updated xxhash lib

2 files changed, 359 insertions, 342 deletions

diff --git a/lib/xxhash.c b/lib/xxhash.c
index 156b484..e9ff2d4 100644
--- a/lib/xxhash.c
+++ b/lib/xxhash.c
@@ -46,7 +46,7 @@
  * Method 2 : direct access. This method doesn't depend on compiler but violate C standard.
  *            It can generate buggy code on targets which do not support unaligned memory accesses.
  *            But in some circumstances, it's the only known way to get the most performance (ie GCC + ARMv6)
- * See https://stackoverflow.com/a/32095106/646947 for details.
+ * See http://stackoverflow.com/a/32095106/646947 for details.
  * Prefer these methods in priority order (0 > 1 > 2)
  */
 #ifndef XXH_FORCE_MEMORY_ACCESS   /* can be defined externally, on command line for example */
@@ -66,10 +66,10 @@
 /* #define XXH_ACCEPT_NULL_INPUT_POINTER 1 */
 
 /*!XXH_FORCE_NATIVE_FORMAT :
- * By default, xxHash library provides endian-independant Hash values, based on little-endian convention.
+ * By default, xxHash library provides endian-independent Hash values, based on little-endian convention.
  * Results are therefore identical for little-endian and big-endian CPU.
  * This comes at a performance cost for big-endian CPU, since some swapping is required to emulate little-endian format.
- * Should endian-independance be of no importance for your application, you may set the #define below to 1,
+ * Should endian-independence be of no importance for your application, you may set the #define below to 1,
  * to improve speed for Big-endian CPU.
  * This option has no impact on Little_Endian CPU.
  */
@@ -95,12 +95,12 @@
 /* *************************************
 *  Includes & Memory related functions
 ***************************************/
-/* Modify the local functions below should you wish to use some other memory routines */
-/* for malloc(), free() */
+/*! Modify the local functions below should you wish to use some other memory routines
+*   for malloc(), free() */
 #include <stdlib.h>
 static void* XXH_malloc(size_t s) { return malloc(s); }
 static void  XXH_free  (void* p)  { free(p); }
-/* for memcpy() */
+/*! and for memcpy() */
 #include <string.h>
 static void* XXH_memcpy(void* dest, const void* src, size_t size) { return memcpy(dest,src,size); }
 
@@ -115,7 +115,7 @@ static void* XXH_memcpy(void* dest, const void* src, size_t size) { return memcp
 #  pragma warning(disable : 4127)      /* disable: C4127: conditional expression is constant */
 #  define FORCE_INLINE static __forceinline
 #else
-#  if defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L   /* C99 */
+#  if defined (__cplusplus) || defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L   /* C99 */
 #    ifdef __GNUC__
 #      define FORCE_INLINE static inline __attribute__((always_inline))
 #    else
@@ -131,45 +131,37 @@ static void* XXH_memcpy(void* dest, const void* src, size_t size) { return memcp
 *  Basic Types
 ***************************************/
 #ifndef MEM_MODULE
-# define MEM_MODULE
-# if defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L   /* C99 */
+# if !defined (__VMS) && (defined (__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) )
 #   include <stdint.h>
     typedef uint8_t  BYTE;
     typedef uint16_t U16;
     typedef uint32_t U32;
     typedef  int32_t S32;
-    typedef uint64_t U64;
-#  else
+# else
     typedef unsigned char      BYTE;
     typedef unsigned short     U16;
     typedef unsigned int       U32;
     typedef   signed int       S32;
-    typedef unsigned long long U64;
-#  endif
+# endif
 #endif
 
-
 #if (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==2))
 
 /* Force direct memory access. Only works on CPU which support unaligned memory access in hardware */
 static U32 XXH_read32(const void* memPtr) { return *(const U32*) memPtr; }
-static U64 XXH_read64(const void* memPtr) { return *(const U64*) memPtr; }
 
 #elif (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==1))
 
 /* __pack instructions are safer, but compiler specific, hence potentially problematic for some compilers */
 /* currently only defined for gcc and icc */
-typedef union { U32 u32; U64 u64; } __attribute__((packed)) unalign;
-
+typedef union { U32 u32; } __attribute__((packed)) unalign;
 static U32 XXH_read32(const void* ptr) { return ((const unalign*)ptr)->u32; }
-static U64 XXH_read64(const void* ptr) { return ((const unalign*)ptr)->u64; }
 
 #else
 
 /* portable and safe solution. Generally efficient.
- * see : https://stackoverflow.com/a/32095106/646947
+ * see : http://stackoverflow.com/a/32095106/646947
  */
-
 static U32 XXH_read32(const void* memPtr)
 {
     U32 val;
@@ -177,20 +169,13 @@ static U32 XXH_read32(const void* memPtr)
     return val;
 }
 
-static U64 XXH_read64(const void* memPtr)
-{
-    U64 val;
-    memcpy(&val, memPtr, sizeof(val));
-    return val;
-}
-
 #endif   /* XXH_FORCE_DIRECT_MEMORY_ACCESS */
 
 
 /* ****************************************
 *  Compiler-specific Functions and Macros
 ******************************************/
-#define GCC_VERSION (__GNUC__ * 100 + __GNUC_MINOR__)
+#define XXH_GCC_VERSION (__GNUC__ * 100 + __GNUC_MINOR__)
 
 /* Note : although _rotl exists for minGW (GCC under windows), performance seems poor */
 #if defined(_MSC_VER)
@@ -203,10 +188,8 @@ static U64 XXH_read64(const void* memPtr)
 
