revisiting hashtables

dtool/src/dtoolbase/Sources.pp

@@ -6,6 +6,7 @@
   #define COMBINED_SOURCES $[TARGET]_composite1.cxx $[TARGET]_composite2.cxx
   
   #define SOURCES \
+    addHash.I addHash.h \
     atomicAdjust.h \
     atomicAdjustDummyImpl.h atomicAdjustDummyImpl.I \
     atomicAdjustI386Impl.h atomicAdjustI386Impl.I \
@@ -31,9 +32,10 @@
     stl_compares.I stl_compares.h \
     pallocator.T pallocator.h \
     pdeque.h plist.h pmap.h pset.h pvector.h \
-    dlmalloc.c
+    dlmalloc.c lookup3.h lookup3.c
 
  #define INCLUDED_SOURCES  \
+    addHash.cxx \
     atomicAdjustDummyImpl.cxx \
     atomicAdjustI386Impl.cxx \
     atomicAdjustNsprImpl.cxx \
@@ -48,6 +50,7 @@
     mutexSpinlockImpl.cxx
 
   #define INSTALL_HEADERS \
+    addHash.I addHash.h \
     atomicAdjust.h \
     atomicAdjustDummyImpl.h atomicAdjustDummyImpl.I \
     atomicAdjustI386Impl.h atomicAdjustI386Impl.I \
@@ -71,6 +74,7 @@
     selectThreadImpl.h \
     stl_compares.I stl_compares.h \
     pallocator.T pallocator.h \
-    pdeque.h plist.h pmap.h pset.h pvector.h
+    pdeque.h plist.h pmap.h pset.h pvector.h \
+    lookup3.h
 
 #end lib_target

dtool/src/dtoolbase/addHash.I

@@ -0,0 +1,49 @@
+// Filename: add_hash.I
+// Created by:  drose (01Sep06)
+//
+////////////////////////////////////////////////////////////////////
+//
+// PANDA 3D SOFTWARE
+// Copyright (c) 2001 - 2004, Disney Enterprises, Inc.  All rights reserved
+//
+// All use of this software is subject to the terms of the Panda 3d
+// Software license.  You should have received a copy of this license
+// along with this source code; you will also find a current copy of
+// the license at http://etc.cmu.edu/panda3d/docs/license/ .
+//
+// To contact the maintainers of this program write to
+// [email protected] .
+//
+////////////////////////////////////////////////////////////////////
+
+
+////////////////////////////////////////////////////////////////////
+//     Function: AddHash::add_hash
+//       Access: Public, Static
+//  Description: Adds a linear sequence of uint32 words to the hash.
+////////////////////////////////////////////////////////////////////
+INLINE size_t AddHash::
+add_hash(size_t start, const PN_uint32 *words, size_t num_words) {
+  return (size_t)hashword(words, num_words, (PN_uint32)start);
+}
+
+////////////////////////////////////////////////////////////////////
+//     Function: AddHash::add_hash
+//       Access: Public, Static
+//  Description: Adds a linear sequence of float32 words to the hash.
+////////////////////////////////////////////////////////////////////
+INLINE size_t AddHash::
+add_hash(size_t start, const PN_float32 *floats, size_t num_floats) {
+  return add_hash(start, (const PN_uint32 *)floats, num_floats);
+}
+
+////////////////////////////////////////////////////////////////////
+//     Function: AddHash::add_hash
+//       Access: Public, Static
+//  Description: Adds a linear sequence of float64 words to the hash.
+////////////////////////////////////////////////////////////////////
+INLINE size_t AddHash::
+add_hash(size_t start, const PN_float64 *floats, size_t num_floats) {
+  return add_hash(start, (const PN_uint32 *)floats, num_floats * 2);
+}
+

dtool/src/dtoolbase/addHash.cxx

@@ -0,0 +1,61 @@
+// Filename: addHash.cxx
+// Created by:  drose (01Sep06)
+//
+////////////////////////////////////////////////////////////////////
+//
+// PANDA 3D SOFTWARE
+// Copyright (c) 2001 - 2004, Disney Enterprises, Inc.  All rights reserved
+//
+// All use of this software is subject to the terms of the Panda 3d
+// Software license.  You should have received a copy of this license
+// along with this source code; you will also find a current copy of
+// the license at http://etc.cmu.edu/panda3d/docs/license/ .
+//
+// To contact the maintainers of this program write to
+// [email protected] .
+//
+////////////////////////////////////////////////////////////////////
+
+#include "addHash.h"
+
+////////////////////////////////////////////////////////////////////
+//     Function: AddHash::add_hash
+//       Access: Public, Static
+//  Description: Adds a linear sequence of bytes to the hash.
+////////////////////////////////////////////////////////////////////
+size_t AddHash::
+add_hash(size_t start, const PN_uint8 *bytes, size_t num_bytes) {
+  size_t num_words = num_bytes >> 2;
+  size_t remaining_bytes = num_bytes - (num_words << 2);
+  size_t hash = (size_t)hashword((const PN_uint32 *)bytes, num_words, (PN_uint32)start);
+
+  switch (remaining_bytes) {
+  case 3:
+    {
+      PN_uint32 remaining;
+      remaining = (bytes[num_bytes - 3] << 16) | (bytes[num_bytes - 2] << 8) | (bytes[num_bytes - 1]);
+      hash = (size_t)hashword(&remaining, 1, (PN_uint32)hash);
+    }
+    break;
+
+  case 2:
+    {
+      PN_uint32 remaining;
+      remaining = (bytes[num_bytes - 2] << 8) | (bytes[num_bytes - 1]);
+      hash = (size_t)hashword(&remaining, 1, (PN_uint32)hash);
+    }
+    break;
+
+  case 1:
+    {
+      PN_uint32 remaining;
+      remaining = (bytes[num_bytes - 1]);
+      hash = (size_t)hashword(&remaining, 1, (PN_uint32)hash);
+    }
+    break;
+
+  default:
+    break;
+  }
+  return hash;
+}

dtool/src/dtoolbase/addHash.h

@@ -0,0 +1,43 @@
+// Filename: addHash.h
+// Created by:  drose (01Sep06)
+//
+////////////////////////////////////////////////////////////////////
+//
+// PANDA 3D SOFTWARE
+// Copyright (c) 2001 - 2004, Disney Enterprises, Inc.  All rights reserved
+//
+// All use of this software is subject to the terms of the Panda 3d
+// Software license.  You should have received a copy of this license
+// along with this source code; you will also find a current copy of
+// the license at http://etc.cmu.edu/panda3d/docs/license/ .
+//
+// To contact the maintainers of this program write to
+// [email protected] .
+//
+////////////////////////////////////////////////////////////////////
+
+#ifndef ADDHASH_H
+#define ADDHASH_H
+
+#include "dtoolbase.h"
+#include "numeric_types.h"
+#include "lookup3.h"
+
+////////////////////////////////////////////////////////////////////
+//       Class : AddHash
+// Description : This class is used just as a namespace scope to
+//               collect together a handful of static functions, which
+//               are used to wrap calls to Bob Jenkins' public-domain
+//               hash generation function (defined in lookup3.c).
+////////////////////////////////////////////////////////////////////
+class EXPCL_DTOOL AddHash {
+public:
+  INLINE static size_t add_hash(size_t start, const PN_uint32 *words, size_t num_words);
+  static size_t add_hash(size_t start, const PN_uint8 *bytes, size_t num_bytes);
+  INLINE static size_t add_hash(size_t start, const PN_float32 *floats, size_t num_floats);
+  INLINE static size_t add_hash(size_t start, const PN_float64 *floats, size_t num_floats);
+};
+
+#include "addHash.I"
+
+#endif

dtool/src/dtoolbase/dtoolbase_composite1.cxx

@@ -1,3 +1,4 @@
+#include "addHash.cxx"
 #include "atomicAdjustDummyImpl.cxx"
 #include "atomicAdjustI386Impl.cxx"
 #include "atomicAdjustNsprImpl.cxx"

