From 210995b867b0b6c1813d34273322fca64cde6f6e Mon Sep 17 00:00:00 2001
From: Jeffrey Walton <noloader@gmail.com>
Date: Wed, 24 Oct 2018 16:55:59 -0400
Subject: [PATCH] Add XOP aware SIMON and SPECK

---
 simon128-simd.cpp | 16 ++++++++++++++++
 simon64-simd.cpp  | 32 ++++++++++++++++++++++++++++++++
 speck128-simd.cpp | 16 ++++++++++++++++
 speck64-simd.cpp  | 20 ++++++++++++++++++++
 4 files changed, 84 insertions(+)

diff --git a/simon128-simd.cpp b/simon128-simd.cpp
index faa81dcb..ffc7cd0a 100644
--- a/simon128-simd.cpp
+++ b/simon128-simd.cpp
@@ -22,6 +22,10 @@
 # include <tmmintrin.h>
 #endif
 
+#if defined(__XOP__)
+# include <ammintrin.h>
+#endif
+
 #if defined(__AVX512F__) && defined(__AVX512VL__)
 # define CRYPTOPP_AVX512_ROTATE 1
 # include <immintrin.h>
@@ -316,6 +320,8 @@ inline __m128i RotateLeft64(const __m128i& val)
 {
 #if defined(CRYPTOPP_AVX512_ROTATE)
     return _mm_rol_epi64(val, R);
+#elif defined(__XOP__)
+    return _mm_roti_epi64(val, R);
 #else
     return _mm_or_si128(
         _mm_slli_epi64(val, R), _mm_srli_epi64(val, 64-R));
@@ -327,6 +333,8 @@ inline __m128i RotateRight64(const __m128i& val)
 {
 #if defined(CRYPTOPP_AVX512_ROTATE)
     return _mm_ror_epi64(val, R);
+#elif defined(__XOP__)
+    return _mm_roti_epi64(val, 64-R);
 #else
     return _mm_or_si128(
         _mm_slli_epi64(val, 64-R), _mm_srli_epi64(val, R));
@@ -337,16 +345,24 @@ inline __m128i RotateRight64(const __m128i& val)
 template <>
 inline __m128i RotateLeft64<8>(const __m128i& val)
 {
+#if defined(__XOP__)
+    return _mm_roti_epi64(val, 8);
+#else
     const __m128i mask = _mm_set_epi8(14,13,12,11, 10,9,8,15, 6,5,4,3, 2,1,0,7);
     return _mm_shuffle_epi8(val, mask);
+#endif
 }
 
 // Faster than two Shifts and an Or. Thanks to Louis Wingers and Bryan Weeks.
 template <>
 inline __m128i RotateRight64<8>(const __m128i& val)
 {
+#if defined(__XOP__)
+    return _mm_roti_epi64(val, 64-8);
+#else
     const __m128i mask = _mm_set_epi8(8,15,14,13, 12,11,10,9, 0,7,6,5, 4,3,2,1);
     return _mm_shuffle_epi8(val, mask);
+#endif
 }
 
 inline __m128i SIMON128_f(const __m128i& v)
diff --git a/simon64-simd.cpp b/simon64-simd.cpp
index 208d85dd..42444180 100644
--- a/simon64-simd.cpp
+++ b/simon64-simd.cpp
@@ -26,6 +26,10 @@
 # include <smmintrin.h>
 #endif
 
