Repository navigation
Rollup of 8 pull requests - #164430
Merged
Merged
Rollup of 8 pull requests#164430
Conversation
…overflow Self-referential RTN expands args on each unrolling step, bypassing the `self.types.insert` cycle check. Track depth against `recursion_limit` and short-circuit with `ControlFlow::Break` when exceeded so we don't blow the stack.
Signed-off-by: Amirhossein Akhlaghpour <m9.akhlaghpoor@gmail.com>
Install LLVM 23 from apt.llvm.org.
…s, r=adwinwhite handle trait solver overflow in `codegen_select_candidate` We now report overflow errors in `codegen_select_candidate` instead of failing to resolve instance and ICE. fixes rust-lang#162343 fixes rust-lang/trait-system-refactor-initiative#306
Fix: Construct byte string valtree array type using literal length closes rust-lang#163904 ## Bug Mentioned in the issue. The reason type checking was not catching the `SizeOverflow` before `const` evaluation starts was: The trait solver just checks that the constant's own type equals the declared type https://github.com/rust-lang/rust/blob/472684ea0c352dad27153b2a9646b0fcd7c8daf8/compiler/rustc_trait_selection/src/traits/fulfill.rs?plain=1#L582-L595 Hence building the value lead to `Err` value from `layout_of` leading panic in `valtree_to_ref`. https://github.com/rust-lang/rust/blob/472684ea0c352dad27153b2a9646b0fcd7c8daf8/compiler/rustc_const_eval/src/const_eval/valtrees.rs?plain=1#L395 ## Solution As suggested by @BoxyUwU and @oli-obk We shouldn't blindly copy the declared type and size! Rather label the value with its real size. Hence now `[u8; 0] != [u8; usize::MAX]`, so the compiler reports normal error. # How to test Run `./x test tests/ui/const-generics/mgca --stage 1` **Manual test** mentioned in the issue
…=oli-obk refine: respect recursion limit in RPITIT collector to prevent stack overflow Fixes rust-lang#163866. Prevents a compiler stack overflow (SIGSEGV) during `refining_impl_trait` checks when an RPITIT return type references its own RTN projection: trait IntFactory { fn stream(&self) -> impl IntFactory<stream(..): Send>; } impl IntFactory for S { fn stream(&self) -> impl IntFactory<stream(..): Send> { todo!() } } ### Cause `ImplTraitInTraitCollector` was detecting cycles using `self.types.insert(proj)`. Because `proj` is an alias projection type that includes generic substitutions, self-referential RTN expands deeper generic arguments on each unrolling step: - `stream::{opaque#0}<S>` - `stream::{opaque#0}<<S as IntFactory>::stream::{opaque#0}>>` - `stream::{opaque#0}<<<S as IntFactory>::stream::{opaque#0} as IntFactory>::stream::{opaque#0}>>>` Since generic args change on every step, `proj` never compares equal to any previously visited item, causing the visitor to loop infinitely until stack exhaustion. ### Fix Instead of outright banning seen `DefId`s (which would cut off valid finitely unrolling types), track the unrolling depth against `tcx.recursion_limit()`. When depth hits the recursion limit, the visitor now short-circuits immediately via `ControlFlow::Break(())` and `?`. This avoids exploring an exponential branching tree when multiple self-referential bounds exist, cleanly terminating the refine check. Added UI regression test in `tests/ui/impl-trait/in-trait/refine-self-referential-rtn.rs`.
…cope-of-object, r=GuillaumeGomez Make the `documentation` scope a sub-scope of the `object` scope ~~First commit from rust-lang#164278.~~ This PR moves the `documentation` scope to be a sub-scope of the `object` scope. The `object` scope is currently defined to be: "This ensures all paths in compiled executables, libraries and objects outputs are remapped but not elsewhere", which can be translated to mean everything written to disk by `rustc` it-self. It is therefore a natural extension for the `documentation` scope to be a sub-scope of it, like `macro`, `debuginfo` and `coverage`, since we use it for things written down on disk, like rustdoc-json (with rust-lang#164278 at-least) or the source code viewer.
