[valenciana-nc]

What

Test-only follow-up to #7672, addressing @wachi35676's feedback. That PR merged at 950eeb847 while this test improvement was being prepared.

CheckReadOnlyDestination used the descriptor that had just read all entries in its writable control. On unpatched main, the read-only buf call therefore returned 0 at EOF instead of exercising the output-buffer access violation.

Close that descriptor and open a fresh one before the read-only checks. After the buf and basep refusals, restore permissions and confirm entries remain unread. No runtime code changes.

Overlap scan: #6044 relocates this test file as part of test-suite organization; it does not implement this fresh-descriptor check. Whichever change lands second must preserve the test adjustment at its resulting location.

Tested

Windows 11 (26300), WinLibs GCC 15.2.0 r7, Release, BUILD_TESTING=ON:

• Rebased onto main b106f6f8c (the merge of fix(libkernel): refuse read-only guest mappings for getdirentries on Windows #7672).
• guest_directory, guest_directory_buffer, guest_directory_rewind, guest_directory_close: 4/4 pass.
• Negative control with only Stdio.cpp restored from pre-fix 8982791bc: the improved test segfaults writing to 0x200000000 during the read-only buf call, rather than merely failing at EOF.

This test binary is Windows-only. No Linux runtime behaviour changes.

Checklist

• Based on current main; no other open PR implements the same functions
• One topic per PR; follow-ups go in a new PR
• No comments in code except technical debt
• Unimplemented paths throw (NotImplemented_nid_no_patch); silent stubs are listed under Silent stubs in their library's technical debt file
• No notes, investigation .md files or images added to the repository (attach them to this PR)
• New third-party code is a submodule built from source
• Generic behaviour, not specific to one title
• Depends on: none
• Fixes: none
• Unblocks: none
• GPUs tested: n/a
• AI-assisted: yes