This reverts commit b6fc00b8f4.
Rust 1.66.0 contains a fix for libiconv being linked unconditionally on macOS, but this only applies to packages that don't depend on older versions of `libc`.
For now, let's go back to including libiconv in `buildInputs` by default for packages that use `buildRustPackage`. As packages bump their `libc` versions, we can eventually stop including it by default, and manually add it where needed.
There are two problems: first that we end up splitting on spaces in the
loop. Even when that is fixed, we still would split on spaces in the
`export` inside the loop. We need to guard against both.
Fixes#199298
Confirmed that it fixes the case mentioned in the ticket:
```console
[nix-develop]$ $(nix-build -I nixpkgs=/home/shana/programming/nixpkgs Cargo.nix -A rootCrate.build --no-out-link)/bin/nix-rustc-env-escape-repro
Expecting three words, got: first second third
```
I think this is going to cause a rebuild of every Rust package even if
they were unaffected, not much we can do here.
This breaks the builder when a nix-shell or keepBuildTree is used. The
issue occurs because paths to cargo lockfiles are read with NIX_BUILD_TOP,
which is not reliable.
This breaks a nix-shell because NIX_BUILD_TOP simply is not set, causing
an invalid path to be used. This can be worked around using
NIX_BUILD_TOP=$PWD, but that obviously is not great.
This breaks keepBuildTree because it changes the working directory to a
different path than NIX_BUILD_TOP. Since the lockfiles are copied based
on the working directory, but read based on NIX_BUILD_TOP, this causes
the hook to not be able to find them.
This was solved by both reading these files based on the working directory,
using absolute paths to avoid having to traverse back in the directory tree.
Fixes: #138554
Rust binaries are unconditionally linked to libiconv on Darwin (see https://github.com/rust-lang/libc/issues/2870). We already add it as a dependency in `buildRustPackage`, so let's go a step further and propagate it.
`toTargetArch` in `pkgs/build-support/rust/lib/default.nix` is used to
set `CARGO_CFG_TARGET_ARCH`. This environment variable is supposed to
be the `<arch>` portion of an LLVM-style platform name:
```
<arch><sub>-<kernel>-<libc><abi>
```
Note that the pointer-width (the "64" in "x86_64" and "mips64") is
part of `<arch>`, but the endianness (the `_be` in `aarch64_be`) is
*not*.
Unfortunately at the moment nixpkgs' parsed `cpuType` has no way to
query for the three subparts (name, pointer-width, and
subarch/endianness), nor any way to ask for just the first two parts.
For now, this commit simply fixes the problem in the two cases that
matter: `mips64el` and `powerpc64le`, which I believe are the only two
platforms supported by both rust and nixpkgs which have a
"subarchitecture".
Tell rust if we want our binaries linked statically or dynamically.
Otherwise the compiler will always produce statically linked binaries for musl
targets, as this is the default.
We are replicating one mechanism behind `-Z build-std`.
There isn't yet crate2nix support for this, but one can (and I do) add
the missing stdlib deps (for this feature to pick up) with overrides.
With Rust 1.61, it is necessary to link to external static/dynamic libaries
when building the rlib that uses them, rather than when linking the final
binary. In fact, it is no longer necessary to specify the libraries to link
when building the final binary, but the library search path flags must still
be included.
The old logic flow had the structure
if ( … ) {
if ( … ) {
…
} else {
…
}
} else {
…
}
which is quite hard to follow in Nix. Instead we ensure that no if
expression is inside a then branch.
This change is zero rebuild, as no logic was changed.
This parameter is being set to `$NIX_BUILD_CORES` by default. This is a
standard practice but there's a suspicion that this can produce broken
builds. For some details see
https://github.com/cargo2nix/cargo2nix/issues/184 . As a
work-around/test, it'd be good if codegen-units can be set to something
constant, such as `1`. This PR allows it.
Note that the default of `$NIX_BUILD_CORES` is preserved so this MR
causes no change in default behaviour and no rebuilds.
These features are internal-only, have special characters that bash
doesn't support in variable names, and aren't normally given
environment variables by cargo as far as I can tell.
