Merge pull request #180 from ComputationalRadiationPhysics/release-2.…

…4.0crp

Release 2.4.0crp

.clang-format

@@ -0,0 +1,77 @@
+---
+AccessModifierOffset: -4
+AlignAfterOpenBracket: AlwaysBreak
+AlignConsecutiveAssignments: false
+AlignConsecutiveDeclarations: false
+AlignEscapedNewlines: DontAlign
+AlignOperands: false
+AlignTrailingComments: false
+AllowAllParametersOfDeclarationOnNextLine: false
+AllowShortBlocksOnASingleLine: false
+AllowShortCaseLabelsOnASingleLine: false
+AllowShortFunctionsOnASingleLine: Empty
+AllowShortIfStatementsOnASingleLine: false
+AllowShortLoopsOnASingleLine: true
+AlwaysBreakAfterReturnType: None
+AlwaysBreakBeforeMultilineStrings: true
+AlwaysBreakTemplateDeclarations: Yes
+BinPackArguments: false
+BinPackParameters: false
+BreakBeforeBraces: Custom
+BraceWrapping:
+  AfterClass: true
+  AfterControlStatement: true
+  AfterEnum: true
+  AfterFunction: true
+  AfterNamespace: true
+  AfterStruct: true
+  AfterUnion: true
+  AfterExternBlock: true
+  BeforeCatch: true
+  BeforeElse: true
+  IndentBraces: false
+  SplitEmptyFunction: false
+  SplitEmptyRecord: false
+  SplitEmptyNamespace: false
+BreakBeforeBinaryOperators: All
+BreakBeforeTernaryOperators: true
+BreakConstructorInitializers: AfterColon
+BreakInheritanceList: AfterColon
+BreakStringLiterals: true
+ColumnLimit: 80
+CompactNamespaces: false
+ConstructorInitializerAllOnOneLineOrOnePerLine: true
+ConstructorInitializerIndentWidth: 8
+ContinuationIndentWidth: 4
+Cpp11BracedListStyle: true
+DerivePointerAlignment: false
+FixNamespaceComments: false
+IncludeBlocks: Regroup
+IndentCaseLabels: false
+IndentPPDirectives: None
+IndentWidth: 4
+IndentWrappedFunctionNames: false
+KeepEmptyLinesAtTheStartOfBlocks: false
+Language: Cpp
+NamespaceIndentation: All
+PointerAlignment: Middle
+ReflowComments: true
+SortIncludes: true
+SortUsingDeclarations: true
+SpaceAfterCStyleCast: false
+SpaceAfterTemplateKeyword: false
+SpaceBeforeAssignmentOperators: true
+SpaceBeforeCpp11BracedList: false
+SpaceBeforeCtorInitializerColon: true
+SpaceBeforeInheritanceColon: true
+SpaceBeforeParens: Never
+SpaceBeforeRangeBasedForLoopColon: true
+SpaceInEmptyParentheses: false
+SpacesInAngles: false
+SpacesInCStyleCastParentheses: false
+SpacesInContainerLiterals: false
+SpacesInParentheses: false
+SpacesInSquareBrackets: false
+Standard: Cpp11
+UseTab: Never
+...

.clang-tidy

@@ -0,0 +1,3 @@
+---
+Checks:            '*,-llvm-header-guard,-fuchsia-default-arguments-declarations,-cppcoreguidelines-no-malloc,-cppcoreguidelines-owning-memory,-misc-non-private-member-variables-in-classes'
+HeaderFilterRegex: '.*'

.gitignore

@@ -14,3 +14,5 @@
 
 *~
 /nbproject
+/.vs
+/build

.travis.yml

@@ -2,7 +2,7 @@ language: cpp
 
 sudo: required
 
-dist: trusty
+dist: bionic
 
 compiler:
   - gcc
@@ -14,6 +14,7 @@ env:
 
 script:
   - mkdir build_tmp && cd build_tmp
+  - CXX=g++-5 && CC=gcc-5
   - cmake -DCMAKE_INSTALL_PREFIX=$INSTALL_DIR $TRAVIS_BUILD_DIR
   - make
   - make install
@@ -28,12 +29,12 @@ before_script:
   - sudo apt-get install -f -qq
   - sudo dpkg --get-selections | grep hold || { echo "All packages OK."; }
   - sudo apt-get install -q -y cmake-data cmake
-  - sudo apt-get install -qq build-essential
-  - gcc --version && g++ --version # 4.8
+  - sudo apt-get install -qq build-essential g++-5
+  - gcc-5 --version && g++-5 --version # 5.5.0
   - apt-cache search nvidia-*
   - sudo apt-get install -qq nvidia-common
-  - sudo apt-get install -qq nvidia-cuda-dev nvidia-cuda-toolkit # 5.5
-  - sudo apt-get install -qq libboost-dev # 1.54.0
+  - sudo apt-get install -qq nvidia-cuda-dev nvidia-cuda-toolkit # 9.1.85
+  - sudo apt-get install -qq libboost-dev # 1.65.1
   - sudo find /usr/ -name libcuda*.so
 
 after_script:

