CDRIVER-2671 swap MD5 with FNV-1a in ObjectIDs

Author: edobranov

src/libbson/CMakeLists.txt

@@ -162,6 +162,7 @@ set (SOURCES
    ${PROJECT_SOURCE_DIR}/src/bson/bson-context.c
    ${PROJECT_SOURCE_DIR}/src/bson/bson-decimal128.c
    ${PROJECT_SOURCE_DIR}/src/bson/bson-error.c
+   ${PROJECT_SOURCE_DIR}/src/bson/bson-fnv.c
    ${PROJECT_SOURCE_DIR}/src/bson/bson-iso8601.c
    ${PROJECT_SOURCE_DIR}/src/bson/bson-iter.c
    ${PROJECT_SOURCE_DIR}/src/bson/bson-json.c

src/libbson/THIRD_PARTY_NOTICES

@@ -52,3 +52,21 @@ freely, subject to the following restrictions:
 
 L. Peter Deutsch
 [email protected]
+
+
+License notice for bson-fnv-private.h, bson-fnv.c and test-fnv.c
+-------------------------------------------------------------------------------
+
+Please do not copyright this code.  This code is in the public domain.
+
+LANDON CURT NOLL DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS SOFTWARE,
+INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS. IN NO
+EVENT SHALL LANDON CURT NOLL BE LIABLE FOR ANY SPECIAL, INDIRECT OR
+CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF
+USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR
+OTHER TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
+PERFORMANCE OF THIS SOFTWARE.
+
+By:
+ chongo <Landon Curt Noll> /\oo/\
+     http://www.isthe.com/chongo/

src/libbson/doc/bson_md5_t.rst

@@ -19,7 +19,7 @@ Synopsis
 Description
 -----------
 
-bson_md5_t encapsulates an implementation of the MD5 algorithm. This is used in OID generation for the MD5(hostname) bytes. It is also used by some libraries such as the MongoDB C driver.
+bson_md5_t encapsulates an implementation of the MD5 algorithm.
 
 .. only:: html
 

src/libbson/doc/oid.rst

@@ -11,7 +11,7 @@ Composition
 -----------
 
 * 4 bytes : The UNIX timestamp in big-endian format.
-* 3 bytes : The first 3 bytes of ``MD5(hostname)``.
+* 3 bytes : A hash of the hostname.
 * 2 bytes : The ``pid_t`` of the current process. Alternatively the task-id if configured.
 * 3 bytes : A 24-bit monotonic counter incrementing from ``rand()`` in big-endian.
 

src/libbson/src/bson/CMakeLists.txt

@@ -25,6 +25,7 @@ set (src_libbson_src_bson_DIST_hs
    bson-private.h
    bson-iso8601-private.h
    bson-context-private.h
+   bson-fnv-private.h
    bson-thread-private.h
    bson-timegm-private.h
 )
@@ -40,6 +41,7 @@ set (src_libbson_src_bson_DIST_cs
    bson-context.c
    bson-decimal128.c
    bson-error.c
+   bson-fnv.c
    bson-iter.c
    bson-iso8601.c
    bson-json.c

src/libbson/src/bson/bson-context-private.h

@@ -31,7 +31,7 @@ struct _bson_context_t {
    int flags : 7;
    bool pidbe_once : 1;
    uint8_t pidbe[2];
-   uint8_t md5[3];
+   uint8_t fnv[3];
    int32_t seq32;
    int64_t seq64;
 

src/libbson/src/bson/bson-context.c

@@ -26,7 +26,7 @@
 #include "bson-clock.h"
 #include "bson-context.h"
 #include "bson-context-private.h"
-#include "bson-md5.h"
+#include "bson-fnv-private.h"
 #include "bson-memory.h"
 #include "bson-thread-private.h"
 
