`ptr_eq` was recently enhanced to lint on more cases of raw pointers
comparison:
- lint on all raw pointer comparison, by proposing to use
`[core|std]::ptr::eq(lhs, rhs)` instead of `lhs == rhs`;
- removing one symetric `as usize` on each size if needed
- peeling any level of `as *[const|mut] _` if the remaining expression
can still be coerced into the original one (i.e., is a ref or raw
pointer to the same type as before)
The current change restricts the lint to the cases where at least one
level of symetric `as usize`, or any conversion to a raw pointer, could
be removed. For example, a direct comparaison of two raw pointers will
not trigger the lint anymore.
changelog: [`ptr_eq`]: do not lint when comparing two raw pointers
directly with no casts involved
Fixesrust-lang/rust-clippy#14525
It's a much better name, more consistent with how we name such things.
Also rename `Lifetime::res` as `Lifetime::kind` to match. I suspect this
field used to have the type `LifetimeRes` and then the type was changed
but the field name remained the same.
fixes#7749.
This issue proposes searching for `DerefMut` impls, which is not done
here: every lifetime parameter (aka `<'a>`) is the input types is
considered to be potentially mutable, and thus deactivates the lint.
changelog: [`mut_from_ref`]: Fixes false positive, where lifetimes
nested in the type (e.g. `Box<&'a mut T>`) were not considered.
`ptr_eq` was recently enhanced to lint on more cases of raw pointers
comparison:
- lint on all raw pointer comparison, by proposing to use
`[core|std]::ptr::eq(lhs, rhs)` instead of `lhs == rhs`;
- removing one symetric `as usize` on each size if needed
- peeling any level of `as *[const|mut] _` if the remaining expression
can still be coerced into the original one (i.e., is a ref or raw
pointer to the same type as before)
The current change restricts the lint to the cases where at least one
level of symetric `as usize`, or any conversion to a raw pointer, could
be removed. For example, a direct comparaison of two raw pointers will
not trigger the lint anymore.
We do not want to remove casts done inside macros. Also, when printing
the suggestion, take it from the same context as the origin expression
(the root context).
Problems found while working on #14526, but should be merged even if
#14526 is not.
changelog: none
We do not want to remove casts done inside macros. Also, when printing
the suggestion, take it from the same context as the origin expression
(the root context).
First of all, note that `Map` has three different relevant meanings.
- The `intravisit::Map` trait.
- The `map::Map` struct.
- The `NestedFilter::Map` associated type.
The `intravisit::Map` trait is impl'd twice.
- For `!`, where the methods are all unreachable.
- For `map::Map`, which gets HIR stuff from the `TyCtxt`.
As part of getting rid of `map::Map`, this commit changes `impl
intravisit::Map for map::Map` to `impl intravisit::Map for TyCtxt`. It's
fairly straightforward except various things are renamed, because the
existing names would no longer have made sense.
- `trait intravisit::Map` becomes `trait intravisit::HirTyCtxt`, so named
because it gets some HIR stuff from a `TyCtxt`.
- `NestedFilter::Map` assoc type becomes `NestedFilter::MaybeTyCtxt`,
because it's always `!` or `TyCtxt`.
- `Visitor::nested_visit_map` becomes `Visitor::maybe_tcx`.
I deliberately made the new trait and associated type names different to
avoid the old `type Map: Map` situation, which I found confusing. We now
have `type MaybeTyCtxt: HirTyCtxt`.
Most notably, this commit changes the `pub use crate::*;` in that file
to `use crate::*;`. This requires a lot of `use` items in other crates
to be adjusted, because everything defined within `rustc_span::*` was
also available via `rustc_span::source_map::*`, which is bizarre.
The commit also removes `SourceMap::span_to_relative_line_string`, which
is unused.
