Use the correct crt*.o files when linking musl targets.
This is supposed to support optionally using the system copy of musl
libc instead of the included one if supported. This currently only
affects the start files, which is enough to allow building rustc on musl
targets.
Most of the changes are analogous to crt-static.
Excluding the start files is something musl based distributions usually patch into their copy of rustc:
- eb064c8/community/rust/musl-fix-linux_musl_base.patch
- 77400fc/srcpkgs/rust/patches/link-musl-dynamically.patch
For third-party distributions that not yet carry those patches it would be nice if it was supported without the need to patch upstream sources.
## Reasons
### What breaks?
Some start files were missed when originally writing the logic to swap in musl start files (gcc comes with its own start files, which are suppressed by -nostdlib, but not manually included later on). This caused #36710, which also affects rustc with the internal llvm copy or any other system libraries that need crtbegin/crtend.
### How is it fixed?
The system linker already has all the logic to decide which start files to include, so we can just defer to it (except of course if it doesn't target musl).
### Why is it optional?
In #40113 it was first tried to remove the start files, which broke compiling musl-targeting static binaries with a glibc-targeting compiler. This is why it eventually landed without removing the start files. Being an option side-steps the issue.
### Why are the start files still installed?
This has the nice side-effect, that the produced rust-std-* binaries can still be used by on a glibc-targeting system with a rustc built against glibc.
## Does it work?
With the following build script (using [musl-cross-make](https://github.com/richfelker/musl-cross-make)): https://shadowice.org/~mixi/rust-musl/build.sh, I was able to cross-compile a musl-host musl-targeting rustc on a glibc-based system. The resulting binaries are at https://shadowice.org/~mixi/rust-musl/binaries/. This also requires #50103 and #50104 (which are also applied to the branch the build script uses).
Add some groundwork for cross-language LTO.
Implements part of #49879:
- Adds a `-Z cross-lang-lto` flag to rustc
- Makes sure that bitcode is embedded in object files if the flag is set.
This should already allow for using cross language LTO for staticlibs (where one has to invoke the linker manually anyway). However, `rustc` will not try to enable LTO for its own linker invocations yet.
r? @alexcrichton
Pass a test directory to rustfmt
Another attempt to fix the rustfmt tests. `RUSTFMT_TEST_DIR` is consumed by Rustfmt in the latext commit (thus the Rustfmt update) because we need a place to create temp files that won't be read-only.
r? @alexcrichton
rustbuild: Allow quick testing of libstd and libcore at stage0
This PR implemented two features:
1. Added a `--no-doc` flag to allow testing a crate *without* doc tests. In this mode, we don't need to build rustdoc, and thus we can skip building the stage2 compiler. (Ideally stage0 test should use the bootstrap rustdoc, but I don't want to mess up the core builder logic here)
2. Moved all libcore tests externally and added a tidy test to ensure we don't accidentally add `#[test]` into libcore.
After this PR, one could run `./x.py test --stage 0 --no-doc src/libstd` to test `libstd` without building the compiler, thus enables us to quickly test new library features.
Misc tweaks
This:
- ~~Add explicit dependencies on `getops`~~
- Fixes the libtest-json test when `RUST_BACKTRACE=1` is set
- ~~Sets `opt-level` to `3`~~
- Removes the use of `staged_api` from `rustc_plugin`
- ~~Enables the Windows Error Reporting dialog when running rustc during bootstrapping~~
- Disables Windows Error Reporting dialog when running compiletest tests
- Enables backtraces when running rustc during bootstrapping
- ~~Removes the `librustc` dependency on `libtest`~~
- Triggers JIT debugging on Windows if rustc panics during bootstrapping
r? @alexcrichton
Add armv5te-unknown-linux-musl target
This PR adds the armv5te-unknown-linux-musl target. The following steps should let you produce a fully statically linked binary now:
1. Running `./src/ci/docker/run.sh dist-armv5te-linux-musl`
2. Changing the run.sh script to start bash instead of the build process and running the container
3.
