UltrafastSecp256k1 Documentation

Version 3.15.0 -- Cross-platform secp256k1 ECC library

Quick Links

Document	Description
API Reference	Complete CPU + CUDA + WASM function reference
Building	Build instructions for all 10+ platforms
Benchmarks	Performance data: CPU, GPU, embedded, mobile, including the 2026-03 x86, Android, CUDA, and OpenCL refresh
Bindings Validation Closure	Final summary of the multi-language binding validation and docs sync pass
Bindings Release Notes Source	Release-time source note for later changelog, release body, and announcement assembly
Bindings Announcement Draft	Ready-to-post announcement draft for the completed binding validation matrix
ESP32 Setup	ESP32-S3/PICO-D4 flashing & testing guide
RISC-V Optimizations	RISC-V assembly & RVV details
Porting Guide	Add new platforms, architectures, GPU backends
Project Graph Reasoning	SQLite code graph, semantic tags, symbol reasoning, and optimization/audit workflows
`scripts/query_graph.py tags`	Semantic tag inventory from the repository knowledge graph
`scripts/query_graph.py symbol/optimize/risk`	Symbol-level crypto reasoning, gain/risk scoring, and optimization candidates

Security & Audit

Document	Description
Audit Guide	Start here -- Auditor navigation, checklist, reproduction commands
Architecture	Technical architecture deep-dive for auditors
CT Verification	Constant-time methodology, dudect, known limitations
Test Matrix	Function -> test coverage map with gap analysis
Security Policy	Vulnerability reporting, audit status, production readiness
Threat Model	Layer-by-layer risk + attack surface analysis
Audit Report	Internal audit: 641,194 checks, 8 suites, 0 failures
GPU Validation Matrix	Backend parity, GPU ABI tests, and the canonical local GPU proof path

Adoption & Integration

Document	Description
Integration Guide	CMake FetchContent, add_subdirectory, vcpkg, migration from libsecp256k1
API Stability	Header stability tiers: Stable / Provisional / Experimental
Backend Guide	CPU, CUDA, ROCm, OpenCL, Metal, WASM, Android matrix
libsecp256k1 Shim	Drop-in C API compatibility layer

External Docs

Document	Description
CUDA / ROCm GPU	CUDA + HIP/ROCm architecture, kernels, benchmarks
WebAssembly	WASM build, JS/TS API, npm package
Contributing	Development workflow, coding standards, PR process
Security	Vulnerability reporting, security model
Changelog	Version history

Getting Started

1. Build

cmake -S . -B build -G Ninja -DCMAKE_BUILD_TYPE=Release
cmake --build build -j

2. Self-Test

ctest --test-dir build --output-on-failure

3. Use in Your Code

#include <secp256k1/field.hpp>
#include <secp256k1/point.hpp>
#include <secp256k1/scalar.hpp>
#include <secp256k1/ecdsa.hpp>
#include <secp256k1/schnorr.hpp>
#include <secp256k1/sha256.hpp>

using namespace secp256k1::fast;

int main() {
    // Key generation
    Scalar private_key = Scalar::from_hex(
        "E9873D79C6D87DC0FB6A5778633389F4453213303DA61F20BD67FC233AA33262"
    );
    Point public_key = Point::generator().scalar_mul(private_key);

    // ECDSA sign / verify
    std::array<uint8_t, 32> msg_hash = sha256::hash("hello", 5);
    auto [r, s] = ecdsa::sign(msg_hash, private_key);
    bool ok = ecdsa::verify(msg_hash, public_key, r, s);

