diff --git a/byterun/extern.c b/byterun/extern.c
index c88179b7b..74a2ee13f 100644
--- a/byterun/extern.c
+++ b/byterun/extern.c
@@ -37,7 +37,7 @@ struct extern_obj {
 static struct extern_obj * extern_table;
 static unsigned long extern_table_size;
 
-#ifdef SIXTYFOUR
+#ifdef ARCH_SIXTYFOUR
 #define Hash(v) (((unsigned long) ((v) >> 3)) % extern_table_size)
 #else
 #define Hash(v) (((unsigned long) ((v) >> 2)) % extern_table_size)
@@ -163,7 +163,7 @@ static void writecode32(code, val)
   extern_ptr += 5;
 }
 
-#ifdef SIXTYFOUR
+#ifdef ARCH_SIXTYFOUR
 static void writecode64(code, val)
      int code;
      long val;
@@ -199,7 +199,7 @@ static void extern_rec(v)
       writecode8(CODE_INT8, n);
     } else if (n >= -(1 << 15) && n < (1 << 15)) {
       writecode16(CODE_INT16, n);
-#ifdef SIXTYFOUR
+#ifdef ARCH_SIXTYFOUR
     } else if (n < -(1L << 31) || n >= (1L << 31)) {
       writecode64(CODE_INT64, n);
 #endif
@@ -334,7 +334,7 @@ static long extern_value(v)
   /* Free the table of shared objects */
   stat_free((char *) extern_table);
   /* Write the sizes */
-#ifdef SIXTYFOUR
+#ifdef ARCH_SIXTYFOUR
   if (size_32 >= (1L << 32) || size_64 >= (1L << 32)) {
     /* The object is so big its size cannot be written in the header.
        Besides, some of the block sizes or string lengths or shared offsets
diff --git a/byterun/fix_code.c b/byterun/fix_code.c
index 20ef44ee6..7ce94095f 100644
--- a/byterun/fix_code.c
+++ b/byterun/fix_code.c
@@ -22,7 +22,7 @@
 
 /* This code is needed only if the processor is big endian */
 
-#ifdef BIG_ENDIAN
+#ifdef ARCH_BIG_ENDIAN
 
 void fixup_endianness(code, len)
      code_t code;
diff --git a/byterun/floats.c b/byterun/floats.c
index ea1999377..fbca35d77 100644
--- a/byterun/floats.c
+++ b/byterun/floats.c
@@ -21,7 +21,7 @@
 #include "misc.h"
 #include "stacks.h"
 
-#ifdef ALIGN_DOUBLE
+#ifdef ARCH_ALIGN_DOUBLE
 
 double Double_val(val)
      value val;
diff --git a/byterun/hash.c b/byterun/hash.c
index f240c16a1..18995ed16 100644
--- a/byterun/hash.c
+++ b/byterun/hash.c
@@ -71,7 +71,7 @@ static void hash_aux(obj)
       /* For doubles, we inspect their binary representation, LSB first.
          The results are consistent among all platforms with IEEE floats. */
       hash_univ_count--;
-#ifdef BIG_ENDIAN
+#ifdef ARCH_BIG_ENDIAN
       for (p = &Byte_u(obj, sizeof(double) - 1), i = sizeof(double);
            i > 0;
            p--, i--)
@@ -85,7 +85,7 @@ static void hash_aux(obj)
     case Double_array_tag:
       hash_univ_count--;
       for (j = 0; j < Bosize_val(obj); j += sizeof(double)) {
-#ifdef BIG_ENDIAN
+#ifdef ARCH_BIG_ENDIAN
       for (p = &Byte_u(obj, j + sizeof(double) - 1), i = sizeof(double);
            i > 0;
            p--, i--)
diff --git a/byterun/intern.c b/byterun/intern.c
index 5c063b3ae..895e1aade 100644
--- a/byterun/intern.c
+++ b/byterun/intern.c
@@ -53,7 +53,7 @@ static value intern_block;
    (Sign_extend(intern_src[-4]) << 24) + (intern_src[-3] << 16) + \
    (intern_src[-2] << 8) + intern_src[-1])
 
