CORDIC_Abs_APFX/README.md
Camille Monière eaa9177962
Modify code to comply to buggy Vivado_HLS 2019.1
- For whatever reasons, v2019.1 can C-Simulate CCordicAbs, but cannot
  C-Synthetize it. So the process member function have been mirrored to
  a static process function in the top level, called directly. It just
  works.
- The TCL script have been amended to support both Vivado_HLS 2019.1 and
  Vitis_HLS 2020.2 (maybe .1 also, but hasn't been tested).
- The CMake project works and is validated for G++ > 6.2, so can't be
  used with vivado less than 2020.1 (which use 4.6.3). This vivado
  version must be used with the TCL script directly.
- A custom target `run_hls` has been created to call vitis_hls 2020.2 if
  ENABLE_XILINX and ENABLE_TESTING have been specified. It is called by
  ctest by default. It can produce an IP (option IP_XILINX) and run Vivado implementation
  design flow (option IMPL_XILINX).
2022-04-19 19:07:06 +02:00

2.7 KiB

CORDIC Abs APFX

A free way to implement a CORDIC-based Complex Absolute Value using HLS, with bit-accurate precision.

Goal

CORDIC (COordinate Rotation DIgital Computer) is an efficient way to implement hardware complex rotations (e.g. z * exp(jw), with z = x + jy a complex and w a real angle). It also can be used to approximate the module of a complex. It is also useful for microcontrollers or microprocessors lacking floating-point units, as such multiplications can consume a noticeable amount of CPU cycles.

Indeed, |z| = sqrt(x² + y²) = |exp(-j*angle(z)) * z|, thus the module of z is the absolute value of the real part of its rotated version. So, ... CORDIC!

This implementation in C++14 (-std=c++14) is suitable for hardware simulation and (with the right headers and maybe a few tweaks) for synthesis.

This repository defines one working CORDIC-based Absolute Value units class, CCordicAbs.

Test suite and dependencies

The Catch2 test framework has been used in conjunction with CTest to provides unit tests.

  • Has been tested successfully compiled with:
    • GNU GCC 6.2 (Xilinx bundled version), 6.5, 9.4, 10.1, 10.2 and 11.2,
    • LLVM Clang 12.0 and 13.0,
  • Uses Catch v2.13.7,
  • Depends on Xilinx HLS arbitrary precision types, available as FOSS here provided by Xilinx or here patched by myself. Note: Xilinx also provides proprietary versions of those headers, suitable for synthesis and implementation, bundled with their products.

A Xilinx C++ HLS testbench is also available, as well as a TCL script to run simulation, synthesis, co-simulation and IP export and implementation if wanted. Xilinx HLS v2020.2 can be directly called from CMake by the target run_hls if the options ENABLE_TESTING, ENABLE_XILINX and COSIM_XILINX are enabled. If CMake feels like black magic, It is advice to use tools like ccmake (NCurses terminal interface to cmake) or cmake-gui.

Copyright 2022 Camille "DrasLorus" Monière.

This program is free software: you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details.

A copy of the license is available here in Markdown or here in plain text.