Update LoongArch64 Linux dist builder to LLVM 23 This is also a defensive fix, as certain linker inputs can trigger an ["address assignment did not converge"](llvm/llvm-project#226712) error during linker relaxation in LLD versions prior to 23.1.3, causing the link to fail.
…smutability, r=ShoyuVanilla Avoid erased region on transmutability It seems like the erasure of regions here wasn't needed anymore since all tests pass and removing it makes the next-solver looses it's lifetime extensions powers. Fixed the last test in rust-lang/trait-system-refactor-initiative#152 Closes rust-lang#164246 [discussed in zulip](https://rust-lang.zulipchat.com/#narrow/channel/618216-t-types.2Fcall-for-participation/topic/next.20solver.20and.20the.20philosopher.20stone/with/631772202) r? @lcnr
…v-libc, r=jieyouxu cargotest: update xsv libc for RISC-V Fixes rust-lang#164339. The xsv repository pinned by `cargotest` has a lockfile from 2018 that uses `libc 0.2.43`. That version fails to compile on RISC-V because it does not define `c_char` for the target. Add a `cargotest` lockfile for xsv with only `libc` updated to `0.2.190` and use the existing lockfile override mechanism. This keeps xsv and its other dependencies at their current pinned versions. Tests: - `cargo update -p libc --precise 0.2.190` in the pinned xsv checkout - `cargo check --target riscv64gc-unknown-linux-gnu --tests` in xsv - `cargotest <cargo> <out-dir> xsv` (394 selected tests passed on x86_64) The full test suite has not been run on RISC-V hardware.
…rame-init-hook, r=RalfJung Implement after-stack-frame-init-hook Adds a new hook to `compiler/rustc_const_eval/src/interpret/machine.rs` that gets called immediately after a stack frame's function arguments are fully initialized. r?<@RalfJung>
Member
Author
Contributor
This comment has been minimized.
This comment has been minimized.
rust-bors Bot
pushed a commit
that referenced
this pull request
Oct 16, 2026
Rollup of 8 pull requests try-job: dist-various-1 try-job: test-various try-job: test-x86_64-gnu-aux try-job: test-x86_64-msvc-1 try-job: test-aarch64-apple-1 try-job: test-aarch64-apple-2 try-job: test-x86_64-mingw-1 try-job: test-i686-msvc try-job: test-armhf-gnu try-job: test-x86_64-gnu-llvm-22-3
This comment has been minimized.
This comment has been minimized.
Contributor
Contributor
Contributor
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 bf9448e (parent) -> 5d8fec0 (this PR) Test differencesShow 46 test diffsStage 1
Stage 2
Additionally, 3 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 5d8fec0669a64607db9ccbf663e7c23125bf4748 --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 |
Collaborator
|
Finished benchmarking commit (5d8fec0): comparison URL. Overall result: no relevant changes - no action needed@rustbot label: -perf-regression Instruction countThis perf run didn't have relevant results for this metric. Max RSS (memory usage)Results (secondary -0.1%)A less reliable metric. May be of interest, but not used to determine the overall result above.
CyclesResults (secondary -2.0%)A less reliable metric. May be of interest, but not used to determine the overall result above.
Binary sizeThis perf run didn't have relevant results for this metric. Bootstrap: 488.875s -> 490.438s (0.32%) |
Contributor
|
📌 Perf builds for each rolled up PR:
parent commit: bf9448e426 In the case of a perf regression, run the following command with the SHAs of each PR you suspect might be the cause: |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
10 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Successful merges:
codegen_select_candidate#162467 (handle trait solver overflow incodegen_select_candidate)documentationscope a sub-scope of theobjectscope #164279 (Make thedocumentationscope a sub-scope of theobjectscope)r? @ghost
Create a similar rollup