regression: `RefCell<LineWriter<std::io::stdio::StdoutRaw>>` cannot be shared between threads safely

[Mark-Simulacrum]

The error seems probably correct, but it's not clear why this would have started failing now.

• https://crater-reports.s3.amazonaws.com/beta-1.80-4/beta-2024-06-22/reg/clap-io-0.1.0/log.txt
• https://crater-reports.s3.amazonaws.com/beta-1.80-4/beta-2024-06-22/reg/fakelogs-0.1.10-75501d4/log.txt

[INFO] [stdout] error[E0277]: `RefCell<LineWriter<std::io::stdio::StdoutRaw>>` cannot be shared between threads safely
[INFO] [stdout]    --> src/lib.rs:205:20
[INFO] [stdout]     |
[INFO] [stdout] 205 |                 Ok(Box::new(stdout))
[INFO] [stdout]     |                    ^^^^^^^^^^^^^^^^ `RefCell<LineWriter<std::io::stdio::StdoutRaw>>` cannot be shared between threads safely
[INFO] [stdout]     |
[INFO] [stdout]     = help: the trait `Sync` is not implemented for `RefCell<LineWriter<std::io::stdio::StdoutRaw>>`, which is required by `StdoutLock<'_>: Sync`
[INFO] [stdout]     = note: if you want to do aliasing and mutation between multiple threads, use `std::sync::RwLock` instead
[INFO] [stdout]     = note: required for `ReentrantLockGuard<'_, RefCell<LineWriter<std::io::stdio::StdoutRaw>>>` to implement `Sync`
[INFO] [stdout] note: required because it appears within the type `StdoutLock<'_>`
[INFO] [stdout]    --> /rustc/64a1fe67112931359c7c9a222f08fd206255c2b5/library/std/src/io/stdio.rs:614:12
[INFO] [stdout]     = note: required for the cast from `Box<StdoutLock<'_>>` to `Box<dyn std::io::Write + Sync>`

[workingjubilee]

Minimized:

use std::io::{self, Write};

pub fn open_stdout() -> io::Result<Box<dyn Write + Sync + 'static>> {
    let stdout = Box::leak(Box::new(io::stdout()));
    let stdout = stdout.lock();
    Ok(Box::new(stdout))
}

[workingjubilee]

cargo-bisect-rustc informs us it is rust-lang-ci@fd225fc

This is collateral damage of

• ReentrantLockGuard's Sync impl is unsound #125526
• Add manual Sync impl for ReentrantLockGuard #125527

...apparently, std said it can have a little bit of unsoundness, as a treat? It seems like exposing #121440 is proving to be a good idea because otherwise we would not have had this actually audited.

cc @joboet @programmerjake