v3.14.3

v3.14.2
v3.14.1
2025-09-17 23:44:27 +00:00 · 2020-06-18 17:30:26 -04:00 · 2020-05-30 21:20:44 -04:00 · 2020-05-21 13:00:29 -04:00 · 2020-05-20 13:56:35 -04:00 · 2020-05-18 18:17:27 -04:00
39 changed files with 915 additions and 891 deletions
--- a/README.txt
+++ b/README.txt
@@ -46,6 +46,15 @@ Icelake is only available on some laptops. Mining with a laptop is not
 recommended. The icelake build is included in anticipation of Intel eventually
 releasing a desktop CPU with a microarchitecture newer than Skylake.

+Notes about included DLL files:
+
+Downloading DLL files from alternative sources presents an inherent
+security risk if their source is unknown. All DLL files included have
+been copied from the Ubuntu-20.04 instalation or compiled by me from
+source code obtained from the author's official repository. The exact
+procedure is documented in the build instructions for Windows:
+https://github.com/JayDDee/cpuminer-opt/wiki/Compiling-from-source
+
 If you like this software feel free to donate:

 BTC: 12tdvfF7KmAsihBXQXynT6E6th2c2pByTT
--- a/43
+++ b/43
@@ -65,6 +65,49 @@ If not what makes it happen or not happen?
 Change Log
 ----------

+v3.14.3
+
+#265: more mutex changes to reduce blocking with high thread count.
+
+#267: fixed hodl algo potential memory alignment issue,
+      add warning when thread count is not valid for mining hodl algo.
+
+v3.14.2
+
+The second line of the Share Accepted log is no longer displayed,
+new Xnonce log is added and other small log tweaks.
+
+#265: Cleanup use of mutex.
+
+v3.14.1
+
+GBT and getwork log changes:
+ fixed missing TTF in New Block log,
+ ntime no longer byte-swapped for display in New Work log,
+ fixed zero effective hash rate in Periodic Report log,
+ deleted "Current block is..." log.
+
+Renamed stratum "New Job" log to "New Work" to be consistent with the solo
+version of the log. Added more data to both versions.
+
+v3.14.0
+
+Changes to solo mining:
+  - segwit is supported by getblocktemplate,
+  - longpolling is not working and is disabled,
+  - Periodic Report log is output,
+  - New Block log includes TTF estimates,
+  - Stratum thread no longer created when using getwork or GBT.
+
+Fixed BUG log mining sha256d.
+
+v3.13.1.1
+
+Fixed Windows crash mining minotaur algo.
+
+Fixed GCC 10 compile again.
+Added -fno-common to testing to be consistent with GCC 10 default.
+
 v3.13.1

 Added minotaur algo for Ringcoin.
--- a/algo-gate-api.c
+++ b/algo-gate-api.c
@@ -105,17 +105,16 @@ int scanhash_generic( struct work *work, uint32_t max_nonce,
   uint32_t hash[8] __attribute__((aligned(64)));
   uint32_t *pdata = work->data;
   uint32_t *ptarget = work->target;
-   uint32_t n = pdata[19];
   const uint32_t first_nonce = pdata[19];
+   const uint32_t last_nonce = max_nonce - 1;
+   uint32_t n = first_nonce;
   const int thr_id = mythr->id;
   const bool bench = opt_benchmark;

   mm128_bswap32_80( edata, pdata );
-
   do
   {
      edata[19] = n;
-
      if ( likely( algo_gate.hash( hash, edata, thr_id ) ) )
      if ( unlikely( valid_hash( hash, ptarget ) && !bench ) )
      {
@@ -123,12 +122,125 @@ int scanhash_generic( struct work *work, uint32_t max_nonce,
         submit_solution( work, hash, mythr );
      }
      n++;
-   } while ( n < max_nonce && !work_restart[thr_id].restart );
+   } while ( n < last_nonce && !work_restart[thr_id].restart );
   *hashes_done = n - first_nonce;
   pdata[19] = n;
   return 0;
 }

+#if defined(__AVX2__)
+
+//int scanhash_4way_64_64( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr )
+
+//int scanhash_4way_64_640( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr )
+
+int scanhash_4way_64in_32out( struct work *work, uint32_t max_nonce,
+                      uint64_t *hashes_done, struct thr_info *mythr )
+{
+   uint32_t hash32[8*4] __attribute__ ((aligned (64)));
+   uint32_t vdata[20*4] __attribute__ ((aligned (64)));
+   uint32_t lane_hash[8] __attribute__ ((aligned (64)));
+   uint32_t *hash32_d7 = &(hash32[ 7*4 ]);
+   uint32_t *pdata = work->data;
+   const uint32_t *ptarget = work->target;
+   const uint32_t first_nonce = pdata[19];
+   const uint32_t last_nonce = max_nonce - 4;
+   __m256i  *noncev = (__m256i*)vdata + 9;
+   uint32_t n = first_nonce;
+   const int thr_id = mythr->id;
+   const uint32_t targ32_d7 = ptarget[7];
+   const bool bench = opt_benchmark;
+
+   mm256_bswap32_intrlv80_4x64( vdata, pdata );
+   *noncev = mm256_intrlv_blend_32(
+                   _mm256_set_epi32( n+3, 0, n+2, 0, n+1, 0, n, 0 ), *noncev );
+   do
+   {
+      if ( likely( algo_gate.hash( hash32, vdata, thr_id ) ) )
+      for ( int lane = 0; lane < 4; lane++ )
+      if ( unlikely( hash32_d7[ lane ] <= targ32_d7 && !bench ) )
+      {
+         extr_lane_4x32( lane_hash, hash32, lane, 256 );
+         if ( valid_hash( lane_hash, ptarget ) )
+         {
+            pdata[19] = bswap_32( n + lane );
+            submit_solution( work, lane_hash, mythr );
+         }
+      }
+      *noncev = _mm256_add_epi32( *noncev,
+                                  m256_const1_64( 0x0000000400000000 ) );
+      n += 4;
+   } while ( likely( ( n <= last_nonce ) && !work_restart[thr_id].restart ) );
+   pdata[19] = n;
+   *hashes_done = n - first_nonce;
+   return 0;
+}
+
+//int scanhash_8way_32_32( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr )
+
+#endif
+
+#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
+
+//int scanhash_8way_64_64( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr )
+
+//int scanhash_8way_64_640( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr )
+
+int scanhash_8way_64in_32out( struct work *work, uint32_t max_nonce,
+                      uint64_t *hashes_done, struct thr_info *mythr )
+{
+   uint32_t hash32[8*8] __attribute__ ((aligned (128)));
+   uint32_t vdata[20*8] __attribute__ ((aligned (64)));
+   uint32_t lane_hash[8] __attribute__ ((aligned (64)));
+   uint32_t *hash32_d7 = &(hash32[7*8]);
+   uint32_t *pdata = work->data;
+   const uint32_t *ptarget = work->target;
+   const uint32_t first_nonce = pdata[19];
+   const uint32_t last_nonce = max_nonce - 8;
+   __m512i  *noncev = (__m512i*)vdata + 9;
+   uint32_t n = first_nonce;
+   const int thr_id = mythr->id;
+   const uint32_t targ32_d7 = ptarget[7];
+   const bool bench = opt_benchmark;
+
+   mm512_bswap32_intrlv80_8x64( vdata, pdata );
+   *noncev = mm512_intrlv_blend_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 );
+   do
+   {
+      if ( likely( algo_gate.hash( hash32, vdata, thr_id ) ) )
+      for ( int lane = 0; lane < 8; lane++ )
+      if ( unlikely( ( hash32_d7[ lane ] <= targ32_d7 ) && !bench ) )
+      {
+         extr_lane_8x32( lane_hash, hash32, lane, 256 );
+         if ( likely( valid_hash( lane_hash, ptarget ) ) )
+         {
+            pdata[19] = bswap_32( n + lane );
+            submit_solution( work, lane_hash, mythr );
+         }
+      }
+      *noncev = _mm512_add_epi32( *noncev,
+                                  m512_const1_64( 0x0000000800000000 ) );
+      n += 8;
+   } while ( likely( ( n < last_nonce ) && !work_restart[thr_id].restart ) );
+   pdata[19] = n;
+   *hashes_done = n - first_nonce;
+   return 0;
+}
+
+//int scanhash_16way_32_32( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr )
+
+#endif
+
+
+
 int null_hash()
 {
   applog(LOG_WARNING,"SWERR: null_hash unsafe null function");
--- a/algo-gate-api.h
+++ b/algo-gate-api.h
@@ -110,10 +110,12 @@ inline bool set_excl ( set_t a, set_t b ) { return (a & b) == 0; }

 typedef struct
 {
-// mandatory function, must be overwritten
+// Mandatory functions, one of these is mandatory. If a generic scanhash
+// is used a custom hash function must be registered, with a custom scanhash
+// the custom hash function can be called directly and doesn't need to be
+// registered in the gate. 
 int ( *scanhash ) ( struct work*, uint32_t, uint64_t*, struct thr_info* );

-//int ( *hash )     ( void*, const void*, uint32_t ) ;
 int ( *hash )     ( void*, const void*, int );

 //optional, safe to use default in most cases
@@ -126,7 +128,7 @@ bool ( *miner_thread_init )     ( int );
 void ( *get_new_work )          ( struct work*, struct work*, int, uint32_t* );

 // Decode getwork blockheader
-bool ( *work_decode )           ( const json_t*, struct work* );
+bool ( *work_decode )           ( struct work* );

 // Extra getwork data
 void ( *decode_extra_data )     ( struct work*, uint64_t* );
@@ -201,19 +203,61 @@ void four_way_not_tested();
 #define STD_WORK_DATA_SIZE 128
 #define STD_WORK_CMP_SIZE 76

-#define JR2_NONCE_INDEX 39  // 8 bit offset
+//#define JR2_NONCE_INDEX 39  // 8 bit offset

 // These indexes are only used with JSON RPC2 and are not gated.
-#define JR2_WORK_CMP_INDEX_2 43
-#define JR2_WORK_CMP_SIZE_2 33
+//#define JR2_WORK_CMP_INDEX_2 43
+//#define JR2_WORK_CMP_SIZE_2 33

 // deprecated, use generic instead
 int null_scanhash();

 // Default generic, may be used in many cases.
+// N-way is more complicated, requires many different implementations
+// depending on architecture, input format, and output format.
+// Naming convention is scanhash_[N]way_[input format]in_[output format]out
+// N = number of lanes
+// input/output format:
+//    32: 32 bit interleaved parallel lanes
+//    64: 64 bit interleaved parallel lanes
+//    640: input only, not interleaved, contiguous serial 640 bit lanes.
+//    256: output only, not interleaved, contiguous serial 256 bit lanes.
+
 int scanhash_generic( struct work *work, uint32_t max_nonce,
                      uint64_t *hashes_done, struct thr_info *mythr );

+#if defined(__AVX2__)
+
+//int scanhash_4way_64in_64out( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr );
+
+//int scanhash_4way_64in_256out( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr );
+
+int scanhash_4way_64in_32out( struct work *work, uint32_t max_nonce,
+                      uint64_t *hashes_done, struct thr_info *mythr );
+
+//int scanhash_8way_32in_32out( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr );
+
+#endif
+
+#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
+
+//int scanhash_8way_64in_64out( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr );
+
+//int scanhash_8way_64in_256out( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr );
+
+int scanhash_8way_64in_32out( struct work *work, uint32_t max_nonce,
+                      uint64_t *hashes_done, struct thr_info *mythr );
+
+//int scanhash_16way_32in_32out( struct work *work, uint32_t max_nonce,
+//                      uint64_t *hashes_done, struct thr_info *mythr );
+
+#endif
+
 // displays warning
 int null_hash    ();

