gpuNUFFT 0.7.4__tar.gz → 0.8.0__tar.gz

gpunufft-0.8.0/.clang-format

@@ -0,0 +1,48 @@
+---
+# BasedOnStyle:  LLVM
+AccessModifierOffset: -1
+ConstructorInitializerIndentWidth: 2
+AlignEscapedNewlinesLeft: false
+AlignTrailingComments: true
+AllowShortFunctionsOnASingleLine: false
+AllowAllParametersOfDeclarationOnNextLine: true
+AllowShortIfStatementsOnASingleLine: false
+AllowShortLoopsOnASingleLine: false
+AlwaysBreakTemplateDeclarations: false
+AlwaysBreakBeforeMultilineStrings: false
+BreakBeforeBinaryOperators: false
+BreakBeforeTernaryOperators: true
+BreakConstructorInitializersBeforeComma: false
+BinPackParameters: true
+ColumnLimit:     80
+ConstructorInitializerAllOnOneLineOrOnePerLine: false
+DerivePointerBinding: false
+ExperimentalAutoDetectBinPacking: false
+IndentCaseLabels: false
+MaxEmptyLinesToKeep: 1
+NamespaceIndentation: None
+ObjCSpaceBeforeProtocolList: true
+PenaltyBreakBeforeFirstCallParameter: 19
+PenaltyBreakComment: 60
+PenaltyBreakString: 1000
+PenaltyBreakFirstLessLess: 120
+PenaltyExcessCharacter: 1000000
+PenaltyReturnTypeOnItsOwnLine: 60
+PointerBindsToType: false
+SpacesBeforeTrailingComments: 2
+Cpp11BracedListStyle: false
+Standard:        Cpp11
+IndentWidth:     2
+TabWidth:        2
+UseTab:          Never
+BreakBeforeBraces: Allman
+IndentFunctionDeclarationAfterType: false
+SpacesInParentheses: false
+SpacesInAngles:  false
+SpaceInEmptyParentheses: false
+SpacesInCStyleCastParentheses: false
+SpaceAfterControlStatementKeyword: true
+SpaceBeforeAssignmentOperators: true
+ContinuationIndentWidth: 4
+...
+

gpunufft-0.8.0/.gitignore

@@ -0,0 +1,49 @@
+# Compiled Object files
+*.slo
+*.lo
+*.o
+
+# Compiled Dynamic libraries
+*.so
+*.dylib
+
+# Compiled Static libraries
+*.lai
+*.la
+*.a
+
+# others
+common/
+lib/
+bin/
+/gpuNUFFT/GPU_gpuNUFFT_3D/data/
+/gpuNUFFT/GPU_gpuNUFFT_3D/functions/trash/
+/gpuNUFFT/gpuNUFFT/build/
+test_03/build/
+gpuNUFFT/NUFFT_recon_demo_2D/kspaceRadial.mat
+gpuNUFFT/gpuNUFFT/bin
+*.asv
+*.zip
+gpuNUFFT/daten/results/
+gpuNUFFT/gpuNUFFT/matlab/script.out.*
+*.mexa64
+*.mexw32
+*.mexw64
+*.m~
+gpuNUFFT/bUTE/results/
+gpuNUFFT/bUTE/data/
+gpuNUFFT/3D_radial_bUTE_test/results/
+gpuNUFFT/bUTE/script.out*
+gpuNUFFT/daten/
+CUDA/inc/*config.hpp
+build/
+*.out
+*.out.*
+inc/*config.hpp
+*.mat
+tmp/
+*.swp
+*.hdr
+*.cfl
+
+matlab/demo/results/