dtool/src/dtoolbase/lookup3.c

@@ -0,0 +1,912 @@
+/* See http://www.burtleburtle.net/bob/hash/ */
+
+/*
+-------------------------------------------------------------------------------
+lookup3.c, by Bob Jenkins, May 2006, Public Domain.
+
+These are functions for producing 32-bit hashes for hash table lookup.
+hashword(), hashlittle(), hashlittle2(), hashbig(), mix(), and final() 
+are externally useful functions.  Routines to test the hash are included 
+if SELF_TEST is defined.  You can use this free for any purpose.  It's in
+the public domain.  It has no warranty.
+
+You probably want to use hashlittle().  hashlittle() and hashbig()
+hash byte arrays.  hashlittle() is is faster than hashbig() on
+little-endian machines.  Intel and AMD are little-endian machines.
+On second thought, you probably want hashlittle2(), which is identical to
+hashlittle() except it returns two 32-bit hashes for the price of one.  
+You could implement hashbig2() if you wanted but I haven't bothered here.
+
+If you want to find a hash of, say, exactly 7 integers, do
+  a = i1;  b = i2;  c = i3;
+  mix(a,b,c);
+  a += i4; b += i5; c += i6;
+  mix(a,b,c);
+  a += i7;
+  final(a,b,c);
+then use c as the hash value.  If you have a variable length array of
+4-byte integers to hash, use hashword().  If you have a byte array (like
+a character string), use hashlittle().  If you have several byte arrays, or
+a mix of things, see the comments above hashlittle().  
+
+Why is this so big?  I read 12 bytes at a time into 3 4-byte integers, 
+then mix those integers.  This is fast (you can do a lot more thorough
+mixing with 12*3 instructions on 3 integers than you can with 3 instructions
+on 1 byte), but shoehorning those bytes into integers efficiently is messy.
+-------------------------------------------------------------------------------
+*/
+/*#define SELF_TEST 1*/
+
+#include "lookup3.h"
+
+#include <stdio.h>
+#include <stddef.h>
+#include <stdlib.h>
+#include <time.h>
+
+#ifdef WORDS_BIGENDIAN
+# define HASH_LITTLE_ENDIAN 0
+# define HASH_BIG_ENDIAN 1
+#else
+# define HASH_LITTLE_ENDIAN 1
+# define HASH_BIG_ENDIAN 0
+#endif
+
+#define hashsize(n) ((uint32_t)1<<(n))
+#define hashmask(n) (hashsize(n)-1)
+#define rot(x,k) (((x)<<(k)) ^ ((x)>>(32-(k))))
+
+/*
+-------------------------------------------------------------------------------
+mix -- mix 3 32-bit values reversibly.
+
+This is reversible, so any information in (a,b,c) before mix() is
+still in (a,b,c) after mix().
+
+If four pairs of (a,b,c) inputs are run through mix(), or through
+mix() in reverse, there are at least 32 bits of the output that
+are sometimes the same for one pair and different for another pair.
+This was tested for:
+* pairs that differed by one bit, by two bits, in any combination
+  of top bits of (a,b,c), or in any combination of bottom bits of
+  (a,b,c).
+* "differ" is defined as +, -, ^, or ~^.  For + and -, I transformed
+  the output delta to a Gray code (a^(a>>1)) so a string of 1's (as
+  is commonly produced by subtraction) look like a single 1-bit
+  difference.
+* the base values were pseudorandom, all zero but one bit set, or 
+  all zero plus a counter that starts at zero.
+
+Some k values for my "a-=c; a^=rot(c,k); c+=b;" arrangement that
+satisfy this are
+    4  6  8 16 19  4
+    9 15  3 18 27 15
+   14  9  3  7 17  3
+Well, "9 15 3 18 27 15" didn't quite get 32 bits diffing
+for "differ" defined as + with a one-bit base and a two-bit delta.  I
+used http://burtleburtle.net/bob/hash/avalanche.html to choose 
+the operations, constants, and arrangements of the variables.
+
+This does not achieve avalanche.  There are input bits of (a,b,c)
+that fail to affect some output bits of (a,b,c), especially of a.  The
+most thoroughly mixed value is c, but it doesn't really even achieve
+avalanche in c.
+
+This allows some parallelism.  Read-after-writes are good at doubling
+the number of bits affected, so the goal of mixing pulls in the opposite
+direction as the goal of parallelism.  I did what I could.  Rotates
+seem to cost as much as shifts on every machine I could lay my hands
+on, and rotates are much kinder to the top and bottom bits, so I used
+rotates.
+-------------------------------------------------------------------------------
+*/
+#define mix(a,b,c) \
+{ \
+  a -= c;  a ^= rot(c, 4);  c += b; \
+  b -= a;  b ^= rot(a, 6);  a += c; \
+  c -= b;  c ^= rot(b, 8);  b += a; \
+  a -= c;  a ^= rot(c,16);  c += b; \
+  b -= a;  b ^= rot(a,19);  a += c; \
+  c -= b;  c ^= rot(b, 4);  b += a; \
+}
+
+/*
+-------------------------------------------------------------------------------
+final -- final mixing of 3 32-bit values (a,b,c) into c
+
+Pairs of (a,b,c) values differing in only a few bits will usually
+produce values of c that look totally different.  This was tested for
+* pairs that differed by one bit, by two bits, in any combination
+  of top bits of (a,b,c), or in any combination of bottom bits of
+  (a,b,c).
+* "differ" is defined as +, -, ^, or ~^.  For + and -, I transformed
+  the output delta to a Gray code (a^(a>>1)) so a string of 1's (as
+  is commonly produced by subtraction) look like a single 1-bit
+  difference.
+* the base values were pseudorandom, all zero but one bit set, or 
+  all zero plus a counter that starts at zero.
+
+These constants passed:
+ 14 11 25 16 4 14 24
+ 12 14 25 16 4 14 24
+and these came close:
+  4  8 15 26 3 22 24
+ 10  8 15 26 3 22 24
+ 11  8 15 26 3 22 24
+-------------------------------------------------------------------------------
+*/
+#define final(a,b,c) \
+{ \
+  c ^= b; c -= rot(b,14); \
+  a ^= c; a -= rot(c,11); \
+  b ^= a; b -= rot(a,25); \
+  c ^= b; c -= rot(b,16); \
+  a ^= c; a -= rot(c,4);  \
+  b ^= a; b -= rot(a,14); \
+  c ^= b; c -= rot(b,24); \
+}
+
+/*
+--------------------------------------------------------------------
+ This works on all machines.  To be useful, it requires
+ -- that the key be an array of PN_uint32's, and
+ -- that the length be the number of PN_uint32's in the key
+
+ The function hashword() is identical to hashlittle() on little-endian
+ machines, and identical to hashbig() on big-endian machines,
+ except that the length has to be measured in PN_uint32s rather than in
+ bytes.  hashlittle() is more complicated than hashword() only because
+ hashlittle() has to dance around fitting the key bytes into registers.
+--------------------------------------------------------------------
+*/
+PN_uint32 hashword(
+const PN_uint32 *k,                   /* the key, an array of PN_uint32 values */
+size_t          length,               /* the length of the key, in PN_uint32s */
+PN_uint32        initval)         /* the previous hash, or an arbitrary value */
+{
+  PN_uint32 a,b,c;
+
+  /* Set up the internal state */
+  a = b = c = 0xdeadbeef + (((PN_uint32)length)<<2) + initval;
+
+  /*------------------------------------------------- handle most of the key */
+  while (length > 3)
+  {
+    a += k[0];
+    b += k[1];
+    c += k[2];
+    mix(a,b,c);
+    length -= 3;
+    k += 3;
+  }
+
+  /*------------------------------------------- handle the last 3 PN_uint32's */
+  switch(length)                     /* all the case statements fall through */
+  { 
+  case 3 : c+=k[2];
+  case 2 : b+=k[1];
+  case 1 : a+=k[0];
+    final(a,b,c);
+  case 0:     /* case 0: nothing left to add */
+    break;
+  }
+  /*------------------------------------------------------ report the result */
+  return c;
+}
+
+
+/*
+-------------------------------------------------------------------------------
+hashlittle() -- hash a variable-length key into a 32-bit value
+  k       : the key (the unaligned variable-length array of bytes)
+  length  : the length of the key, counting by bytes
+  initval : can be any 4-byte value
+Returns a 32-bit value.  Every bit of the key affects every bit of
+the return value.  Two keys differing by one or two bits will have
+totally different hash values.
+
+The best hash table sizes are powers of 2.  There is no need to do
+mod a prime (mod is sooo slow!).  If you need less than 32 bits,
+use a bitmask.  For example, if you need only 10 bits, do
