Ball X Pit (PPSA31872)

[BunnyFl]

Existing thread

• I searched this category for the title ID and no thread exists for this title on any platform.

Title

Ball X Pit (PPSA31872)

Content version tested

01.301.000

AnyPS5 commit

main 7ce3b1d, no pull requests or local changes on top. The controller test below also ran on main 86b8424.

System tested

Windows 11 Pro 10.0.26200, AMD Radeon RX 9070 XT (AMD proprietary driver 26.10.1), Ryzen 7 5800X3D, GCC 15.2.0 (WinLibs 15.2.0posix-14.0.0-ucrt-r7)

Furthest point

In game

What is shown

Played in game with a DualSense, from a new save and with Continue: picture, audio and controller input work, and the game runs smoothly. Output 1920x1080. In game the log reports 310-600 presents per 10 s (mostly 500-600), with the GPU busy about 5 ms per frame; queue worker 0x0 is 97-98 % busy (about 54,000 DRAW_INDEX_2 per 10 s taking 6.9 s of it), so the frame rate is bound on the CPU (#7895).

• Hitches of 1.3 to 1.9 s now and then in play: the 10 s windows that hold them are the only ones with Vulkan.GraphicsPipeline / Vulkan.GraphicsPipelineLibrary in the top phases (6-10 ms per frame), so they are graphics pipelines built on first use. The pipeline journal grew from 90 to 118 to 205 pipelines over the runs, so they fade as the cache fills.
• One freeze in four runs, at the first level load (level3, Level_Graveyard.bank) about 74 s into the first run (empty caches, new save): presents stopped, and so did the queue worker's 10 s statistics, while the title kept opening Master.bank every 9.4 s. Three later runs, two from a new save and one with Continue, loaded the same level without it. No dump of the frozen state yet.
• A controller plugged in again is not seen on Windows: after unplugging the DualSense and plugging it in again the title gets no input, and the log shows no new Pad: connected game controller line. Fixed by fix(videoout): start the game controller subsystem on the thread that pumps events #8964, which is closed for now under the open pull request limit; I will reopen it once fix(agc): drop pending storage-image results when the guest unmaps their memory #8932 is merged or closed.
• Once on main 86b8424, about 5 s after start: AGC graphics: storage texture has no committed guest pages (abort). Two more runs on that commit did not show it.

Current blocker

None for play. The freeze at the first level load (one run in four) and the first-use hitches remain. Reconnecting a controller needs #8964 (closed for now, see above).

How it was run

relinker --windows --registry --sce-module-path app0 --module-dir Media/Modules --module-dir Media/Plugins app0/eboot.bin app.exe

The dump's eboot.bin and its 10 .prx files (sce_module, Media/Modules, Media/Plugins) are SELF containers, unwrapped with tools/unwrap_self.py first. Unity title: the modules in Media/Modules and Media/Plugins need --module-dir. tools/import_audit.py (with --modules for each of the three folders) reports 0 absent imports and 27 throwing stubs. No APS5_* variables are needed to play; the runs used APS5_PROFILE_DRAW=1 APS5_LOG_TIMESTAMPS=1 APS5_TRACE_SUMMARY_EVERY=10 and, to dump a freeze, APS5_HANG_DUMP_SECS. Default in-game settings.

Conversion, log reading and this report were prepared with AI assistance (Claude Code); the in-game result (smooth play, the hitches, the freeze, the controller) was observed by me.