Some crates do not have a Cargo.toml at the top-level, but only in
nested directories. Before this change importCargoLock used to fail with:
error: manifest path `/nix/store/some-store-path/Cargo.toml` does not exist
- `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
In restricted mode (and therefore with flakes) `builtins.readFile` may not be the result of `builtins.toFile`,
making it impossible to use a generated lockFile (with or without IFD),
and thereby causing evaluation to fail if `system != builtins.currentSystem` on Hydra
so the jobs are not delegated to eligible build machines that support that system.
This is done in a way that avoids rebuilds.
I currently do not have much time to work on nixpkgs. Remove
myself as a maintainer from a bunch of packages to avoid that
people are waiting on me for a review.
near the end of 2019, the default Cargo.lock format was changed to
[[package]]
checksum = ...
This is what importCargoLock assumes. If the crate had not been `cargo
update`'d with a more recent toolchain than the one with the new
format as default, importCargoLock would fail when trying to access
pkg.checksum.
I ran into such a case (shamefully, in my own crate) and it took me a
while to figure out what was going on, so here is an assert with a
more user friendly message and a hint.
According to rustc implementation[1], `-C incremental=no` enables
incremental builds with directory name `no`. This patch removes the
`-C incremental` argument to disable incremental builds.
[1]: ee86f96ba1/compiler/rustc_session/src/options.rs (L918-L919)
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.
Also begin to start work on cross compilation, though that will have to
be finished later.
The patches are based on the first version of
https://reviews.llvm.org/D99484. It's very annoying to do the
back-porting but the review has uncovered nothing super major so I'm
fine sticking with what I've got.
Beyond making the outputs work, I also strove to re-sync the packages,
as they have been drifting pointlessly apart for some time.
----
Other misc notes, highly incomplete
- lvm-config-native and llvm-config are put in `dev` because they are
tools just for build time.
- Clang no longer has an lld dep. That was introduced in
db29857eb3, but if clang needs help
finding lld when it is used we should just pass it flags / put in the
resource dir. Providing it at build time increases critical path
length for no good reason.
----
A note on `nativeCC`:
`stdenv` takes tools from the previous stage, so:
1. `pkgsBuildBuild`: `(?1, x, x)`
2. `pkgsBuildBuild.stdenv.cc`: `(?0, ?1, x)`
while:
1. `pkgsBuildBuild`: `(?1, x, x)`
2. `pkgsBuildBuild.targetPackages`: `(x, x, ?2)`
3. `pkgsBuildBuild.targetPackages.stdenv.cc`: `(?1, x, x)`
The `checkType` argument of buildRustPackage was not used anymore
since the refactoring of `buildRustPackage` into hooks. This was
an oversight that is fixed by this change.
The check type can also be passed directly to cargoCheckHook using the
`cargoCheckType` environment variable.
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`.
The directory in the tarball of vendored dependencies contains `name`,
which is by default set to `${pname}-${version}`. This adds an
additional attribute to permit setting the name to something of the
user's choosing.
Since `cargoSha256`/`cargoHash` depend on the name of the directory of
vendored dependencies, `cargoDepsName` can be used to e.g. make the
hash invariant to the package version by setting `cargoDepsName =
pname`.
- 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.
`buildRustPackage` currently accepts `cargoSha256` as a hash for
vendored dependencies. This change adds `cargoHash` which accepts SRI
hashes, setting `outputHashAlgo` to `null`.
The hash mismatch message still uses `cargoSha256` as an example,
which it probably should until we completely switch to SRI hashes.
As @lopsided98 points out in #105305, since the hashes are now target
sensative, and until we find reason to actually care to test what they
are exactly, we are best just normalizing them away in the tests.
Bofore this change, NUM_JOBS was set to 1. Some crates for building
C/C++ code (e.g. the cc and cmake crates), rely on this variable to
set the number of jobs. As a consequence, we were compiling embedded
libraries serially. Change this to NIX_BUILD_CORES to permit parallel
builds.
