Rollup of 5 pull requests
Successful merges:
- #75938 (Added some `min_const_generics` revisions into `const_generics` tests)
- #76050 (Remove unused function)
- #76075 (datastructures: replace `once_cell` crate with an impl from std)
- #76115 (Restore public visibility on some parsing functions for rustfmt)
- #76127 (rustbuild: Remove one LLD workaround)
Failed merges:
r? @ghost
Restore public visibility on some parsing functions for rustfmt
In #74826 the visibility of several parsing functions was reduced. However, rustfmt is an external consumer of some of these functions as well and needs the visibility to be public, similar to other elements in rustc_parse such as `parse_ident`
db534b3ac2/src/librustc_parse/parser/mod.rs (L433-L436)
Update MinGW instructions to include ninja
Rust now requires `ninja` to build, so the MinGW build instructions are updated to reflect this.
Like for `python` and `cmake`, the `mingw-w64-x86_64-ninja` package should be used. The default package from the `msys2` subsystem doesn't handle paths correctly on windows.
GH-66816: Remove disable attr before return
Passing --disable-per-crate-search removes the create search inputs so moved code around so that the search input is enabled
first before the function returns.
Fixes#66816
This commit changes `uncached_llvm_type` so that a named struct type
(with an empty name) is always created when the `fewer_names` option
is enabled. By skipping the generation of names, we can improve perf.
Giving `LLVMStructCreateNamed` an empty name works because LLVM will
perform random renames to avoid collisions.
Signed-off-by: David Wood <david@davidtw.co>
Use intra-doc links for `core/src/slice.mod.rs`
partial help in #75080
r? @jyn514
- most are using primitive types links, which cannot be used with intra links at the moment
- also `std` cannot be referenced in any link, `std::ptr::NonNull` and `std::slice` could not be referenced
ty: remove obsolete pretty printer
Fixes#61139.
This PR removes the obsolete printer and replaces all uses of it with `FmtPrinter`. Of the replaced uses, all but one use was in `debug!` logging, two cases were notable:
- `MonoItem::to_string` is used in `-Z print-mono-items` and therefore affects the output of all codegen-units tests (which have been updated).
- `DefPathBasedNames` was used in `librustc_codegen_llvm/type_of.rs` with `LLVMStructCreateNamed` and that'll now get different values, but nothing will break as a result of this.
cc @eddyb (whom I've discussed this with)
Add `-Z proc-macro-backtrace` to allow showing proc-macro panics
Fixes#75050
Previously, we would unconditionally suppress the panic hook during
proc-macro execution. This commit adds a new flag
`-Z proc-macro-backtrace`, which allows running the panic hook for
easier debugging.
Constify the following methods of `std::net::Ipv4Addr`:
- `octets`
- `is_loopback`
- `is_private`
- `is_link_local`
- `is_shared`
- `is_ietf_protocol_assignment`
- `is_benchmarking`
- `is_multicast`
- `is_documentation`
Also insta-stabilizes these methods as const.
Possible because of the stabilization of const integer arithmetic and control flow.
Fixes#75050
Previously, we would unconditionally suppress the panic hook during
proc-macro execution. This commit adds a new flag
-Z proc-macro-backtrace, which allows running the panic hook for
easier debugging.
Constify the following methods of `alloc::borrow::Cow`:
- `is_borrowed`
- `is_owned`
These methods are still unstable under `cow_is_borrowed`.
Possible because of #49146 (Allow if and match in constants).
Tracking issue: #65143
Make some Ordering methods const
Constify the following methods of `core::cmp::Ordering`:
- `reverse`
- `then`
Possible because of #49146 (Allow `if` and `match` in constants).
Tracking issue: #76113
Added the `mingw-w64-x86_64-ninja` package to the build guide for MinGW, as well as a note not to use the `ninja` package from the `msys2` subsystem (doesn't handle paths correctly on windows).
Improve error message when typo is made in format!
The expansion of the format! built-in macro is roughly done in two steps:
- the format expression is parsed, the arguments are parsed,
- the format expression is checked to be a string literal, code is expanded.
The problem is that the expression parser can eat too much tokens, which invalidates the parsing of the next format arguments. As the format expression check happens next, the error emitted concerns the format arguments, whereas the problem is about the format expression.
This PR contains two commits. The first one actually checks that the formatting expression is a string literal before raising any error about the formatting arguments, and the second one contains some simple heuristics which allow to suggest, when the format expression is followed by a dot instead of a comma, to suggest to replace the dot with a comma.
This pull request should fix#75492.
Note: this is my first non-doc contribution to the rust ecosystem. Feel free to make any comment about my code, or whatever. I'll be very happy to fix it :)
LLVM can't figure out in
let rem = self.len() % chunk_size;
let len = self.len() - rem;
let (fst, snd) = self.split_at(len);
and
let rem = self.len() % chunk_size;
let (fst, snd) = self.split_at(rem);
that the index passed to split_at() is smaller than the slice length and
adds a bounds check plus panic for it.
Apart from removing the overhead of the bounds check this also allows
LLVM to optimize code around the ChunksExact iterator better.
These are unsafe variants of the non-unchecked functions and don't do
any bounds checking.
For the time being these are not public and only a preparation for the
following commit. Making it public and stabilization can follow later
and be discussed in https://github.com/rust-lang/rust/issues/76014 .
Previous implementation used the `Parser::parse_expr` function in order
to extract the format expression. If the first comma following the
format expression was mistakenly replaced with a dot, then the next
format expression was eaten by the function, because it looked as a
syntactically valid expression, which resulted in incorrectly spanned
error messages.
The way the format expression is exctracted is changed: we first look at
the first available token in the first argument supplied to the
`format!` macro call. If it is a string literal, then it is promoted as
a format expression immediatly, otherwise we fall back to the original
`parse_expr`-related method.
This allows us to ensure that the parser won't consume too much tokens
when a typo is made.
A test has been created so that it is ensured that the issue is properly
fixed.
