From 6a1071ccd370adf04fef27f59ecf93ce3c0a7b82 Mon Sep 17 00:00:00 2001
From: Howard Hinnant <howard.hinnant@gmail.com>
Date: Mon, 28 Jul 2014 08:59:16 -0700
Subject: [PATCH] Tidy up hardened_hash: * Added siphash as a HashAlgorithm *
 Refactored class responsibilities

---
 beast/container/Container.unity.cpp          |   1 +
 beast/container/hardened_hash.h              | 174 +++++++++----------
 beast/container/hash_append.h                |  54 +++++-
 beast/container/impl/siphash.cpp             | 166 ++++++++++++++++++
 beast/container/tests/hardened_hash.test.cpp |  12 +-
 beast/crypto/UnsignedInteger.h               |   2 +-
 beast/utility/static_initializer.h           |  36 ++++
 7 files changed, 344 insertions(+), 101 deletions(-)
 create mode 100644 beast/container/impl/siphash.cpp

diff --git a/beast/container/Container.unity.cpp b/beast/container/Container.unity.cpp
index 994129dff..d631d45ea 100644
--- a/beast/container/Container.unity.cpp
+++ b/beast/container/Container.unity.cpp
@@ -22,6 +22,7 @@
 #endif
 
 #include <beast/container/impl/spookyv2.cpp>
+#include <beast/container/impl/siphash.cpp>
 
 #include <beast/container/tests/aged_associative_container.test.cpp>
 #include <beast/container/tests/buffer_view.test.cpp>
diff --git a/beast/container/hardened_hash.h b/beast/container/hardened_hash.h
index 2c632d06a..612fb3c9f 100644
--- a/beast/container/hardened_hash.h
+++ b/beast/container/hardened_hash.h
@@ -21,16 +21,17 @@
 #define BEAST_CONTAINER_HARDENED_HASH_H_INCLUDED
 
 #include <beast/container/hash_append.h>
-
+#include <beast/cxx14/utility.h> // <utility>
+#include <beast/cxx14/type_traits.h> // <type_traits>
 #include <beast/utility/noexcept.h>
+#include <beast/utility/static_initializer.h>
+
 #include <cstdint>
 #include <functional>
 #include <mutex>
 #include <random>
-#include <beast/cxx14/type_traits.h> // <type_traits>
 #include <unordered_map>
 #include <unordered_set>
-#include <beast/cxx14/utility.h> // <utility>
 
 // When set to 1, makes the seed per-process instead
 // of per default-constructed instance of hardened_hash
@@ -44,78 +45,86 @@
 #endif
 
 namespace beast {
-namespace detail {
 
-template <class Result>
-class hardened_hash_base
+using seed_pair = std::pair<std::uint64_t, std::uint64_t>;
+
+template <bool = true>
+seed_pair
+get_seed_pair() noexcept
 {
-public:
-    typedef Result result_type;
+    struct state_t
+    {
+        std::mutex mutex;
+        std::random_device rng;
+        std::mt19937_64 gen {rng()};
+        std::uniform_int_distribution <std::uint64_t> dist;
 
