2026-08-31 (later) - second payload run absorbed: five vaddrs confirmed (D419)
The D274 obSCEne changes came back from hardware exactly as hoped.
The 139-exports section confirmed twelve vaddrs (up from eight): the
previously-refuted sceKernelGetTscFrequency (0x1cf30) now confirms under
the widened band, and the four new candidates - getsid 0x1590,
sceKernelReadTsc 0x1cfa0, sceKernelGetProcessTimeCounter 0x1d010,
sceKernelGetProcessTimeCounterFrequency 0x1d030 - all pass. Each already
sat in libkernel-vaddrs.txt as a candidate with the exact vaddr hardware
reported, so this was confirmation, not correction. Promoted all five to
confirmed (7→12); firmware tests pass.
The trap to not fall into: this payload run shows 363 fails, all under 900-surface/* ("none of this library is present"). That is the payload path, not a regression - elfldr resolves only base+vaddr exports, so the whole-surface census sees no import tables. The number that matters is 139-exports: 12/12, 0xc confirmed. No byte dumps this run, so no new struct layouts; SwVersion stays the last one.
Next data still wants hardware, not code: a native-PS5-mode run to un-refuse GetModuleInfo and give PS5-native module-info layout. Nothing more to crunch here unattended.