[p3n1cillin]

What

PS5SX2 (PPSA99203, catalog release) stops right after its "starting" notification on Windows:

terminate called after throwing an instance of 'std::runtime_error'
  what():  sysctlbyname_nid_postfix: unsupported name machdep.tsc_freq

The caller is its boot-time clock survey (guest 0x7428E, next to its [cpuclk] boot: measured TSC ... and [console] CPU: %llu Hz strings). Read with gdb at sysctlbyname and disassembled, it calls sysctlbyname("machdep.tsc_freq", &u64, &len = 8, NULL, 0). It uses the value when the call returns 0 and the value is above 100 MHz; otherwise it measures the TSC against its monotonic clock for 50 ms. The repository's main-boot.cpp has the other call, which also reads machdep.tsc_freq as an unsigned long long. That one runs only when sceKernelGetCpuFrequency returns 0, which AnyPS5's never does.

machdep.tsc_freq is FreeBSD's 64-bit sysctl_machdep_tsc_freq, present on every amd64 kernel with a TSC. sysctlbyname now reports the host TSC frequency that sceKernelGetTscFrequency already measures, because guest rdtsc reads the host counter. It uses the existing CopyOut, so a null buffer gets the length 8 and a short buffer gets a partial copy and ENOMEM, as sysctl_handle_64 does. A write, which FreeBSD would take as a new frequency, throws. The console's value is unknown; this is recorded under Unknown function info in docs/dev/debt/libkernel.md.

Tested

• PS5SX2 PPSA99203 on Windows 11 (MinGW GCC 15.2.0), from the catalog release. The relinker refuses it without --trap-sdk-syscall-stubs (a forbidden inline syscall), so this run used that debugging flag; no trapped instruction was reached. Before, it stops as above within 1 s. After, it runs for the whole 90 s run and keeps rewriting its data/PCSX2/ps5sx2-alive.txt heartbeat; nothing else reached was checked.
• guest_system_configuration: the length query, the value (nonzero and equal to sceKernelGetTscFrequency, with errno untouched), a 4-byte buffer (ENOMEM, 4 bytes copied, the rest untouched) and a throwing write. All 124 libkernel tests pass on Windows; Linux was not built locally.
• python tools/check_conventions.py --base origin/main and tools/pr_overlap.py pass.

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
• Unblocks: PS5SX2 (PPSA99203) start-up, when relinked with --trap-sdk-syscall-stubs
• AI-assisted: yes

🤖 Generated with Claude Code

[p3n1cillin]

Linux run of this change: Ubuntu in WSL2, GCC 15.2.0, Ninja Release, BUILD_TESTING=ON, full build. Built from a local integration head a14a49d8d, which is main 5d50886ee with this PR and my other open library PRs merged (#5013, #5009, #5061, #5078, #5098; no conflicts). All 121 libkernel tests and all 68 libc tests pass, including guest_system_configuration, guest_getrlimit and guest_select. Not run alone on Linux; no title run on Linux.

[coltken95]

Tested on Linux x86-64 (Release build), with 46f3d77 cherry-picked onto main 14036cb:

• With the PR: guest_system_configuration passes.
• With main's SystemConfiguration.cpp and the PR's test it aborts with
  sysctlbyname_nid_postfix: unsupported name machdep.tsc_freq

It also cherry-picks cleanly together with #5098, which edits the same files.

[oneandonlydean]

Closing: same change as #4776, which came first and is now merged. Reopen if something here is still missing on current main.