Add files via upload

FOUND.txt

@@ -0,0 +1 @@
+0x000000000000000000000000000000000000000000000000006ABE1F9B67E114

IntGroup.cpp

@@ -5,7 +5,10 @@ using namespace std;
 
 IntGroup::IntGroup(int size) {
   this->size = size;
-  subp = (Int *)malloc(size * sizeof(Int));
+  // Use aligned allocation for better SIMD performance
+  if (posix_memalign((void**)&subp, 64, size * sizeof(Int)) != 0) {
+    subp = (Int *)malloc(size * sizeof(Int));
+  }
 }
 
 IntGroup::~IntGroup() {
@@ -23,18 +26,72 @@ void IntGroup::ModInv() {
   Int inverse;
 
   subp[0].Set(&ints[0]);
-  for (int i = 1; i < size; i++) {
-    subp[i].ModMulK1(&subp[i - 1], &ints[i]);
+
+  // Unroll small loops for better performance
+  if (size >= 8) {
+    for (int i = 1; i < size; i += 8) {
+      if (i < size) subp[i].ModMulK1(&subp[i - 1], &ints[i]);
+      if (i+1 < size) subp[i+1].ModMulK1(&subp[i], &ints[i+1]);
+      if (i+2 < size) subp[i+2].ModMulK1(&subp[i+1], &ints[i+2]);
+      if (i+3 < size) subp[i+3].ModMulK1(&subp[i+2], &ints[i+3]);
+      if (i+4 < size) subp[i+4].ModMulK1(&subp[i+3], &ints[i+4]);
+      if (i+5 < size) subp[i+5].ModMulK1(&subp[i+4], &ints[i+5]);
+      if (i+6 < size) subp[i+6].ModMulK1(&subp[i+5], &ints[i+6]);
+      if (i+7 < size) subp[i+7].ModMulK1(&subp[i+6], &ints[i+7]);
+    }
+  } else {
+    for (int i = 1; i < size; i++) {
+      subp[i].ModMulK1(&subp[i - 1], &ints[i]);
+    }
   }
 
   // Do the inversion
   inverse.Set(&subp[size - 1]);
   inverse.ModInv();
 
-  for (int i = size - 1; i > 0; i--) {
-    newValue.ModMulK1(&subp[i - 1], &inverse);
-    inverse.ModMulK1(&ints[i]);
-    ints[i].Set(&newValue);
+  // Unroll the back-substitution loop
+  if (size >= 8) {
+    for (int i = size - 1; i > 0; i -= 8) {
+      if (i >= 8) {
+        newValue.ModMulK1(&subp[i - 1], &inverse);
+        inverse.ModMulK1(&ints[i]);
+        ints[i].Set(&newValue);
+
+        newValue.ModMulK1(&subp[i - 2], &inverse);
+        inverse.ModMulK1(&ints[i-1]);
+        ints[i-1].Set(&newValue);
+
+        newValue.ModMulK1(&subp[i - 3], &inverse);
+        inverse.ModMulK1(&ints[i-2]);
+        ints[i-2].Set(&newValue);
+
+        newValue.ModMulK1(&subp[i - 4], &inverse);
+        inverse.ModMulK1(&ints[i-3]);
+        ints[i-3].Set(&newValue);
+
+        newValue.ModMulK1(&subp[i - 5], &inverse);
+        inverse.ModMulK1(&ints[i-4]);
+        ints[i-4].Set(&newValue);
+
+        newValue.ModMulK1(&subp[i - 6], &inverse);
+        inverse.ModMulK1(&ints[i-5]);
+        ints[i-5].Set(&newValue);
+
+        newValue.ModMulK1(&subp[i - 7], &inverse);
+        inverse.ModMulK1(&ints[i-6]);
+        ints[i-6].Set(&newValue);
+
+        newValue.ModMulK1(&subp[i - 8], &inverse);
+        inverse.ModMulK1(&ints[i-7]);
+        ints[i-7].Set(&newValue);
+      }
+    }
+  } else {
+    for (int i = size - 1; i > 0; i--) {
+      newValue.ModMulK1(&subp[i - 1], &inverse);
+      inverse.ModMulK1(&ints[i]);
+      ints[i].Set(&newValue);
+    }
   }
 
   ints[0].Set(&inverse);

