chore: Update readme (#85)

Author: ch4r10t33r

README.md

@@ -129,41 +129,6 @@ const keypair = try scheme.keyGen(0, 256);
 
 > **Performance Note**: For lifetimes `2^18` and `2^32`, always compile with `zig build -Doptimize=ReleaseFast` for acceptable performance. Debug builds will be extremely slow for larger lifetimes.
 
-## Cross-Language Compatibility Tests
-
-**Cross-language compatibility status:**
-
-- ✅ **Lifetime `2^8`**: All combinations pass (Rust sign → Rust/Zig verify, Zig sign → Zig/Rust verify).
-- ⚠️ **Lifetime `2^18`**: Rust sign → Rust/Zig verify and Zig sign → Zig verify pass; **Zig sign → Rust verify currently fails** due to a mismatch in the 2^18 instantiation on the Rust side.
-- ✅ **Lifetime `2^32`**: All combinations pass (Rust sign → Rust/Zig verify, Zig sign → Zig/Rust verify).
-
-### Quick Start
-
-```bash
-# Run all supported lifetimes
-python3 benchmark/benchmark.py --lifetime "2^8,2^18,2^32"
-
-# Run specific lifetime
-python3 benchmark/benchmark.py --lifetime 2^32
-
-# Enable verbose debug logs
-BENCHMARK_DEBUG_LOGS=1 python3 benchmark/benchmark.py --lifetime 2^8
-```
-
-The script automatically builds both implementations and runs comprehensive cross-language tests (Rust ↔ Zig signing and verification).
-
-### Test Results
-
-**Lifetime 2^32 (256 active epochs):**
-- ✅ Rust keygen: ~2.0s, Zig keygen: ~1.3s
-- ✅ All verification tests pass
-
-**Lifetime 2^32 (1024 active epochs):**
-- ✅ Zig keygen: ~51.7s (46.5% faster with parallel optimization)
-- ✅ All verification tests pass
-
-See [Performance Benchmarks](#performance-benchmarks) for detailed timing information.
-
 ## Performance Benchmarks
 
 Performance measurements are taken using ReleaseFast builds with debug logging disabled. For lifetime `2^32` with 1024 active epochs, key generation time is measured using the dedicated profiler:

scripts/benchmark_hash_function.zig

@@ -6,9 +6,7 @@ const hash_zig = @import("hash-zig");
 const Allocator = std.mem.Allocator;
 const FieldElement = hash_zig.FieldElement;
 const Poseidon2RustCompat = hash_zig.Poseidon2RustCompat;
-// Import directly to avoid module conflicts (benchmark is separate executable)
 const poseidon2_simd = @import("../src/hash/poseidon2_hash_simd.zig");
-const simd_utils = @import("../src/signature/native/simd_utils.zig");
 
 pub fn main() !void {
     var gpa = std.heap.GeneralPurposeAllocator(.{}){};
@@ -26,27 +24,27 @@ pub fn main() !void {
     // Test parameters
     const num_iterations = 10000;
     const hash_len = 8; // HASH_LEN_FE
-    const SIMD_WIDTH = simd_utils.SIMD_WIDTH;
+    const SIMD_WIDTH = poseidon2_simd.SIMD_WIDTH_CONST;
 
     std.debug.print("Configuration:\n", .{});
     std.debug.print("  SIMD Width: {}\n", .{SIMD_WIDTH});
     std.debug.print("  Hash Length: {}\n", .{hash_len});
     std.debug.print("  Iterations: {}\n\n", .{num_iterations});
 
     // Benchmark: compress16SIMD with SIMD-packed input
-    var packed_input: [16]simd_utils.PackedF = undefined;
+    var packed_input: [16]poseidon2_simd.PackedF = undefined;
 
     // Initialize with test data
     for (0..16) |i| {
         var values: [SIMD_WIDTH]u32 = undefined;
         for (0..SIMD_WIDTH) |lane| {
             values[lane] = @as(u32, @intCast((i * 1000) + lane));
         }
-        packed_input[i] = simd_utils.PackedF{ .values = values };
+        packed_input[i] = poseidon2_simd.PackedF{ .values = values };
     }
 
     // Warmup
-    var output_stack: [8]simd_utils.PackedF = undefined;
+    var output_stack: [8]poseidon2_simd.PackedF = undefined;
     for (0..100) |_| {
         _ = try simd_poseidon2.compress16SIMD(&packed_input, hash_len, output_stack[0..hash_len]);
     }

src/hash/poseidon2_hash_simd.zig

@@ -16,6 +16,10 @@ const Poseidon2KoalaBear24 = @import("../poseidon2/root.zig").Poseidon2KoalaBear
 const WIDTH_16 = 16;
 const SIMD_WIDTH = simd_utils.SIMD_WIDTH;
 
+// Re-export SIMD utilities for external benchmarks without importing simd_utils directly
+pub const PackedF = simd_utils.PackedF;
+pub const SIMD_WIDTH_CONST: u32 = SIMD_WIDTH;
+
 /// SIMD-aware Poseidon2 compression function
 /// Processes multiple states simultaneously using SIMD operations
 pub const Poseidon2SIMD = struct {