fsl-base

This project contains the "base" of an FSL installation. It includes:

• Configuration scripts for setting up an environment to use FSL
• Makefile infrastructure used to compile and install FSL projects
• Various other small configuration, data, and auxillary files

Prior to FSL 6.0.6, the contents of this project were provided by the etc and config projects.

Tests

The .ci directory contains a collection of tests which validate the behaviour of various aspects of the base project. All tests are written in Python, and are executed with pytest.

Standard conventions for C/C++/CUDA projects.

Commands for compilation of intermediate object files should have the form:

$(CC)   $(CFLAGS)       <input/output files> # for .c files
$(CXX)  $(CXXFLAGS)     <input/output files> # for .cc files
$(CXX)  $(CUDACXXFLAGS) <input/output files> # for .cc files which are to
                                             # be linked into CUDA
                                             # libraries/executables
$(NVCC) $(NVCCFLAGS)    <input/output files> # for .cu files

And commands for compilation and linking of executables and libraries should have the form:

$(CC)   $(CFLAGS)            <input/output files> ${LDFLAGS}     # for c libs/exes
$(CXX)  $(CXXFLAGS)          <input/output files> ${LDFLAGS}     # for c++ exes
$(CXX)  $(CXXFLAGS)  -shared <input/output files> ${LDFLAGS}     # for c++ libs
$(NVCC) $(NVCCFLAGS) -shared <input/output files> ${NVCCLDFLAGS} # for CUDA libs
$(NVCC) $(NVCCFLAGS)         <input/output files> ${NVCCLDFLAGS} # for CUDA exes

LDFLAGS must come at the end, to ensure proper linking.

An example C++ Makefile might look like this:

include ${FSLCONFDIR}/default.mk

PROJNAME = myproject
SOFILES  = libmylib.so
XFILES   = myexe1 myexe2

# Instead of explicitly defining this rule,
# We could instead rely on implicit Make
# rules for generating *.o files from *.cc
# files.
%.o: %.cc
    $(CXX) $(CXXFLAGS) -c -o $@ $<

libmylib.so: myobj1.o myobj2.o
    $(CXX) $(CXXFLAGS) -shared -o $@ $^ $(LDFLAGS)

myexe1: myexe1.o myobj1.o
    $(CXX) $(CXXFLAGS) -o $@ $^ $(LDFLAGS)

myexe2: myexe2.o myobj2.o
    $(CXX) $(CXXFLAGS) -o $@ $^ $(LDFLAGS)

Building FSL CUDA projects.

Some FSL projects (e.g. [fsl/eddy(https://git.fmrib.ox.ac.uk/fsl/eddy) use CUDA for GPU acceleration. In order to compile these projects, the nvcc compiler must be available on your $PATH, or a variable called $NVCC must refer to the nvcc executable to use.

The following additional options may be used to control compilation of a CUDA project:

• CUDA_STATIC: Statically link against the CUDA runtime and certain; CUDA Toolkit libraris (default is to use dynamic linking).
• GENCODEFLAGS: -gencode options specifying the device architectures to compile device code for (default is to use the config/supportedGencodes.sh script).

For example, to compile a CUDA project with a specific CUDA installation, and using static linking:

export PATH=/usr/local/cuda-10.2/bin:$PATH
make CUDA_STATIC=1

The Makefile for each FSL CUDA project may provide additional options for controlling compilation.

An example Makefile for a C++/CUDA project might look like this:

include ${FSLCONFDIR}/default.mk

PROJNAME = myproject
XFILES   = myexe

%.o: %.cu
    $(NVCC) $(NVCCFLAGS) -c -o $@ $<

%.o: %.cc
    $(CXX) $(CUDACXXFLAGS) -c -o $@ $<

myexe:
    $(NVCC) $(NVCCFLAGS) -o $@ $< $(NVCCLDFLAGS)