| Age | Commit message (Collapse) | Author | Lines |
|---|
|
This reverts commit 399a258f46074740862568b124c02f7b7d04638c.
|
|
|
|
Rollup of 4 pull requests
Successful merges:
- #123541 (remove miri-test-libstd hacks that are no longer needed)
- #123552 (Add missing -Zquery-dep-graph to the spike-neg incr comp tests)
- #123553 (Miri subtree update)
- #123554 (Simplify/cleanup `search-result-color.goml`)
r? `@ghost`
`@rustbot` modify labels: rollup
|
|
r=notriddle
Simplify/cleanup `search-result-color.goml`
Greatly shorten code of `search-result-color.goml` GUI test.
I split the changes into smaller commits to allow to more easily see what changed.
r? `@notriddle`
|
|
Miri subtree update
r? `@ghost`
|
|
Add missing -Zquery-dep-graph to the spike-neg incr comp tests
This ensures that the tests actually test what they are meant to test rather than exitting immediately with an error that -Zquery-dep-graph has to be passed.
|
|
remove miri-test-libstd hacks that are no longer needed
In https://github.com/rust-lang/rust/pull/123317 we developed a different approach to testing the standard library in Miri, and with https://github.com/rust-lang/miri-test-libstd/pull/56 the out-of-tree miri-test-libstd has been switched to that approach. That makes these hacks here no longer necessary.
|
|
optimize tidy check on `src/tools/tidy/src/issues.txt`
This change applies to the following:
- Handles `is_sorted` in the first iteration without needing a second.
- Fixes line sorting on `--bless`.
- Reads `issues.txt` as str rather than making it part of the source code.
Fixes #123002
|
|
|
|
|
|
This ensures that the tests actually test what they are meant to test
rather than exitting immediately with an error that -Zquery-dep-graph
has to be passed.
|
|
`tests/rustdoc-gui/search-result-color.goml`
|
|
Rollup of 4 pull requests
Successful merges:
- #114788 (impl get_mut_or_init and get_mut_or_try_init for OnceCell and OnceLock)
- #122291 (Stabilize `const_caller_location` and `const_location_fields`)
- #123357 (CI: Redirect stderr to stdout to order GHA logs)
- #123504 (bootstrap: split cargo-miri test into separate Step)
r? `@ghost`
`@rustbot` modify labels: rollup
|
|
|
|
make 'missing extern static' error consistent with missing shim
What's relevant is mostly the link name of the external symbol, not its Rust path.
|
|
|
|
This change applies to the following:
- Handles `is_sorted` in the first iteration without needing a second.
- Fixes line sorting on `--bless`.
- Reads `issues.txt` as str rather than making it part of the source code.
Signed-off-by: onur-ozkan <work@onurozkan.dev>
|
|
bootstrap: split cargo-miri test into separate Step
This makes it easier to test just the driver or the cargo-miri integration.
````@rust-lang/miri```` this means to test both you now need to do `./x.py test miri cargo-miri`.
|
|
CI: Redirect stderr to stdout to order GHA logs
This PR modifies the main CI workflow so that its stderr is redirected to stdout. This should make it so that stderr and stdout output is not interleaved in the wrong order in GHA logs (see discussion [here](https://rust-lang.zulipchat.com/#narrow/stream/242791-t-infra/topic/Github.20actions.20logs.20show.20lines.20in.20the.20wrong.20order)).
|
|
Stabilize `const_caller_location` and `const_location_fields`
Closes #102911. Closes #76156.
tests: [library/core/tests/panic/location.rs](https://github.com/rust-lang/rust/blob/3521a2f2f317cb978063842485c7d1bc86ec82b6/library/core/tests/panic/location.rs)
API:
```rust
// core::panic::location
impl Location {
pub const fn caller() -> &'static Location<'static>;
pub const fn file(&self) -> &str;
pub const fn line(&self) -> u32;
pub const fn column(&self) -> u32;
}
```
|
|
impl get_mut_or_init and get_mut_or_try_init for OnceCell and OnceLock
See also https://github.com/rust-lang/rust/issues/74465#issuecomment-1676522051
I'm trying to understand the process for such proposal. And I'll appreciate it if anyone can guide me the next step for consensus or adding tests.