Prior discussion:
https://github.com/NixOS/nixpkgs/pull/50452#issuecomment-439407547
This enables short argument attrsets similar to fetchPypi:
src = fetchCrate {
inherit pname version;
sha256 = "02h8pikmk19ziqw9jgxxf7kjhnb3792vz9is446p1xfvlh4mzmyx";
};
While the artifacts from `buildPhase` should be used for testing as
well, it should be avoided that those are modified during `checkPhase`.
This can happen if a package is built e.g. with special
`cargoBuildFlags` that don't apply to the `checkPhase`. In that case, a
binary would be installed into `$out` without those flags since
`checkPhase` overrides the binary in the `target`-directory.
This patch copies the state of `target/release` into a temporary
location at the end of the `buildPhase` and installs the results from
that temporary directory into `$out` while `checkPhase` can continue
using the configured build-dir.
cc #91689Closes#93119Closes#91191
When features were supplied in cargoBuildFlags, the binaries were built
with these features enabled. Unless checking was disabled, `cargo test`
was executed without these build flags, meaning the binaries were
rebuilt and overwritten without the specified features.
Fix this bug by running tests after the installation phase.
Cargo sets `CARGO_FEATURE_*` for all features when running a build
script:
https://doc.rust-lang.org/cargo/reference/environment-variables.html#environment-variables-cargo-sets-for-build-scripts
Some crates have build scripts (e.g. openblas-src) that rely on the
feature variables being properly set.
Since we now need several representations of features, this change
also updates `createFeatures` to be a list of features, rather than
`rustc` feature arguments. `configureCrate` and `buildCrate` then
build the required representations as-needed.
Fixes#68978
There are several tarballs (such as the `rust-lang/rust`-source) with a
`Cargo.toml` at root and several sub-packages (with their own Cargo.toml)
without using workspaces[1].
In such a case it's needed to move into a subdir to only build the
specified sub-package (e.g. `rustfmt` or `rsl`), however the artifacts
are at `/target` in the root-dir of the build environment. This breaks
the build since `buildRustPackage` searches for executables in `target`
(which is at the build-env's root) at the end of the `buildPhase`.
With the optional `buildAndTestSubdir`-argument, the builder moves into
the specified subdir using `pushd`/`popd` during `buildPhase` and
`checkPhase`.
Also moved the logic to find executables and libs to the end of the `buildPhase`
from a custom `postBuild`-hook to fix packages with custom `build`/`install`-procedures
such as `uutils-coreutils`.
[1] https://doc.rust-lang.org/book/ch14-03-cargo-workspaces.html
If a user provides `nativeBuildInputs = [ llvmPackages.bintools ]` or any other
package containing a `${prefix}/bin/diff`, the builder could use it instead
of the standard unix `diff`, causing a build failure.
This updates the call to specify an abspath to `diff` and avoid reliance on `PATH`.
Resolves#87081
When running `cargo test --release`, the artifacts from `buildPhase`
will be reused here. Previously, most of the stuff had to be recompiled
without optimizations.
I know, heretic, but...
I also know that this is not perfect but it is a good start, I think. It
would be nice if this were part of the automatic "nixdoc" function
reference. I'd like guidance if this should be part of the rust section
or something else.
As it turns out Darwin does most of the things differently then "normal"
systems. They are using a different shared library extension and require
an obscure commandline parameter that has to be added to every build
system out there. That issue seems to be with clang on Darwin as on
Linux that flag isn't required to build the very same tests (when using
clang).
After adjusting these two details the tests are running fine on the
darwin box that I was able to obtain.
According to the Cargo documentation:
> The build script does not have access to the dependencies listed in
> the dependencies or dev-dependencies section (they’re not built
> yet!). Also, build dependencies are not available to the package
> itself unless also explicitly added in the [dependencies] table.
https://doc.rust-lang.org/cargo/reference/build-scripts.html
This change separates linkage of regular dependencies and build
dependencies.
* Make errors include the crate name and make them much more prominent.
* Move more code into lib.sh
* Already source generated logging code and lib.sh in configure
The inlined readme that we were iterating on has been moved to GitHub
issue #79975, and the default is now the new cargo fetcher, so this
doc comment is out of date.
Previously, we would asssert that the lockfiles are consistent during the
unpackPhase, but if the pkg has a patch for the lockfile itself then we must
wait until the patchPhase is complete to check.