CHANGELOG.md

@@ -1,6 +1,27 @@
 Change Log / Release Log for mallocMC
 ================================================================
 
+2.4.0crp
+--------
+**Date:** 2020-05-28
+
+This release removes the Boost dependency and switched to C++11.
+
+### Changes to mallocMC 2.3.1crp
+
+**Features**
+  - Cleaning, remove Boost dependency & C++11 Migration #169
+
+**Bug fixes**
+  - Choose the value for the -arch nvcc flag depending on CUDA version #164 #165
+
+**Misc:**
+  - Travis CI: GCC 5.5.0 + CUDA 9.1.85 #170
+  - Adding headers to projects and applied clang-tidy #171
+  - clang-format #172
+
+Thanks to  Sergei Bastrakov, Bernhard Manfred Gruber and Axel Huebl for contributing to this release!
+
 2.3.1crp
 --------
 **Date:** 2019-02-14

CMakeLists.txt

@@ -1,7 +1,12 @@
-project(mallocMC)
-cmake_minimum_required(VERSION 2.8.12.2)
+project(mallocMC LANGUAGES CUDA CXX)
+cmake_minimum_required(VERSION 3.8)
+
+set(CMAKE_CXX_STANDARD 11)
+set(CMAKE_CXX_STANDARD_REQUIRED ON)
 
 # helper for libs and packages
+set(CMAKE_CUDA_STANDARD 11)
+set(CMAKE_CUDA_STANDARD_REQUIRED ON)
 set(CMAKE_PREFIX_PATH "/usr/lib/x86_64-linux-gnu/"
     "$ENV{CUDA_ROOT}" "$ENV{BOOST_ROOT}")
 
@@ -14,64 +19,37 @@ set(CMAKE_PREFIX_PATH "/usr/lib/x86_64-linux-gnu/"
 ################################################################################
 
 if(POLICY CMP0074)
-    cmake_policy(SET CMP0074 NEW)
+  cmake_policy(SET CMP0074 NEW)
 endif()
 
 
 ###############################################################################
 # CUDA
 ###############################################################################
-find_package(CUDA REQUIRED)
-set(CUDA_NVCC_FLAGS "-arch=sm_20;-use_fast_math;")
-set(CUDA_INCLUDE_DIRS ${CMAKE_CURRENT_SOURCE_DIR})
-include_directories(${CUDA_INCLUDE_DIRS})
-cuda_include_directories(${CUDA_INCLUDE_DIRS})
+if(NOT DEFINED COMPUTE_CAPABILITY)
+  set(COMPUTE_CAPABILITY "30")
+endif()
+set(CMAKE_CUDA_FLAGS "${CMAKE_CUDA_FLAGS} -arch=sm_${COMPUTE_CAPABILITY} -use_fast_math")
 
 OPTION(CUDA_OUTPUT_INTERMEDIATE_CODE "Output ptx code" OFF)
 if(CUDA_OUTPUT_INTERMEDIATE_CODE)
-set(CUDA_NVCC_FLAGS "${CUDA_NVCC_FLAGS};-Xptxas;-v;--keep")
-endif(CUDA_OUTPUT_INTERMEDIATE_CODE)
+  set(CMAKE_CUDA_FLAGS "${CMAKE_CUDA_FLAGS} -Xptxas -v --keep")
+endif()
 
 SET(CUDA_OPTIMIZATION_TYPE "unset" CACHE STRING "CUDA Optimization")
 set_property(CACHE CUDA_OPTIMIZATION_TYPE PROPERTY STRINGS "unset;-G0;-O0;-O1;-O2;-O3")
-if(NOT ${CUDA_OPTIMIZATION_TYPE} STREQUAL  "unset")
-  set(CUDA_NVCC_FLAGS "${CUDA_NVCC_FLAGS};${CUDA_OPTIMIZATION_TYPE}")
+if(NOT ${CUDA_OPTIMIZATION_TYPE} STREQUAL "unset")
+  set(CMAKE_CUDA_FLAGS "${CMAKE_CUDA_FLAGS} ${CUDA_OPTIMIZATION_TYPE}")
 endif()
 
