[StoneDC]

What

Adds strncat exact-capacity checks for bounded counts, several existing-prefix lengths, empty/short/long sources, high bytes and the required final NUL. Output terminators sit immediately before a protected page.

Implementation basis: Expected strings come from an owned prefix plus a bounded substring. It checks the original return pointer, prefix guards and unchanged source. Contract: https://github.com/freebsd/freebsd-src/blob/833d39bd2e38421a14ea489a956261a0fa993fbc/lib/libc/string/strcat.3 . The actual production function is in StringExtensions.cpp.

Known limits: Only legal nonoverlapping inputs with sufficient destination capacity are used. This does not define strncat_s or invalid-pointer/error policy.

Existing implementation/schema: core/libs/prx/libc/src/StringExtensions.cpp

Exact topic head e5e0c73fdaf8442dc129b46c87be0d18d343e607 is one independent commit on main 5454cd2e455994d992aca39335ce60507eeda1eb. The repository overlap checker against 157b02c4f1803e9d2cd757677b9a805961203fc4 reported no confirmed overlap. It could not complete shared-file comparisons for older PRs already conflicting with main: #5021, #6085, #7081, #7447, #7715, #8150, #10324, #10615, #11625, #12271, #12682, #12688, #12700. Scoped file diffs were reviewed separately. 11 file changes projected onto main and then combined with this topic without a new conflict; 2 peer file changes already conflict with main and are not claimed automatically mergeable. They do not provide this feature or test matrix. The new test has its own source and registration, with no required landing order. If shared registration code changes before landing, this branch needs refreshed merge and affected test checks. Depends on: none. The 20 independent topics also merged together into tested integration 81a0f73e5972352ba2b50ce53077921c51112273 on the same main.

Tested

Tested topic head: e5e0c73fdaf8442dc129b46c87be0d18d343e607. Final combined integration: 81a0f73e5972352ba2b50ce53077921c51112273, main 5454cd2e455994d992aca39335ce60507eeda1eb. Ubuntu/WSL GCC 15.2, CMake 4.2.3, Python 3.14.4; Windows MinGW GCC 15.2, CMake 4.3.1, Python 3.14.3.

Built and ran registered guest_strncat_protected against the actual unpatched libc on Linux and Windows, no skips. A separate control compiled the whole relevant production source unit and this fixture: the unchanged source passed and an intentional production-function defect was caught. Those source-unit and fixture bytes are unchanged after the main refresh; current-head CMake runtime was rechecked separately. This control is not described as a full product build.

cmake --build <full-linux-build> --parallel 8: full Linux default product build passed, including a final build at the integration above. Windows used the full configuration and built the affected libc/test targets and relinker with --parallel 8; a complete default Windows build is not claimed. ctest --test-dir <build> -R "^(tools_sanitizer_json|tools_annotation_symbols_json|tools_compile_plan|tools_catalog_json|tools_nid_offline|tools_audit_strict_partial|tools_intervals_complete|tools_pad_preview|tools_release_notes_check|tools_progress_diff_json|guest_memmove_snapshot|guest_memcpy_protected|guest_memcmp_protected|guest_span_protected|guest_substring_protected|guest_tokenizer_protected|guest_strncpy_protected|guest_strncat_protected|guest_wmemmove_snapshot|guest_wide_length_protected)$" -j 8 -V --output-on-failure: 20/20 passed on each platform, zero skips. Conventions, whitespace, script registration and the public-header audit passed. No PS5/title execution or GPU validation protocol was run; macOS and hosted full-build CI are not reported as passed.

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

[github-actions]

Some comparisons could not be completed because a pull request conflicts with the base branch. Review these shared files manually before concluding that there is no overlap.

Pull request	Shared files not compared
PR 5021	core/libs/prx/libc/CMakeLists.txt
PR 6819	core/libs/prx/libc/CMakeLists.txt
PR 7081	core/libs/prx/libc/CMakeLists.txt
PR 7447	core/libs/prx/libc/CMakeLists.txt
PR 7715	core/libs/prx/libc/CMakeLists.txt
PR 8150	core/libs/prx/libc/CMakeLists.txt
PR 8965	core/libs/prx/libc/CMakeLists.txt
PR 9929	core/libs/prx/libc/CMakeLists.txt
PR 11831	core/libs/prx/libc/CMakeLists.txt
PR 13390	core/libs/prx/libc/CMakeLists.txt
PR 13486	core/libs/prx/libc/CMakeLists.txt
PR 14585	core/libs/prx/libc/CMakeLists.txt

[bunbeans2-arch]

Windows run (Windows 11, MSYS2 UCRT64 GCC 16.1), this PR (e5e0c73) merged onto main f7f03da, ctest -R ^guest_strncat_protected$, 3x each:

• PR as is: 3/3 pass (here strncat_nid_postfix forwards to the UCRT strncat)
• mutation check, strncat_nid_postfix changed to ignore the limit (strcat): 3/3 fail with SEGFAULT at the guard page

So on Windows the test does catch an unbounded concatenation, not only pass.

AI-assisted: yes (Claude Code)