This makes 'nix flake' less cluttered and more consistent (it's only
subcommands that operator on a flake). Also, the registry is not
inherently flake-related (e.g. fetchTree could also use it to remap
inputs).
Most functions now take a StorePath argument rather than a Path (which
is just an alias for std::string). The StorePath constructor ensures
that the path is syntactically correct (i.e. it looks like
<store-dir>/<base32-hash>-<name>). Similarly, functions like
buildPaths() now take a StorePathWithOutputs, rather than abusing Path
by adding a '!<outputs>' suffix.
Note that the StorePath type is implemented in Rust. This involves
some hackery to allow Rust values to be used directly in C++, via a
helper type whose destructor calls the Rust type's drop()
function. The main issue is the dynamic nature of C++ move semantics:
after we have moved a Rust value, we should not call the drop function
on the original value. So when we move a value, we set the original
value to bitwise zero, and the destructor only calls drop() if the
value is not bitwise zero. This should be sufficient for most types.
Also lots of minor cleanups to the C++ API to make it more modern
(e.g. using std::optional and std::string_view in some places).
A command like
$ nix run nixpkgs#hello
will now build the attribute 'packages.${system}.hello' rather than
'packages.hello'. Note that this does mean that the flake needs to
export an attribute for every system type it supports, and you can't
build on unsupported systems. So 'packages' typically looks like this:
packages = nixpkgs.lib.genAttrs ["x86_64-linux" "i686-linux"] (system: {
hello = ...;
});
The 'checks', 'defaultPackage', 'devShell', 'apps' and 'defaultApp'
outputs similarly are now attrsets that map system types to
derivations/apps. 'nix flake check' checks that the derivations for
all platforms evaluate correctly, but only builds the derivations in
'checks.${system}'.
Fixes#2861. (That issue also talks about access to ~/.config/nixpkgs
and --arg, but I think it's reasonable to say that flakes shouldn't
support those.)
The alternative to attribute selection is to pass the system type as
an argument to the flake's 'outputs' function, e.g. 'outputs = { self,
nixpkgs, system }: ...'. However, that approach would be at odds with
hermetic evaluation and make it impossible to enumerate the packages
provided by a flake.
Some kernels disable "unpriveleged user namespaces". This is
unfortunate, but we can still use mount namespaces. Anyway, since each
builder has its own nixbld user, we already have most of the benefits
of user namespaces.
This is like 'nix run', except that the command to execute is defined
in a flake output, e.g.
defaultApp = {
type = "app";
program = "${packages.blender_2_80}/bin/blender";
};
Thus you can do
$ nix app blender-bin
to start Blender from the 'blender-bin' flake.
In the future, we can extend this with sandboxing. (For example we
would want to be able to specify that Blender should not have network
access by default and should only have access to certain paths in the
user's home directory.)
This reverts commit a0ef21262f. This
doesn't work in 'nix run' and nix-shell because setns() fails in
multithreaded programs, and Boehm GC mark threads are uncancellable.
Fixes#2646.
This is useful for testing commands in isolation.
For example,
$ nix run nixpkgs.geeqie -i -k DISPLAY -k XAUTHORITY -c geeqie
runs geeqie in an empty environment, except for $DISPLAY and
$XAUTHORITY.
E.g.
nix run nixpkgs.hello -c hello --greeting Hallo
Note that unlike "nix-shell --command", no quoting of arguments is
necessary.
"-c" (short for "--command") cannot be combined with "--" because they
both consume all remaining arguments. But since installables shouldn't
start with a dash, this is unlikely to cause problems.
Running "nix run" with a diverted store, e.g.
$ nix run --store local?root=/tmp/nix nixpkgs.hello
stopped working when Nix became multithreaded, because
unshare(CLONE_NEWUSER) doesn't work in multithreaded processes. The
obvious solution is to terminate all other threads first, but 1) there
is no way to terminate Boehm GC marker threads; and 2) it appears that
the kernel has a race where unshare(CLONE_NEWUSER) will still fail for
some indeterminate amount of time after joining other threads.
So instead, "nix run" will now exec() a single-threaded helper ("nix
__run_in_chroot") that performs the actual unshare()/chroot()/exec().