 
-###############################################################################
-# Boost
-###############################################################################
-find_package(Boost 1.48.0 REQUIRED)
-include_directories(SYSTEM ${Boost_INCLUDE_DIRS})
-set(LIBS ${LIBS} ${Boost_LIBRARIES})
-
-# nvcc + boost 1.55 work around
-if(Boost_VERSION EQUAL 105500)
-  set(CUDA_NVCC_FLAGS "${CUDA_NVCC_FLAGS} \"-DBOOST_NOINLINE=__attribute__((noinline))\" ")
-endif(Boost_VERSION EQUAL 105500)
-
-
 ################################################################################
 # Warnings
 ################################################################################
-# GNU
 if(CMAKE_COMPILER_IS_GNUCXX)
-  set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wall")
-  set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wshadow")
-  set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wno-unknown-pragmas")
-  set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wextra")
-  set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wno-unused-parameter")
-  # new warning in gcc 4.8 (flag ignored in previous version)
-  set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wno-unused-local-typedefs")
-# ICC
+  set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wall -Wshadow -Wno-unknown-pragmas -Wextra -Wno-unused-parameter -Wno-unused-local-typedefs")
 elseif("${CMAKE_CXX_COMPILER_ID}" STREQUAL "Intel")
-  set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wall")
-  set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wshadow")
-  set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -DBOOST_NO_VARIADIC_TEMPLATES")
-  set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -DBOOST_NO_CXX11_VARIADIC_TEMPLATES")
-  set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -DBOOST_NO_FENV_H")
-# PGI
+  set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wall -Wshadow")
 elseif("${CMAKE_CXX_COMPILER_ID}" STREQUAL "PGI")
   set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Minform=inform")
 endif()
@@ -87,28 +65,19 @@ INSTALL(
   DESTINATION include
   PATTERN ".git" EXCLUDE
   PATTERN "mallocMC_config.hpp" EXCLUDE
-  )
+)
 
 
 ###############################################################################
 # Executables
 ###############################################################################
+file(GLOB_RECURSE headers src/include/**)
+add_custom_target(mallocMC SOURCES ${headers}) # create a target with the header files for IDE projects
+source_group(TREE ${CMAKE_CURRENT_LIST_DIR}/src/include FILES ${headers})
+
+include_directories(${CMAKE_CURRENT_LIST_DIR}/src/include)
+add_executable(mallocMC_Example01 EXCLUDE_FROM_ALL examples/mallocMC_example01.cu examples/mallocMC_example01_config.hpp)
+add_executable(mallocMC_Example02 EXCLUDE_FROM_ALL examples/mallocMC_example02.cu)
+add_executable(mallocMC_Example03 EXCLUDE_FROM_ALL examples/mallocMC_example03.cu)
+add_executable(VerifyHeap EXCLUDE_FROM_ALL tests/verify_heap.cu tests/verify_heap_config.hpp)
 add_custom_target(examples DEPENDS mallocMC_Example01 mallocMC_Example02 mallocMC_Example03 VerifyHeap)
-
-cuda_add_executable(mallocMC_Example01
-                    EXCLUDE_FROM_ALL
-                    examples/mallocMC_example01.cu )
-cuda_add_executable(mallocMC_Example02
-                    EXCLUDE_FROM_ALL
-                    examples/mallocMC_example02.cu )
-cuda_add_executable(mallocMC_Example03
-                    EXCLUDE_FROM_ALL
-                    examples/mallocMC_example03.cu )
-cuda_add_executable(VerifyHeap
-                    EXCLUDE_FROM_ALL
-                    tests/verify_heap.cu )
-
-target_link_libraries(mallocMC_Example01 ${LIBS})
-target_link_libraries(mallocMC_Example02 ${LIBS})
-target_link_libraries(mallocMC_Example03 ${LIBS})
-target_link_libraries(VerifyHeap ${LIBS})

CONTRIBUTING.md

@@ -0,0 +1,15 @@
+# Contributing
+
+## Formatting
+
+Please format your code before before opening pull requests using clang-format and the .clang-format file placed in the repository root.
+
+### Visual Studio and CLion
+Suport for clang-format is built-in since Visual Studio 2017 15.7 and CLion 2019.1.
+The .clang-format file in the repository will be automatically detected and formatting is done as you type, or triggered when pressing the format hotkey.
+
+### Bash
+First install clang-format. Instructions therefore can be found on the web. To format you can run this command in bash:
+```
+find -iname *.cu -o -iname *.hpp | xargs clang-format-10 -i
+```

README.md

@@ -22,6 +22,11 @@ mallocMC is header-only, but requires a few other C++ libraries to be
 available. Our installation notes can be found in [INSTALL.md](INSTALL.md).
 