    return ok ? 0 : 1;
}

Architecture Overview

UltrafastSecp256k1/
+-- cpu/                         # CPU library (C++20, header-only + compiled)
|   +-- include/secp256k1/       # Public headers
|   |   +-- field.hpp            #   Field element (mod p)
|   |   +-- scalar.hpp           #   Scalar (mod n)
|   |   +-- point.hpp            #   EC point operations
|   |   +-- ecdsa.hpp            #   ECDSA sign/verify (RFC 6979)
|   |   +-- schnorr.hpp          #   Schnorr BIP-340 sign/verify
|   |   +-- sha256.hpp           #   SHA-256 hash
|   |   +-- glv.hpp              #   GLV endomorphism
|   |   +-- precompute.hpp       #   Generator table
|   |   +-- ct/                  #   Constant-time variants
|   |   +-- types.hpp            #   Cross-backend POD types
|   +-- src/                     # Implementation + platform ASM
|   |   +-- field.cpp
|   |   +-- field_asm_x64.asm    #   x86-64 BMI2/ADX
|   |   +-- field_asm_riscv64.S  #   RISC-V RV64GC + RVV
|   |   +-- field_asm_arm64.cpp  #   ARM64 MUL/UMULH
|   |   +-- ecdsa.cpp
|   |   +-- schnorr.cpp
|   |   +-- ...
|   +-- tests/                   # CTest unit tests
|   +-- bench/                   # Benchmarks
|   +-- fuzz/                    # libFuzzer harnesses
|
+-- cuda/                        # CUDA + ROCm/HIP GPU library
|   +-- include/
|   |   +-- secp256k1.cuh        #   All device functions (field/point/scalar)
|   |   +-- ptx_math.cuh         #   PTX inline asm (with __int128 fallback)
|   |   +-- gpu_compat.h         #   CUDA <-> HIP API mapping
|   |   +-- batch_inversion.cuh  #   Montgomery trick batch inverse
|   |   +-- bloom.cuh            #   Device-side Bloom filter
|   |   +-- hash160.cuh          #   SHA-256 + RIPEMD-160
|   +-- app/                     #   Experimental search kernels
|   +-- src/                     #   Kernel wrappers, tests, benchmarks
|
+-- opencl/                      # OpenCL GPU library
|   +-- kernels/                 #   .cl kernel sources
|   +-- ...
|
+-- wasm/                        # WebAssembly (Emscripten)
|   +-- secp256k1_wasm.h         #   C API (11 functions)
|   +-- secp256k1_wasm.cpp       #   Implementation
|   +-- secp256k1.mjs            #   JS wrapper
|   +-- secp256k1.d.ts           #   TypeScript declarations
|   +-- package.json             #   npm: @ultrafastsecp256k1/wasm
|
+-- examples/
|   +-- basic_usage/             #   Desktop C++ example
|   +-- esp32_test/              #   ESP32-S3 / ESP32-PICO-D4
|   +-- stm32_test/              #   STM32F103ZET6 ARM Cortex-M3
|
+-- cmake/
|   +-- version.hpp.in           #   Auto-generated version header
|   +-- ios.toolchain.cmake      #   iOS cross-compilation toolchain
|
+-- scripts/
|   +-- build_wasm.sh            #   Emscripten WASM build
|   +-- build_xcframework.sh     #   iOS XCFramework build
|
+-- .github/workflows/
|   +-- ci.yml                   #   CI: Linux/Win/macOS/iOS/WASM/Android/ROCm
|   +-- docs.yml                 #   Doxygen -> GitHub Pages
|
+-- Package.swift                # Swift Package Manager
+-- UltrafastSecp256k1.podspec   # CocoaPods
+-- Doxyfile                     # Doxygen config
+-- CMakeLists.txt               # Top-level CMake (v3.0.0)

Supported Platforms

Platform	Architecture	Assembly	Status
Linux x86-64	BMI2/ADX	x86-64 ASM	[OK] Production
Windows x86-64	BMI2/ADX	x86-64 ASM	[OK] Production
macOS x86-64 / ARM64	Native	ARM64 ASM	[OK] Production
RISC-V 64	RV64GC + RVV	RISC-V ASM	[OK] Production
Android ARM64	Cortex-A55/A76	ARM64 ASM	[OK] Production
iOS 17+	Apple Silicon	ARM64 ASM	[OK] CI (testers wanted)
CUDA (sm_75+)	PTX	PTX inline	[OK] Production path
ROCm / HIP	GCN / RDNA	Portable	[..] Planned hardware validation
OpenCL 3.0	PTX	PTX inline	[OK] Broader backend, partial unified GPU C ABI (4/6)
WebAssembly	Emscripten	Portable C++	[OK] Production
ESP32-S3	Xtensa LX7	Portable C++	[OK] Tested
ESP32-PICO-D4	Xtensa LX6	Portable C++	[OK] Tested
STM32F103	Cortex-M3	ARM Thumb ASM	[OK] Tested

License

MIT -- See LICENSE

Integration consulting available -- contact payysoon@gmail.com

8.4 KiB Raw Permalink Blame History

UltrafastSecp256k1 Documentation

Quick Links

Security & Audit

Adoption & Integration

External Docs

Getting Started

1. Build

2. Self-Test

3. Use in Your Code

Architecture Overview

Supported Platforms

License

8.4 KiB

Raw Permalink Blame History