rbnacl-libsodium 1.0.5 → 1.0.6

data/vendor/libsodium/src/libsodium/crypto_generichash/blake2/ref/blake2b-load-sse41.h

@@ -0,0 +1,402 @@
+/*
+   BLAKE2 reference source code package - optimized C implementations
+
+   Written in 2012 by Samuel Neves <sneves@dei.uc.pt>
+
+   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/>.
+*/
+
+#ifndef blake2b_load_sse41_H
+#define blake2b_load_sse41_H
+
+#define LOAD_MSG_0_1(b0, b1) \
+do \
+{ \
+b0 = _mm_unpacklo_epi64(m0, m1); \
+b1 = _mm_unpacklo_epi64(m2, m3); \
+} while(0)
+
+
+#define LOAD_MSG_0_2(b0, b1) \
+do \
+{ \
+b0 = _mm_unpackhi_epi64(m0, m1); \
+b1 = _mm_unpackhi_epi64(m2, m3); \
+} while(0)
+
+
+#define LOAD_MSG_0_3(b0, b1) \
+do \
+{ \
+b0 = _mm_unpacklo_epi64(m4, m5); \
+b1 = _mm_unpacklo_epi64(m6, m7); \
+} while(0)
+
+
+#define LOAD_MSG_0_4(b0, b1) \
+do \
+{ \
+b0 = _mm_unpackhi_epi64(m4, m5); \
+b1 = _mm_unpackhi_epi64(m6, m7); \
+} while(0)
+
+
+#define LOAD_MSG_1_1(b0, b1) \
+do \
+{ \
+b0 = _mm_unpacklo_epi64(m7, m2); \
+b1 = _mm_unpackhi_epi64(m4, m6); \
+} while(0)
+
+
+#define LOAD_MSG_1_2(b0, b1) \
+do \
+{ \
+b0 = _mm_unpacklo_epi64(m5, m4); \
+b1 = _mm_alignr_epi8(m3, m7, 8); \
+} while(0)
+
+
+#define LOAD_MSG_1_3(b0, b1) \
+do \
+{ \
+b0 = _mm_shuffle_epi32(m0, _MM_SHUFFLE(1,0,3,2)); \
+b1 = _mm_unpackhi_epi64(m5, m2); \
+} while(0)
+
+
+#define LOAD_MSG_1_4(b0, b1) \
+do \
+{ \
+b0 = _mm_unpacklo_epi64(m6, m1); \
+b1 = _mm_unpackhi_epi64(m3, m1); \
+} while(0)
+
+
+#define LOAD_MSG_2_1(b0, b1) \
+do \
+{ \
+b0 = _mm_alignr_epi8(m6, m5, 8); \
+b1 = _mm_unpackhi_epi64(m2, m7); \
+} while(0)
+
+
+#define LOAD_MSG_2_2(b0, b1) \
+do \
+{ \
+b0 = _mm_unpacklo_epi64(m4, m0); \
+b1 = _mm_blend_epi16(m1, m6, 0xF0); \
+} while(0)
+
+
+#define LOAD_MSG_2_3(b0, b1) \
+do \
+{ \
+b0 = _mm_blend_epi16(m5, m1, 0xF0); \
+b1 = _mm_unpackhi_epi64(m3, m4); \
+} while(0)
+
+
+#define LOAD_MSG_2_4(b0, b1) \
+do \
+{ \
+b0 = _mm_unpacklo_epi64(m7, m3); \
+b1 = _mm_alignr_epi8(m2, m0, 8); \
+} while(0)
+
+
+#define LOAD_MSG_3_1(b0, b1) \
+do \
+{ \
+b0 = _mm_unpackhi_epi64(m3, m1); \
+b1 = _mm_unpackhi_epi64(m6, m5); \
+} while(0)
+
+
+#define LOAD_MSG_3_2(b0, b1) \
+do \
+{ \
+b0 = _mm_unpackhi_epi64(m4, m0); \
+b1 = _mm_unpacklo_epi64(m6, m7); \
+} while(0)