@@ -225,8 +269,8 @@ void std_get_new_work( struct work *work, struct work *g_work, int thr_id,
 void sha256d_gen_merkle_root( char *merkle_root, struct stratum_ctx *sctx );
 void SHA256_gen_merkle_root ( char *merkle_root, struct stratum_ctx *sctx );

-bool std_le_work_decode( const json_t *val, struct work *work );
-bool std_be_work_decode( const json_t *val, struct work *work );
+bool std_le_work_decode( struct work *work );
+bool std_be_work_decode( struct work *work );

 bool std_le_submit_getwork_result( CURL *curl, struct work *work );
 bool std_be_submit_getwork_result( CURL *curl, struct work *work );
@@ -261,7 +305,7 @@ int std_get_work_data_size();
 // by calling the algo's register function.
 bool register_algo_gate( int algo, algo_gate_t *gate );

-// Called by algos toverride any default gate functions that are applicable
+// Called by algos to verride any default gate functions that are applicable
 // and do any other algo-specific initialization.
 // The register functions for all the algos can be declared here to reduce
 // compiler warnings but that's just more work for devs adding new algos.
--- a/algo/blake/decred-gate.c
+++ b/algo/blake/decred-gate.c
@@ -78,7 +78,6 @@ void decred_build_extraheader( struct work* g_work, struct stratum_ctx* sctx )
   uint32_t extraheader[32] = { 0 };
   int headersize = 0;
   uint32_t* extradata = (uint32_t*) sctx->xnonce1;
-   size_t t;
   int i;

   // getwork over stratum, getwork merkle + header passed in coinb1
@@ -87,9 +86,6 @@ void decred_build_extraheader( struct work* g_work, struct stratum_ctx* sctx )
                  sizeof(extraheader) );
   memcpy( extraheader, &sctx->job.coinbase[32], headersize );

-   // Increment extranonce2 
-   for ( t = 0; t < sctx->xnonce2_size && !( ++sctx->job.xnonce2[t] ); t++ );
-
   // Assemble block header 
   memset( g_work->data, 0, sizeof(g_work->data) );
   g_work->data[0] = le32dec( sctx->job.version );
--- a/algo/hodl/hodl-gate.c
+++ b/algo/hodl/hodl-gate.c
@@ -99,9 +99,13 @@ void hodl_build_block_header( struct work* g_work, uint32_t version,
 // called only by thread 0, saves a backup of g_work
 void hodl_get_new_work( struct work* work, struct work* g_work)
 {
-     work_free( &hodl_work );
-     work_copy( &hodl_work, g_work );
-     hodl_work.data[ algo_gate.nonce_index ] = ( clock() + rand() ) % 9999;
+   pthread_rwlock_rdlock( &g_work_lock );
+
+   work_free( &hodl_work );
+   work_copy( &hodl_work, g_work );
+   hodl_work.data[ algo_gate.nonce_index ] = ( clock() + rand() ) % 9999;
+
+   pthread_rwlock_unlock( &g_work_lock );
 }

 json_t *hodl_longpoll_rpc_call( CURL *curl, int *err, char* lp_url )
@@ -155,11 +159,10 @@ bool register_hodl_algo( algo_gate_t* gate )
  applog( LOG_ERR, "Only CPUs with AES are supported, use legacy version.");
  return false;
 #endif
-//  if ( TOTAL_CHUNKS % opt_n_threads )
-//  {
-//     applog(LOG_ERR,"Thread count must be power of 2.");
-//     return false;
-//  }
+
+  if ( GARBAGE_SIZE % opt_n_threads )
+     applog( LOG_WARNING,"WARNING: Thread count must be power of 2. Miner may crash or produce invalid hash!" );
+
  pthread_barrier_init( &hodl_barrier, NULL, opt_n_threads );
  gate->optimizations         = SSE42_OPT | AES_OPT | AVX2_OPT;
  gate->scanhash              = (void*)&hodl_scanhash;
@@ -171,7 +174,7 @@ bool register_hodl_algo( algo_gate_t* gate )
  gate->resync_threads        = (void*)&hodl_resync_threads;
  gate->do_this_thread        = (void*)&hodl_do_this_thread;
  gate->work_cmp_size         = 76;
-  hodl_scratchbuf = (unsigned char*)malloc( 1 << 30 );
+  hodl_scratchbuf = (unsigned char*)_mm_malloc( 1 << 30, 64 );
  allow_getwork = false;
  opt_target_factor = 8388608.0;
  return ( hodl_scratchbuf != NULL );
--- a/algo/hodl/hodl-wolf.c
+++ b/algo/hodl/hodl-wolf.c
@@ -70,7 +70,7 @@ int scanhash_hodl_wolf( struct work* work, uint32_t max_nonce,
    uint32_t *ptarget = work->target;
    int threadNumber = mythr->id;
    CacheEntry *Garbage = (CacheEntry*)hodl_scratchbuf;
-    CacheEntry Cache[AES_PARALLEL_N];
+    CacheEntry Cache[AES_PARALLEL_N] __attribute__ ((aligned (64)));
    __m128i* data[AES_PARALLEL_N];
    const __m128i* next[AES_PARALLEL_N];
    uint32_t CollisionCount = 0;
--- a/algo/lyra2/lyra2-gate.c
+++ b/algo/lyra2/lyra2-gate.c
@@ -215,9 +215,6 @@ void phi2_build_extraheader( struct work* g_work, struct stratum_ctx* sctx )
   size_t t;

   algo_gate.gen_merkle_root( merkle_tree, sctx );
-   // Increment extranonce2
-   for ( t = 0; t < sctx->xnonce2_size && !( ++sctx->job.xnonce2[t] ); t++ );
-   // Assemble block header
   algo_gate.build_block_header( g_work, le32dec( sctx->job.version ),
                  (uint32_t*) sctx->job.prevhash, (uint32_t*) merkle_tree,
                  le32dec( sctx->job.ntime ), le32dec(sctx->job.nbits), NULL );
@@ -225,7 +222,6 @@ void phi2_build_extraheader( struct work* g_work, struct stratum_ctx* sctx )
      g_work->data[ 20+t ] = ((uint32_t*)sctx->job.extra)[t];
 }

-
 bool register_phi2_algo( algo_gate_t* gate )
 {
   gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT | VAES_OPT;
--- a/algo/nist5/zr5.c
+++ b/algo/nist5/zr5.c
@@ -156,6 +156,8 @@ int scanhash_zr5( struct work *work, uint32_t max_nonce,
 void zr5_get_new_work( struct work* work, struct work* g_work, int thr_id,
                       uint32_t* end_nonce_ptr )
 {
+   pthread_rwlock_rdlock( &g_work_lock );
+
   // ignore POK in first word
   const int wkcmp_sz = 72;  // (19-1) * sizeof(uint32_t)
   uint32_t *nonceptr = work->data + algo_gate.nonce_index;
@@ -171,6 +173,8 @@ void zr5_get_new_work( struct work* work, struct work* g_work, int thr_id,
   }
   else
       ++(*nonceptr);
+
+   pthread_rwlock_unlock( &g_work_lock );
 }

 void zr5_display_pok( struct work* work )
--- a/algo/ripemd/lbry-gate.c
+++ b/algo/ripemd/lbry-gate.c
@@ -69,13 +69,9 @@ void lbry_build_block_header( struct work* g_work, uint32_t version,
 void lbry_build_extraheader( struct work* g_work, struct stratum_ctx* sctx )
 {
   unsigned char merkle_root[64] = { 0 };
-   size_t t;
   int i;

   algo_gate.gen_merkle_root( merkle_root, sctx );
-   // Increment extranonce2 
-   for ( t = 0; t < sctx->xnonce2_size && !( ++sctx->job.xnonce2[t] ); t++ );
-   // Assemble block header 

   memset( g_work->data, 0, sizeof(g_work->data) );
   g_work->data[0] = le32dec( sctx->job.version );
--- a/algo/sha/sha2.c
+++ b/algo/sha/sha2.c
@@ -479,8 +479,8 @@ static inline void sha256d_ms(uint32_t *hash, uint32_t *W,
 void sha256d_ms_4way(uint32_t *hash,  uint32_t *data,
 	const uint32_t *midstate, const uint32_t *prehash);

-static inline int scanhash_sha256d_4way(int thr_id, struct work *work,
-             uint32_t max_nonce, uint64_t *hashes_done)
+static inline int scanhash_sha256d_4way( struct work *work,
+             uint32_t max_nonce, uint64_t *hashes_done, struct thr_info *mythr )
 {
        uint32_t *pdata = work->data;
        uint32_t *ptarget = work->target;
@@ -492,6 +492,7 @@ static inline int scanhash_sha256d_4way(int thr_id, struct work *work,
 	uint32_t n = pdata[19] - 1;
 	const uint32_t first_nonce = pdata[19];
 	const uint32_t Htarg = ptarget[7];
+   int thr_id = mythr->id;
 	int i, j;
 	
 	memcpy(data, pdata + 16, 64);
@@ -521,10 +522,8 @@ static inline int scanhash_sha256d_4way(int thr_id, struct work *work,
 			if (swab32(hash[4 * 7 + i]) <= Htarg) {
 				pdata[19] = data[4 * 3 + i];
 				sha256d_80_swap(hash, pdata);
-				if (fulltest(hash, ptarget)) {
-					*hashes_done = n - first_nonce + 1;
-					return 1;
-				}
+            if ( fulltest( hash, ptarget ) && !opt_benchmark )
+               submit_solution( work, hash, mythr );
 			}
 		}
 	} while (n < max_nonce && !work_restart[thr_id].restart);
@@ -541,8 +540,8 @@ static inline int scanhash_sha256d_4way(int thr_id, struct work *work,
 void sha256d_ms_8way(uint32_t *hash,  uint32_t *data,
 	const uint32_t *midstate, const uint32_t *prehash);

-static inline int scanhash_sha256d_8way(int thr_id, struct work *work,
-                              uint32_t max_nonce, uint64_t *hashes_done)
+static inline int scanhash_sha256d_8way( struct work *work,
+            uint32_t max_nonce, uint64_t *hashes_done, struct thr_info *mythr )
 {
        uint32_t *pdata = work->data;
        uint32_t *ptarget = work->target;
@@ -554,6 +553,7 @@ static inline int scanhash_sha256d_8way(int thr_id, struct work *work,
 	uint32_t n = pdata[19] - 1;
 	const uint32_t first_nonce = pdata[19];
 	const uint32_t Htarg = ptarget[7];
+   int thr_id = mythr->id;
 	int i, j;
 	
 	memcpy(data, pdata + 16, 64);
@@ -583,10 +583,8 @@ static inline int scanhash_sha256d_8way(int thr_id, struct work *work,
 			if (swab32(hash[8 * 7 + i]) <= Htarg) {
 				pdata[19] = data[8 * 3 + i];
 				sha256d_80_swap(hash, pdata);
-				if (fulltest(hash, ptarget)) {
-					*hashes_done = n - first_nonce + 1;
-					return 1;
-				}
+            if ( fulltest( hash, ptarget ) && !opt_benchmark )
+               submit_solution( work, hash, mythr );
 			}
 		}
 	} while (n < max_nonce && !work_restart[thr_id].restart);
@@ -614,13 +612,11 @@ int scanhash_sha256d( struct work *work,

