These attributes were deprecated in #230951 because they confused things
with splicing, e.g. passing `rustPlatform.rust.rustc` to
`nativeBuildInputs` would get the wrong derivation when cross-compiling.
This was a consequence of the fact that they come from the buildPackages
set. But we still want to expose these packages as otherwise Rust
packages have no way to reference the rustc and cargo that build them,
which means they can't e.g. use `disallowedReferences` to ensure they
don't end up in the runtime closure.
Just as `stdenv.cc` itself isn't spliced, we can solve this problem by
unsplicing the packages.
This fixes a long standing issue where rust hooks behave differently when used inside buildRustPackage vs inside mkDerivation, which lead to surprising behavior, like for example the package being built for the wrong paltform or the linker not being found especially in cross compilation scenarios.
The reason for this inconsitency was, that buildRustPackage consumed the hooks in a non-spliced form, via [this inherit statement](4506ece030/pkgs/development/compilers/rust/make-rust-platform.nix (L60)), and therefore the usual platform shift on the hooks introduced by putting them in `nativeBuildInputs` was not applied here.
Thoug whenever the hook was used inside other builders like `mkDerivation` the platform shift did apply correctly as the hook was consumed via the spliced package set, introducing the inconsitecy.
Because of the wrong (non-spliced) use in buildRustPackage, most rust hooks have been designed with the wrong build/host/target shift in mind which is fixed by this change.
Due to the inconsitent behavior between different builders, workarounds like `rust.envVars`, which were previously introduced, likely become obsolete by this change.
This likely fixes a bunch of cross compilation issues for rust packages that are not based on `buildRustPackage` but instead consume the hooks directly.
Done:
- ensure that `buildRustPackage` consumes spliced hooks by using makeScopeWithSplicing' in make-rust-platform.nix.
- refactor hooks to make them refer to correct build/host/target packages.
- remove `rust.envVars` workaround from all rust hooks
- implement tests for most rust hooks in /pkgs/test/rut-hooks
The newly added tests can be executed for native as well as cross compilation via:
```
nix-build -A tests.rust-hooks -A pkgsCross.riscv64.tests.rust-hooks
```
After final improvements to the official formatter implementation,
this commit now performs the first treewide reformat of Nix files using it.
This is part of the implementation of RFC 166.
Only "inactive" files are reformatted, meaning only files that
aren't being touched by any PR with activity in the past 2 months.
This is to avoid conflicts for PRs that might soon be merged.
Later we can do a full treewide reformat to get the rest,
which should not cause as many conflicts.
A CI check has already been running for some time to ensure that new and
already-formatted files are formatted, so the files being reformatted here
should also stay formatted.
This commit was automatically created and can be verified using
nix-build a08b3a4d19.tar.gz \
--argstr baseRev b32a094368
result/bin/apply-formatting $NIXPKGS_PATH
Previously, trying to use `.override {}` on one of the hooks,
specifically `bindgenHook` would yield in an error.
By replacing `callPackage` with `callPackages`, this error is fixed and
the overrides are propagated to the hooks.
Co-Authored-By: Atemu <atemu.main+nixpkgs@gmail.com>
It is almost never correct to use these attributes, because they don't
work correctly with splicing. Compare:
% nix repl -f . --argstr crossSystem aarch64-linux
Welcome to Nix 2.10.3. Type :? for help.
Loading installable ''...
Added 18988 variables.
nix-repl> callPackage ({ stdenv, rustc }: (stdenv.mkDerivation { name = ""; nativeBuildInputs = [ rustc ]; }).nativeBuildInputs) {}
«derivation /nix/store/bjrkg8kcq3hvg5kb03ivb856zy91qpbk-aarch64-unknown-linux-gnu-rustc-1.69.0.drv» ]
nix-repl> callPackage ({ stdenv, rustPlatform }: (stdenv.mkDerivation { name = ""; nativeBuildInputs = [ rustPlatform.rust.rustc ]; }).nativeBuildInputs) {}
«derivation /nix/store/ra5r07j52y7akclr827r3dzxzvqnvfbl-rustc-1.69.0.drv» ]
I'm not sure this is fixable. I don't think it's worth keeping them
around considering we have top level attributes. It makes overriding
slightly more annoying, but only slightly.
The main purpose of `makeRustPlatform` is to enable users to override
the `rustc` and `cargo` versions used by the `rustPlatform` derivations.
In all attributes of the result of `makeRustPlatform`, `rustc` and/or
`cargo` are overriden, except in `importCargoLock`. I think this is an
oversight / bug, and passing the received cargo derivation is the right
behaviour.
If `importCargoLock` always using the global cargo package even in
`makeRustPlatform` is the intended behaviour, I think it should be
documented at least in a comment.
- `toRustTarget` and friends pulled out from rust tools into rust
library. Since they don't depend on any packages they can be more
widely useable.
- `build-rust-package` gets its own directory
- `fetch-cargo-tarball` gets its own directory
This change introduces the cargoLock argument to buildRustPackage,
which can be used in place of cargo{Sha256,Hash} or cargoVendorDir. It
uses the importCargoLock function to build the vendor
directory. Differences compared to cargo{Sha256,Hash}:
- Requires a Cargo.lock file.
- Does not require a Cargo hash.
- Retrieves all dependencies as fixed-output derivations.
This makes buildRustPackage much easier to use as part of a Rust
project, since it does not require updating cargo{Sha256,Hash} for
every change to the lock file.
This function can be used to create an output path that is a cargo
vendor directory. In contrast to e.g. fetchCargoTarball all the
dependent crates are fetched using fixed-output derivations. The
hashes for the fixed-output derivations are gathered from the
Cargo.lock file.
Usage is very simple, e.g.:
importCargoLock {
lockFile = ./Cargo.lock;
}
would use the lockfile from the current directory.
The implementation of this function is based on Eelco Dolstra's
import-cargo:
https://github.com/edolstra/import-cargo/blob/master/flake.nix
Compared to upstream:
- We use fetchgit in place of builtins.fetchGit.
- Sync to current cargo vendoring.
API change:
`cargoParallelTestThreads` suggests that this attribute sets the
number of threads used during tests, while it is actually a boolean
option (use 1 thread or NIX_BUILD_CORES threads). In the hook, this
is replaced by a more canonical name `dontUseCargoParallelTests`.
- API change: remove the `target` argument of `buildRustPackage`, the
target should always be in sync with the C/C++ compiler that is used.
- Gathering of binaries has moved from `buildPhase` to `installPhase`,
this simplifies the hook and orders this functionality logically
with the installation logic.
We expose it on the top level, but I don't think it makes sense to pull
it from a specific version of the rust tools when it is in fact version
agnostic.
This reverts a tiny portion of 912dca193a.
