Use https for sourceforge during CI
I saw that we use http during CI opening up the CI process to on the wire tampering.
based on #86573
r? `@pietroalbini`
Do not spawn blacklisted_name lint in test context
---
fixed#7305
*Please write a short comment explaining your change (or "none" for internal only changes)*
changelog: `blacklisted_name` lint is not spawned in the test context anymore.
Fix detecting of the 'test' attribute
Update UI test to actually check that warning is not triggered in the test code
Fix approach for detecting the test module
Add nested test case
Remove code duplication by extracting 'is_test_module_or_function' into 'clippy_utils'
Cleanup the code
Added `cargo dev setup git-hook` and updated `cargo dev setup intellij` including a `remove` command
This PR enables our dev tool to install a git hook that formats the code before each commit and also runs `update_lints` to make sure that everything is registered correctly. The script is located at `util/etc/pre-commit.sh`. I found it reasonable to locate it in the `util` folder and decided to add a `etc` in correlation to the main rust repo and to bring a bit of structure into it.
* The hook can be installed via: `cargo dev setup git-hook`
* And removed via: `cargo dev remove git-hook`
cc: #5394
The refactoring of `src/ide_setup.rs` to `src/setup/intellij.rs` is an extra commit to simplify the review.
---
Changes:
* Added `cargo dev setup git-hook` for formatting before every commit
* Added `cargo dev remove git-hook` to remove the hook again
* Added `cargo dev remove intellij` to remove rustc source path dependencies
* Changed `cargo dev ide_setup` to `cargo dev setup intellij`
changelog: none
This is only an internal change and therefore not worth an entry in the general change log.
---
Tested on:
* [x] Linux (by `@xFrednet)`
* [ ] Windows (All used commands run inside the git bash, so it's very likely to work as well `@xFrednet)`
* [ ] macOS
rustc_data_structures has a dependency on crossbeam-utils but never uses
it. It appears to have originally had this dependency in order to set
the "nightly" feature; however, its other dependencies use a different
version of crossbeam-utils, so this doesn't actually affect anything.
Furthermore, in current crossbeam-utils, the "nightly" feature has
become a no-op.
The sha-1 and md-5 packages contain crates named sha1 and md5,
respectively. This discrepancy makes it somewhat more challenging to
automate detection of unused crates. Explicitly rename the packages to
the names of the crates they contain, to simplify such detection.
Add `BuildHasher::hash_one` as unstable
Inspired by https://github.com/rust-lang/rust/pull/86140/files#diff-246941135168fbc44fce120385ee9c3156e08a1c3e2697985b56dcb8d728eedeR2416, where I wanted to write a quick test for a `Hash` implementation and it took more of a dance than I'd hoped.
It looks like this would be handy in hashtable implementations, too -- a quick look at hashbrown found two places where it needs to do the same dance:
6302512a8a/src/map.rs (L247-L270)
I wanted to get a "seems plausible" from a libs member before making a tracking issue, so random-sampling the intersection of highfive and governance gave me...
r? `@joshtriplett`
(As always, bikeshed away! And let me know if I missed something obvious again that I should have used instead.)
Don't lint :pat when re-parsing a macro from another crate.
`compile_macro` is used both when compiling the original definition in the crate that defines it, and to compile the macro when loading it when compiling a crate that uses it. We should only emit lints in the first case.
This adds a `is_definition: bool` to pass this information in, so we don't warn about things that only concern the definition site.
Fixes#86567
Even if the content from box is used in a sharef-ref context,
we capture the box entirerly.
This is motivated by:
1) We only capture data that is on the stack.
2) Capturing data from within the box might end up moving more data than
the user anticipated.
tidy: verify that test revisions with --target have associated needs-llvm-components directives
This ensures that people who tend to write `--target` `#[no_core]` tests don't miss specifying the `needs-llvm-components` directive. This is necessary for the test suite to pass when LLVM is compiled with a subset of components enabled.
While here I also took the opportunity to implement a more fine-grained handling of the ignore directives, so that they are evaluated for each revision, rather than for the entire test. With this even if people have `arm` component disabled, only the revision that depends on the arm component will not run.
Fixes https://github.com/rust-lang/rust/issues/82405
Add import_rename lint, this adds a field on the Conf struct
fixes#7276
changelog: Add ``[`import_rename`]`` a lint that enforces import renaming defined in the config file.
Otherwise something that ought to seemingly work like `//[x86]
needs-llvm-components: x86` or `//[nll_beyond]should-fail` do not get
evaluated properly.