 #if defined(_MSC_VER)     /* Visual Studio */
 #  define XXH_swap32 _byteswap_ulong
-#  define XXH_swap64 _byteswap_uint64
-#elif GCC_VERSION >= 403
+#elif XXH_GCC_VERSION >= 403
 #  define XXH_swap32 __builtin_bswap32
-#  define XXH_swap64 __builtin_bswap64
 #else
 static U32 XXH_swap32 (U32 x)
 {
@@ -215,17 +198,6 @@ static U32 XXH_swap32 (U32 x)
             ((x >>  8) & 0x0000ff00 ) |
             ((x >> 24) & 0x000000ff );
 }
-static U64 XXH_swap64 (U64 x)
-{
-    return  ((x << 56) & 0xff00000000000000ULL) |
-            ((x << 40) & 0x00ff000000000000ULL) |
-            ((x << 24) & 0x0000ff0000000000ULL) |
-            ((x << 8)  & 0x000000ff00000000ULL) |
-            ((x >> 8)  & 0x00000000ff000000ULL) |
-            ((x >> 24) & 0x0000000000ff0000ULL) |
-            ((x >> 40) & 0x000000000000ff00ULL) |
-            ((x >> 56) & 0x00000000000000ffULL);
-}
 #endif
 
 
@@ -264,67 +236,23 @@ static U32 XXH_readBE32(const void* ptr)
     return XXH_CPU_LITTLE_ENDIAN ? XXH_swap32(XXH_read32(ptr)) : XXH_read32(ptr);
 }
 
-FORCE_INLINE U64 XXH_readLE64_align(const void* ptr, XXH_endianess endian, XXH_alignment align)
-{
-    if (align==XXH_unaligned)
-        return endian==XXH_littleEndian ? XXH_read64(ptr) : XXH_swap64(XXH_read64(ptr));
-    else
-        return endian==XXH_littleEndian ? *(const U64*)ptr : XXH_swap64(*(const U64*)ptr);
-}
-
-FORCE_INLINE U64 XXH_readLE64(const void* ptr, XXH_endianess endian)
-{
-    return XXH_readLE64_align(ptr, endian, XXH_unaligned);
-}
-
-static U64 XXH_readBE64(const void* ptr)
-{
-    return XXH_CPU_LITTLE_ENDIAN ? XXH_swap64(XXH_read64(ptr)) : XXH_read64(ptr);
-}
-
 
 /* *************************************
 *  Macros
 ***************************************/
 #define XXH_STATIC_ASSERT(c)   { enum { XXH_static_assert = 1/(int)(!!(c)) }; }    /* use only *after* variable declarations */
+XXH_PUBLIC_API unsigned XXH_versionNumber (void) { return XXH_VERSION_NUMBER; }
 
 
-/* *************************************
-*  Constants
-***************************************/
+/* *******************************************************************
+*  32-bits hash functions
+*********************************************************************/
 static const U32 PRIME32_1 = 2654435761U;
 static const U32 PRIME32_2 = 2246822519U;
 static const U32 PRIME32_3 = 3266489917U;
 static const U32 PRIME32_4 =  668265263U;
 static const U32 PRIME32_5 =  374761393U;
 
-static const U64 PRIME64_1 = 11400714785074694791ULL;
-static const U64 PRIME64_2 = 14029467366897019727ULL;
-static const U64 PRIME64_3 =  1609587929392839161ULL;
-static const U64 PRIME64_4 =  9650029242287828579ULL;
-static const U64 PRIME64_5 =  2870177450012600261ULL;
-
-XXH_PUBLIC_API unsigned XXH_versionNumber (void) { return XXH_VERSION_NUMBER; }
-
-
-/* **************************
-*  Utils
-****************************/
-XXH_PUBLIC_API void XXH32_copyState(XXH32_state_t* restrict dstState, const XXH32_state_t* restrict srcState)
-{
-    memcpy(dstState, srcState, sizeof(*dstState));
-}
-
-XXH_PUBLIC_API void XXH64_copyState(XXH64_state_t* restrict dstState, const XXH64_state_t* restrict srcState)
-{
-    memcpy(dstState, srcState, sizeof(*dstState));
-}
-
-
-/* ***************************
-*  Simple Hash Functions
-*****************************/
-
 static U32 XXH32_round(U32 seed, U32 input)
 {
     seed += input * PRIME32_2;
@@ -394,10 +322,10 @@ XXH_PUBLIC_API unsigned int XXH32 (const void* input, size_t len, unsigned int s
 {
 #if 0
     /* Simple version, good for code maintenance, but unfortunately slow for small inputs */
-    XXH32_CREATESTATE_STATIC(state);
-    XXH32_reset(state, seed);
-    XXH32_update(state, input, len);
-    return XXH32_digest(state);
+    XXH32_state_t state;
+    XXH32_reset(&state, seed);
+    XXH32_update(&state, input, len);
+    return XXH32_digest(&state);
 #else
     XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN;
 
@@ -417,121 +345,8 @@ XXH_PUBLIC_API unsigned int XXH32 (const void* input, size_t len, unsigned int s
 }
 