@@ -60,7 +60,7 @@ bson_gettid (void)
  *
  * _bson_context_get_oid_host --
  *
- *       Retrieves the first three bytes of MD5(hostname) and assigns them
+ *       Retrieves a 3 byte hash of the hostname and assigns those bytes
  *       to the host portion of oid.
  *
  * Returns:
@@ -77,8 +77,7 @@ _bson_context_get_oid_host (bson_context_t *context, /* IN */
                             bson_oid_t *oid)         /* OUT */
 {
    uint8_t *bytes = (uint8_t *) oid;
-   uint8_t digest[16];
-   bson_md5_t md5;
+   uint32_t fnv1a_hash;
    char hostname[HOST_NAME_MAX];
 
    BSON_ASSERT (context);
@@ -87,14 +86,11 @@ _bson_context_get_oid_host (bson_context_t *context, /* IN */
    gethostname (hostname, sizeof hostname);
    hostname[HOST_NAME_MAX - 1] = '\0';
 
-   bson_md5_init (&md5);
-   bson_md5_append (
-      &md5, (const uint8_t *) hostname, (uint32_t) strlen (hostname));
-   bson_md5_finish (&md5, &digest[0]);
+   fnv1a_hash = _mongoc_fnv_24a_str (hostname);
 
-   bytes[4] = digest[0];
-   bytes[5] = digest[1];
-   bytes[6] = digest[2];
+   bytes[4] = (uint8_t) ((fnv1a_hash >> (0 * 8)) & 0xff);
+   bytes[5] = (uint8_t) ((fnv1a_hash >> (1 * 8)) & 0xff);
+   bytes[6] = (uint8_t) ((fnv1a_hash >> (2 * 8)) & 0xff);
 }
 
 
@@ -103,7 +99,7 @@ _bson_context_get_oid_host (bson_context_t *context, /* IN */
  *
  * _bson_context_get_oid_host_cached --
  *
- *       Fetch the cached copy of the MD5(hostname).
+ *       Fetch the cached copy of the 3 byte hash of the hostname
  *
  * Returns:
  *       None.
@@ -121,9 +117,9 @@ _bson_context_get_oid_host_cached (bson_context_t *context, /* IN */
    BSON_ASSERT (context);
    BSON_ASSERT (oid);
 
-   oid->bytes[4] = context->md5[0];
-   oid->bytes[5] = context->md5[1];
-   oid->bytes[6] = context->md5[2];
+   oid->bytes[4] = context->fnv[0];
+   oid->bytes[5] = context->fnv[1];
+   oid->bytes[6] = context->fnv[2];
 }
 
 
@@ -364,9 +360,9 @@ _bson_context_init (bson_context_t *context,    /* IN */
       context->oid_get_host = _bson_context_get_oid_host;
    } else {
       _bson_context_get_oid_host (context, &oid);
-      context->md5[0] = oid.bytes[4];
-      context->md5[1] = oid.bytes[5];
-      context->md5[2] = oid.bytes[6];
+      context->fnv[0] = oid.bytes[4];
+      context->fnv[1] = oid.bytes[5];
+      context->fnv[2] = oid.bytes[6];
    }
 
    if ((flags & BSON_CONTEXT_THREAD_SAFE)) {

src/libbson/src/bson/bson-fnv-private.h

@@ -0,0 +1,25 @@
+/*
+ * Please do not copyright this code.  This code is in the public domain.
+ *
+ * LANDON CURT NOLL DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS SOFTWARE,
+ * INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS. IN NO
+ * EVENT SHALL LANDON CURT NOLL BE LIABLE FOR ANY SPECIAL, INDIRECT OR
+ * CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF
+ * USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR
+ * OTHER TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
+ * PERFORMANCE OF THIS SOFTWARE.
+ *
+ * By:
+ *  chongo <Landon Curt Noll> /\oo/\
+ *      http://www.isthe.com/chongo/
+ */
+
+#ifndef MONGO_C_DRIVER_BSON_FNV_PRIVATE_H
+#define MONGO_C_DRIVER_BSON_FNV_PRIVATE_H
+
+#include "bson.h"
+
+uint32_t
+_mongoc_fnv_24a_str (char *str);
+
+#endif /* MONGO_C_DRIVER_BSON_FNV_PRIVATE_H */

