finn-plus 1.0.0__tar.gz

finn_plus-1.0.0/AUTHORS.rst

@@ -0,0 +1,40 @@
+============
+Contributors
+============
+
+* Felix Jentzsch (@fpjentzsch) (FINN+ maintainer)
+* Christoph Berganski (@iksnagreb) (FINN+ maintainer)
+* Linus Jungemann (@LinusJungemann) (FINN+ maintainer)
+* Bjarne Wintermann (@bwintermann) (FINN+ maintainer)
+* Jakoba Petri-Koenig (@auphelia) (FINN maintainer)
+* Thomas Preusser (@preusser)
+* Yaman Umuroglu (@maltanar)
+* Andrea Rigoni (@AndreaRigoni)
+* Hendrik Borras (@HenniOVP)
+* Lucian Petrica (@quetric)
+* Tobias Alonso (@Tobi-Alonso)
+* Mirza Mrahorovic (@mmrahorovic)
+* Suranga Mahesh (@surangamh)
+* Peter Lehnhardt (@pete-lennart)
+* Neil Kim Nielsen (@neilkimn)
+* Jon Ander Lezeta (@jalezeta)
+* John Terry (@jterry-x)
+* Alina Vasilciuc (@alinavalinav)
+* Alessandro Pappalardo (@volcacius)
+* Giuseppe Franco (@Giuseppe5)
+* Syed Asad Alam (@asadalam)
+* Javier Duarte (@jmduarte)
+* Uma Maheshwari (@umav1511)
+* José Rosa (@pinxau1000)
+* Aziz Bahri (@azizb-xlnx)
+* Fionn O'Donohoe (@fionnodonohoe-xlnx)
+* Matthias Gehre (@mgehre-amd)
+* Hugo Le Blevec (@hleblevec)
+* Patrick Geel (@patrickgeel)
+* John Monks (@jmonks-amd)
+* Tim Paine (@timkpaine)
+* Shashwat Khandelwal (@shashwat1198)
+* Ian Colbert (@i-colbert)
+* Rachit Garg (@rstar900)
+* Jonas Kuehle (@vopade)
+* Aditya S (@Adityasrinivas24)

finn_plus-1.0.0/LICENSE.txt

@@ -0,0 +1,33 @@
+Copyright for portions of project FINN+ is held by AMD as part of project FINN
+and is provided under the BSD license. All other copyright for project FINN+
+is held by Paderborn University and is provided under the same license.
+
+Copyright (C) 2020-2022, Xilinx, Inc.
+Copyright (C) 2022-2025, Advanced Micro Devices, Inc.
+Copyright (C) 2025, Paderborn University
+All rights reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are met:
+
+* Redistributions of source code must retain the above copyright notice, this
+  list of conditions and the following disclaimer.
+
+* Redistributions in binary form must reproduce the above copyright notice,
+  this list of conditions and the following disclaimer in the documentation
+  and/or other materials provided with the distribution.
+
+* Neither the name of FINN/FINN+ nor the names of its
+  contributors may be used to endorse or promote products derived from
+  this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
+FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
+CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
+OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