-#ifdef SIXTYFOUR
+#ifdef ARCH_SIXTYFOUR
 static long read64s()
 {
   long res;
@@ -136,7 +136,7 @@ static void intern_rec(dest)
         v = Val_long(read32s());
         break;
       case CODE_INT64:
-#ifdef SIXTYFOUR
+#ifdef ARCH_SIXTYFOUR
         v = Val_long(read64s());
         break;
 #else
@@ -255,7 +255,7 @@ value input_value(chan)         /* ML */
   }
   intern_src = intern_input;
   /* Allocate result */
-#ifdef SIXTYFOUR
+#ifdef ARCH_SIXTYFOUR
   whsize = size_64;
 #else
   whsize = size_32;
@@ -286,7 +286,7 @@ value input_value_from_string(str, ofs) /* ML */
   size_32 = read32u();
   size_64 = read32u();
   /* Allocate result */
-#ifdef SIXTYFOUR
+#ifdef ARCH_SIXTYFOUR
   whsize = size_64;
 #else
   whsize = size_32;
diff --git a/byterun/intext.h b/byterun/intext.h
index 9768b6db4..bf80ca20e 100644
--- a/byterun/intext.h
+++ b/byterun/intext.h
@@ -46,7 +46,7 @@
 #define CODE_DOUBLE_ARRAY32_BIG 0xF
 #define CODE_DOUBLE_ARRAY32_LITTLE 0x7
 
-#ifdef BIG_ENDIAN
+#ifdef ARCH_BIG_ENDIAN
 #define CODE_DOUBLE_NATIVE CODE_DOUBLE_BIG
 #define CODE_DOUBLE_ARRAY8_NATIVE CODE_DOUBLE_ARRAY8_BIG
 #define CODE_DOUBLE_ARRAY32_NATIVE CODE_DOUBLE_ARRAY32_BIG
diff --git a/byterun/io.c b/byterun/io.c
index 398bbe898..869b518d3 100644
--- a/byterun/io.c
+++ b/byterun/io.c
@@ -117,7 +117,9 @@ static int do_write(fd, p, n)
   retcode = ui_write(fd, p, n);
 #else
 again:
+  enter_blocking_section();
   retcode = write(fd, p, n);
+  leave_blocking_section();
   if (retcode == -1) {
     if (errno == EINTR) goto again;
     if (errno == EAGAIN || errno == EWOULDBLOCK) {
@@ -342,7 +344,7 @@ value input_int(channel)        /* ML */
 {
   long i;
   i = getword(channel);
-#ifdef SIXTYFOUR
+#ifdef ARCH_SIXTYFOUR
   i = (i << 32) >> 32;          /* Force sign extension */
 #endif
   return Val_long(i);
diff --git a/byterun/lexing.c b/byterun/lexing.c
index 86387a911..e5e6ab907 100644
--- a/byterun/lexing.c
+++ b/byterun/lexing.c
@@ -38,7 +38,7 @@ struct lexing_table {
   value lex_check;
 };
 
-#ifdef BIG_ENDIAN
+#ifdef ARCH_BIG_ENDIAN
 #define Short(tbl,n) \
   (*((unsigned char *)((tbl) + (n) * sizeof(short))) + \
           (*((schar *)((tbl) + (n) * sizeof(short) + 1)) << 8))
diff --git a/byterun/main.c b/byterun/main.c
index 45e759c4a..9337a9b46 100644
--- a/byterun/main.c
+++ b/byterun/main.c
@@ -227,7 +227,7 @@ int main(argc, argv)
     if (read(fd, (char *) start_code, code_size) != code_size)
       fatal_error("Fatal error: truncated bytecode file.\n");
 
-#ifdef BIG_ENDIAN
+#ifdef ARCH_BIG_ENDIAN
     fixup_endianness(start_code, code_size);
 #endif
 
diff --git a/byterun/major_gc.c b/byterun/major_gc.c
index 58376071b..1e683d8f6 100644
--- a/byterun/major_gc.c
+++ b/byterun/major_gc.c
@@ -24,7 +24,7 @@
 #ifdef __STDC__
 #include <limits.h>
 #else
-#ifdef SIXTYFOUR
+#ifdef ARCH_SIXTYFOUR
 #define LONG_MAX 0x7FFFFFFFFFFFFFFF
 #else
 #define LONG_MAX 0x7FFFFFFF
diff --git a/byterun/md5.c b/byterun/md5.c
index ed6abcdfb..bdaa9244b 100644
--- a/byterun/md5.c
+++ b/byterun/md5.c
@@ -83,7 +83,7 @@ value md5_chan(chan, len)       /* ML */
  * will fill a supplied 16-byte array with the digest.
  */
 
