std::process::Command::spawn cause panic on Windows 10 21H2 19044.2846

[Aobanana-chan]

I tried this code:

#[test]
fn test2() {
    std::process::Command::new("adb")
    .args(["devices"])
    .stdout(Stdio::piped())
    .stderr(Stdio::piped())
    .spawn()
    .unwrap();
}

Only panics when using use std::process::Command; in Test Debug mode, but not in other cases.
And only panic on Windows 10 21H2 19044.2846.

Here is an issue discussing this problem, and it includes two videos of about ten seconds to demonstrate the situation.
tokio-rs/tokio#6934

Meta

rustc --version --verbose:

rustc 1.82.0 (f6e511eec 2024-10-15)
binary: rustc
commit-hash: f6e511eec7342f59a25f7c0534f1dbea00d01b14
commit-date: 2024-10-15
host: x86_64-pc-windows-msvc
release: 1.82.0
LLVM version: 19.1.1

Backtrace

test test2 ... FAILED

successes:

successes:

failures:

---- test2 stdout ----
thread 'test2' panicked at src/main.rs:80:6:
called `Result::unwrap()` on an `Err` value: Os { code: 6, kind: Uncategorized, message: "鍙ユ焺鏃犳晥銆? }
stack backtrace:
   0: std::panicking::begin_panic_handler
             at /rustc/f6e511eec7342f59a25f7c0534f1dbea00d01b14\library/std\src\panicking.rs:662
   1: core::panicking::panic_fmt
             at /rustc/f6e511eec7342f59a25f7c0534f1dbea00d01b14\library/core\src\panicking.rs:74
   2: core::result::unwrap_failed
             at /rustc/f6e511eec7342f59a25f7c0534f1dbea00d01b14\library/core\src\result.rs:1677
   3: enum2$<core::result::Result<std::process::Child,std::io::error::Error> >::unwrap
             at /rustc/f6e511eec7342f59a25f7c0534f1dbea00d01b14\library\core\src\result.rs:1102
   4: static_lib::test2
             at .\src\main.rs:77
   5: static_lib::test2::closure$0
             at .\src\main.rs:76
   6: core::ops::function::FnOnce::call_once<static_lib::test2::closure_env$0,tuple$<> >
             at /rustc/f6e511eec7342f59a25f7c0534f1dbea00d01b14\library\core\src\ops\function.rs:250
   7: core::ops::function::FnOnce::call_once
             at /rustc/f6e511eec7342f59a25f7c0534f1dbea00d01b14\library/core\src\ops\function.rs:250
note: Some details are omitted, run with `RUST_BACKTRACE=full` for a verbose backtrace.
2

failures:
    test2

test result: FAILED. 0 passed; 1 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.07s

I deciphered the garbled message, and it seems to mean "The handle is invalid."

[jieyouxu]

I'm going to mark this as S-needs-repro because uhh that seems very specific. Since as you mentioned

This issue did not occur on my other PC(Windows 11 23H2 22631.4317)

[ChrisDenton]

This could use some help narrowing it down. I couldn't immediately reproduce it.

Could you try only running the tests on one thread? E.g. cargo test -- --test-threads=1

Can you reproduce it on nightly?

[Aobanana-chan]

Could you try only running the tests on one thread? E.g. cargo test -- --test-threads=1

As I mentioned, panic only occurs during Debug Test. Using cargo test or cargo test -- --test-threads=1, or clicking the Rust Analyzer's Run Test button, all pass the test normally. Panic only occurs when clicking the Rust Analyzer's Debug button. Also, if it's not a test and you debug this code directly in the main function, no panic occurs. Panic only happens during Debug Test.

Can you reproduce it on nightly?

I tried switching to nightly, but the result is the same as stable
rustc 1.84.0-nightly (df4ca44 2024-10-27)