std: Tweak expansion of thread-local const

[alexcrichton]

This commit tweaks the expansion of thread_local! when combined with a
const { ... } value to help ensure that the rules which apply to
const { ... } blocks will be the same as when they're stabilized.
Previously with this invocation:

thread_local!(static NAME: Type = const { init_expr });

this would generate (on supporting platforms):

#[thread_local]
static NAME: Type = init_expr;

instead the macro now expands to:

const INIT_EXPR: Type = init_expr;
#[thread_local]
static NAME: Type = INIT_EXPR;

with the hope that because init_expr is defined as a const item then
it's not accidentally allowing more behavior than if it were put into a
static. For example on the stabilization issue this example now
gives the same error both ways.

[rust-highfive]

r? @yaahc

(rust-highfive has picked a reviewer for you, use r? to override)

[nbdd0121]

Probably INIT_VAL would be better, since it isn't an expression.

[m-ou-se]

@bors r+

[bors]

📌 Commit 1ac5d7d has been approved by m-ou-se

[hkratz]

Could this have caused the following aarch64 assertion failure in the rollup #90975 CI (CI log):

fatal runtime error: assertion failed: thread_info.is_none()

apparently stemming from:

rust/library/std/src/sys_common/thread_info.rs

Line 13 in 1149193

thread_local! { static THREAD_INFO: RefCell<Option<ThreadInfo>> = const { RefCell::new(None) } }

rust/library/std/src/sys_common/thread_info.rs

Lines 41 to 47 in 1149193

	pub fn set(stack_guard: Option<Guard>, thread: Thread) {
	THREAD_INFO.with(move |thread_info| {
	let mut thread_info = thread_info.borrow_mut();
	rtassert!(thread_info.is_none());
	*thread_info = Some(ThreadInfo { stack_guard, thread });
	});
	}

[alexcrichton]

@bors: r-

That seems likely, yes, I will work on investigating.

[alexcrichton]

@bors: r=m-ou-se rollup=never

Hm ok so I can't reproduce this locally and the IR looks the same before/after this change, so I'm going to put this back in the queue, but I'll exclude it from rollups in case it's still the culprit.

[bors]

📌 Commit 1ac5d7d has been approved by m-ou-se

[bors]

⌛ Testing commit 1ac5d7d with merge 66fdedc993e987dce337cbd88324cff0b3488559...

[bors]

💥 Test timed out

[rust-log-analyzer]

A job failed! Check out the build log: (web) (plain)

Click to see the possible cause of the failure (guessed by this bot)

[alexcrichton]

It looks like dist-aarch64-apple took over an hour to rebuild LLVM, so presumably caches were being primed.

@bors: retry

[bors]

⌛ Testing commit 1ac5d7d with merge 548c108...

[bors]

☀️ Test successful - checks-actions
Approved by: m-ou-se
Pushing 548c108 to master...

[rust-timer]

Finished benchmarking commit (548c108): comparison url.

Summary: This benchmark run did not return any relevant changes.

If you disagree with this performance assessment, please file an issue in rust-lang/rustc-perf.

@rustbot label: -perf-regression