+  h = (h & hashmask(10));
+In which case, the hash table should have hashsize(10) elements.
+
+If you are hashing n strings (PN_uint8 **)k, do it like this:
+  for (i=0, h=0; i<n; ++i) h = hashlittle( k[i], len[i], h);
+
+By Bob Jenkins, 2006.  [email protected].  You may use this
+code any way you wish, private, educational, or commercial.  It's free.
+
+Use for hash table lookup, or anything where one collision in 2^^32 is
+acceptable.  Do NOT use for cryptographic purposes.
+-------------------------------------------------------------------------------
+*/
+
+PN_uint32 hashlittle( const void *key, size_t length, PN_uint32 initval)
+{
+  PN_uint32 a,b,c;                                          /* internal state */
+  union { const void *ptr; size_t i; } u;     /* needed for Mac Powerbook G4 */
+
+  /* Set up the internal state */
+  a = b = c = 0xdeadbeef + ((PN_uint32)length) + initval;
+
+  u.ptr = key;
+  if (HASH_LITTLE_ENDIAN && ((u.i & 0x3) == 0)) {
+    const PN_uint32 *k = key;                           /* read 32-bit chunks */
+#ifdef VALGRIND
+    const PN_uint8  *k8;
+#endif
+
+    /*------ all but last block: aligned reads and affect 32 bits of (a,b,c) */
+    while (length > 12)
+    {
+      a += k[0];
+      b += k[1];
+      c += k[2];
+      mix(a,b,c);
+      length -= 12;
+      k += 3;
+    }
+
+    /*----------------------------- handle the last (probably partial) block */
+    /* 
+     * "k[2]&0xffffff" actually reads beyond the end of the string, but
+     * then masks off the part it's not allowed to read.  Because the
+     * string is aligned, the masked-off tail is in the same word as the
+     * rest of the string.  Every machine with memory protection I've seen
+     * does it on word boundaries, so is OK with this.  But VALGRIND will
+     * still catch it and complain.  The masking trick does make the hash
+     * noticably faster for short strings (like English words).
+     */
+#ifndef VALGRIND
+
+    switch(length)
+    {
+    case 12: c+=k[2]; b+=k[1]; a+=k[0]; break;
+    case 11: c+=k[2]&0xffffff; b+=k[1]; a+=k[0]; break;
+    case 10: c+=k[2]&0xffff; b+=k[1]; a+=k[0]; break;
+    case 9 : c+=k[2]&0xff; b+=k[1]; a+=k[0]; break;
+    case 8 : b+=k[1]; a+=k[0]; break;
+    case 7 : b+=k[1]&0xffffff; a+=k[0]; break;
+    case 6 : b+=k[1]&0xffff; a+=k[0]; break;
+    case 5 : b+=k[1]&0xff; a+=k[0]; break;
+    case 4 : a+=k[0]; break;
+    case 3 : a+=k[0]&0xffffff; break;
+    case 2 : a+=k[0]&0xffff; break;
+    case 1 : a+=k[0]&0xff; break;
+    case 0 : return c;              /* zero length strings require no mixing */
+    }
+
+#else /* make valgrind happy */
+
+    k8 = (const PN_uint8 *)k;
+    switch(length)
+    {
+    case 12: c+=k[2]; b+=k[1]; a+=k[0]; break;
+    case 11: c+=((PN_uint32)k8[10])<<16;  /* fall through */
+    case 10: c+=((PN_uint32)k8[9])<<8;    /* fall through */
+    case 9 : c+=k8[8];                   /* fall through */
+    case 8 : b+=k[1]; a+=k[0]; break;
+    case 7 : b+=((PN_uint32)k8[6])<<16;   /* fall through */
+    case 6 : b+=((PN_uint32)k8[5])<<8;    /* fall through */
+    case 5 : b+=k8[4];                   /* fall through */
+    case 4 : a+=k[0]; break;
+    case 3 : a+=((PN_uint32)k8[2])<<16;   /* fall through */
+    case 2 : a+=((PN_uint32)k8[1])<<8;    /* fall through */
+    case 1 : a+=k8[0]; break;
+    case 0 : return c;
+    }
+
+#endif /* !valgrind */
+
+  } else if (HASH_LITTLE_ENDIAN && ((u.i & 0x1) == 0)) {
+    const PN_uint16 *k = key;                           /* read 16-bit chunks */
+    const PN_uint8  *k8;
+
+    /*--------------- all but last block: aligned reads and different mixing */
+    while (length > 12)
+    {
+      a += k[0] + (((PN_uint32)k[1])<<16);
+      b += k[2] + (((PN_uint32)k[3])<<16);
+      c += k[4] + (((PN_uint32)k[5])<<16);
+      mix(a,b,c);
+      length -= 12;
+      k += 6;
+    }
+
+    /*----------------------------- handle the last (probably partial) block */
+    k8 = (const PN_uint8 *)k;
+    switch(length)
+    {
+    case 12: c+=k[4]+(((PN_uint32)k[5])<<16);
+             b+=k[2]+(((PN_uint32)k[3])<<16);
+             a+=k[0]+(((PN_uint32)k[1])<<16);
+             break;
+    case 11: c+=((PN_uint32)k8[10])<<16;     /* fall through */
+    case 10: c+=k[4];
+             b+=k[2]+(((PN_uint32)k[3])<<16);
+             a+=k[0]+(((PN_uint32)k[1])<<16);
+             break;
+    case 9 : c+=k8[8];                      /* fall through */
+    case 8 : b+=k[2]+(((PN_uint32)k[3])<<16);
+             a+=k[0]+(((PN_uint32)k[1])<<16);
+             break;
+    case 7 : b+=((PN_uint32)k8[6])<<16;      /* fall through */
+    case 6 : b+=k[2];
+             a+=k[0]+(((PN_uint32)k[1])<<16);
+             break;
+    case 5 : b+=k8[4];                      /* fall through */
+    case 4 : a+=k[0]+(((PN_uint32)k[1])<<16);
+             break;
+    case 3 : a+=((PN_uint32)k8[2])<<16;      /* fall through */
+    case 2 : a+=k[0];
+             break;
+    case 1 : a+=k8[0];
+             break;
+    case 0 : return c;                     /* zero length requires no mixing */
+    }
+
+  } else {                        /* need to read the key one byte at a time */
+    const PN_uint8 *k = key;
+
+    /*--------------- all but the last block: affect some 32 bits of (a,b,c) */
+    while (length > 12)
+    {
+      a += k[0];
+      a += ((PN_uint32)k[1])<<8;
+      a += ((PN_uint32)k[2])<<16;
+      a += ((PN_uint32)k[3])<<24;
+      b += k[4];
+      b += ((PN_uint32)k[5])<<8;
+      b += ((PN_uint32)k[6])<<16;
+      b += ((PN_uint32)k[7])<<24;
+      c += k[8];
+      c += ((PN_uint32)k[9])<<8;
+      c += ((PN_uint32)k[10])<<16;
+      c += ((PN_uint32)k[11])<<24;
+      mix(a,b,c);
+      length -= 12;
+      k += 12;
+    }
+
+    /*-------------------------------- last block: affect all 32 bits of (c) */
+    switch(length)                   /* all the case statements fall through */
+    {
+    case 12: c+=((PN_uint32)k[11])<<24;
+    case 11: c+=((PN_uint32)k[10])<<16;
+    case 10: c+=((PN_uint32)k[9])<<8;
+    case 9 : c+=k[8];
+    case 8 : b+=((PN_uint32)k[7])<<24;
+    case 7 : b+=((PN_uint32)k[6])<<16;
+    case 6 : b+=((PN_uint32)k[5])<<8;
+    case 5 : b+=k[4];
+    case 4 : a+=((PN_uint32)k[3])<<24;
+    case 3 : a+=((PN_uint32)k[2])<<16;
+    case 2 : a+=((PN_uint32)k[1])<<8;
+    case 1 : a+=k[0];
+             break;
+    case 0 : return c;
+    }
+  }
+
+  final(a,b,c);
+  return c;
+}
+
+
+/*
+ * hashlittle2: return 2 32-bit hash values
+ *
+ * This is identical to hashlittle(), except it returns two 32-bit hash
+ * values instead of just one.  This is good enough for hash table
+ * lookup with 2^^64 buckets, or if you want a second hash if you're not
+ * happy with the first, or if you want a probably-unique 64-bit ID for
+ * the key.  *pc is better mixed than *pb, so use *pc first.  If you want
+ * a 64-bit value do something like "*pc + (((uint64_t)*pb)<<32)".
+ */
+void hashlittle2( 
+  const void *key,       /* the key to hash */
+  size_t      length,    /* length of the key */
+  PN_uint32   *pc,        /* IN: primary initval, OUT: primary hash */
+  PN_uint32   *pb)        /* IN: secondary initval, OUT: secondary hash */
+{
+  PN_uint32 a,b,c;                                          /* internal state */
+  union { const void *ptr; size_t i; } u;     /* needed for Mac Powerbook G4 */
+
+  /* Set up the internal state */
+  a = b = c = 0xdeadbeef + ((PN_uint32)length) + *pc;
+  c += *pb;
+
+  u.ptr = key;
+  if (HASH_LITTLE_ENDIAN && ((u.i & 0x3) == 0)) {
+    const PN_uint32 *k = key;                           /* read 32-bit chunks */
+#ifdef VALGRIND
+    const PN_uint8  *k8;
+#endif
+
+    /*------ all but last block: aligned reads and affect 32 bits of (a,b,c) */
+    while (length > 12)
+    {
+      a += k[0];
+      b += k[1];
+      c += k[2];
+      mix(a,b,c);
+      length -= 12;
+      k += 3;
+    }
+
+    /*----------------------------- handle the last (probably partial) block */
+    /* 
+     * "k[2]&0xffffff" actually reads beyond the end of the string, but
+     * then masks off the part it's not allowed to read.  Because the