gpunufft-0.8.0/.ycm_extra_conf.py

@@ -0,0 +1,162 @@
+# This file is NOT licensed under the GPLv3, which is the license for the rest
+# of YouCompleteMe.
+#
+# Here's the license text for this file:
+#
+# This is free and unencumbered software released into the public domain.
+#
+# Anyone is free to copy, modify, publish, use, compile, sell, or
+# distribute this software, either in source code form or as a compiled
+# binary, for any purpose, commercial or non-commercial, and by any
+# means.
+#
+# In jurisdictions that recognize copyright laws, the author or authors
+# of this software dedicate any and all copyright interest in the
+# software to the public domain. We make this dedication for the benefit
+# of the public at large and to the detriment of our heirs and
+# successors. We intend this dedication to be an overt act of
+# relinquishment in perpetuity of all present and future rights to this
+# software under copyright law.
+#
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+# EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+# MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+# IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY CLAIM, DAMAGES OR
+# OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+# ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+# OTHER DEALINGS IN THE SOFTWARE.
+#
+# For more information, please refer to <http://unlicense.org/>
+
+import os
+import ycm_core
+
+# These are the compilation flags that will be used in case there's no
+# compilation database set (by default, one is not set).
+# CHANGE THIS LIST OF FLAGS. YES, THIS IS THE DROID YOU HAVE BEEN LOOKING FOR.
+flags = [
+'-Wall',
+'-Wextra',
+'-Werror',
+#'-Wc++98-compat',
+'-Wno-long-long',
+'-Wno-variadic-macros',
+'-fexceptions',
+'-DNDEBUG',
+'-DUSE_CLANG_COMPLETER',
+# THIS IS IMPORTANT! Without a "-std=<something>" flag, clang won't know which
+# language to use when compiling headers. So it will guess. Badly. So C++
+# headers will be compiled as C headers. You don't want that so ALWAYS specify
+# a "-std=<something>".
+# For a C project, you would set this to something like 'c99' instead of
+# 'c++11'.
+#'-std=c++',
+# ...and the same thing goes for the magic -x option which specifies the
+# language that the files to be compiled are written in. This is mostly
+# relevant for c++ headers.
+# For a C project, you would set this to 'c' instead of 'c++'.
+'-x',
+'c++',
+'-isystem',
+'../BoostParts',
+'-isystem',
+# This path will only work on OS X, but extra paths that don't exist are not
+# harmful
+'/System/Library/Frameworks/Python.framework/Headers',
+'-isystem',
+'../llvm/include',
+'-isystem',
+'../llvm/tools/clang/include',
+'-I',
+'.',
+'-I',
+'./ClangCompleter',
+'-isystem',
+'./tests/gmock/gtest',
+'-isystem',
+'./tests/gmock/gtest/include',
+'-isystem',
+'./tests/gmock',
+'-isystem',
+'./tests/gmock/include',
+'-isystem',
+'/usr/include/flycapture',
+'-isystem',
+'./test/',
+'-isystem',
+'./dependencies/gtest',
+'-isystem',
+'./dependencies/gtest/include'
+]
+
+# Set this to the absolute path to the folder (NOT the file!) containing the
+# compile_commands.json file to use that instead of 'flags'. See here for
+# more details: http://clang.llvm.org/docs/JSONCompilationDatabase.html
+#
+# Most projects will NOT need to set this to anything; you can just change the
+# 'flags' list of compilation flags. Notice that YCM itself uses that approach.
+compilation_database_folder = ''
+
+if compilation_database_folder:
+  database = ycm_core.CompilationDatabase( compilation_database_folder )
+else:
+  database = None
+
+
+def DirectoryOfThisScript():
+  return os.path.dirname( os.path.abspath( __file__ ) )
+
+
+def MakeRelativePathsInFlagsAbsolute( flags, working_directory ):
+  if not working_directory:
+    return list( flags )
+  new_flags = []
+  make_next_absolute = False
+  path_flags = [ '-isystem', '-I', '-iquote', '--sysroot=' ]
+  for flag in flags:
+    new_flag = flag
+
+    if make_next_absolute:
+      make_next_absolute = False
+      if not flag.startswith( '/' ):
+        new_flag = os.path.join( working_directory, flag )
+
+    for path_flag in path_flags:
+      if flag == path_flag:
+        make_next_absolute = True
+        break
+
+      if flag.startswith( path_flag ):
+        path = flag[ len( path_flag ): ]
+        new_flag = path_flag + os.path.join( working_directory, path )
+        break
+
+    if new_flag:
+      new_flags.append( new_flag )
+  return new_flags
+
+
+def FlagsForFile( filename ):
+  if database:
+    # Bear in mind that compilation_info.compiler_flags_ does NOT return a
+    # python list, but a "list-like" StringVec object
+    compilation_info = database.GetCompilationInfoForFile( filename )
+    final_flags = MakeRelativePathsInFlagsAbsolute(
+      compilation_info.compiler_flags_,
+      compilation_info.compiler_working_dir_ )
+
+    # NOTE: This is just for YouCompleteMe; it's highly likely that your project
+    # does NOT need to remove the stdlib flag. DO NOT USE THIS IN YOUR
+    # ycm_extra_conf IF YOU'RE NOT 100% YOU NEED IT.
+    try:
+      final_flags.remove( '-stdlib=libc++' )
+    except ValueError:
+      pass
+  else:
+    relative_to = DirectoryOfThisScript()
+    final_flags = MakeRelativePathsInFlagsAbsolute( flags, relative_to )
+
+  return {
+    'flags': final_flags,
+    'do_cache': True
+  }