+
+
+#define LOAD_MSG_3_3(b0, b1) \
+do \
+{ \
+b0 = _mm_blend_epi16(m1, m2, 0xF0); \
+b1 = _mm_blend_epi16(m2, m7, 0xF0); \
+} while(0)
+
+
+#define LOAD_MSG_3_4(b0, b1) \
+do \
+{ \
+b0 = _mm_unpacklo_epi64(m3, m5); \
+b1 = _mm_unpacklo_epi64(m0, m4); \
+} while(0)
+
+
+#define LOAD_MSG_4_1(b0, b1) \
+do \
+{ \
+b0 = _mm_unpackhi_epi64(m4, m2); \
+b1 = _mm_unpacklo_epi64(m1, m5); \
+} while(0)
+
+
+#define LOAD_MSG_4_2(b0, b1) \
+do \
+{ \
+b0 = _mm_blend_epi16(m0, m3, 0xF0); \
+b1 = _mm_blend_epi16(m2, m7, 0xF0); \
+} while(0)
+
+
+#define LOAD_MSG_4_3(b0, b1) \
+do \
+{ \
+b0 = _mm_blend_epi16(m7, m5, 0xF0); \
+b1 = _mm_blend_epi16(m3, m1, 0xF0); \
+} while(0)
+
+
+#define LOAD_MSG_4_4(b0, b1) \
+do \
+{ \
+b0 = _mm_alignr_epi8(m6, m0, 8); \
+b1 = _mm_blend_epi16(m4, m6, 0xF0); \
+} while(0)
+
+
+#define LOAD_MSG_5_1(b0, b1) \
+do \
+{ \
+b0 = _mm_unpacklo_epi64(m1, m3); \
+b1 = _mm_unpacklo_epi64(m0, m4); \
+} while(0)
+
+
+#define LOAD_MSG_5_2(b0, b1) \
+do \
+{ \
+b0 = _mm_unpacklo_epi64(m6, m5); \
+b1 = _mm_unpackhi_epi64(m5, m1); \
+} while(0)
+
+
+#define LOAD_MSG_5_3(b0, b1) \
+do \
+{ \
+b0 = _mm_blend_epi16(m2, m3, 0xF0); \
+b1 = _mm_unpackhi_epi64(m7, m0); \
+} while(0)
+
+
+#define LOAD_MSG_5_4(b0, b1) \
+do \
+{ \
+b0 = _mm_unpackhi_epi64(m6, m2); \
+b1 = _mm_blend_epi16(m7, m4, 0xF0); \
+} while(0)
+
+
+#define LOAD_MSG_6_1(b0, b1) \
+do \
+{ \
+b0 = _mm_blend_epi16(m6, m0, 0xF0); \
+b1 = _mm_unpacklo_epi64(m7, m2); \
+} while(0)
+
+
+#define LOAD_MSG_6_2(b0, b1) \
+do \
+{ \
+b0 = _mm_unpackhi_epi64(m2, m7); \
+b1 = _mm_alignr_epi8(m5, m6, 8); \
+} while(0)
+
+
+#define LOAD_MSG_6_3(b0, b1) \
+do \
+{ \
+b0 = _mm_unpacklo_epi64(m0, m3); \
+b1 = _mm_shuffle_epi32(m4, _MM_SHUFFLE(1,0,3,2)); \
+} while(0)
+
+
+#define LOAD_MSG_6_4(b0, b1) \
+do \
+{ \
+b0 = _mm_unpackhi_epi64(m3, m1); \
+b1 = _mm_blend_epi16(m1, m5, 0xF0); \
+} while(0)
+
+
+#define LOAD_MSG_7_1(b0, b1) \
+do \
+{ \
+b0 = _mm_unpackhi_epi64(m6, m3); \
+b1 = _mm_blend_epi16(m6, m1, 0xF0); \
+} while(0)
+
+
+#define LOAD_MSG_7_2(b0, b1) \
+do \
+{ \
+b0 = _mm_alignr_epi8(m7, m5, 8); \
+b1 = _mm_unpackhi_epi64(m0, m4); \
+} while(0)
+
+
+#define LOAD_MSG_7_3(b0, b1) \
+do \
+{ \
+b0 = _mm_unpackhi_epi64(m2, m7); \
+b1 = _mm_unpacklo_epi64(m4, m1); \
