Nils had an excellent idea the other day: the same way that rustdoc is
able to load `rustc_driver` from the sysroot, ui-fulldeps tests should
also be able to load it from the sysroot. That allows us to run fulldeps
tests with stage1, without having to fully rebuild the compiler twice.
It does unfortunately have the downside that we're running the tests on
the *bootstrap* compiler, not the in-tree sources, but since most of the
fulldeps tests are for the *API* of the compiler, that seems ok.
I think it's possible to extend this to `run-make-fulldeps`, but I've
run out of energy for tonight.
- Move `plugin` tests into a subdirectory.
Plugins are loaded at runtime with `dlopen` and so require the ABI of
the running compile to match the ABI of the compiler linked with
`rustc_driver`. As a result they can't be supported in stage 1 and have
to use `// ignore-stage1`.
- Remove `ignore-stage1` from most non-plugin tests
- Ignore diagnostic tests in stage 1. Even though this requires a stage
2 build to load rustc_driver, it's primarily testing the error message
that the *running* compiler emits when the diagnostic struct is malformed.
- Pass `-Zforce-unstable-if-unmarked` in stage1, not just stage2. That
allows running `hash-stable-is-unstable` in stage1, since it now
suggests adding `rustc_private` to enable loading the crates.
- Add libLLVM.so to the stage0 target sysroot, to allow fulldeps tests
that act as custom drivers to load it at runtime.
- Pass `--sysroot stage0-sysroot` in compiletest so that we use the
correct version of std.
Do not attempt to commute comparison and cast to codegen discriminants
The general algorithm to compute a discriminant is:
```
relative_tag = tag - niche_start
is_niche = relative_tag <= (ule) relative_max
discr = if is_niche {
cast(relative_tag) + niche_variants.start()
} else {
untagged_variant
}
```
We have an optimization branch which attempts to merge the addition and the subtraction by commuting them with the cast. We currently get this optimization wrong.
This PR takes the easiest and safest way: remove the optimization, and let LLVM handle it. (Perf may not agree with that course of action 😅)
There may be a less invasive solution, but I don't have the necessary knowledge of LLVM semantics to find it. Cranelift has the same optimization, which should be handled similarly.
cc `@nikic` and `@bjorn3` if you have a better solution.
Fixes https://github.com/rust-lang/rust/issues/110128
Create "suggested tests" tool in `rustbuild`
Not the claimed person in #97339 but:
I've done a very rough implementation of this feature in-tree. I'm very new to `rustc` development (outside of docs) so some help would be greatly appreciated. The UI of this new subcommand obviously will change and I need some mentoring with the `--run` flag.
r? ```@jyn514```
resolve: Pre-compute non-reexport module children
Instead of repeating the same logic by walking HIR during metadata encoding.
The only difference is that we are no longer encoding `macro_rules` items, but we never currently need them as a part of this list. They can be encoded separately if this need ever arises.
`module_reexports` is also un-querified, because I don't see any reasons to make it a query, only overhead.
Rollup of 8 pull requests
Successful merges:
- #110153 (Fix typos in compiler)
- #110165 (rustdoc: use CSS `overscroll-behavior` instead of JavaScript)
- #110175 (Symbol cleanups)
- #110203 (Remove `..` from return type notation)
- #110205 (rustdoc: make settings radio and checks thicker, less contrast)
- #110222 (Improve the error message when forwarding a matched fragment to another macro)
- #110237 (Split out a separate feature gate for impl trait in associated types)
- #110241 (tidy: Issue an error when UI test limits are too high)
Failed merges:
- #110218 (Remove `ToRegionVid`)
r? `@ghost`
`@rustbot` modify labels: rollup
Instead of repeating the same logic by walking HIR during metadata encoding.
The only difference is that we are no longer encoding `macro_rules` items, but we never currently need them as a part of this list.
They can be encoded separately if this need ever arises.
`module_reexports` is also un-querified, because I don't see any reasons to make it a query, only overhead.
Pass host linker to compiletest.
Tests marked `// force-host` were using the default linker, even if a custom linker was configured in config.toml.
This change adds a new flag, --host-linker, to compiletest, and renames --linker to --target-linker.
The panic test is now counted as an error test; we encounter a Terminate
terminator, and emit an interpreter error, as opposed to just
terminating due to a panic. So this test should have broken with
https://github.com/rust-lang/rust/pull/102906 but wasn't because the Miri
test suite is currently broken in rust-lang/rust:
https://github.com/rust-lang/rust/issues/110102
Rollup of 6 pull requests
Successful merges:
- #109724 (prioritize param env candidates if they don't guide type inference)
- #110021 (Fix a couple ICEs in the new `CastKind::Transmute` code)
- #110044 (Avoid some manual slice length calculation)
- #110115 (compiletest: Use remap-path-prefix only in CI)
- #110121 (Fix `x check --stage 1` when download-rustc is enabled)
- #110124 (Some clippy fixes in the compiler)
Failed merges:
- #109752 (Stall auto trait assembly in new solver for int/float vars)
r? `@ghost`
`@rustbot` modify labels: rollup
compiletest: Use remap-path-prefix only in CI
This makes jump-to-definition work in most IDEs, as well as being easier to understand for contributors.
Fixes https://github.com/rust-lang/rust/issues/109725. cc `@TimNN`
Make elaboration generic over input
Combines all the `elaborate_*` family of functions into just one, which is an iterator over the same type that you pass in (e.g. elaborating `Predicate` gives `Predicate`s, elaborating `Obligation`s gives `Obligation`s, etc.)
Refactor unwind in MIR
This makes unwinding from current `Option<BasicBlock>` into
```rust
enum UnwindAction {
Continue,
Cleanup(BasicBlock),
Unreachable,
Terminate,
}
```
cc `@JakobDegen` `@RalfJung` `@Amanieu`
Tests marked `// force-host` were using the default linker, even if a
custom linker was configured in config.toml.
This change adds a new flag, --host-linker, to compiletest, and renames
--linker to --target-linker.