-private:
+        state_t() : gen(rng()) {}
+        // state_t(state_t const&) = delete;
+        // state_t& operator=(state_t const&) = delete;
+    };
+    static static_initializer <state_t> state;
+    std::lock_guard <std::mutex> lock (state->mutex);
+    return {state->dist(state->gen), state->dist(state->gen)};
+}
+
+template <class HashAlgorithm, bool ProcessSeeded>
+class basic_hardened_hash;
+
+/**
+ * Seed functor once per process
+*/
+template <class HashAlgorithm>
+class basic_hardened_hash<HashAlgorithm, true>
+{
     static
+    seed_pair const&
+    init_seed_pair()
+    {
+        static static_initializer <seed_pair, basic_hardened_hash> const
+                                                           p(get_seed_pair<>());
+        return *p;
+    }
+
+public:
+    using result_type = typename HashAlgorithm::result_type;
+
+    template <class T>
     result_type
-    next_seed() noexcept
+    operator()(T const& t) const noexcept
     {
-        static std::mutex mutex;
-        static std::random_device rng;
-        static std::mt19937_64 gen (rng());
-        std::lock_guard <std::mutex> lock (mutex);
-        std::uniform_int_distribution <result_type> dist;
-        result_type value;
-        for(;;)
-        {
-            value = dist (gen);
-            // VFALCO Do we care if 0 is picked?
-            if (value != 0)
-                break;
-        }
-        return value;
+        std::uint64_t seed0;
+        std::uint64_t seed1;
+        std::tie(seed0, seed1) = init_seed_pair();
+        HashAlgorithm h(seed0, seed1);
+        hash_append(h, t);
+        return static_cast<result_type>(h);
     }
-
-#if BEAST_NO_HARDENED_HASH_INSTANCE_SEED
-protected:
-    hardened_hash_base() noexcept = default;
-
-    hardened_hash_base(result_type) noexcept
-    {
-    }
-
-    result_type
-    seed() const noexcept
-    {
-        static result_type const value (next_seed());
-        return value;
-    }
-
-#else
-protected:
-    hardened_hash_base() noexcept
-        : m_seed (next_seed())
-    {
-    }
-
-    hardened_hash_base(result_type seed) noexcept
-        : m_seed (seed)
-    {
-    }
-
-    result_type
-    seed() const noexcept
-    {
-        return m_seed;
-    }
-
-private:
-    // VFALCO Should seed be per process or per hash function?
-    result_type m_seed;
-
-#endif
 };
 
-//------------------------------------------------------------------------------
+/**
+ * Seed functor once per construction
+*/
+template <class HashAlgorithm>
+class basic_hardened_hash<HashAlgorithm, false>
+{
+    seed_pair m_seeds;
+public:
+    using result_type = typename HashAlgorithm::result_type;
 
-} // detail
+    basic_hardened_hash()
+        : m_seeds(get_seed_pair<>())
+    {}
+
+    template <class T>
+    result_type
+    operator()(T const& t) const noexcept
+    {
+        HashAlgorithm h(m_seeds.first, m_seeds.second);
+        hash_append(h, t);
+        return static_cast<result_type>(h);
+    }
+};
 
 //------------------------------------------------------------------------------
 
@@ -140,31 +149,18 @@ private:
     }
 
     @endcode
+
+    Do not use any version of Murmur or CityHash for the Hasher
+    template parameter (the hashing algorithm).  For details
+    see https://131002.net/siphash/#at
 */
-template <class T, class Hasher = detail::spooky_wrapper>
-class hardened_hash
-    : public detail::hardened_hash_base <std::size_t>
-{
-    typedef detail::hardened_hash_base <std::size_t> base;
-public:
-    typedef T argument_type;
-    using detail::hardened_hash_base <std::size_t>::result_type;
-
-public:
-    hardened_hash() = default;
-    explicit hardened_hash(result_type seed)
-        : base (seed)
-    {
-    }
-
-    result_type
-    operator() (argument_type const& key) const noexcept
-    {
-        Hasher h {base::seed()};
-        hash_append (h, key);
-        return static_cast<result_type> (h);
-    }
-};
+#if BEAST_NO_HARDENED_HASH_INSTANCE_SEED
+template <class HashAlgorithm = siphash>
+    using hardened_hash = basic_hardened_hash<HashAlgorithm, true>;
+#else
+template <class HashAlgorithm = siphash>
+    using hardened_hash = basic_hardened_hash<HashAlgorithm, false>;
+#endif
 
 } // beast
 
diff --git a/beast/container/hash_append.h b/beast/container/hash_append.h
index fa5a73454..b8d195154 100644
--- a/beast/container/hash_append.h
+++ b/beast/container/hash_append.h
@@ -656,16 +656,38 @@ hash_append (Hasher& h, T0 const& t0, T1 const& t1, T const& ...t) noexcept
     hash_append (h, t1, t...);
 }
 