+} while(0)
+
+
+#define LOAD_MSG_7_4(b0, b1) \
+do \
+{ \
+b0 = _mm_unpacklo_epi64(m0, m2); \
+b1 = _mm_unpacklo_epi64(m3, m5); \
+} while(0)
+
+
+#define LOAD_MSG_8_1(b0, b1) \
+do \
+{ \
+b0 = _mm_unpacklo_epi64(m3, m7); \
+b1 = _mm_alignr_epi8(m0, m5, 8); \
+} while(0)
+
+
+#define LOAD_MSG_8_2(b0, b1) \
+do \
+{ \
+b0 = _mm_unpackhi_epi64(m7, m4); \
+b1 = _mm_alignr_epi8(m4, m1, 8); \
+} while(0)
+
+
+#define LOAD_MSG_8_3(b0, b1) \
+do \
+{ \
+b0 = m6; \
+b1 = _mm_alignr_epi8(m5, m0, 8); \
+} while(0)
+
+
+#define LOAD_MSG_8_4(b0, b1) \
+do \
+{ \
+b0 = _mm_blend_epi16(m1, m3, 0xF0); \
+b1 = m2; \
+} while(0)
+
+
+#define LOAD_MSG_9_1(b0, b1) \
+do \
+{ \
+b0 = _mm_unpacklo_epi64(m5, m4); \
+b1 = _mm_unpackhi_epi64(m3, m0); \
+} while(0)
+
+
+#define LOAD_MSG_9_2(b0, b1) \
+do \
+{ \
+b0 = _mm_unpacklo_epi64(m1, m2); \
+b1 = _mm_blend_epi16(m3, m2, 0xF0); \
+} while(0)
+
+
+#define LOAD_MSG_9_3(b0, b1) \
+do \
+{ \
+b0 = _mm_unpackhi_epi64(m7, m4); \
+b1 = _mm_unpackhi_epi64(m1, m6); \
+} while(0)
+
+
+#define LOAD_MSG_9_4(b0, b1) \
+do \
+{ \
+b0 = _mm_alignr_epi8(m7, m5, 8); \
+b1 = _mm_unpacklo_epi64(m6, m0); \
+} while(0)
+
+
+#define LOAD_MSG_10_1(b0, b1) \
+do \
+{ \
+b0 = _mm_unpacklo_epi64(m0, m1); \
+b1 = _mm_unpacklo_epi64(m2, m3); \
+} while(0)
+
+
+#define LOAD_MSG_10_2(b0, b1) \
+do \
+{ \
+b0 = _mm_unpackhi_epi64(m0, m1); \
+b1 = _mm_unpackhi_epi64(m2, m3); \
+} while(0)
+
+
+#define LOAD_MSG_10_3(b0, b1) \
+do \
+{ \
+b0 = _mm_unpacklo_epi64(m4, m5); \
+b1 = _mm_unpacklo_epi64(m6, m7); \
+} while(0)
+
+
+#define LOAD_MSG_10_4(b0, b1) \
+do \
+{ \
+b0 = _mm_unpackhi_epi64(m4, m5); \
+b1 = _mm_unpackhi_epi64(m6, m7); \
+} while(0)
+
+
+#define LOAD_MSG_11_1(b0, b1) \
+do \
+{ \
+b0 = _mm_unpacklo_epi64(m7, m2); \
+b1 = _mm_unpackhi_epi64(m4, m6); \
+} while(0)
+
+
+#define LOAD_MSG_11_2(b0, b1) \
+do \
+{ \
+b0 = _mm_unpacklo_epi64(m5, m4); \
+b1 = _mm_alignr_epi8(m3, m7, 8); \
+} while(0)
+
+
+#define LOAD_MSG_11_3(b0, b1) \
+do \
+{ \
+b0 = _mm_shuffle_epi32(m0, _MM_SHUFFLE(1,0,3,2)); \
+b1 = _mm_unpackhi_epi64(m5, m2); \
+} while(0)
+
+
+#define LOAD_MSG_11_4(b0, b1) \
+do \
+{ \
+b0 = _mm_unpacklo_epi64(m6, m1); \
+b1 = _mm_unpackhi_epi64(m3, m1); \
+} while(0)
+
+
+#endif
+

