Allow more MIR SROA
This removes some guards on SROA that are no longer needed:
- With https://github.com/rust-lang/compiler-team/issues/838 it no longer needs to check for SIMD
- With https://github.com/rust-lang/compiler-team/issues/807 it no longer needs to check for niches
- This means that `Wrapper(char)` and `Pin<&mut T>` can get SRoA'd now, where previously they weren't because the check was banning SRaA for anything with a niche -- not just things with `#[rustc_layout_scalar_valid_range_*]`.
- Technically rust-lang/rust#133652 isn't complete yet, but `NonZero` and `NonNull` have already moved over, so this is fine. At worst this will mean that LLVM gets less `!range` metadata on something that wasn't already fixed by rust-lang/rust#133651 or rust-lang/rust#135236, but that's still sound, and unblocking general SRoA is worth that tradeoff.
Rollup of 7 pull requests
Successful merges:
- rust-lang/rust#144072 (update `Atomic*::from_ptr` and `Atomic*::as_ptr` docs)
- rust-lang/rust#144151 (`tests/ui/issues/`: The Issues Strike Back [1/N])
- rust-lang/rust#144300 (Clippy fixes for miropt-test-tools)
- rust-lang/rust#144399 (Add a ratchet for moving all standard library tests to separate packages)
- rust-lang/rust#144472 (str: Mark unstable `round_char_boundary` feature functions as const)
- rust-lang/rust#144503 (Various refactors to the codegen coordinator code (part 3))
- rust-lang/rust#144530 (coverage: Infer `instances_used` from `pgo_func_name_var_map`)
r? `@ghost`
`@rustbot` modify labels: rollup
coverage: Infer `instances_used` from `pgo_func_name_var_map`
In obscure circumstances involving macro-expanded spans, we would sometimes emit a covfun record for a function with no physical coverage counters, and therefore no corresponding entry in the “PGO names” section of the binary. The absence of that name entry causes `llvm-cov` to fail with the cryptic error message:
```text
malformed instrumentation profile data: function name is empty
```
We can eliminate this mismatch by removing `instances_used` entirely, and instead inferring its contents from the keys of `pgo_func_name_var_map`.
This makes it impossible for a "used" function to lack a PGO name entry.
---
This is an attempt to eliminate the cause of rust-lang/rust#141577 when re-landing changes like rust-lang/rust#144298 in the future.
I haven't been able to reproduce the underlying issue in an in-tree test, because the only known repro involves a non-trivial derive proc-macro that relies on `syn` and `proc-macro2`. But I have manually verified in a separate branch that this change would have prevented the reoccurrence of https://github.com/rust-lang/rust/issues/141577#issuecomment-3120667286.
Various refactors to the codegen coordinator code (part 3)
Continuing from https://github.com/rust-lang/rust/pull/144062 this removes an option without any known users, uses the object crate in favor of LLVM for getting the LTO bitcode and improves the coordinator channel handling.
str: Mark unstable `round_char_boundary` feature functions as const
Mark `floor_char_boundary`, `ceil_char_boundary` const
Simplify the implementations, reducing the number of arithmetic operations
It seems unnecessary to do the lower/upper bounds calculations and extra slicing when we can jump straight to inspecting the bytes, assuming the underlying data is valid UTF-8.
Tracking issue https://github.com/rust-lang/rust/issues/93743
`tests/ui/issues/`: The Issues Strike Back [1/N]
I believe I’ve finally brought [my program](https://github.com/Kivooeo/test-manager) to life -- it now handles multiple test moves in one go: plain moves first, then a gentle touch on each file depends on given options. The process should be much smoother now.
Of course, I won’t rush through everything in a few days -- that would be unkind to `@Oneirical.` I’ll pace myself. And also I can't have more than one such PR because `issues.txt` will conflict with previous parts after merging them which is not fun as well.
This PR is just that: first commit - moves; second - regression comments and the occasional .stderr reblesses, also issue.txt and tidy changes. Nothing special, but progress nonetheless. This is for the purpose of preserving test file history during restructuring
Part of https://github.com/rust-lang/rust/issues/133895.
r? `@jieyouxu`
rustdoc: save target modifiers
`rustdoc` was filling a `target_modifiers` variable, but it was not using the result.
In turn, that means that trying to use a dependency that set a target modifier fails.
