Repository navigation
Open
Labels
T-releaseRelevant to the release subteam, which will review and decide on the PR/issue.Relevant to the release subteam, which will review and decide on the PR/issue.relnotes-tracking-issueMarks issues tracking what text to put in release notes.Marks issues tracking what text to put in release notes.
Description
NOTE: Use the 📝 links to edit those that have a relnotes-tracking-issue,
and they will be updated when we regenerate the notes periodically (manually).
Version 1.100.0 (2026-11-12)
Language
- Stabilize never type
📝 - Lint against repeated
#[repr(..)]attributes
📝 - Stabilize
extern "custom"
📝 - Extend
dropping_references,dropping_copy_typesandundropped_manually_dropslints todrop_in_place
📝 unfulfilled_lint_expectationsno longer fires on#[expect(dead_code)]inside code that is covered by a dead parents' diagnostic
📝- riscv: stabilize 'd' and 'f' target features
📝 - Turn the error for
repr(align)fields in arepr(packed)type into a (deny-by-default) lint,aligned_fields_in_packed.
📝
Compiler
Platform Support
i686-pc-windows-gnuis now a tier 2 without host tools target
📝- Promote
wasm32-wasip3to Tier 2
📝 - Add new Tier 3 target
powerpc64-sony-ps3
📝
Refer to Rust's platform support page
for more information on Rust's tiered platform support.
Libraries
- Stabilize custom allocators
📝 - Map
ENOTSUPtoErrorKind::Unsupported
📝 - volatile: allow accesses to non-AM memory to trap
- Fix soundness issues in std::sys::pal::sgx::waitqueue::unsafe_list
📝
Stabilized APIs
impl PartialEq<Cow<'_, [U]>> for Vec<T, A>impl PartialEq<Cow<'_, [U]>> for &[T]impl PartialEq<Cow<'_, [U]>> for &mut [T]
📝std::alloc::Allocatorstd::alloc::Globalstd::alloc::Systemimpl Allocator for Globalimpl Allocator for Systemimpl<A: Allocator> for &Aimpl<T: Allocator, A: Allocator> for Box<T, A>impl<A: Allocator> for &mut Aimpl<T: Allocator, A: Allocator> for Arc<T, A>impl<T: Allocator, A: Allocator> for Rc<T, A>Box::new_inBox::from_raw_inBox::from_non_null_inBox::into_raw_with_allocatorBox::into_non_null_with_allocatorBox::allocatorVec::new_inVec::with_capacity_inVec::from_raw_parts_inVec::from_parts_inVec::into_raw_parts_with_allocatorVec::into_parts_with_allocatorVec::allocator
📝Child::main_thread_handle
📝impl Default for std::sync::Once
📝bool::toggle
📝Box::take
📝Box::mapRc::mapArc::map
📝core::mem::DropGuard
📝std::mem::conjure_zst
📝
These previously stable APIs are now stable in const contexts:
Cargo
- feat(diag): Stabilize cargo-lints
- docs: adopt rust-lang/rust's LLM usage policy
- fix(update)!: Remove unstable
--breaking - feat(resolver): Stabilize min-publish-age
- Re-stabilize build-dir layout v2
- fix(install): use packaged lockfile by default
Rustdoc
Compatibility Notes
- More precise captures for closure with uses of variables in
if lets may cause new errors
📝 - Always require trait objects to be dyn-compatible
📝 Infallibleis now a type alias to!- The never type fallback was changed to
!on all editions
📝 boxpatterns are no longer syntactically allowed
📝- Report Query Cycle errors in more cases when computing the layout of enums with explicit discriminants
📝 - Always escape grapheme extenders in
str::escape_debug
📝 - Make let-else respect
macro_rulesexprmetavariable grouping
📝 Iterator::{min,max}(_by_key)changed whichOrdmethods they use. This cannot change behaviour (other than being potentially faster) for types correctly following theOrdrules, but may impact types with inconsistent trait implementations.
📝- Emit an error when trying to enable SSE/AVX on an x86 (32bit or 64bit) softfloat target
📝 - Require windowed (and exception) for Xtensa ABI
📝 - windows-gnullvm: always link libunwind statically
📝 - From version 1.94 up to 1.100 a number of ARMv7-A hardfloat targets required the
VFPv3-D32FPU as a minimum, as they erroneously had thed32feature in their baseline list of features. Breakage introduced as a result of correcting this can be fixed by passing-Ctarget-feature=+d32to Rust.
📝 - Change "Rust" ABI on x86 to use SSE registers if they are always-available for the current target
📝 - Make
ambiguous_glob_imported_traitsfire in slightly more cases
📝 - Make
+fix-cortex-a53-835769a default feature foraarch64-unknown-fuchsia
📝 - Some impls on
Boxhave been generalised to arbitrary allocators which may cause some new type inference errors
📝 - On ARM targets,
-Ctarget-feature/-Ctarget-cpuvalues that would alter the ABI are rejected as hard errors now.
📝
Internal Changes
These changes do not affect any public interfaces of Rust, but they represent
significant improvements to the performance or internals of rustc and related
tools.
Other
Metadata
Metadata
Assignees
Labels
T-releaseRelevant to the release subteam, which will review and decide on the PR/issue.Relevant to the release subteam, which will review and decide on the PR/issue.relnotes-tracking-issueMarks issues tracking what text to put in release notes.Marks issues tracking what text to put in release notes.
Activity
BoxyUwU commentedon Oct 5, 2026
cc @rust-lang/release
@rustbot ping relnotes-interest-group
rustbot commentedon Oct 5, 2026
Hi relnotes-interest-group, this issue/PR could use some help in reviewing /
adjusting release notes. Could you take a look if available? Thanks <3
cc @alex-semenyuk @jieyouxu @joshtriplett @lcnr @traviscross @weihanglo
theemathas commentedon Oct 6, 2026
This belongs under "language", not "library". However... the link to the relnotes issue is missing, so I can't edit that.
theemathas commentedon Oct 6, 2026
I would link to https://doc.rust-lang.org/cargo/reference/lints.html, and also mention that users will now get warnings about unused dependencies, among other things.
I don't think this needs relnotes
This should also have a link to https://blog.rust-lang.org/2026/03/13/call-for-testing-build-dir-layout-v2/
weihanglo commentedon Oct 6, 2026
@theemathas agreed with you on all your points. For cargo we probably only need these things.
And prehaps the build dir layout as a compat note:
See also Cargo's changelog: https://doc.rust-lang.org/nightly/cargo/CHANGELOG.html#cargo-1100-2026-11-12
pthariensflame commentedon Oct 22, 2026
These entries all have the typo of missing the trait name
Allocatorin the position of the trait being implemented; we think maybe it was erroneously "deduplicated" with the occurrence(s) in the type parameter list beforehand. Corrected: