Break out bucket 0 (containing `idx < 4096`) as an early return, which
simplifies the remainder of the function, and allows optimizing the
`checked_ilog2` since it can no longer return `None`.
This reduces the runtime of `vec_cache::tests::slot_index_exhaustive`
(which calls `SlotIndex::from_index` for every `u32`, twice) from ~15.5s
to ~13.3s.
Before this change we had two different ways to attempt to locate the
sysroot which are inconsistently used:
* get_or_default_sysroot which tries to locate based on the 0th cli
argument and if that doesn't work falls back to locating it using the
librustc_driver.so location and returns a single path.,
* sysroot_candidates which takes the former and additionally does
another attempt at locating using librustc_driver.so except without
linux multiarch handling and then returns both paths.,
The latter was originally introduced to be able to locate the codegen
backend back when cg_llvm was dynamically linked even for a custom
driver when the --sysroot passed in does not contain a copy of cg_llvm.
Back then get_or_default_sysroot did not attempt to locate the sysroot
based on the location of librustc_driver.so yet. Because that is now
done, the only case where removing sysroot_candidates can break things
is if you have a custom driver inside what looks like a sysroot
including the lib/rustlib directory, but which is missing some parts of
the full sysroot like eg rust-lld.
#[used] currently is an alias for #[used(linker)] on all platforms
except ELF based ones where it is an alias for #[used(compiler)]. The
latter has surprising behavior and the LLVM LangRef explicitly states
that it "should only be used in rare circumstances, and should not be
exposed to source languages."
The reason #[used] still was an alias to #[used(compiler)] on ELF is
because the gold linker has issues with it. Luckily gold has been
deprecated with GCC 15 and seems to be unable to bootstrap rustc anyway.
As such we shouldn't really care about supporting gold.
Move placeholder handling to a proper preprocessing step
This commit breaks out the logic of placheolder rewriting into its own preprocessing step. It's one of the more boring
parts of #130227.
The only functional change from this is that the preprocessing step (where extra `r: 'static` constraints are added) is performed upstream of Polonius legacy, finally affecting Polonius. That is mostly a by-product, though.
This should be reviewable by anyone in the compiler team, so
r? rust-lang/compiler
Rollup of 11 pull requests
Successful merges:
- rust-lang/rust#141890 (Add link to correct documentation in htmldocck.py)
- rust-lang/rust#141932 (Fix for async drop inside async gen fn)
- rust-lang/rust#141960 (Use non-2015 edition paths in tests that do not test for their resolution)
- rust-lang/rust#141968 (Run wfcheck in one big loop instead of per module)
- rust-lang/rust#141969 (Triagebot: Remove `assign.users_on_vacation`)
- rust-lang/rust#141985 (Ensure query keys are printed with reduced queries)
- rust-lang/rust#141999 (Visit the ident in `PreciseCapturingNonLifetimeArg`.)
- rust-lang/rust#142005 (Change `tag_field` to `FieldIdx` in `Variants::Multiple`)
- rust-lang/rust#142017 (Fix incorrect use of "recommend" over "recommended")
- rust-lang/rust#142024 (Don't refer to 'this tail expression' in expansion.)
- rust-lang/rust#142025 (Don't refer to 'local binding' in extern macro.)
r? `@ghost`
`@rustbot` modify labels: rollup
This solve a stack overflow found on Fedora s390x when building
`tests/ui/parser/survive-peano-lesson-queue.rs`. Note that the singular
`mirror_expr` method already has this stack check, but in this case the
plural method was the one recursing too deeply.
```
error[E0747]: type provided when a constant was expected
--> $DIR/invalid-const-arguments.rs:10:19
|
LL | impl<N> Foo for B<N> {}
| ^
|
help: consider changing this type parameter to a const parameter
|
LL - impl<N> Foo for B<N> {}
LL + impl<const N: u8> Foo for B<N> {}
|
```
x86 (32/64): go back to passing SIMD vectors by-ptr
Fixes https://github.com/rust-lang/rust/issues/139029 by partially reverting https://github.com/rust-lang/rust/pull/135408 and going back to passing SIMD vectors by-ptr on x86. Sadly, by-val confuses the LLVM inliner so much that it's not worth it...
