Remove `default_hidden_visibility: false` from wasm targets
To the best of my ability I believe that this is no longer necessary. I don't fully recall why this was first added but I believe it had to do with symbols all being exported by default and this was required to undo that. Regardless nowadays the default output of rustc seems suitable so it seems best to keep wasm in line with other targets.
Prefer identity equality over equating types during coercion.
These types are always generic only over their own generic parameters with no inference variables involved.
r? `@compiler-errors`
I love touching code that [hasn't changed meaningfully since 2016](https://github.com/rust-lang/rust/pull/41937)
Box::into_raw: make Miri understand that this is a box-to-raw cast
Turns out https://github.com/rust-lang/rust/pull/122647 went a bit too far in cleaning up `Box`... we still need a hack in `Box::into_raw`. The nicer fix would be to make Stacked Borrows not care about reference-to-raw-pointer casts, but it's unclear whether that will ever be possible without going to full Tree Borrows.
Fixes https://github.com/rust-lang/miri/issues/3473.
Use `/* value */` as a placeholder
The expression `value` isn't a valid suggestion; let's use `/* value */` as a placeholder (which is also invalid) since it more clearly signals to the user that they need to fill it in with something meaningful. This parallels the suggestions we have in a couple other places, like arguments.
We could also print the type name instead of `/* value */`, especially if it's suggestable, but I don't care strongly about that.
Port the 2 `rust-lld` run-make tests to `rmake`
In preparation for finalizing most of the `rust-lld` work, this PR ports the following tests to `rmake`:
- `tests/run-make/rust-lld`
- `tests/run-make/rust-lld-custom-target`
As they use `$(CGREP) -e` I added `regex` as an exported dependency to the `run_make_support` library.
Unfortunately, the most recent versions depend on `memchr` 2.6.0 but it's currently pinned at 2.5.0 in the workspace, and therefore had to settle for the older `regex-1.8.0`.
r? `@jieyouxu`
Taint const qualifs if a static is referenced that didn't pass wfcheck
It is correct to only check the signature here, as the ICE is caused by `USE_WITH_ERROR` trying to allocate memory to store the result of `WITH_ERROR` before evaluating it.
fixes#123153
Rollup of 7 pull requests
Successful merges:
- #122811 (Move `SourceMap` initialization)
- #123512 (Match ergonomics 2024: Implement eat-one-layer)
- #123811 (Use queue-based `RwLock` on more platforms)
- #123859 (Remove uneeded clones now that TrustedStep implies Copy)
- #123979 (Subtype predicates only exist on inference types, so we can allow them to register opaque types within them.)
- #124016 (Outline default query and hook provider function implementations)
- #124023 (Allow workproducts without object files.)
r? `@ghost`
`@rustbot` modify labels: rollup
To the best of my ability I believe that this is no longer necessary. I
don't fully recall why this was first added but I believe it had to do
with symbols all being exported by default and this was required to undo
that. Regardless nowadays the default output of rustc seems suitable so
it seems best to keep wasm in line with other targets.
Allow workproducts without object files.
This pull request partially reverts changes from e16c3b4a44
Original motivation for this assert was described with "A WorkProduct without a saved file is useless"
which was true at the time but now it is possible to have work products with other types of files
(llvm-ir, asm, etc) and there are bugreports for this failure:
For example: https://github.com/rust-lang/rust/issues/123695
Fixes https://github.com/rust-lang/rust/issues/123234
Now existing `assert` and `.unwrap_or_else` are unified into a single
check that emits slightly more user friendly error message if an object
files was meant to be produced but it's missing
Outline default query and hook provider function implementations
The default query and hook provider functions call `bug!` with a decently long message.
Due to argument inlining in `format_args!` ([`flatten_format_args`](https://github.com/rust-lang/rust/issues/78356)), this ends up duplicating the message for each query, adding ~90KB to `librustc_driver.so` of unreachable panic messages.
To avoid this, we can outline the common `bug!` logic.
Subtype predicates only exist on inference types, so we can allow them to register opaque types within them.
We were unable to come up with an example where this could be reached (subtype predicates with either side not being an infer var gets consumed during any `select_where_possible` invocation, of which we have a lot in typeck). To ensure we don't silently accept new behaviour in case we missed a situation where this could occur, I have added an assert that prevents opaque types from having their hidden type constrained.
r? `@compiler-errors`
Use queue-based `RwLock` on more platforms
This switches over Windows 7, SGX and Xous to the queue-based `RwLock` implementation added in #110211, thereby fixing #121949 for Windows 7 and partially resolving #114581 on SGX. TEEOS can't currently be switched because it doesn't have a good thread parking implementation.
CC `@roblabla` `@raoulstrackx` `@xobs` Could you help me test this, please?
r? `@ChrisDenton` the Windows stuff should be familiar to you
ide: improve ReferenceCategoryType
It is bitset semantically --- many categorical things can be true about a reference at the same time.
In parciular, a reference can be a "test" and a "write" at the same time.
Previously this command was linting compiler and library together.
As we no longer run clippy on the entire tree unless it's explicitly
requested, we need to update this command by adding `library` path.
Signed-off-by: onur-ozkan <work@onurozkan.dev>
It is bitset semantically --- many categorical things can be true about
a reference at the same time.
In parciular, a reference can be a "test" and a "write" at the same
time.
