Miri reports leak in thread local allocation

[Mark-Simulacrum]

This failed in #123941 (I'm hoping it's not related to that PR's changes) on gnu-aux:

2024-04-15T05:56:37.4909233Z �[1m�[32m     Running�[0m unittests src/lib.rs (obj/build/x86_64-unknown-linux-gnu/stage1-std/miri/i686-pc-windows-gnu/debug/deps/std-8c5441287b78bd95.exe)
[...]
2024-04-15T06:03:09.6754201Z error: memory leaked: alloc3166557 (Rust heap, size: 16, align: 4), allocated here:
2024-04-15T06:03:09.6760527Z ##[error]   --> /checkout/library/alloc/src/alloc.rs:100:9
2024-04-15T06:03:09.6762309Z     |
2024-04-15T06:03:09.6764983Z 100 |         __rust_alloc(layout.size(), layout.align())
2024-04-15T06:03:09.6765800Z     |         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2024-04-15T06:03:09.6766357Z     |
2024-04-15T06:03:09.6766612Z     = note: BACKTRACE:
2024-04-15T06:03:09.6767207Z     = note: inside `realstd::alloc::alloc` at /checkout/library/alloc/src/alloc.rs:100:9: 100:52
2024-04-15T06:03:09.6768136Z     = note: inside `realstd::alloc::Global::alloc_impl` at /checkout/library/alloc/src/alloc.rs:183:73: 183:86
2024-04-15T06:03:09.6769447Z     = note: inside `<realstd::alloc::Global as core::alloc::Allocator>::allocate` at /checkout/library/alloc/src/alloc.rs:243:9: 243:39
2024-04-15T06:03:09.6770476Z     = note: inside `alloc_crate::alloc::exchange_malloc` at /checkout/library/alloc/src/alloc.rs:332:11: 332:34
2024-04-15T06:03:09.6771753Z     = note: inside `realstd::boxed::Box::<realstd::sys::thread_local::os_local::Value<core::cell::Cell<(usize, bool)>>>::new` at /checkout/library/alloc/src/boxed.rs:218:9: 218:20
2024-04-15T06:03:09.6773482Z     = note: inside `realstd::thread::local_impl::Key::<core::cell::Cell<(usize, bool)>>::try_initialize::<{closure@realstd::rt::panic_count::LOCAL_PANIC_COUNT::__getit::{closure#0}}>` at /checkout/library/std/src/sys/thread_local/os_local.rs:145:37: 145:94
2024-04-15T06:03:09.6775495Z     = note: inside `realstd::thread::local_impl::Key::<core::cell::Cell<(usize, bool)>>::get::<{closure@realstd::rt::panic_count::LOCAL_PANIC_COUNT::__getit::{closure#0}}>` at /checkout/library/std/src/sys/thread_local/os_local.rs:127:18: 127:43
2024-04-15T06:03:09.6777050Z     = note: inside `realstd::rt::panic_count::LOCAL_PANIC_COUNT::__getit` at /checkout/library/std/src/sys/thread_local/os_local.rs:28:17: 37:19
2024-04-15T06:03:09.6778549Z     = note: inside `realstd::thread::LocalKey::<core::cell::Cell<(usize, bool)>>::try_with::<{closure@realstd::rt::panic_count::is_zero_slow_path::{closure#0}}, bool>` at /checkout/library/std/src/thread/local.rs:285:37: 285:55
2024-04-15T06:03:09.6780283Z     = note: inside `realstd::thread::LocalKey::<core::cell::Cell<(usize, bool)>>::with::<{closure@realstd::rt::panic_count::is_zero_slow_path::{closure#0}}, bool>` at /checkout/library/std/src/thread/local.rs:262:9: 262:25
2024-04-15T06:03:09.6781634Z     = note: inside `realstd::rt::panic_count::is_zero_slow_path` at /checkout/library/std/src/panicking.rs:457:9: 457:51
2024-04-15T06:03:09.6782624Z     = note: inside `realstd::rt::panic_count::count_is_zero` at /checkout/library/std/src/panicking.rs:448:13: 448:32
2024-04-15T06:03:09.6783568Z     = note: inside `realstd::panicking::panicking` at /checkout/library/std/src/panicking.rs:587:6: 587:34
2024-04-15T06:03:09.6784553Z     = note: inside `realstd::thread::panicking` at /checkout/library/std/src/thread/mod.rs:830:5: 830:27
2024-04-15T06:03:09.6785529Z     = note: inside `realstd::sync::poison::Flag::guard` at /checkout/library/std/src/sync/poison.rs:45:24: 45:43
2024-04-15T06:03:09.6786817Z     = note: inside `realstd::sync::MutexGuard::<'_, realstd::sync::mpmc::waker::Waker>::new` at /checkout/library/std/src/sync/mutex.rs:525:28: 525:47
2024-04-15T06:03:09.6788017Z     = note: inside `realstd::sync::Mutex::<realstd::sync::mpmc::waker::Waker>::lock` at /checkout/library/std/src/sync/mutex.rs:318:13: 318:34
2024-04-15T06:03:09.6789155Z     = note: inside `realstd::sync::mpmc::waker::SyncWaker::notify` at /checkout/library/std/src/sync/mpmc/waker.rs:174:29: 174:46