gpunufft-0.8.0/CUDA/bin/cvp/Gridding3D.cvp

@@ -0,0 +1,34 @@
+<!DOCTYPE cvp>
+<cvp version="4.0" >
+ <session>
+  <name>Session1</name>
+  <progpath>"E:/Andi/workspaces/cuda/cuda-test/gridding3D/gridding3D/bin/runGPUUnitTests.exe"</progpath>
+  <workdirpath>E:/Andi/workspaces/cuda/cuda-test/gridding3D/gridding3D/bin</workdirpath>
+  <datetime>16 Aug 2012 16:01:37</datetime>
+  <normalizedcounter>false</normalizedcounter>
+  <device multProcessorCount="2" label="Device_0" id="0" computeCapability="1.1" maxGridDim="65535 65535" GpuOverlap="0" maxPitchMem="2147483647" totalConstMem="65536" name="GeForce 9500M (GF 9400MG + GF 9200MGS)" clockFreq="1100000" L2CacheSize="0" maxRegPerBlock="8192" chip="172" maxThreadPerSM="768" maxShareMemPerBlock="16384" maxBlockDim="512 512 64" MemClockRate="1064000" asyncEngineCount="0" warpSize="32" maxThreadPerBlock="512" MemBandwidth="272384000" textureAlignment="256" >
+   <context totalcountercount="0" invalidcountercount="0" id="0" label="Context_0" truncatedcountercount="0" trace="0" >
+    <run number="2" >branch,divergent_branch,instructions,warp_serialize</run>
+    <run number="3" >cta_launched,gld_incoherent,gld_coherent,gst_incoherent</run>
+    <run number="4" >gst_coherent,local_load,local_store</run>
+    <run number="5" >tex_cache_hit,tex_cache_miss</run>
+   </context>
+  </device>
+ </session>
+ <session>
+  <name>Session2</name>
+  <progpath>"E:/Andi/workspaces/cuda/cuda-test/gridding3D/gridding3D/bin/runGPUUnitTests.exe"</progpath>
+  <workdirpath>E:/Andi/workspaces/cuda/cuda-test/gridding3D/gridding3D/bin</workdirpath>
+  <datetime>16 Aug 2012 19:25:24</datetime>
+  <normalizedcounter>false</normalizedcounter>
+  <device multProcessorCount="2" label="Device_0" id="0" computeCapability="1.1" maxGridDim="65535 65535" GpuOverlap="0" maxPitchMem="2147483647" totalConstMem="65536" name="GeForce 9500M (GF 9400MG + GF 9200MGS)" clockFreq="1100000" L2CacheSize="0" maxRegPerBlock="8192" chip="172" maxThreadPerSM="768" maxShareMemPerBlock="16384" maxBlockDim="512 512 64" MemClockRate="1064000" asyncEngineCount="0" warpSize="32" maxThreadPerBlock="512" MemBandwidth="272384000" textureAlignment="256" >
+   <context totalcountercount="0" invalidcountercount="0" id="0" label="Context_0" truncatedcountercount="0" trace="0" >
+    <run number="2" >branch,divergent_branch,instructions,warp_serialize</run>
+    <run number="3" >cta_launched,gld_incoherent,gld_coherent,gst_incoherent</run>
+    <run number="4" >gst_coherent,local_load,local_store</run>
+    <run number="5" >tex_cache_hit,tex_cache_miss</run>
+   </context>
+  </device>
+ </session>
+</cvp>
+

