repl_cwd is flaky

[nathanwhit]

---- repl::repl_cwd stdout ----
command /Users/runner/work/deno/deno/target/release/deno repl -A
command cwd /var/folders/3m/p59k4qdj0f17st0gn2cmj3640000gn/T/deno-cli-test7nNeWZ
------ Start Full Text ------
""
------- End Full Text -------
Next text: ""
thread 'repl::repl_cwd' panicked at tests/util/server/src/pty.rs:41:11:
Timed out.
stack backtrace:
   0:        0x100eda73c - <std::sys_common::backtrace::_print::DisplayBacktrace as core::fmt::Display>::fmt::h0aa20ca08aeb683c
   1:        0x100d7c51c - core::fmt::write::h168dbafcf35bac68
   2:        0x100edce40 - std::io::Write::write_fmt::h6542afca710d2aa1
   3:        0x100edc878 - std::sys_common::backtrace::print::h57b289e4b951ee17
   4:        0x100edc5a4 - std::panicking::default_hook::***closure***::h783b6c512154ec65
   5:        0x100edc0d8 - std::panicking::default_hook::hcdfa9e1e0f234a4f
   6:        0x100edd214 - std::panicking::rust_panic_with_hook::h9aea678ca49d64cf
   7:        0x100edcfa8 - std::panicking::begin_panic_handler::***closure***::ha16c3377e66deceb
   8:        0x100edcf38 - std::sys_common::backtrace::__rust_end_short_backtrace::hea8fdda1ea8a4c0e
   9:        0x100edcf2c - _rust_begin_unwind
  10:        0x100d7b478 - core::panicking::panic_fmt::h1cb43b60f5788132
  11:        0x100f3d7f0 - test_server::pty::Pty::new::h86f54c2420a93d58
  12:        0x100c3fae8 - core::ops::function::FnOnce::call_once::he12bac19933d1e37
  13:        0x100ef476c - test::__rust_begin_short_backtrace::he4ac465ee2112f70
  14:        0x100ef2548 - test::run_test::***closure***::h596d0554a4a78e26
  15:        0x100ef4af0 - std::sys_common::backtrace::__rust_begin_short_backtrace::ha5f46acead689572
  16:        0x100ef4904 - core::ops::function::FnOnce::call_once***vtable.shim***::hbfacf00afa1541e7
  17:        0x100edf164 - std::sys::pal::unix::thread::Thread::new::thread_start::h207e3c76161182d1
  18:        0x183506f94 - __pthread_joiner_wake

https://github.com/denoland/deno/actions/runs/8919724267/job/24496399647

[dsherret]

Looks like the same issue as #23625

I think we just need to increase the timeout on the CI for these tests or limit the amount of parallelism that cargo test does. I believe this would occur if the CI was too busy