 #ifdef HAVE_SHA256_8WAY
 	if (sha256_use_8way())
-		return scanhash_sha256d_8way(thr_id, work,
-			max_nonce, hashes_done);
+		return scanhash_sha256d_8way( work,	max_nonce, hashes_done, mythr );
 #endif
 #ifdef HAVE_SHA256_4WAY
 	if (sha256_use_4way())
-		return scanhash_sha256d_4way(thr_id, work,
-			max_nonce, hashes_done);
+		return scanhash_sha256d_4way( work,	max_nonce, hashes_done, mythr );
 #endif
 	
 	memcpy(data, pdata + 16, 64);
@@ -657,7 +653,7 @@ int scanhash_SHA256d( struct work *work, const uint32_t max_nonce,
   uint32_t n = pdata[19] - 1;
   const uint32_t first_nonce = pdata[19];
   const uint32_t Htarg = ptarget[7];
-   int thr_id = mythr->id;  // thr_id arg is deprecated
+   int thr_id = mythr->id;

   memcpy( data, pdata, 80 );

--- a/algo/skein/skein-hash-4way.c
+++ b/algo/skein/skein-hash-4way.c
@@ -731,7 +731,7 @@ void skein512_8way_full( skein512_8way_context *sc, void *out, const void *data,
 void
 skein512_8way_prehash64( skein512_8way_context *sc, const void *data )
 {
-   __m512i *vdata = (__m512*)data;
+   __m512i *vdata = (__m512i*)data;
   __m512i *buf = sc->buf;
   buf[0] = vdata[0];
   buf[1] = vdata[1];
--- a/algo/x16/minotaur.c
+++ b/algo/x16/minotaur.c
@@ -7,7 +7,6 @@
 #include <stdio.h>
 #include "algo/blake/sph_blake.h"
 #include "algo/bmw/sph_bmw.h"
-#include "algo/groestl/sph_groestl.h"
 #include "algo/jh/sph_jh.h"
 #include "algo/keccak/sph_keccak.h"
 #include "algo/skein/sph_skein.h"
@@ -15,7 +14,6 @@
 #include "algo/luffa/luffa_for_sse2.h"
 #include "algo/cubehash/cubehash_sse2.h"
 #include "algo/simd/nist.h"
-#include "algo/echo/sph_echo.h"
 #include "algo/hamsi/sph_hamsi.h"
 #include "algo/fugue/sph_fugue.h"
 #include "algo/shabal/sph_shabal.h"
@@ -24,9 +22,11 @@
 #if defined(__AES__)
  #include "algo/echo/aes_ni/hash_api.h"
  #include "algo/groestl/aes_ni/hash-groestl.h"
+#else
+  #include "algo/echo/sph_echo.h"
+  #include "algo/groestl/sph_groestl.h"
 #endif

-
 // Config
 #define MINOTAUR_ALGO_COUNT	16

@@ -39,8 +39,8 @@ struct TortureGarden {
        hashState_echo          echo;
        hashState_groestl       groestl;
 #else
-        sph_groestl512_context   groestl;
        sph_echo512_context      echo;
+        sph_groestl512_context   groestl;
 #endif
        sph_blake512_context    blake;
        sph_bmw512_context      bmw;
@@ -62,13 +62,13 @@ struct TortureGarden {
        TortureNode *childLeft;
        TortureNode *childRight;
    } nodes[22];
-};
+} __attribute__ ((aligned (64)));

 // Get a 64-byte hash for given 64-byte input, using given TortureGarden contexts and given algo index
 static void get_hash( void *output, const void *input, TortureGarden *garden,
 	              unsigned int algo )
 {    
-	unsigned char _ALIGN(64) hash[64];
+	unsigned char hash[64] __attribute__ ((aligned (64)));

    switch (algo) {
        case 0:
@@ -170,7 +170,7 @@ static void get_hash( void *output, const void *input, TortureGarden *garden,
 static void traverse_garden( TortureGarden *garden, void *hash,
 	                     TortureNode *node )
 {
-    unsigned char _ALIGN(64) partialHash[64];
+    unsigned char partialHash[64] __attribute__ ((aligned (64)));
    get_hash(partialHash, hash, garden, node->algo);

    if ( partialHash[63] % 2 == 0 )
@@ -195,9 +195,9 @@ static inline void link_nodes( TortureNode *parent, TortureNode *childLeft,
    parent->childRight = childRight;
 }

-static TortureGarden garden;
+static __thread TortureGarden garden;

-void initialize_torture_garden()
+bool initialize_torture_garden()
 {
    // Create torture garden nodes. Note that both sides of 19 and 20 lead to 21, and 21 has no children (to make traversal complete).
    link_nodes(&garden.nodes[0], &garden.nodes[1], &garden.nodes[2]);
@@ -223,12 +223,13 @@ void initialize_torture_garden()
    link_nodes(&garden.nodes[20], &garden.nodes[21], &garden.nodes[21]);
    garden.nodes[21].childLeft = NULL;
    garden.nodes[21].childRight = NULL;
+    return true;
 }

 // Produce a 32-byte hash from 80-byte input data
-int minotaur_hash( void *output, const void *input )
+int minotaur_hash( void *output, const void *input, int thr_id )
 {    
-    unsigned char _ALIGN(64) hash[64];
+    unsigned char hash[64] __attribute__ ((aligned (64)));

    // Find initial sha512 hash
    SHA512_Init( &garden.sha512 );
@@ -251,7 +252,7 @@ bool register_minotaur_algo( algo_gate_t* gate )
 {
  gate->hash      = (void*)&minotaur_hash;
  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT;
-  initialize_torture_garden();
+  gate->miner_thread_init = (void*)&initialize_torture_garden;
  return true;
 };

--- a/algo/x16/x16r-gate.c
+++ b/algo/x16/x16r-gate.c
@@ -135,18 +135,16 @@ void x16rt_getAlgoString( const uint32_t *timeHash, char *output)

 void veil_build_extraheader( struct work* g_work, struct stratum_ctx* sctx )
 {
+   uint32_t merkleroothash[8];
+   uint32_t witmerkleroothash[8];
+   uint32_t denom10[8];
+   uint32_t denom100[8];
+   uint32_t denom1000[8];
+   uint32_t denom10000[8];
+   int i;
   uchar merkle_tree[64] = { 0 };
-   size_t t;

   algo_gate.gen_merkle_root( merkle_tree, sctx );
-   // Increment extranonce2
-   for ( t = 0; t < sctx->xnonce2_size && !( ++sctx->job.xnonce2[t] ); t++ );
-
-   // Assemble block header
-//   algo_gate.build_block_header( g_work, le32dec( sctx->job.version ),
-//          (uint32_t*) sctx->job.prevhash, (uint32_t*) merkle_tree,
-//          le32dec( sctx->job.ntime ), le32dec(sctx->job.nbits) );
-   int i;

   memset( g_work->data, 0, sizeof(g_work->data) );
   g_work->data[0] = le32dec( sctx->job.version );
@@ -164,35 +162,35 @@ void veil_build_extraheader( struct work* g_work, struct stratum_ctx* sctx )
   g_work->data[31] = 0x00000280;

   for ( i = 0; i < 8; i++ )
-      g_work->merkleroothash[7 - i] = be32dec((uint32_t *)merkle_tree + i);
+      merkleroothash[7 - i] = be32dec((uint32_t *)merkle_tree + i);
   for ( i = 0; i < 8; i++ )
-      g_work->witmerkleroothash[7 - i] = be32dec((uint32_t *)merkle_tree + i);
+      witmerkleroothash[7 - i] = be32dec((uint32_t *)merkle_tree + i);
   for ( i = 0; i < 8; i++ )
-      g_work->denom10[i] =    le32dec((uint32_t *)sctx->job.denom10 + i);
+      denom10[i] =    le32dec((uint32_t *)sctx->job.denom10 + i);
   for ( i = 0; i < 8; i++ )
-      g_work->denom100[i] =   le32dec((uint32_t *)sctx->job.denom100 + i);
+      denom100[i] =   le32dec((uint32_t *)sctx->job.denom100 + i);
   for ( i = 0; i < 8; i++ )
-      g_work->denom1000[i] =  le32dec((uint32_t *)sctx->job.denom1000 + i);
+      denom1000[i] =  le32dec((uint32_t *)sctx->job.denom1000 + i);
   for ( i = 0; i < 8; i++ )
-      g_work->denom10000[i] = le32dec((uint32_t *)sctx->job.denom10000 + i);
+      denom10000[i] = le32dec((uint32_t *)sctx->job.denom10000 + i);

   uint32_t pofnhash[8];
   memset(pofnhash, 0x00, 32);

-   char denom10_str      [ 2 * sizeof( g_work->denom10 )           + 1 ];
-   char denom100_str     [ 2 * sizeof( g_work->denom100 )          + 1 ];
-   char denom1000_str    [ 2 * sizeof( g_work->denom1000 )         + 1 ];
-   char denom10000_str   [ 2 * sizeof( g_work->denom10000 )        + 1 ];
-   char merkleroot_str   [ 2 * sizeof( g_work->merkleroothash )    + 1 ];
-   char witmerkleroot_str[ 2 * sizeof( g_work->witmerkleroothash ) + 1 ];
+   char denom10_str      [ 2 * sizeof( denom10 )           + 1 ];
+   char denom100_str     [ 2 * sizeof( denom100 )          + 1 ];
+   char denom1000_str    [ 2 * sizeof( denom1000 )         + 1 ];
+   char denom10000_str   [ 2 * sizeof( denom10000 )        + 1 ];
+   char merkleroot_str   [ 2 * sizeof( merkleroothash )    + 1 ];
+   char witmerkleroot_str[ 2 * sizeof( witmerkleroothash ) + 1 ];
   char pofn_str         [ 2 * sizeof( pofnhash )                  + 1 ];

-   cbin2hex( denom10_str,       (char*) g_work->denom10,           32 );
-   cbin2hex( denom100_str,      (char*) g_work->denom100,          32 );
-   cbin2hex( denom1000_str,     (char*) g_work->denom1000,         32 );
-   cbin2hex( denom10000_str,    (char*) g_work->denom10000,        32 );
-   cbin2hex( merkleroot_str,    (char*) g_work->merkleroothash,    32 );
-   cbin2hex( witmerkleroot_str, (char*) g_work->witmerkleroothash, 32 );
+   cbin2hex( denom10_str,       (char*) denom10,           32 );
+   cbin2hex( denom100_str,      (char*) denom100,          32 );
+   cbin2hex( denom1000_str,     (char*) denom1000,         32 );
+   cbin2hex( denom10000_str,    (char*) denom10000,        32 );
+   cbin2hex( merkleroot_str,    (char*) merkleroothash,    32 );
+   cbin2hex( witmerkleroot_str, (char*) witmerkleroothash, 32 );
   cbin2hex( pofn_str,          (char*) pofnhash,                  32 );

   if ( true )
--- a/algo/x17/sonoa-4way.c
+++ b/algo/x17/sonoa-4way.c
@@ -58,7 +58,7 @@ union _sonoa_8way_context_overlay

 typedef union _sonoa_8way_context_overlay sonoa_8way_context_overlay;