gpunufft-0.8.0/CUDA/bin/cvp/README

@@ -0,0 +1,4 @@
+USAGE of CUDA Visual Profiler under LINUX
+
+export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda/computeprof/bin
+

gpunufft-0.8.0/CUDA/bin/gtest_options

@@ -0,0 +1,3 @@
+repeat tests: --gtest_repeat=n --gtest_break_on_failure
+select single tests, filter: --gtest_filter=TestGridding.GPUTest_8SectorsKernel3nDataw32
+write output to xml file: --gtest_output="xml:<file name>"  

gpunufft-0.8.0/CUDA/bin/run_cpu_filtered_unit_test.cmd

@@ -0,0 +1 @@
+runUnitTests.exe --gtest_filter=TestGridding.CPUTest_8SectorsKernel3nDataw32 --gtest_repeat=100 --gtest_break_on_failure

gpunufft-0.8.0/CUDA/bin/run_cuda_memcheck

@@ -0,0 +1,4 @@
+export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda/lib64;
+
+/usr/local/cuda/bin/cuda-memcheck runGPUUnitTests
+ 

gpunufft-0.8.0/CUDA/bin/run_gpu_filtered_unit_test.cmd

@@ -0,0 +1 @@
+runGPUUnitTests.exe --gtest_filter=TestGridding.GPUTest_8SectorsKernel3nDataw32 --gtest_repeat=100 --gtest_break_on_failure

gpunufft-0.8.0/CUDA/build/configure_double_prec

@@ -0,0 +1 @@
+cmake -G "Visual Studio 10" .. -DGPU_DOUBLE_PREC=ON

gpunufft-0.8.0/CUDA/build/configure_single_prec

@@ -0,0 +1 @@
+cmake -G "Visual Studio 10" ..

gpunufft-0.8.0/CUDA/cmake_modules/FindGPUNUFFT.cmake

@@ -0,0 +1,24 @@
+INCLUDE(FindPackageHandleStandardArgs)
+
+IF (NOT DEFINED GPUNUFFT_ROOT_DIR)
+  MESSAGE(FATAL_ERROR "GPUNUFFT_ROOT_DIR not defined!")
+ELSE()
+  MESSAGE("Searching for GPUNUFFT in ${GPUNUFFT_ROOT_DIR}")
+
+  SET(GPUNUFFT_IncludeSearchPaths ${GPUNUFFT_ROOT_DIR}/CUDA/inc ${GPUNUFFT_ROOT_DIR}/inc) #/usr/local/include
+  SET(GPUNUFFT_LibrarySearchPaths ${GPUNUFFT_ROOT_DIR}/CUDA/bin ${GPUNUFFT_ROOT_DIR}/bin) #/usr/local/lib
+
+  FIND_PATH(GPUNUFFT_INCLUDE_DIRS NAMES gpuNUFFT_operator_factory.hpp PATHS ${GPUNUFFT_IncludeSearchPaths})
+
+  FIND_LIBRARY(GPUNUFFT_LIBRARIES NAMES gpuNUFFT gpuNUFFT_ATM_f gpuNUFFT_f  PATHS ${GPUNUFFT_LibrarySearchPaths})
+
+  #Handle the REQUIRED argument and set the < UPPERCASED_NAME > _FOUND variable
+  FIND_PACKAGE_HANDLE_STANDARD_ARGS(GPUNUFFT "Could NOT find GPUNUFFT." GPUNUFFT_LIBRARIES GPUNUFFT_INCLUDE_DIRS)
+  IF(GPUNUFFT_FOUND)
+    FIND_PACKAGE_MESSAGE(
+        GPUNUFFT      
+        "Found GPUNUFFT ${GPUNUFFT_LIBRARIES}"
+        "[${GPUNUFFT_LIBRARIES}][${GPUNUFFT_INCLUDE_DIRS}]")
+    ENDIF(GPUNUFFT_FOUND)
+  MARK_AS_ADVANCED(GPUNUFFT_INCLUDE_DIRS GPUNUFFT_LIBRARIES)
+ENDIF()

