+// align_ptr_up() : get the first aligned element of an array.
+// ptr must point to an array of size at least `sizeof(T) * N + alignment` bytes,
+// where N is the number of elements in the array.
+template <uintptr_t Alignment, typename T>
+T* align_ptr_up(T* ptr)
+{
+ static_assert(alignof(T) < Alignment);
+
+ const uintptr_t ptrint = reinterpret_cast<uintptr_t>(reinterpret_cast<char*>(ptr));
+ return reinterpret_cast<T*>(reinterpret_cast<char*>((ptrint + (Alignment - 1)) / Alignment * Alignment));
+}
+
+
+// IsLittleEndian : true if and only if the binary is compiled on a little endian machine
+static inline const union { uint32_t i; char c[4]; } Le = { 0x01020304 };
+static inline const bool IsLittleEndian = (Le.c[0] == 4);
+
+
+// RunningAverage : a class to calculate a running average of a series of values.
+// For efficiency, all computations are done with integers.
+class RunningAverage {
+ public:
+
+ // Reset the running average to rational value p / q
+ void set(int64_t p, int64_t q)
+ { average = p * PERIOD * RESOLUTION / q; }
+
+ // Update average with value v
+ void update(int64_t v)
+ { average = RESOLUTION * v + (PERIOD - 1) * average / PERIOD; }
+
+ // Test if average is strictly greater than rational a / b
+ bool is_greater(int64_t a, int64_t b) const
+ { return b * average > a * (PERIOD * RESOLUTION); }
+
+ int64_t value() const
+ { return average / (PERIOD * RESOLUTION); }
+
+ private :
+ static constexpr int64_t PERIOD = 4096;
+ static constexpr int64_t RESOLUTION = 1024;
+ int64_t average;
+};
+
+template <typename T, std::size_t MaxSize>
+class ValueList {
+
+public:
+ std::size_t size() const { return size_; }
+ void push_back(const T& value) { values_[size_++] = value; }
+ const T* begin() const { return values_; }
+ const T* end() const { return values_ + size_; }
+
+private:
+ T values_[MaxSize];
+ std::size_t size_ = 0;
+};
+
+
+/// xorshift64star Pseudo-Random Number Generator
+/// This class is based on original code written and dedicated
+/// to the public domain by Sebastiano Vigna (2014).
+/// It has the following characteristics:
+///
+/// - Outputs 64-bit numbers
+/// - Passes Dieharder and SmallCrush test batteries
+/// - Does not require warm-up, no zeroland to escape
+/// - Internal state is a single 64-bit integer
+/// - Period is 2^64 - 1
+/// - Speed: 1.60 ns/call (Core i7 @3.40GHz)
+///
+/// For further analysis see
+/// <http://vigna.di.unimi.it/ftp/papers/xorshift.pdf>
+
+class PRNG {
+
+ uint64_t s;
+
+ uint64_t rand64() {
+
+ s ^= s >> 12, s ^= s << 25, s ^= s >> 27;
+ return s * 2685821657736338717LL;
+ }
+
+public:
+ PRNG(uint64_t seed) : s(seed) { assert(seed); }