finn_plus-1.0.0/PKG-INFO

@@ -0,0 +1,144 @@
+Metadata-Version: 2.3
+Name: finn-plus
+Version: 1.0.0
+Summary: Fork of FINN, a QNN dataflow compiler for AMD/Xilinx FPGAs
+Maintainer: Felix Jentzsch
+Maintainer-email: felix.jentzsch@upb.de
+Requires-Python: >=3.10,<3.12
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Requires-Dist: bitstring (==3.1.7)
+Requires-Dist: click (>=8.1.8,<9.0.0)
+Requires-Dist: clize (==5.0.1)
+Requires-Dist: deap (==1.3.1)
+Requires-Dist: dependencies (==2.0.1)
+Requires-Dist: dvc[webdav] (==3.59.1)
+Requires-Dist: dvclive[image] (==3.48.2)
+Requires-Dist: future-annotations (==1.0.0)
+Requires-Dist: gspread (==3.6.0)
+Requires-Dist: importlib-resources (==6.1.0)
+Requires-Dist: ipykernel (==6.21.2)
+Requires-Dist: ipython (==8.12.2)
+Requires-Dist: jupyter (==1.0.0)
+Requires-Dist: markupsafe (==2.0.1)
+Requires-Dist: mashumaro[yaml] (>=3.15,<4.0)
+Requires-Dist: matplotlib (==3.7.0)
+Requires-Dist: mip (==1.13.0)
+Requires-Dist: netron (>5.0.0)
+Requires-Dist: networkx (==2.8)
+Requires-Dist: numpy (==1.24.1)
+Requires-Dist: onnx (==1.17.0)
+Requires-Dist: onnxoptimizer (>=0.3.13,<0.4.0)
+Requires-Dist: onnxruntime (==1.18.1)
+Requires-Dist: pandas (==1.5.3)
+Requires-Dist: pre-commit (==3.3.2)
+Requires-Dist: protobuf (==3.20.3)
+Requires-Dist: psutil (==5.9.4)
+Requires-Dist: pygments (==2.14.0)
+Requires-Dist: pyscaffold (==4.6)
+Requires-Dist: pytest (==8.3.4)
+Requires-Dist: pytest-cov (==6.0.0)
+Requires-Dist: pytest-dependency (==0.6.0)
+Requires-Dist: pytest-forked (==1.6.0)
+Requires-Dist: pytest-html (==4.1.1)
+Requires-Dist: pytest-html-merger (==0.1.0)
+Requires-Dist: pytest-metadata (==3.1.1)
+Requires-Dist: pytest-parallel (==0.1.1)
+Requires-Dist: pytest-rerunfailures (==15.0)
+Requires-Dist: pytest-xdist[setproctitle] (==3.6.1)
+Requires-Dist: pyyaml (==6.0.1)
+Requires-Dist: rich (>=13.9.4,<14.0.0)
+Requires-Dist: scikit-learn (==1.2.1)
+Requires-Dist: scipy (==1.10.1)
+Requires-Dist: setupext-janitor (>=1.1.2,<2.0.0)
+Requires-Dist: setuptools (>40.8.0)
+Requires-Dist: sigtools (==4.0.1)
+Requires-Dist: tclwrapper (==0.0.1)
+Requires-Dist: termcolor (>=2.5.0,<3.0.0)
+Requires-Dist: tokenize-rt (==4.2.1)
+Requires-Dist: toposort (==1.7.0)
+Requires-Dist: torch (==2.6.0)
+Requires-Dist: torchvision (==0.21.0)
+Requires-Dist: tqdm (==4.64.1)
+Requires-Dist: vcdvcd (==1.0.5)
+Requires-Dist: wget (==3.2)
+Project-URL: EKI Project, https://www.eki-project.tech/
+Project-URL: GitHub, https://github.com/eki-project/finn-plus
+Description-Content-Type: text/markdown
+
+<img src=https://cs.uni-paderborn.de/fileadmin-eim/informatik/fg/ce/MiscImages/finn-plus_logo.png width=196/>
+
+# Dataflow Compiler for Fast, Scalable Quantized Neural Network Inference on FPGAs
+
+**FINN+** is a fork of **FINN**, an experimental framework from the Integrated Communications and AI Lab of AMD Research & Advanced Development to explore deep neural network inference on FPGAs.
+It specifically targets quantized neural networks, with emphasis on generating dataflow-style architectures customized for each network.
+The resulting FPGA accelerators are highly efficient and can yield high throughput and low latency.
+The framework is fully open-source in order to give a higher degree of flexibility, and is intended to enable neural network research spanning several layers of the software/hardware abstraction stack.
+
+**To get an overview of how FINN+ is used, take a look at the Getting Started section below!**
+
+**While our [Wiki](https://github.com/eki-project/finn-plus/wiki) is under construction, we refer to the original [FINN homepage](https://xilinx.github.io/finn/) for further information.**
+
+## FINN+ Extensions
+**FINN+** aims to incorporate all development from the upstream repository (dev branch) while extending **FINN** in all directions, including the following list of features that are either in progress or already completed:
+- Transformer/Attention support
+- Improved streamlining
+- Improved automatic folding and FIFO-sizing
+- Empirical quality-of-result (QoR) estimation
+- Back-end extensions
+    - Instrumentation for accurate performance profiling in simulation and on hardware
+    - Improved Alveo build flow
+    - Multi-FPGA support
+    - Optimized C++ driver
+- Quality-of-live improvements
+    - Better logging and error handling
+    - Type hinting/checking
+    - Alternative YAML-based build configuration
+    - Containerless setup
+
+Please refer to our [**Feature Tracker**](https://github.com/orgs/eki-project/projects/1) for the current status of individual features.
+While some items are already on-track to be merged into the upstream repository, we try to merge them into the **FINN+** dev branch as early as possible to increase development pace and drive our research forward.
+
+## Getting Started
+This is a quick overview of how to get started, for additional information please refer to our [**Wiki**](https://github.com/eki-project/finn-plus/wiki)!
+
+### Requirements
+The primary dependencies currently are:
+- Python >= 3.10 (< 3.12)
+- Vivado, Vitis, Vitis HLS (2022.2 or 2024.2)
+- Some basic system-level packages, refer to the [**dependency installation script**](https://github.com/eki-project/finn-plus/blob/main/installDependencies.sh)
+
+### Installing via pip
+After preparing the dependencies mentioned above, simply run the following to start a build flow:
+```
+# Make sure to create a fresh virtual environment for FINN+
+pip install finn-plus          # Install FINN+ and its Python dependencies via pip
+finn deps update               # Ensure FINN+ pulled all further dependencies (this might update packages in your venv!)
+finn build build_config.yaml   # Run a FINN+ build defined in a YAML file
+```
+
+### Installing from the repository
+To install directly from the repository, you'll need Poetry (>= 2.0) for dependency management. After cloning the repo and setting up the system-level dependencies, run the following to start a build flow:
+```
+cd finn-plus
+poetry install                 # Install Python packages into a Poetry-managed virtual environment
+source <your-poetry-venv>      # Use "poetry env info" to find the path to your Poetry venv. For further information visit the Poetry documentation
+finn config create             # Create a default configuration in ~/.finn/settings.yaml. Optional but recommended
+finn deps update               # Ensure FINN+ pulled all further dependencies (this might update packages in your venv!)
+finn build build_config.yaml   # Run a FINN+ build defined in a YAML file
+```
+
+## About Us
+FINN+ is maintained by researchers from the [Computer Engineering Group](https://en.cs.uni-paderborn.de/ceg) (CEG) and [Paderborn Center for Parallel Computing](https://pc2.uni-paderborn.de/) (PC²) at Paderborn University, Germany as part of the [eki research project](https://www.eki-project.tech/).
+
+<p align="left">
+<a href="https://en.cs.uni-paderborn.de/ceg"><img align="top" src="https://cs.uni-paderborn.de/fileadmin-eim/informatik/fg/ce/MiscImages/UPB_Logo_ENG_coloured_RGB.jpg" alt="logo" style="margin-right: 20px" width="250"/></a>
+<a href="https://pc2.uni-paderborn.de/"><img align="top" src="https://cs.uni-paderborn.de/fileadmin-eim/informatik/fg/ce/MiscImages/PC2_logo.png" alt="logo" style="margin-right: 20px" width="250"/></a>
+</p>
+
+<p align="left">
+<a href="https://www.eki-project.tech/"><img align="top" src="https://cs.uni-paderborn.de/fileadmin-eim/informatik/fg/ce/MiscImages/eki-RGB-EN-s.png" alt="logo" style="margin-right: 20px" width="250"/></a>
+<a href="https://www.bmuv.de/"><img align="top" src="https://cs.uni-paderborn.de/fileadmin-eim/informatik/fg/ce/MiscImages/BMUV_Fz_2021_Office_Farbe_en.png" alt="logo" style="margin-right: 20px" width="250"/></a>
+</p>
+