+#if defined(__XOP__)
+# include <ammintrin.h>
+#endif
+
 #if defined(__AVX512F__) && defined(__AVX512VL__)
 # define CRYPTOPP_AVX512_ROTATE 1
 # include <immintrin.h>
@@ -322,6 +326,8 @@ inline __m128i RotateLeft64(const __m128i& val)
 {
 #if defined(CRYPTOPP_AVX512_ROTATE)
     return _mm_rol_epi64(val, R);
+#elif defined(__XOP__)
+    return _mm_roti_epi64(val, R);
 #else
     return _mm_or_si128(
         _mm_slli_epi64(val, R), _mm_srli_epi64(val, 64-R));
@@ -333,6 +339,8 @@ inline __m128i RotateRight64(const __m128i& val)
 {
 #if defined(CRYPTOPP_AVX512_ROTATE)
     return _mm_ror_epi64(val, R);
+#elif defined(__XOP__)
+    return _mm_roti_epi64(val, 64-R);
 #else
     return _mm_or_si128(
         _mm_slli_epi64(val, 64-R), _mm_srli_epi64(val, R));
@@ -343,16 +351,24 @@ inline __m128i RotateRight64(const __m128i& val)
 template <>
 inline __m128i RotateLeft64<8>(const __m128i& val)
 {
+#if defined(__XOP__)
+    return _mm_roti_epi64(val, 8);
+#else
     const __m128i mask = _mm_set_epi8(14,13,12,11, 10,9,8,15, 6,5,4,3, 2,1,0,7);
     return _mm_shuffle_epi8(val, mask);
+#endif
 }
 
 // Faster than two Shifts and an Or. Thanks to Louis Wingers and Bryan Weeks.
 template <>
 inline __m128i RotateRight64<8>(const __m128i& val)
 {
+#if defined(__XOP__)
+    return _mm_roti_epi64(val, 64-8);
+#else
     const __m128i mask = _mm_set_epi8(8,15,14,13, 12,11,10,9, 0,7,6,5, 4,3,2,1);
     return _mm_shuffle_epi8(val, mask);
+#endif
 }
 
 inline __m128i SIMON128_f(const __m128i& v)
@@ -527,31 +543,47 @@ inline void SIMON128_Dec_6_Blocks(__m128i &block0, __m128i &block1,
 template <unsigned int R>
 inline __m128i RotateLeft32(const __m128i& val)
 {
+#if defined(__XOP__)
+    return _mm_roti_epi32(val, R);
+#else
     return _mm_or_si128(
         _mm_slli_epi32(val, R), _mm_srli_epi32(val, 32-R));
+#endif
 }
 
 template <unsigned int R>
 inline __m128i RotateRight32(const __m128i& val)
 {
+#if defined(__XOP__)
+    return _mm_roti_epi32(val, 32-R);
+#else
     return _mm_or_si128(
         _mm_slli_epi32(val, 32-R), _mm_srli_epi32(val, R));
+#endif
 }
 
 // Faster than two Shifts and an Or. Thanks to Louis Wingers and Bryan Weeks.
 template <>
 inline __m128i RotateLeft32<8>(const __m128i& val)
 {
+#if defined(__XOP__)
+    return _mm_roti_epi32(val, 8);
+#else
     const __m128i mask = _mm_set_epi8(14,13,12,15, 10,9,8,11, 6,5,4,7, 2,1,0,3);
     return _mm_shuffle_epi8(val, mask);
+#endif
 }
 
 // Faster than two Shifts and an Or. Thanks to Louis Wingers and Bryan Weeks.
 template <>
 inline __m128i RotateRight32<8>(const __m128i& val)
 {
+#if defined(__XOP__)
+    return _mm_roti_epi32(val, 32-8);
+#else
     const __m128i mask = _mm_set_epi8(12,15,14,13, 8,11,10,9, 4,7,6,5, 0,3,2,1);
     return _mm_shuffle_epi8(val, mask);
+#endif
 }
 
 inline __m128i SIMON64_f(const __m128i& v)
diff --git a/speck128-simd.cpp b/speck128-simd.cpp
index 171640a1..5a503064 100644
--- a/speck128-simd.cpp
+++ b/speck128-simd.cpp
@@ -22,6 +22,10 @@
 # include <tmmintrin.h>
 #endif
 
