[Danigamer05]

What

Add the missing libc export __std_min_8u (i8DeU1sI6I4). It reads a nonempty range of unsigned 64-bit elements and returns the minimum value, rather than a pointer. Reads use the existing memcpy helper, so the range need not be aligned; the input is left untouched. Null, empty, reversed or partial-element ranges call NotImplemented_nid_no_patch; their native failure behavior is not reproduced and remains in debt. The test does not assert those unsupported paths.

The contract comes from the libc.prx bundled with PPSA03671 v01.001.005: export RVA 0x62a30, size 0x2a4, ELF SHA256 483961e086b8a1fffa8138b3c0badf798987fb5f35ec908d828946408be2d8a6. Its implementation takes rdi/rsi range pointers, reads 8-byte elements, selects the unsigned smaller value (cmovb at 0x62cc6 in the scalar tail), and returns the value in rax. The isolated original function was also executed on this Windows host from its mapped load segments, using its image-relative constants without relocations, module initialization or other module exports. This is a host differential, not a console measurement. Microsoft STL's public implementation corroborates the unsigned value-returning signature.

Declare the export once in libc's public Api.hpp; tests use that header. No third-party code, module bytes or title files are added. No current title dependency or title unblock is claimed.

Tested

Base: acb900c194325b9ee8c23c10768a9a491454f0b2. Published/tested head: 94e88881d482031d47178adcbe87a58212d3a304.

• Windows 11 build 26200, AMD Ryzen 7 3700X, WinLibs MinGW-w64 GCC 15.2.0 r7, CMake 4.3.1, Release, bundled dependencies, BUILD_TESTING=ON, ANYPS5_RELINKER_ONLY=OFF, SDL_CCACHE=OFF. Built production libs, relinker and guest_vector_algorithms_tests.
• The original guest_vector_algorithms passes on the unmodified base. Added cases cover unsigned/sign and 32/64-bit boundaries, a single value, repeated minima, lengths around vector/tail boundaries, offsets 0–31, input/canary preservation, all-maximum ranges and a unique minimum at every position. All existing cases remain.
• Three independently rebuilt negative controls fail as expected: signed comparison, truncating later elements to 32 bits, and skipping the last element. Restored exact sources and binaries afterward.
• ctest --test-dir build -R "^(guest_vector_algorithms|public_headers)$" -V --output-on-failure: both pass on Windows. The public-header audit verifies unique declarations and umbrella coverage; this is not a claim that all standalone-header compilation batches or the full CTest suite ran.
• Isolated original / NID-patched production candidate / unsigned-minimum reference: 8,609 cases, zero mismatches. This includes lengths 1–257 at offsets 0–31, 256 cases against inaccessible guard pages, every single-bit boundary, zero/UINT64_MAX, and all 64 pairs of eight unsigned edge values. Inputs remain unchanged. Repeated against the final production binary after restoration.
• Ubuntu 24.04 under WSL2 (kernel 6.6.114.1), GCC 13.3.0, CMake 3.28.3, Release, bundled dependencies: full-project configuration, production libc and guest_vector_algorithms_tests build; the same CTest expression passes both tests. A separate Git checkout fixed to the published commit is clean for the tested sources before and after the run. The existing runtime reported unavailable asynchronous userfaultfd write protection and used its memory-comparison fallback.
• Conventions and PR-diff whitespace checks pass. No console, title run, macOS, full CTest suite, GPU execution or performance claim.

Overlap: the required scan against acb900c194325b9ee8c23c10768a9a491454f0b2 reports no confirmed overlap. Manually inspected all 29 stale shared-file comparisons against their merge bases: #5021, #5303, #6713, #7081, #7386, #7447, #7544, #7648, #7715, #8150, #8417, #9269, #9371, #9447, #9526, #9669, #9771, #9863, #9873, #9929, #10007, #10113, #10226, #10405, #10467, #10610, #10742, #10788, #10898. Their changes touch other debt entries; #10742 adds _LCosh/_LSinh declarations and #10788 adds _Thrd_lt, away from this declaration. None changes this export or the adjacent vector-algorithm entry. Those stale branches are not claimed merge-clean. A trial merge-tree with #10460 (62c85f44a6a50b0ffb6ae72e7d6405c5a8c7769f, unsigned maximum) succeeds without content conflicts; the two exports are independent.