2024-04-15T06:03:09.6790635Z     = note: inside `realstd::sync::mpmc::list::Channel::<test::event::CompletedTest>::write` at /checkout/library/std/src/sync/mpmc/list.rs:273:9: 273:32
2024-04-15T06:03:09.6791880Z     = note: inside `realstd::sync::mpmc::list::Channel::<test::event::CompletedTest>::send` at /checkout/library/std/src/sync/mpmc/list.rs:403:18: 403:40
2024-04-15T06:03:09.6793099Z     = note: inside `realstd::sync::mpmc::Sender::<test::event::CompletedTest>::send` at /checkout/library/std/src/sync/mpmc/mod.rs:128:41: 128:61
2024-04-15T06:03:09.6794290Z     = note: inside `realstd::sync::mpsc::Sender::<test::event::CompletedTest>::send` at /checkout/library/std/src/sync/mpsc/mod.rs:613:9: 613:27
2024-04-15T06:03:09.6795291Z     = note: inside `test::run_test_in_process` at /checkout/library/test/src/lib.rs:661:5: 661:29
2024-04-15T06:03:09.6796040Z     = note: inside closure at /checkout/library/test/src/lib.rs:570:43: 578:18
2024-04-15T06:03:09.6796709Z     = note: inside closure at /checkout/library/test/src/lib.rs:598:41: 598:83
2024-04-15T06:03:09.6797917Z     = note: inside `realstd::sys_common::backtrace::__rust_begin_short_backtrace::<{closure@test::run_test::{closure#1}}, ()>` at /checkout/library/std/src/sys_common/backtrace.rs:155:18: 155:21
2024-04-15T06:03:09.6799058Z     = note: inside closure at /checkout/library/std/src/thread/mod.rs:542:17: 542:78
2024-04-15T06:03:09.6800731Z     = note: inside `<core::panic::AssertUnwindSafe<{closure@realstd::thread::Builder::spawn_unchecked_<'_, '_, {closure@test::run_test::{closure#1}}, ()>::{closure#2}::{closure#0}}> as core::ops::FnOnce<()>>::call_once` at /checkout/library/core/src/panic/unwind_safe.rs:272:9: 272:19
2024-04-15T06:03:09.6803054Z     = note: inside `realstd::panicking::r#try::do_call::<core::panic::AssertUnwindSafe<{closure@realstd::thread::Builder::spawn_unchecked_<'_, '_, {closure@test::run_test::{closure#1}}, ()>::{closure#2}::{closure#0}}>, ()>` at /checkout/library/std/src/panicking.rs:552:40: 552:43
2024-04-15T06:03:09.6805259Z     = note: inside `realstd::panicking::r#try::<(), core::panic::AssertUnwindSafe<{closure@realstd::thread::Builder::spawn_unchecked_<'_, '_, {closure@test::run_test::{closure#1}}, ()>::{closure#2}::{closure#0}}>>` at /checkout/library/std/src/panicking.rs:516:19: 516:88
2024-04-15T06:03:09.6807415Z     = note: inside `realstd::panic::catch_unwind::<core::panic::AssertUnwindSafe<{closure@realstd::thread::Builder::spawn_unchecked_<'_, '_, {closure@test::run_test::{closure#1}}, ()>::{closure#2}::{closure#0}}>, ()>` at /checkout/library/std/src/panic.rs:149:14: 149:33
2024-04-15T06:03:09.6808783Z     = note: inside closure at /checkout/library/std/src/thread/mod.rs:541:30: 543:16
2024-04-15T06:03:09.6810256Z     = note: inside `<{closure@realstd::thread::Builder::spawn_unchecked_<'_, '_, {closure@test::run_test::{closure#1}}, ()>::{closure#2}} as core::ops::FnOnce<()>>::call_once - shim(vtable)` at /checkout/library/core/src/ops/function.rs:250:5: 250:71
2024-04-15T06:03:09.6811816Z     = note: inside `<realstd::boxed::Box<dyn core::ops::FnOnce()> as core::ops::FnOnce<()>>::call_once` at /checkout/library/alloc/src/boxed.rs:2018:9: 2018:52
2024-04-15T06:03:09.6813179Z     = note: inside `<realstd::boxed::Box<realstd::boxed::Box<dyn core::ops::FnOnce()>> as core::ops::FnOnce<()>>::call_once` at /checkout/library/alloc/src/boxed.rs:2018:9: 2018:52
2024-04-15T06:03:09.6814546Z     = note: inside `realstd::sys::pal::windows::thread::Thread::new::thread_start` at /checkout/library/std/src/sys/pal/windows/thread.rs:52:13: 52:60
2024-04-15T06:03:09.6815241Z 
2024-04-15T06:03:09.6815699Z note: the evaluated program leaked memory, pass `-Zmiri-ignore-leaks` to disable this check

[bjorn3]

cc @rust-lang/miri

[saethlin]

I suspect this is another duplicate report of #123583, though it has a slightly different backtrace. See #123937

[Mark-Simulacrum]

Ah, yes, that seems plausible. I ignored that issue since it mentions reentrant locks and this backtrace at least doesn't seem like it involves those (maybe I missed it).

[RalfJung]

Yeah the issue would affect all thread-local storage. So seems plausible that it is the same issue. So hopefully this is fixed by #123937.

Closing as a duplicate.