skiplist codefrag prng)
 
 NATIVE_C_SOURCES := $(addsuffix .c, \
   startup_aux startup_nat main fail_nat roots signals \
@@ -37,7 +37,7 @@ NATIVE_C_SOURCES := $(addsuffix .c, \
   globroots backtrace_nat backtrace dynlink_nat debugger meta \
   platform fiber shared_heap addrmap frame_descriptors \
   dynlink clambda_checks afl bigarray \
-  memprof domain sync skiplist lf_skiplist codefrag)
+  memprof domain sync skiplist lf_skiplist codefrag prng)
 
 # Header files generated by configure
 CONFIGURED_HEADERS := caml/m.h caml/s.h caml/version.h

runtime/gen_primitives.sh

@@ -26,7 +26,7 @@ export LC_ALL=C
       lexing md5 meta memprof obj parsing signals str sys callback weak \
       finalise domain platform fiber memory startup_aux sync \
       dynlink backtrace_byt backtrace afl \
-      bigarray eventlog
+      bigarray eventlog prng
   do
       sed -n -e 's/^CAMLprim value \([a-z0-9_][a-z0-9_]*\).*/\1/p' "$prim.c"
   done

runtime/prng.c

@@ -0,0 +1,69 @@
+/**************************************************************************/
+/*                                                                        */
+/*                                 OCaml                                  */
+/*                                                                        */
+/*          Xavier Leroy, projet Cambium, College de France and Inria     */
+/*                                                                        */
+/*   Copyright 2021 Institut National de Recherche en Informatique et     */
+/*     en Automatique.                                                    */
+/*                                                                        */
+/*   All rights reserved.  This file is distributed under the terms of    */
+/*   the GNU Lesser General Public License version 2.1, with the          */
+/*   special exception on linking described in the file LICENSE.          */
+/*                                                                        */
+/**************************************************************************/
+
+#include <string.h>
+#include "caml/alloc.h"
+#include "caml/bigarray.h"
+#include "caml/mlvalues.h"
+
+/* The L64X128 member of the LXM family.  Taken from figure 1 in
+   "LXM: Better Splittable Pseudorandom Number Generators
+    (and Almost as Fast)" by Guy L. Steele Jr. and Sebastiano Vigna,
+    OOPSLA 2021.  */
+
+static const uint64_t M = 0xd1342543de82ef95;
+
+struct LXM_state {
+  uint64_t a;            /* per-instance additive parameter (odd) */
+  uint64_t s;            /* state of the LCG subgenerator */
+  uint64_t x[2];         /* state of the XBG subgenerator (not 0) */
+};
+
+/* In OCaml, states are represented as a 1D big array of 64-bit integers */
+
+#define LXM_val(v) ((struct LXM_state *) Caml_ba_data_val(v))
+
+Caml_inline uint64_t rotl(const uint64_t x, int k) {
+  return (x << k) | (x >> (64 - k));
+}
+
+CAMLprim uint64_t caml_lxm_next_unboxed(value v)
+{
+  uint64_t z, q0, q1;
+  struct LXM_state * st = LXM_val(v);
+
+  /* Combining operation */
+  z = st->s + st->x[0];
+  /* Mixing function */
+  z = (z ^ (z >> 32)) * 0xdaba0b6eb09322e3;
+  z = (z ^ (z >> 32)) * 0xdaba0b6eb09322e3;
+  z = (z ^ (z >> 32));
+  /* LCG update */
+  st->s = st->s * M + st->a;
+  /* XBG update */
+  q0 = st->x[0]; q1 = st->x[1];
+  q1 ^= q0;
+  q0 = rotl(q0, 24);
+  q0 = q0 ^ q1 ^ (q1 << 16);
+  q1 = rotl(q1, 37);
+  st->x[0] = q0; st->x[1] = q1;
+  /* Return result */
+  return z;
+}
+
+CAMLprim value caml_lxm_next(value v)
+{
+  return caml_copy_int64(caml_lxm_next_unboxed(v));
+}

stdlib/.depend

@@ -623,27 +623,23 @@ stdlib__Queue.cmx : queue.ml \
 stdlib__Queue.cmi : queue.mli \
     stdlib__Seq.cmi
 stdlib__Random.cmo : random.ml \
-    stdlib__String.cmi \
-    stdlib.cmi \
     stdlib__Nativeint.cmi \
     stdlib__Int64.cmi \
     stdlib__Int32.cmi \
-    stdlib__Int.cmi \
     stdlib__Domain.cmi \
     stdlib__Digest.cmi \
-    stdlib__Char.cmi \
+    stdlib__Bytes.cmi \
+    stdlib__Bigarray.cmi \
     stdlib__Array.cmi \
     stdlib__Random.cmi
 stdlib__Random.cmx : random.ml \
-    stdlib__String.cmx \
-    stdlib.cmx \
     stdlib__Nativeint.cmx \
     stdlib__Int64.cmx \
     stdlib__Int32.cmx \
-    stdlib__Int.cmx \
     stdlib__Domain.cmx \
     stdlib__Digest.cmx \
-    stdlib__Char.cmx \
+    stdlib__Bytes.cmx \
+    stdlib__Bigarray.cmx \
     stdlib__Array.cmx \
     stdlib__Random.cmi
 stdlib__Random.cmi : random.mli \

stdlib/StdlibModules

@@ -42,10 +42,10 @@ STDLIB_MODULE_BASENAMES = \
   lexing parsing set map stack queue stream buffer \
   atomic mutex condition semaphore domain \
   camlinternalFormat printf arg \
-  printexc fun gc digest random hashtbl weak \
+  printexc fun gc digest bigarray random hashtbl weak \
   format scanf callback camlinternalOO oo camlinternalMod genlex ephemeron \
   filename complex arrayLabels listLabels bytesLabels stringLabels moreLabels \
-  stdLabels bigarray in_channel out_channel effect
+  stdLabels in_channel out_channel effect
 
 STDLIB_PREFIXED_MODULES = \
   $(filter-out stdlib camlinternal%, $(STDLIB_MODULE_BASENAMES))