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]

Progress report (898a7fe → 94e8888)

📉 System libraries: 50.22% (-0.01%, +1 declared functions)

🆕 1 declared as stubs

Library	Name
libc	__std_min_8u_nid_postfix

Changed libraries (implemented exports)

Library	Before	After	Change
libc	1357/1367 (99.27%)	1357/1368 (99.20%)	-0.07 pp

[MarcZX10]

Re-reviewed exact head 94e88881d482031d47178adcbe87a58212d3a304. The unsigned 64-bit minimum implementation handles the nonempty range through the existing Read helper, including unaligned inputs; the added tests cover boundaries, random vectors, offsets, and guard/canary behavior. I found no concrete correctness issue in the diff.

The author reports Windows and WSL builds/differential tests (8,609 cases). Hosted conventions, eligibility, intake, labels, Ubuntu/macOS/Windows relinker, ASan, register-context, and write-watch checks pass. Full Linux and Windows product builds were path-skipped. I did not independently rerun the author’s local differential tests.

[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	docs/dev/debt/libc.md
PR 5303	docs/dev/debt/libc.md
PR 6713	docs/dev/debt/libc.md
PR 7081	docs/dev/debt/libc.md
PR 7386	docs/dev/debt/libc.md
PR 7447	docs/dev/debt/libc.md
PR 7544	docs/dev/debt/libc.md
PR 7648	docs/dev/debt/libc.md
PR 7715	docs/dev/debt/libc.md
PR 7851	core/libs/prx/libc/include/Api.hpp
PR 8150	docs/dev/debt/libc.md
PR 9371	docs/dev/debt/libc.md
PR 9447	docs/dev/debt/libc.md
PR 9526	docs/dev/debt/libc.md
PR 9648	docs/dev/debt/libc.md
PR 9677	docs/dev/debt/libc.md
PR 9772	core/libs/prx/libc/include/Api.hpp docs/dev/debt/libc.md
PR 9863	docs/dev/debt/libc.md
PR 9929	docs/dev/debt/libc.md
PR 9953	docs/dev/debt/libc.md
PR 9964	docs/dev/debt/libc.md
PR 10002	docs/dev/debt/libc.md
PR 10007	docs/dev/debt/libc.md
PR 10234	docs/dev/debt/libc.md
PR 10315	docs/dev/debt/libc.md
PR 10317	docs/dev/debt/libc.md
PR 10321	docs/dev/debt/libc.md
PR 10324	core/libs/prx/libc/include/Api.hpp docs/dev/debt/libc.md
PR 10325	docs/dev/debt/libc.md
PR 10328	docs/dev/debt/libc.md
PR 10345	docs/dev/debt/libc.md
PR 10349	docs/dev/debt/libc.md
PR 10405	docs/dev/debt/libc.md
PR 10467	docs/dev/debt/libc.md
PR 10506	docs/dev/debt/libc.md
PR 10510	docs/dev/debt/libc.md
PR 10512	docs/dev/debt/libc.md
PR 10529	docs/dev/debt/libc.md
PR 10565	docs/dev/debt/libc.md
PR 10749	core/libs/prx/libc/include/Api.hpp docs/dev/debt/libc.md
PR 10895	docs/dev/debt/libc.md
PR 10903	docs/dev/debt/libc.md
PR 11121	docs/dev/debt/libc.md
PR 11332	docs/dev/debt/libc.md
PR 11341	docs/dev/debt/libc.md

[wachi35676]

Windows run of head 94e88881d482 merged onto main 0f34ebb85, WinLibs GCC 15.2.0, Release. guest_vector_algorithms passes 2/2 with no warnings. I also merged this together with #11295 and #11300 onto main. All three merge cleanly in any order, including docs/dev/debt/libc.md, and the combined guest_vector_algorithms passes 3/3. So the three can land in any order.

AI-assisted: yes (Claude Code); results above were run on this machine.