+#if defined(__XOP__)
+# include <ammintrin.h>
+#endif
+
 #if defined(__AVX512F__) && defined(__AVX512VL__)
 # define CRYPTOPP_AVX512_ROTATE 1
 # include <immintrin.h>
@@ -278,6 +282,8 @@ inline __m128i RotateLeft64(const __m128i& val)
 {
 #if defined(CRYPTOPP_AVX512_ROTATE)
     return _mm_rol_epi64(val, R);
+#elif defined(__XOP__)
+    return _mm_roti_epi64(val, R);
 #else
     return _mm_or_si128(
         _mm_slli_epi64(val, R), _mm_srli_epi64(val, 64-R));
@@ -289,6 +295,8 @@ inline __m128i RotateRight64(const __m128i& val)
 {
 #if defined(CRYPTOPP_AVX512_ROTATE)
     return _mm_ror_epi64(val, R);
+#elif defined(__XOP__)
+    return _mm_roti_epi64(val, 64-R);
 #else
     return _mm_or_si128(
         _mm_slli_epi64(val, 64-R), _mm_srli_epi64(val, R));
@@ -299,16 +307,24 @@ inline __m128i RotateRight64(const __m128i& val)
 template <>
 inline __m128i RotateLeft64<8>(const __m128i& val)
 {
+#if defined(__XOP__)
+    return _mm_roti_epi64(val, 8);
+#else
     const __m128i mask = _mm_set_epi8(14,13,12,11, 10,9,8,15, 6,5,4,3, 2,1,0,7);
     return _mm_shuffle_epi8(val, mask);
+#endif
 }
 
 // Faster than two Shifts and an Or. Thanks to Louis Wingers and Bryan Weeks.
 template <>
 inline __m128i RotateRight64<8>(const __m128i& val)
 {
+#if defined(__XOP__)
+    return _mm_roti_epi64(val, 64-8);
+#else
     const __m128i mask = _mm_set_epi8(8,15,14,13, 12,11,10,9, 0,7,6,5, 4,3,2,1);
     return _mm_shuffle_epi8(val, mask);
+#endif
 }
 
 inline void SPECK128_Enc_Block(__m128i &block0, __m128i &block1,
diff --git a/speck64-simd.cpp b/speck64-simd.cpp
index 79a46bfb..c0e32c5d 100644
--- a/speck64-simd.cpp
+++ b/speck64-simd.cpp
@@ -26,6 +26,10 @@
 # include <smmintrin.h>
 #endif
 
+#if defined(__XOP__)
+# include <ammintrin.h>
+#endif
+
 #if defined(__AVX512F__) && defined(__AVX512VL__)
 # define CRYPTOPP_AVX512_ROTATE 1
 # include <immintrin.h>
@@ -266,31 +270,47 @@ inline void SPECK64_Dec_6_Blocks(uint32x4_t &block0, uint32x4_t &block1,
 template <unsigned int R>
 inline __m128i RotateLeft32(const __m128i& val)
 {
+#if defined(__XOP__)
+    return _mm_roti_epi32(val, R);
+#else
     return _mm_or_si128(
         _mm_slli_epi32(val, R), _mm_srli_epi32(val, 32-R));
+#endif
 }
 
 template <unsigned int R>
 inline __m128i RotateRight32(const __m128i& val)
 {
+#if defined(__XOP__)
+    return _mm_roti_epi32(val, 32-R);
+#else
     return _mm_or_si128(
         _mm_slli_epi32(val, 32-R), _mm_srli_epi32(val, R));
+#endif
 }
 
 // Faster than two Shifts and an Or. Thanks to Louis Wingers and Bryan Weeks.
 template <>
 inline __m128i RotateLeft32<8>(const __m128i& val)
 {
+#if defined(__XOP__)
+    return _mm_roti_epi32(val, 8);
+#else
     const __m128i mask = _mm_set_epi8(14,13,12,15, 10,9,8,11, 6,5,4,7, 2,1,0,3);
     return _mm_shuffle_epi8(val, mask);
+#endif
 }
 
 // Faster than two Shifts and an Or. Thanks to Louis Wingers and Bryan Weeks.
 template <>
 inline __m128i RotateRight32<8>(const __m128i& val)
 {
+#if defined(__XOP__)
+    return _mm_roti_epi32(val, 32-8);
+#else
     const __m128i mask = _mm_set_epi8(12,15,14,13, 8,11,10,9, 4,7,6,5, 0,3,2,1);
     return _mm_shuffle_epi8(val, mask);
+#endif
 }
 
 inline void SPECK64_Enc_Block(__m128i &block0, __m128i &block1,