{ fetchurl , stdenv , lib , gfortran , perl , llvmPackages , precision ? "double" , enableAvx ? stdenv.hostPlatform.avxSupport , enableAvx2 ? stdenv.hostPlatform.avx2Support , enableAvx512 ? stdenv.hostPlatform.avx512Support , enableFma ? stdenv.hostPlatform.fmaSupport , enableMpi ? false , mpi , withDoc ? stdenv.cc.isGNU }: assert lib.elem precision [ "single" "double" "long-double" "quad-precision" ]; stdenv.mkDerivation rec { pname = "fftw-${precision}"; version = "3.3.10"; src = fetchurl { urls = [ "https://fftw.org/fftw-${version}.tar.gz" "ftp://ftp.fftw.org/pub/fftw/fftw-${version}.tar.gz" ]; sha256 = "sha256-VskyVJhSzdz6/as4ILAgDHdCZ1vpIXnlnmIVs0DiZGc="; }; outputs = [ "out" "dev" "man" ] ++ lib.optional withDoc "info"; # it's dev-doc only outputBin = "dev"; # fftw-wisdom nativeBuildInputs = [ gfortran ]; buildInputs = lib.optionals stdenv.cc.isClang [ # TODO: This may mismatch the LLVM version sin the stdenv, see #79818. llvmPackages.openmp ] ++ lib.optional enableMpi mpi; configureFlags = [ "--enable-shared" "--enable-threads" ] ++ lib.optional (precision != "double") "--enable-${precision}" # all x86_64 have sse2 # however, not all float sizes fit ++ lib.optional (stdenv.isx86_64 && (precision == "single" || precision == "double") ) "--enable-sse2" ++ lib.optional enableAvx "--enable-avx" ++ lib.optional enableAvx2 "--enable-avx2" ++ lib.optional enableAvx512 "--enable-avx512" ++ lib.optional enableFma "--enable-fma" ++ [ "--enable-openmp" ] ++ lib.optional enableMpi "--enable-mpi" # doc generation causes Fortran wrapper generation which hard-codes gcc ++ lib.optional (!withDoc) "--disable-doc"; enableParallelBuilding = true; nativeCheckInputs = [ perl ]; meta = with lib; { description = "Fastest Fourier Transform in the West library"; homepage = "http://www.fftw.org/"; license = licenses.gpl2Plus; maintainers = [ maintainers.spwhitt ]; platforms = platforms.unix; }; }