Deal with EINTR in net timeout tests
We've seen sporadic QE failures in the timeout tests on this assertion:
assert!(kind == ErrorKind::WouldBlock || kind == ErrorKind::TimedOut);
So there's an error, but not either of the expected kinds. Adding a
format to show the kind revealed `ErrorKind::Interrupted` (`EINTR`).
For the cases that were using `read`, we can just use `read_exact` to
keep trying after interruption. For those using `recv_from`, we have to
manually loop until we get a non-interrupted result.
Enable -mergefunc-use-aliases
If the Rust LLVM fork is used, enable the -mergefunc-use-aliases
flag, which will create aliases for merged functions, rather than
inserting a call from one to the other.
A number of codegen tests needed to be adjusted, because functions
that previously fell below the thunk limit are now being merged.
Merging is prevented in various ways now.
I expect that this is going to break something, somewhere, because
it isn't able to deal with aliases properly, but we won't find out
until we try :)
This fixes#52651.
r? @rkruppe
dropping the param-env on the floor is obviously the wrong thing to do.
The ICE was probably exposed by #54490 adding the problem-exposing use of
`traits::find_associated_item`.
Fixes#55380.
Consider references and unions potentially inhabited during privacy-respecting inhabitedness checks
It isn't settled exactly how references to uninhabited types and unions of uninhabited types should act, but we should be more conservative here, as it's likely it will be permitted to soundly have values of such types.
This will also be more important in light of the changes at https://github.com/rust-lang/rust/pull/54125.
cc @RalfJung
* Create output directories for crates beforehand so rustdoc uses relative links
* Readd rustc_codegen_ssa
* Don't build out of tree dependencies for rustdoc like we don't for rustc
name-anon-globals should always be run at the very end of the pass
pipeline, as optimization passes (in particular mergefunc) may
introduce new anonymous globals.
I believe we did not run into this earlier because it requires the
rather specific combination of a) mergefunc merging two weak functions
b) compilation not using thinlto.
arena: speed up TypedArena::clear and improve common patterns
- speed up `TypedArena::clear`: improves its performance by up to **33%** (in case of a single entry)
- simplify `DroplessArena::in_arena`
experiment: Support aliasing local crate root in extern prelude
This PR provides some minimally invasive solution for the 2018 edition migration issue described in https://github.com/rust-lang/rust/issues/54647 and affecting proc macro crates.
`extern crate NAME as RENAME;` now accepts `NAME`=`self` and interprets it as referring to the local crate.
As with other `extern crate` items, `RENAME` in this case gets into extern prelude in accordance with https://github.com/rust-lang/rust/pull/54658, thus resolving https://github.com/rust-lang/rust/issues/54647.
```rust
extern crate self as serde; // Adds local crate to extern prelude as `serde`
```
This solution doesn't introduce any new syntax and has minimal maintenance cost, so it can be easily deprecated if something better is found in the future.
Closes https://github.com/rust-lang/rust/issues/54647
drop glue takes in mutable references, it should reflect that in its type
When drop glue begins, it should retag, like all functions taking references do. But to do that, it needs to take the reference at a proper type: `&mut T`, not `*mut T`.
Failing to retag can mean that the memory the reference points to remains frozen, and `EscapeToRaw` on a frozen location is a NOP, meaning later mutations cause a Stacked Borrows violation.
Cc @nikomatsakis @Gankro because Stacked Borrows
Cc @eddyb for the changes to miri argument passing (the intention is to allow passing `*mut [u8]` when `&mut [u8]` is expected and vice versa)
ci: Only run compare-mode tests on one builder
The run-pass test suite currently takes 30 minutes on Windows, and
that appears to be roughly split between two 15 minute runs of the test
suite: one without NLL and one with NLL. In discussion on Discord the
platform coverage of the NLL compare mode may not necessarily be worth
it, so this commit removes the NLL compare mode from tests by default,
and then reenables it on only one builder.
We've seen sporadic QE failures in the timeout tests on this assertion:
assert!(kind == ErrorKind::WouldBlock || kind == ErrorKind::TimedOut);
So there's an error, but not either of the expected kinds. Adding a
format to show the kind revealed `ErrorKind::Interrupted` (`EINTR`).
For the cases that were using `read`, we can just use `read_exact` to
keep trying after interruption. For those using `recv_from`, we have to
manually loop until we get a non-interrupted result.
The run-pass test suite currently takes 30 minutes on Windows, and
that appears to be roughly split between two 15 minute runs of the test
suite: one without NLL and one with NLL. In discussion on Discord the
platform coverage of the NLL compare mode may not necessarily be worth
it, so this commit removes the NLL compare mode from tests by default,
and then reenables it on only one builder.
