import project

.gitignore

@@ -0,0 +1,7 @@
+bin
+build
+resources/skel
+*.log
+*.aux
+*.toc
+*.synctex.gz

Image Skeletonisation - data.numbers/Metadata/BuildVersionHistory.plist

@@ -0,0 +1,8 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
+<plist version="1.0">
+<array>
+	<string>Template: Blank (2014-02-28 10:53)</string>
+	<string>M3.2.2-1878-1</string>
+</array>
+</plist>

Image Skeletonisation - data.numbers/Metadata/DocumentIdentifier

@@ -0,0 +1 @@
+620FD7AF-D9E8-471B-BB48-D3F36C68B59E

Image Skeletonisation - report.pages/Metadata/BuildVersionHistory.plist

@@ -0,0 +1,8 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
+<plist version="1.0">
+<array>
+	<string>Template: 09_School_Report (2014-02-28 10:44)</string>
+	<string>M5.2.2-1877-1</string>
+</array>
+</plist>

Image Skeletonisation - report.pages/Metadata/DocumentIdentifier

@@ -0,0 +1 @@
+04814554-5C89-4945-8F68-49A9666A4801

Image Skeletonisation - slides.key/Metadata/BuildVersionHistory.plist

@@ -0,0 +1,8 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
+<plist version="1.0">
+<array>
+	<string>Template: Black (2014-02-28 09:37)</string>
+	<string>M6.2.2-1878-1</string>
+</array>
+</plist>

Image Skeletonisation - slides.key/Metadata/DocumentIdentifier

@@ -0,0 +1 @@
+4547E3A7-6DD3-4BD3-A656-C85376E4581C

LICENSE

@@ -0,0 +1,19 @@
+Image Skeletonisation - Copyright (c) 2014-2015 Marco Antognini <[email protected]>
+
+This software is provided 'as-is', without any express or implied warranty.
+In no event will the authors be held liable for any damages arising from
+the use of this software.
+
+Permission is granted to anyone to use this software for any purpose,
+including commercial applications, and to alter it and redistribute it
+freely, subject to the following restrictions:
+
+1. The origin of this software must not be misrepresented; you must not claim
+   that you wrote the original software. If you use this software in a product,
+   an acknowledgement in the product documentation would be appreciated but is
+   not required.
+
+2. Altered source versions must be plainly marked as such, and must not be
+   misrepresented as being the original software.
+
+3. This notice may not be removed or altered from any source distribution.