IntMod.cpp

@@ -909,10 +909,7 @@ void Int::ModMulK1(Int *a, Int *b) {
   uint64_t ah, al;
   uint64_t t[NB64BLOCK];
 #if BISIZE==256
-  uint64_t r512[8];
-  r512[5] = 0;
-  r512[6] = 0;
-  r512[7] = 0;
+  uint64_t r512[8] = {0};  // Initialize to zero for better performance
 #else
   uint64_t r512[12];
   r512[5] = 0;

Makefile

@@ -0,0 +1,121 @@
+# ===================================================================
+# Pollard-Kangaroo Solver (Mark1) - Optimized Makefile
+# ===================================================================
+
+# Compiler and flags
+CXX = g++
+CXXFLAGS_COMMON = -std=c++17 -march=native -pthread -fopenmp
+CXXFLAGS_OPT = -O3 -funroll-loops -ftree-vectorize -fstrict-aliasing \
+               -fno-semantic-interposition -fvect-cost-model=unlimited \
+               -fno-trapping-math -fipa-ra -fipa-modref -flto \
+               -fassociative-math -fomit-frame-pointer -ffast-math \
+               -malign-data=cacheline -floop-nest-optimize \
+               -floop-unroll-and-jam -fpeel-loops \
+               -fvariable-expansion-in-unroller
+CXXFLAGS_SIMD = -mavx2 -mbmi2 -madx
+CXXFLAGS_DEBUG = -g -O0 -DDEBUG
+
+# Target names
+TARGET_MARK1 = Mark1
+TARGET_DP_ANALYZER = DP-analyzer
+
+# Source files
+SRCS_MARK1 = Mark1.cpp Int.cpp SECP256K1.cpp Point.cpp Random.cpp IntMod.cpp IntGroup.cpp Timer.cpp
+SRCS_DP_ANALYZER = DP-analyzer/DP-analyzer.cpp Int.cpp
+HDRS = Int.h Point.h SECP256K1.h IntGroup.h Timer.h Random.h hashutil.h simd_block_bloom.h
+
+# Build directories
+BUILD_DIR = build
+OBJ_DIR = $(BUILD_DIR)/obj
+BIN_DIR = $(BUILD_DIR)/bin
+
+# Object files
+OBJS_MARK1 = $(OBJ_DIR)/Mark1.o $(OBJ_DIR)/Int.o $(OBJ_DIR)/SECP256K1.o $(OBJ_DIR)/Point.o $(OBJ_DIR)/Random.o $(OBJ_DIR)/IntMod.o $(OBJ_DIR)/IntGroup.o $(OBJ_DIR)/Timer.o
+OBJS_DP_ANALYZER = $(OBJ_DIR)/DP-analyzer.o $(OBJ_DIR)/Int.o
+
+# Default target
+all: release
+
+# Release build (optimized)
+release: CXXFLAGS = $(CXXFLAGS_COMMON) $(CXXFLAGS_OPT) $(CXXFLAGS_SIMD)
+release: $(BIN_DIR)/$(TARGET_MARK1) $(BIN_DIR)/$(TARGET_DP_ANALYZER)
+
+# Debug build
+debug: CXXFLAGS = $(CXXFLAGS_COMMON) $(CXXFLAGS_DEBUG) $(CXXFLAGS_SIMD)
+debug: $(BIN_DIR)/$(TARGET_MARK1)_debug $(BIN_DIR)/$(TARGET_DP_ANALYZER)_debug
+
+# Profile build
+profile: CXXFLAGS = $(CXXFLAGS_COMMON) $(CXXFLAGS_OPT) $(CXXFLAGS_SIMD) -pg
+profile: $(BIN_DIR)/$(TARGET_MARK1)_profile
+
+# Create directories
