DRsolve: Dixon Resultant & Polynomial System Solver

A cross-platform C library and command-line tool for computing Dixon resultants and solving polynomial systems over finite fields and the rationals ℚ, based on the FLINT and PML libraries.

Library & Platform Support

DRSolve is distributed as both a shared library (libdrsolve.so / libdrsolve.dylib / libdrsolve-1.dll) and a static library (libdrsolve.a / libdrsolve-1.a), alongside the drsolve command-line executable.

Platform	Shared lib	Static lib	CLI
Linux (x86-64, ARM64)	libdrsolve.so	libdrsolve.a	drsolve
macOS (x86-64, Apple Silicon)	libdrsolve.dylib	libdrsolve.a	drsolve
Windows (x86-64)	libdrsolve-1.dll	libdrsolve-1.a	drsolve.exe + drsolve_win_gui.exe

---

Features

• Dixon resultant computation for variable elimination
• Polynomial system solver for n×n systems
• Dixon with triangular ideal reduction
• Finite fields:
  • Prime fields F_p (any size): Implemented with FLINT modular arithmetic, optionally accelerated by PML.
  • Extension fields F_{p^k}: Further optimized for binary fields F_{2^n} with n in {8, 16, 32, 64, 128}.
• Rational field ℚ: Rational reconstruction via multi-prime CRT. Set field_size = 0 to enable.
• Complexity analysis — estimates Dixon matrix size, Bezout degree bound, and operation count before computing.
• Command-line input or file input. Automatic output to solution files.

---

Dependencies

• FLINT (recommended version: 3.5.0)
  https://github.com/flintlib/flint

Optional:

• PML (used automatically if available)
  https://github.com/vneiger/pml

---

Build

DRSolve uses CMake (≥ 3.16) as its primary build system.

Linux / macOS

git clone https://github.com/drsolve/drsolve-cross.git && cd drsolve-cross
cmake -B build
cmake --build build -j$(nproc)
ctest --test-dir build          # optional: run tests
sudo cmake --install build      # optional: install to /usr/local

FLINT must be installed (e.g. via your package manager) before configuring.
If FLINT is in a non-standard location, pass -DFLINT_ROOT=/path/to/flint.

# Example: FLINT installed under $HOME/.local
cmake -B build -DFLINT_ROOT=$HOME/.local
cmake --build build -j$(nproc)

Windows

Option A — GUI installer (recommended)
Download from drsolve/drsolve-win.

Option B — MSYS2/UCRT64

pacman -S mingw-w64-ucrt-x86_64-cmake \
          mingw-w64-ucrt-x86_64-gcc \
          mingw-w64-ucrt-x86_64-flint
cmake -B build -G "MinGW Makefiles"
cmake --build build -j$(nproc)

Option C — cross-compile from Linux/macOS (MinGW-w64)
Bundled third_party/ libraries are used automatically:

cmake -B build-win \
      -DCMAKE_TOOLCHAIN_FILE="$(pwd)/cmake/toolchain-mingw64.cmake"
cmake --build build-win -j$(nproc)

For full build options and advanced configurations, see BUILDING.md.

---

Usage

Dixon Resultant (Basic)

./drsolve "polynomials" "eliminate_vars" field_size

Examples:

./drsolve "x+y+z, x*y+y*z+z*x, x*y*z+1" "x,y" 257
./drsolve "x^2+y^2+z^2-1, x^2+y^2-2*z^2, x+y+z" "x,y" 0

---

Polynomial System Solver (n equations in n variables)

./drsolve --solve "polynomials" field_size

Example:

./drsolve --solve "x^2 + y^2 + z^2 - 6, x + y + z - 4, x*y*z - x - 1" 257

---

Complexity Analysis

Estimates the difficulty of a Dixon resultant computation without performing it. Reports equation count, variable count, degree sequence, Dixon matrix size (via Hessenberg recurrence), Bezout degree bound, and complexity in bits.

./drsolve --comp "polynomials" "eliminate_vars" field_size
./drsolve -c     "polynomials" "eliminate_vars" field_size

Examples:

./drsolve --comp "x^3+y^3+z^3, x^2*y+y^2*z+z^2*x, x+y+z-1" "x,y" 257

Custom omega — set the matrix-multiplication exponent used in the complexity formula (default: 2.81):

./drsolve --comp --omega 2.81 "polynomials" "eliminate_vars" field_size
./drsolve -c -w 2.0            "polynomials" "eliminate_vars" field_size

File input uses the same format as Dixon mode:

./drsolve --comp example.dr          # output: example_comp.dr

---

Extension Fields

./drsolve "x + y^2 + t, x*y + t*y + 1" "x" 2^8

The default settings use t as the extension field generator and FLINT's built-in field polynomial.