Makefile

@@ -0,0 +1,131 @@
+CXX                 = clang++-3.5
+CXXCFLAGS           = -std=c++11 -stdlib=libc++ -Wall -Ofast \
+                      -nostdinc++ -I/usr/local/lib/llvm-3.5/include/c++/v1
+CXXLDFLAGS          = -L lib/libc++/ -stdlib=libc++ \
+                      -lsfml-system-s -lsfml-graphics-s -lboost_program_options-s -ljpeg-s \
+                      -L/usr/local/lib/llvm-3.5/usr/lib
+
+GXX                 = g++-4.9
+GXXCFLAGS           = -std=c++11 -Wall -Ofast
+GXXLDFLAGS          = -L lib/stdlibc++/ \
+                      -lsfml-system-s -lsfml-graphics-s -lboost_program_options-s -ljpeg-s
+
+NVCC                = nvcc
+NVCCCFLAGS          = -Xcompiler -stdlib=libstdc++ -O3
+NVCCLDFLAGS         = -L lib/stdlibc++/ -Xcompiler -stdlib=libstdc++ \
+                      -lsfml-system-s -lsfml-graphics-s -lboost_program_options-s -ljpeg-s
+
+SERIALCFLAGS        =
+SERIALLDFLAGS       =
+
+CUDACFLAGS          = -DTHRUST_DEVICE_SYSTEM=THRUST_DEVICE_SYSTEM_CUDA
+CUDALDFLAGS         =
+
+TBBCFLAGS           = -x c++ -DTHRUST_DEVICE_SYSTEM=THRUST_DEVICE_SYSTEM_TBB \
+                      -I /Developer/NVIDIA/CUDA-6.5/include
+TBBLDFLAGS          = -ltbb
+
+OMPCFLAGS           = -x c++ -fopenmp -DTHRUST_DEVICE_SYSTEM=THRUST_DEVICE_SYSTEM_OMP \
+                      -I /Developer/NVIDIA/CUDA-6.5/include
+OMPLDFLAGS          = -lgomp
+
+BUILD_DIR           = build/
+BIN_DIR             = bin/
+
+COMMON_DIR          = src/common/
+SERIAL_DIR          = src/serial/
+PAR_DIR             = src/parallel/
+
+SOURCES_COMMON      = main.cpp
+SOURCES_SERIAL      = Image.cpp Skeleton.cpp
+SOURCES_PAR         = Image.cu Skeleton.cu
+EXECUTABLE_SERIAL   = skeleton_serial
+EXECUTABLE_CUDA     = skeleton_cuda
+EXECUTABLE_TBB      = skeleton_tbb
+EXECUTABLE_OMP      = skeleton_omp
+
+SOURCES_COMMON     := $(addprefix $(COMMON_DIR)/, $(SOURCES_COMMON))
+SOURCES_SERIAL     := $(addprefix $(SERIAL_DIR)/, $(SOURCES_SERIAL)) $(SOURCES_COMMON)
+SOURCES_PAR        := $(addprefix $(PAR_DIR)/, $(SOURCES_PAR)) $(SOURCES_COMMON)
+OBJECTS_SERIAL     := $(addsuffix .serial.o, $(notdir $(SOURCES_SERIAL)))
+OBJECTS_PAR        := $(addsuffix .cuda.o, $(notdir $(SOURCES_PAR)))
+OBJECTS_TBB        := $(addsuffix .tbb.o, $(notdir $(SOURCES_PAR)))
+OBJECTS_OMP        := $(addsuffix .omp.o, $(notdir $(SOURCES_PAR)))
+OBJECTS_SERIAL     := $(addprefix $(BUILD_DIR)/, $(OBJECTS_SERIAL))
+OBJECTS_PAR        := $(addprefix $(BUILD_DIR)/, $(OBJECTS_PAR))
+OBJECTS_TBB        := $(addprefix $(BUILD_DIR)/, $(OBJECTS_TBB))
+OBJECTS_OMP        := $(addprefix $(BUILD_DIR)/, $(OBJECTS_OMP))
+
+EXECUTABLES         = $(EXECUTABLE_SERIAL) $(EXECUTABLE_CUDA) $(EXECUTABLE_TBB) $(EXECUTABLE_OMP)
+
+
+## This hack fixes foreach commands below
+define \n
+
+
+endef
+
+all: $(EXECUTABLES)
+
+serial: $(EXECUTABLE_SERIAL)
+
+cuda: $(EXECUTABLE_CUDA)
+
+tbb: $(EXECUTABLE_TBB)
+
+omp: $(EXECUTABLE_OMP)
+
+$(EXECUTABLE_SERIAL): $(OBJECTS_SERIAL)
+	$(CXX) $(CXXLDFLAGS) $(SERIALLDFLAGS) $(OBJECTS_SERIAL) -o $(BIN_DIR)/$@
+
+$(EXECUTABLE_CUDA): $(OBJECTS_PAR)
+	$(NVCC) $(NVCCLDFLAGS) $(CUDALDFLAGS) $(OBJECTS_PAR) -o $(BIN_DIR)/$@
+
+$(EXECUTABLE_TBB): $(OBJECTS_TBB)
+	$(CXX) $(CXXLDFLAGS) $(TBBLDFLAGS) $(OBJECTS_TBB) -o $(BIN_DIR)/$@
+
+$(EXECUTABLE_OMP): $(OBJECTS_OMP)
+	$(GXX) $(GXXLDFLAGS) $(OMPLDFLAGS) $(OBJECTS_OMP) -o $(BIN_DIR)/$@
+
+$(OBJECTS_SERIAL): $(SOURCES_SERIAL) | dirs
+	$(foreach file, $?, \
+		$(CXX) $(CXXCFLAGS) $(SERIALCFLAGS) -I $(COMMON_DIR) -I $(SERIAL_DIR) \
+		-c $(file) \
+		-o $(BUILD_DIR)/$(notdir $(addsuffix .serial.o, $(file))) \
+		${\n} \
+	)
+
+$(OBJECTS_PAR): $(SOURCES_PAR) | dirs
+	$(foreach file, $?, \
+		$(NVCC) $(NVCCCFLAGS) $(CUDACFLAGS) -I $(COMMON_DIR) -I $(PAR_DIR) \
+		-c $(file) \
+		-o $(BUILD_DIR)/$(notdir $(addsuffix .cuda.o, $(file))) \
+		${\n} \
+	)
+
+$(OBJECTS_TBB): $(SOURCES_PAR) | dirs
+	$(foreach file, $?, \
+		$(CXX) $(CXXCFLAGS) $(TBBCFLAGS) -I $(COMMON_DIR) -I $(SERIAL_DIR) \
+		-c $(file) \
+		-o $(BUILD_DIR)/$(notdir $(addsuffix .tbb.o, $(file))) \
+		${\n} \
+	)
+
+$(OBJECTS_OMP): $(SOURCES_PAR) | dirs
+	$(foreach file, $?, \
+		$(GXX) $(GXXCFLAGS) $(OMPCFLAGS) -I $(COMMON_DIR) -I $(SERIAL_DIR) \
+		-c $(file) \
+		-o $(BUILD_DIR)/$(notdir $(addsuffix .omp.o, $(file))) \
+		${\n} \
+	)
+
+dirs:
+	mkdir -p $(BUILD_DIR) $(BIN_DIR)
+
+clean:
+	rm -f $(addprefix $(BIN_DIR)/, $(EXECUTABLES)) $(BUILD_DIR)/*.o
+
+run: $(EXECUTABLES)
+	./scripts/run.sh
+
+

