[LeandroLP23]

What

PS5VR (PPSA99177) stops at start-up, right after it loads libSceVrTracker2:

what():  sysctl_nid_postfix: unsupported MIB 1.14.35.12828

That MIB is {CTL_KERN, KERN_PROC, KERN_PROC_ENV, pid}, and the title asks for its own pid. sysctl now answers it as FreeBSD 11's sysctl_kern_proc_env does for the calling process:

• It returns the process's start-up environment strings. AnyPS5 starts a title with an empty environment: the environ libkernel exports and libc's getenv both begin empty. So the call returns 0 bytes. FreeBSD reads the start-up strings, not later setenv changes, so those changes don't show up there either.
• EINVAL for a MIB without exactly one pid. ESRCH for any other pid, -1 included, as pget gives. EPERM for a write. These are the same rules as the kern.proc.vmmap handler next to it.

Which variables, if any, the console starts an application with is unverified. That is recorded in docs/dev/debt/libkernel.md.

Tested

Windows 11, Ryzen AI 7 350 / Radeon 860M, WinLibs GCC 15.2, Release.

• PS5VR (PPSA99177, catalog 1.2.0), 40 s runs:
  • main cc00785: the title ends with the unsupported MIB 1.14.35.<pid> exception above.
  • This PR, rebased on main 8b3a6ba (run again after the rebase): it gets past the call and lists its library (library: list Library -> ok, 7 items). It then shows its TV screen: the PS5VR logo, "Put on your PlayStation VR2" and the phone-link QR code. It also logs vr: headset setup failed, as expected with no PlayStation VR2 attached (sceHmd2Initialize -> 0x81110016).
• guest_system_configuration has a new CheckEnvSysctl, which covers:
  • the size query and the read for the own pid: 0 bytes, buffer untouched, errno unchanged;
  • a write;
  • a foreign pid and pid -1;
  • a MIB that is too short and one that is too long.
    It fails on main's SystemConfiguration.cpp and passes with this change. ctest -R "sysctl|system_config|guest_process": 6 of 6 pass.
• python tools/check_conventions.py --base origin/main passes.
• tools/pr_overlap.py on the rebased branch:
  • fix(libkernel): return ENOENT for unknown sysctl names #8910 edits the docs/dev/debt/libkernel.md line after kern.proc.vmmap, so the new entry sits before kern.proc.vmmap instead. A local merge of fix(libkernel): return ENOENT for unknown sysctl names #8910 into this branch is clean.
  • feat(libs): add four sysctl MIB mappings and an IME setter #8553 edits the same three files for other MIBs (kern.smp.maxcpus, kern.stackprot, kern.sched.name, kern.timecounter.hardware). It already conflicts with main in docs/dev/debt/libSceIme.md, a file this PR doesn't touch.

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: PPSA99177 Homebrew catalog on main d8888df9b (Windows, AMD): 23 titles reach their own UI #5745
• 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	docs/dev/debt/libkernel.md
PR 6044	core/libs/tests/GuestSystemConfiguration.cpp
PR 6149	docs/dev/debt/libkernel.md
PR 6469	docs/dev/debt/libkernel.md
PR 6740	docs/dev/debt/libkernel.md
PR 6830	docs/dev/debt/libkernel.md
PR 6936	docs/dev/debt/libkernel.md
PR 7049	docs/dev/debt/libkernel.md
PR 7088	docs/dev/debt/libkernel.md
PR 7105	docs/dev/debt/libkernel.md
PR 7121	docs/dev/debt/libkernel.md
PR 7149	docs/dev/debt/libkernel.md
PR 7509	docs/dev/debt/libkernel.md
PR 7511	docs/dev/debt/libkernel.md
PR 7565	docs/dev/debt/libkernel.md
PR 7653	docs/dev/debt/libkernel.md
PR 8010	docs/dev/debt/libkernel.md
PR 8448	docs/dev/debt/libkernel.md
PR 8520	docs/dev/debt/libkernel.md
PR 8553	core/libs/prx/libkernel/System/src/SystemConfiguration.cpp core/libs/tests/GuestSystemConfiguration.cpp docs/dev/debt/libkernel.md
PR 9133	docs/dev/debt/libkernel.md
PR 9160	docs/dev/debt/libkernel.md
PR 9284	docs/dev/debt/libkernel.md
PR 9293	docs/dev/debt/libkernel.md
PR 9297	docs/dev/debt/libkernel.md
PR 9302	docs/dev/debt/libkernel.md
PR 9561	docs/dev/debt/libkernel.md
PR 9575	docs/dev/debt/libkernel.md
PR 9669	docs/dev/debt/libkernel.md
PR 9693	docs/dev/debt/libkernel.md

[MarcZX10]

I checked this against current main: the PR is 11 commits behind, although it still merges cleanly. Please refresh the branch and rerun the focused sysctl/system-configuration tests and overlap check against the current tree before merge. The title-specific Unblocks: PPSA99177 #5745 entry is present; the hosted full Linux/Windows jobs are skipped on PRs by repository policy.

[LeandroLP23]

Rebased on current main 8b3a6ba (0 behind), head 6e1df42. Full build on Windows (WinLibs GCC 15.2, Release), then:

• ctest -R "sysctl|system_config|guest_process": 6 of 6 pass, guest_system_configuration included.
• PS5VR (PPSA99177), 40 s: it gets past the sysctl call and reaches library: list Library -> ok, 7 items, as before the rebase.
• check_conventions.py --base origin/main passes.
• pr_overlap.py: fix(libkernel): return ENOENT for unknown sysctl names #8910 changed the debt line right after kern.proc.vmmap, where the new entry was, so I moved the entry one line up, before kern.proc.vmmap. A local merge of fix(libkernel): return ENOENT for unknown sysctl names #8910 into this branch is now clean. feat(libs): add four sysctl MIB mappings and an IME setter #8553 already conflicts with main in docs/dev/debt/libSceIme.md, which this PR doesn't touch.

The PR description is updated to match.

[harisibrar1]

Checked against main 3ca07d0 and FreeBSD releng/11.4 sys/kern/kern_proc.c (sysctl_kern_proc_env, lines 1975-2001, node flags at 2956):

• The errors match FreeBSD 11.4: EINVAL when namelen != 1, ESRCH from pget for a foreign pid and for -1, and EPERM for a write from the CTLFLAG_RD node before the handler runs. The EPERM-before-EINVAL order in the new branch matches too.
• 0 bytes is consistent with main: environ_nid_postfix starts as emptyEnvironment (libkernel/System/src/Process.cpp:224) and libc's table starts empty (libc/src/Environment.cpp:26). The console's real start-up environment is recorded in the debt entry.
• No commit since cc00785 touches the three files, so the refresh MarcZX10 asked for should be conflict-free.
• Optional nit: CopyOut(nullptr, 0, ...) ends in std::memcpy(oldValue, nullptr, 0), which is formally UB even with a zero length. A one-byte static empty buffer as the source would sidestep it. Not blocking.

Reviewed with AI assistance (Claude Code); the checks above were run locally on my machine.