./drsolve --solve "x^2 + t*y, x*y + t^2" "2^8: t^8 + t^4 + t^3 + t + 1"

(with AES custom polynomial for F_256)

---

Dixon with Ideal Reduction

./drsolve --ideal "ideal_generators" "polynomials" "eliminate_vars" field_size

Example:

./drsolve --ideal "a2^3=2*a1+1, a3^3=a1*a2+3" "a1^2+a2^2+a3^2-10, a3^3-a1*a2-3" "a3" 257

Field-equation reduction mode

After each multiplication, reduces x^q → x for every variable.

./drsolve --field-equation "polynomials" "eliminate_vars" field_size
./drsolve --field-equation -r "[d1,d2,...,dn]" field_size

Example:

./drsolve --field-equation "x0*x2+x1, x0*x1*x2+x2+1, x1*x2+x0+1" "x0,x1" 2
./drsolve --field-equation -r [3]*5 2

---

Silent Mode

./drsolve --silent [--solve|--comp|-c] <arguments>

No console output is produced; the solution/report file is still generated.

---

Method Selection

./drsolve --method <num> <args>
./drsolve --step1 <num> --step4 <num> <args>

Available determinant methods:

• 0 = Recursive
• 1 = Kronecker+HNF
• 2 = Interpolation
• 3 = Sparse interpolation

--method sets both step 1 and step 4 for backward compatibility.

---

Random Mode

Generate random polynomial systems with specified degrees for testing and benchmarking.

Basic Usage

./drsolve --random "[d1,d2,...,dn]" field_size
./drsolve -r       "[d]*n"          field_size

• [d1,d2,...,dn]: degree list (comma-separated) for n polynomials
• [d]*n: all n polynomials have same degree d
• field_size: field size (prime or extension); use 0 for Q

Combine with Compute Flags

# Random + Dixon elimination
./drsolve -r --solve "[d1,...,dn]" field_size

# Random + complexity analysis
./drsolve -r --comp  "[d]*n" field_size
./drsolve -r -c --omega 2.81 "[4]*5" 257   # custom omega

# Random + Dixon with ideal reduction
./drsolve -r "[d1,d2,d3]" "ideal_generators" field_size

Examples

# 3 polynomials (deg 3,3,2) in GF(257)
./drsolve --random "[3,3,2]" 257

# 3 polynomials (deg 3,3,2) over Q
./drsolve --random "[3,3,2]" 0

# Solve 3 quadratic system in GF(257)
./drsolve -r --solve "[2]*3" 257

# Complexity analysis of 4 quartic polynomials
./drsolve -r --comp --omega 2.81 "[4]*4" 257

# GF(2^8) with degrees 3 and 2
./drsolve -r "[3,2]" 2^8

File Input Format

Dixon Mode / Complexity Mode (multiline)

Line 1 : field size (prime or p^k; use 0 for Q; generator defaults to 't')
Line 2+: polynomials (comma-separated, may span multiple lines)
Last   : variables to ELIMINATE (comma-separated)
         (#eliminate = #equations - 1)

Example:

./drsolve       example.dr
./drsolve --comp example.dr

Polynomial Solver Mode (multiline)

Line 1 : field size
Line 2+: polynomials
         (n equations in n variables)

---

Output

Mode	Command-line input	File input example.dr
Dixon / Solver	solution_YYYYMMDD_HHMMSS.dr	example_solution.dr
Complexity	comp_YYYYMMDD_HHMMSS.dr	example_comp.dr

Each output file contains field information, input polynomials, computation time, and the resultant, solutions, or complexity report.

Complexity report contents

• Equation count, variable list, elimination variable list, remaining variables
• Degree sequence of input polynomials
• Bezout bound (product of degrees)
• Dixon matrix size (Hessenberg recurrence)
• Resultant degree estimate
• Complexity in log₂ bits (with the omega value used)

---

Notes

• All computation modes generate a solution/report file by default
• Extension fields are slower than prime fields due to polynomial arithmetic
• The optional PML library only accelerates well-determined systems over prime fields
• Complexity analysis does not run any polynomial arithmetic; it parses only
• Over Q (field_size=0), --solve, --ideal, and --field-equation are not yet supported

---

Feature Support by Field

Feature	F_p (p<2^63)	F_p (p>2^63)	F_{p^k} (p<2^63)	Q
Dixon resultant	✅	✅	✅	✅
Complexity analysis (--comp)	✅	✅	✅	✅
Random mode (-r)	✅	✅	✅	✅
Polynomial solver (--solve)	✅	✅	✅	✅
Ideal reduction (--ideal)	✅	❌	✅	❌
Field-equation reduction	✅	❌	✅	❌
PML acceleration	✅	✅	❌	✅

---

License

DRSolve is distributed under the GNU General Public License version 2.0 (GPL-2.0-or-later). See the file COPYING.