data/vendor/libsodium/src/libsodium/crypto_generichash/blake2/ref/blake2b-ref.c

@@ -1,5 +1,5 @@
 /*
-   BLAKE2 reference source code package - reference C implementations
+   BLAKE2 reference source code package - C implementations
 
    Written in 2012 by Samuel Neves <sneves@dei.uc.pt>
 
@@ -12,11 +12,14 @@
 */
 
 #include <stdint.h>
+#include <stdlib.h>
 #include <string.h>
-#include <stdio.h>
 
 #include "blake2.h"
 #include "blake2-impl.h"
+#include "runtime.h"
+
+static blake2b_compress_fn blake2b_compress = blake2b_compress_ref;
 
 static const uint64_t blake2b_IV[8] =
 {
@@ -26,23 +29,6 @@ static const uint64_t blake2b_IV[8] =
   0x1f83d9abfb41bd6bULL, 0x5be0cd19137e2179ULL
 };
 
-static const uint8_t blake2b_sigma[12][16] =
-{
-  {  0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15 } ,
-  { 14, 10,  4,  8,  9, 15, 13,  6,  1, 12,  0,  2, 11,  7,  5,  3 } ,
-  { 11,  8, 12,  0,  5,  2, 15, 13, 10, 14,  3,  6,  7,  1,  9,  4 } ,
-  {  7,  9,  3,  1, 13, 12, 11, 14,  2,  6,  5, 10,  4,  0, 15,  8 } ,
-  {  9,  0,  5,  7,  2,  4, 10, 15, 14,  1, 11, 12,  6,  8,  3, 13 } ,
-  {  2, 12,  6, 10,  0, 11,  8,  3,  4, 13,  7,  5, 15, 14,  1,  9 } ,
-  { 12,  5,  1, 15, 14, 13,  4, 10,  0,  7,  6,  3,  9,  2,  8, 11 } ,
-  { 13, 11,  7, 14, 12,  1,  3,  9,  5,  0, 15,  4,  8,  6,  2, 10 } ,
-  {  6, 15, 14,  9, 11,  3,  0,  8, 12,  2, 13,  7,  1,  4, 10,  5 } ,
-  { 10,  2,  8,  4,  7,  6,  1,  5, 15, 11,  9, 14,  3, 12, 13 , 0 } ,
-  {  0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15 } ,
-  { 14, 10,  4,  8,  9, 15, 13,  6,  1, 12,  0,  2, 11,  7,  5,  3 }
-};
-
-
 /* LCOV_EXCL_START */
 static inline int blake2b_set_lastnode( blake2b_state *S )
 {
@@ -57,7 +43,7 @@ static inline int blake2b_clear_lastnode( blake2b_state *S )
   return 0;
 }
 #endif
-/* Some helper functions, not necessarily useful */
+
 static inline int blake2b_set_lastblock( blake2b_state *S )
 {
   if( S->last_node ) blake2b_set_lastnode( S );
@@ -76,13 +62,18 @@ static inline int blake2b_clear_lastblock( blake2b_state *S )
 #endif
 static inline int blake2b_increment_counter( blake2b_state *S, const uint64_t inc )
 {
+#if defined(__x86_64__) && defined(__SIZEOF_INT128__)
+  __uint128_t t = ( ( __uint128_t )S->t[1] << 64 ) | S->t[0];
+  t += inc;
+  S->t[0] = ( uint64_t )( t >>  0 );
+  S->t[1] = ( uint64_t )( t >> 64 );
+#else
   S->t[0] += inc;
   S->t[1] += ( S->t[0] < inc );
+#endif
   return 0;
 }
 