+     * string is aligned, the masked-off tail is in the same word as the
+     * rest of the string.  Every machine with memory protection I've seen
+     * does it on word boundaries, so is OK with this.  But VALGRIND will
+     * still catch it and complain.  The masking trick does make the hash
+     * noticably faster for short strings (like English words).
+     */
+#ifndef VALGRIND
+
+    switch(length)
+    {
+    case 12: c+=k[2]; b+=k[1]; a+=k[0]; break;
+    case 11: c+=k[2]&0xffffff; b+=k[1]; a+=k[0]; break;
+    case 10: c+=k[2]&0xffff; b+=k[1]; a+=k[0]; break;
+    case 9 : c+=k[2]&0xff; b+=k[1]; a+=k[0]; break;
+    case 8 : b+=k[1]; a+=k[0]; break;
+    case 7 : b+=k[1]&0xffffff; a+=k[0]; break;
+    case 6 : b+=k[1]&0xffff; a+=k[0]; break;
+    case 5 : b+=k[1]&0xff; a+=k[0]; break;
+    case 4 : a+=k[0]; break;
+    case 3 : a+=k[0]&0xffffff; break;
+    case 2 : a+=k[0]&0xffff; break;
+    case 1 : a+=k[0]&0xff; break;
+    case 0 : *pc=c; *pb=b; return;  /* zero length strings require no mixing */
+    }
+
+#else /* make valgrind happy */
+
+    k8 = (const PN_uint8 *)k;
+    switch(length)
+    {
+    case 12: c+=k[2]; b+=k[1]; a+=k[0]; break;
+    case 11: c+=((PN_uint32)k8[10])<<16;  /* fall through */
+    case 10: c+=((PN_uint32)k8[9])<<8;    /* fall through */
+    case 9 : c+=k8[8];                   /* fall through */
+    case 8 : b+=k[1]; a+=k[0]; break;
+    case 7 : b+=((PN_uint32)k8[6])<<16;   /* fall through */
+    case 6 : b+=((PN_uint32)k8[5])<<8;    /* fall through */
+    case 5 : b+=k8[4];                   /* fall through */
+    case 4 : a+=k[0]; break;
+    case 3 : a+=((PN_uint32)k8[2])<<16;   /* fall through */
+    case 2 : a+=((PN_uint32)k8[1])<<8;    /* fall through */
+    case 1 : a+=k8[0]; break;
+    case 0 : *pc=c; *pb=b; return;  /* zero length strings require no mixing */
+    }
+
+#endif /* !valgrind */
+
+  } else if (HASH_LITTLE_ENDIAN && ((u.i & 0x1) == 0)) {
+    const PN_uint16 *k = key;                           /* read 16-bit chunks */
+    const PN_uint8  *k8;
+
+    /*--------------- all but last block: aligned reads and different mixing */
+    while (length > 12)
+    {
+      a += k[0] + (((PN_uint32)k[1])<<16);
+      b += k[2] + (((PN_uint32)k[3])<<16);
+      c += k[4] + (((PN_uint32)k[5])<<16);
+      mix(a,b,c);
+      length -= 12;
+      k += 6;
+    }
+
+    /*----------------------------- handle the last (probably partial) block */
+    k8 = (const PN_uint8 *)k;
+    switch(length)
+    {
+    case 12: c+=k[4]+(((PN_uint32)k[5])<<16);
+             b+=k[2]+(((PN_uint32)k[3])<<16);
+             a+=k[0]+(((PN_uint32)k[1])<<16);
+             break;
+    case 11: c+=((PN_uint32)k8[10])<<16;     /* fall through */
+    case 10: c+=k[4];
+             b+=k[2]+(((PN_uint32)k[3])<<16);
+             a+=k[0]+(((PN_uint32)k[1])<<16);
+             break;
+    case 9 : c+=k8[8];                      /* fall through */
+    case 8 : b+=k[2]+(((PN_uint32)k[3])<<16);
+             a+=k[0]+(((PN_uint32)k[1])<<16);
+             break;
+    case 7 : b+=((PN_uint32)k8[6])<<16;      /* fall through */
+    case 6 : b+=k[2];
+             a+=k[0]+(((PN_uint32)k[1])<<16);
+             break;
+    case 5 : b+=k8[4];                      /* fall through */
+    case 4 : a+=k[0]+(((PN_uint32)k[1])<<16);
+             break;
+    case 3 : a+=((PN_uint32)k8[2])<<16;      /* fall through */
+    case 2 : a+=k[0];
+             break;
+    case 1 : a+=k8[0];
+             break;
+    case 0 : *pc=c; *pb=b; return;  /* zero length strings require no mixing */
+    }
+
+  } else {                        /* need to read the key one byte at a time */
+    const PN_uint8 *k = key;
+
+    /*--------------- all but the last block: affect some 32 bits of (a,b,c) */
+    while (length > 12)
+    {
+      a += k[0];
+      a += ((PN_uint32)k[1])<<8;
+      a += ((PN_uint32)k[2])<<16;
+      a += ((PN_uint32)k[3])<<24;
+      b += k[4];
+      b += ((PN_uint32)k[5])<<8;
+      b += ((PN_uint32)k[6])<<16;
+      b += ((PN_uint32)k[7])<<24;
+      c += k[8];
+      c += ((PN_uint32)k[9])<<8;
+      c += ((PN_uint32)k[10])<<16;
+      c += ((PN_uint32)k[11])<<24;
+      mix(a,b,c);
+      length -= 12;
+      k += 12;
+    }
+
+    /*-------------------------------- last block: affect all 32 bits of (c) */
+    switch(length)                   /* all the case statements fall through */
+    {
+    case 12: c+=((PN_uint32)k[11])<<24;
+    case 11: c+=((PN_uint32)k[10])<<16;
+    case 10: c+=((PN_uint32)k[9])<<8;
+    case 9 : c+=k[8];
+    case 8 : b+=((PN_uint32)k[7])<<24;
+    case 7 : b+=((PN_uint32)k[6])<<16;
+    case 6 : b+=((PN_uint32)k[5])<<8;
+    case 5 : b+=k[4];
+    case 4 : a+=((PN_uint32)k[3])<<24;
+    case 3 : a+=((PN_uint32)k[2])<<16;
+    case 2 : a+=((PN_uint32)k[1])<<8;
+    case 1 : a+=k[0];
+             break;
+    case 0 : *pc=c; *pb=b; return;  /* zero length strings require no mixing */
+    }
+  }
+
+  final(a,b,c);
+  *pc=c; *pb=b; return;             /* zero length strings require no mixing */
+}
+
+
+
+/*
+ * hashbig():
+ * This is the same as hashword() on big-endian machines.  It is different
+ * from hashlittle() on all machines.  hashbig() takes advantage of
+ * big-endian byte ordering. 
+ */
+PN_uint32 hashbig( const void *key, size_t length, PN_uint32 initval)
+{
+  PN_uint32 a,b,c;
+  union { const void *ptr; size_t i; } u; /* to cast key to (size_t) happily */
+
+  /* Set up the internal state */
+  a = b = c = 0xdeadbeef + ((PN_uint32)length) + initval;
+
+  u.ptr = key;
+  if (HASH_BIG_ENDIAN && ((u.i & 0x3) == 0)) {
+    const PN_uint32 *k = key;                           /* read 32-bit chunks */
+#ifdef VALGRIND
+    const PN_uint8  *k8;
+#endif
+
+    /*------ all but last block: aligned reads and affect 32 bits of (a,b,c) */
+    while (length > 12)
+    {
+      a += k[0];
+      b += k[1];
+      c += k[2];
+      mix(a,b,c);
+      length -= 12;
+      k += 3;
+    }
+
+    /*----------------------------- handle the last (probably partial) block */
+    /* 
+     * "k[2]<<8" actually reads beyond the end of the string, but
+     * then shifts out the part it's not allowed to read.  Because the
+     * string is aligned, the illegal read is in the same word as the
+     * rest of the string.  Every machine with memory protection I've seen
+     * does it on word boundaries, so is OK with this.  But VALGRIND will
+     * still catch it and complain.  The masking trick does make the hash
+     * noticably faster for short strings (like English words).
+     */
+#ifndef VALGRIND
+
+    switch(length)
+    {
+    case 12: c+=k[2]; b+=k[1]; a+=k[0]; break;
+    case 11: c+=k[2]<<8; b+=k[1]; a+=k[0]; break;
+    case 10: c+=k[2]<<16; b+=k[1]; a+=k[0]; break;
+    case 9 : c+=k[2]<<24; b+=k[1]; a+=k[0]; break;
+    case 8 : b+=k[1]; a+=k[0]; break;
+    case 7 : b+=k[1]<<8; a+=k[0]; break;
+    case 6 : b+=k[1]<<16; a+=k[0]; break;
+    case 5 : b+=k[1]<<24; a+=k[0]; break;
+    case 4 : a+=k[0]; break;
+    case 3 : a+=k[0]<<8; break;
+    case 2 : a+=k[0]<<16; break;
+    case 1 : a+=k[0]<<24; break;
+    case 0 : return c;              /* zero length strings require no mixing */
+    }
+
+#else  /* make valgrind happy */
+
+    k8 = (const PN_uint8 *)k;
+    switch(length)                   /* all the case statements fall through */
+    {
+    case 12: c+=k[2]; b+=k[1]; a+=k[0]; break;
+    case 11: c+=((PN_uint32)k8[10])<<8;  /* fall through */
+    case 10: c+=((PN_uint32)k8[9])<<16;  /* fall through */
+    case 9 : c+=((PN_uint32)k8[8])<<24;  /* fall through */
+    case 8 : b+=k[1]; a+=k[0]; break;
+    case 7 : b+=((PN_uint32)k8[6])<<8;   /* fall through */
+    case 6 : b+=((PN_uint32)k8[5])<<16;  /* fall through */
+    case 5 : b+=((PN_uint32)k8[4])<<24;  /* fall through */
+    case 4 : a+=k[0]; break;
+    case 3 : a+=((PN_uint32)k8[2])<<8;   /* fall through */
+    case 2 : a+=((PN_uint32)k8[1])<<16;  /* fall through */
+    case 1 : a+=((PN_uint32)k8[0])<<24; break;
+    case 0 : return c;
+    }
+
+#endif /* !VALGRIND */
+
+  } else {                        /* need to read the key one byte at a time */
+    const PN_uint8 *k = key;
+
