StateInfo is usually allocated on the stack by search()

And passed in do_move(), this ensures maximum efficiency and
speed and at the same time unlimited move numbers.

The draw back is that to handle Position init we need to
reserve a StateInfo inside Position itself and use at
init time and when copying from another Position.

After lazy SMP we don't need anymore this gimmick and we can
get rid of this special case and always pass an external
StateInfo to Position object.

Also rewritten and simplified Position constructors.

Verified it does not regress with a 3 threads SMP test:
ELO: -0.00 +-12.7 (95%) LOS: 50.0%
Total: 1000 W: 173 L: 173 D: 654

No functional change.

src/benchmark.cpp

@@ -144,10 +144,12 @@ void benchmark(const Position& current, istream& is) {
 
   uint64_t nodes = 0;
   TimePoint elapsed = now();
+  Position pos;
 
   for (size_t i = 0; i < fens.size(); ++i)
   {
-      Position pos(fens[i], Options["UCI_Chess960"], Threads.main());
+      StateListPtr states(new std::vector<StateInfo>(1));
+      pos.set(fens[i], Options["UCI_Chess960"], &states->back(), Threads.main());
 
       cerr << "\nPosition: " << i + 1 << '/' << fens.size() << endl;
 
@@ -156,9 +158,8 @@ void benchmark(const Position& current, istream& is) {
 
       else
       {
-          Search::StateStackPtr st;
           limits.startTime = now();
-          Threads.start_thinking(pos, limits, st);
+          Threads.start_thinking(pos, states, limits);
           Threads.main()->wait_for_search_finished();
           nodes += Threads.nodes_searched();
       }