+$(BUILD_DIR) $(OBJ_DIR) $(BIN_DIR):
+	mkdir -p $@
+
+# Compile object files (general rule, specific dependencies override)
+$(OBJ_DIR)/%.o: %.cpp | $(OBJ_DIR)
+	$(CXX) $(CXXFLAGS) -c $< -o $@
+
+# Special rule for DP-analyzer in subdirectory
+$(OBJ_DIR)/DP-analyzer.o: DP-analyzer/DP-analyzer.cpp | $(OBJ_DIR)
+	$(CXX) $(CXXFLAGS) -c $< -o $@
+
+# Link Mark1
+$(BIN_DIR)/$(TARGET_MARK1): $(OBJS_MARK1) | $(BIN_DIR)
+	$(CXX) $(CXXFLAGS) $^ -o $@ -fopenmp -pthread
+
+# Link DP-analyzer
+$(BIN_DIR)/$(TARGET_DP_ANALYZER): $(OBJS_DP_ANALYZER) | $(BIN_DIR)
+	$(CXX) $(CXXFLAGS) $^ -o $@
+
+# Debug versions
+$(BIN_DIR)/$(TARGET_MARK1)_debug: $(OBJS_MARK1) | $(BIN_DIR)
+	$(CXX) $(CXXFLAGS) $^ -o $@ -fopenmp -pthread
+
+$(BIN_DIR)/$(TARGET_DP_ANALYZER)_debug: $(OBJS_DP_ANALYZER) | $(BIN_DIR)
+	$(CXX) $(CXXFLAGS) $^ -o $@
+
+# Profile version
+$(BIN_DIR)/$(TARGET_MARK1)_profile: $(OBJS_MARK1) | $(BIN_DIR)
+	$(CXX) $(CXXFLAGS) $^ -o $@ -fopenmp -pthread -pg
+
+# Clean
+clean:
+	rm -rf $(BUILD_DIR)
+
+# Deep clean (removes DP table files too)
+distclean: clean
+	rm -f dp_table.bin DP.bin
+
+# Dependencies
+$(OBJ_DIR)/Mark1.o: Mark1.cpp Int.h Point.h SECP256K1.h IntGroup.h Timer.h Random.h hashutil.h simd_block_bloom.h
+$(OBJ_DIR)/Int.o: Int.cpp Int.h
+$(OBJ_DIR)/SECP256K1.o: SECP256K1.cpp SECP256K1.h Int.h Point.h
+$(OBJ_DIR)/Point.o: Point.cpp Point.h Int.h
+$(OBJ_DIR)/Random.o: Random.cpp Random.h
+$(OBJ_DIR)/IntMod.o: IntMod.cpp Int.h
+$(OBJ_DIR)/IntGroup.o: IntGroup.cpp IntGroup.h Int.h
+$(OBJ_DIR)/Timer.o: Timer.cpp Timer.h
+$(OBJ_DIR)/DP-analyzer.o: DP-analyzer/DP-analyzer.cpp Int.h
+
+# Help
+help:
+	@echo "Pollard-Kangaroo Solver (Mark1) - Makefile"
+	@echo ""
+	@echo "Targets:"
+	@echo "  all          - Build optimized release version (default)"
+	@echo "  release      - Build optimized release version"
+	@echo "  debug        - Build debug version with symbols"
+	@echo "  profile      - Build with profiling support"
+	@echo "  clean        - Remove build artifacts"
+	@echo "  distclean    - Remove build artifacts and data files"
+	@echo "  help         - Show this help"
+	@echo ""
+	@echo "Executables will be placed in $(BIN_DIR)/"
+	@echo ""
+	@echo "Usage examples:"
+	@echo "  make release    # Build optimized version"
+	@echo "  make debug      # Build debug version"
+	@echo "  ./build/bin/Mark1 --help"
+
+.PHONY: all release debug profile clean distclean help