+    /*--------------- all but the last block: affect some 32 bits of (a,b,c) */
+    while (length > 12)
+    {
+      a += ((PN_uint32)k[0])<<24;
+      a += ((PN_uint32)k[1])<<16;
+      a += ((PN_uint32)k[2])<<8;
+      a += ((PN_uint32)k[3]);
+      b += ((PN_uint32)k[4])<<24;
+      b += ((PN_uint32)k[5])<<16;
+      b += ((PN_uint32)k[6])<<8;
+      b += ((PN_uint32)k[7]);
+      c += ((PN_uint32)k[8])<<24;
+      c += ((PN_uint32)k[9])<<16;
+      c += ((PN_uint32)k[10])<<8;
+      c += ((PN_uint32)k[11]);
+      mix(a,b,c);
+      length -= 12;
+      k += 12;
+    }
+
+    /*-------------------------------- last block: affect all 32 bits of (c) */
+    switch(length)                   /* all the case statements fall through */
+    {
+    case 12: c+=k[11];
+    case 11: c+=((PN_uint32)k[10])<<8;
+    case 10: c+=((PN_uint32)k[9])<<16;
+    case 9 : c+=((PN_uint32)k[8])<<24;
+    case 8 : b+=k[7];
+    case 7 : b+=((PN_uint32)k[6])<<8;
+    case 6 : b+=((PN_uint32)k[5])<<16;
+    case 5 : b+=((PN_uint32)k[4])<<24;
+    case 4 : a+=k[3];
+    case 3 : a+=((PN_uint32)k[2])<<8;
+    case 2 : a+=((PN_uint32)k[1])<<16;
+    case 1 : a+=((PN_uint32)k[0])<<24;
+             break;
+    case 0 : return c;
+    }
+  }
+
+  final(a,b,c);
+  return c;
+}
+
+
+#ifdef SELF_TEST
+
+/* used for timings */
+void driver1()
+{
+  PN_uint8 buf[256];
+  PN_uint32 i;
+  PN_uint32 h=0;
+  time_t a,z;
+
+  time(&a);
+  for (i=0; i<256; ++i) buf[i] = 'x';
+  for (i=0; i<1; ++i) 
+  {
+    h = hashlittle(&buf[0],1,h);
+  }
+  time(&z);
+  if (z-a > 0) printf("time %d %.8x\n", z-a, h);
+}
+
+/* check that every input bit changes every output bit half the time */
+#define HASHSTATE 1
+#define HASHLEN   1
+#define MAXPAIR 60
+#define MAXLEN  70
+void driver2()
+{
+  PN_uint8 qa[MAXLEN+1], qb[MAXLEN+2], *a = &qa[0], *b = &qb[1];
+  PN_uint32 c[HASHSTATE], d[HASHSTATE], i=0, j=0, k, l, m=0, z;
+  PN_uint32 e[HASHSTATE],f[HASHSTATE],g[HASHSTATE],h[HASHSTATE];
+  PN_uint32 x[HASHSTATE],y[HASHSTATE];
+  PN_uint32 hlen;
+
+  printf("No more than %d trials should ever be needed \n",MAXPAIR/2);
+  for (hlen=0; hlen < MAXLEN; ++hlen)
+  {
+    z=0;
+    for (i=0; i<hlen; ++i)  /*----------------------- for each input byte, */
+    {
+      for (j=0; j<8; ++j)   /*------------------------ for each input bit, */
+      {
+	for (m=1; m<8; ++m) /*------------ for serveral possible initvals, */
+	{
+	  for (l=0; l<HASHSTATE; ++l)
+	    e[l]=f[l]=g[l]=h[l]=x[l]=y[l]=~((PN_uint32)0);
+
+      	  /*---- check that every output bit is affected by that input bit */
+	  for (k=0; k<MAXPAIR; k+=2)
+	  { 
+	    PN_uint32 finished=1;
+	    /* keys have one bit different */
+	    for (l=0; l<hlen+1; ++l) {a[l] = b[l] = (PN_uint8)0;}
+	    /* have a and b be two keys differing in only one bit */
+	    a[i] ^= (k<<j);
+	    a[i] ^= (k>>(8-j));
+	     c[0] = hashlittle(a, hlen, m);
+	    b[i] ^= ((k+1)<<j);
+	    b[i] ^= ((k+1)>>(8-j));
+	     d[0] = hashlittle(b, hlen, m);
+	    /* check every bit is 1, 0, set, and not set at least once */
+	    for (l=0; l<HASHSTATE; ++l)
+	    {
+	      e[l] &= (c[l]^d[l]);
+	      f[l] &= ~(c[l]^d[l]);
+	      g[l] &= c[l];
+	      h[l] &= ~c[l];
+	      x[l] &= d[l];
+	      y[l] &= ~d[l];
+	      if (e[l]|f[l]|g[l]|h[l]|x[l]|y[l]) finished=0;
+	    }
+	    if (finished) break;
+	  }
+	  if (k>z) z=k;
+	  if (k==MAXPAIR) 
+	  {
+	     printf("Some bit didn't change: ");
+	     printf("%.8x %.8x %.8x %.8x %.8x %.8x  ",
+	            e[0],f[0],g[0],h[0],x[0],y[0]);
+	     printf("i %d j %d m %d len %d\n", i, j, m, hlen);
+	  }
+	  if (z==MAXPAIR) goto done;
+	}
+      }
+    }
+   done:
+    if (z < MAXPAIR)
+    {
+      printf("Mix success  %2d bytes  %2d initvals  ",i,m);
+      printf("required  %d  trials\n", z/2);
+    }
+  }
+  printf("\n");
+}
+
+/* Check for reading beyond the end of the buffer and alignment problems */
+void driver3()
+{
+  PN_uint8 buf[MAXLEN+20], *b;
+  PN_uint32 len;
+  PN_uint8 q[] = "This is the time for all good men to come to the aid of their country...";
+  PN_uint32 h;
+  PN_uint8 qq[] = "xThis is the time for all good men to come to the aid of their country...";
+  PN_uint32 i;
+  PN_uint8 qqq[] = "xxThis is the time for all good men to come to the aid of their country...";
+  PN_uint32 j;
+  PN_uint8 qqqq[] = "xxxThis is the time for all good men to come to the aid of their country...";
+  PN_uint32 ref,x,y;
+  PN_uint8 *p;
+
+  printf("Endianness.  These lines should all be the same (for values filled in):\n");
+  printf("%.8x                            %.8x                            %.8x\n",
+         hashword((const PN_uint32 *)q, (sizeof(q)-1)/4, 13),
+         hashword((const PN_uint32 *)q, (sizeof(q)-5)/4, 13),
+         hashword((const PN_uint32 *)q, (sizeof(q)-9)/4, 13));
+  p = q;
+  printf("%.8x %.8x %.8x %.8x %.8x %.8x %.8x %.8x %.8x %.8x %.8x %.8x\n",
+         hashlittle(p, sizeof(q)-1, 13), hashlittle(p, sizeof(q)-2, 13),
+         hashlittle(p, sizeof(q)-3, 13), hashlittle(p, sizeof(q)-4, 13),
+         hashlittle(p, sizeof(q)-5, 13), hashlittle(p, sizeof(q)-6, 13),
+         hashlittle(p, sizeof(q)-7, 13), hashlittle(p, sizeof(q)-8, 13),
+         hashlittle(p, sizeof(q)-9, 13), hashlittle(p, sizeof(q)-10, 13),
+         hashlittle(p, sizeof(q)-11, 13), hashlittle(p, sizeof(q)-12, 13));
+  p = &qq[1];
+  printf("%.8x %.8x %.8x %.8x %.8x %.8x %.8x %.8x %.8x %.8x %.8x %.8x\n",
+         hashlittle(p, sizeof(q)-1, 13), hashlittle(p, sizeof(q)-2, 13),
+         hashlittle(p, sizeof(q)-3, 13), hashlittle(p, sizeof(q)-4, 13),
+         hashlittle(p, sizeof(q)-5, 13), hashlittle(p, sizeof(q)-6, 13),
+         hashlittle(p, sizeof(q)-7, 13), hashlittle(p, sizeof(q)-8, 13),
+         hashlittle(p, sizeof(q)-9, 13), hashlittle(p, sizeof(q)-10, 13),
+         hashlittle(p, sizeof(q)-11, 13), hashlittle(p, sizeof(q)-12, 13));
+  p = &qqq[2];
+  printf("%.8x %.8x %.8x %.8x %.8x %.8x %.8x %.8x %.8x %.8x %.8x %.8x\n",
+         hashlittle(p, sizeof(q)-1, 13), hashlittle(p, sizeof(q)-2, 13),
+         hashlittle(p, sizeof(q)-3, 13), hashlittle(p, sizeof(q)-4, 13),
+         hashlittle(p, sizeof(q)-5, 13), hashlittle(p, sizeof(q)-6, 13),
+         hashlittle(p, sizeof(q)-7, 13), hashlittle(p, sizeof(q)-8, 13),
+         hashlittle(p, sizeof(q)-9, 13), hashlittle(p, sizeof(q)-10, 13),
+         hashlittle(p, sizeof(q)-11, 13), hashlittle(p, sizeof(q)-12, 13));
+  p = &qqqq[3];
+  printf("%.8x %.8x %.8x %.8x %.8x %.8x %.8x %.8x %.8x %.8x %.8x %.8x\n",
+         hashlittle(p, sizeof(q)-1, 13), hashlittle(p, sizeof(q)-2, 13),
+         hashlittle(p, sizeof(q)-3, 13), hashlittle(p, sizeof(q)-4, 13),
+         hashlittle(p, sizeof(q)-5, 13), hashlittle(p, sizeof(q)-6, 13),
+         hashlittle(p, sizeof(q)-7, 13), hashlittle(p, sizeof(q)-8, 13),
+         hashlittle(p, sizeof(q)-9, 13), hashlittle(p, sizeof(q)-10, 13),
+         hashlittle(p, sizeof(q)-11, 13), hashlittle(p, sizeof(q)-12, 13));
+  printf("\n");
+  for (h=0, b=buf+1; h<8; ++h, ++b)
+  {
+    for (i=0; i<MAXLEN; ++i)
+    {
+      len = i;
+      for (j=0; j<i; ++j) *(b+j)=0;
+
+      /* these should all be equal */
+      ref = hashlittle(b, len, (PN_uint32)1);
+      *(b+i)=(PN_uint8)~0;
+      *(b-1)=(PN_uint8)~0;
+      x = hashlittle(b, len, (PN_uint32)1);
+      y = hashlittle(b, len, (PN_uint32)1);
+      if ((ref != x) || (ref != y)) 
+      {
+	printf("alignment error: %.8x %.8x %.8x %d %d\n",ref,x,y,
+               h, i);
+      }
+    }
+  }
+}
+
+/* check for problems with nulls */
+ void driver4()
+{
+  PN_uint8 buf[1];
+  PN_uint32 h,i,state[HASHSTATE];
+
+
+  buf[0] = ~0;
+  for (i=0; i<HASHSTATE; ++i) state[i] = 1;
+  printf("These should all be different\n");
+  for (i=0, h=0; i<8; ++i)
+  {
+    h = hashlittle(buf, 0, h);
+    printf("%2ld  0-byte strings, hash is  %.8x\n", i, h);
+  }
+}
+
+
+int main()
+{
+  driver1();   /* test that the key is hashed: used for timings */
+  driver2();   /* test that whole key is hashed thoroughly */
+  driver3();   /* test that nothing but the key is hashed */
+  driver4();   /* test hashing multiple buffers (all buffers are null) */
+  return 1;
+}
+
+#endif  /* SELF_TEST */