```sh
export USER=root
export PATH=/checkout/obj/build/x86_64-unknown-linux-gnu/stage2/bin:/checkout/obj/build/x86_64-unknown-linux-gnu/stage0/bin:$PATH
```
4. Configuring Cargo
```yaml
[target.armv5te-unknown-linux-musl]
linker = "arm-linux-gnueabi-gcc"
```
5. Building a project
```sh
cargo new --bin hello
cd hello
cargo build --target=armv5te-unknown-linux-musl --release
```
This commit adds a dedicated mode to compiletest for running rustfix tests,
adding a new `src/test/rustfix` directory which will execute all tests as a
"rustfix" test, namely requiring that a `*.fixed` is next to the main file which
is the result of the rustfix project's application of fixes.
The `rustfix` crate is pulled in to actually perform the fixing, and the rustfix
compiletest mode will assert a few properties about the fixing:
* The expected fixed output must be the same as rustc's output suggestions
applied to the original code.
* The fixed code must compile successfully
* The fixed code must have no further diagnostics emitted about it
This is the first small step towards testing auto-fixable compiler
suggestions using compiletest. Currently, it only checks if next to a
UI test there also happens to a `*.rs.fixed` file, and then uses rustfix
(added as external crate) on the original file, and asserts that it
produces the fixed version.
To show that this works, I've included one such test. I picked this test
case at random (and because it was simple) -- It is not relevant to the
2018 edition. Indeed, in the near future, we want to be able to restrict
rustfix to edition-lints, so this test cast might go away soon.
In case you still think this is somewhat feature-complete, here's a
quick list of things currently missing that I want to add before telling
people they can use this:
- [ ] Make this an actual compiletest mode, with `test [fix] …` output
and everything
- [ ] Assert that fixed files still compile
- [ ] Assert that fixed files produce no (or a known set of) diagnostics
output
- [ ] Update `update-references.sh` to support rustfix
- [ ] Use a published version of rustfix (i.e.: publish a new version
rustfix that exposes a useful API for this)
Immutably and implicitly borrow all pattern ids for their guards (NLL only)
This is an important piece of rust-lang/rust#27282.
It applies only to NLL mode. It is a change to MIR codegen that is currently toggled on only when NLL is turned on. It thus affect MIR-borrowck but not the earlier static analyses (such as the type checker).
This change makes it so that any pattern bindings of type T for a match arm will map to a `&T` within the context of the guard expression for that arm, but will continue to map to a `T` in the context of the arm body.
To avoid surfacing this type distinction in the user source code (which would be a severe change to the language and would also require far more revision to the compiler internals), any occurrence of such an identifier in the guard expression will automatically get a deref op applied to it.
So an input like:
```rust
let place = (1, Foo::new());
match place {
(1, foo) if inspect(foo) => feed(foo),
...
}
```
will be treated as if it were really something like:
```rust
let place = (1, Foo::new());
match place {
(1, Foo { .. }) if { let tmp1 = &place.1; inspect(*tmp1) }
=> { let tmp2 = place.1; feed(tmp2) },
...
}
```
And an input like:
```rust
let place = (2, Foo::new());
match place {
(2, ref mut foo) if inspect(foo) => feed(foo),
...
}
```
will be treated as if it were really something like:
```rust
let place = (2, Foo::new());
match place {
(2, Foo { .. }) if { let tmp1 = & &mut place.1; inspect(*tmp1) }
=> { let tmp2 = &mut place.1; feed(tmp2) },
...
}
```
In short, any pattern binding will always look like *some* kind of `&T` within the guard at least in terms of how the MIR-borrowck views it, and this will ensure that guard expressions cannot mutate their the match inputs via such bindings. (It also ensures that guard expressions can at most *copy* values from such bindings; non-Copy things cannot be moved via these pattern bindings in guard expressions, since one cannot move out of a `&T`.)