 
-static U64 XXH64_round(U64 acc, U64 input)
-{
-    acc += input * PRIME64_2;
-    acc  = XXH_rotl64(acc, 31);
-    acc *= PRIME64_1;
-    return acc;
-}
-
-static U64 XXH64_mergeRound(U64 acc, U64 val)
-{
-    val  = XXH64_round(0, val);
-    acc ^= val;
-    acc  = acc * PRIME64_1 + PRIME64_4;
-    return acc;
-}
-
-FORCE_INLINE U64 XXH64_endian_align(const void* input, size_t len, U64 seed, XXH_endianess endian, XXH_alignment align)
-{
-    const BYTE* p = (const BYTE*)input;
-    const BYTE* const bEnd = p + len;
-    U64 h64;
-#define XXH_get64bits(p) XXH_readLE64_align(p, endian, align)
-
-#ifdef XXH_ACCEPT_NULL_INPUT_POINTER
-    if (p==NULL) {
-        len=0;
-        bEnd=p=(const BYTE*)(size_t)32;
-    }
-#endif
-
-    if (len>=32) {
-        const BYTE* const limit = bEnd - 32;
-        U64 v1 = seed + PRIME64_1 + PRIME64_2;
-        U64 v2 = seed + PRIME64_2;
-        U64 v3 = seed + 0;
-        U64 v4 = seed - PRIME64_1;
-
-        do {
-            v1 = XXH64_round(v1, XXH_get64bits(p)); p+=8;
-            v2 = XXH64_round(v2, XXH_get64bits(p)); p+=8;
-            v3 = XXH64_round(v3, XXH_get64bits(p)); p+=8;
-            v4 = XXH64_round(v4, XXH_get64bits(p)); p+=8;
-        } while (p<=limit);
-
-        h64 = XXH_rotl64(v1, 1) + XXH_rotl64(v2, 7) + XXH_rotl64(v3, 12) + XXH_rotl64(v4, 18);
-        h64 = XXH64_mergeRound(h64, v1);
-        h64 = XXH64_mergeRound(h64, v2);
-        h64 = XXH64_mergeRound(h64, v3);
-        h64 = XXH64_mergeRound(h64, v4);
-
-    } else {
-        h64  = seed + PRIME64_5;
-    }
-
-    h64 += (U64) len;
-
-    while (p+8<=bEnd) {
-        U64 const k1 = XXH64_round(0, XXH_get64bits(p));
-        h64 ^= k1;
-        h64  = XXH_rotl64(h64,27) * PRIME64_1 + PRIME64_4;
-        p+=8;
-    }
-
-    if (p+4<=bEnd) {
-        h64 ^= (U64)(XXH_get32bits(p)) * PRIME64_1;
-        h64 = XXH_rotl64(h64, 23) * PRIME64_2 + PRIME64_3;
-        p+=4;
-    }
-
-    while (p<bEnd) {
-        h64 ^= (*p) * PRIME64_5;
-        h64 = XXH_rotl64(h64, 11) * PRIME64_1;
-        p++;
-    }
-
-    h64 ^= h64 >> 33;
-    h64 *= PRIME64_2;
-    h64 ^= h64 >> 29;
-    h64 *= PRIME64_3;
-    h64 ^= h64 >> 32;
-
-    return h64;
-}
-
-
-XXH_PUBLIC_API unsigned long long XXH64 (const void* input, size_t len, unsigned long long seed)
-{
-#if 0
-    /* Simple version, good for code maintenance, but unfortunately slow for small inputs */
-    XXH64_CREATESTATE_STATIC(state);
-    XXH64_reset(state, seed);
-    XXH64_update(state, input, len);
-    return XXH64_digest(state);
-#else
-    XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN;
 
-    if (XXH_FORCE_ALIGN_CHECK) {
-        if ((((size_t)input) & 7)==0) {  /* Input is aligned, let's leverage the speed advantage */
-            if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT)
-                return XXH64_endian_align(input, len, seed, XXH_littleEndian, XXH_aligned);
-            else
-                return XXH64_endian_align(input, len, seed, XXH_bigEndian, XXH_aligned);
-    }   }
-
-    if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT)
-        return XXH64_endian_align(input, len, seed, XXH_littleEndian, XXH_unaligned);
-    else
-        return XXH64_endian_align(input, len, seed, XXH_bigEndian, XXH_unaligned);
-#endif
-}
-
-
-/* **************************************************
-*  Advanced Hash Functions
-****************************************************/
+/*======   Hash streaming   ======*/
 
 XXH_PUBLIC_API XXH32_state_t* XXH32_createState(void)
 {
@@ -543,24 +358,15 @@ XXH_PUBLIC_API XXH_errorcode XXH32_freeState(XXH32_state_t* statePtr)
     return XXH_OK;
 }
 
-XXH_PUBLIC_API XXH64_state_t* XXH64_createState(void)
+XXH_PUBLIC_API void XXH32_copyState(XXH32_state_t* dstState, const XXH32_state_t* srcState)
 {
-    return (XXH64_state_t*)XXH_malloc(sizeof(XXH64_state_t));
-}
-XXH_PUBLIC_API XXH_errorcode XXH64_freeState(XXH64_state_t* statePtr)
-{
-    XXH_free(statePtr);
-    return XXH_OK;
+    memcpy(dstState, srcState, sizeof(*dstState));
 }
 
-
-/*** Hash feed ***/
-
 XXH_PUBLIC_API XXH_errorcode XXH32_reset(XXH32_state_t* statePtr, unsigned int seed)
 {
     XXH32_state_t state;   /* using a local state to memcpy() in order to avoid strict-aliasing warnings */
-    memset(&state, 0, sizeof(state));
-    state.seed = seed;
+    memset(&state, 0, sizeof(state)-4);   /* do not write into reserved, for future removal */
     state.v1 = seed + PRIME32_1 + PRIME32_2;
     state.v2 = seed + PRIME32_2;
     state.v3 = seed + 0;
@@ -570,20 +376,6 @@ XXH_PUBLIC_API XXH_errorcode XXH32_reset(XXH32_state_t* statePtr, unsigned int s
 }
 
 
-XXH_PUBLIC_API XXH_errorcode XXH64_reset(XXH64_state_t* statePtr, unsigned long long seed)
-{
-    XXH64_state_t state;   /* using a local state to memcpy() in order to avoid strict-aliasing warnings */
-    memset(&state, 0, sizeof(state));
-    state.seed = seed;
-    state.v1 = seed + PRIME64_1 + PRIME64_2;
-    state.v2 = seed + PRIME64_2;
-    state.v3 = seed + 0;
-    state.v4 = seed - PRIME64_1;
-    memcpy(statePtr, &state, sizeof(state));
-    return XXH_OK;
-}
-
-
 FORCE_INLINE XXH_errorcode XXH32_update_endian (XXH32_state_t* state, const void* input, size_t len, XXH_endianess endian)
 {
     const BYTE* p = (const BYTE*)input;
@@ -593,11 +385,12 @@ FORCE_INLINE XXH_errorcode XXH32_update_endian (XXH32_state_t* state, const void
     if (input==NULL) return XXH_ERROR;
 #endif
 
-    state->total_len += len;
+    state->total_len_32 += (unsigned)len;
+    state->large_len |= (len>=16) | (state->total_len_32>=16);
 
     if (state->memsize + len < 16)  {   /* fill in tmp buffer */
         XXH_memcpy((BYTE*)(state->mem32) + state->memsize, input, len);
-        state->memsize += (U32)len;
+        state->memsize += (unsigned)len;
         return XXH_OK;
     }
 