This also removes an implicity dependency on the src attribute coming from
`fetchzip` / `fetchFromGitHub`, which happens to name the source directory
"source". Now we glob for it, so different fetchers will work consistently.
Changes the default fetcher in the Rust Platform to be the newer
`fetchCargoTarball`, and changes every application using the current default to
instead opt out.
This commit does not change any hashes or cause any rebuilds. Once integrated,
we will start deleting the opt-outs and recomputing hashes.
See #79975 for details.
The readme was nice to discuss in the implementation PR, but now that this is
merged it's better to have an issue that can be linked against in PRs and
doesn't require further merges to update status.
Ported with a status update in #79975
By overriding each dependency on every level of the dependency tree we
are creating a lot of unnecessary instances of the same derivation
Looking at the output size of `nix-instantiate --trace-function-calls
-vvvv …` and the execution time I got about a 10x improvement after
applying this change.
It was probably good intentions that lead to these overrides but in
practice no tooling (that I know of) really needs this. `carnix` and
`crate2nix` are fine without those overrides. Furthermore I believe that
it is the job of the tooling around `buildRustCrate` to provide a
coherent set of overrides. By not enforcing all of the overrides, debug
flags, verbosity, … to be the same throughout the closure we also allow
consumers to override specific aspects of the crates. Some (older?)
crates might need different `crateOverrides` then newer crates with the
same name. Currently such situations can not (easily) be implemented
with the override in-place.
This has several advantages:
1. It takes up less space on disk in-between builds in the nix store.
2. It uses less space in the binary cache for vendor derivation packages.
3. It uses less network traffic downloading from the binary cache.
4. It plays nicely with hashed mirrors like tarballs.nixos.org, which only
substitute --flat hashes on single files (not recursive directory hashes).
5. It's consistent with how simple `fetchurl` src derivations work.
6. It provides a stronger abstraction between input src-package and output
package, e.g., it's harder to accidentally depend on the src derivation at
runtime by referencing something like `${src}/etc/index.html`. Likewise, in
the store it's harder to get confused with something that is just there as a
build-time dependency vs. a runtime dependency, since the build-time
src dependencies are tarred up.
Disadvantages are:
1. It takes slightly longer to untar at the start of a build.
As currently implemented, this attaches the compacted vendor.tar.gz feature as a
rider on `verifyCargoDeps`, since both of them are relatively newly implemented
behavior that change the `cargoSha256`.
If this PR is accepted, I will push forward the remaining rust packages with a
series of treewide PRs to update the `cargoSha256`s.
Previously I did use `runCommand` to do the same. Using
releaseTools.aggregate seems a lot saner and we might get nicer hydra
output of the tests that are failing.
It used to be the case (ref missing) that cargo did treat
`src/$libName.rs` as an alternative to `src/lib.rs` when the latter
wasn't present. Recently I failed to reproduce that with vanilla cargo
and it started to cause pain with some crates of the form:
some_crate/
`- src
`- main.rs
`- some_crate.rs
We would build `src/some_crate.rs` and thing it is a library while that
might not be the actual case. This crate is a valid `bin` crate not a
`lib` crate as far as I can tell from the samples I took.
I removed support for the previously required heuristic and commented
out the test cases in case we will need them again. We could crawl in
the Git history but chances are that the next person looking into this
doesn't know about the history.
When this fails, the user may want to copy-paste the path to the "bad"
Cargo.lock file to inspect. The trailing `.` on `$cargoDeps.` gets caught in
most terminal copy-pastes. Since half the lines already don't have it, this
removes it from all of them for consistent output.
This helps us instruct rustc to build tests instead of binaries. The
actual build will then ONLY produce test executables. This is a first
step towards having rust crate tests within nixpkgs.
We default back to only a single output in test cases since that is the
only reasonable thing to do here.
Producing libraries or binaries in addition to tests would theoretically
be feasible but usually generates different dependency trees. It is very
common to have some libraries in `[dev-depdendencies]` within Cargo.toml
just for your tests. To not start mixing things up going with a
dedicated derivation for the test build sounds like the best choice for
now.
To use this you must provide a proper test dependency chain to
`buildRustCrate` (as you would usually do with your non-test inputs).