-int sonoa_8way_hash( void *state, const void *input, int thrid )
+int sonoa_8way_hash( void *state, const void *input, int thr_id )
 {
     uint64_t vhash[8*8] __attribute__ ((aligned (128)));
     uint64_t vhashA[8*8] __attribute__ ((aligned (64)));
@@ -186,7 +186,7 @@ int sonoa_8way_hash( void *state, const void *input, int thrid )

 #endif

-     if ( work_restart[thrid].restart ) return 0;
+     if ( work_restart[thr_id].restart ) return 0;
 // 2

     bmw512_8way_full( &ctx.bmw, vhash, vhash, 64 );
@@ -302,7 +302,7 @@ int sonoa_8way_hash( void *state, const void *input, int thrid )
     hamsi512_8way_update( &ctx.hamsi, vhash, 64 );
     hamsi512_8way_close( &ctx.hamsi, vhash );

-     if ( work_restart[thrid].restart ) return 0;
+     if ( work_restart[thr_id].restart ) return 0;
 // 3

     bmw512_8way_full( &ctx.bmw, vhash, vhash, 64 );
@@ -432,7 +432,7 @@ int sonoa_8way_hash( void *state, const void *input, int thrid )
     sph_fugue512_full( &ctx.fugue, hash6, hash6, 64 );
     sph_fugue512_full( &ctx.fugue, hash7, hash7, 64 );

-     if ( work_restart[thrid].restart ) return 0;
+     if ( work_restart[thr_id].restart ) return 0;
 // 4

     intrlv_8x64_512( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
@@ -630,7 +630,7 @@ int sonoa_8way_hash( void *state, const void *input, int thrid )

 #endif

-     if ( work_restart[thrid].restart ) return 0;
+     if ( work_restart[thr_id].restart ) return 0;
 // 5

     bmw512_8way_full( &ctx.bmw, vhash, vhash, 64 );
@@ -783,7 +783,7 @@ int sonoa_8way_hash( void *state, const void *input, int thrid )
     sph_whirlpool512_full( &ctx.whirlpool, hash6, hash6, 64 );
     sph_whirlpool512_full( &ctx.whirlpool, hash7, hash7, 64 );

-     if ( work_restart[thrid].restart ) return 0;
+     if ( work_restart[thr_id].restart ) return 0;
 // 6

     intrlv_8x64_512( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
@@ -952,7 +952,7 @@ int sonoa_8way_hash( void *state, const void *input, int thrid )
     sph_whirlpool512_full( &ctx.whirlpool, hash6, hash6, 64 );
     sph_whirlpool512_full( &ctx.whirlpool, hash7, hash7, 64 );

-     if ( work_restart[thrid].restart ) return 0;
+     if ( work_restart[thr_id].restart ) return 0;
 // 7

     intrlv_8x64_512( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
@@ -1117,49 +1117,6 @@ int sonoa_8way_hash( void *state, const void *input, int thrid )

     return 1;
 }
-     
-int scanhash_sonoa_8way( struct work *work, uint32_t max_nonce,
-                       uint64_t *hashes_done, struct thr_info *mythr )
-{
-   uint32_t hash[8*16] __attribute__ ((aligned (128)));
-   uint32_t vdata[20*8] __attribute__ ((aligned (64)));
-   uint32_t lane_hash[8] __attribute__ ((aligned (64)));
-   uint32_t *hashd7 = &(hash[7<<3]);
-   uint32_t *pdata = work->data;
-   const uint32_t *ptarget = work->target;
-   const uint32_t first_nonce = pdata[19];
-   const uint32_t last_nonce = max_nonce - 8;
-   __m512i  *noncev = (__m512i*)vdata + 9;   // aligned
-   uint32_t n = first_nonce;
-   const int thr_id = mythr->id;
-   const uint32_t targ32 = ptarget[7];
-
-   mm512_bswap32_intrlv80_8x64( vdata, pdata );
-   *noncev = mm512_intrlv_blend_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 );
-
-   do
-   {
-      if ( sonoa_8way_hash( hash, vdata, thr_id ) )
-      for ( int lane = 0; lane < 8; lane++ )
-      if unlikely( ( hashd7[ lane ] <= targ32 ) )
-      {
-         extr_lane_8x32( lane_hash, hash, lane, 256 );
-         if ( likely( valid_hash( lane_hash, ptarget ) && !opt_benchmark ) )
-         {
-            pdata[19] = bswap_32( n + lane );
-            submit_solution( work, lane_hash, mythr );
-         }
-      }
-      *noncev = _mm512_add_epi32( *noncev,
-                                  m512_const1_64( 0x0000000800000000 ) );
-      n += 8;
-   } while ( likely( ( n < last_nonce ) && !work_restart[thr_id].restart ) );
-   pdata[19] = n;
-   *hashes_done = n - first_nonce;
-   return 0;
-}

 #elif defined(SONOA_4WAY)

@@ -1186,7 +1143,7 @@ union _sonoa_4way_context_overlay

 typedef union _sonoa_4way_context_overlay sonoa_4way_context_overlay;

-int sonoa_4way_hash( void *state, const void *input, int thrid )
+int sonoa_4way_hash( void *state, const void *input, int thr_id )
 {
     uint64_t hash0[8] __attribute__ ((aligned (64)));
     uint64_t hash1[8] __attribute__ ((aligned (64)));
@@ -1250,7 +1207,7 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     echo_full( &ctx.echo, (BitSequence *)hash3, 512,
                     (const BitSequence *)hash3, 64 );
     
-     if ( work_restart[thrid].restart ) return 0;
+     if ( work_restart[thr_id].restart ) return 0;
 // 2

     intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );
@@ -1310,7 +1267,7 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     hamsi512_4way_update( &ctx.hamsi, vhash, 64 );
     hamsi512_4way_close( &ctx.hamsi, vhash );

-     if ( work_restart[thrid].restart ) return 0;
+     if ( work_restart[thr_id].restart ) return 0;
 // 3

     bmw512_4way_init( &ctx.bmw );
@@ -1375,7 +1332,7 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     sph_fugue512_full( &ctx.fugue, hash2, hash2, 64 );
     sph_fugue512_full( &ctx.fugue, hash3, hash3, 64 );

-     if ( work_restart[thrid].restart ) return 0;
+     if ( work_restart[thr_id].restart ) return 0;
 // 4
     intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );

@@ -1472,7 +1429,7 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     shavite512_2way_init( &ctx.shavite );
     shavite512_2way_update_close( &ctx.shavite, vhashB, vhashB, 64 );

-     if ( work_restart[thrid].restart ) return 0;
+     if ( work_restart[thr_id].restart ) return 0;
 // 5
     rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );

@@ -1557,7 +1514,7 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     sph_whirlpool512_full( &ctx.whirlpool, hash2, hash2, 64 );
     sph_whirlpool512_full( &ctx.whirlpool, hash3, hash3, 64 );

-     if ( work_restart[thrid].restart ) return 0;
+     if ( work_restart[thr_id].restart ) return 0;
 // 6

     intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );
@@ -1650,7 +1607,7 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     sph_whirlpool512_full( &ctx.whirlpool, hash2, hash2, 64 );
     sph_whirlpool512_full( &ctx.whirlpool, hash3, hash3, 64 );

-     if ( work_restart[thrid].restart ) return 0;    
+     if ( work_restart[thr_id].restart ) return 0;    
 // 7

     intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );
@@ -1745,46 +1702,4 @@ int sonoa_4way_hash( void *state, const void *input, int thrid )
     return 1;
 }

-int scanhash_sonoa_4way( struct work *work, const uint32_t max_nonce,
-	            uint64_t *hashes_done, struct thr_info *mythr )
-{
-     uint32_t hash[4*16] __attribute__ ((aligned (64)));
-     uint32_t vdata[24*4] __attribute__ ((aligned (64)));
-     uint32_t lane_hash[8] __attribute__ ((aligned (32)));
-     uint32_t *hashd7 = &( hash[7<<2] );
-     uint32_t *pdata = work->data;
-     const uint32_t *ptarget = work->target;
-     const uint32_t first_nonce = pdata[19];
-     const uint32_t last_nonce = max_nonce - 4;
-     const uint32_t targ32 = ptarget[7];
-     uint32_t n = first_nonce;
-     __m256i  *noncev = (__m256i*)vdata + 9;  
-     const int thr_id = mythr->id;
-
-     mm256_bswap32_intrlv80_4x64( vdata, pdata );
-     *noncev = mm256_intrlv_blend_32(
-                   _mm256_set_epi32( n+3, 0, n+2, 0, n+1, 0, n, 0 ), *noncev );
-
-     do
-     {
-        if ( sonoa_4way_hash( hash, vdata, thr_id ) )
-        for ( int lane = 0; lane < 4; lane++ )
-        if ( unlikely( hashd7[ lane ] <= targ32 ) )
-        {
-           extr_lane_4x32( lane_hash, hash, lane, 256 );
-           if ( likely( valid_hash( lane_hash, ptarget ) && !opt_benchmark ) )
-           {
-              pdata[19] = bswap_32( n + lane );
-              submit_solution( work, lane_hash, mythr );
-           }
-        }
-        *noncev = _mm256_add_epi32( *noncev,
-                                    m256_const1_64( 0x0000000400000000 ) );
-        n += 4;
-     } while ( likely( ( n < last_nonce ) && !work_restart[thr_id].restart ) );
-     pdata[19] = n;
-     *hashes_done = n - first_nonce;
-     return 0;
-}
-
 #endif
--- a/algo/x17/sonoa-gate.c
+++ b/algo/x17/sonoa-gate.c
@@ -3,15 +3,14 @@
 bool register_sonoa_algo( algo_gate_t* gate )
 {
 #if defined (SONOA_8WAY)
-  gate->scanhash  = (void*)&scanhash_sonoa_8way;
-//  gate->hash      = (void*)&sonoa_8way_hash;
+  gate->scanhash  = (void*)&scanhash_8way_64in_32out;
+  gate->hash      = (void*)&sonoa_8way_hash;
 #elif defined (SONOA_4WAY)
-  gate->scanhash  = (void*)&scanhash_sonoa_4way;
-//  gate->hash      = (void*)&sonoa_4way_hash;
+  gate->scanhash  = (void*)&scanhash_4way_64in_32out;
+  gate->hash      = (void*)&sonoa_4way_hash;
 #else
  init_sonoa_ctx();
-  gate->scanhash  = (void*)&scanhash_sonoa;
-//  gate->hash      = (void*)&sonoa_hash;
+  gate->hash      = (void*)&sonoa_hash;
 #endif
  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT | VAES_OPT;
  return true;
--- a/algo/x17/sonoa-gate.h
+++ b/algo/x17/sonoa-gate.h
@@ -14,21 +14,15 @@ bool register_sonoa_algo( algo_gate_t* gate );

 #if defined(SONOA_8WAY)

-int sonoa_8way_hash( void *state, const void *input, int thrid );
-int scanhash_sonoa_8way( struct work *work, uint32_t max_nonce,
-                         uint64_t *hashes_done, struct thr_info *mythr );
+int sonoa_8way_hash( void *state, const void *input, int thr_id );

