c4e0bff9d7
No hardware support for 80-bit floating point (long double) on aarch64-darwin; patch upstreamed
56 lines
1.3 KiB
Nix
56 lines
1.3 KiB
Nix
{ lib, stdenv
|
|
, fetchFromGitHub
|
|
, fetchpatch
|
|
, pkg-config
|
|
, gettext
|
|
, autoreconfHook
|
|
, gmp
|
|
, mpfr
|
|
}:
|
|
|
|
stdenv.mkDerivation rec {
|
|
pname = "fplll";
|
|
version = "5.4.5";
|
|
|
|
src = fetchFromGitHub {
|
|
owner = "fplll";
|
|
repo = "fplll";
|
|
rev = version;
|
|
sha256 = "sha256-taSS7jpVyjVfNe6kSuUDXMD2PgKmtG64V5MjZyQzorI=";
|
|
};
|
|
|
|
patches = [
|
|
(fetchpatch {
|
|
url = "https://github.com/fplll/fplll/commit/317cf70893eebfb2625da12e5377189908c36240.diff";
|
|
sha256 = "sha256-GbYSolBgv/he4QzjuRFdg93wHJABVHvA9x3PjpJTSRE=";
|
|
})
|
|
];
|
|
|
|
nativeBuildInputs = [
|
|
pkg-config
|
|
gettext
|
|
autoreconfHook
|
|
];
|
|
|
|
buildInputs = [
|
|
gmp
|
|
mpfr
|
|
];
|
|
|
|
meta = with lib; {
|
|
description = "Lattice algorithms using floating-point arithmetic";
|
|
changelog = [
|
|
# Some release notes are added to the github tags, though they are not
|
|
# always complete.
|
|
"https://github.com/fplll/fplll/releases/tag/${version}"
|
|
# Releases are announced on this mailing list. Unfortunately it is not
|
|
# possible to generate a direct link to the most recent announcement, but
|
|
# this search should find it.
|
|
"https://groups.google.com/forum/#!searchin/fplll-devel/FPLLL$20${version}"
|
|
];
|
|
license = licenses.lgpl21Plus;
|
|
maintainers = teams.sage.members;
|
|
platforms = platforms.unix;
|
|
};
|
|
}
|