For instance, running:
```sh
RUSTC_BOOTSTRAP=1 rustc --edition=2024 --target=aarch64-unknown-none-softfloat --sysroot=/dev/null --emit=metadata -Zfixed-x18 --crate-type rlib --crate-name core $(rustc --print sysroot)/lib/rustlib/src/rust/library/core/src/lib.rs
echo '#![allow(internal_features)]
' | RUSTC_BOOTSTRAP=1 rustdoc --edition=2021 --target=aarch64-unknown-none-softfloat --sysroot=/dev/null -Zfixed-x18 --extern core=libcore.rmeta -
```
will fail with:
```text
error: mixing `-Zfixed-x18` will cause an ABI mismatch in crate `rust_out`
|
= help: the `-Zfixed-x18` flag modifies the ABI so Rust crates compiled with different values of this flag cannot be used together safely
= note: unset `-Zfixed-x18` in this crate is incompatible with `-Zfixed-x18=` in dependency `core`
= help: set `-Zfixed-x18=` in this crate or unset `-Zfixed-x18` in `core`
= help: if you are sure this will not cause problems, you may use `-Cunsafe-allow-abi-mismatch=fixed-x18` to silence this error
```
Thus save the targets modifiers in `Options` to then pass it to the session options, so that eventually the diff can be performed as expected in `report_incompatible_target_modifiers()`.
Cc: ``@azhogin``
Fixes: https://github.com/rust-lang/rust/issues/144521
Remove `compiler-builtins-{no-asm,mangled-names}`
Remove `compiler-builtins-no-asm`
This feature used to be for when Cranelift didn't support inline
assembly, but its last uses were removed in 52933e0bd2 ("Don't disable
inline asm usage in compiler-builtins when the cranelift backend is
enabled"). and cba05a7a14 ("Support naked functions").
This doesn't remove the feature from the `compiler-builtins` crate, that
will be done separately in the subtree repo.
---
Remove `compiler-builtins-mangled-names`
This config was added in 207de019dc ("libary: Forward
compiler-builtins "asm" and "mangled-names" feature") but it does not
appear this has ever been used. The PR adding it (rust-lang/rust#78472) says that
this was exposed to help with configuration and points at the [Hermit
Cargo config], but as far as I can tell, this feature name has never
been mentioned in that repository's git history.
Thus, clean up a seemingly unneeded feature.
[Hermit Cargo config]: https://github.com/hermit-os/hermit-rs/blob/ab2b830930e6a9a98c8294997a8183feeabeda4a/.cargo/config
Port the proc macro attributes to the new attribute parsing infrastructure
Ports `#[proc_macro]`, `#[proc_macro_attribute]`, `#[proc_macro_derive]` and `#[rustc_builtin_macro]` to the new attribute parsing infrastructure for https://github.com/rust-lang/rust/issues/131229#issuecomment-2971351163
I've split this PR into commits for reviewability, and left some comments to clarify things
I did 4 related attributes in one PR because they share a lot of their code and logic, and doing them separately is kind of annoying as I need to leave both the old and new parsing in place then.
r? ``@oli-obk``
cc ``@jdonszelmann``
Don't special-case llvm.* as nounwind
Certain LLVM intrinsics, such as `llvm.wasm.throw`, can unwind. Marking them as nounwind causes us to skip cleanup of locals and optimize out `catch_unwind` under inlining or when `llvm.wasm.throw` is used directly by user code.
The motivation for forcibly marking llvm.* as nounwind is no longer present: most intrinsics are linked as `extern "C"` or other non-unwinding ABIs, so we won't codegen `invoke` for them anyway.
Closesrust-lang/rust#132416.
`@rustbot` label +T-compiler +A-panic
Avoid unnecessary `new_adt`/`new_fn_def` calls.
They can be skipped if there are no arguments, avoiding the "relate" operation work and also the subsequent interning.
r? `@ghost`
In obscure circumstances, we would sometimes emit a covfun record for a
function with no physical coverage counters, causing `llvm-cov` to fail with
the cryptic error message:
malformed instrumentation profile data: function name is empty
We can eliminate this mismatch by removing `instances_used` entirely, and
instead inferring its contents from the keys of `pgo_func_name_var_map`.
This makes it impossible for a "used" function to lack a PGO name entry.
Unify LLVM ctlz/cttz intrinsic generation
The type signature for `llvm.ctlz` is the same as the one for `llvm.cttz`, which enables a small simplification.
move uefi test to run-make
Turn the `uefi` test into a more standard `run-make` test, and execute it using the `test-various` CI job like before.