@@ -634,8 +427,8 @@ FORCE_INLINE XXH_errorcode XXH32_update_endian (XXH32_state_t* state, const void
     }
 
     if (p < bEnd) {
-        XXH_memcpy(state->mem32, p, bEnd-p);
-        state->memsize = (int)(bEnd-p);
+        XXH_memcpy(state->mem32, p, (size_t)(bEnd-p));
+        state->memsize = (unsigned)(bEnd-p);
     }
 
     return XXH_OK;
@@ -659,13 +452,13 @@ FORCE_INLINE U32 XXH32_digest_endian (const XXH32_state_t* state, XXH_endianess
     const BYTE* const bEnd = (const BYTE*)(state->mem32) + state->memsize;
     U32 h32;
 
-    if (state->total_len >= 16) {
+    if (state->large_len) {
         h32 = XXH_rotl32(state->v1, 1) + XXH_rotl32(state->v2, 7) + XXH_rotl32(state->v3, 12) + XXH_rotl32(state->v4, 18);
     } else {
-        h32 = state->seed + PRIME32_5;
+        h32 = state->v3 /* == seed */ + PRIME32_5;
     }
 
-    h32 += (U32) state->total_len;
+    h32 += state->total_len_32;
 
     while (p+4<=bEnd) {
         h32 += XXH_readLE32(p, endian) * PRIME32_3;
@@ -700,8 +493,257 @@ XXH_PUBLIC_API unsigned int XXH32_digest (const XXH32_state_t* state_in)
 }
 
 
+/*======   Canonical representation   ======*/
+
+/*! Default XXH result types are basic unsigned 32 and 64 bits.
+*   The canonical representation follows human-readable write convention, aka big-endian (large digits first).
+*   These functions allow transformation of hash result into and from its canonical format.
+*   This way, hash values can be written into a file or buffer, and remain comparable across different systems and programs.
+*/
+
+XXH_PUBLIC_API void XXH32_canonicalFromHash(XXH32_canonical_t* dst, XXH32_hash_t hash)
+{
+    XXH_STATIC_ASSERT(sizeof(XXH32_canonical_t) == sizeof(XXH32_hash_t));
+    if (XXH_CPU_LITTLE_ENDIAN) hash = XXH_swap32(hash);
+    memcpy(dst, &hash, sizeof(*dst));
+}
+
+XXH_PUBLIC_API XXH32_hash_t XXH32_hashFromCanonical(const XXH32_canonical_t* src)
+{
+    return XXH_readBE32(src);
+}
+
+
+#ifndef XXH_NO_LONG_LONG
 
