update makefile to include all builds

.gitignore

@@ -60,3 +60,8 @@ CMakeUserPresets.json
 *.ptx
 *.cubin
 *.fatbin
+
+# Coin miner executables
+coin_miner_cpu
+coin_miner_simd
+coin_miner_cuda

aad_coin_miner_simd.c

@@ -30,39 +30,61 @@ static int is_valid_coin(u32_t *hash)
   return hash[0] == 0xAAD20250u;
 }
 
+//
+// increment coin variable part using the same logic as CPU miner
+// returns 0 if overflow (all positions wrapped around), 1 otherwise
+//
+static int increment_coin(u32_t coin[14])
+{
+  // Increment the variable part using byte-by-byte logic with carry
+  // Increment from the end to beginning (positions 53 down to 12)
+  int pos = 53;
+  while(pos >= 12)
+  {
+    u08_t *byte = &((u08_t *)coin)[pos ^ 3];
+    if(*byte == '\n' || *byte == 0x80)
+      *byte = 32; // Start from space
+
+    (*byte)++;
+
+    // Skip newline character
+    if(*byte == '\n')
+      (*byte)++;
+
+    // Wrap around at 127 (printable ASCII limit)
+    if(*byte >= 127)
+    {
+      *byte = 32; // Reset to space
+      pos--; // Carry to next position
+    }
+    else
+    {
+      break; // No carry needed
+    }
+  }
+  
+  // Return 0 if we carried all the way through (overflow), 1 otherwise
+  return (pos >= 12);
+}
+
 //
 // prepare interleaved data for SIMD processing
 //
-static void prepare_coins(u32_t base_coin[14], u32_t *interleaved_data, int simd_width, u64_t base_counter)
+static void prepare_coins(u32_t base_coin[14], u32_t *interleaved_data, int simd_width)
 {
   for(int lane = 0; lane < simd_width; lane++)
   {
     u32_t coin[14];
     memcpy(coin, base_coin, sizeof(coin));
 
-    // Modify the coin for this lane (encode counter in the variable part)
-    u64_t counter = base_counter + lane;
-    for(int i = 12; i < 20 && i < 54; i++)
-    {
-      int shift = (19 - i) * 8;
-      if(shift >= 0 && shift < 64)
-      {
-        u08_t byte = (counter >> shift) & 0xFF;
-        // Map to ASCII printable range (32-126, excluding newline position)
-        if(byte == '\n' || byte >= 0x80)
-          byte = 'X';
-        // Ensure byte is in printable ASCII range (32-126)
-        // Map all values to ASCII printable characters: space (32) to tilde (126)
-        // byte = 32 + (byte % 95);
-        ((u08_t *)coin)[i ^ 3] = byte;
-      }
-    }
-
     // Interleave the data
     for(int idx = 0; idx < 14; idx++)
     {
       interleaved_data[idx * simd_width + lane] = coin[idx];
     }
+    
+    // Increment the base coin for the next lane
+    increment_coin(base_coin);
   }
 }
 
@@ -125,19 +147,21 @@ static void mine_coins_avx(u64_t max_attempts)
   ((u08_t *)base_coin)[0x36 ^ 3] = '\n';
   ((u08_t *)base_coin)[0x37 ^ 3] = 0x80;
 
-  // Initialize variable part
+  // Initialize variable part with A-Z cycling pattern (same as CPU miner)
   for(int i = 12; i < 54; i++)
-    ((u08_t *)base_coin)[i ^ 3] = 'A';
+    ((u08_t *)base_coin)[i ^ 3] = 'A' + (i - 12) % 26;
 
   printf("Mining DETI coins using AVX (4-way SIMD)...\n");
   printf("Press Ctrl+C to stop\n\n");
 
   time_measurement();
+  time_measurement();
+  double start_time = wall_time_delta();
 
   while(keep_running && (max_attempts == 0 || attempts < max_attempts))
   {
     // Prepare coins for this batch
-    prepare_coins(base_coin, interleaved_data, SIMD_WIDTH, attempts);
+    prepare_coins(base_coin, interleaved_data, SIMD_WIDTH);
 
     // Compute SHA1 hashes
     sha1_avx((v4si *)interleaved_data, (v4si *)interleaved_hash);
@@ -163,13 +187,16 @@ static void mine_coins_avx(u64_t max_attempts)
     // Print progress every 1M attempts
     if(attempts % 1000000 < SIMD_WIDTH)
     {
-      printf("Attempts: %llu, Coins: %u\n",
-             (unsigned long long)attempts, coins_found);
+      time_measurement();
+      double current_time = wall_time_delta() - start_time;
+      double rate = attempts / current_time;
+      printf("Attempts: %llu, Rate: %.2f MH/s, Coins: %u\n",
+             (unsigned long long)attempts, rate / 1e6, coins_found);
     }
   }
 
   time_measurement();
-  double total_time = wall_time_delta();
+  double total_time = wall_time_delta() - start_time;
 
   printf("\n=== Mining Statistics ===\n");
   printf("Total attempts: %llu\n", (unsigned long long)attempts);
@@ -212,17 +239,20 @@ static void mine_coins_avx2(u64_t max_attempts)
   ((u08_t *)base_coin)[0x36 ^ 3] = '\n';
   ((u08_t *)base_coin)[0x37 ^ 3] = 0x80;
 