And then set the `buildTests` attribute to `true`. The derivation will
then contain all tests that were built in `$out/tests`. All common test
patterns and directories should be supported and tested by this change.
Below is an example how you would run a single test from the derivation.
This commit contains some more examples in the `buildRustCrateTests`
attribute set that might be helpful.
```
let
drv = buildRustCrate {
…
buildTests true;
};
in runCommand "test-my-crate" {} ''
touch $out
exec ${drv}/tests/my-test
''
```
While unifying most of the lib function calls I accidentially changed
the filterSource functions as well. Since there were no tests I ended
up forgetting about this case (even thought I ran into it…).
Most stdenv wrappers already work like this -- it allows greater
customisation. We just have to be careful to remove arguments we're
using that shouldn't be passed to stdenv. I've been conservative
here, because fetchcargo checksums shouldn't change lightly.
The previous lines were only different in the kind of dependencies but
otherwise exactly the same. It makes the entire thing a bit more
readable by moving this into a function that takes care of this.
We can get rid of a bunch of workarounds that were in the build script
before by just passing on the `crateBin` attribute.
Before we converted the list of attributes to a string only to convert
it back in bash during the build phase. We can do the entire looping
through builds in Nix and thus need no conversion and parsing of
attributes over and over again.
The big part that still remains bash is the heuristic that cargo
introduced and that we can't do at eval time.
That code had been in the derivation for a while but no explanation was
given why that is needed. It might be helpful to our future selfs to
document why things are done the way they are.
The expression is already long and confusing enough without the color
stuff sprinkled in. Moving it to a dedicated file makes sense.
I switched a bit of the color support code to pure Nix since there
wasn't much point in doing that in bash while we can just do it in Nix.
We can just use `lib` instead of `builtins` in all cases but the
`hashString` case. Also changed a few lines to make use of some optional
helpers from lib.
This cuts down the dependency tree on some rust builds where a crate not
just exposes a binary but also a library. `$out/lib` contained a bunch
of extra support files that among other information carry linker flags
(including the full path to link-time dependencies). Worst case this led
to some binary outputs depending on the full build closure of rust
crates.
Moving all the `$out/lib` files to `$lib/lib` solves this nicely.
`lib` might be a bit weird here as they are most of the time just rlib
files (rust libraries). Those are essential only required during
compilation but they can also be shared objects (like with traditional
C-style packages). Which is why I went with `lib` for the new output.
One of the caveats we are running into here is that we do not (always)
know ahead of time of a crate produces just a library or just a binary.
Cargo allows for some ambiguity regarding whether or not a crate
provides one, two, … binaries and libraries as it's outputs. Ideally we
would be able to rely on the `crateType` entirely but so far that isn't
the case. More work on that area might show how difficult that actually
is.
One issue with cargoSha256 is that it's hard to detect when it needs to
be updated or not. It's possible to upgrade a package and forget to
update cargoSha256 and run with old versions of the program or
libraries.
This commit introduces `verifyCargoDeps` which, when enabled, will check
that the Cargo.lock is not out of date in the cargoDeps by comparing it
with the package source.
Quoting from the splitString docstring:
NOTE: this function is not performant and should never be used.
This replaces trivial uses of splitString for splitting version
strings with the (potentially builtin) splitVersion.
Before this change, buildRustCrate always called rustc with
--extern libName=[...]libName[...]
However, Cargo permits using a different name under which a dependency
is known to a crate. For example, rand 0.7.0 uses:
[dependencies]
getrandom_package = { version = "0.1.1", package = "getrandom", optional = true }
Which introduces the getrandom dependency such that it is known as
getrandom_package to the rand crate. In this case, the correct extern
flag is of the form
--extern getrandom_package=[...]getrandom[...]
which is currently not supported. In order to support such cases, this
change introduces a crateRenames argument to buildRustCrate. This
argument is an attribute set of dependencies that should be renamed. In
this case, crateRenames would be:
{
"getrandom" = "getrandom_package";
}
The extern options are then built such that if the libName occurs as
an attribute in this set, it value will be used as the local
name. Otherwise libName will be used as before.
