Revert #131669 due to ICEs
Revert [lint: change help for pointers to dyn types in FFI #131669](https://github.com/rust-lang/rust/pull/131669) due to ICE reports:
- <https://github.com/rust-lang/rust/issues/134059> (real-world)
- <https://github.com/rust-lang/rust/issues/134060> (fuzzing)
Closes#134060.
The revert criteria I used to assess whether to post this revert was:
1. It's not trivial to fix-forward. (1) The implementation itself is tricky due to `tcx.is_sized` query not being very trivial. (2) It will need more extensive test coverage for different ty kinds.
2. It is impacting real-world crates, i.e. #134059.
3. `improper_ctypes_definitions` is a warn-by-default lint.
This revert is without prejudice to relanding the changes. The changes can be re-landed with those cases addressed and stronger test coverage.
A rough regression test corresponding to the fuzzed example reported in #134060 is added to check that the revert worked, it is not sufficient for the lint test coverage when the lint improvements are to be relanded. Please feel free to improve the test in the reland.
r? `@workingjubilee` (or compiler)
cc `@niacdoial` (PR author)
The part about zero-sized structures is totally wrong. The rest of
it has almost no explanatory value; there are better explanations in
comments elsewhere.
Introduce `MixedBitSet`
`ChunkedBitSet` is good at avoiding excessive memory usage for programs with very large functgions where dataflow bitsets have very large domain sizes. But it's overly heavyweight for small bitsets, because any non-empty `ChunkedBitSet` takes up at least 256 bytes.
This PR introduces `MixedBitSet`, which is a simple bitset that uses `BitSet` for small/medium bitsets and `ChunkedBitSet` for large bitsets. It's a speed and memory usage win.
r? `@Mark-Simulacrum`
Remove ignored tests for hangs w/ new solver
As asked on zulip [here](https://rust-lang.zulipchat.com/#narrow/channel/364551-t-types.2Ftrait-system-refactor/topic/needs_help.3A.20look.20through.20compare-mode.20hangs).
As far as I can tell there are no more UI tests that hang anymore, so this removes the ignore directives for the compare mode.
(As I was using `--compare-mode new-solver` and that failed in an obscure way without any info about what to do, I've also fixed its error handling in `compiletest`: it didn't show the invalid `--compare-mode`, nor the valid values one can pass).
r? lcnr
Fix some triagebot mentions paths
This fixes some mentions paths in `triagebot.toml` that are no longer valid.
* rustdoc themes were merged into rustdoc.css (which already has a mention, and unfortunately can't independently mention the ayu theme). https://github.com/rust-lang/rust/pull/115829 cc `@GuillaumeGomez` `@Cldfire`
* The entry for `inspect_obligations.rs` was never correct, from https://github.com/rust-lang/rust/pull/122385. cc `@lcnr` `@compiler-errors`
* The entry for `need_type_info.rs` was moved in https://github.com/rust-lang/rust/pull/127501
A bunch of cleanups
These are all extracted from a branch I have to get rid of driver queries. Most of the commits are not directly necessary for this, but were found in the process of implementing the removal of driver queries.
Previous PR: https://github.com/rust-lang/rust/pull/132410
A `ChunkedBitSet` has to be at least 2048 bits for it to outperform a
`BitSet`, because that's the chunk size. The largest `SparseBitMatrix`
encountered when compiling the compiler and the entire rustc-perf
benchmark suite is less than 600 bits.
This change is a tiny perf win, but the motivation is more about
avoiding uses of `ChunkedBitSet` outside of `MixedBitSet`.
The test change is necessary to avoid hitting the `<BitSet<T> as
BitRelations<ChunkedBitSet<T>>>::subtract` method that has
`unimplemented!` in its body and isn't otherwise used.