-namespace detail
+// See http://www.isthe.com/chongo/tech/comp/fnv/
+class fnv1a
 {
+    std::uint64_t state_ = 14695981039346656037ULL;
+public:
 
-class spooky_wrapper
+    using result_type = std::size_t;
+
+    void
+    append (void const* key, std::size_t len) noexcept
+    {
+        unsigned char const* p = static_cast<unsigned char const*>(key);
+        unsigned char const* const e = p + len;
+        for (; p < e; ++p)
+            state_ = (state_ ^ *p) * 1099511628211ULL;
+    }
+
+    explicit
+    operator std::size_t() noexcept
+    {
+        return static_cast<std::size_t>(state_);
+    }
+};
+
+// See http://burtleburtle.net/bob/hash/spooky.html
+class spooky
 {
     SpookyHash state_;
 public:
     using result_type = std::size_t;
 
-    spooky_wrapper (std::size_t seed1 = 1, std::size_t seed2 = 2) noexcept
+    spooky (std::size_t seed1 = 1, std::size_t seed2 = 2) noexcept
     {
         state_.Init (seed1, seed2);
     }
@@ -685,9 +707,30 @@ public:
     }
 };
 
-} // detail
+// See https://131002.net/siphash/
+class siphash
+{
+    std::uint64_t v0_ = 0x736f6d6570736575ULL;
+    std::uint64_t v1_ = 0x646f72616e646f6dULL;
+    std::uint64_t v2_ = 0x6c7967656e657261ULL;
+    std::uint64_t v3_ = 0x7465646279746573ULL;
+    unsigned char buf_[8];
+    unsigned bufsize_ = 0;
+    unsigned total_length_ = 0;
+public:
+    using result_type = std::size_t;
 
-template <class Hasher = detail::spooky_wrapper>
+    siphash() = default;
+    explicit siphash(std::uint64_t k0, std::uint64_t k1 = 0) noexcept;
+
+    void
+    append (void const* key, std::size_t len) noexcept;
+
+    explicit
+    operator std::size_t() noexcept;
+};
+
+template <class Hasher = spooky>
 struct uhash
 {
     using result_type = typename Hasher::result_type;
@@ -702,6 +745,7 @@ struct uhash
     }
 };
 