|
|
Bump dependencies
Follow up for #123252
Unfortunately this file needs to be manually bumped when any dependencies are bumped in the main lockfile
```
Updating autocfg v1.1.0 -> v1.2.0
Updating chrono v0.4.35 -> v0.4.37
Updating clap v4.5.3 -> v4.5.4
Updating clap_derive v4.5.3 -> v4.5.4
Updating handlebars v5.1.0 -> v5.1.2
Updating itoa v1.0.10 -> v1.0.11
Updating memoffset v0.9.0 -> v0.9.1
Updating openssl-sys v0.9.101 -> v0.9.102
Updating pin-project-lite v0.2.13 -> v0.2.14
Updating r-efi v4.3.0 -> v4.4.0
Updating regex-syntax v0.8.2 -> v0.8.3
Updating security-framework v2.9.2 -> v2.10.0
Updating security-framework-sys v2.9.1 -> v2.10.0
Updating serde_json v1.0.114 -> v1.0.115
Updating syn v2.0.55 -> v2.0.57
Updating tokio v1.36.0 -> v1.37.0
```
|
|
Rollup of 7 pull requests
Successful merges:
- #123294 (Require LLVM_CONFIG to be set in rustc_llvm/build.rs)
- #123467 (MSVC targets should use COFF as their archive format)
- #123498 (explaining `DefKind::Field`)
- #123519 (Improve cfg and check-cfg configuration)
- #123525 (CFI: Don't rewrite ty::Dynamic directly)
- #123526 (Do not ICE when calling incorrectly defined `transmute` intrinsic)
- #123528 (Hide async_gen_internals from standard library documentation)
r? `@ghost`
`@rustbot` modify labels: rollup
|
|
|
|
Hide async_gen_internals from standard library documentation
These are from https://github.com/rust-lang/rust/pull/118420. It doesn't appear that there is any intention to ever make these APIs available to user code. These are just conveniences meant for the compiler's implementation of `async gen`. I don't think having them featured in documentation in <https://doc.rust-lang.org/1.77.1/core/task/enum.Poll.html> is appropriate.
|
|
Do not ICE when calling incorrectly defined `transmute` intrinsic
Fix #123442
|
|
CFI: Don't rewrite ty::Dynamic directly
Now that we're using a type folder, the arguments in predicates are processed automatically - we don't need to descend manually.
We also want to keep projection clauses around, and this does so.
r? `@compiler-errors`
|
|
r=wesleywiser
Improve cfg and check-cfg configuration
This PR improves cfg and check-cfg configuration by:
1. Extracting both logic under a common module (to improve the connection between the two)
2. Adding more documentation, in particular some steps when adding a new cfg
I also added my-self as mention in our triagebot conf for the new module.
Inspired by https://github.com/rust-lang/rust/pull/123411#discussion_r1554056681
|
|
explaining `DefKind::Field`
|
|
MSVC targets should use COFF as their archive format
While adding support for Arm64EC I ran into an issue where the standard library's rlib was missing the "EC Symbol Table" which is required for the MSVC linker to find import library symbols (generated by Rust's `raw-dylib` feature) when building for EC.