 #elif defined(SONOA_4WAY)

-int sonoa_4way_hash( void *state, const void *input, int thrid );
-int scanhash_sonoa_4way( struct work *work, uint32_t max_nonce,
-                         uint64_t *hashes_done, struct thr_info *mythr );
+int sonoa_4way_hash( void *state, const void *input, int thr_id );

 #else

-int sonoa_hash( void *state, const void *input, int thrid );
-int scanhash_sonoa( struct work *work, uint32_t max_nonce,
-                  uint64_t *hashes_done, struct thr_info *mythr );
+int sonoa_hash( void *state, const void *input, int thr_id );
 void init_sonoa_ctx();

 #endif
--- a/algo/x17/sonoa.c
+++ b/algo/x17/sonoa.c
@@ -83,7 +83,7 @@ void init_sonoa_ctx()
        sph_haval256_5_init(&sonoa_ctx.haval);
 };

-int sonoa_hash( void *state, const void *input, int thrid )
+int sonoa_hash( void *state, const void *input, int thr_id )
 {
 	uint8_t hash[128] __attribute__ ((aligned (64)));
   sonoa_ctx_holder ctx __attribute__ ((aligned (64)));
@@ -132,7 +132,7 @@ int sonoa_hash( void *state, const void *input, int thrid )
   sph_echo512_close(&ctx.echo, hash);
 #endif

-   if ( work_restart[thrid].restart ) return 0;
+   if ( work_restart[thr_id].restart ) return 0;
 //

   sph_bmw512_init( &ctx.bmw);
@@ -190,7 +190,7 @@ int sonoa_hash( void *state, const void *input, int thrid )
   sph_hamsi512(&ctx.hamsi, hash, 64);
   sph_hamsi512_close(&ctx.hamsi, hash);
 	
-   if ( work_restart[thrid].restart ) return 0;
+   if ( work_restart[thr_id].restart ) return 0;
 //

   sph_bmw512_init( &ctx.bmw);
@@ -252,7 +252,7 @@ int sonoa_hash( void *state, const void *input, int thrid )
   sph_fugue512(&ctx.fugue, hash, 64);
   sph_fugue512_close(&ctx.fugue, hash);

-   if ( work_restart[thrid].restart ) return 0;
+   if ( work_restart[thr_id].restart ) return 0;
 //

   sph_bmw512_init( &ctx.bmw);
@@ -336,7 +336,7 @@ int sonoa_hash( void *state, const void *input, int thrid )
   sph_shavite512(&ctx.shavite, hash, 64);
   sph_shavite512_close(&ctx.shavite, hash);

-   if ( work_restart[thrid].restart ) return 0;
+   if ( work_restart[thr_id].restart ) return 0;
 //

   sph_bmw512_init( &ctx.bmw);
@@ -410,7 +410,7 @@ int sonoa_hash( void *state, const void *input, int thrid )
   sph_whirlpool(&ctx.whirlpool, hash, 64);
   sph_whirlpool_close(&ctx.whirlpool, hash);

-   if ( work_restart[thrid].restart ) return 0;
+   if ( work_restart[thr_id].restart ) return 0;
 //
   sph_bmw512_init( &ctx.bmw);
   sph_bmw512(&ctx.bmw, hash, 64);
@@ -487,7 +487,7 @@ int sonoa_hash( void *state, const void *input, int thrid )
   sph_whirlpool(&ctx.whirlpool, hash, 64);
   sph_whirlpool_close(&ctx.whirlpool, hash);

-   if ( work_restart[thrid].restart ) return 0;
+   if ( work_restart[thr_id].restart ) return 0;
 //

   sph_bmw512_init( &ctx.bmw);
@@ -569,34 +569,4 @@ int sonoa_hash( void *state, const void *input, int thrid )
   return 1;
 }

-int scanhash_sonoa( struct work *work, uint32_t max_nonce,
-             uint64_t *hashes_done, struct thr_info *mythr)
-{
-   uint32_t edata[20] __attribute__((aligned(64)));
-   uint32_t hash64[8] __attribute__((aligned(64)));
-   uint32_t *pdata = work->data;
-   uint32_t *ptarget = work->target;
-   uint32_t n = pdata[19];
-   const uint32_t first_nonce = pdata[19];
-   const int thr_id = mythr->id;
-   const bool bench = opt_benchmark;
-
-   mm128_bswap32_80( edata, pdata );
-
-   do
-   {
-      edata[19] = n;
-      if ( sonoa_hash( hash64, edata, thr_id ) )
-      if ( unlikely( valid_hash( hash64, ptarget ) && !bench ) )
-      {
-         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;
-   pdata[19] = n;
-   return 0;
-}
-
 #endif
--- a/algo/x17/x17-4way.c
+++ b/algo/x17/x17-4way.c
@@ -57,7 +57,7 @@ union _x17_8way_context_overlay
 } __attribute__ ((aligned (64)));
 typedef union _x17_8way_context_overlay x17_8way_context_overlay;

-void x17_8way_hash( void *state, const void *input )
+int x17_8way_hash( void *state, const void *input, int thr_id )
 {
     uint64_t vhash[8*8] __attribute__ ((aligned (128)));
     uint64_t vhashA[8*8] __attribute__ ((aligned (64)));
@@ -230,50 +230,8 @@ void x17_8way_hash( void *state, const void *input )
     haval256_5_8way_init( &ctx.haval );
     haval256_5_8way_update( &ctx.haval, vhashA, 64 );
     haval256_5_8way_close( &ctx.haval, state );
-}

-int scanhash_x17_8way( struct work *work, uint32_t max_nonce,
-                       uint64_t *hashes_done, struct thr_info *mythr )
-{
-   uint32_t hash32[8*8] __attribute__ ((aligned (128)));
-   uint32_t vdata[20*8] __attribute__ ((aligned (64)));
-   uint32_t lane_hash[8] __attribute__ ((aligned (64)));
-   uint32_t *hash32_d7 = &(hash32[7*8]);
-   uint32_t *pdata = work->data;
-   const uint32_t *ptarget = work->target;
-   const uint32_t first_nonce = pdata[19];
-   const uint32_t last_nonce = max_nonce - 8;
-   __m512i  *noncev = (__m512i*)vdata + 9; 
-   uint32_t n = first_nonce;
-   const int thr_id = mythr->id;
-   const uint32_t targ32_d7 = ptarget[7];
-   const bool bench = opt_benchmark;
-
-   mm512_bswap32_intrlv80_8x64( vdata, pdata );
-   *noncev = mm512_intrlv_blend_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 );
-   do
-   {
-      x17_8way_hash( hash32, vdata );
-
-      for ( int lane = 0; lane < 8; lane++ )
-      if ( unlikely( ( hash32_d7[ lane ] <= targ32_d7 ) && !bench ) )
-      {
-         extr_lane_8x32( lane_hash, hash32, lane, 256 );
-         if ( likely( valid_hash( lane_hash, ptarget ) ) )
-         {
-            pdata[19] = bswap_32( n + lane );
-            submit_solution( work, lane_hash, mythr );
-         }
-      }
-      *noncev = _mm512_add_epi32( *noncev,
-                                  m512_const1_64( 0x0000000800000000 ) );
-      n += 8;
-   } while ( likely( ( n < last_nonce ) && !work_restart[thr_id].restart ) );
-   pdata[19] = n;
-   *hashes_done = n - first_nonce;
-   return 0;
+     return 1;
 }

 #elif defined(X17_4WAY)
@@ -300,7 +258,7 @@ union _x17_4way_context_overlay
 };  
 typedef union _x17_4way_context_overlay x17_4way_context_overlay;

-void x17_4way_hash( void *state, const void *input )
+int x17_4way_hash( void *state, const void *input, int thr_id )
 {
     uint64_t vhash[8*4] __attribute__ ((aligned (64)));
     uint64_t vhashA[8*4] __attribute__ ((aligned (64)));
@@ -399,49 +357,8 @@ void x17_4way_hash( void *state, const void *input )
     haval256_5_4way_init( &ctx.haval );
     haval256_5_4way_update( &ctx.haval, vhashB, 64 );
     haval256_5_4way_close( &ctx.haval, state );
-}

-int scanhash_x17_4way( struct work *work, uint32_t max_nonce,
-                       uint64_t *hashes_done, struct thr_info *mythr )
-{
-   uint32_t hash32[8*4] __attribute__ ((aligned (64)));
-   uint32_t vdata[20*4] __attribute__ ((aligned (64)));
-   uint32_t lane_hash[8] __attribute__ ((aligned (64)));
-   uint32_t *hash32_d7 = &(hash32[ 7*4 ]);
-   uint32_t *pdata = work->data;
-   const uint32_t *ptarget = work->target;
-   const uint32_t first_nonce = pdata[19];
-   const uint32_t last_nonce = max_nonce - 4;
-   __m256i  *noncev = (__m256i*)vdata + 9;
-   uint32_t n = first_nonce;
-   const int thr_id = mythr->id;
-   const uint32_t targ32_d7 = ptarget[7];
-   const bool bench = opt_benchmark;
-
-   mm256_bswap32_intrlv80_4x64( vdata, pdata );
-   *noncev = mm256_intrlv_blend_32(
-                   _mm256_set_epi32( n+3, 0, n+2, 0, n+1, 0, n, 0 ), *noncev );
-   do
-   {
-      x17_4way_hash( hash32, vdata );
-
-      for ( int lane = 0; lane < 4; lane++ )
-      if ( unlikely( hash32_d7[ lane ] <= targ32_d7 && !bench ) )
-      {  
-         extr_lane_4x32( lane_hash, hash32, lane, 256 );
-         if ( valid_hash( lane_hash, ptarget ) )
-         {
-            pdata[19] = bswap_32( n + lane );
-            submit_solution( work, lane_hash, mythr );
-         }            
-      }
-      *noncev = _mm256_add_epi32( *noncev,
-                                  m256_const1_64( 0x0000000400000000 ) );
-      n += 4;
-   } while ( likely( ( n <= last_nonce ) && !work_restart[thr_id].restart ) );
-   pdata[19] = n;
-   *hashes_done = n - first_nonce;
-   return 0;
+     return 1;
 }

 #endif
--- a/algo/x17/x17-gate.c
+++ b/algo/x17/x17-gate.c
@@ -3,13 +3,12 @@
 bool register_x17_algo( algo_gate_t* gate )
 {
 #if defined (X17_8WAY)
-  gate->scanhash  = (void*)&scanhash_x17_8way;
+  gate->scanhash  = (void*)&scanhash_8way_64in_32out;
  gate->hash      = (void*)&x17_8way_hash;
 #elif defined (X17_4WAY)
-  gate->scanhash  = (void*)&scanhash_x17_4way;
+  gate->scanhash  = (void*)&scanhash_4way_64in_32out;
  gate->hash      = (void*)&x17_4way_hash;
 #else
-  gate->scanhash  = (void*)&scanhash_x17;
  gate->hash      = (void*)&x17_hash;
 #endif
  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT | VAES_OPT;
--- a/algo/x17/x17-gate.h
+++ b/algo/x17/x17-gate.h
@@ -14,20 +14,15 @@ bool register_x17_algo( algo_gate_t* gate );

 #if defined(X17_8WAY)