-#ifndef BIG_ENDIAN
+#ifndef ARCH_BIG_ENDIAN
 #define byteReverse(buf, len)	/* Nothing */
 #else
 void byteReverse(buf, longs)
diff --git a/byterun/meta.c b/byterun/meta.c
index a2df1572c..03a9ea0ef 100644
--- a/byterun/meta.c
+++ b/byterun/meta.c
@@ -35,7 +35,7 @@ value reify_bytecode(prog, len) /* ML */
      value prog, len;
 {
   value clos;
-#ifdef BIG_ENDIAN
+#ifdef ARCH_BIG_ENDIAN
   fixup_endianness((code_t) prog, (asize_t) Long_val(len));
 #endif
 #ifdef THREADED_CODE
diff --git a/byterun/mlvalues.h b/byterun/mlvalues.h
index f6a768da3..96b632d3c 100644
--- a/byterun/mlvalues.h
+++ b/byterun/mlvalues.h
@@ -104,7 +104,7 @@ bits  63    10 9     8 7   0
 #define Bp_hp(hp) ((char *) Val_hp (hp))
 
 #define Num_tags (1 << 8)
-#ifdef SIXTYFOUR
+#ifdef ARCH_SIXTYFOUR
 #define Max_wosize ((1L << 54) - 1)
 #else
 #define Max_wosize ((1 << 22) - 1)
@@ -132,7 +132,7 @@ bits  63    10 9     8 7   0
 #define Bhsize_hp(hp) (Bsize_wsize (Whsize_hp (hp)))
 #define Bhsize_hd(hd) (Bsize_wsize (Whsize_hd (hd)))
 
-#ifdef BIG_ENDIAN
+#ifdef ARCH_BIG_ENDIAN
 #define Tag_val(val) (((unsigned char *) (val)) [-1])
                                                  /* Also an l-value. */
 #define Tag_hp(hp) (((unsigned char *) (hp)) [sizeof(value)-1])
@@ -195,7 +195,7 @@ value callback3 P((value closure, value arg1, value arg2, value arg3));
 /* Floating-point numbers. */
 #define Double_tag 253
 #define Double_wosize ((sizeof(double) / sizeof(value)))
-#ifndef ALIGN_DOUBLE
+#ifndef ARCH_ALIGN_DOUBLE
 #define Double_val(v) (* (double *)(v))
 #define Store_double_val(v,d) (* (double *)(v) = (d))
 #else
diff --git a/byterun/parsing.c b/byterun/parsing.c
index 1c3de01c5..4fb163490 100644
--- a/byterun/parsing.c
+++ b/byterun/parsing.c
@@ -57,7 +57,7 @@ struct parser_env {       /* Mirrors parser_env in ../stdlib/parsing.ml */
   value errflag;
 };
 
-#ifdef BIG_ENDIAN
+#ifdef ARCH_BIG_ENDIAN
 #define Short(tbl,n) \
   (*((unsigned char *)((tbl) + (n) * sizeof(short))) + \
           (*((schar *)((tbl) + (n) * sizeof(short) + 1)) << 8))
diff --git a/byterun/reverse.h b/byterun/reverse.h
index 1b404f561..124b90888 100644
--- a/byterun/reverse.h
+++ b/byterun/reverse.h
@@ -56,7 +56,7 @@
   _p[4] = _a;                                                                 \
 }
 
-#ifdef SIXTYFOUR
+#ifdef ARCH_SIXTYFOUR
 #define Reverse_word Reverse_int64
 #else
 #define Reverse_word Reverse_int32
diff --git a/config/m-nt.h b/config/m-nt.h
index b3d88107f..6f50bf70b 100644
--- a/config/m-nt.h
+++ b/config/m-nt.h
@@ -13,7 +13,7 @@
 
 /* Machine configuration, Intel x86 processors, Windows NT */
 
-#undef SIXTYFOUR
-#undef BIG_ENDIAN
-#undef ALIGN_DOUBLE
+#undef ARCH_SIXTYFOUR
+#undef ARCH_BIG_ENDIAN
+#undef ARCH_ALIGN_DOUBLE
 
diff --git a/config/m-templ.h b/config/m-templ.h
index 98afad177..43307e01f 100644
--- a/config/m-templ.h
+++ b/config/m-templ.h
@@ -13,23 +13,24 @@
 
