Previously, `x.py` would unconditionally run `x.py build` to get the
help message. After https://github.com/rust-lang/rust/issues/76165,
when checking the CI stage was moved into `Config`, that would cause an
assertion failure (but only only in CI!):
```
thread 'main' panicked at 'assertion failed: `(left == right)`
left: `1`,
right: `2`', src/bootstrap/config.rs:619:49
```
This changes bootstrap to only generate a help message when it needs
to (when someone passes `--help`).
Rollup of 10 pull requests
Successful merges:
- #76056 (Add more info for Vec Drain doc)
- #76062 (Vec slice example fix style and show type elision)
- #76262 (Use inline(never) instead of cold)
- #76335 (Make all methods of `Duration` unstably const)
- #76366 (Add Arith Tests in Library)
- #76369 (Move Various str tests in library)
- #76534 (Add doc comments for From impls)
- #76622 (Update bootstrap readme)
- #76641 (Some cleanup changes and commenting)
- #76662 (Fix liballoc test suite for Miri)
Failed merges:
r? `@ghost`
Update bootstrap readme
- Reflect changes in x.py defaults
- Remove recommendation to use nightly for incremental; it works fine on
beta
- Remove note that incremental chooses stage 1 by default; stage 1 is
already the default
- Update Discord -> Zulip
r? @Mark-Simulacrum
Make all methods of `Duration` unstably const
Make the following methods of `Duration` unstable const under `duration_const_2`:
- `from_secs_f64`
- `from_secs_f32`
- `mul_f64`
- `mul_f32`
- `div_f64`
- `div_f32`
This results in all methods of `Duration` being (unstable) const.
Moved the tests to `library` as part of #76268.
Possible because of #72449, which made the relevant `f32` and `f64` methods const.
Tracking issue: #72440
r? @ecstatic-morse
Make the default stage for x.py configurable
This also allows configuring each sub-command individually.
Possibly #76617 should land before this? I don't feel strongly either way, I don't mind waiting.
Closes https://github.com/rust-lang/rust/issues/76165.
r? `@Mark-Simulacrum`
Rollup of 10 pull requests
Successful merges:
- #73955 (deny(unsafe_op_in_unsafe_fn) in libstd/process.rs)
- #75146 (Detect overflow in proc_macro_server subspan)
- #75304 (Note when a a move/borrow error is caused by a deref coercion)
- #75749 (Consolidate some duplicate code in the sys modules.)
- #75882 (Use translated variable for test string)
- #75886 (Test that bounds checks are elided for [..index] after .position())
- #76048 (Initial support for riscv32gc_unknown_linux_gnu)
- #76198 (Make some Ordering methods const)
- #76689 (Upgrade to pulldown-cmark 0.8.0)
- #76763 (Update cargo)
Failed merges:
r? `@ghost`
Update cargo
6 commits in 875e0123259b0b6299903fe4aea0a12ecde9324f..8777a6b1e8834899f51b7e09cc9b8d85b2417110
2020-09-08 20:17:21 +0000 to 2020-09-15 19:11:03 +0000
- updated yank error message (rust-lang/cargo#8697)
- Fix non-determinism with new feature resolver. (rust-lang/cargo#8701)
- Display formatted output for JSON diffing in tests. (rust-lang/cargo#8692)
- Add --name suggestion for cargo new (rust-lang/cargo#8675)
- Sweep unrelated message from unnecessary workspace infromation (rust-lang/cargo#8681)
- Docs: Make it more clear we have two types of workspaces (rust-lang/cargo#8666)
Note when a a move/borrow error is caused by a deref coercion
Fixes#73268
When a deref coercion occurs, we may end up with a move error if the
base value has been partially moved out of. However, we do not indicate
anywhere that a deref coercion is occuring, resulting in an error
message with a confusing span.
This PR adds an explicit note to move errors when a deref coercion is
involved. We mention the name of the type that the deref-coercion
resolved to, as well as the `Deref::Target` associated type being used.
Gate macOS on both Azure and GHA
As discussed in the previous infrastructure team meeting, this PR gates macOS builds on both GHA and Azure. Once this is merged we'll wait a week or two to see if there is a troublesome rate of spurious failures, and if not we'll remove the builds on the Azure side.
r? `@Mark-Simulacrum`
cc #71988
improve diagnostics for lifetime after `&mut`
If, when parsing a borrow pointee type, we see a lifetime after `mut`, suggest placing the lifetime before `mut` and eat the lifetime to avoid a large number of unhelpful diagnostics.
There are some subtleties to avoid false positives in cases like `&mut 'a + Trait`, where `&mut ('a + Trait)` is a better suggestion.
fixes#73568
This allows configuring the default stage for each sub-command individually.
- Normalize the stage as early as possible, so there's no confusion
about which to use.
- Don't add an explicit `stage` option in config.toml
This offers no more flexibility than `*_stage` and makes it confusing
which takes precedence.
- Always give `--stage N` precedence over config.toml
- Fix bootstrap tests
This changes the tests to go through `Config::parse` so that they test
the actual defaults, not the dummy ones provided by `default_opts`. To
make this workable (and independent of the environment), it does not
read `config.toml` for tests.