-void x17_8way_hash( void *state, const void *input );
-int scanhash_x17_8way( struct work *work, uint32_t max_nonce,
-                       uint64_t *hashes_done, struct thr_info *mythr );
+int x17_8way_hash( void *state, const void *input, int thr_id );
+
 #elif defined(X17_4WAY)

-void x17_4way_hash( void *state, const void *input );
-int scanhash_x17_4way( struct work *work, uint32_t max_nonce,
-                       uint64_t *hashes_done, struct thr_info *mythr );
+int x17_4way_hash( void *state, const void *input, int thr_id );

 #endif

-void x17_hash( void *state, const void *input );
-int scanhash_x17( struct work *work, uint32_t max_nonce,
-                  uint64_t *hashes_done, struct thr_info *mythr );
+int x17_hash( void *state, const void *input, int thr_id );

 #endif

--- a/algo/x17/x17.c
+++ b/algo/x17/x17.c
@@ -56,7 +56,7 @@ union _x17_context_overlay
 };
 typedef union _x17_context_overlay x17_context_overlay;

-void x17_hash(void *output, const void *input)
+int x17_hash(void *output, const void *input, int thr_id )
 {
 //    unsigned char hash[64 * 4] __attribute__((aligned(64))) = {0};
    unsigned char hash[64] __attribute__((aligned(64)));
@@ -143,36 +143,8 @@ void x17_hash(void *output, const void *input)
    sph_haval256_5_init(&ctx.haval);
    sph_haval256_5( &ctx.haval, (const void*)hash, 64 );
    sph_haval256_5_close( &ctx.haval, output );
-}

-int scanhash_x17( struct work *work, uint32_t max_nonce,
-	          uint64_t *hashes_done, struct thr_info *mythr)
-{
-   uint32_t edata[20] __attribute__((aligned(64)));
-   uint32_t hash64[8] __attribute__((aligned(64)));
-   uint32_t *pdata = work->data;
-   uint32_t *ptarget = work->target;
-   uint32_t n = pdata[19] - 1;
-   const uint32_t first_nonce = pdata[19];
-   const int thr_id = mythr->id;
-   const bool bench = opt_benchmark;
-
-   mm128_bswap32_80( edata, pdata );
-   
-   do
-   {
-      edata[19] = n;
-      x17_hash( hash64, edata );
-      if ( unlikely( valid_hash( hash64, ptarget ) && !bench ) )
-      {
-         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;
-   pdata[19] = n;
-   return 0;
+    return 1;
 }

 #endif
--- a/algo/x17/xevan-4way.c
+++ b/algo/x17/xevan-4way.c
@@ -57,7 +57,7 @@ union _xevan_8way_context_overlay
 } __attribute__ ((aligned (64)));
 typedef union _xevan_8way_context_overlay xevan_8way_context_overlay;

-void xevan_8way_hash( void *output, const void *input )
+int xevan_8way_hash( void *output, const void *input, int thr_id )
 {
     uint64_t vhash[16<<3] __attribute__ ((aligned (128)));
     uint64_t vhashA[16<<3] __attribute__ ((aligned (64)));
@@ -395,50 +395,8 @@ void xevan_8way_hash( void *output, const void *input )
     haval256_5_8way_init( &ctx.haval );
     haval256_5_8way_update( &ctx.haval, vhashA, dataLen );
     haval256_5_8way_close( &ctx.haval, output );
-}

-int scanhash_xevan_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)));
-   uint32_t *hashd7 = &(hash[7*8]);
-   uint32_t *pdata = work->data;
-   const uint32_t *ptarget = work->target;
-   const uint32_t first_nonce = pdata[19];
-   const uint32_t last_nonce = max_nonce - 8;
-   __m512i  *noncev = (__m512i*)vdata + 9;
-   uint32_t n = first_nonce;
-   const int thr_id = mythr->id;
-   const uint32_t targ32 = ptarget[7];
-   const bool bench = opt_benchmark;
-
-   mm512_bswap32_intrlv80_8x64( vdata, pdata );
-   *noncev = mm512_intrlv_blend_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 );
-   do
-   {
-      xevan_8way_hash( hash, vdata );
-
-      for ( int lane = 0; lane < 8; lane++ )
-      if ( unlikely( ( hashd7[ lane ] <= targ32 ) && !bench ) )
-      {
-         extr_lane_8x32( lane_hash, hash, lane, 256 );
-         if ( likely( valid_hash( lane_hash, ptarget ) ) )
-         {
-            pdata[19] = bswap_32( n + lane );
-            submit_solution( work, lane_hash, mythr );
-         }
-      }
-      *noncev = _mm512_add_epi32( *noncev,
-                                  m512_const1_64( 0x0000000800000000 ) );
-      n += 8;
-   } while ( likely( ( n < last_nonce ) && !work_restart[thr_id].restart ) );
-   pdata[19] = n;
-   *hashes_done = n - first_nonce;
-   return 0;
+     return 1;
 }

 #elif defined(XEVAN_4WAY)
@@ -465,7 +423,7 @@ union _xevan_4way_context_overlay
 };
 typedef union _xevan_4way_context_overlay xevan_4way_context_overlay;

-void xevan_4way_hash( void *output, const void *input )
+int xevan_4way_hash( void *output, const void *input, int thr_id )
 {
     uint64_t hash0[16] __attribute__ ((aligned (64)));
     uint64_t hash1[16] __attribute__ ((aligned (64)));
@@ -666,49 +624,8 @@ void xevan_4way_hash( void *output, const void *input )
     haval256_5_4way_init( &ctx.haval );
     haval256_5_4way_update( &ctx.haval, vhashA, dataLen );
     haval256_5_4way_close( &ctx.haval, output );
-}

-int scanhash_xevan_4way( struct work *work, uint32_t max_nonce,
-                         uint64_t *hashes_done, struct thr_info *mythr )
-{
-   uint32_t hash[16*4] __attribute__ ((aligned (128)));
-   uint32_t vdata[20*4] __attribute__ ((aligned (64)));
-   uint32_t lane_hash[8] __attribute__ ((aligned (64)));
-   uint32_t *hashd7 = &(hash[7<<2]);
-   uint32_t *pdata = work->data;
-   uint32_t *ptarget = work->target;
-   int thr_id = mythr->id;
-   __m256i  *noncev = (__m256i*)vdata + 9; 
-   const uint32_t targ32 = ptarget[7];
-   const uint32_t first_nonce = pdata[19];
-   const uint32_t last_nonce = max_nonce - 4;
-   uint32_t n = first_nonce;
-   const bool bench = opt_benchmark;
-
-   if ( bench )  ptarget[7] = 0x0cff;
-
-   mm256_bswap32_intrlv80_4x64( vdata, pdata );
-   *noncev = mm256_intrlv_blend_32(
-                   _mm256_set_epi32( n+3, 0, n+2, 0, n+1, 0, n, 0 ), *noncev );
-   do {
-      xevan_4way_hash( hash, vdata );
-      for ( int lane = 0; lane < 4; lane++ )
-      if ( unlikely( hashd7[ lane ] <= targ32 ) && ! bench )
-      {
-         extr_lane_4x32( lane_hash, hash, lane, 256 );
-	      if ( valid_hash( lane_hash, ptarget ) )
-         {
-             pdata[19] = bswap_32( n + lane );
-             submit_solution( work, lane_hash, mythr );
-         }
-      }
-      *noncev = _mm256_add_epi32( *noncev,
-                                  m256_const1_64( 0x0000000400000000 ) );
-      n += 4;
-   } while ( likely( ( n < last_nonce ) && !work_restart[thr_id].restart ) );
-   pdata[19] = n;
-   *hashes_done = n - first_nonce;
-   return 0;
+     return 1;
 }

 #endif
--- a/algo/x17/xevan-gate.c
+++ b/algo/x17/xevan-gate.c
@@ -3,14 +3,13 @@
 bool register_xevan_algo( algo_gate_t* gate )
 {
 #if defined (XEVAN_8WAY)
-  gate->scanhash  = (void*)&scanhash_xevan_8way;
+  gate->scanhash  = (void*)&scanhash_8way_64in_32out;
  gate->hash      = (void*)&xevan_8way_hash;
 #elif defined (XEVAN_4WAY)
-  gate->scanhash  = (void*)&scanhash_xevan_4way;
+  gate->scanhash  = (void*)&scanhash_4way_64in_32out;
  gate->hash      = (void*)&xevan_4way_hash;
 #else
  init_xevan_ctx();
-  gate->scanhash  = (void*)&scanhash_xevan;
  gate->hash      = (void*)&xevan_hash;
 #endif
  gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT | VAES_OPT;
--- a/algo/x17/xevan-gate.h
+++ b/algo/x17/xevan-gate.h
@@ -14,26 +14,15 @@ bool register_xevan_algo( algo_gate_t* gate );

 #if defined(XEVAN_8WAY)

-void xevan_8way_hash( void *state, const void *input );
+int xevan_8way_hash( void *state, const void *input, int thr_id );

-int scanhash_xevan_8way( struct work *work, uint32_t max_nonce,
-                       uint64_t *hashes_done, struct thr_info *mythr );
 #elif defined(XEVAN_4WAY)

-void xevan_4way_hash( void *state, const void *input );
-
-int scanhash_xevan_4way( struct work *work, uint32_t max_nonce,
-                       uint64_t *hashes_done, struct thr_info *mythr );
-
-//void init_xevan_4way_ctx();
+int xevan_4way_hash( void *state, const void *input, int thr_id );

 #else

-void xevan_hash( void *state, const void *input );
-
-int scanhash_xevan( struct work *work, uint32_t max_nonce,
-                  uint64_t *hashes_done, struct thr_info *mythr );
-
+int xevan_hash( void *state, const void *input, int trh_id );
 void init_xevan_ctx();

 #endif
--- a/algo/x17/xevan.c
+++ b/algo/x17/xevan.c
@@ -83,7 +83,7 @@ void init_xevan_ctx()
 #endif
 };

-void xevan_hash(void *output, const void *input)
+int xevan_hash(void *output, const void *input, int thr_id )
 {
   uint32_t _ALIGN(64) hash[32]; // 128 bytes required
 	const int dataLen = 128;
@@ -218,36 +218,8 @@ void xevan_hash(void *output, const void *input)
 	sph_haval256_5_close(&ctx.haval, hash);

 	memcpy(output, hash, 32);
-}

-int scanhash_xevan( struct work *work, uint32_t max_nonce,
-             uint64_t *hashes_done, struct thr_info *mythr)
-{
-   uint32_t edata[20] __attribute__((aligned(64)));
-   uint32_t hash64[8] __attribute__((aligned(64)));
-   uint32_t *pdata = work->data;
-   uint32_t *ptarget = work->target;
-   uint32_t n = pdata[19];
-   const uint32_t first_nonce = pdata[19];
-   const int thr_id = mythr->id;
-   const bool bench = opt_benchmark;
-
-   mm128_bswap32_80( edata, pdata );
-
-   do
-   {
-      edata[19] = n;
-      xevan_hash( hash64, edata );
-      if ( unlikely( valid_hash( hash64, ptarget ) && !bench ) )
-      {
-         pdata[19] = bswap_32( n );
-         submit_solution( work, hash64, mythr );
-      }
-      n++;
-   } while ( n < max_nonce && !work_restart[thr_id].restart );
-   pdata[19] = n;
-   *hashes_done = n - first_nonce;
-   return 0;
+   return 1;
 }

