| Age | Commit message (Collapse) | Author | Lines |
|---|
|
RFC-2841: add codegen flag export symbols from executable
Closes #84161
r? `@nikomatsakis` `@Mark-Simulacrum`
|
|
Rollup of 7 pull requests
Successful merges:
- #95040 (protect `std::io::Take::limit` from overflow in `read`)
- #95916 (kmc-solid: Use `libc::abort` to abort a program)
- #99494 (Use non-relocatable code in nofile-limit.rs test)
- #99581 (Improve error messages involving `derive` and `packed`.)
- #99643 (Add `sign-ext` target feature to the WASM target)
- #99659 (Use `VecMap::get` in `ConstraintLocator::check`)
- #99690 (add miri-track-caller to more intrinsic-exposing methods)
Failed merges:
r? `@ghost`
`@rustbot` modify labels: rollup
|
|
add miri-track-caller to more intrinsic-exposing methods
Follow-up to https://github.com/rust-lang/rust/pull/98674: I went through the Miri test suite to find more functions that would benefit from Miri backtrace pruning, and this is what I found.
Basically anything that just exposes a potentially-UB intrinsic to the user should get this treatment.
|
|
Use `VecMap::get` in `ConstraintLocator::check`
Also rename the `def_id` param to `item_def_id` because that's easily confused with `self.def_id` (which is the opaque ty did).
|
|
Add `sign-ext` target feature to the WASM target
Some target features are still missing from that list.
See #97808 for basically the same PR by `@alexcrichton.`
Related issue: #96472.
PR introducing this issue: #87402.
|
|
Improve error messages involving `derive` and `packed`.
There are two errors involving `derive` and `packed`.
```
`#[derive]` can't be derived on a `#[repr(packed)]` struct with type or const parameters
`#[derive]` can't be derived on a `#[repr(packed)]` struct that does not derive Copy
```
The second one overstates things. It is possible to use derive on a
repr(packed) struct that doesn't derive Copy in two cases.
- If all the fields within the struct meet the required alignment: 1 for
`repr(packed)`, or `N` for `repr(packed(N))`.
- If `Default` is the only trait derived.
This commit improves things in a few ways.
- Changes the errors to say `this trait can't be derived on this ...`.
This is more accurate, because it's just *this* trait and *this*
packed struct that are a problem, not *all* derived traits on *all*
packed structs.
- Adds more details to the "ERROR" lines in the test case, enough to
distinguish between the two error messages.
- Adds more cases to the test case that don't cause errors, e.g. `Default`
derives.
- Uses a wider variety of builtin traits in the test case, for better coverage.
r? `@estebank`
|
|
Use non-relocatable code in nofile-limit.rs test
To avoid using static-pie which is not essential to the test but which
was reported to cause problems on Void Linux where glibc is build
without support for static-pie.
|
|
kmc-solid: Use `libc::abort` to abort a program
This PR updates the target-specific abort subroutine for the [`*-kmc-solid_*`](https://doc.rust-lang.org/nightly/rustc/platform-support/kmc-solid.html) Tier 3 targets.
The current implementation uses a `hlt` instruction, which is the most direct way to notify a connected debugger but is not the most flexible way. This PR changes it to call the `abort` libc function, making it possible for a system designer to override its behavior as they see fit.
|
|
protect `std::io::Take::limit` from overflow in `read`
Resolves #94981
|
|
Tweak `SubstFolder` implementation
|
|
rmeta: avoid embedding `StabilityLevel::Unstable` reason multiple times into .rmeta\.rlib files
Avoids bloating size of some rmeta\rlib files by not placing default string for `StabilityLevel::Unstable` reason multiple times, affects only stdlib\rustc artifacts. For stdlib cuts about 3% (diff of total size for patched\unpatched *.rmeta files of stage1-std) of file size, depending on crates.
fixes #88180
|
|
|
|
Don't overwrite CodegenFnAttrs when adding used compiler/linker
Pointed out by `@tmiasko` in https://github.com/rust-lang/rust/pull/93718#discussion_r928259163. Probably a better fix than https://github.com/rust-lang/rust/pull/99674 for https://github.com/rust-lang/rust/issues/99596, assuming it does fix the latter.
I don't have the ability to write a regression test for this at the moment, and am quite surprised this didn't cause other failures. The old code is clearly wrong though, but I'm not picky as to which of these is better to go with.
|
|
There are two errors involving `derive` and `packed`.
```
`#[derive]` can't be derived on a `#[repr(packed)]` struct with type or const parameters
`#[derive]` can't be derived on a `#[repr(packed)]` struct that does not derive Copy
```
The second one overstates things. It is possible to use derive on a
repr(packed) struct that doesn't derive Copy in two cases.
- If all the fields within the struct meet the required alignment: 1 for
`repr(packed)`, or `N` for `repr(packed(N))`.
- If `Default` is the only trait derived.
This commit improves things in a few ways.
- Changes the errors to say `$TRAIT can't be derived on this ...`.
This is more accurate, because it's just $TRAIT and *this* packed
struct that are a problem, not *all* derived traits on *all* packed
structs.
- Adds more details to the "ERROR" lines in the test case, enough to
distinguish between the two error messages.
- Adds more cases to the test case that don't cause errors, e.g. `Default`
derives.
