crypto-native: calculate ghash using vpclmulqdq instructions

vpclmulqdq is introduced on intel icelake architecture and
allows computing 4 carry-less multiplications in paralled by using
512-bit SIMD registers

Type: feature
Change-Id: Idb09d6f51ba6f116bba11649b2d99f649356d449
Signed-off-by: Damjan Marion <damjan.marion@gmail.com>
diff --git a/src/vppinfra/vector_avx512.h b/src/vppinfra/vector_avx512.h
index e27c7c0..efdc78b 100644
--- a/src/vppinfra/vector_avx512.h
+++ b/src/vppinfra/vector_avx512.h
@@ -113,6 +113,18 @@
   return (u32x8) _mm512_extracti64x4_epi64 ((__m512i) v, 1);
 }
 
+static_always_inline u8x32
+u8x64_extract_lo (u8x64 v)
+{
+  return (u8x32) _mm512_extracti64x4_epi64 ((__m512i) v, 0);
+}
+
+static_always_inline u8x32
+u8x64_extract_hi (u8x64 v)
+{
+  return (u8x32) _mm512_extracti64x4_epi64 ((__m512i) v, 1);
+}
+
 static_always_inline u32
 u32x16_min_scalar (u32x16 v)
 {
@@ -149,6 +161,9 @@
 #define u8x64_extract_u8x16(a, n) \
   (u8x16) _mm512_extracti64x2_epi64 ((__m512i) (a), n)
 
+#define u8x64_word_shift_left(a,n)  (u8x64) _mm512_bslli_epi128((__m512i) a, n)
+#define u8x64_word_shift_right(a,n) (u8x64) _mm512_bsrli_epi128((__m512i) a, n)
+
 static_always_inline u8x64
 u8x64_xor3 (u8x64 a, u8x64 b, u8x64 c)
 {