 #endif
--- a/algo/yescrypt/yescrypt.c
+++ b/algo/yescrypt/yescrypt.c
@@ -445,7 +445,7 @@ bool register_yescrypt_algo( algo_gate_t* gate )

   YESCRYPT_P = 1;

-   applog( LOG_NOTICE,"Yescrypt parameters: N= %d, R= %d.", YESCRYPT_N,
+   applog( LOG_NOTICE,"Yescrypt parameters: N= %d, R= %d", YESCRYPT_N,
                                                            YESCRYPT_R );
   if ( yescrypt_client_key )
     applog( LOG_NOTICE,"Key= \"%s\"\n", yescrypt_client_key );
--- a/algo/yespower/yespower-gate.c
+++ b/algo/yespower/yespower-gate.c
@@ -139,7 +139,7 @@ bool register_yespower_algo( algo_gate_t* gate )
     yespower_params.perslen = 0;
  }

-  applog( LOG_NOTICE,"Yespower parameters: N= %d, R= %d.", yespower_params.N,
+  applog( LOG_NOTICE,"Yespower parameters: N= %d, R= %d", yespower_params.N,
                                                           yespower_params.r );
  if ( yespower_params.pers )
     applog( LOG_NOTICE,"Key= \"%s\"\n", yespower_params.pers );
@@ -264,7 +264,7 @@ bool register_power2b_algo( algo_gate_t* gate )
  yespower_params.pers = "Now I am become Death, the destroyer of worlds";
  yespower_params.perslen = 46;

-  applog( LOG_NOTICE,"yespower-b2b parameters: N= %d, R= %d.", yespower_params.N,
+  applog( LOG_NOTICE,"yespower-b2b parameters: N= %d, R= %d", yespower_params.N,
                                                           yespower_params.r );
  applog( LOG_NOTICE,"Key= \"%s\"", yespower_params.pers );
  applog( LOG_NOTICE,"Key length= %d\n", yespower_params.perslen );
--- a/algo/yespower/yespower.h
+++ b/algo/yespower/yespower.h
@@ -76,7 +76,7 @@ typedef struct {
 	unsigned char uc[32];
 } yespower_binary_t __attribute__ ((aligned (64)));

-yespower_params_t yespower_params;
+extern yespower_params_t yespower_params;

 //SHA256_CTX sha256_prehash_ctx;
 extern __thread SHA256_CTX sha256_prehash_ctx;
--- a/build-allarch.sh
+++ b/build-allarch.sh
@@ -9,15 +9,15 @@ rm cpuminer-avx512-sha-vaes cpuminer-avx512 cpuminer-avx2 cpuminer-aes-avx cpumi
 make distclean || echo clean
 rm -f config.status
 ./autogen.sh || echo done
-CFLAGS="-O3 -march=icelake-client -Wall" ./configure --with-curl
-make -j 16
+CFLAGS="-O3 -march=icelake-client -Wall -fno-common" ./configure --with-curl
+make -j 8
 strip -s cpuminer.exe
 mv cpuminer.exe cpuminer-avx512-sha-vaes.exe
 strip -s cpuminer
 mv cpuminer cpuminer-avx512-sha-vaes

-CFLAGS="-O3 -march=skylake-avx512 -Wall" ./configure --with-curl
-make -j 16
+CFLAGS="-O3 -march=skylake-avx512 -Wall -fno-common" ./configure --with-curl
+make -j 8
 strip -s cpuminer.exe
 mv cpuminer.exe cpuminer-avx512.exe
 strip -s cpuminer
@@ -26,8 +26,8 @@ mv cpuminer cpuminer-avx512
 make clean || echo clean
 rm -f config.status
 # GCC 9 doesn't include AES with core-avx2
-CFLAGS="-O3 -march=core-avx2 -maes -Wall" ./configure --with-curl
-make -j 16
+CFLAGS="-O3 -march=core-avx2 -maes -Wall -fno-common" ./configure --with-curl
+make -j 8
 strip -s cpuminer.exe
 mv cpuminer.exe cpuminer-avx2.exe
 strip -s cpuminer
@@ -35,17 +35,17 @@ mv cpuminer cpuminer-avx2

 make clean || echo clean
 rm -f config.status
-CFLAGS="-O3 -march=corei7-avx -maes -Wall" ./configure --with-curl
-make -j 16
+CFLAGS="-O3 -march=corei7-avx -maes -Wall -fno-common" ./configure --with-curl
+make -j 8
 strip -s cpuminer.exe
 mv cpuminer.exe cpuminer-avx.exe
 strip -s cpuminer
-mv cpuminer cpuminer-aes-avx
+mv cpuminer cpuminer-avx

 make clean || echo clean
 rm -f config.status
-CFLAGS="-O3 -maes -msse4.2 -Wall" ./configure --with-curl
-make -j 16
+CFLAGS="-O3 -maes -msse4.2 -Wall -fno-common" ./configure --with-curl
+make -j 8
 strip -s cpuminer.exe
 mv cpuminer.exe cpuminer-aes-sse42.exe
 strip -s cpuminer
@@ -53,8 +53,8 @@ mv cpuminer cpuminer-aes-sse42

 #make clean || echo clean
 #rm -f config.status
-#CFLAGS="-O3 -march=corei7 -Wall" ./configure --with-curl
-#make -j 16
+#CFLAGS="-O3 -march=corei7 -Wall -fno-common" ./configure --with-curl
+#make -j 8
 #strip -s cpuminer.exe
 #mv cpuminer.exe cpuminer-sse42.exe
 #strip -s cpuminer
@@ -62,8 +62,8 @@ mv cpuminer cpuminer-aes-sse42

 #make clean || echo clean
 #rm -f config.status
-#CFLAGS="-O3 -march=core2 -Wall" ./configure --with-curl
-#make -j 16
+#CFLAGS="-O3 -march=core2 -Wall -fno-common" ./configure --with-curl
+#make -j 8
 #strip -s cpuminer.exe
 #mv cpuminer.exe cpuminer-ssse3.exe
 #strip -s cpuminer
@@ -71,8 +71,8 @@ mv cpuminer cpuminer-aes-sse42

 make clean || echo clean
 rm -f config.status
-CFLAGS="-O3 -msse2 -Wall" ./configure --with-curl
-make -j 16
+CFLAGS="-O3 -msse2 -Wall -fno-common" ./configure --with-curl
+make -j 8
 strip -s cpuminer.exe
 mv cpuminer.exe cpuminer-sse2.exe
 strip -s cpuminer
@@ -80,8 +80,8 @@ mv cpuminer cpuminer-sse2

 make clean || echo done
 rm -f config.status
-CFLAGS="-O3 -march=znver1 -Wall" ./configure --with-curl
-make -j 16
+CFLAGS="-O3 -march=znver1 -Wall -fno-common" ./configure --with-curl
+make -j 8
 strip -s cpuminer.exe
 mv cpuminer.exe cpuminer-zen.exe
 strip -s cpuminer
@@ -89,8 +89,8 @@ mv cpuminer cpuminer-zen

 make clean || echo done
 rm -f config.status
-CFLAGS="-O3 -march=native -Wall" ./configure --with-curl
-make -j 16
+CFLAGS="-O3 -march=native -Wall -fno-common" ./configure --with-curl
+make -j 8
 strip -s cpuminer.exe
 strip -s cpuminer

--- a/build-no-common.sh
+++ b/build-no-common.sh
@@ -0,0 +1,27 @@
+#!/bin/bash
+
+#if [ "$OS" = "Windows_NT" ]; then
+#    ./mingw64.sh
+#    exit 0
+#fi
+
+# Linux build
+
+make distclean || echo clean
+
+rm -f config.status
+./autogen.sh || echo done
+
+# Ubuntu 10.04 (gcc 4.4)
+# extracflags="-O3 -march=native -Wall -D_REENTRANT -funroll-loops -fvariable-expansion-in-unroller -fmerge-all-constants -fbranch-target-load-optimize2 -fsched2-use-superblocks -falign-loops=16 -falign-functions=16 -falign-jumps=16 -falign-labels=16"
+
+# Debian 7.7 / Ubuntu 14.04 (gcc 4.7+)
+#extracflags="$extracflags -Ofast -flto -fuse-linker-plugin -ftree-loop-if-convert-stores"
+
+#CFLAGS="-O3 -march=native -Wall" ./configure --with-curl --with-crypto=$HOME/usr
+CFLAGS="-O3 -march=native -Wall -fno-common" ./configure --with-curl
+#CFLAGS="-O3 -march=native -Wall" CXXFLAGS="$CFLAGS -std=gnu++11" ./configure --with-curl
+
+make -j 4
+
+strip -s cpuminer
--- a/clean-all.sh
+++ b/clean-all.sh
@@ -3,8 +3,8 @@
 # imake clean and rm all the targetted executables.
 # tips to users.

-rm cpuminer-avx512-sha-vaes cpuminer-avx512 cpuminer-avx2 cpuminer-aes-avx cpuminer-aes-sse42 cpuminer-sse2 cpuminer-zen  > /dev/null
+rm cpuminer-avx512-sha-vaes cpuminer-avx512 cpuminer-avx2 cpuminer-avx cpuminer-aes-sse42 cpuminer-sse2 cpuminer-zen  > /dev/null

-rm cpuminer-avx512-sha-vaes.exe cpuminer-avx512.exe cpuminer-avx2.exe cpuminer-aes-avx.exe cpuminer-aes-sse42.exe cpuminer-sse2.exe cpuminer-zen.exe  > /dev/null
+rm cpuminer-avx512-sha-vaes.exe cpuminer-avx512.exe cpuminer-avx2.exe cpuminer-avx.exe cpuminer-aes-sse42.exe cpuminer-sse2.exe cpuminer-zen.exe  > /dev/null

 make distclean > /dev/null
--- a/20
+++ b/20
@@ -1,6 +1,6 @@
 #! /bin/sh
 # Guess values for system-dependent variables and create Makefiles.
-# Generated by GNU Autoconf 2.69 for cpuminer-opt 3.13.1.
+# Generated by GNU Autoconf 2.69 for cpuminer-opt 3.14.3.
 #
 #
 # Copyright (C) 1992-1996, 1998-2012 Free Software Foundation, Inc.
@@ -577,8 +577,8 @@ MAKEFLAGS=
 # Identity of this package.
 PACKAGE_NAME='cpuminer-opt'
 PACKAGE_TARNAME='cpuminer-opt'
-PACKAGE_VERSION='3.13.1'
-PACKAGE_STRING='cpuminer-opt 3.13.1'
+PACKAGE_VERSION='3.14.3'
+PACKAGE_STRING='cpuminer-opt 3.14.3'
 PACKAGE_BUGREPORT=''
 PACKAGE_URL=''

@@ -1332,7 +1332,7 @@ if test "$ac_init_help" = "long"; then
  # Omit some internal or obsolete options to make the list less imposing.
  # This message is too long to be a string in the A/UX 3.1 sh.
  cat <<_ACEOF
-\`configure' configures cpuminer-opt 3.13.1 to adapt to many kinds of systems.
+\`configure' configures cpuminer-opt 3.14.3 to adapt to many kinds of systems.

 Usage: $0 [OPTION]... [VAR=VALUE]...

