[Celegans12]

What

On Windows the guest arena ends at 0x72_0000_0000, so a fixed direct memory mapping above it throws shared mapping outside the guest arena.

• The arena now ends at the application map area end, 0xFC_0000_0000: KytyPS5's user range ends there (HOST_USER_MAX), as does SCE_KERNEL_APP_MAP_AREA_END_ADDR in GPCS4 (PS4 reference).
• Boot Demon's Souls to its first level: AGC recorder driver, runtime, audio, Bink fixes #5 kept the arena below that end; fix(libc): reserve the guest arena at its fixed range around what Windows already placed in it #231 fixed it at 448 GiB from 0x2_0000_0000, the start feat: boot PPSA04263 with its own sce_module, into the story mode load (supersedes #144) #153 had chosen.
• PPSA26344 v1.008 sets up its own virtual range at 0x1489dee (base 0x90_0000_0000, size 0x6c_0000_0000) and maps direct memory into it with fixed mappings, first 0x20000 bytes at 0x90_0000_0000 (prot 0xd3, flags 0x10).
• The out-of-arena errors name the address and length; a fixed mapping over a used range says whether it hit a host region.
• Host allocations made before libc loads (thread stacks, with high-entropy ASLR) may now sit anywhere in the arena; they stay holes, as since fix(libc): reserve the guest arena at its fixed range around what Windows already placed in it #231, and fixed mappings over them throw. TechnicalDebt records this and the assumed end.
• libSceAgcDriver's write-block counters (zero-filled) grow from 84 to 188 MiB, its committed page-state table from 112 to 250 MiB.
• fix(libc): track writes to singly mapped direct memory without faults on Windows #637 changes the line after this branch's message change in GuestArenaMap_nid_postfix; whichever merges second needs a one-line rebase. No behavioural overlap.

Tested

• Windows 11, MinGW-w64 GCC 15.2.0, based on main 53dc19c: configure -DBUILD_TESTING=ON, full build, --target libs, ctest:
• 284/284 tests pass (ctest --output-on-failure)
• guest_memory gains CheckFixedMappingsReachTheApplicationAreaEnd: a fixed mapping of the last two pages below 0xFC_0000_0000, read through an alias, and a refused one at the end; it fails on main at the range check.
• PPSA26344 v1.008, past the VideoOut checks that stop it on main: without this change it aborts at that first mapping; with it, it runs on to sceAudioOut2PortCreate: data format 0xc00 is not implemented.
• Not tested: Linux (CI), console behaviour.

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

[github-actions]

Other open pull requests touch the same code. Merging one will break the other, or both implement the same thing. If this one builds on them, list them in Depends on.

Pull request	Conflicting files (lines on main)	Same exports	Same new files
#637	core/libs/prx/libc/src/GuestArena.cpp L211-212