-/* **** XXH64 **** */
+/* *******************************************************************
+*  64-bits hash functions
+*********************************************************************/
+
+/*======   Memory access   ======*/
+
+#ifndef MEM_MODULE
+# define MEM_MODULE
+# if !defined (__VMS) && (defined (__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) )
+#   include <stdint.h>
+    typedef uint64_t U64;
+# else
+    typedef unsigned long long U64;   /* if your compiler doesn't support unsigned long long, replace by another 64-bit type here. Note that xxhash.h will also need to be updated. */
+# endif
+#endif
+
+
+#if (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==2))
+
+/* Force direct memory access. Only works on CPU which support unaligned memory access in hardware */
+static U64 XXH_read64(const void* memPtr) { return *(const U64*) memPtr; }
+
+#elif (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==1))
+
+/* __pack instructions are safer, but compiler specific, hence potentially problematic for some compilers */
+/* currently only defined for gcc and icc */
+typedef union { U32 u32; U64 u64; } __attribute__((packed)) unalign64;
+static U64 XXH_read64(const void* ptr) { return ((const unalign64*)ptr)->u64; }
+
+#else
+
+/* portable and safe solution. Generally efficient.
+ * see : http://stackoverflow.com/a/32095106/646947
+ */
+
+static U64 XXH_read64(const void* memPtr)
+{
+    U64 val;
+    memcpy(&val, memPtr, sizeof(val));
+    return val;
+}
+
+#endif   /* XXH_FORCE_DIRECT_MEMORY_ACCESS */
+
+#if defined(_MSC_VER)     /* Visual Studio */
+#  define XXH_swap64 _byteswap_uint64
+#elif XXH_GCC_VERSION >= 403
+#  define XXH_swap64 __builtin_bswap64
+#else
+static U64 XXH_swap64 (U64 x)
+{
+    return  ((x << 56) & 0xff00000000000000ULL) |
+            ((x << 40) & 0x00ff000000000000ULL) |
+            ((x << 24) & 0x0000ff0000000000ULL) |
+            ((x << 8)  & 0x000000ff00000000ULL) |
+            ((x >> 8)  & 0x00000000ff000000ULL) |
+            ((x >> 24) & 0x0000000000ff0000ULL) |
+            ((x >> 40) & 0x000000000000ff00ULL) |
+            ((x >> 56) & 0x00000000000000ffULL);
+}
+#endif
+
+FORCE_INLINE U64 XXH_readLE64_align(const void* ptr, XXH_endianess endian, XXH_alignment align)
+{
+    if (align==XXH_unaligned)
+        return endian==XXH_littleEndian ? XXH_read64(ptr) : XXH_swap64(XXH_read64(ptr));
+    else
+        return endian==XXH_littleEndian ? *(const U64*)ptr : XXH_swap64(*(const U64*)ptr);
+}
+
+FORCE_INLINE U64 XXH_readLE64(const void* ptr, XXH_endianess endian)
+{
+    return XXH_readLE64_align(ptr, endian, XXH_unaligned);
+}
+
+static U64 XXH_readBE64(const void* ptr)
+{
+    return XXH_CPU_LITTLE_ENDIAN ? XXH_swap64(XXH_read64(ptr)) : XXH_read64(ptr);
+}
+
+
+/*======   xxh64   ======*/
+
+static const U64 PRIME64_1 = 11400714785074694791ULL;
+static const U64 PRIME64_2 = 14029467366897019727ULL;
+static const U64 PRIME64_3 =  1609587929392839161ULL;
+static const U64 PRIME64_4 =  9650029242287828579ULL;
+static const U64 PRIME64_5 =  2870177450012600261ULL;
+
+static U64 XXH64_round(U64 acc, U64 input)
+{
+    acc += input * PRIME64_2;
+    acc  = XXH_rotl64(acc, 31);
+    acc *= PRIME64_1;
+    return acc;
+}
+
+static U64 XXH64_mergeRound(U64 acc, U64 val)
+{
+    val  = XXH64_round(0, val);
+    acc ^= val;
+    acc  = acc * PRIME64_1 + PRIME64_4;
+    return acc;
+}
+
+FORCE_INLINE U64 XXH64_endian_align(const void* input, size_t len, U64 seed, XXH_endianess endian, XXH_alignment align)
+{
+    const BYTE* p = (const BYTE*)input;
+    const BYTE* const bEnd = p + len;
+    U64 h64;
+#define XXH_get64bits(p) XXH_readLE64_align(p, endian, align)
+
+#ifdef XXH_ACCEPT_NULL_INPUT_POINTER
+    if (p==NULL) {
+        len=0;
+        bEnd=p=(const BYTE*)(size_t)32;
+    }
+#endif
+
+    if (len>=32) {
+        const BYTE* const limit = bEnd - 32;
+        U64 v1 = seed + PRIME64_1 + PRIME64_2;
+        U64 v2 = seed + PRIME64_2;
+        U64 v3 = seed + 0;
+        U64 v4 = seed - PRIME64_1;
+
+        do {
+            v1 = XXH64_round(v1, XXH_get64bits(p)); p+=8;
+            v2 = XXH64_round(v2, XXH_get64bits(p)); p+=8;
+            v3 = XXH64_round(v3, XXH_get64bits(p)); p+=8;
+            v4 = XXH64_round(v4, XXH_get64bits(p)); p+=8;
+        } while (p<=limit);
+
+        h64 = XXH_rotl64(v1, 1) + XXH_rotl64(v2, 7) + XXH_rotl64(v3, 12) + XXH_rotl64(v4, 18);
+        h64 = XXH64_mergeRound(h64, v1);
+        h64 = XXH64_mergeRound(h64, v2);
+        h64 = XXH64_mergeRound(h64, v3);
+        h64 = XXH64_mergeRound(h64, v4);
+
+    } else {
+        h64  = seed + PRIME64_5;
+    }
+
+    h64 += (U64) len;
+
+    while (p+8<=bEnd) {
+        U64 const k1 = XXH64_round(0, XXH_get64bits(p));
+        h64 ^= k1;
+        h64  = XXH_rotl64(h64,27) * PRIME64_1 + PRIME64_4;
+        p+=8;
+    }
+
+    if (p+4<=bEnd) {
+        h64 ^= (U64)(XXH_get32bits(p)) * PRIME64_1;
+        h64 = XXH_rotl64(h64, 23) * PRIME64_2 + PRIME64_3;
+        p+=4;
+    }
+
+    while (p<bEnd) {
+        h64 ^= (*p) * PRIME64_5;
+        h64 = XXH_rotl64(h64, 11) * PRIME64_1;
+        p++;
+    }
+
+    h64 ^= h64 >> 33;
+    h64 *= PRIME64_2;
+    h64 ^= h64 >> 29;
+    h64 *= PRIME64_3;
+    h64 ^= h64 >> 32;
+
+    return h64;
+}
+
+
+XXH_PUBLIC_API unsigned long long XXH64 (const void* input, size_t len, unsigned long long seed)
+{
+#if 0
+    /* Simple version, good for code maintenance, but unfortunately slow for small inputs */
+    XXH64_state_t state;
+    XXH64_reset(&state, seed);
+    XXH64_update(&state, input, len);
+    return XXH64_digest(&state);
+#else
+    XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN;
+
+    if (XXH_FORCE_ALIGN_CHECK) {
+        if ((((size_t)input) & 7)==0) {  /* Input is aligned, let's leverage the speed advantage */
+            if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT)
+                return XXH64_endian_align(input, len, seed, XXH_littleEndian, XXH_aligned);
+            else
+                return XXH64_endian_align(input, len, seed, XXH_bigEndian, XXH_aligned);
+    }   }
+
+    if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT)
+        return XXH64_endian_align(input, len, seed, XXH_littleEndian, XXH_unaligned);
+    else
+        return XXH64_endian_align(input, len, seed, XXH_bigEndian, XXH_unaligned);
+#endif
+}
+
+/*======   Hash Streaming   ======*/
+
+XXH_PUBLIC_API XXH64_state_t* XXH64_createState(void)
+{
+    return (XXH64_state_t*)XXH_malloc(sizeof(XXH64_state_t));
+}
+XXH_PUBLIC_API XXH_errorcode XXH64_freeState(XXH64_state_t* statePtr)
+{
+    XXH_free(statePtr);
+    return XXH_OK;
+}
+
+XXH_PUBLIC_API void XXH64_copyState(XXH64_state_t* dstState, const XXH64_state_t* srcState)
+{
+    memcpy(dstState, srcState, sizeof(*dstState));
+}
+
+XXH_PUBLIC_API XXH_errorcode XXH64_reset(XXH64_state_t* statePtr, unsigned long long seed)
+{
+    XXH64_state_t state;   /* using a local state to memcpy() in order to avoid strict-aliasing warnings */
+    memset(&state, 0, sizeof(state)-8);   /* do not write into reserved, for future removal */
+    state.v1 = seed + PRIME64_1 + PRIME64_2;
+    state.v2 = seed + PRIME64_2;
+    state.v3 = seed + 0;
+    state.v4 = seed - PRIME64_1;
+    memcpy(statePtr, &state, sizeof(state));
+    return XXH_OK;
+}
 
