ff6sym.py resolves 7E:7233 to UpdatePPU+$42 instead of CutsceneNMI #25

New issue

Closed

opened 2026-10-29 16:47:15 +00:00 by Gatchan · 2 comments

Gatchan commented

2026-10-29 16:47:15 +00:00

Owner

Found during #8: tools/ff6sym.py 7E:7233 prints UpdatePPU+$42, but the US .dbg defines CutsceneNMI at $7E:7233. Label selection for RAM-resident (relocated) code picks a wrong neighbour; check the scope/segment handling of labels whose segment runs from WRAM. Also: tools/ram_trace_report.py does not list the RAM->RAM jump $1500 -> CutsceneNMI as a transition.

Found during #8: `tools/ff6sym.py 7E:7233` prints `UpdatePPU+$42`, but the US `.dbg` defines `CutsceneNMI` at `$7E:7233`. Label selection for RAM-resident (relocated) code picks a wrong neighbour; check the scope/segment handling of labels whose segment runs from WRAM. Also: `tools/ram_trace_report.py` does not list the RAM->RAM jump `$1500 -> CutsceneNMI` as a transition.

Gatchan added the

tools

bug

labels

2026-10-29 16:47:15 +00:00

Gatchan commented

2026-10-29 17:22:43 +00:00

Author

Owner

Starting: delegated to a fresh agent (background), one branch and PR per issue.

Gatchan referenced this issue from a commit

2026-10-29 18:15:17 +00:00

tools: ff6sym labels scoped to their segment, RAM regression tests (refs #25)

Gatchan referenced this issue from a commit

2026-10-29 18:15:17 +00:00

tools: RAM trace records jumps between RAM banks ($1500 -> CutsceneNMI), fixes #25

Gatchan referenced this issue from a pull request that will close it,

2026-10-29 18:15:34 +00:00

ff6sym: segment-scoped labels; RAM trace lists $1500 -> CutsceneNMI jumps #28

Gatchan closed this issue

2026-10-29 19:03:59 +00:00

Gatchan commented

2026-10-29 19:07:20 +00:00

Author

Owner

Merged (PR #28). Main-session check: US 7E:7233 and JP 7E:71BF both resolve to CutsceneNMI+$0; release --fast 20/20. The original report was a US address resolved with the JP .dbg; the real defect fixed here is label lookup crossing segment boundaries. RAM->RAM jumps now appear in the trace.

Merged (PR #28). Main-session check: US `7E:7233` and JP `7E:71BF` both resolve to `CutsceneNMI+$0`; release `--fast` 20/20. The original report was a US address resolved with the JP .dbg; the real defect fixed here is label lookup crossing segment boundaries. RAM->RAM jumps now appear in the trace.