Drop the last two lint expectations clippy 1.98 no longer fulfils - #8568

• I did not use AI to write the code of this patch.
• This PR follows our AI policy

Summary

#8564 removed four #[expect(clippy::std_instead_of_core)] that clippy 1.98 stopped
fulfilling. Two more were left behind, and each hid behind a cfg that the job which
would have reported it did not compile.

crates/vm/src/stdlib/_signal.rs is gated on #[cfg(windows)], so only the Windows job
sees it. It showed up on #8564's own last run, once the other four were gone:

error: this lint expectation is unfulfilled
   --> crates\vm\src\stdlib\_signal.rs:341:13
    = note: false positive: core::io::ErrorKind is unstable (core_io)

crates/host_env/src/posix.rs is gated on not(any(linux, haiku, solaris, illumos, hurd)),
so among the CI runners only macOS compiles it. It stayed quiet until #8564 landed because
clippy reports one error per crate, and the fopen expectation in the same crate came
first. It surfaced on main right after the merge:

error: this lint expectation is unfulfilled
   --> crates/host_env/src/posix.rs:1410:13

Only the attributes go; the cfg on each statement stays.

Testing

cargo fmt --check and the workspace clippy with the flags from the CI job both pass on
Linux, so the side that was already green is undisturbed. I do not have a Rust toolchain on
Windows or macOS, so I did not reproduce either of those two jobs myself. The quoted errors
are from your CI rather than from my machine, and your CI is what will confirm the fix.

AI assistance

Claude Code (claude-opus-5) was used throughout: reading the CI logs to locate the two
expectations, making the deletions, and running fmt and clippy on Linux. I reviewed the
diff and the policy before opening this.

Summary by CodeRabbit

• Chores
  • Removed unnecessary code-quality suppressions from Windows-specific and POSIX platform implementations.
  • No user-visible behavior or functionality changed.