Use an std::map instead of a custom hash table for MAC address prefix

lookup. The hash table used linear probing which got very slow as the
hash table got full. Using std::map is about 10 times faster. The hash
table was slow enough that it took the majority of the time for me in an
ARP scan of a single address.

# nmap -sP 192.168.0.190
mac_prefix_init took 0.49261 s.
Nmap done: 1 IP address (1 host up) scanned in 0.59 seconds

# nmap -sP 192.168.0.190
mac_prefix_init took 0.04392 s.
Nmap done: 1 IP address (1 host up) scanned in 0.13 seconds

The memory usage of std::map is probably greater. The hash table used
19037 pointers and about 13000 structures of size 8 (on a 32-bit
architecture), or about 176 KB. Assuming the map has left, right, and
parent pointers, and a red-black indicator per node, the usage is 16
bytes per prefix plus 8 bytes for the structure data, or 304 KB total.
But this makes fingerdiff so much faster, I want to leave it in place at
least until this round of OS integration is done.

MACLookup.cc

@@ -91,6 +91,7 @@
 
 /* $Id$ */
 
+#include <map>
 
 /* Character pool memory allocation */
 #include "MACLookup.h"
@@ -101,28 +102,12 @@
 
 extern NmapOps o;
 
-struct MAC_entry {
+std::map<int, char *> MacTable;
-  int prefix; /* -1 means none set */
-  char *vendor;
-};
-
-struct MAC_hash_table {
-  int table_capacity; /* How many members the table can hold */
-  int table_members; /* How many members it has now */
-  struct MAC_entry **table;
-} MacTable;
 
 static inline int MacCharPrefix2Key(const u8 *prefix) {
   return (prefix[0] << 16) + (prefix[1] << 8) + prefix[2];
 }
 
-/* Hashes the prefix into a position from 0 to table_capacity - 1.  Does not
-   check if the position is free or anything */
-static inline int MACTableHash(int prefix, int table_capacity) {
-  // Maybe I should think about changing this sometime.
-  return prefix % table_capacity;
-}
-
 static void mac_prefix_init() {
   static int initialized = 0;
   if (initialized) return;
@@ -130,14 +115,9 @@ static void mac_prefix_init() {
   char filename[256];
   FILE *fp;
   char line[128];
-  int pfx, pos;
+  int pfx;
-  char *endptr, *p;
+  char *endptr, *vendor;
   int lineno = 0;
-  struct MAC_entry *ME;
-
-  MacTable.table_capacity = 19037;
-  MacTable.table_members = 0;
-  MacTable.table = (struct MAC_entry **) safe_zalloc(MacTable.table_capacity * sizeof(struct MAC_entry *));
 
   /* Now it is time to read in all of the entries ... */
   if (nmap_fetchfile(filename, sizeof(filename), "nmap-mac-prefixes") != 1){
@@ -169,26 +149,17 @@ static void mac_prefix_init() {
     /* Now grab the vendor */
     while(*endptr && isspace((int) (unsigned char) *endptr)) endptr++;
     assert(*endptr);
-    p = endptr;
+    vendor = endptr;
     while(*endptr && *endptr != '\n' && *endptr != '\r') endptr++;
     *endptr = '\0';
 
-    // Create the new MAC_entry
+    if (MacTable.find(pfx) == MacTable.end()) {
-    ME = (struct MAC_entry *) cp_alloc(sizeof(struct MAC_entry));
+      MacTable[pfx] = cp_strdup(vendor);
-    ME->prefix = pfx;
+    } else {
-    ME->vendor = cp_strdup(p);
+      if (o.debugging > 1)
+	error("MAC prefix %06X is duplicated in %s; ignoring duplicates.", pfx, filename);
+    }
 
-    if (MacTable.table_members > MacTable.table_capacity * 0.8)
-      error("WARNING:  nmap-mac-prefixes has grown to more than 80%% of our hash table size.  MacTable.table_capacity should be increased");
-
-    // Now insert it into the table
-    if (MacTable.table_members >= MacTable.table_capacity)
-      fatal("No space for further MAC prefixes from nmap-mac-prefixes.  Increase MacTable.table_capacity");
-
-    pos = MACTableHash(pfx, MacTable.table_capacity); 
-    while (MacTable.table[pos]) pos = (pos + 1) % MacTable.table_capacity;
-    MacTable.table[pos] = ME;
-    MacTable.table_members++;
   }
 
   fclose(fp);
@@ -196,16 +167,14 @@ static void mac_prefix_init() {
 }
 
 
-static struct MAC_entry *findMACEntry(int prefix) {
+static const char *findMACEntry(int prefix) {
-  int pos = MACTableHash(prefix, MacTable.table_capacity);
+  std::map<int, char *>::iterator i;
-
-  while (MacTable.table[pos]) {
-    if (MacTable.table[pos]->prefix == prefix)
-      return MacTable.table[pos];
-    pos = (pos + 1) % MacTable.table_capacity;
-  }
 
+  i = MacTable.find(prefix);
+  if (i == MacTable.end())
     return NULL;
+
+  return i->second;
 }
 
 /* Takes a three byte MAC address prefix (passing the whole MAC is OK
@@ -213,13 +182,10 @@ static struct MAC_entry *findMACEntry(int prefix) {
    NULL is returned if no vendor is found for the given prefix or if there
    is some other error. */
 const char *MACPrefix2Corp(const u8 *prefix) {
-  struct MAC_entry *ent;
-
   if (!prefix) fatal("%s called with a NULL prefix", __func__);
   mac_prefix_init();
 
-  ent = findMACEntry(MacCharPrefix2Key(prefix));
+  return findMACEntry(MacCharPrefix2Key(prefix));
-  return (ent)? ent->vendor : NULL;
 }
 
 /* Takes a string and looks through the table for a vendor name which
@@ -229,16 +195,17 @@ const char *MACPrefix2Corp(const u8 *prefix) {
    is not particularly efficient and so should be rewriteen if it is
    called often */
 bool MACCorp2Prefix(const char *vendorstr, u8 *mac_data) {
+  std::map<int, char *>::iterator i;
+
   if (!vendorstr) fatal("%s: vendorstr is NULL", __func__);
   if (!mac_data) fatal("%s: mac_data is NULL", __func__);
   mac_prefix_init();
 
-  for(int i = 0; i < MacTable.table_capacity; i++ ) {
+  for (i = MacTable.begin(); i != MacTable.end(); i++) {
-    if (MacTable.table[i])
+    if (strcasestr(i->second, vendorstr)) {
-      if (strcasestr(MacTable.table[i]->vendor, vendorstr)) {
+      mac_data[0] = i->first >> 16;
-	mac_data[0] = MacTable.table[i]->prefix >> 16;
+      mac_data[1] = (i->first >> 8) & 0xFF;
-	mac_data[1] = (MacTable.table[i]->prefix >> 8) & 0xFF;
+      mac_data[2] = i->first & 0xFF;
-	mac_data[2] = MacTable.table[i]->prefix & 0xFF;
       return true;
     }
   }