The root cause of the issue is that LLVM only generated symbol tables (including the EC Symbol Table) if the `ArchiveKind` is `COFF`, but the MSVC targets didn't set their archive format, so it was defaulting to GNU.
|
|
Require LLVM_CONFIG to be set in rustc_llvm/build.rs
This environment variable should always be set by bootstrap in `rustc_llvm_env`. The fallback is quite ugly and complicated, so removing it is nice.
https://github.com/rust-lang/rust/blob/bf71daedc29e7a240261acd1516378047e311a6f/src/bootstrap/src/core/build_steps/compile.rs#L1166
I tried finding when this was added in git history, but it pointed all the way to "add build scripts" at which point I stopped digging more. This has always been here.
cc `@nikic` `@cuviper` in case you happen to be aware of a deeper reason behind this
r? bootstrap
|
|
Check def id before calling `match_projection_projections`
When I "inlined" `assemble_candidates_from_predicates` into `for_each_item_bound` in #120584, I forgot to copy over the check that actually made sure the def id of the candidate was equal to the def id of the obligation. This means that we normalize goal a bit too often even if it's not productive to do so.
This PR adds that def id check back.
Fixes #123448
|
|
Rustup
|
|
|
|
|
|
chore: fix some typos
|
|
Signed-off-by: findseat <penglili@outlook.com>
|
|
Update cargo
9 commits in 0637083df5bbdcc951845f0d2eff6999cdb6d30a..28e7b2bc0a812f90126be30f48a00a4ada990eaa
2024-04-02 23:55:05 +0000 to 2024-04-05 19:31:01 +0000
- refactor(toml): Decouple target discovery from Target creation (rust-lang/cargo#13701)
- Don't depend on `?` affecting type inference in weird ways (rust-lang/cargo#13706)
- test(metadata): Show behavior with TOML-specific types (rust-lang/cargo#13703)
- fix: adjust tracing verbosity in list_files_git (rust-lang/cargo#13704)
- doc: comments on `PackageRegistry` (rust-lang/cargo#13698)
- Switch to using gitoxide by default for listing files (rust-lang/cargo#13696)
- Allow precise update to prerelease. (rust-lang/cargo#13626)
- refactor(toml): Split out an explicit step to resolve `Cargo.toml` (rust-lang/cargo#13693)
- chore(deps): update rust crate base64 to 0.22.0 (rust-lang/cargo#13675)
r? ghost
|
|
|
|
|
|
Fix #123442
|
|
Remove rt::init allocation for thread name
This removes one of the allocations in a `fn main() {}` program.
|
|
Now that we're using a type folder, the arguments in predicates are
processed automatically - we don't need to descend manually.
We also want to keep projection clauses around, and this does so.
|
|
Rollup of 8 pull requests
Successful merges:
- #121419 (Add aarch64-apple-visionos and aarch64-apple-visionos-sim tier 3 targets)
- #123159 (Fix target-cpu fpu features on Arm R/M-profile)
- #123487 (CFI: Restore typeid_for_instance default behavior)
- #123500 (Revert removing miri jobserver workaround)
- #123505 (Revert "Use OS thread name by default")
- #123509 (Add jieyouxu to compiler review rotation and as a reviewer for `tests/run-make`, `src/tools/run-make-support` and `src/tools/compiletest`)
- #123514 (Fix typo in `compiler/rustc_middle/src/traits/solve/inspect.rs`)
- #123515 (Use `include` command to reduce code duplication)
r? `@ghost`
`@rustbot` modify labels: rollup
|
|
|
|
|
|
Use `include` command to reduce code duplication
Since we updated `browser-ui-test` version, let's make use of the new commands. :)
r? `@notriddle`
|
|
Fix typo in `compiler/rustc_middle/src/traits/solve/inspect.rs`
r? lcnr
(typo from #123363)
`@bors` rollup=always
|
|
r=Mark-Simulacrum
Add jieyouxu to compiler review rotation and as a reviewer for `tests/run-make`, `src/tools/run-make-support` and `src/tools/compiletest`
|
|
Revert "Use OS thread name by default"
This reverts #121666 (Use the OS thread name by default if `THREAD_INFO` has not been initialized) due to #123495 (Thread names are not always valid UTF-8).
It's not a direct revert because there have been other changes since that PR.
|
