From 12673c5a228f282dc34693cc12d5290603f60d59 Mon Sep 17 00:00:00 2001 From: Cyd Date: Sat, 18 Jul 2026 03:26:09 -0500 Subject: [PATCH] IGC decode: convergence experiment negative; OPF2(145) = next suspect Co-Authored-By: Claude Opus 4.8 --- emulator/firmware-decomp/IGC-ENCODING-DERIVATION.md | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/emulator/firmware-decomp/IGC-ENCODING-DERIVATION.md b/emulator/firmware-decomp/IGC-ENCODING-DERIVATION.md index 7e8d229..fe5e072 100644 --- a/emulator/firmware-decomp/IGC-ENCODING-DERIVATION.md +++ b/emulator/firmware-decomp/IGC-ENCODING-DERIVATION.md @@ -264,3 +264,10 @@ MEMintoENAB, SHIFTL-family, send_em's DMA emit). Output: the op table 2. execute per-tile via igc_array.py (pixel-memory model exists there); 3. acceptance 1 = SMPTE bars from the cap7 bench stream; acceptance 2 = the fxtest asset frame. + +## Overnight experiment log (2026-07-18) +- Multi-frame convergence: NEGATIVE — 5 iterations of the 4-block tile program + are state-stable; nothing implemented propagates s145(x) into texz/texu. +- Next suspect: OPF2 = 0x3a90f291 reads addr 145 (the SCMEMA x-target!) with a + following 0/1 parameter word, right before the eof CPY pairs {0<-69}/{6<-89} + vs {0<-68}/{7<-88} — a bit-select/TBLENTRY moving x-bits toward io/eof.