dtool/src/dtoolbase/lookup3.h

@@ -0,0 +1,40 @@
+// Filename: lookup3.h
+// Created by:  drose (01Sep06)
+//
+////////////////////////////////////////////////////////////////////
+//
+// PANDA 3D SOFTWARE
+// Copyright (c) 2001 - 2004, Disney Enterprises, Inc.  All rights reserved
+//
+// All use of this software is subject to the terms of the Panda 3d
+// Software license.  You should have received a copy of this license
+// along with this source code; you will also find a current copy of
+// the license at http://etc.cmu.edu/panda3d/docs/license/ .
+//
+// To contact the maintainers of this program write to
+// [email protected] .
+//
+////////////////////////////////////////////////////////////////////
+
+#ifndef LOOKUP3_H
+#define LOOKUP3_H
+
+#include "dtoolbase.h"
+#include "numeric_types.h"
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+EXPCL_DTOOL PN_uint32 hashword(const PN_uint32 *k,                   /* the key, an array of PN_uint32 values */
+                               size_t          length,               /* the length of the key, in PN_uint32s */
+                               PN_uint32        initval);
+
+#ifdef __cplusplus
+};  /* end of extern "C" */
+#endif
+
+#endif
+
+
+

dtool/src/dtoolbase/pallocator.h

@@ -32,7 +32,7 @@
 //
 //               pvector, pmap, etc. are all defined in this directory
 //               to use a pallocator.
-
+//
 //               pallocator actually comes it two flavors now:
 //               pallocator_single, which can only allocate single
 //               instances of an object, and pallocator_array, which

dtool/src/dtoolbase/pmap.h

@@ -35,8 +35,8 @@
 #define pmultimap multimap
 
 #ifdef HAVE_STL_HASH
-#define phash_map hash_map
-#define phash_multimap hash_multimap
+#define phash_map stdext::hash_map
+#define phash_multimap stdext::hash_multimap
 #else  // HAVE_STL_HASH
 #define phash_map map
 #define phash_multimap multimap
@@ -137,11 +137,11 @@ public:
 //               memory.
 ////////////////////////////////////////////////////////////////////
 template<class Key, class Value, class Compare = method_hash<Key, less<Key> > >
-class phash_map : public hash_map<Key, Value, Compare, pallocator_single<pair<const Key, Value> > > {
+class phash_map : public stdext::hash_map<Key, Value, Compare, pallocator_array<pair<const Key, Value> > > {
 public:
-  phash_map() : hash_map<Key, Value, Compare, pallocator_single<pair<const Key, Value> > >() { }
-  phash_map(const phash_map<Key, Value, Compare> &copy) : hash_map<Key, Value, Compare, pallocator_single<pair<const Key, Value> > >(copy) { }
-  phash_map(const Compare &comp) : hash_map<Key, Value, Compare, pallocator_single<pair<const Key, Value> > >(comp) { }
+  phash_map() : stdext::hash_map<Key, Value, Compare, pallocator_array<pair<const Key, Value> > >() { }
+  phash_map(const phash_map<Key, Value, Compare> &copy) : stdext::hash_map<Key, Value, Compare, pallocator_array<pair<const Key, Value> > >(copy) { }
+  phash_map(const Compare &comp) : stdext::hash_map<Key, Value, Compare, pallocator_array<pair<const Key, Value> > >(comp) { }
 };
 
 ////////////////////////////////////////////////////////////////////
@@ -152,11 +152,11 @@ public:
 //               memory.
 ////////////////////////////////////////////////////////////////////
 template<class Key, class Value, class Compare = method_hash<Key, less<Key> > >
-class phash_multimap : public hash_multimap<Key, Value, Compare, pallocator_single<pair<const Key, Value> > > {
+class phash_multimap : public stdext::hash_multimap<Key, Value, Compare, pallocator_array<pair<const Key, Value> > > {
 public:
-  phash_multimap() : hash_multimap<Key, Value, Compare, pallocator_single<pair<const Key, Value> > >() { }
-  phash_multimap(const phash_multimap<Key, Value, Compare> &copy) : hash_multimap<Key, Value, Compare, pallocator_single<pair<const Key, Value> > >(copy) { }
-  phash_multimap(const Compare &comp) : hash_multimap<Key, Value, Compare, pallocator_single<pair<const Key, Value> > >(comp) { }
+  phash_multimap() : stdext::hash_multimap<Key, Value, Compare, pallocator_array<pair<const Key, Value> > >() { }
+  phash_multimap(const phash_multimap<Key, Value, Compare> &copy) : stdext::hash_multimap<Key, Value, Compare, pallocator_array<pair<const Key, Value> > >(copy) { }
+  phash_multimap(const Compare &comp) : stdext::hash_multimap<Key, Value, Compare, pallocator_array<pair<const Key, Value> > >(comp) { }
 };
 
