[parl0v]

What

tools/nid_names.py resolves 14 libcCyberUnknown* stubs, and the earlier pull requests that implemented them (#492, #1952, #1981) are closed. Two of them are being implemented in their own pull requests, abs (Ye20uNnlglA) in #2316 and __powidf2 (H+8UBOwfScI) in #2158, so this leaves those two alone. As in #1145, it renames the other 12 so a NotImplemented failure names the real function. Behaviour is unchanged: every one still throws.

NID	Old	New
5Lf51jvohTQ	libcCyberUnknown03	_Mtx_destroy
7yMFgcS8EPA	libcCyberUnknown05	_Cnd_destroy
CyXs2l-1kNA	libcCyberUnknown07	_Iswctype
JhVR7D4Ax6Y	libcCyberUnknown09	_WStoul
SreZybSRWpU	libcCyberUnknown10	_Cnd_init
VsP3daJgmVA	libcCyberUnknown11	_Cnd_broadcast
YaHc3GS7y7g	libcCyberUnknown12	_Mtx_init
gTuXQwP9rrs	libcCyberUnknown13	_Mtx_unlock
iS4aWbUonl0	libcCyberUnknown14	_Mtx_lock
vEaqE-7IZYc	libcCyberUnknown16	_Cnd_wait
rWSuTWY2JN0	libcCyberUnknown18	vsnprintf_s
tfNbpqL3D0M	libcCyberUnknown19	vsscanf_s

Each name hashes to its NID with compute_nid. The functions follow the _nid_postfix convention, so the APS5_EXPORT aliases go.

Side effect on the export table: the stubs were also exported under their own names, which nid_patcher hashed into 12 extra NIDs (for example compute_nid("libcCyberUnknown03")) that no Sony module has. Those are gone. The 12 real NIDs above are still exported: libc.prx goes from 1483 to 1471 exports, and the 12 removed are exactly the hashes of the old stub names.

The comments above vsnprintf_s and vsscanf_s (file-size idiom from #358) are left as they are; #1981 discussed that they don't match the names.

Tested

Windows 10 (19045), MinGW-w64 GCC 15.2.0 (WinLibs r7), Release, BUILD_TESTING=ON, on d70b899.

• python tools/nid_names.py: 22 unknowns / 14 suggestions before, 10 unknowns / 2 suggestions after (the two left to feat(libc): implement __powidf2 #2158 and feat(libc): implement abs #2316).
• libc.prx export table compared with objdump -p before and after, as above.
• Full ctest: 506/507 passed (1 skipped by design); guest_sce_net fails on this host on main too, fixed by test(net): reset the closed peer with unread data so the send fails at once #2295.
• tools/check_conventions.py --base origin/main passes.

Checklist

• Based on current main; no other open PR implements the same functions (tools/pr_overlap.py: no shared exports; feat(libc): implement __powidf2 #2158 and feat(libc): implement abs #2316 edit the neighbouring stubs, so whichever lands second needs a rebase)
• One topic per PR; follow-ups go in a new PR
• No comments in code except technical debt
• Unimplemented paths throw; silent stubs are listed in TechnicalDebt
• No notes, investigation .md files or images added to the repository
• New third-party code is a submodule built from source
• Generic behaviour, not specific to one title
• Depends on: none
• AI-assisted: yes

[adriadam10]

Checked all 12 names with compute_nid (each matches its NID), the diff (rename only, every function still throws, no leftover references to the old names), build + ctest (517/517) on top of current main, and a clean merge-tree. Same pattern as #1145. Merging once CI is green; whichever of #2158 / #2316 lands second needs a rebase.

[github-actions]

Progress report (d70b899 → 9cea268)

➖ System libraries: 84.69% (+0.0%, +12 declared, -12 removed functions)

🆕 12 declared as stubs

Library	Name
libc	_Cnd_broadcast_nid_postfix
libc	_Cnd_destroy_nid_postfix
libc	_Cnd_init_nid_postfix
libc	_Cnd_wait_nid_postfix
libc	_Iswctype_nid_postfix
libc	_Mtx_destroy_nid_postfix
libc	_Mtx_init_nid_postfix
libc	_Mtx_lock_nid_postfix
libc	_Mtx_unlock_nid_postfix
libc	_WStoul_nid_postfix
libc	vsnprintf_s_nid_postfix
libc	vsscanf_s_nid_postfix

🗑️ 12 removed

Library	Name
libc	libcCyberUnknown03
libc	libcCyberUnknown05
libc	libcCyberUnknown07
libc	libcCyberUnknown09
libc	libcCyberUnknown10
libc	libcCyberUnknown11
libc	libcCyberUnknown12
libc	libcCyberUnknown13
libc	libcCyberUnknown14
libc	libcCyberUnknown16
libc	libcCyberUnknown18
libc	libcCyberUnknown19