Rework target checking for built-in attributes #145085
Conversation
rustbot
added
A-attributes
JonathanBrouwer
force-pushed
the
target_checking
branch
2 times, most recently
from
66feb37 to
a1c6f2a
Compare
- Added a few more variants which are needed for various attributes - Previously a trait method with default block had the same target representation as a method in a `impl trait for` block, this has been changed (See `MethodKind`) - Added both `singular_name` and `plural_name` for more precision on the form of the name
This comment has been minimized.
This comment has been minimized.
Every acceptor gets an `ALLOWED_TARGETS` specification which can specify per target whether it is allowed, warned, or errored.
JonathanBrouwer
force-pushed
the
target_checking
branch
from
a1c6f2a to
df39f8d
Compare
|
|
rustbot
added
S-waiting-on-review
|
These commits modify the If this was unintentional then you should revert the changes before this PR is merged. This PR modifies Some changes occurred in compiler/rustc_attr_parsing Some changes occurred in compiler/rustc_passes/src/check_attr.rs |
| @@ -64,27 +71,32 @@ pub(crate) struct ExportStableParser; | |||
| impl<S: Stage> NoArgsAttributeParser<S> for ExportStableParser { | |||
| const PATH: &[Symbol] = &[sym::export_stable]; | |||
| const ON_DUPLICATE: OnDuplicate<S> = OnDuplicate::Warn; | |||
| const ALLOWED_TARGETS: AllowedTargets = AllowedTargets::AllowAll; //FIXME Still checked fully in `check_attr.rs` | |||
There was a problem hiding this comment.
Some attributes were a little non-trivial to check, I will do these soon but I think let's get all the easy ones done first
| Allow(Target::AssocTy), | ||
| Allow(Target::AssocConst), | ||
| Allow(Target::Variant), | ||
| Allow(Target::Impl { of_trait: false }), //FIXME This does not make sense |
There was a problem hiding this comment.
We can make breaking changes in a future PR, for example to fix this.
| | | ||
| = note: `#[deny(useless_deprecated)]` on by default | ||
| = warning: this was previously accepted by the compiler but is being phased out; it will become a hard error in a future release! |
There was a problem hiding this comment.
I've made all attributes that were previously warnings a future error. I can try to replicate the previous behaviour if you wish but I think this should be fine.
| --> $DIR/issue-31769.rs:2:5 | ||
| | | ||
| LL | #[inline] struct Foo; | ||
| | ^^^^^^^^^ |
There was a problem hiding this comment.
Hmmm I've just noticed this is super ugly with the help getting its own codespan, let me see if I can fix that
| @@ -57,7 +57,7 @@ LL | #[lang = "fn"] | |||
| | ^^^^^^^^^^^^^^ | |||
| ... | |||
| LL | trait Fn { | |||
| | - this trait has 0 generic arguments | |||
| | - this a trait has 0 generic arguments | |||
There was a problem hiding this comment.
Whoops, I'll have to fix that
|
The job Click to see the possible cause of the failure (guessed by this bot) |
This is a refactoring of target checking for built-in attributes.
This PR has the following goals:
#[repr]will be in a future PR. Tho I have written the code in such a way that this will be possible to add in the future.I've split it in commits and left a description on some of the commits to help review.
r? @jdonszelmann