Always try to promote shared LLVM to the sysroot
Even when LLVM is not generally participating in a shared link with rustc, we
will likely still link to the shared dylib from rust-lld, so we still need to
promote it.
This reverts part of #76349; my expectation that the link-shared rule was sufficient was likely wrong.
Hopefully fixes#76698.
r? `@alexcrichton`
Windows doesn't quite support dynamic linking to LLVM yet, but on other
platforms we do. In #76708, it was discovered that we dynamically link to LLVM
from the LLVM tools (e.g., rust-lld), so we need the shared LLVM library to link
against. That means that if we do not have a shared link to LLVM, and want LLVM
tools to work, we'd be shipping two copies of LLVM on all of these platforms:
one in librustc_driver and one in libLLVM.
Also introduce an error into rustbuild if we do end up configured for shared
linking on Windows.
Split `core::slice` to smaller mods
Unfortunately the `#[lang = "slice"]` is too big (3003 lines), I cannot split it further.
Note for reviewer:
* I split to multiple commits for easier reviewing, but I could git squash them all to one if requested.
* Recommend pulling this change locally and using advanced git diff viewer or this command:
```
git show --reverse --color-moved=dimmed-zebra master..
```
---
I split core/slice/mod.rs to these modules:
* `ascii`: For operations on `[u8]`.
* `cmp`: For comparison operations on `[T]`, like PartialEq and SliceContains impl.
* `index`: For indexing operations like Index/IndexMut and SliceIndex.
* `iter`: For Iterator definitions and implementation on `[T]`.
- `macros`: For iterator! and forward_iterator! macros.
* `raw`: For free function to create `&[T]` or `&mut [T]` from pointer + length or a reference.
The heapsort wrapper in mod.rs is removed in favor of reexport from `sort::heapsort`.
Refactor intra doc link code
I got tired of `fold_item` being 500 lines long.
This is best reviewed one commit at a time with whitespace changes hidden.
There are no logic changes other than the last commit making a parameter checked by the caller instead of the callee.
r? `@Manishearth`
More structured suggestions for boxed trait objects instead of impl Trait on non-coerceable tail expressions
When encountering a `match` or `if` as a tail expression where the
different arms do not have the same type *and* the return type of that
`fn` is an `impl Trait`, check whether those arms can implement `Trait`
and if so, suggest using boxed trait objects.
Use structured suggestion for `impl T` to `Box<dyn T>`.
Fix https://github.com/rust-lang/rust/issues/69107
Stabilize doc_alias feature
Fixes#50146.
This PR intend to stabilize the `doc_alias` feature. The last remaining bits were missing checks on the attribute usage and on its arguments. Both have been added so I think we can now move to the next step.
r? `@ollie27`
cc `@rust-lang/rustdoc`
Ignore rustc_private items from std docs
By ignoring rustc_private items for non local impl block,
this may fix#74672 and fix#75588 .
This might suppress #76529 if it is simple enough for backport.
Auto-generate lint documentation.
This adds a tool which will generate the lint documentation in the rustc book automatically. This is motivated by keeping the documentation up-to-date, and consistently formatted. It also ensures the examples are correct and that they actually generate the expected lint. The lint groups table is also auto-generated. See https://github.com/rust-lang/compiler-team/issues/349 for the original proposal.
An outline of how this works:
- The `declare_lint!` macro now accepts a doc comment where the documentation is written. This is inspired by how clippy works.
- A new tool `src/tools/lint-docs` scrapes the documentation and adds it to the rustc book during the build.
- It runs each example and verifies its output and embeds the output in the book.
- It does a few formatting checks.
- It verifies that every lint is documented.
- Groups are collected from `rustc -W help`.
I updated the documentation for all the missing lints. I have also added an "Explanation" section to each lint providing a reason for the lint and suggestions on how to resolve it.
This can lead towards a future enhancement of possibly showing these docs via the `--explain` flag to make them easily accessible and discoverable.
allow concrete self types in consts
This is quite a bad hack to fix#75486. There might be a better way to check if the self type depends on generic parameters, but I wasn't able to come up with one.
r? `@varkor` cc `@petrochenkov`
inliner: Emit storage markers for introduced arg temporaries
When introducing argument temporaries during inlining, emit storage
marker statements just before the assignment and in the beginning of
the return block.
This ensures that such temporaries will not be considered live across
yield points after inlining inside a generator.
Fixes#71793.
Thanks to marcusklaas' hard work in https://github.com/raphlinus/pulldown-cmark/pull/469, this fixes a lot of rustdoc bugs!
- Get rid of unnecessary `RefCell`
- Fix duplicate warnings for broken implicit reference link
- Remove unnecessary copy of links
Previously, `resolve` would immediately check that `module_id` was
non-empty and give an error if not. This had two downsides:
- It introduced `Option`s everywhere, even if the calling function knew
it had a valid module, and
- It checked the module on each namespace, which is unnecessary: it only
needed to be checked once.
This makes the caller responsible for checking the module exists, making
the code a lot simpler.