 FORCE_INLINE XXH_errorcode XXH64_update_endian (XXH64_state_t* state, const void* input, size_t len, XXH_endianess endian)
 {
@@ -751,8 +793,8 @@ FORCE_INLINE XXH_errorcode XXH64_update_endian (XXH64_state_t* state, const void
     }
 
     if (p < bEnd) {
-        XXH_memcpy(state->mem64, p, bEnd-p);
-        state->memsize = (int)(bEnd-p);
+        XXH_memcpy(state->mem64, p, (size_t)(bEnd-p));
+        state->memsize = (unsigned)(bEnd-p);
     }
 
     return XXH_OK;
@@ -768,8 +810,6 @@ XXH_PUBLIC_API XXH_errorcode XXH64_update (XXH64_state_t* state_in, const void*
         return XXH64_update_endian(state_in, input, len, XXH_bigEndian);
 }
 
-
-
 FORCE_INLINE U64 XXH64_digest_endian (const XXH64_state_t* state, XXH_endianess endian)
 {
     const BYTE * p = (const BYTE*)state->mem64;
@@ -788,7 +828,7 @@ FORCE_INLINE U64 XXH64_digest_endian (const XXH64_state_t* state, XXH_endianess
         h64 = XXH64_mergeRound(h64, v3);
         h64 = XXH64_mergeRound(h64, v4);
     } else {
-        h64  = state->seed + PRIME64_5;
+        h64  = state->v3 + PRIME64_5;
     }
 
     h64 += (U64) state->total_len;
@@ -821,7 +861,6 @@ FORCE_INLINE U64 XXH64_digest_endian (const XXH64_state_t* state, XXH_endianess
     return h64;
 }
 
-
 XXH_PUBLIC_API unsigned long long XXH64_digest (const XXH64_state_t* state_in)
 {
     XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN;
@@ -833,22 +872,7 @@ XXH_PUBLIC_API unsigned long long XXH64_digest (const XXH64_state_t* state_in)
 }
 
 
-/* **************************
-*  Canonical representation
-****************************/
-
-/*! Default XXH result types are basic unsigned 32 and 64 bits.
-*   The canonical representation follows human-readable write convention, aka big-endian (large digits first).
-*   These functions allow transformation of hash result into and from its canonical format.
-*   This way, hash values can be written into a file or buffer, and remain comparable across different systems and programs.
-*/
-
-XXH_PUBLIC_API void XXH32_canonicalFromHash(XXH32_canonical_t* dst, XXH32_hash_t hash)
-{
-    XXH_STATIC_ASSERT(sizeof(XXH32_canonical_t) == sizeof(XXH32_hash_t));
-    if (XXH_CPU_LITTLE_ENDIAN) hash = XXH_swap32(hash);
-    memcpy(dst, &hash, sizeof(*dst));
-}
+/*====== Canonical representation   ======*/
 
 XXH_PUBLIC_API void XXH64_canonicalFromHash(XXH64_canonical_t* dst, XXH64_hash_t hash)
 {
@@ -857,12 +881,9 @@ XXH_PUBLIC_API void XXH64_canonicalFromHash(XXH64_canonical_t* dst, XXH64_hash_t
     memcpy(dst, &hash, sizeof(*dst));
 }
 
-XXH_PUBLIC_API XXH32_hash_t XXH32_hashFromCanonical(const XXH32_canonical_t* src)
-{
-    return XXH_readBE32(src);
-}
-
 XXH_PUBLIC_API XXH64_hash_t XXH64_hashFromCanonical(const XXH64_canonical_t* src)
 {
     return XXH_readBE64(src);
 }
+
+#endif  /* XXH_NO_LONG_LONG */
diff --git a/lib/xxhash.h b/lib/xxhash.h
index 4d7feff..870a6d9 100644
--- a/lib/xxhash.h
+++ b/lib/xxhash.h
@@ -83,20 +83,20 @@ typedef enum { XXH_OK=0, XXH_ERROR } XXH_errorcode;
 *  API modifier
 ******************************/
 /** XXH_PRIVATE_API
-*   This is useful if you want to include xxhash functions in `static` mode
+*   This is useful to include xxhash functions in `static` mode
 *   in order to inline them, and remove their symbol from the public list.
 *   Methodology :
 *     #define XXH_PRIVATE_API
 *     #include "xxhash.h"
-*   `xxhash.c` is automatically included, so the file is still needed,
-*   but it's not useful to compile and link it anymore.
+*   `xxhash.c` is automatically included.
+*   It's not useful to compile and link it as a separate module.
 */
 #ifdef XXH_PRIVATE_API
 #  ifndef XXH_STATIC_LINKING_ONLY
 #    define XXH_STATIC_LINKING_ONLY
 #  endif
 #  if defined(__GNUC__)
-#    define XXH_PUBLIC_API static __attribute__((unused))
+#    define XXH_PUBLIC_API static __inline __attribute__((unused))
 #  elif defined (__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */)
 #    define XXH_PUBLIC_API static inline
 #  elif defined(_MSC_VER)
@@ -111,10 +111,10 @@ typedef enum { XXH_OK=0, XXH_ERROR } XXH_errorcode;
 /*!XXH_NAMESPACE, aka Namespace Emulation :
 
 If you want to include _and expose_ xxHash functions from within your own library,
-but also want to avoid symbol collisions with another library which also includes xxHash,
+but also want to avoid symbol collisions with other libraries which may also include xxHash,
 
 you can use XXH_NAMESPACE, to automatically prefix any public symbol from xxhash library
-with the value of XXH_NAMESPACE (so avoid to keep it NULL and avoid numeric values).
+with the value of XXH_NAMESPACE (therefore, avoid NULL and numeric values).
 
