Rollup of 13 pull requests#157991
Conversation
Fixes <rust-lang#147714>. The changes in `const_to_pat.rs` are entirely to avoid regressing diagnostic quality, and should not make any difference to what code is accepted. The change in `compute_applicable_impls_for_diagnostics` and its callers is entirely to be able to reuse that algorithm for this purpose.
These functions make it easy to read a fixed-size type as little-endian or big-endian. They're trivial wrappers around the combination of `io::Read::read_array` and `T::from_le_bytes`/`T::from_be_bytes`. The implementation uses a sealed trait `FromEndianBytes`. That trait is currently in `std` and accepts a `&mut io::Read`. Once we can use associated consts in the types of method parameters, we can change this trait to have `from_le_bytes` and `from_be_bytes` methods, move it to `core`, and make it public.
…ct-return The new Rust symbol mangling[0] uses prefix `_R` instead of `@_ZN`. [0] https://rust-lang.github.io/rfcs/2603-rust-symbol-name-mangling-v0.html
Instead of keeping it as `only-bpf`, use minicore. This way we make sure it will never get outdated again.
connect_timeout cast timeout.as_secs() (u64) straight to tv_sec, which is c_long (i32 on 32-bit Solid), so timeouts over ~68 years truncated to a wrong or negative value. Clamp to c_long::MAX, matching set_timeout in the same file.
Removes the `enabledManagers` override so all managers run, but requires manual approval from the Dependency Dashboard before any PRs are opened. The dashboard then acts as a preview without cluttering the PR list.
Subtrees are updated in their upstream repositories and synced into the tree, so Renovate should not manage their dependencies here. Without this, Renovate would surface pinned actions from every subtree workflow file (see rust-lang#134127). Mirrors the subtree list in `external-repos.md`.
Pins GitHub Actions to their commit SHA digests and keeps them updated. This lets Renovate handle the pinning that was proposed in rust-lang#155089 instead of maintaining it manually.
* fix(rustc_codegen_ssa): Use cg_operand for Freeze check * chore: remove unnecessary comment * test: add test for ssa rval monomorphization issue * fix: move ssa-rval-monomorphization test * tweak comment Co-authored-by: Ralf Jung <post@ralfj.de>
…liasTerm, r=lcnr Refactor `AliasTy`. `AliasTerm` & `UnevaluatedConst` to use `Alias` Refactors `AliasTy`, `AliasTerm` & `UnevaluatedConst` to use `Alias`. Part of rust-lang#156181 r? @lcnr
Add `T: PartialEq` bounds to derived `StructuralPartialEq` impls. Fixes <rust-lang#147714>. This is a breaking change to fix a bug, so it will need a crater run if it is to be done at all. The changes in `const_to_pat.rs` are entirely to avoid regressing diagnostic quality, and should not make any difference to what code is accepted. The changes in `compute_applicable_impls_for_diagnostics` and its callers are entirely to be able to reuse that algorithm for this purpose. @rustbot label +T-lang
…-symbols, r=petrochenkov Staticlib rename internal symbols Follow-up to rust-lang#155338. `-Zstaticlib-rename-internal-symbols` appends a crate-specific suffix (`_rs{StableCrateId}`) to non-exported symbols, resolving duplicate symbol conflicts when linking multiple Rust staticlibs into the same binary. The implementation collects all defined `GLOBAL/WEAK` symbol names not in the exported set across all .o files, then renames them by extending the strtab and patching symbol name offsets. When combined with `-Zstaticlib-hide-internal-symbols`, the renamed symbols also receive `STV_HIDDEN` visibility. Supported on ELF targets (Linux, BSD, etc.) and Apple targets (macOS, iOS, etc.). On unsupported targets (Windows), a warning is emitted and the flag has no effect. r?@bjorn3 @petrochenkov
Add `io::Read::read_le` and `io::Read::read_be` These functions make it easy to read a fixed-size type as little-endian or big-endian. They're trivial wrappers around the combination of `io::Read::read_array` and `T::from_le_bytes`/`T::from_be_bytes`. The implementation uses a sealed trait `FromEndianBytes`. That trait is currently in `std` and accepts a `&mut io::Read`. Once we can use associated consts in the types of method parameters, we can change this trait to have `from_le_bytes` and `from_be_bytes` methods, move it to `core`, and make it public. --- As discussed at RustWeek.