 
+Contributing
+------------
+
+Rules for contributions are found in [CONTRIBUTING.md](CONTRIBUTING.md).
+
 On the ScatterAlloc Algorithm
 -----------------------------
 

Usage.md

@@ -19,15 +19,15 @@ Currently, there are the following policy classes available:
 
 |Policy                 | Policy Classes (implementations) | description |
 |-------                |----------------------------------| ----------- |
-|**CreationPolicy**     | Scatter`<conf1,conf2>`           | A scattered allocation to tradeoff fragmentation for allocation time, as proposed in [ScatterAlloc](http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6339604). `conf1` configures the heap layout, `conf2` determines the hashing parameters|
+|**CreationPolicy**     | Scatter`<conf1,conf2>`         | A scattered allocation to tradeoff fragmentation for allocation time, as proposed in [ScatterAlloc](http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6339604). `conf1` configures the heap layout, `conf2` determines the hashing parameters|
 |                       | OldMalloc                        | device-side malloc/new and free/delete syscalls as implemented on NVidia CUDA graphics cards with compute capability sm_20 and higher |
-|**DistributionPolicy** | XMallocSIMD`<conf>`              | SIMD optimization for warp-wide allocation on NVIDIA CUDA accelerators, as proposed by [XMalloc](http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=5577907). `conf` is used to determine the pagesize. If used in combination with *Scatter*, the pagesizes must match |
+|**DistributionPolicy** | XMallocSIMD`<conf>`             | SIMD optimization for warp-wide allocation on NVIDIA CUDA accelerators, as proposed by [XMalloc](http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=5577907). `conf` is used to determine the pagesize. If used in combination with *Scatter*, the pagesizes must match |
 |                       | Noop                             | no workload distribution at all |
-|**OOMPolicy**          | ReturnNull                       | pointers will be *NULL*, if the request could not be fulfilled |
+|**OOMPolicy**          | ReturnNull                       | pointers will be *nullptr*, if the request could not be fulfilled |
 |                       | ~~BadAllocException~~            | will throw a `std::bad_alloc` exception. The accelerator has to support exceptions |
 |**ReservePoolPolicy**  | SimpleCudaMalloc                 | allocate a fixed heap with `CudaMalloc` |
 |                       | CudaSetLimits                    | call to `CudaSetLimits` to increase the available Heap (e.g. when using *OldMalloc*) |
-|**AlignmentPolicy**    | Shrink`<conf>`                   | shrinks the pool so that the starting pointer is well aligned, applies padding to requested memory chunks. `conf` is used to determine the alignment|
+|**AlignmentPolicy**    | Shrink`<conf>`                  | shrinks the pool so that the starting pointer is well aligned, applies padding to requested memory chunks. `conf` is used to determine the alignment|
 |                       | Noop                             | no alignment at all |
 
 The user has to choose one of each policy that will form a useful allocator
@@ -45,7 +45,7 @@ to the policy class:
 ```c++
 // configure the AlignmentPolicy "Shrink"
 struct ShrinkConfig : mallocMC::AlignmentPolicies::Shrink<>::Properties {
-  typedef boost::mpl::int_<16> dataAlignment;
+  static constexpr auto dataAlignment = 16;
 };
 ```
 
@@ -57,29 +57,29 @@ parameters to create the desired allocator type:
 ```c++
 using namespace mallocMC;
 
-typedef mallocMC::Allocator<
+using Allocator1 = mallocMC::Allocator<
   CreationPolicy::OldMalloc,
   DistributionPolicy::Noop,
   OOMPolicy::ReturnNull,
   ReservePoolPolicy::CudaSetLimits,
   AlignmentPolicy::Noop
-> Allocator1;
+>;
 ```
 
 `Allocator1` will resemble the behaviour of classical device-side allocation known
 from NVIDIA CUDA since compute capability sm_20. To get a more novel allocator, one
-could create the following typedef instead:
+could create the following alias instead:
 
 ```c++
 using namespace mallocMC;
 
-typedef mallocMC::Allocator<
+using ScatterAllocator = mallocMC::Allocator<
   CreationPolicies::Scatter<>,
   DistributionPolicies::XMallocSIMD<>,
   OOMPolicies::ReturnNull,
   ReservePoolPolicies::SimpleCudaMalloc,
   AlignmentPolicies::Shrink<ShrinkConfig>
-> ScatterAllocator;
+>;
 ```
 
 Notice, how the policy classes `Scatter` and `XMallocSIMD` are instantiated without
@@ -122,13 +122,13 @@ A simplistic example would look like this:
 
 namespace mallocMC = MC;
 
-typedef MC::Allocator<
+using ScatterAllocator = MC::Allocator<
   MC::CreationPolicies::Scatter<>,
   MC::DistributionPolicies::XMallocSIMD<>,
   MC::OOMPolicies::ReturnNull,
   MC::ReservePoolPolicies::SimpleCudaMalloc,
   MC::AlignmentPolicies::Shrink<ShrinkConfig>
-  > ScatterAllocator;
+>;
 
 __global__ exampleKernel(ScatterAllocator::AllocatorHandle sah)
 {