+  // Initialize variable part with A-Z cycling pattern (same as CPU miner)
   for(int i = 12; i < 54; i++)
-    ((u08_t *)base_coin)[i ^ 3] = 'A';
+    ((u08_t *)base_coin)[i ^ 3] = 'A' + (i - 12) % 26;
 
   printf("Mining DETI coins using AVX2 (8-way SIMD)...\n");
   printf("Press Ctrl+C to stop\n\n");
 
   time_measurement();
+  time_measurement();
+  double start_time = wall_time_delta();
 
   while(keep_running && (max_attempts == 0 || attempts < max_attempts))
   {
-    prepare_coins(base_coin, interleaved_data, SIMD_WIDTH, attempts);
+    prepare_coins(base_coin, interleaved_data, SIMD_WIDTH);
     sha1_avx2((v8si *)interleaved_data, (v8si *)interleaved_hash);
     attempts += SIMD_WIDTH;
 
@@ -244,13 +274,16 @@ static void mine_coins_avx2(u64_t max_attempts)
 
     if(attempts % 1000000 < SIMD_WIDTH)
     {
-      printf("Attempts: %llu, Coins: %u\n",
-             (unsigned long long)attempts, coins_found);
+      time_measurement();
+      double current_time = wall_time_delta() - start_time;
+      double rate = attempts / current_time;
+      printf("Attempts: %llu, Rate: %.2f MH/s, Coins: %u\n",
+             (unsigned long long)attempts, rate / 1e6, coins_found);
     }
   }
 
   time_measurement();
-  double total_time = wall_time_delta();
+  double total_time = wall_time_delta() - start_time;
 
   printf("\n=== Mining Statistics ===\n");
   printf("Total attempts: %llu\n", (unsigned long long)attempts);

makefile

@@ -32,7 +32,7 @@ OPENCL_DIR = $(CUDA_DIR)
 #   RTX 4070 -------------- sm_89
 #
 
-CUDA_ARCH = sm_75
+CUDA_ARCH = sm_86
 
 
 #
@@ -42,6 +42,8 @@ CUDA_ARCH = sm_75
 clean:
 	rm -f sha1_tests
 	rm -f sha1_cuda_test sha1_cuda_kernel.cubin
+	rm -f coin_miner_cpu coin_miner_simd coin_miner_cuda coin_miner_cuda_kernel.cubin
+	rm -f benchmark
 	rm -f a.out
 
 
@@ -64,3 +66,29 @@ sha1_cuda_kernel.cubin:			aad_sha1_cuda_kernel.cu aad_sha1.h makefile
 	nvcc -arch=$(CUDA_ARCH) --compiler-options -O2,-Wall -I$(CUDA_DIR)/include --cubin $< -o $@
 
 all: sha1_tests sha1_cuda_test sha1_cuda_kernel.cubin
+
+
+#
+# DETI coin miners
+#
+
+coin_miner_cpu:	aad_coin_miner_cpu.c aad_sha1.h aad_sha1_cpu.h aad_data_types.h aad_utilities.h aad_vault.h makefile
+	cc -march=native -Wall -Wshadow -Werror -O3 $< -o $@
+
+coin_miner_simd:	aad_coin_miner_simd.c aad_sha1.h aad_sha1_cpu.h aad_data_types.h aad_utilities.h aad_vault.h makefile
+	cc -march=native -Wall -Wshadow -Werror -O3 $< -o $@
+
+coin_miner_cuda_kernel.cubin:	aad_coin_miner_cuda_kernel.cu aad_sha1.h makefile
+	nvcc -arch=$(CUDA_ARCH) --compiler-options -O2,-Wall -I$(CUDA_DIR)/include --cubin $< -o $@
+
+coin_miner_cuda:	aad_coin_miner_cuda.c coin_miner_cuda_kernel.cubin aad_sha1.h aad_sha1_cpu.h aad_data_types.h aad_utilities.h aad_vault.h aad_cuda_utilities.h makefile
+	cc -march=native -Wall -Wshadow -Werror -O3 -I$(CUDA_DIR)/include $< -o $@ -lcuda
+
+benchmark:	aad_benchmark.c aad_sha1.h aad_sha1_cpu.h aad_data_types.h aad_utilities.h makefile
+	cc -march=native -Wall -Wshadow -Werror -O3 $< -o $@
+
+miners: coin_miner_cpu coin_miner_simd coin_miner_cuda benchmark
+
+all: 	sha1_tests sha1_cuda_test sha1_cuda_kernel.cubin \
+		coin_miner_cpu coin_miner_simd coin_miner_cuda coin_miner_cuda_kernel.cubin \
+		benchmark