{gpuNUFFT-0.7.4 → gpunufft-0.8.0}/CUDA/inc/gpuNUFFT_operator.hpp

@@ -53,13 +53,13 @@ class GpuNUFFTOperator
   GpuNUFFTOperator(IndType kernelWidth, IndType sectorWidth, DType osf,
                    Dimensions imgDims, bool loadKernel = true,
                    OperatorType operatorType = DEFAULT,
-                   bool matlabSharedMem = false)
+                   bool matlabSharedMem = false, bool grad_mode = false)
     : operatorType(operatorType), osf(osf), kernelWidth(kernelWidth),
       sectorWidth(sectorWidth), imgDims(imgDims), gpuMemAllocated(false),
       debugTiming(DEBUG), sens_d(NULL), crds_d(NULL), density_comp_d(NULL),
       deapo_d(NULL), gdata_d(NULL), sector_centers_d(NULL), sectors_d(NULL),
       data_indices_d(NULL), data_sorted_d(NULL), allocatedCoils(0),
-      matlabSharedMem(matlabSharedMem)
+      matlabSharedMem(matlabSharedMem), grad_mode(grad_mode)
   {
     if (loadKernel)
       initKernel();
@@ -342,6 +342,14 @@ class GpuNUFFTOperator
                                           GpuNUFFTOutput gpuNUFFTOut);
 
   void clean_memory();
+
+  void setGradMode(bool grad_mode) {
+    this->grad_mode = grad_mode;
+  }
+
+  bool getGradMode() {
+    return this->grad_mode;
+  }
   /** \brief Check if density compensation data is available. */
   bool applyDensComp()
   {
@@ -452,6 +460,10 @@ class GpuNUFFTOperator
   */
   bool matlabSharedMem;
 
+  /** \brief Flag for changing the isign, mainly used for gradients 
+  */
+  bool grad_mode;
+
   /** \brief Return Grid Width (ImageWidth * osf) */
   IndType getGridWidth()
   {

{gpuNUFFT-0.7.4 → gpunufft-0.8.0}/CUDA/src/gpu/python/gpuNUFFT_operator_python_factory.cpp

@@ -141,7 +141,7 @@ class GpuNUFFTPythonOperator
 
         // density compensation weights
         gpuNUFFT::Array<DType> density_compArray;
-        if(density_comp == Py_None)
+        if(density_comp != Py_None)
         {
             density_compArray = readNumpyArray(density_comp);
             density_compArray.dim.length = trajectory_length;
@@ -291,6 +291,12 @@ class GpuNUFFTPythonOperator
        gpuNUFFTOp->clean_memory();
     }
 