README

@@ -0,0 +1,51 @@
+Image Skeletonisation
+=====================
+
+A performance comparison with Thrust
+
+
+Description
+-----------
+
+This project was realised during the fall semester 2014 at EPFL for CS-425,
+Program parallelisation on PC clusters.
+
+It aims to compare a theoretical analysis of the parallelisation of the image
+skeletonisation problem against an actual implementation in C++ using Thrust,
+a high level algorithm library with CUDA, TBB and OpenMP backends. The goal is
+to compare the speedup of the parallel implementation and highlight the effect
+of some hardware properties on the scalability of the performances.
+
+The full report is available in `Image Skeletonisation - report.pdf`.
+
+
+Build Instructions
+------------------
+
+To run all implementations and measure running time, use `make run`. You will
+need the software described in the `Hardware & Software Specifications` section
+of the report.
+
+Binaries of the required boost, SFML and libjpeg dependencies for Mac OS X are
+provided for convenience. If you are not using this operating system or the same
+compilers, you will need to build those yourself and potentially edit the
+`Makefile`.
+
+Adding `-DENABLE_SAVE_IMAGE` to the compilation flags will enable saving the
+skeletonised image and include I/O times in the running time measurement.
+
+Similarly, `-DENABLE_PARALLEL_SHORT_CIRCUIT_WHITE` enable some optimisations
+in the parallel implementation. Note that on CUDA this might decrease the
+speedup due to thread divergence.
+
+
+License
+-------
+
+The source code included in this project is distributed under the zlib/png
+license. Please refer to LICENSE for the full text.
+
+The boost library is distributed under Boost Software License, which is
+available in lib/BOOST_LICENSE. The SFML library is distributed under
+zlib/png license, which is available in lib/SFML_LICENSE. License for libjpeg is
+available in lib/LIBJPEG_LICENSE.