 #else // HAVE_STL_HASH

dtool/src/dtoolbase/pset.h

@@ -35,8 +35,8 @@
 #define pmultiset multiset
 
 #ifdef HAVE_STL_HASH
-#define phash_set hash_set
-#define phash_multiset hash_multiset
+#define phash_set stdext::hash_set
+#define phash_multiset stdext::hash_multiset
 #else  // HAVE_STL_HASH
 #define phash_set set
 #define phash_multiset multiset
@@ -129,11 +129,11 @@ public:
 //               memory.
 ////////////////////////////////////////////////////////////////////
 template<class Key, class Compare = method_hash<Key, less<Key> > >
-class phash_set : public hash_set<Key, Compare, pallocator_single<Key> > {
+class phash_set : public stdext::hash_set<Key, Compare, pallocator_array<Key> > {
 public:
-  phash_set() : hash_set<Key, Compare, pallocator_single<Key> >() { }
-  phash_set(const phash_set<Key, Compare> &copy) : hash_set<Key, Compare, pallocator_single<Key> >(copy) { }
-  phash_set(const Compare &comp) : hash_set<Key, Compare, pallocator_single<Key> >(comp) { }
+  phash_set() : stdext::hash_set<Key, Compare, pallocator_array<Key> >() { }
+  phash_set(const phash_set<Key, Compare> &copy) : stdext::hash_set<Key, Compare, pallocator_array<Key> >(copy) { }
+  phash_set(const Compare &comp) : stdext::hash_set<Key, Compare, pallocator_array<Key> >(comp) { }
 };
 
 ////////////////////////////////////////////////////////////////////
@@ -144,11 +144,11 @@ public:
 //               memory.
 ////////////////////////////////////////////////////////////////////
 template<class Key, class Compare = method_hash<Key, less<Key> > >
-class phash_multiset : public hash_multiset<Key, Compare, pallocator_single<Key> > {
+class phash_multiset : public stdext::hash_multiset<Key, Compare, pallocator_array<Key> > {
 public:
-  phash_multiset() : hash_multiset<Key, Compare, pallocator_single<Key> >() { }
-  phash_multiset(const phash_multiset<Key, Compare> &copy) : hash_multiset<Key, Compare, pallocator_single<Key> >(copy) { }
-  phash_multiset(const Compare &comp) : hash_multiset<Key, Compare, pallocator_single<Key> >(comp) { }
+  phash_multiset() : stdext::hash_multiset<Key, Compare, pallocator_array<Key> >() { }
+  phash_multiset(const phash_multiset<Key, Compare> &copy) : stdext::hash_multiset<Key, Compare, pallocator_array<Key> >(copy) { }
+  phash_multiset(const Compare &comp) : stdext::hash_multiset<Key, Compare, pallocator_array<Key> >(comp) { }
 };
 
 #else // HAVE_STL_HASH

dtool/src/dtoolbase/stl_compares.I

@@ -91,7 +91,8 @@ operator () (const Key &a, const Key &b) const {
 template<class Key, class Compare>
 INLINE size_t integer_hash<Key, Compare>::
 add_hash(size_t hash, const Key &key) {
-  return (hash * 31) + (size_t)key;
+  PN_uint32 key32 = (PN_uint32)key;
+  return AddHash::add_hash(hash, &key32, 1);
 }
 
 ////////////////////////////////////////////////////////////////////
@@ -136,7 +137,8 @@ operator () (const Key &a, const Key &b) const {
 template<class Key>
 INLINE size_t floating_point_hash<Key>::
 add_hash(size_t hash, const Key &key) const {
-  return (hash * 31) + (size_t)(cfloor(key / _threshold + 0.5f));
+  PN_uint32 key32 = (PN_uint32)(key / _threshold + 0.5f);
+  return AddHash::add_hash(hash, &key32, 1);
 }
 
 ////////////////////////////////////////////////////////////////////
@@ -160,11 +162,12 @@ operator () (const Key &key) const {
 template<class Key, class Compare>
 INLINE size_t sequence_hash<Key, Compare>::
 add_hash(size_t hash, const Key &key) {
-  TYPENAME Key::const_iterator ki;
-  for (ki = key.begin(); ki != key.end(); ++ki) {
-    hash = (hash * 31) + (size_t)(*ki);
-  }
-  return hash;
+#ifdef _DEBUG
+  // We assume that the sequence is laid out sequentially in memory.
+  assert(&(*key.end()) - &key[0] == key.size());
+#endif
+  size_t num_bytes = (key.size() * sizeof(key[0]));
+  return AddHash::add_hash(hash, (const PN_uint8 *)&key[0], num_bytes);
 }
 
 ////////////////////////////////////////////////////////////////////

dtool/src/dtoolbase/stl_compares.h

@@ -22,11 +22,14 @@
 #include "dtoolbase.h"
 #include "cmath.h"
 #include "nearly_zero.h"
+#include "addHash.h"
+
+#include <assert.h>
 
 #ifdef HAVE_STL_HASH
 #include <hash_map>  // for hash_compare
 
-#define stl_hash_compare hash_compare
+#define stl_hash_compare stdext::hash_compare
 
 #else
 

dtool/src/parser-inc/hash_map

@@ -28,30 +28,32 @@
 #include <stdcompare.h>
 #include <pair>
 
-template<class key, class element, class compare = hash_compare<key, less<key> > >
-class hash_map {
-public:
-  typedef key key_type;
-  typedef element data_type;
-  typedef element mapped_type;
-  typedef pair<const key, element> value_type;
-  typedef compare key_compare;
-
-  typedef element *pointer;
-  typedef const element *const_pointer;
-  typedef element &reference;
-  typedef const element &const_reference;
-
-  class iterator;
-  class const_iterator;
-  class reverse_iterator;
-  class const_reverse_iterator;
-  typedef size_t size_type;
-  class difference_type;
-};
-
-template<class key, class element, class compare = hash_compare<key, less<key> > >
-class hash_multimap : public hash_map<key, element, compare> {
+namespace stdext {
+  template<class key, class element, class compare = hash_compare<key, less<key> > >
+  class hash_map {
+  public:
+    typedef key key_type;
+    typedef element data_type;
+    typedef element mapped_type;
+    typedef pair<const key, element> value_type;
+    typedef compare key_compare;
+
+    typedef element *pointer;
+    typedef const element *const_pointer;
+    typedef element &reference;
+    typedef const element &const_reference;
+
+    class iterator;
+    class const_iterator;
+    class reverse_iterator;
+    class const_reverse_iterator;
+    typedef size_t size_type;
+    class difference_type;
+  };
+
+  template<class key, class element, class compare = hash_compare<key, less<key> > >
+  class hash_multimap : public hash_map<key, element, compare> {
+  };
 };
 
 #endif

dtool/src/parser-inc/hash_set

@@ -27,29 +27,33 @@
 #include <stdtypedefs.h>
 #include <stdcompare.h>
 
-template<class key, class compare = hash_compare<key, less<key> > >
-class hash_set {
-public:
-  typedef key key_type;
-  typedef key value_type;
-  typedef compare key_compare;
-  typedef compare value_compare;
-
-  typedef key *pointer;
-  typedef const key *const_pointer;
-  typedef key &reference;
-  typedef const key &const_reference;
-
-  class iterator;
-  class const_iterator;
-  class reverse_iterator;
-  class const_reverse_iterator;
-  typedef size_t size_type;
-  class difference_type;
-};
+namespace stdext {
+
+  template<class key, class compare = hash_compare<key, less<key> > >
+  class hash_set {
+  public:
+    typedef key key_type;
+    typedef key value_type;
+    typedef compare key_compare;
+    typedef compare value_compare;
+
+    typedef key *pointer;
+    typedef const key *const_pointer;
+    typedef key &reference;
+    typedef const key &const_reference;
+
+    class iterator;
+    class const_iterator;
+    class reverse_iterator;
+    class const_reverse_iterator;
+    typedef size_t size_type;
+    class difference_type;
+  };
+
+  template<class key, class compare = hash_compare<key, less<key> > >
+  class hash_multiset : public hash_set<key, compare> {
+  };
 