 Note that no change is required within the calling program as long as it includes `xxhash.h` :
 regular symbol name will be automatically translated by this header.
@@ -122,21 +122,25 @@ regular symbol name will be automatically translated by this header.
 #ifdef XXH_NAMESPACE
 #  define XXH_CAT(A,B) A##B
 #  define XXH_NAME2(A,B) XXH_CAT(A,B)
-#  define XXH32 XXH_NAME2(XXH_NAMESPACE, XXH32)
-#  define XXH64 XXH_NAME2(XXH_NAMESPACE, XXH64)
 #  define XXH_versionNumber XXH_NAME2(XXH_NAMESPACE, XXH_versionNumber)
+#  define XXH32 XXH_NAME2(XXH_NAMESPACE, XXH32)
 #  define XXH32_createState XXH_NAME2(XXH_NAMESPACE, XXH32_createState)
-#  define XXH64_createState XXH_NAME2(XXH_NAMESPACE, XXH64_createState)
 #  define XXH32_freeState XXH_NAME2(XXH_NAMESPACE, XXH32_freeState)
-#  define XXH64_freeState XXH_NAME2(XXH_NAMESPACE, XXH64_freeState)
 #  define XXH32_reset XXH_NAME2(XXH_NAMESPACE, XXH32_reset)
-#  define XXH64_reset XXH_NAME2(XXH_NAMESPACE, XXH64_reset)
 #  define XXH32_update XXH_NAME2(XXH_NAMESPACE, XXH32_update)
-#  define XXH64_update XXH_NAME2(XXH_NAMESPACE, XXH64_update)
 #  define XXH32_digest XXH_NAME2(XXH_NAMESPACE, XXH32_digest)
-#  define XXH64_digest XXH_NAME2(XXH_NAMESPACE, XXH64_digest)
 #  define XXH32_copyState XXH_NAME2(XXH_NAMESPACE, XXH32_copyState)
+#  define XXH32_canonicalFromHash XXH_NAME2(XXH_NAMESPACE, XXH32_canonicalFromHash)
+#  define XXH32_hashFromCanonical XXH_NAME2(XXH_NAMESPACE, XXH32_hashFromCanonical)
+#  define XXH64 XXH_NAME2(XXH_NAMESPACE, XXH64)
+#  define XXH64_createState XXH_NAME2(XXH_NAMESPACE, XXH64_createState)
+#  define XXH64_freeState XXH_NAME2(XXH_NAMESPACE, XXH64_freeState)
+#  define XXH64_reset XXH_NAME2(XXH_NAMESPACE, XXH64_reset)
+#  define XXH64_update XXH_NAME2(XXH_NAMESPACE, XXH64_update)
+#  define XXH64_digest XXH_NAME2(XXH_NAMESPACE, XXH64_digest)
 #  define XXH64_copyState XXH_NAME2(XXH_NAMESPACE, XXH64_copyState)
+#  define XXH64_canonicalFromHash XXH_NAME2(XXH_NAMESPACE, XXH64_canonicalFromHash)
+#  define XXH64_hashFromCanonical XXH_NAME2(XXH_NAMESPACE, XXH64_hashFromCanonical)
 #endif
 
 
@@ -145,58 +149,33 @@ regular symbol name will be automatically translated by this header.
 ***************************************/
 #define XXH_VERSION_MAJOR    0
 #define XXH_VERSION_MINOR    6
-#define XXH_VERSION_RELEASE  1
+#define XXH_VERSION_RELEASE  2
 #define XXH_VERSION_NUMBER  (XXH_VERSION_MAJOR *100*100 + XXH_VERSION_MINOR *100 + XXH_VERSION_RELEASE)
 XXH_PUBLIC_API unsigned XXH_versionNumber (void);
 
 
-/* ****************************
-*  Simple Hash Functions
-******************************/
+/*-**********************************************************************
+*  32-bits hash
+************************************************************************/
 typedef unsigned int       XXH32_hash_t;
-typedef unsigned long long XXH64_hash_t;
-
-XXH_PUBLIC_API XXH32_hash_t XXH32 (const void* input, size_t length, unsigned int seed);
-XXH_PUBLIC_API XXH64_hash_t XXH64 (const void* input, size_t length, unsigned long long seed);
 
-/*!
-XXH32() :
+/*! XXH32() :
     Calculate the 32-bits hash of sequence "length" bytes stored at memory address "input".
     The memory between input & input+length must be valid (allocated and read-accessible).
     "seed" can be used to alter the result predictably.
-    Speed on Core 2 Duo @ 3 GHz (single thread, SMHasher benchmark) : 5.4 GB/s
-XXH64() :
-    Calculate the 64-bits hash of sequence of length "len" stored at memory address "input".
-    "seed" can be used to alter the result predictably.
-    This function runs 2x faster on 64-bits systems, but slower on 32-bits systems (see benchmark).
-*/
-
+    Speed on Core 2 Duo @ 3 GHz (single thread, SMHasher benchmark) : 5.4 GB/s */
+XXH_PUBLIC_API XXH32_hash_t XXH32 (const void* input, size_t length, unsigned int seed);
 
-/* ****************************
-*  Streaming Hash Functions
-******************************/
+/*======   Streaming   ======*/
 typedef struct XXH32_state_s XXH32_state_t;   /* incomplete type */
-typedef struct XXH64_state_s XXH64_state_t;   /* incomplete type */
-
-/*! State allocation, compatible with dynamic libraries */
-
 XXH_PUBLIC_API XXH32_state_t* XXH32_createState(void);
 XXH_PUBLIC_API XXH_errorcode  XXH32_freeState(XXH32_state_t* statePtr);
-
-XXH_PUBLIC_API XXH64_state_t* XXH64_createState(void);
-XXH_PUBLIC_API XXH_errorcode  XXH64_freeState(XXH64_state_t* statePtr);
-
-
-/* hash streaming */
+XXH_PUBLIC_API void XXH32_copyState(XXH32_state_t* dst_state, const XXH32_state_t* src_state);
 
 XXH_PUBLIC_API XXH_errorcode XXH32_reset  (XXH32_state_t* statePtr, unsigned int seed);
 XXH_PUBLIC_API XXH_errorcode XXH32_update (XXH32_state_t* statePtr, const void* input, size_t length);
 XXH_PUBLIC_API XXH32_hash_t  XXH32_digest (const XXH32_state_t* statePtr);
 