+
 } // beast
 
 #endif
diff --git a/beast/container/impl/siphash.cpp b/beast/container/impl/siphash.cpp
new file mode 100644
index 000000000..ef5181b4d
--- /dev/null
+++ b/beast/container/impl/siphash.cpp
@@ -0,0 +1,166 @@
+//------------------------------- siphash.h ------------------------------------
+// 
+// This software is in the public domain.  The only restriction on its use is
+// that no one can remove it from the public domain by claiming ownership of it,
+// including the original authors.
+// 
+// There is no warranty of correctness on the software contained herein.  Use
+// at your own risk.
+//
+// Derived from:
+// 
+// SipHash reference C implementation
+// 
+// Written in 2012 by Jean-Philippe Aumasson <jeanphilippe.aumasson@gmail.com>
+// Daniel J. Bernstein <djb@cr.yp.to>
+// 
+// To the extent possible under law, the author(s) have dedicated all copyright
+// and related and neighboring rights to this software to the public domain
+// worldwide. This software is distributed without any warranty.
+// 
+// You should have received a copy of the CC0 Public Domain Dedication along
+// with this software. If not, see
+// <http://creativecommons.org/publicdomain/zero/1.0/>.
+// 
+//------------------------------------------------------------------------------
+
+#include <beast/container/hash_append.h>
+#include <algorithm>
+#include <cstddef>
+#include <cstdint>
+
+// namespace acme is used to demonstrate example code.  It is not proposed.
+
+namespace beast
+{
+
+namespace
+{
+
+typedef std::uint64_t u64;
+typedef std::uint32_t u32;
+typedef std::uint8_t u8;
+
+inline
+u64
+rotl(u64 x, u64 b)
+{
+    return (x << b) | (x >> (64 - b));
+}
+
+inline
+u64
+u8to64_le(const u8* p)
+{
+#if BEAST_LITTLE_ENDIAN
+    return *static_cast<u64 const*>(static_cast<void const*>(p));
+#else
+    return static_cast<u64>(p[7]) << 56 | static_cast<u64>(p[6]) << 48 |
+           static_cast<u64>(p[5]) << 40 | static_cast<u64>(p[4]) << 32 |
+           static_cast<u64>(p[3]) << 24 | static_cast<u64>(p[2]) << 16 |
+           static_cast<u64>(p[1]) <<  8 | static_cast<u64>(p[0]);
+#endif
+}
+
+inline
+void
+sipround(u64& v0, u64& v1, u64& v2, u64& v3)
+{
+    v0 += v1;
+    v1 = rotl(v1, 13);
+    v1 ^= v0;
+    v0 = rotl(v0, 32);
+    v2 += v3;
+    v3 = rotl(v3, 16);
+    v3 ^= v2;
+    v0 += v3;
+    v3 = rotl(v3, 21);
+    v3 ^= v0;
+    v2 += v1;
+    v1 = rotl(v1, 17);
+    v1 ^= v2;
+    v2 = rotl(v2, 32);
+}
+
+}  // unnamed
+
+siphash::siphash(std::uint64_t k0, std::uint64_t k1) noexcept
+{
+    v3_ ^= k1;
+    v2_ ^= k0;
+    v1_ ^= k1;
+    v0_ ^= k0;
+}
+
+void
+siphash::append (void const* key, std::size_t inlen) noexcept
+{
+    u8 const* in = static_cast<const u8*>(key);
+    total_length_ += inlen;
+    if (bufsize_ + inlen < 8)
+    {
+        std::copy(in, in+inlen, buf_ + bufsize_);
+        bufsize_ += inlen;
+        return;
+    }
+    if (bufsize_ > 0)
+    {
+        auto t = 8 - bufsize_;
+        std::copy(in, in+t, buf_ + bufsize_);
+        u64 m = u8to64_le( buf_ );
+        v3_ ^= m;
+        sipround(v0_, v1_, v2_, v3_);
+        sipround(v0_, v1_, v2_, v3_);
+        v0_ ^= m;
+        in += t;
+        inlen -= t;
+    }
+    bufsize_ = inlen & 7;
+    u8 const* const end = in + (inlen - bufsize_);
+    for ( ; in != end; in += 8 )
+    {
+        u64 m = u8to64_le( in );
+        v3_ ^= m;
+        sipround(v0_, v1_, v2_, v3_);
+        sipround(v0_, v1_, v2_, v3_);
+        v0_ ^= m;
+    }
+    std::copy(end, end + bufsize_, buf_);
+}
+
+siphash::operator std::size_t() noexcept
+{
+    std::size_t b = static_cast<u64>(total_length_) << 56;
+    switch(bufsize_)
+    {
+    case 7:
+        b |= static_cast<u64>(buf_[6]) << 48;
+    case 6:
+        b |= static_cast<u64>(buf_[5]) << 40;
+    case 5:
+        b |= static_cast<u64>(buf_[4]) << 32;
+    case 4:
+        b |= static_cast<u64>(buf_[3]) << 24;
+    case 3:
+        b |= static_cast<u64>(buf_[2]) << 16;
+    case 2:
+        b |= static_cast<u64>(buf_[1]) << 8;
+    case 1:
+        b |= static_cast<u64>(buf_[0]);
+    case 0:
+        break;
+    }
+    v3_ ^= b;
+    sipround(v0_, v1_, v2_, v3_);
+    sipround(v0_, v1_, v2_, v3_);
+    v0_ ^= b;
+    v2_ ^= 0xff;
+    sipround(v0_, v1_, v2_, v3_);
+    sipround(v0_, v1_, v2_, v3_);
+    sipround(v0_, v1_, v2_, v3_);
+    sipround(v0_, v1_, v2_, v3_);
+    b = v0_ ^ v1_ ^ v2_  ^ v3_;
+    return b;
+}
+
+}  // beast
diff --git a/beast/container/tests/hardened_hash.test.cpp b/beast/container/tests/hardened_hash.test.cpp
index f60f18bed..a90aaa1ee 100644
--- a/beast/container/tests/hardened_hash.test.cpp
+++ b/beast/container/tests/hardened_hash.test.cpp
@@ -90,19 +90,19 @@ namespace detail {
 