finn_plus-1.0.0/README.md

@@ -0,0 +1,74 @@
+<img src=https://cs.uni-paderborn.de/fileadmin-eim/informatik/fg/ce/MiscImages/finn-plus_logo.png width=196/>
+
+# Dataflow Compiler for Fast, Scalable Quantized Neural Network Inference on FPGAs
+
+**FINN+** is a fork of **FINN**, an experimental framework from the Integrated Communications and AI Lab of AMD Research & Advanced Development to explore deep neural network inference on FPGAs.
+It specifically targets quantized neural networks, with emphasis on generating dataflow-style architectures customized for each network.
+The resulting FPGA accelerators are highly efficient and can yield high throughput and low latency.
+The framework is fully open-source in order to give a higher degree of flexibility, and is intended to enable neural network research spanning several layers of the software/hardware abstraction stack.
+
+**To get an overview of how FINN+ is used, take a look at the Getting Started section below!**
+
+**While our [Wiki](https://github.com/eki-project/finn-plus/wiki) is under construction, we refer to the original [FINN homepage](https://xilinx.github.io/finn/) for further information.**
+
+## FINN+ Extensions
+**FINN+** aims to incorporate all development from the upstream repository (dev branch) while extending **FINN** in all directions, including the following list of features that are either in progress or already completed:
+- Transformer/Attention support
+- Improved streamlining
+- Improved automatic folding and FIFO-sizing
+- Empirical quality-of-result (QoR) estimation
+- Back-end extensions
+    - Instrumentation for accurate performance profiling in simulation and on hardware
+    - Improved Alveo build flow
+    - Multi-FPGA support
+    - Optimized C++ driver
+- Quality-of-live improvements
+    - Better logging and error handling
+    - Type hinting/checking
+    - Alternative YAML-based build configuration
+    - Containerless setup
+
+Please refer to our [**Feature Tracker**](https://github.com/orgs/eki-project/projects/1) for the current status of individual features.
+While some items are already on-track to be merged into the upstream repository, we try to merge them into the **FINN+** dev branch as early as possible to increase development pace and drive our research forward.
+
+## Getting Started
+This is a quick overview of how to get started, for additional information please refer to our [**Wiki**](https://github.com/eki-project/finn-plus/wiki)!
+
+### Requirements
+The primary dependencies currently are:
+- Python >= 3.10 (< 3.12)
+- Vivado, Vitis, Vitis HLS (2022.2 or 2024.2)
+- Some basic system-level packages, refer to the [**dependency installation script**](https://github.com/eki-project/finn-plus/blob/main/installDependencies.sh)
+
+### Installing via pip
+After preparing the dependencies mentioned above, simply run the following to start a build flow:
+```
+# Make sure to create a fresh virtual environment for FINN+
+pip install finn-plus          # Install FINN+ and its Python dependencies via pip
+finn deps update               # Ensure FINN+ pulled all further dependencies (this might update packages in your venv!)
+finn build build_config.yaml   # Run a FINN+ build defined in a YAML file
+```
+
+### Installing from the repository
+To install directly from the repository, you'll need Poetry (>= 2.0) for dependency management. After cloning the repo and setting up the system-level dependencies, run the following to start a build flow:
+```
+cd finn-plus
+poetry install                 # Install Python packages into a Poetry-managed virtual environment
+source <your-poetry-venv>      # Use "poetry env info" to find the path to your Poetry venv. For further information visit the Poetry documentation
+finn config create             # Create a default configuration in ~/.finn/settings.yaml. Optional but recommended
+finn deps update               # Ensure FINN+ pulled all further dependencies (this might update packages in your venv!)
+finn build build_config.yaml   # Run a FINN+ build defined in a YAML file
+```
+
+## About Us
+FINN+ is maintained by researchers from the [Computer Engineering Group](https://en.cs.uni-paderborn.de/ceg) (CEG) and [Paderborn Center for Parallel Computing](https://pc2.uni-paderborn.de/) (PC²) at Paderborn University, Germany as part of the [eki research project](https://www.eki-project.tech/).
+
+<p align="left">
+<a href="https://en.cs.uni-paderborn.de/ceg"><img align="top" src="https://cs.uni-paderborn.de/fileadmin-eim/informatik/fg/ce/MiscImages/UPB_Logo_ENG_coloured_RGB.jpg" alt="logo" style="margin-right: 20px" width="250"/></a>
+<a href="https://pc2.uni-paderborn.de/"><img align="top" src="https://cs.uni-paderborn.de/fileadmin-eim/informatik/fg/ce/MiscImages/PC2_logo.png" alt="logo" style="margin-right: 20px" width="250"/></a>
+</p>
+
+<p align="left">
+<a href="https://www.eki-project.tech/"><img align="top" src="https://cs.uni-paderborn.de/fileadmin-eim/informatik/fg/ce/MiscImages/eki-RGB-EN-s.png" alt="logo" style="margin-right: 20px" width="250"/></a>
+<a href="https://www.bmuv.de/"><img align="top" src="https://cs.uni-paderborn.de/fileadmin-eim/informatik/fg/ce/MiscImages/BMUV_Fz_2021_Office_Farbe_en.png" alt="logo" style="margin-right: 20px" width="250"/></a>
+</p>