-
-
 // Parameter-related functions
 #if 0
 static inline int blake2b_param_set_digest_length( blake2b_param *P, const uint8_t digest_length )
@@ -165,8 +156,6 @@ int blake2b_init_param( blake2b_state *S, const blake2b_param *P )
   return 0;
 }
 
-
-
 int blake2b_init( blake2b_state *S, const uint8_t outlen )
 {
   blake2b_param P[1];
@@ -289,69 +278,6 @@ int blake2b_init_key_salt_personal( blake2b_state *S, const uint8_t outlen, cons
   return 0;
 }
 
-static int blake2b_compress( blake2b_state *S, const uint8_t block[BLAKE2B_BLOCKBYTES] )
-{
-  uint64_t m[16];
-  uint64_t v[16];
-  int      i;
-
-  for( i = 0; i < 16; ++i )
-    m[i] = load64( block + i * sizeof( m[i] ) );
-
-  for( i = 0; i < 8; ++i )
-    v[i] = S->h[i];
-
-  v[ 8] = blake2b_IV[0];
-  v[ 9] = blake2b_IV[1];
-  v[10] = blake2b_IV[2];
-  v[11] = blake2b_IV[3];
-  v[12] = S->t[0] ^ blake2b_IV[4];
-  v[13] = S->t[1] ^ blake2b_IV[5];
-  v[14] = S->f[0] ^ blake2b_IV[6];
-  v[15] = S->f[1] ^ blake2b_IV[7];
-#define G(r,i,a,b,c,d) \
-  do { \
-    a = a + b + m[blake2b_sigma[r][2*i+0]]; \
-    d = rotr64(d ^ a, 32); \
-    c = c + d; \
-    b = rotr64(b ^ c, 24); \
-    a = a + b + m[blake2b_sigma[r][2*i+1]]; \
-    d = rotr64(d ^ a, 16); \
-    c = c + d; \
-    b = rotr64(b ^ c, 63); \
-  } while(0)
-#define ROUND(r)  \
-  do { \
-    G(r,0,v[ 0],v[ 4],v[ 8],v[12]); \
-    G(r,1,v[ 1],v[ 5],v[ 9],v[13]); \
-    G(r,2,v[ 2],v[ 6],v[10],v[14]); \
-    G(r,3,v[ 3],v[ 7],v[11],v[15]); \
-    G(r,4,v[ 0],v[ 5],v[10],v[15]); \
-    G(r,5,v[ 1],v[ 6],v[11],v[12]); \
-    G(r,6,v[ 2],v[ 7],v[ 8],v[13]); \
-    G(r,7,v[ 3],v[ 4],v[ 9],v[14]); \
-  } while(0)
-  ROUND( 0 );
-  ROUND( 1 );
-  ROUND( 2 );
-  ROUND( 3 );
-  ROUND( 4 );
-  ROUND( 5 );
-  ROUND( 6 );
-  ROUND( 7 );
-  ROUND( 8 );
-  ROUND( 9 );
-  ROUND( 10 );
-  ROUND( 11 );
-
-  for( i = 0; i < 8; ++i )
-    S->h[i] = S->h[i] ^ v[i] ^ v[i + 8];
-
-#undef G
-#undef ROUND
-  return 0;
-}
-
 /* inlen now in bytes */
 int blake2b_update( blake2b_state *S, const uint8_t *in, uint64_t inlen )
 {
@@ -383,13 +309,9 @@ int blake2b_update( blake2b_state *S, const uint8_t *in, uint64_t inlen )
   return 0;
 }
 
