2026-08-30 - obSCEne drew its own report on the console's display

OBS|display|ready|1920x1080 framebuffer|0x0
OBS|display|presenting|a submitted frame reached the display|0x0

Evidence: the records above, and docs/screenshots/ps5-hardware.png - a screen legible enough to read every check name. Build: DISPLAY_MEM=0 (write-back onion), framebuffer aligned to 0x10000, two buffers.

presenting is measured rather than assumed - the frame counter moved, which is a stronger statement than the flip having been accepted (D187).

The defect was an alignment, three steps earlier than anyone was looking. 0x4000 is refused for a scanout buffer and 0x10000 is accepted, and 0x80290015 is returned identically whatever else is right or wrong - so five sessions of varying display arguments never moved it. It was an argument to sceKernelAllocateDirectMemory. (D253)

Then the picture tore, because one buffer means drawing into the memory being scanned out. Two buffers and a bounded wait for the flip to take effect fixed it. Not an SDK problem - an ordinary choice about how many buffers to allocate, made the wrong way. (D256, D257)

---