…i-indirect-return, r=nagisa tests: codegen-llvm: Expect the new mangling scheme in bpf-abi-indirect-return [The new Rust symbol mangling](https://rust-lang.github.io/rfcs/2603-rust-symbol-name-mangling-v0.html) uses prefix `_R` instead of `@_ZN`. Instead of keeping the test as `only-bpf`, use minicore. This way we make sure it will never get outdated again. r? @nagisa
…info, r=davidtwco,estebank Add expansion info to implied bounds r? @estebank maybe? cc @davidtwco for implied bound diagnostics We used to have various fun ways for testing whether a `Sized` bound was implicit or not (usually by checking if the span was equal to the span of a generic parameter). That felt a bit hacky, so I checked what happens if we explicitly mark the span as being desugared. Turns out a lot of diagnostics break. You can see in the diff where I changed things to make them work again. Then it became clear that we were missing out on various other implied bound sites, and we just got all of them basically for free now. I think there are various other good refactorings we can do (probably not special casing `Sized` and not special casing implied bound desugarings, but just reporting all desugarings), but for now this PR is in a state where I'd rather merge it than add more to it
…-vec, r=oli-obk constify `TryFrom<Vec>` for array Related issues rust-lang#79597, rust-lang#143773, rust-lang#133214 this makes `Vec` `const Destruct` along the way.
|
@bors r+ rollup=never p=5 |
This comment has been minimized.
This comment has been minimized.
What is this?This is an experimental post-merge analysis report that shows differences in test outcomes between the merged PR and its parent PR.Comparing c712ea9 (parent) -> 9e2abe0 (this PR) Test differencesShow 640 test diffsStage 1
Stage 2
Additionally, 618 doctest diffs were found. These are ignored, as they are noisy. Job group index
Test dashboardRun cargo run --manifest-path src/ci/citool/Cargo.toml -- \
test-dashboard 9e2abe0c6ab27fcbb95c30695188a75776e2feb1 --output-dir test-dashboardAnd then open Job duration changes
How to interpret the job duration changes?Job durations can vary a lot, based on the actual runner instance |
|
Finished benchmarking commit (9e2abe0): comparison URL. Overall result: ❌ regressions - please read:Our benchmarks found a performance regression caused by this PR. Next Steps:
@rustbot label: +perf-regression Instruction countOur most reliable metric. Used to determine the overall result above. However, even this metric can be noisy.
Max RSS (memory usage)Results (primary 1.8%, secondary 2.6%)A less reliable metric. May be of interest, but not used to determine the overall result above.
CyclesResults (primary 2.6%, secondary 0.4%)A less reliable metric. May be of interest, but not used to determine the overall result above.
Binary sizeResults (primary 1.8%, secondary 1.8%)A less reliable metric. May be of interest, but not used to determine the overall result above.
Bootstrap: 519.996s -> 522.871s (0.55%) |
Successful merges:
AliasTy.AliasTerm&UnevaluatedConstto useAlias#156538 (RefactorAliasTy.AliasTerm&UnevaluatedConstto useAlias)T: PartialEqbounds to derivedStructuralPartialEqimpls. #156807 (AddT: PartialEqbounds to derivedStructuralPartialEqimpls.)io::Read::read_leandio::Read::read_be#156983 (Addio::Read::read_leandio::Read::read_be)TryFrom<Vec>for array #155616 (constifyTryFrom<Vec>for array)-Zlint-rust-version#157958 (doc: Document-Zlint-rust-version)errors.rsfile todiagnostics.rs(11/N) #157974 (Renameerrors.rsfile todiagnostics.rs(11/N))r? @ghost
Create a similar rollup