-template<class key, class compare = hash_compare<key, less<key> > >
-class hash_multiset : public hash_set<key, compare> {
 };
 
 #endif

panda/src/pgraph/renderState.cxx

@@ -138,6 +138,26 @@ operator < (const RenderState &other) const {
                                  CompareTo<Attribute>());
 }
 
+////////////////////////////////////////////////////////////////////
+//     Function: RenderState::get_hash
+//       Access: Published
+//  Description: Returns a suitable hash value for phash_map.
+////////////////////////////////////////////////////////////////////
+size_t RenderState::
+get_hash() const {
+  size_t hash = 0;
+
+  //  hash = sequence_hash<Attributes>::add_hash(hash, _attributes);
+  Attributes::const_iterator ai;
+  for (ai = _attributes.begin(); ai != _attributes.end(); ++ai) {
+    const Attribute &attrib = *ai;
+    hash = pointer_hash::add_hash(hash, attrib._attrib);
+    hash = int_hash::add_hash(hash, attrib._override);
+  }
+
+  return hash;
+}
+
 ////////////////////////////////////////////////////////////////////
 //     Function: RenderState::cull_callback
 //       Access: Published

panda/src/pgraph/renderState.h

@@ -74,6 +74,7 @@ public:
 
 PUBLISHED:
   bool operator < (const RenderState &other) const;
+  size_t get_hash() const;
 
   INLINE bool is_empty() const;
   INLINE int get_num_attribs() const;
@@ -211,7 +212,7 @@ private:
   // to the cache, which is encoded in _composition_cache and
   // _invert_composition_cache.
   static ReMutex *_states_lock;
-  typedef pset<const RenderState *, indirect_less<const RenderState *> > States;
+  typedef phash_set<const RenderState *, indirect_less_hash<const RenderState *> > States;
   static States *_states;
   static CPT(RenderState) _empty_state;
 

panda/src/pgraph/transformState.I

@@ -33,6 +33,17 @@ operator < (const TransformState &other) const {
   return sorts_less(other, uniquify_matrix);
 }
 
+////////////////////////////////////////////////////////////////////
+//     Function: TransformState::get_hash
+//       Access: Published
+//  Description: Returns a suitable hash value for phash_map.
+////////////////////////////////////////////////////////////////////
+INLINE size_t TransformState::
+get_hash() const {
+  check_hash();
+  return _hash;
+}
+
 ////////////////////////////////////////////////////////////////////
 //     Function: TransformState::make_pos
 //       Access: Published, Static
@@ -751,6 +762,20 @@ flush_level() {
   _cache_counter.flush_level();
 }
 
+////////////////////////////////////////////////////////////////////
+//     Function: TransformState::check_hash
+//       Access: Private
+//  Description: Ensures that we know the hash value.
+////////////////////////////////////////////////////////////////////
+INLINE void TransformState::
+check_hash() const {
+  // This pretends to be a const function, even though it's not,
+  // because it only updates a transparent cache value.
+  if ((_flags & F_hash_known) == 0) {
+    ((TransformState *)this)->calc_hash();
+  }
+}
+
 ////////////////////////////////////////////////////////////////////
 //     Function: TransformState::check_singular
 //       Access: Private
@@ -824,6 +849,17 @@ check_mat() const {
   }
 }
 
+////////////////////////////////////////////////////////////////////
+//     Function: TransformState::calc_hash
+//       Access: Private
+//  Description: Computes the hash value.
+////////////////////////////////////////////////////////////////////
+INLINE void TransformState::
+calc_hash() {
+  MutexHolder holder(_lock);
+  do_calc_hash();
+}
+
 ////////////////////////////////////////////////////////////////////
 //     Function: TransformState::calc_components
 //       Access: Private

panda/src/pgraph/transformState.cxx

@@ -181,52 +181,6 @@ sorts_less(const TransformState &other, bool uniquify_matrix) const {
   }
 }
 
-////////////////////////////////////////////////////////////////////
-//     Function: TransformState::get_hash()
-//       Access: Published
-//  Description: Computes a suitable hash value for phash_map.
-////////////////////////////////////////////////////////////////////
-size_t TransformState::
-get_hash() const {
-  size_t hash = 0;
-
-  static const int significant_flags = 
-    (F_is_invalid | F_is_identity | F_components_given | F_hpr_given | F_is_2d);
-
-  int flags = (_flags & significant_flags);
-  hash = int_hash::add_hash(hash, flags);
-
-  if ((_flags & (F_is_invalid | F_is_identity)) == 0) {
-    // Only bother to put the rest of the stuff in the hash if the
-    // transform is not invalid or empty.
-    
-    if ((_flags & (F_components_given | F_hpr_given | F_quat_given)) == 
-        (F_components_given | F_hpr_given | F_quat_given)) {
-      // If the transform was specified componentwise, hash it
-      // componentwise.
-      hash = _pos.add_hash(hash);
-      if ((_flags & F_hpr_given) != 0) {
-        hash = _hpr.add_hash(hash);
-
-      } else if ((_flags & F_quat_given) != 0) {
-        hash = _quat.add_hash(hash);
-      }
-
-      hash = _scale.add_hash(hash);
-      hash = _shear.add_hash(hash);
-
-    } else {
-      // Otherwise, hash the pointer only--any two different
-      // matrix-based TransformStates are considered to be different,
-      // even if their matrices have the same values.
-
-      hash = pointer_hash::add_hash(hash, this);
-    }
-  }
-  
-  return hash;
-}
-
 ////////////////////////////////////////////////////////////////////
 //     Function: TransformState::make_identity
 //       Access: Published, Static
@@ -1736,6 +1690,52 @@ remove_cache_pointers() {
   }
 }
 
+////////////////////////////////////////////////////////////////////
+//     Function: TransformState::do_calc_hash
+//       Access: Private
+//  Description: Computes a suitable hash value for phash_map.
+////////////////////////////////////////////////////////////////////
+void TransformState::
+do_calc_hash() {
+  _hash = 0;
+
+  static const int significant_flags = 
+    (F_is_invalid | F_is_identity | F_components_given | F_hpr_given | F_is_2d);
+
+  int flags = (_flags & significant_flags);
+  _hash = int_hash::add_hash(_hash, flags);
+
+  if ((_flags & (F_is_invalid | F_is_identity)) == 0) {
+    // Only bother to put the rest of the stuff in the hash if the
+    // transform is not invalid or empty.
+    
+    if ((_flags & (F_components_given | F_hpr_given | F_quat_given)) == 
+        (F_components_given | F_hpr_given | F_quat_given)) {
+      // If the transform was specified componentwise, hash it
+      // componentwise.
+      _hash = _pos.add_hash(_hash);
+      if ((_flags & F_hpr_given) != 0) {
+        _hash = _hpr.add_hash(_hash);
+
+      } else if ((_flags & F_quat_given) != 0) {
+        _hash = _quat.add_hash(_hash);
+      }
+
+      _hash = _scale.add_hash(_hash);
+      _hash = _shear.add_hash(_hash);
+
+    } else {
+      // Otherwise, hash the pointer only--any two different
+      // matrix-based TransformStates are considered to be different,
+      // even if their matrices have the same values.
+
+      _hash = pointer_hash::add_hash(_hash, this);
+    }
+  }
+
+  _flags |= F_hash_known;
+}
+
 ////////////////////////////////////////////////////////////////////
 //     Function: TransformState::calc_singular
 //       Access: Private

panda/src/pgraph/transformState.h

@@ -74,7 +74,7 @@ public:
 PUBLISHED:
   INLINE bool operator < (const TransformState &other) const;
   bool sorts_less(const TransformState &other, bool uniquify_matrix) const;
-  size_t get_hash() const;
+  INLINE size_t get_hash() const;
 
   static CPT(TransformState) make_identity();
   static CPT(TransformState) make_invalid();
@@ -266,11 +266,14 @@ private:
 
 private:
   // This is the actual data within the TransformState.
+  INLINE void check_hash() const;
   INLINE void check_singular() const;
   INLINE void check_components() const;
   INLINE void check_hpr() const;
   INLINE void check_quat() const;
   INLINE void check_mat() const;
+  INLINE void calc_hash();
+  void do_calc_hash();
   void calc_singular();
   INLINE void calc_components();
   void do_calc_components();
@@ -307,12 +310,14 @@ private:
     F_has_nonzero_shear  = 0x00004000,
     F_is_destructing     = 0x00008000,
     F_is_2d              = 0x00010000,
+    F_hash_known         = 0x00020000,
   };
   LPoint3f _pos;
   LVecBase3f _hpr, _scale, _shear;
   LQuaternionf _quat;
   LMatrix4f _mat;
   LMatrix4f *_inv_mat;
+  size_t _hash;
   
   unsigned int _flags;