 /* Processor dependencies */
 
-#define SIXTYFOUR
+#define ARCH_SIXTYFOUR
 
-/* Define SIXTYFOUR if the processor has a natural word size of 64 bits.
+/* Define ARCH_SIXTYFOUR if the processor has a natural word size of 64 bits.
    That is, both sizeof(long) = 8 and sizeof(char *) = 8.
-   Otherwise, leave SIXTYFOUR undefined. This assumes
+   Otherwise, leave ARCH_SIXTYFOUR undefined. This assumes
    sizeof(long) = sizeof(char *) = 4. */
 
-#define BIG_ENDIAN
+#define ARCH_BIG_ENDIAN
 
-/* Define BIG_ENDIAN if the processor is big endian (the most significant
-   byte of an integer stored in memory comes first). Leave BIG_ENDIAN undefined
-   if the processor is little-endian (the least significant byte comes first).
+/* Define ARCH_BIG_ENDIAN if the processor is big endian (the most
+   significant byte of an integer stored in memory comes first).
+   Leave ARCH_BIG_ENDIAN undefined if the processor is little-endian
+   (the least significant byte comes first).
 */
 
-#define ALIGN_DOUBLE
+#define ARCH_ALIGN_DOUBLE
 
-/* Define ALIGN_DOUBLE if the processor requires doubles to be
-   doubleword-aligned. Leave ALIGN_DOUBLE undefined if the processor supports
-   word-aligned doubles. */
+/* Define ARCH_ALIGN_DOUBLE if the processor requires doubles to be
+   doubleword-aligned. Leave ARCH_ALIGN_DOUBLE undefined if the processor
+   supports word-aligned doubles. */
 
diff --git a/configure b/configure
index cc16d6bec..88e16ce24 100755
--- a/configure
+++ b/configure
@@ -99,9 +99,9 @@ echo "Checking the sizes of integers and pointers..."
 set `sh runtest sizes.c`
 case "$1,$2,$3" in
   4,4,4) echo "OK, this is a regular 32 bit architecture."
-         echo "#undef SIXTYFOUR" >> m.h;;
+         echo "#undef ARCH_SIXTYFOUR" >> m.h;;
   4,8,8) echo "Wow! A 64 bit architecture!"
-         echo "#define SIXTYFOUR" >> m.h;;
+         echo "#define ARCH_SIXTYFOUR" >> m.h;;
   8,*,*) echo "Wow! A 64 bit architecture!"
          echo "Unfortunately, Objective Caml does not handle the case"
          echo "sizeof(int) = 8."
@@ -125,15 +125,15 @@ esac
 sh runtest endian.c
 case $? in
   0) echo "This is a big-endian architecture."
-     echo "#define BIG_ENDIAN" >> m.h;;
+     echo "#define ARCH_BIG_ENDIAN" >> m.h;;
   1) echo "This is a little-endian architecture."
-     echo "#undef BIG_ENDIAN" >> m.h;;
+     echo "#undef ARCH_BIG_ENDIAN" >> m.h;;
   2) echo "This architecture seems to be neither big endian nor little endian."
      echo "Objective Caml won't run on this architecture."
      exit 2;;
   *) echo "Something went wrong during endianness determination."
      echo "You'll have to figure out endianness yourself"
-     echo "(option BIG_ENDIAN in m.h).";;
+     echo "(option ARCH_BIG_ENDIAN in m.h).";;
 esac
 
 # Determine alignment constraints
@@ -141,14 +141,14 @@ esac
 sh runtest dblalign.c
 case $? in
   0) echo "Doubles can be word-aligned."
-     echo "#undef ALIGN_DOUBLE" >> m.h;;
+     echo "#undef ARCH_ALIGN_DOUBLE" >> m.h;;
   1) echo "Doubles must be doubleword-aligned."
-     echo "#define ALIGN_DOUBLE" >> m.h;;
+     echo "#define ARCH_ALIGN_DOUBLE" >> m.h;;
   *) echo "Something went wrong during alignment determination for doubles."
      echo "I'm going to assume this architecture has alignment constraints over doubles."
      echo "That's a safe bet: Objective Caml will work even if"
      echo "this architecture has actually no alignment constraints."
-     echo "#define ALIGN_DOUBLE" >> m.h;;
+     echo "#define ARCH_ALIGN_DOUBLE" >> m.h;;
 esac
 
 # Configure the bytecode compiler