-/* Is this correct? */
 int blake2b_final( blake2b_state *S, uint8_t *out, uint8_t outlen )
 {
-  uint8_t buffer[BLAKE2B_OUTBYTES];
-  int     i;
-
-  if( outlen > BLAKE2B_OUTBYTES ) {
+  if( !outlen || outlen > BLAKE2B_OUTBYTES ) {
     return -1;
   }
   if( S->buflen > BLAKE2B_BLOCKBYTES )
@@ -405,10 +327,18 @@ int blake2b_final( blake2b_state *S, uint8_t *out, uint8_t outlen )
   memset( S->buf + S->buflen, 0, 2 * BLAKE2B_BLOCKBYTES - S->buflen ); /* Padding */
   blake2b_compress( S, S->buf );
 
-  for( i = 0; i < 8; ++i ) /* Output full hash to temp buffer */
-    store64( buffer + sizeof( S->h[i] ) * i, S->h[i] );
+#ifdef NATIVE_LITTLE_ENDIAN
+  memcpy( out, &S->h[0], outlen );
+#else
+  {
+    uint8_t buffer[BLAKE2B_OUTBYTES];
+    int     i;
 
-  memcpy( out, buffer, outlen );
+    for( i = 0; i < 8; ++i ) /* Output full hash to temp buffer */
+      store64( buffer + sizeof( S->h[i] ) * i, S->h[i] );
+    memcpy( out, buffer, outlen );
+  }
+#endif
   return 0;
 }
 
@@ -418,14 +348,14 @@ int blake2b( uint8_t *out, const void *in, const void *key, const uint8_t outlen
   blake2b_state S[1];
 
   /* Verify parameters */
-  if ( NULL == in && inlen > 0 ) return -1;
+  if( NULL == in && inlen > 0 ) return -1;
 
-  if ( NULL == out ) return -1;
-
-  if( NULL == key && keylen > 0 ) return -1;
+  if( NULL == out ) return -1;
 
   if( !outlen || outlen > BLAKE2B_OUTBYTES ) return -1;
 
+  if( NULL == key && keylen > 0 ) return -1;
+
   if( keylen > BLAKE2B_KEYBYTES ) return -1;
 
   if( keylen > 0 )
@@ -448,11 +378,15 @@ int blake2b_salt_personal( uint8_t *out, const void *in, const void *key, const
   blake2b_state S[1];
 
   /* Verify parameters */
-  if ( NULL == in ) return -1;
+  if( NULL == in && inlen > 0 ) return -1;
 
-  if ( NULL == out ) return -1;
+  if( NULL == out ) return -1;
 
-  if( NULL == key ) keylen = 0;
+  if( !outlen || outlen > BLAKE2B_OUTBYTES ) return -1;
+
+  if( NULL == key && keylen > 0 ) return -1;
+
+  if( keylen > BLAKE2B_KEYBYTES ) return -1;
 
   if( keylen > 0 )
   {
@@ -467,3 +401,25 @@ int blake2b_salt_personal( uint8_t *out, const void *in, const void *key, const
   blake2b_final( S, out, outlen );
   return 0;
 }
+
+int
+blake2b_pick_best_implementation(void)
+{
+#if (defined(HAVE_EMMINTRIN_H) && defined(HAVE_TMMINTRIN_H) && defined(HAVE_SMMINTRIN_H)) || \
+    (defined(_MSC_VER) && (defined(_M_X64) || defined(_M_AMD64) || defined(_M_IX86)))
+  if (sodium_runtime_has_sse41()) {
+    blake2b_compress = blake2b_compress_sse41;
+    return 0;
+  }
+#endif
+#if (defined(HAVE_EMMINTRIN_H) && defined(HAVE_TMMINTRIN_H)) || \
+    (defined(_MSC_VER) && (defined(_M_X64) || defined(_M_AMD64)))
+  if (sodium_runtime_has_ssse3()) {
+    blake2b_compress = blake2b_compress_ssse3;
+    return 0;
+  }
+#endif
+  blake2b_compress = blake2b_compress_ref;
+
+  return 0;
+}