Rollup of 7 pull requests
Successful merges:
- #60287 (Use references for variances_of)
- #60327 (Search for incompatible universes in borrow errors)
- #60330 (Suggest using an inclusive range instead of an exclusive range when the endpoint overflows by 1)
- #60366 (build-gcc: Create missing cc symlink)
- #60369 (Support ZSTs in DispatchFromDyn)
- #60404 (Implement `BorrowMut<str>` for `String`)
- #60417 (Rename hir::ExprKind::Use to ::DropTemps and improve docs.)
Failed merges:
r? @ghost
Extend the std::process::Child docs with warning about possible zombies.
The previous version mentioned that when dropping the Child, the
process is not killed. However, the wording gave the impression that
such behaviour is fine to do (leaving the reader believe low-level
details like reaping zombies of the dead processes is taken over by std
somehow; or simply leaving the reader unaware about the problem).
This commit gensyms the generated ident for replacement arguments so
that users cannot refer to them. It also ensures that levenshtein
distance suggestions do not suggest gensymed identifiers.
This commit modifies the lowering of `async fn` arguments so that the
drop order matches the equivalent `fn`.
Previously, async function arguments were lowered as shown below:
async fn foo(<pattern>: <ty>) {
async move {
}
} // <-- dropped as you "exit" the fn
// ...becomes...
fn foo(__arg0: <ty>) {
async move {
let <pattern> = __arg0;
} // <-- dropped as you "exit" the async block
}
After this PR, async function arguments will be lowered as:
async fn foo(<pattern>: <ty>, <pattern>: <ty>, <pattern>: <ty>) {
async move {
}
} // <-- dropped as you "exit" the fn
// ...becomes...
fn foo(__arg0: <ty>, __arg1: <ty>, __arg2: <ty>) {
async move {
let __arg2 = __arg2;
let <pattern> = __arg2;
let __arg1 = __arg1;
let <pattern> = __arg1;
let __arg0 = __arg0;
let <pattern> = __arg0;
} // <-- dropped as you "exit" the async block
}
If `<pattern>` is a simple ident, then it is lowered to a single
`let <pattern> = <pattern>;` statement as an optimization.
rustdoc: remove def_ctor hack.
~~No longer necessary since we have `describe_def`.~~
Turns out `def_ctor` was used in conjunction with abusing `tcx.type_of(def_id)` working on both type definitions and `impl`s (specifically, of builtin types), but also reimplementing a lot of the logic that `Clean` already provides on `Ty` / `ty::TraitRef`.
The first commit now does the minimal refactor to keep it working, while the second commit contains the rest of the refactor I started (parts of which I'm not sure we need to keep).
The commit moves metadata writing from `link_binary` to
`encode_metadata` (and renames the latter as
`encode_and_write_metadata`). This is at the very start of code
generation.
Support ZSTs in DispatchFromDyn
Allows to use ZSTs with 1 byte alignment in `DispatchFromDyn` implementation. This is required for `Box<T, A: Alloc>`
cc #58457
build-gcc: Create missing cc symlink
This PR mostly fixes build error caused by using rustc docker images
to build [rustup][1] (with the error: ``linker `cc` not found``).
I don't know why gcc build script doesn't install cc hard link by default.
In build log, with `make SHELL='sh -x' install`, gcc build script installs c++
hard link but not `cc` one:
```bash
if test "" != "yes" ; then \
rm -f /rustroot/bin/g++; \
/usr/bin/install -c xg++ /rustroot/bin/g++; \
chmod a+x /rustroot/bin/g++; \
rm -f /rustroot/bin/c++; \
( cd /rustroot/bin && \
ln g++ c++ ); \
if [ -f cc1plus ] ; then \
if [ ! -f g++-cross ] ; then \
rm -f /rustroot/bin/x86_64-unknown-linux-gnu-g++; \
( cd /rustroot/bin && \
ln g++ x86_64-unknown-linux-gnu-g++ ); \
rm -f /rustroot/bin/x86_64-unknown-linux-gnu-c++; \
( cd /rustroot/bin && \
ln c++ x86_64-unknown-linux-gnu-c++ ); \
fi ; \
fi; \
fi
```
This might be fixed downstream by manually creating cc hard link
or setting `RUSTFLAGS="-C linker=gcc"` as [suggested by @mati865][2].
But I find it better to fix it upstream in this PR.
[1]: https://github.com/rust-lang/rustup.rs/pull/1815
[2]: https://github.com/rust-lang/rustup.rs/pull/1815#discussion_r279192797
Search for incompatible universes in borrow errors
If we have a borrow that has to live for `'static` we need to check for
any regions in incompatible universes when trying to find the cause.
closes#60274
Debug-print error when using rtunwrap
When I added this macro a while back I didn't have a way to make it print the failure for all types that you might want to unwrap. Now, I came up with a solution.
PGO: Add a run-make test that makes sure that PGO profiling data is used by the compiler during optimizations.
From the tests comment section:
```
# This test makes sure that PGO profiling data leads to cold functions being
# marked as `cold` and hot functions with `inlinehint`.
# The test program contains an `if` were actual execution only ever takes the
# `else` branch. Accordingly, we expect the function that is never called to
# be marked as cold.
```
r? @alexcrichton
Revert "Update Source Code Pro fonts to version 2.030"
Temporary fix for https://github.com/rust-lang/rust/issues/60365https://github.com/rust-lang/rust/pull/60146 updated all fonts (not just the fonts that were buggy), however it looks like the new Source Code Pro is buggy.
We should test this out of tree: `cargo doc` on anything, replace the font in `target/doc`, and use `python -m SimpleHTTPServer` to see what's going on.
Till we figure that out, I'll just back this out.
r? @GuillaumeGomez
This reverts commit 6bafc58ced.
Don't try to render auto-trait bounds with any inference variables
Previously, we checked if the target of a projection type was itself an
inference variable. However, for Rustdoc rendering purposes, there's no
distinction between an inference variable ('_') and a type containing
one (e.g. (MyStruct<u8, _>)) - we don't want to render either of them.
Fixes#60269
Due to the complexity of the original bug, which spans three different
crates (hyper, tower-hyper, and tower), I have been unable to create a
minimized reproduction for the issue.