finn_plus-1.0.0/custom_hls/checksum.cpp

@@ -0,0 +1,36 @@
+/******************************************************************************
+ *  Copyright (c) 2022, Xilinx, Inc.
+ *  All rights reserved.
+ *
+ *  Redistribution and use in source and binary forms, with or without
+ *  modification, are permitted provided that the following conditions are met:
+ *
+ *  1.  Redistributions of source code must retain the above copyright notice,
+ *     this list of conditions and the following disclaimer.
+ *
+ *  2.  Redistributions in binary form must reproduce the above copyright
+ *      notice, this list of conditions and the following disclaimer in the
+ *      documentation and/or other materials provided with the distribution.
+ *
+ *  3.  Neither the name of the copyright holder nor the names of its
+ *      contributors may be used to endorse or promote products derived from
+ *      this software without specific prior written permission.
+ *
+ *  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ *  AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO,
+ *  THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+ *  PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR
+ *  CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ *  EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ *  PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
+ *  OR BUSINESS INTERRUPTION). HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
+ *  WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
+ *  OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
+ *  ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ * @brief	Validation top-level module for checksum component.
+ * @author	Thomas B. Preußer <tpreusse@amd.com>
+ *
+ *******************************************************************************/
+#include "checksum.hpp"
+CHECKSUM_TOP(WORDS_PER_FRAME, WORD_SIZE, ITEMS_PER_WORD)

