jqueuniet/dm2-tools
1
0
Fork 0
Code Issues 2 Pull requests Projects Releases Packages Wiki Activity Actions

dm2-saves: phase-6 timer-actuator link model is lossy — empty chain re-encodes as a full record, record type hardcoded to 7 #173

New issue
Closed
opened 2026-07-04 15:15:17 +02:00 by jqueuniet · 1 comment
jqueuniet commented 2026-07-04 15:15:17 +02:00 (Migrated from codeberg.org)
Copy link

Problem

The phase-6 timer-actuator link model (records.rs ~839-913, decode/encode_timer_actuator_links) is lossy, and encode is not decode's inverse on two edges:

  1. Empty chain: decode maps an empty link chain to val_b = 0xfffe, but encode unconditionally writes a one-record chain. A save where a 0x3c/0x3d timer's link chain is empty round-trips 1 bit into 21 bits — desyncing every phase after 6.
  2. Record type/payload: decode discards the linked record's rec_type and all payload except bytes 2-3; encode hardcodes type 7 with zeros elsewhere. A non-type-7 record in that slot (different mask width, e.g. type 3's body) desyncs on re-encode.

Current fixtures evidently never contain either shape, so the round-trip tests pass.

Also, the doc comment's claim that "the engine writes back the chain's 16-bit value stored in bytes 2-3" doesn't match the engine: wvalueB receives the newly allocated record ID via DM2_APPEND_RECORD_TO, not the record's data word. The crate model works today only because it captures the raw stream bits.

Fix direction

Preserve the whole decoded CheckcodeRecord (including the empty-chain state) in TimerActuatorLink instead of a synthesized type-7 val_b pair, so re-encode replays exactly what was decoded.

## Problem The phase-6 timer-actuator link model (`records.rs` ~839-913, `decode/encode_timer_actuator_links`) is lossy, and encode is not decode's inverse on two edges: 1. **Empty chain**: decode maps an empty link chain to `val_b = 0xfffe`, but encode unconditionally writes a one-record chain. A save where a 0x3c/0x3d timer's link chain is empty round-trips 1 bit into 21 bits — desyncing every phase after 6. 2. **Record type/payload**: decode discards the linked record's `rec_type` and all payload except bytes 2-3; encode hardcodes type 7 with zeros elsewhere. A non-type-7 record in that slot (different mask width, e.g. type 3's body) desyncs on re-encode. Current fixtures evidently never contain either shape, so the round-trip tests pass. Also, the doc comment's claim that "the engine writes back the chain's 16-bit value stored in bytes 2-3" doesn't match the engine: `wvalueB` receives the newly *allocated record ID* via `DM2_APPEND_RECORD_TO`, not the record's data word. The crate model works today only because it captures the raw stream bits. ## Fix direction Preserve the whole decoded `CheckcodeRecord` (including the empty-chain state) in `TimerActuatorLink` instead of a synthesized type-7 `val_b` pair, so re-encode replays exactly what was decoded.
jqueuniet commented 2026-07-06 08:47:21 +02:00 (Migrated from codeberg.org)
Copy link

Fixed by #209 (merged into main).

Fixed by #209 (merged into `main`).
Sign in to join this conversation.
No Branch/Tag specified
Branches Tags
main
pages
No results found.
Labels
Clear labels
No items
No labels
Milestone
Clear milestone
No items
No milestone
Projects
Clear projects
No items
No project
Assignees
Clear assignees
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
jqueuniet/dm2-tools#173
No description provided.