fix(libkernel): accept null results in gettimeofday and clock_getres

[AchrafReyani]

What

FreeBSD treats both result pointers as optional:

• gettimeofday(tp, tzp) stores the time only when tp is not null (sys_gettimeofday: if (uap->tp) ...), so gettimeofday(NULL, &tz) or gettimeofday(NULL, NULL) succeeds.
• clock_getres(id, res) stores the resolution only when res is not null, so clock_getres(id, NULL) succeeds. POSIX says the same, and it is the usual way to check that a clock exists.

Both functions threw std::invalid_argument for a null result, which aborts the title. They now write into a local value instead, and every other path is unchanged.

Overlap: #1676 adds a branch right after the clock_getres null check that this replaces, so whichever lands second needs a two-line rebase. I'll do it here if #1676 goes first.

Tested

Windows 11, WinLibs GCC 15.2.0 r7, Release, -DBUILD_TESTING=ON:

• New checks in guest_freebsd_clocks_tests (at the start of main):
  • gettimeofday(NULL, NULL) succeeds;
  • gettimeofday(NULL, &tz) succeeds and fills tz;
  • clock_getres(id, NULL) succeeds for REALTIME, VIRTUAL, MONOTONIC, SECOND and PROCESS_CPUTIME_ID.
• On main the test aborts with gettimeofday_nid_postfix: invalid argument. Restoring only the old clock_getres check makes it abort in clock_getres, so both halves are covered.
• ctest -R "freebsd_clocks|cpu_time_clocks": 2/2 passed.
• tools/check_conventions.py passes, and tools/pr_overlap.py --branch reports no overlap. It merges cleanly with fix(libkernel): return guest errors for invalid clock IDs #2089, fix(libkernel): initialize the full local-time conversion state #2157, fix(libkernel): preserve large nanosleep durations #2067, test(libkernel): cover cancellation recovery for #1668 #1739 and fix(libkernel): read the process start before the clock in process time #1501, which touch the same files. Apart from feat(libkernel): implement pthread_getcpuclockid and per-thread CPU clocks #1676 above, feat(libkernel): implement thread cancellation #1141 conflicts only in TechnicalDebt.md, where it already conflicts with main.

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 in TechnicalDebt
• 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
• AI-assisted: yes

🤖 Generated with Claude Code

[AchrafReyani]

Full Linux and Windows builds and tests for this head (7d04df2e), using the project's Build workflow. I can't start the upstream manual run, so I ran the same build.yml in my fork through a mirror PR. Like the upstream run, the jobs check out the PR head merged into current main (d70b8998):

Job	Result
linux	success; full build, ctest: 516/516 passed
windows	success; full build, ctest: 507/507 passed
relinker (Linux, Windows, macOS)	success

Run: https://github.com/AchrafReyani/AnyPS5/actions/runs/37947367784

[oneandonlydean]

This conflicts with main after today's merges; could you rebase? It passed testing and is otherwise ready.

[AchrafReyani]

Rebased onto current main (70d6382). The only conflict was in clock_getres: main added the CPU-clock redirect after the old null check. The null result is now replaced first, then the redirect runs, so clock_getres(cpuClock, NULL) also succeeds.

On Windows, guest_freebsd_clocks, guest_cpu_time_clocks and guest_thread_cpu_clock pass, ctest -E "^agc_|video_out" is 297/297, and check_conventions.py passes. The full Linux and Windows build for the new head is running and will be posted here.

[AchrafReyani]

Full Linux and Windows builds and tests for this head (70d6382e), using the project's Build workflow. I can't start the upstream manual run, so I ran the same build.yml in my fork through a mirror PR. Like the upstream run, the jobs check out the PR head merged into current main (c72d5ce1):

Job	Result
linux	success; full build, ctest: 587/587 passed
windows	success; full build, ctest: 573/573 passed
relinker (Linux, Windows, macOS)	success

Run: https://github.com/AchrafReyani/AnyPS5/actions/runs/38021702524