-XXH_PUBLIC_API XXH_errorcode XXH64_reset  (XXH64_state_t* statePtr, unsigned long long seed);
-XXH_PUBLIC_API XXH_errorcode XXH64_update (XXH64_state_t* statePtr, const void* input, size_t length);
-XXH_PUBLIC_API XXH64_hash_t  XXH64_digest (const XXH64_state_t* statePtr);
-
 /*
 These functions generate the xxHash of an input provided in multiple segments.
 Note that, for small input, they are slower than single-call functions, due to state management.
@@ -219,29 +198,11 @@ and generate some new hashes later on, by calling again XXH*_digest().
 When done, free XXH state space if it was allocated dynamically.
 */
 
+/*======   Canonical representation   ======*/
 
-/* **************************
-*  Utils
-****************************/
-#if !(defined(__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L))   /* ! C99 */
-#  define restrict   /* disable restrict */
-#endif
-
-XXH_PUBLIC_API void XXH32_copyState(XXH32_state_t* restrict dst_state, const XXH32_state_t* restrict src_state);
-XXH_PUBLIC_API void XXH64_copyState(XXH64_state_t* restrict dst_state, const XXH64_state_t* restrict src_state);
-
-
-/* **************************
-*  Canonical representation
-****************************/
 typedef struct { unsigned char digest[4]; } XXH32_canonical_t;
-typedef struct { unsigned char digest[8]; } XXH64_canonical_t;
-
 XXH_PUBLIC_API void XXH32_canonicalFromHash(XXH32_canonical_t* dst, XXH32_hash_t hash);
-XXH_PUBLIC_API void XXH64_canonicalFromHash(XXH64_canonical_t* dst, XXH64_hash_t hash);
-
 XXH_PUBLIC_API XXH32_hash_t XXH32_hashFromCanonical(const XXH32_canonical_t* src);
-XXH_PUBLIC_API XXH64_hash_t XXH64_hashFromCanonical(const XXH64_canonical_t* src);
 
 /* Default result type for XXH functions are primitive unsigned 32 and 64 bits.
 *  The canonical representation uses human-readable write convention, aka big-endian (large digits first).
@@ -250,41 +211,76 @@ XXH_PUBLIC_API XXH64_hash_t XXH64_hashFromCanonical(const XXH64_canonical_t* src
 */
 
 
+#ifndef XXH_NO_LONG_LONG
+/*-**********************************************************************
+*  64-bits hash
+************************************************************************/
+typedef unsigned long long XXH64_hash_t;
+
+/*! XXH64() :
+    Calculate the 64-bits hash of sequence of length "len" stored at memory address "input".
+    "seed" can be used to alter the result predictably.
+    This function runs faster on 64-bits systems, but slower on 32-bits systems (see benchmark).
+*/
+XXH_PUBLIC_API XXH64_hash_t XXH64 (const void* input, size_t length, unsigned long long seed);
+
+/*======   Streaming   ======*/
+typedef struct XXH64_state_s XXH64_state_t;   /* incomplete type */
+XXH_PUBLIC_API XXH64_state_t* XXH64_createState(void);
+XXH_PUBLIC_API XXH_errorcode  XXH64_freeState(XXH64_state_t* statePtr);
+XXH_PUBLIC_API void XXH64_copyState(XXH64_state_t* dst_state, const XXH64_state_t* src_state);
+
+XXH_PUBLIC_API XXH_errorcode XXH64_reset  (XXH64_state_t* statePtr, unsigned long long seed);
+XXH_PUBLIC_API XXH_errorcode XXH64_update (XXH64_state_t* statePtr, const void* input, size_t length);
+XXH_PUBLIC_API XXH64_hash_t  XXH64_digest (const XXH64_state_t* statePtr);
+
+/*======   Canonical representation   ======*/
+typedef struct { unsigned char digest[8]; } XXH64_canonical_t;
+XXH_PUBLIC_API void XXH64_canonicalFromHash(XXH64_canonical_t* dst, XXH64_hash_t hash);
+XXH_PUBLIC_API XXH64_hash_t XXH64_hashFromCanonical(const XXH64_canonical_t* src);
+#endif  /* XXH_NO_LONG_LONG */
+
+
 #ifdef XXH_STATIC_LINKING_ONLY
 
 /* ================================================================================================
    This section contains definitions which are not guaranteed to remain stable.
-   They could change in a future version, becoming incompatible with a different version of the library.
+   They may change in future versions, becoming incompatible with a different version of the library.
    They shall only be used with static linking.
+   Never use these definitions in association with dynamic linking !
 =================================================================================================== */
 
-/* These definitions allow allocating XXH state statically (on stack) */
+/* These definitions are only meant to allow allocation of XXH state
+   statically, on stack, or in a struct for example.
+   Do not use members directly. */
 
    struct XXH32_state_s {
-       unsigned long long total_len;
-       unsigned seed;
+       unsigned total_len_32;
+       unsigned large_len;
        unsigned v1;
        unsigned v2;
        unsigned v3;
        unsigned v4;
        unsigned mem32[4];   /* buffer defined as U32 for alignment */
        unsigned memsize;
+       unsigned reserved;   /* never read nor write, will be removed in a future version */
    };   /* typedef'd to XXH32_state_t */
 
+#ifndef XXH_NO_LONG_LONG
    struct XXH64_state_s {
        unsigned long long total_len;
-       unsigned long long seed;
        unsigned long long v1;
        unsigned long long v2;
        unsigned long long v3;
        unsigned long long v4;
        unsigned long long mem64[4];   /* buffer defined as U64 for alignment */
        unsigned memsize;
+       unsigned reserved[2];          /* never read nor write, will be removed in a future version */
    };   /* typedef'd to XXH64_state_t */
-
+#endif
 
 #  ifdef XXH_PRIVATE_API
-#    include "xxhash.c"   /* include xxhash functions as `static`, for inlining */
+#    include "xxhash.c"   /* include xxhash function bodies as `static`, for inlining */
 #  endif
 
 #endif /* XXH_STATIC_LINKING_ONLY */