…tting.

Consider the situation where we have two variants of the same crate:
```
rust_library(
    name = "foo",
    srcs = ["foo.rs"],
)

rust_library(
    name = "foo2",
    crate_name = "foo",
    srcs = ["foo.rs"],
)
```

A use case for this is for example when we want to define different variants
of a crate using different feature sets, e.g., define a rustc_private non-std
variant of a vendored crate of the standard library together with a more
standard version of the vendored crate that requires the standard library.

Currently, this is a fragile definition since the output rlib filename of both
of these is the same, e.g., `libfoo-$HASH.rlib`, where `$HASH` is
derived from the crate root source filename, in this case `foo.rs`. So whenever
we have a Bazel query that includes both of these, the build may fail with
errors like:
```
ERROR: file 'test/unit/crate_variants/libfoo-717083168.rlib' is generated by these conflicting actions:
Label: //test/unit/crate_variants:foo2, //test/unit/crate_variants:foo
```

This patch reduces the risk of such output filename collisions by mixing in
the rule label in the output hash calculation.

)

…1073)

When building on an Apple Silicon machine, a different target triple
is required to target the simulator:
https://doc.rust-lang.org/rustc/platform-support/aarch64-apple-ios-sim.html

As rules_rust currently ignores abi in triples, listing device
and simulator triples in a rust_repository_set() ends up always using
the first one. Adding this constraint allows a build to be switched
between targeting a device and targeting the simulator.

…build#1088)

…azelbuild#1089)

* Fix proc_macro_dylib_path in cases where they are built in both opt and debug

* Resolve rustfmt issues

* Update test comment

* optimization: switch fastbuild opt default to 1 from 0

This will very slightly increase build times but applies basic
optimizations that will do things like make iterators _much_
faster. Since fastbuild is how people iterate most of the time, this
seems like a sensible default value.

* Regenerate documentation

Co-authored-by: Augie Fackler <augie@google.com>

…build#1091)" (bazelbuild#1094)

This reverts commit 2048a65.

…azelbuild#1095)

This variable is set for the default profiles as described in https://doc.rust-lang.org/cargo/reference/profiles.html -- true for dev and false for release.

This is an environment variable cargo sets for build scripts documented here https://doc.rust-lang.org/cargo/reference/environment-variables.html#environment-variables-cargo-sets-for-build-scripts

* Added Rust 1.58.1

* Regenerate documentation

* Updated rust_test docs

* Regenerate documentation

)

…build#1102)

…bazelbuild#1101)

* Allow users to configure the timeout of `cargo_bootstrap_repository`.

* Regenerate documentation

…azelbuild#1104)

* Made components of `rustc` bundles mandatory for `rust_toolchain`.

* Regenerate documentation

bazelbuild#1100)

* Allow cargo_bootstrap_repository to specify rustc and cargo separately

* Regenerate documentation

…build#1106)

…azelbuild#1107)

…bazelbuild#1105)

* Fixed typo

* Regenerate documentation

…azelbuild#1109)

* Replace lists of files and Targets in `rust_toolchain` with depsets

* Use `find_sysroot` helper.

* Simply use `rust_std` files directly

* Revert "Use `find_sysroot` helper."

This reverts commit 914d159.

* files

Works around bazelbuild/bazel#14681

…1115)

* Calculate more things in the toolchain construction

* Update sysroot logic

…d#1124)

Closes bazelbuild#1051

We use the equivalent `toolchain._cc_toolchain` attribute instead

…ild#1129)