src/libbson/src/bson/bson-fnv.c

@@ -0,0 +1,90 @@
+/*
+ * Please do not copyright this code.  This code is in the public domain.
+ *
+ * LANDON CURT NOLL DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS SOFTWARE,
+ * INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS. IN NO
+ * EVENT SHALL LANDON CURT NOLL BE LIABLE FOR ANY SPECIAL, INDIRECT OR
+ * CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF
+ * USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR
+ * OTHER TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
+ * PERFORMANCE OF THIS SOFTWARE.
+ *
+ * By:
+ *  chongo <Landon Curt Noll> /\oo/\
+ *      http://www.isthe.com/chongo/
+ */
+
+/*
+ * Sources: http://www.isthe.com/chongo/src/fnv/hash_32a.c
+ *          http://www.isthe.com/chongo/tech/comp/fnv/index.html#xor-fold
+ */
+
+#include "bson-fnv-private.h"
+
+/*
+ * 32 bit FNV-1a non-zero initial basis
+ *
+ * The FNV-1 initial basis is the FNV-0 hash of the following 32 octets:
+ *
+ *              chongo <Landon Curt Noll> /\../\
+ *
+ * NOTE: The \'s above are not back-slashing escape characters.
+ * They are literal ASCII  backslash 0x5c characters.
+ *
+ * NOTE: The FNV-1a initial basis is the same value as FNV-1 by definition.
+ */
+#define FNV1_32A_INIT ((uint32_t) 0x811c9dc5)
+
+/*
+ * 32 bit magic FNV-1a prime
+ * NOTE: define NO_FNV_GCC_OPTIMIZATION to use this prime
+ */
+#define FNV_32_PRIME ((Fnv32_t) 0x01000193)
+
+/*
+ * For producing 24 bit FNV-1a hash using xor-fold on a 32 bit FNV-1a hash
+ */
+#define MASK_24 (((uint32_t) 1 << 24) - 1) /* i.e., (uint32_t) 0xffffff */
+
+/*---------------------------------------------------------------------------
+ *
+ * _mongoc_fnv_24a_str --
+ *
+ *       perform a 32 bit Fowler/Noll/Vo FNV-1a hash on a string and
+ *       xor-fold it into a 24 bit hash
+ *
+ * Return:
+ *       24 bit hash as a static hash type
+ *
+ * Note:
+ *       input strings with multiple null characters will stop being
+ *       processed at the first null
+ *
+ *--------------------------------------------------------------------------
+ */
+uint32_t
+_mongoc_fnv_24a_str (char *str)
+{
+   uint32_t hval = FNV1_32A_INIT;            /* initial 32 bit hash basis */
+   unsigned char *s = (unsigned char *) str; /* unsigned string */
+
+   /* FNV-1a hash each octet in the buffer */
+   while (*s) {
+      /* xor the bottom with the current octet */
+      hval ^= (uint32_t) *s++;
+
+/* multiply by the 32 bit FNV magic prime mod 2^32 */
+#if defined(NO_FNV_GCC_OPTIMIZATION)
+      hval *= FNV_32_PRIME;
+#else
+      hval +=
+         (hval << 1) + (hval << 4) + (hval << 7) + (hval << 8) + (hval << 24);
+#endif
+   }
+
+   /* xor-fold the result to a 24 bit value */
+   hval = (hval >> 24) ^ (hval & MASK_24);
+
+   /* return our new 24 bit hash value */
+   return hval;
+}

src/libbson/src/bson/bson-oid.c

@@ -22,7 +22,6 @@
 #include <string.h>
 
 #include "bson-context-private.h"
-#include "bson-md5.h"
 #include "bson-oid.h"
 #include "bson-string.h"
 
@@ -134,7 +133,7 @@ bson_oid_init_sequence (bson_oid_t *oid,         /* OUT */
  *
  *       Generates bytes for a new bson_oid_t and stores them in @oid. The
  *       bytes will be generated according to the specification and includes
- *       the current time, first 3 bytes of MD5(hostname), pid (or tid), and
+ *       the current time, a 3 byte hash of the hostname, pid (or tid), and
  *       monotonic counter.
  *
  *       The bson_oid_t generated by this function is not guaranteed to be