@@ -1404,7 +1404,7 @@ fi

 if test -n "$ac_init_help"; then
  case $ac_init_help in
-     short | recursive ) echo "Configuration of cpuminer-opt 3.13.1:";;
+     short | recursive ) echo "Configuration of cpuminer-opt 3.14.3:";;
   esac
  cat <<\_ACEOF

@@ -1509,7 +1509,7 @@ fi
 test -n "$ac_init_help" && exit $ac_status
 if $ac_init_version; then
  cat <<\_ACEOF
-cpuminer-opt configure 3.13.1
+cpuminer-opt configure 3.14.3
 generated by GNU Autoconf 2.69

 Copyright (C) 2012 Free Software Foundation, Inc.
@@ -2012,7 +2012,7 @@ cat >config.log <<_ACEOF
 This file contains any messages produced by compilers while
 running configure, to aid debugging if configure makes a mistake.

-It was created by cpuminer-opt $as_me 3.13.1, which was
+It was created by cpuminer-opt $as_me 3.14.3, which was
 generated by GNU Autoconf 2.69.  Invocation command line was

  $ $0 $@
@@ -2993,7 +2993,7 @@ fi

 # Define the identity of the package.
 PACKAGE='cpuminer-opt'
- VERSION='3.13.1'
+ VERSION='3.14.3'


 cat >>confdefs.h <<_ACEOF
@@ -6690,7 +6690,7 @@ cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1
 # report actual input values of CONFIG_FILES etc. instead of their
 # values after options handling.
 ac_log="
-This file was extended by cpuminer-opt $as_me 3.13.1, which was
+This file was extended by cpuminer-opt $as_me 3.14.3, which was
 generated by GNU Autoconf 2.69.  Invocation command line was

  CONFIG_FILES    = $CONFIG_FILES
@@ -6756,7 +6756,7 @@ _ACEOF
 cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1
 ac_cs_config="`$as_echo "$ac_configure_args" | sed 's/^ //; s/[\\""\`\$]/\\\\&/g'`"
 ac_cs_version="\\
-cpuminer-opt config.status 3.13.1
+cpuminer-opt config.status 3.14.3
 configured by $0, generated by GNU Autoconf 2.69,
  with options \\"\$ac_cs_config\\"

--- a/configure.ac
+++ b/configure.ac
@@ -1,4 +1,4 @@
-AC_INIT([cpuminer-opt], [3.13.1])
+AC_INIT([cpuminer-opt], [3.14.3])

 AC_PREREQ([2.59c])
 AC_CANONICAL_SYSTEM
--- a/cpu-miner.c
+++ b/cpu-miner.c
--- a/miner.h
+++ b/miner.h
@@ -83,6 +83,8 @@ enum {
 };
 #endif

+extern bool is_power_of_2( int n );
+
 static inline bool is_windows(void)
 {
 #ifdef WIN32
@@ -313,6 +315,10 @@ size_t address_to_script( unsigned char *out, size_t outsz, const char *addr );
 int    timeval_subtract( struct timeval *result, struct timeval *x,
                           struct timeval *y);

+// Segwit BEGIN
+extern void memrev(unsigned char *p, size_t len);
+// Segwit END
+
 // Bitcoin formula for converting difficulty to an equivalent
 // number of hashes.
 //
@@ -324,12 +330,12 @@ int    timeval_subtract( struct timeval *result, struct timeval *x,

 #define EXP16 65536.
 #define EXP32 4294967296.
-const long double exp32;  // 2**32
-const long double exp48;  // 2**48
-const long double exp64;  // 2**64
-const long double exp96;  // 2**96
-const long double exp128; // 2**128
-const long double exp160; // 2**160
+extern const long double exp32;  // 2**32
+extern const long double exp48;  // 2**48
+extern const long double exp64;  // 2**64
+extern const long double exp96;  // 2**96
+extern const long double exp128; // 2**128
+extern const long double exp160; // 2**160

 bool   fulltest( const uint32_t *hash, const uint32_t *target );
 bool   valid_hash( const void*, const void* );
@@ -374,36 +380,25 @@ void   cpu_brand_string( char* s );
 float cpu_temp( int core );
 */

-struct work {
+struct work
+{
+   uint32_t target[8] __attribute__ ((aligned (64)));
 	uint32_t data[48] __attribute__ ((aligned (64)));
-	uint32_t target[8] __attribute__ ((aligned (64)));
-
 	double targetdiff;
-//	double shareratio;
 	double sharediff;
   double stratum_diff;
-
 	int height;
 	char *txs;
 	char *workid;
-
 	char *job_id;
 	size_t xnonce2_len;
 	unsigned char *xnonce2;
   bool sapling;
   bool stale;
-
-   // x16rt
-   uint32_t merkleroothash[8];
-   uint32_t witmerkleroothash[8];
-   uint32_t denom10[8];
-   uint32_t denom100[8];
-   uint32_t denom1000[8];
-   uint32_t denom10000[8];
-
 } __attribute__ ((aligned (64)));

-struct stratum_job {
+struct stratum_job
+{
 	unsigned char prevhash[32];
   unsigned char final_sapling_hash[32];
   char *job_id;
@@ -417,7 +412,7 @@ struct stratum_job {
 	unsigned char ntime[4];
 	double diff;
   bool clean;
-   // for x16rt
+   // for x16rt-veil
   unsigned char extra[64];
   unsigned char denom10[32];
   unsigned char denom100[32];
@@ -752,6 +747,7 @@ extern double opt_diff_factor;
 extern double opt_target_factor;
 extern bool opt_randomize;
 extern bool allow_mininginfo;
+extern pthread_rwlock_t g_work_lock;
 extern time_t g_work_time;
 extern bool opt_stratum_stats;
 extern int num_cpus;
--- a/simd-utils/simd-512.h
+++ b/simd-utils/simd-512.h
@@ -375,10 +375,10 @@ static inline void memcpy_512( __m512i *dst, const __m512i *src, const int n )

 // Generic for odd rotations
 #define mm512_ror_x64( v, n )      _mm512_alignr_epi64( v, v, n )
-#define mm512_rol_x64( v, n )      _mm512_alignr_epi64( v, v, 8-n )
+#define mm512_rol_x64( v, n )      _mm512_alignr_epi64( v, v, 8-(n) )

 #define mm512_ror_x32( v, n )      _mm512_alignr_epi32( v, v, n )
-#define mm512_rol_x32( v, n )      _mm512_alignr_epi32( v, v, 16-n )
+#define mm512_rol_x32( v, n )      _mm512_alignr_epi32( v, v, 16-(n) )

 #define mm512_ror_1x16( v ) \
   _mm512_permutexvar_epi16( m512_const_64( \
--- a/util.c
+++ b/util.c
@@ -81,6 +81,15 @@ struct thread_q {
 	pthread_cond_t		cond;
 };

+bool is_power_of_2( int n ) 
+{ 
+  while ( n > 1 ) 
+  { 
+      if ( n % 2 != 0 ) return false; 
+      n = n / 2; 
+  } 
+  return true; 
+} 

 void applog2( int prio, const char *fmt, ... )
 {
@@ -609,6 +618,8 @@ json_t *json_rpc_call(CURL *curl, const char *url,
 		goto err_out;
 	}

+// want_stratum is useless, and so is this code it seems. Nothing in
+// hi appears to be set.   
 	/* If X-Stratum was found, activate Stratum */
 	if (want_stratum && hi.stratum_url &&
 	    !strncasecmp(hi.stratum_url, "stratum+tcp://", 14)) {
@@ -747,6 +758,19 @@ err_out:
 	return cfg;
 }

+// Segwit BEGIN
+void memrev(unsigned char *p, size_t len)
+{
+   unsigned char c, *q;
+   for (q = p + len - 1; p < q; p++, q--) {
+      c = *p;
+      *p = *q;
+      *q = c;
+   }
+}
+// Segwit END
+
+
 void cbin2hex(char *out, const char *in, size_t len)
 {
   if (out) {
--- a/winbuild-cross.sh
+++ b/winbuild-cross.sh
@@ -44,14 +44,14 @@ cp $LOCAL_LIB/curl/lib/.libs/libcurl-4.dll release/
 rm -f config.status
 ./autogen.sh || echo done
 CFLAGS="-O3 -march=icelake-client -Wall" ./configure $CONFIGURE_ARGS
-make -j 16
+make -j 8
 strip -s cpuminer.exe
 mv cpuminer.exe release/cpuminer-avx512-sha-vaes.exe

 make clean || echo clean
 rm -f config.status
 CFLAGS="-O3 -march=znver1 -Wall" ./configure $CONFIGURE_ARGS
-make -j 16
+make -j 8
 strip -s cpuminer.exe
 mv cpuminer.exe release/cpuminer-zen.exe

@@ -60,7 +60,7 @@ make clean || echo clean
 rm -f config.status
 CFLAGS="-O3 -march=skylake-avx512 -Wall" ./configure $CONFIGURE_ARGS
 #CFLAGS="-O3 -march=skylake-avx512 -Wall -fno-asynchronous-unwind-tables" ./configure $CONFIGURE_ARGS
-make -j 16
+make -j 8
 strip -s cpuminer.exe
 mv cpuminer.exe release/cpuminer-avx512.exe

@@ -68,7 +68,7 @@ make clean || echo clean
 rm -f config.status
 # GCC 9 doesn't include AES in -march=core-avx2
 CFLAGS="-O3 -march=core-avx2 -maes -Wall" ./configure $CONFIGURE_ARGS
-make -j 16
+make -j 8
 strip -s cpuminer.exe
 mv cpuminer.exe release/cpuminer-avx2.exe

@@ -76,7 +76,7 @@ make clean || echo clean
 rm -f config.status
 # -march=corei7-avx still includes aes, but just in case
 CFLAGS="-O3 -march=corei7-avx -maes -Wall" ./configure $CONFIGURE_ARGS 
-make -j 16
+make -j 8
 strip -s cpuminer.exe
 mv cpuminer.exe release/cpuminer-avx.exe

@@ -85,7 +85,7 @@ make clean || echo clean
 rm -f config.status
 CFLAGS="-O3 -march=westmere -Wall" ./configure $CONFIGURE_ARGS
 #CFLAGS="-O3 -maes -msse4.2 -Wall" ./configure $CONFIGURE_ARGS
-make -j 16
+make -j 8
 strip -s cpuminer.exe
 mv cpuminer.exe release/cpuminer-aes-sse42.exe

@@ -107,7 +107,7 @@ mv cpuminer.exe release/cpuminer-aes-sse42.exe
 make clean || echo clean
 rm -f config.status
 CFLAGS="-O3 -msse2 -Wall" ./configure $CONFIGURE_ARGS
-make -j 16
+make -j 8
 strip -s cpuminer.exe
 mv cpuminer.exe release/cpuminer-sse2.exe
 make clean || echo clean
Author	SHA1	Message	Date
Jay D Dee	cdd587537e	v3.14.3	2020-06-18 17:30:26 -04:00
Jay D Dee	51a1d91abd	v3.14.2	2020-05-30 21:20:44 -04:00
Jay D Dee	13563e2598	v3.14.1	2020-05-21 13:00:29 -04:00
Jay D Dee	9571f85d53	v3.14.0	2020-05-20 13:56:35 -04:00
Jay D Dee	0e69756634	v3.13.2-segwit-test	2020-05-18 18:17:27 -04:00
Jay D Dee	9653bca1e2	v3.13.1.1	2020-05-17 19:21:37 -04:00