finn_plus-1.0.0/custom_hls/checksum.hpp

@@ -0,0 +1,134 @@
+/******************************************************************************
+ *  Copyright (c) 2022, Xilinx, Inc.
+ *  All rights reserved.
+ *
+ *  Redistribution and use in source and binary forms, with or without
+ *  modification, are permitted provided that the following conditions are met:
+ *
+ *  1.  Redistributions of source code must retain the above copyright notice,
+ *     this list of conditions and the following disclaimer.
+ *
+ *  2.  Redistributions in binary form must reproduce the above copyright
+ *      notice, this list of conditions and the following disclaimer in the
+ *      documentation and/or other materials provided with the distribution.
+ *
+ *  3.  Neither the name of the copyright holder nor the names of its
+ *      contributors may be used to endorse or promote products derived from
+ *      this software without specific prior written permission.
+ *
+ *  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ *  AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO,
+ *  THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+ *  PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR
+ *  CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ *  EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ *  PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
+ *  OR BUSINESS INTERRUPTION). HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
+ *  WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
+ *  OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
+ *  ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ * @brief	Checksum over stream-carried data frames.
+ * @author	Thomas B. Preußer <tpreusse@amd.com>
+ *
+ *******************************************************************************/
+#include <hls_stream.h>
+#include <ap_int.h>
+
+
+/**
+ * Computes a checksum over a forwarded stream assumed to carry frames of
+ * N words further subdivided into K subwords.
+ *	- Subword slicing can be customized typically by using a lambda.
+ *	  The provided DefaultSubwordSlicer assumes an `ap_(u)int`-like word
+ *	  type with a member `width` and a range-based slicing operator. It
+ *	  further assumes a little-endian arrangement of subwords within words
+ *	  for the canonical subword stream order.
+ *	- Subwords wider than 23 bits are folded using bitwise XOR across
+ *	  slices of 23 bits starting from the LSB.
+ *	- The folded subword values are weighted according to their position
+ *	  in the stream relative to the start of frame by a periodic weight
+ *	  sequence 1, 2, 3, ...
+ *	- The weighted folded subword values are reduced to a checksum by an
+ *	  accumulation module 2^24.
+ *	- A checksum is emitted for each completed frame. It is the concatenation
+ *	  of an 8-bit (modulo 256) frame counter and the 24-bit frame checksum.
+ */
+template<typename T, unsigned K> class DefaultSubwordSlicer {
+	static_assert(T::width%K == 0, "Word size must be subword multiple.");
+	static constexpr unsigned  W = T::width/K;
+public:
+	ap_uint<W> operator()(T const &x, unsigned const  j) const {
+#pragma HLS inline
+		return  x((j+1)*W-1, j*W);
+	}
+};
+
+template<
+  unsigned N,	// number of data words in a frame
+  unsigned K,	// subword count per data word
+  typename T,	// type of stream-carried data words
+  typename F = DefaultSubwordSlicer<T, K>	// f(T(), j) to extract subwords
+>
+void checksum(
+	hls::stream<T> &src,
+	hls::stream<T> &dst,
+	ap_uint<32>    &chk,
+	ap_uint<1>     drain,	// drain data after checksuming without forward to `dst`
+	F&& f = F()
+) {
+	ap_uint<2>  coeff[3] = { 1, 2, 3 };
+	ap_uint<24>  s = 0;
+
+	for(unsigned  i = 0; i < N; i++) {
+#pragma HLS pipeline II=1 style=flp
+		T const  x = src.read();
+
+		// Pass-thru copy
+		if(!drain)  dst.write(x);
+
+		// Actual checksum update
+		for(unsigned  j = 0; j < K; j++) {
+#pragma HLS unroll
+			auto const   v0 = f(x, j);
+			constexpr unsigned  W = 1 + (decltype(v0)::width-1)/23;
+			ap_uint<K*23>  v = v0;
+			ap_uint<  23>  w = 0;
+			for(unsigned  k = 0; k < W; k++) {
+				w ^= v(23*k+22, 23*k);
+			}
+			s += (coeff[j%3][1]? (w, ap_uint<1>(0)) : ap_uint<24>(0)) + (coeff[j%3][0]? w : ap_uint<23>(0));
+		}
+
+		// Re-align coefficients
+		for(unsigned  j = 0; j < 3; j++) {
+#pragma HLS unroll
+			ap_uint<3> const  cc = coeff[j] + ap_uint<3>(K%3);
+			coeff[j] = cc(1, 0) + cc[2];
+		}
+	}
+
+	// Frame counter & output
+	static ap_uint<8>  cnt = 0;
+#pragma HLS reset variable=cnt
+	chk = (cnt++, s);
+}
+
+#define CHECKSUM_TOP_(WORDS_PER_FRAME, WORD_SIZE, ITEMS_PER_WORD) \
+	using  T = ap_uint<WORD_SIZE>; \
+	void checksum_ ## WORDS_PER_FRAME ## _ ## WORD_SIZE ## _ ## ITEMS_PER_WORD ( \
+		hls::stream<T> &src, \
+		hls::stream<T> &dst, \
+		ap_uint<32>    &chk, \
+		ap_uint< 1>    drain \
+	) { \
+	_Pragma("HLS interface port=src axis") \
+	_Pragma("HLS interface port=dst axis") \
+	_Pragma("HLS interface port=chk s_axilite") \
+	_Pragma("HLS interface port=drain s_axilite") \
+	_Pragma("HLS interface port=return ap_ctrl_none") \
+	_Pragma("HLS dataflow disable_start_propagation") \
+		checksum<WORDS_PER_FRAME, ITEMS_PER_WORD>(src, dst, chk, drain); \
+	}
+#define CHECKSUM_TOP(WORDS_PER_FRAME, WORD_SIZE, ITEMS_PER_WORD) \
+	CHECKSUM_TOP_(WORDS_PER_FRAME, WORD_SIZE, ITEMS_PER_WORD)