This is just a straightforward translation of the python code, but using `run-make` to supply the target (hence the 3 separate calls in the docker file).
r? ```@jieyouxu```
cc ```@nicholasbishop```
try-job: test-various
No longer need `alloca`s for consuming `Result<!, i32>` and similar
In optimized builds GVN gets rid of these already, but in `opt-level=0` we actually make `alloca`s for this, which particularly impacts `?`-style things that use actually-only-one-variant types like this.
While doing so, rewrite `LocalAnalyzer::process_place` to be non-recursive, solving a 6+ year old FIXME.
r? codegen
`rustdoc` was filling a `target_modifiers` variable, but it was not
using the result.
In turn, that means that trying to use a dependency that set a target
modifier fails.
For instance, running:
```sh
RUSTC_BOOTSTRAP=1 rustc --edition=2024 --target=aarch64-unknown-none-softfloat --sysroot=/dev/null --emit=metadata -Zfixed-x18 --crate-type rlib --crate-name core $(rustc --print sysroot)/lib/rustlib/src/rust/library/core/src/lib.rs
echo '#![allow(internal_features)]
' | RUSTC_BOOTSTRAP=1 rustdoc --edition=2021 --target=aarch64-unknown-none-softfloat --sysroot=/dev/null -Zfixed-x18 --extern core=libcore.rmeta -
```
will fail with:
```text
error: mixing `-Zfixed-x18` will cause an ABI mismatch in crate `rust_out`
|
= help: the `-Zfixed-x18` flag modifies the ABI so Rust crates compiled with different values of this flag cannot be used together safely
= note: unset `-Zfixed-x18` in this crate is incompatible with `-Zfixed-x18=` in dependency `core`
= help: set `-Zfixed-x18=` in this crate or unset `-Zfixed-x18` in `core`
= help: if you are sure this will not cause problems, you may use `-Cunsafe-allow-abi-mismatch=fixed-x18` to silence this error
```
Thus save the targets modifiers in `Options` to then pass it to the
session options, so that eventually the diff can be performed as expected
in `report_incompatible_target_modifiers()`.
Fixes: https://github.com/rust-lang/rust/issues/144521
Signed-off-by: Miguel Ojeda <ojeda@kernel.org>
canonicalize build root in `tests/run-make/linker-warning`
r? jieyouxu
This is similar to rust-lang/rust#139823 -- the test fails for me because my build dir is a symlink.
tests: aarch64-outline-atomics: Remove hardcoded target
Since this test is limited to aarch64 and linux hosts, the `--target` flag is entirely unnecessary and only breaks this on musl hosts. Let the compiler use the default target instead.
Enable outline-atomics for aarch64-unknown-linux-musl
They were disabled in bd287fa508 and haven't been causing problems for a while anymore, see https://github.com/rust-lang/rust/issues/89626 for details.
The entire testsuite still passes on `aarch64-unknown-linux-musl` with this feature enabled.
library/windows_targets: Fix macro expansion error in 'link' macro
A recent change altered the definition of the link! macro when the windows_raw_dylib feature is enabled, changing its syntax from pub macro {..} to pub macro($tt:tt) {..} in rust-lang/rust#143592
This change introduced a build failure with the error: "macros that expand to items must be delimited with braces or followed by a semicolon".
We add a semicolon to the line causing the issue as we also modify the non windows_raw_dylib link to make use of the link_dylib macro
Stop compilation early if macro expansion failed
Fixesrust-lang/rust#116180.
So there isn't really a type that is central for macro expansion and some errors are actually emitted (because the resolution happens after the expansion I suppose) after the expansion pass (like "not found macro"). Sometimes, errors are only emitted on the second "try" (to improve error output). So I couldn't reach a similar solution than what was done in https://github.com/rust-lang/rust/pull/133937 and suggested by ````@estebank```` in https://github.com/rust-lang/rust/issues/116180#issuecomment-3109468922. But maybe I missed something?
So in the end, I realized that there is method called every time (except one, described below) a macro error is actually emitted: `ExtCtxt::trace_macros_diag`. Considering I updated what it did, I renamed it into `macro_error_and_trace_macros_diag` to better reflect it.
There is only one call of `trace_macros_diag` which isn't reporting an error but just used for `macro_trace` feature, so I kept it as is.
r? ````@oli-obk````
binarycat