 template <class T>
 using test_hardened_unordered_set =
-    std::unordered_set <T, hardened_hash <T>>;
+    std::unordered_set <T, hardened_hash <>>;
 
 template <class T>
 using test_hardened_unordered_map =
-    std::unordered_map <T, int, hardened_hash <T>>;
+    std::unordered_map <T, int, hardened_hash <>>;
 
 template <class T>
 using test_hardened_unordered_multiset =
-    std::unordered_multiset <T, hardened_hash <T>>;
+    std::unordered_multiset <T, hardened_hash <>>;
 
 template <class T>
 using test_hardened_unordered_multimap =
-    std::unordered_multimap <T, int, hardened_hash <T>>;
+    std::unordered_multimap <T, int, hardened_hash <>>;
 
 } // beast
 
@@ -196,7 +196,7 @@ public:
     check ()
     {
         T t{};
-        hardened_hash <T>() (t);
+        hardened_hash <>() (t);
         pass();
     }
 
@@ -280,7 +280,7 @@ public:
         log <<
             "sizeof(std::size_t) == " << sizeof(std::size_t);
 
-        hardened_hash <sha256_t> h;
+        hardened_hash <> h;
         for (int i = 0; i < 100; ++i)
         {
             sha256_t v (sha256_t::from_number (i));
diff --git a/beast/crypto/UnsignedInteger.h b/beast/crypto/UnsignedInteger.h
index 99346be6b..de2adbfd4 100644
--- a/beast/crypto/UnsignedInteger.h
+++ b/beast/crypto/UnsignedInteger.h
@@ -65,7 +65,7 @@ public:
         attackers from exploiting crafted inputs to produce degenerate
         containers.
     */
-    typedef hardened_hash <UnsignedInteger> hasher;
+    typedef hardened_hash <> hasher;
 
     /** Determins if two UnsignedInteger objects are equal. */
     class equal
diff --git a/beast/utility/static_initializer.h b/beast/utility/static_initializer.h
index 2033dd39c..237af40b4 100644
--- a/beast/utility/static_initializer.h
+++ b/beast/utility/static_initializer.h
@@ -83,17 +83,35 @@ public:
     T&
     get() noexcept;
 
+    T const&
+    get() const noexcept
+    {
+        return const_cast<static_initializer&>(*this).get();
+    }
+
     T&
     operator*() noexcept
     {
         return get();
     }
 
+    T const&
+    operator*() const noexcept
+    {
+        return get();
+    }
+
     T*
     operator->() noexcept
     {
         return &get();
     }
+
+    T const*
+    operator->() const noexcept
+    {
+        return &get();
+    }
 };
 
 //------------------------------------------------------------------------------
@@ -187,17 +205,35 @@ public:
         return *instance_;
     }
 
+    T const&
+    get() const noexcept
+    {
+        return const_cast<static_initializer&>(*this).get();
+    }
+
     T&
     operator*() noexcept
     {
         return get();
     }
 
+    T const&
+    operator*() const noexcept
+    {
+        return get();
+    }
+
     T*
     operator->() noexcept
     {
         return &get();
     }
+
+    T const*
+    operator->() const noexcept
+    {
+        return &get();
+    }
 };
 
 template <class T, class Tag>