finn_plus-1.0.0/custom_hls/checksum_tb.sv

@@ -0,0 +1,136 @@
+/******************************************************************************
+ *  Copyright (c) 2022, Xilinx, Inc.
+ *  All rights reserved.
+ *
+ *  Redistribution and use in source and binary forms, with or without
+ *  modification, are permitted provided that the following conditions are met:
+ *
+ *  1.  Redistributions of source code must retain the above copyright notice,
+ *     this list of conditions and the following disclaimer.
+ *
+ *  2.  Redistributions in binary form must reproduce the above copyright
+ *      notice, this list of conditions and the following disclaimer in the
+ *      documentation and/or other materials provided with the distribution.
+ *
+ *  3.  Neither the name of the copyright holder nor the names of its
+ *      contributors may be used to endorse or promote products derived from
+ *      this software without specific prior written permission.
+ *
+ *  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ *  AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO,
+ *  THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+ *  PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR
+ *  CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ *  EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ *  PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
+ *  OR BUSINESS INTERRUPTION). HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
+ *  WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
+ *  OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
+ *  ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ * @brief	Testbench for checksum component.
+ * @author	Thomas B. Preußer <tpreusse@amd.com>
+ *
+ *******************************************************************************/
+module checksum_tb;
+
+	//-----------------------------------------------------------------------
+	// Global Control
+	logic  clk = 0;
+	always #5ns clk = !clk;
+	logic  rst;
+
+	//-----------------------------------------------------------------------
+	// DUT
+	localparam int unsigned  N = 60;	// words per frame
+	localparam int unsigned  K = 4;		// subwords per word
+	localparam int unsigned  W = 8;		// subword size
+
+	logic [K-1:0][W-1:0]  src_TDATA;
+	logic  src_TVALID;
+	uwire  src_TREADY;
+
+	uwire [K-1:0][W-1:0]  dst_TDATA;
+	uwire  dst_TVALID;
+	logic  dst_TREADY;
+
+	uwire [31:0]  chk;
+	uwire         chk_vld;
+
+	checksum_top dut (
+		.ap_clk(clk), .ap_rst_n(!rst),
+		.src_TDATA, .src_TVALID, .src_TREADY,
+		.dst_TDATA, .dst_TVALID, .dst_TREADY,
+		.chk, .chk_ap_vld(chk_vld),
+		.ap_local_block(), .ap_local_deadlock()
+	);
+
+	//-----------------------------------------------------------------------
+	// Stimulus
+	logic [K-1:0][W-1:0]  Bypass  [$] = {};
+	logic [31:0]          Checksum[$] = {};
+	initial begin
+		src_TDATA  = 'x;
+		src_TVALID =  0;
+
+		rst = 1;
+		repeat(9) @(posedge clk);
+		rst <= 0;
+
+		for(int unsigned  r = 0; r < 311; r++) begin
+			automatic logic [23:0]  sum = 0;
+			src_TVALID <= 1;
+			for(int unsigned  i = 0; i < N; i++) begin
+				for(int unsigned  k = 0; k < K; k++) begin
+					automatic logic [W-1:0]  v = $urandom()>>17;
+					src_TDATA[k] <= v;
+					sum += ((K*i+k)%3 + 1) * v;
+				end
+				@(posedge clk iff src_TREADY);
+				Bypass.push_back(src_TDATA);
+			end
+			src_TVALID <= 0;
+			$display("Expect: %02x:%06x", r[7:0], sum);
+			Checksum.push_back({r, sum});
+		end
+
+		repeat(8) @(posedge clk);
+		$finish;
+	end
+
+	//-----------------------------------------------------------------------
+	// Output Validation
+
+	// Drain and check pass-thru stream
+	assign	dst_TREADY = 1;
+	always_ff @(posedge clk iff dst_TVALID) begin
+		assert(Bypass.size()) begin
+			automatic logic [K-1:0][W-1:0]  exp = Bypass.pop_front();
+			assert(dst_TDATA === exp) else begin
+				$error("Unexpected output %0x instead of %0x.", dst_TDATA, exp);
+				$stop;
+			end
+		end
+		else begin
+			$error("Spurious data output.");
+			$stop;
+		end
+	end
+
+	// Validate checksum reports
+	always_ff @(posedge clk iff chk_vld) begin
+		$display("Check:  %02x:%06x", chk[31:24], chk[23:0]);
+		assert(Checksum.size()) begin
+			automatic logic [31:0]  exp = Checksum.pop_front();
+			assert(chk === exp) else begin
+				$error("Unexpected checksum %0x instead of %0x.", chk, exp);
+				$stop;
+			end
+		end
+		else begin
+			$error("Spurious checksum output.");
+			$stop;
+		end
+	end
+
+endmodule : checksum_tb