- Uses a wider variety of builtin traits in the test case, for better coverage.
|
|
AngelicosPhosphoros:improve_calloc_check_in_vec_macro_for_tuples, r=Mark-Simulacrum
Support vec zero-alloc optimization for tuples and byte arrays
* Implement IsZero trait for tuples up to 8 IsZero elements;
* Implement IsZero for u8/i8, leading to implementation of it for arrays of them too;
* Add more codegen tests for this optimization.
* Lower size of array for IsZero trait because it fails to inline checks
|
|
Convert rust-analyzer to an in-tree tool
This re-adds `rust-lang/rust-analyzer` as a git subtree rather than a submodule.
Closes https://github.com/rust-lang/rust-analyzer/issues/12815.
Prior attempt (research PR): https://github.com/rust-lang/rust/pull/99465
* [x] Remove submodule: `git rm -f src/tools/rust-analyzer`
* [x] Add subtree: `git subtree add -P src/tools/rust-analyzer https://github.com/rust-lang/rust-analyzer.git master`
* [x] Move to `SourceType::InTree`,
* [x] Enable `rust-analyzer/in-rust-tree` feature when built through `x.py`
* [x] Add 'check' step
* [x] Add 'test' step
With this PR, rust-analyzer becomes an "in-tree" tool. Syncs can happen in both directions, see [clippy's relevant book section](https://doc.rust-lang.org/nightly/clippy/development/infrastructure/sync.html).
Making sure `proc-macro-srv` doesn't break when the proc_macro bridge changes effectively becomes the responsibility of `rust-lang/rust` contributors. These days, that's mostly `@mystor,` who has been consulted throughout the process. I'm also making myself available in case there's questions / work needed that nobody else signed up for.
This doesn't change rust-analyzer's release cycle. After this PR is merged and the next nightly goes out, one can point `rust-analyzer.procMacro.server` to the rustup-provided `rust-analyzer` binary. Changes to improve the situation further (auto-discovery/install of the rust-analyzer component) will happen in `rust-lang/rust-analyzer` and be synced here eventually.
|
|
* Implement IsZero trait for tuples up to 8 IsZero elements;
* Implement IsZero for u8/i8, leading to implementation of it for arrays of them too;
* Add more codegen tests for this optimization.
* Lower size of array for IsZero trait because it fails to inline checks
|
|
@jyn514, and @lnicola)
|
|
Rollup of 4 pull requests
Successful merges:
- #99644 (remove some provenance-related machine hooks that Miri no longer needs)
- #99657 (Docs - remove unnecessary `mut` that gives a warning)
- #99672 (Remove Clean trait implementation for more items)
- #99678 (Update doc comments that refer to config parameter)
Failed merges:
r? `@ghost`
`@rustbot` modify labels: rollup
|
|
|
|
Update doc comments that refer to config parameter
This commit updates the `source_file_to_parser` and the
`maybe_source_file_to_parse` function's doc comments which currently
refer to a `config` parameter. The doc comments have been updated to
refer to the `session` parameter similar to the doc comment for
`try_file_to_source_file`, which also takes a `&Session` parameter.
|
|
Remove Clean trait implementation for more items
Follow up of #99638.
cc `@camelid`
r? `@notriddle`
|
|
Docs - remove unnecessary `mut` that gives a warning
Fixes #99654.
A trivial linting fix for a Stdio example.
`@rustbot` label +T-lib
|
|
remove some provenance-related machine hooks that Miri no longer needs
Then we can make `scalar_to_ptr` a method on `Scalar`. :)
Fixes https://github.com/rust-lang/miri/issues/2188
r? `@oli-obk`
|
|
Test codegen of atomic compare-exchange with additional memory orderings
* Add a test for atomic operations introduced in #97423 & #98383.
* Add a test for fallback code generation strategy used on LLVM 12 introduced in #98385. Use a separate test case instead of a revision system since test will be gone once LLVM 12 is no longer supported.
|
|
|
|
This commit updates the source_file_to_parser and the
maybe_source_file_to_parse function's doc comments which currently
refer to a config parameter. The doc comments have been updated to
refer to the 'session' parameter similar to the doc comment for
try_file_to_source_file, which also takes a &Session parameter.
|
|
|
|
|
|
order isn't guaranteed
(And, in fact, it failed on i686)
|
|
Revert "Mark atomics as unsupported on thumbv6m"
This is a breaking change for the `thumbv6m` target. See #99668 for discussion on how we can proceed forward from here.
This reverts commit 75146102197c7f35e6c38cb402b2bc1065858c54.
cc `@nikic`
|
|
|
|
|
|
|
|
This reverts commit 75146102197c7f35e6c38cb402b2bc1065858c54.
|
|
|
|
r=oli-obk
Allow `trait A: ~const B`
What's included: a minimal working change set for `~const` supertraits to work.
r? `@oli-obk`
|
|
- use `path` instead of `paths`
- don't mark rust-analyzer as an optional tool
- print the cargo command that's run in the proc-macro-test build script
this originally was part of a change to fix `test --stage 0 rust-analyzer`,
but I'm going to leave that for a separate PR so it's easier to review.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Co-authored-by: Joshua Nelson <github@jyn.dev>
|
|
|
|
|
|
|
|
|
|
'977e12a0bdc3e329af179ef3a9d466af9eb613bb'
git-subtree-dir: src/tools/rust-analyzer
git-subtree-mainline: 3c98486a0cdb6d92f0fca34ffb1fd46c0e498653
git-subtree-split: 977e12a0bdc3e329af179ef3a9d466af9eb613bb
|
