GCM and CTR optimizations

In CTR, special case for counter widths of special interest.

In GHASH, uses a 4x reduction technique suggested by Intel.

Split out GHASH to its own source file and header.

With these changes GCM is over twice as fast on Skylake and about
50% faster on Westmere.

1 files changed, 9 insertions, 0 deletions

diff --git a/src/lib/mac/gmac/gmac.cpp b/src/lib/mac/gmac/gmac.cpp
index be27aba4a..a4e84f57b 100644
--- a/src/lib/mac/gmac/gmac.cpp
+++ b/src/lib/mac/gmac/gmac.cpp
@@ -7,6 +7,8 @@
  */
 
 #include <botan/gmac.h>
+#include <botan/ghash.h>
+#include <botan/block_cipher.h>
 
 namespace Botan {
 
@@ -28,6 +30,13 @@ void GMAC::clear()
    m_initialized = false;
    }
 
+GMAC::~GMAC() { /* for unique_ptr */ }
+
+Key_Length_Specification GMAC::key_spec() const
+   {
+   return m_cipher->key_spec();
+   }
+
 std::string GMAC::name() const
    {
    return "GMAC(" + m_cipher->name() + ")";