+    void toggle_grad_mode()
+    {
+        bool current_mode = gpuNUFFTOp->getGradMode();
+        gpuNUFFTOp->setGradMode(!current_mode);
+    }
+
     void set_smaps(py::array_t<std::complex<DType>> sense_maps)
     {
         CAST_POINTER_VARNAME(sense_maps, sensArray);
@@ -436,6 +442,7 @@ PYBIND11_MODULE(gpuNUFFT, m) {
         .def("clean_memory", &GpuNUFFTPythonOperator::clean_memory)
         .def("estimate_density_comp", &GpuNUFFTPythonOperator::estimate_density_comp, py::arg("max_iter") = 10)
         .def("set_smaps", &GpuNUFFTPythonOperator::set_smaps)
+        .def("toggle_grad_mode", &GpuNUFFTPythonOperator::toggle_grad_mode)
         .def("get_spectral_radius", &GpuNUFFTPythonOperator::get_spectral_radius, py::arg("max_iter") = 20, py::arg("tolerance") = 1e-6);
 }
 #endif  // GPUNUFFT_OPERATOR_PYTHONFACTORY_H_INCLUDED

{gpuNUFFT-0.7.4 → gpunufft-0.8.0}/CUDA/src/gpuNUFFT_operator.cpp

@@ -1026,7 +1026,7 @@ void gpuNUFFT::GpuNUFFTOperator::performForwardGpuNUFFT(
     {
       if ((err = pt2CufftExec(fft_plan, gdata_d + c * gi_host->gridDims_count,
                               gdata_d + c * gi_host->gridDims_count,
-                              CUFFT_FORWARD)) != CUFFT_SUCCESS)
+                              grad_mode?CUFFT_INVERSE:CUFFT_FORWARD)) != CUFFT_SUCCESS)
       {
         fprintf(stderr, "cufft has failed with err %i \n", err);
         showMemoryInfo(true, stderr);
@@ -1037,7 +1037,7 @@ void gpuNUFFT::GpuNUFFTOperator::performForwardGpuNUFFT(
     if (DEBUG && (cudaStreamSynchronize(new_stream) != cudaSuccess))
       printf("error at thread synchronization 5: %s\n",
              cudaGetErrorString(cudaGetLastError()));
-    performFFTShift(gdata_d, FORWARD, getGridDims(), gi_host);
+    performFFTShift(gdata_d, grad_mode?INVERSE:FORWARD, getGridDims(), gi_host);
 
     if (DEBUG && (cudaStreamSynchronize(new_stream) != cudaSuccess))
       printf("error at thread synchronization 6: %s\n",
@@ -1240,7 +1240,7 @@ void gpuNUFFT::GpuNUFFTOperator::performForwardGpuNUFFT(
     {
       if ((err = pt2CufftExec(fft_plan, gdata_d + c * gi_host->gridDims_count,
                               gdata_d + c * gi_host->gridDims_count,
-                              CUFFT_FORWARD)) != CUFFT_SUCCESS)
+                              grad_mode?CUFFT_INVERSE:CUFFT_FORWARD)) != CUFFT_SUCCESS)
       {
         fprintf(stderr, "cufft has failed with err %i \n", err);
         showMemoryInfo(true, stderr);
@@ -1251,7 +1251,7 @@ void gpuNUFFT::GpuNUFFTOperator::performForwardGpuNUFFT(
     if (DEBUG && (cudaStreamSynchronize(new_stream) != cudaSuccess))
       printf("error at thread synchronization 5: %s\n",
              cudaGetErrorString(cudaGetLastError()));
-    performFFTShift(gdata_d, FORWARD, getGridDims(), gi_host);
+    performFFTShift(gdata_d, grad_mode?INVERSE:FORWARD, getGridDims(), gi_host);
 
     if (DEBUG && (cudaStreamSynchronize(new_stream) != cudaSuccess))
       printf("error at thread synchronization 6: %s\n",

