d14ba81572
Data safety > valgrind.
38 lines
1.2 KiB
Nix
38 lines
1.2 KiB
Nix
{ stdenv, fetchurl, valgrind }:
|
|
|
|
stdenv.mkDerivation rec {
|
|
# The r127 source still calls itself r126 everywhere, but I'm not going to
|
|
# patch over such silly cosmetic oversights in an official release. -- nckx
|
|
version = "127";
|
|
name = "lz4-${version}";
|
|
|
|
src = fetchurl {
|
|
url = "https://github.com/Cyan4973/lz4/archive/r${version}.tar.gz";
|
|
sha256 = "0hvbbr07j4hfix4dn4xw4fsmkr5s02bj596fn0i15d1i49xby2aj";
|
|
};
|
|
|
|
# valgrind is required only by `make test`
|
|
buildInputs = [ valgrind ];
|
|
|
|
enableParallelBuilding = true;
|
|
|
|
makeFlags = "PREFIX=$(out)";
|
|
|
|
doCheck = true;
|
|
checkTarget = "test";
|
|
|
|
meta = with stdenv.lib; {
|
|
description = "Extremely fast compression algorithm";
|
|
longDescription = ''
|
|
Very fast lossless compression algorithm, providing compression speed
|
|
at 400 MB/s per core, with near-linear scalability for multi-threaded
|
|
applications. It also features an extremely fast decoder, with speed in
|
|
multiple GB/s per core, typically reaching RAM speed limits on
|
|
multi-core systems.
|
|
'';
|
|
homepage = https://code.google.com/p/lz4/;
|
|
license = with licenses; [ bsd2 gpl2Plus ];
|
|
platforms = with platforms; linux;
|
|
maintainers = with maintainers; [ nckx ];
|
|
};
|
|
}
|