cpuminer-opt-gpu/algo/blake/blake2b.c

#include "algo-gate-api.h"
#include <string.h>
#include <stdint.h>
#include "blake2b-hash.h"

#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
  #define BLAKE2B_8WAY
#elif defined(__AVX2__)
  #define BLAKE2B_4WAY
#endif

#if defined(BLAKE2B_8WAY)

int scanhash_blake2b_8way( struct work *work, uint32_t max_nonce,
                           uint64_t *hashes_done, struct thr_info *mythr )
{
   uint32_t hash[8*8] __attribute__ ((aligned (128)));;
   uint32_t vdata[20*8] __attribute__ ((aligned (64)));;
   uint32_t lane_hash[8] __attribute__ ((aligned (64)));
   blake2b_8way_ctx ctx __attribute__ ((aligned (64)));
   uint32_t *hash7 = &(hash[49]);   // 3*16+1
   uint32_t *pdata = work->data;
   uint32_t *ptarget = work->target;
   int thr_id = mythr->id;
   __m512i  *noncev = (__m512i*)vdata + 9;   // aligned
   const uint32_t Htarg = ptarget[7];
   const uint32_t first_nonce = pdata[19];

   uint32_t n = first_nonce;

   mm512_bswap32_intrlv80_8x64( vdata, pdata );

   do {
      *noncev = mm512_intrlv_blend_32( mm512_bswap_32(
                _mm512_set_epi32( n+7, 0, n+6, 0, n+5, 0, n+4, 0,
                                  n+3, 0, n+2, 0, n+1, 0, n  , 0 ) ), *noncev );

      blake2b_8way_init( &ctx );
      blake2b_8way_update( &ctx, vdata, 80 );
      blake2b_8way_final( &ctx, hash );

      for ( int lane = 0; lane < 8; lane++ )
      if ( hash7[ lane<<1 ] <= Htarg )
      {
          extr_lane_8x64( lane_hash, hash, lane, 256 );
          if ( fulltest( lane_hash, ptarget ) && !opt_benchmark )
          {
              pdata[19] = n + lane;
              submit_solution( work, lane_hash, mythr );
          }
      }
      n += 8;
   } while ( (n < max_nonce-8) && !work_restart[thr_id].restart);

   *hashes_done = n - first_nonce + 1;
   return 0;
}

#elif defined(BLAKE2B_4WAY)

// Function not used, code inlined.
void blake2b_4way_hash(void *output, const void *input)
{
    blake2b_4way_ctx ctx;
    blake2b_4way_init( &ctx );
    blake2b_4way_update( &ctx, input, 80 );
    blake2b_4way_final( &ctx, output );
}

int scanhash_blake2b_4way( struct work *work, uint32_t max_nonce,
                           uint64_t *hashes_done, struct thr_info *mythr )
{
	uint32_t hash[8*4] __attribute__ ((aligned (64)));;
   uint32_t vdata[20*4] __attribute__ ((aligned (32)));;
   uint32_t lane_hash[8] __attribute__ ((aligned (32)));
   blake2b_4way_ctx ctx __attribute__ ((aligned (32)));
   uint32_t *hash7 = &(hash[25]);   // 3*8+1
	uint32_t *pdata = work->data;
	uint32_t *ptarget = work->target;
   int thr_id = mythr->id;
   __m256i  *noncev = (__m256i*)vdata + 9;   // aligned
	const uint32_t Htarg = ptarget[7];
	const uint32_t first_nonce = pdata[19];

	uint32_t n = first_nonce;

   mm256_bswap32_intrlv80_4x64( vdata, pdata );

	do {
      *noncev = mm256_intrlv_blend_32( mm256_bswap_32(
                _mm256_set_epi32( n+3, 0, n+2, 0, n+1, 0, n, 0 ) ), *noncev );

      blake2b_4way_init( &ctx ); 
      blake2b_4way_update( &ctx, vdata, 80 );
      blake2b_4way_final( &ctx, hash );

      for ( int lane = 0; lane < 4; lane++ )
      if ( hash7[ lane<<1 ] <= Htarg )
      {
          extr_lane_4x64( lane_hash, hash, lane, 256 );
          if ( fulltest( lane_hash, ptarget ) && !opt_benchmark )
          {
              pdata[19] = n + lane;
              submit_solution( work, lane_hash, mythr );
          }
      }
      n += 4;
   } while ( (n < max_nonce-4) && !work_restart[thr_id].restart);

   *hashes_done = n - first_nonce + 1;
   return 0;
}

#else

#include "algo/blake/sph_blake2b.h"

void blake2b_hash(void *output, const void *input)
{
   uint8_t _ALIGN(32) hash[32];
   sph_blake2b_ctx ctx __attribute__ ((aligned (32)));

   sph_blake2b_init(&ctx, 32, NULL, 0);
   sph_blake2b_update(&ctx, input, 80);
   sph_blake2b_final(&ctx, hash);

   memcpy(output, hash, 32);
}

int scanhash_blake2b( struct work *work, uint32_t max_nonce,
                      uint64_t *hashes_done, struct thr_info *mythr )
{
   uint32_t _ALIGN(32) hash64[8];
   uint32_t _ALIGN(32) endiandata[20];
   uint32_t *pdata = work->data;
   uint32_t *ptarget = work->target;
   int thr_id = mythr->id; 
   const uint32_t first_nonce = pdata[19];
   uint32_t n = first_nonce;

   mm128_bswap32_80( endiandata, pdata );

   do {
      endiandata[19] = n;
      blake2b_hash( hash64, endiandata );
      if ( unlikely( valid_hash( hash64, ptarget ) ) && !opt_benchmark )
      {
         pdata[19] = bswap_32( n );
         submit_solution( work, hash64, mythr );
      }
      n++;
   } while (n < max_nonce && !work_restart[thr_id].restart);
   *hashes_done = n - first_nonce + 1;
   pdata[19] = n;

   return 0;
}

#endif

bool register_blake2b_algo( algo_gate_t* gate )
{
#if defined(BLAKE2B_8WAY)
  gate->scanhash  = (void*)&scanhash_blake2b_8way;
#elif defined(BLAKE2B_4WAY)
  gate->scanhash  = (void*)&scanhash_blake2b_4way;
  gate->hash      = (void*)&blake2b_4way_hash;
#else
  gate->scanhash  = (void*)&scanhash_blake2b;
  gate->hash      = (void*)&blake2b_hash;
#endif
  gate->optimizations =  AVX2_OPT | AVX512_OPT;
  return true;
};