Also fixes https://github.com/rust-lang/rust/issues/141848 by no longer actually using vector registers with the "Rust" ABI.
r? `@tgross35`
Cc `@nikic`
try-job: `test-various*`
try-job: dist-i586-gnu-i586-i686-musl
try-job: x86_64-gnu-nopt
try-job: `x86_64-msvc*`
try-job: `i686-msvc*`
When a method is not present because of a trait bound not being met, and that trait bound is on a tuple, we check if making the tuple have no borrowed types makes the method to be found and highlight it if it does. This is a common problem for Bevy in particular and ORMs in general.
Don't refer to 'local binding' in extern macro.
When it comes from a macro expansion, the user has no clue what 'local binding' the compiler is talking about, if they don't know the expansion of the macro. Better to just say 'temporary value'.
Don't refer to 'this tail expression' in expansion.
The user has no clue what the compiler is talking about when it says "this tail expression". It is an implementation detail of the macro that it uses a block with tail expression.
Change `tag_field` to `FieldIdx` in `Variants::Multiple`
It was already available as a generic parameter anyway, and it's not like we'll ever put a tag in the 5-billionth field.
This is a first part of pulling smaller pieces out of rust-lang/rust#138759, so
r? workingjubilee
Ensure query keys are printed with reduced queries
Using `-Z query-dep-graph` and debug assertions leads to an ICE that was originally discovered in rust-lang/rust#141700:
> This isn't an incremental bug per se, but instead a bug that has to do with debug printing query keys when debug assertions and `-Z query-dep-graph` is enabled. We end up printing a const (b/c we're using generic const args here) whose debug printing for -Z query-dep-graph requires invoking the same query cyclically 😃
>
> I've pushed a commit which should fix this.
This isn't related to the standard library changes, but instead b/c it seems to be the first usage of `feature(adt_const_params)` in the standard library that ends up being triggered in incremental tests.
r? oli-obk
Run wfcheck in one big loop instead of per module
Maybe we can merge this big loop in the future with the `par_hir_body_owners` call below and run typeck only on items that didn't fail wfcheck. For now let's just see if perf likes it, as it by itself should be beneficial to parallel rustc
Fix for async drop inside async gen fn
Return value (for yield) is corrected for async drop inside async gen function.
In CFG, when internal async drop future is polled and returned `Poll<()>::Pending`, then async gen resume function returns `Poll<(OptRet)>::Pending`.
Fixesrust-lang/rust#140530
Fix incorrect eq_unspanned in TokenStream
Fixesrust-lang/rust#141522
r? ``@workingjubilee``
should we remove this function?
since it's used in several places, i'd prefer to keep it.
The user has no clue what tail expression the compiler is talking
about: it is an implementation detail of the macro that it uses a block
with tail expression.
Rollup of 8 pull requests
Successful merges:
- rust-lang/rust#136687 (Improve the documentation of `Display` and `FromStr`, and their interactions)
- rust-lang/rust#137306 (Remove `i128` and `u128` from `improper_ctypes_definitions`)
- rust-lang/rust#138699 (build dist for x86_64-pc-solaris and sparcv9-sun-solaris)
- rust-lang/rust#141250 (add s390x z17 target features)
- rust-lang/rust#141467 (make `OsString::new` and `PathBuf::new` unstably const)
- rust-lang/rust#141871 (index: add method for checking range on DenseBitSet)
- rust-lang/rust#141888 (Use non-2015 edition paths in tests that do not test for their resolution)
- rust-lang/rust#142000 (bootstrap: don't symlink source dir into stage0 sysroot)
r? `@ghost`
`@rustbot` modify labels: rollup
