Remove std::io::lazy::Lazy in favour of SyncOnceCell

[m-ou-se]

The (internal) std::io::lazy::Lazy was used to lazily initialize the stdout and stdin buffers (and mutexes). It uses atexit() to register a destructor to flush the streams on exit, and mark the streams as 'closed'. Using the stream afterwards would result in a panic.

Stdout uses a LineWriter which contains a BufWriter that will flush the buffer on drop. This one is important to be executed during shutdown, to make sure no buffered output is lost. It also forbids access to stdout afterwards, since the buffer is already flushed and gone.

Stdin uses a BufReader, which does not implement Drop. It simply forgets any previously read data that was not read from the buffer yet. This means that in the case of stdin, the atexit() function's only effect is making stdin inaccessible to the program, such that later accesses result in a panic. This is uncessary, as it'd have been safe to access stdin during shutdown of the program.

---

This change removes the entire io::lazy module in favour of SyncOnceCell. SyncOnceCell's fast path is much faster (a single atomic operation) than locking a sys_common::Mutex on every access like Lazy did.

However, SyncOnceCell does not use atexit() to drop the contained object during shutdown.

As noted above, this is not a problem for stdin. It simply means stdin is now usable during shutdown.

The atexit() call for stdout is moved to the stdio module. Unlike the now-removed Lazy struct, SyncOnceCell does not have a 'gone and unusable' state that panics. Instead of adding this again, this simply replaces the buffer with one with zero capacity. This effectively flushes the old buffer and makes any writes afterwards pass through directly without touching a buffer, making print!() available during shutdown without panicking.

---

In addition, because the contents of the SyncOnceCell are no longer dropped, we can now use &'static instead of Arc in Stdout and Stdin. This also saves two levels of indirection in stdin() and stdout(), since Lazy effectively stored a Box<Arc<T>>, and SyncOnceCell stores the T directly.

[rust-highfive]

r? @dtolnay

(rust_highfive has picked a reviewer for you, use r? to override)

[m-ou-se]

Two interesting minor differences before/after:

---

1:

use std::{thread, io, mem, time::Duration};

fn main() {
    print!("hello");
    thread::spawn(|| mem::forget(io::stdout()));
    thread::sleep(Duration::from_secs(1));
}

Before this change, this outputs nothing: The Arc count does not drop back to 0, the stdout buffer is not dropped/flushed.
After this change, this properly flushes the stream, and outputs hello. Improvement!

---

2:

use std::{thread, io, mem, time::Duration};

fn main() {
    print!("hello");
    thread::spawn(|| mem::forget(io::stdout().lock()));
    thread::sleep(Duration::from_secs(1));
}

Before this change, this outputs hello, while after this change, this no longer flushes the stream and outputs nothing. :( However: This only worked because the whole ReentrantMutex in the Arc was dropped (thus dropping/flushing the contained buffer), even though the ReentrantMutex inside there was still locked (since a guard was leaked). Strictly speaking, this would've been unsound: destroying a locked mutex is undefined behaviour according to posix. (Although, this seems safe on all existing implementations. See #31936.)

(Adding a print!() at the end of main here would've deadlocked, both before and after this change.)

[m-ou-se]

@rustbot modify labels: +T-libs +A-io

[dtolnay]

This looks amazing to me. Thank you @m-ou-se!

[dtolnay]

To get some eyes on the externally observable behavior changes in #77154 (comment):
@rfcbot fcp merge

[rfcbot]

Team member @dtolnay has proposed to merge this. The next step is review by the rest of the tagged team members:

• @Amanieu
• @BurntSushi
• @KodrAus
• @dtolnay
• @sfackler
• @withoutboats

No concerns currently listed.

Once a majority of reviewers approve (and at most 2 approvals are outstanding), this will enter its final comment period. If you spot a major issue that hasn't been raised at any point in this process, please speak up!

See this document for info about what commands tagged team members can give me.

[withoutboats]

IMO all the externally visible differences are bug fixes. This is great!

[rfcbot]

🔔 This is now entering its final comment period, as per the review above. 🔔

[dtolnay]

@bors r+

[bors]

📌 Commit 6f9c132 has been approved by dtolnay

[RalfJung]

Maybe add comments explaining why this is ignored?

[m-ou-se]

Maybe add comments explaining why this is ignored?

Done.

It might be the case that adding this in the test would also solve it:

#![cfg_attr(target_os = "emscripten", feature(link_args))]
#[cfg(target_os = "emscripten")]
#[allow(unused_attributes)]
#[link_args = "-s EXIT_RUNTIME=1"]
extern "C" {}

But I'm unable to test this locally. And I don't think this test really matters on emscripten.

Edit: Tested it. Kind of works, but still fails: with EXIT_RUNTIME enabled Emscripten appends a newline by itself, failing the test. So I'll just leave the // ignore-emscripten there.

[RalfJung]

@bors r=dtolnay rollup=iffy

[bors]

📌 Commit 6b8b9c4 has been approved by dtolnay

[bors]

⌛ Testing commit 6b8b9c4 with merge c9e5e6a...

[bors]

☀️ Test successful - checks-actions, checks-azure
Approved by: dtolnay
Pushing c9e5e6a to master...