gpunufft-0.8.0/CUDA/test/CMakeLists.txt

@@ -0,0 +1,30 @@
+#setup testframework
+MESSAGE("------setting up testframework-------")
+
+set(GTEST_DIR "${CMAKE_SOURCE_DIR}/../../gtest/googletest")
+MESSAGE("GTEST_DIR " ${GTEST_DIR})
+
+include_directories("${GTEST_DIR}/include")
+
+find_library(GTEST_LIB NAMES gtest libgtest HINTS "${GTEST_DIR}/build" "${GTEST_DIR}/build/Debug")
+find_library(GTESTMAIN_LIB NAMES gtest_main libgtest_main HINTS "${GTEST_DIR}/build" "${GTEST_DIR}/build/Debug")
+
+MESSAGE("GTEST_LIB = ${GTEST_LIB}")
+MESSAGE("GTESTMAIN_LIB = ${GTESTMAIN_LIB}")
+
+#gpu test sources
+set(GPU_SOURCES gpu/gpuNUFFT_conv_tests.cpp
+				gpu/gpuNUFFT_2D_conv_tests.cpp
+				gpu/gpuNUFFT_fft_tests.cpp
+				gpu/gpuNUFFT_deapo_tests.cpp
+				gpu/forward/gpuNUFFT_forward_tests.cpp)
+
+#add CPU Tests
+add_subdirectory(cpu)
+
+cuda_add_executable(runGPUATMUnitTests ${GPU_SOURCES} ${GPUNUFFT_SOURCES} ../inc/gpuNUFFT_utils.hpp ../inc/gpuNUFFT_operator_factory.hpp ../inc/gpuNUFFT_operator.hpp ../inc/gpuNUFFT_kernels.hpp)
+if(WIN32)
+	target_link_libraries(runGPUATMUnitTests ${GRID_LIB_ATM_NAME} ${GTEST_LIB} ${GTESTMAIN_LIB})
+else(WIN32)
+	target_link_libraries(runGPUATMUnitTests pthread ${GRID_LIB_ATM_NAME} ${GTEST_LIB} ${GTESTMAIN_LIB})
+endif(WIN32)

gpunufft-0.8.0/CUDA/test/cpu/CMakeLists.txt

@@ -0,0 +1,16 @@
+#setup testframework
+MESSAGE("------ generating cpu tests -------")
+
+#cpu test sources
+set(CPU_SOURCES gpuNUFFT_tests.cpp 
+				gpuNUFFT_kernel_tests.cpp
+				gpuNUFFT_precomputation_tests.cpp
+				gpuNUFFT_operator_factory_tests.cpp
+				../../src/gpuNUFFT_utils.cpp 
+				../../src/cpu/gpuNUFFT_cpu.cpp)
+
+include_directories(${CUDA_INCLUDE_DIRS})
+#add source dir
+add_executable(runUnitTests ${CPU_SOURCES} ${GPUNUFFT_SOURCES} ../../inc/gpuNUFFT_cpu.hpp ../../inc/gpuNUFFT_utils.hpp ../../inc/gpuNUFFT_operator_factory.hpp ../../inc/gpuNUFFT_operator.hpp ../../inc/gpuNUFFT_kernels.hpp)
+target_link_libraries(runUnitTests ${GRID_LIB_ATM_NAME} ${GTEST_LIB} ${GTESTMAIN_LIB})
+set_target_properties(runUnitTests PROPERTIES LINK_FLAGS -lpthread)

gpunufft-0.8.0/CUDA/test/cpu/gpuNUFFT_kernel_tests.cpp

@@ -0,0 +1,27 @@
+#include <limits.h>
+#include "gpuNUFFT_cpu.hpp"
+
+#include "gtest/gtest.h"
+
+#define epsilon 0.0001f
+
+TEST(TestKernel, LoadKernel) {
+	if (DEBUG)
+		printf("start creating kernel...\n");
+	long kernel_entries = calculateGrid3KernelSize();
+	
+	assert(kernel_entries > 0);
+
+	DType *kern = (DType*) calloc(kernel_entries,sizeof(DType));
+	if (kern != NULL)
+	{
+		loadGrid3Kernel(kern,kernel_entries);
+		EXPECT_EQ(1.0f,kern[0]);
+		EXPECT_LT(0.9940f-kern[1],epsilon);
+		EXPECT_LT(0.0621f-kern[401],epsilon);
+		EXPECT_LT(0.0041f-kern[665],epsilon);
+		EXPECT_EQ(0.0f,kern[kernel_entries-1]);
+		free(kern);
+	}
+	EXPECT_EQ(1, 1);
+}