jan/rrt
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Box in region class-mix rebuild seam

Browse source
This commit is contained in:
Jan Petykiewicz 2026-04-18 20:41:08 -07:00
commit a105c387c3
2 changed files with 5 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

2
crates/rrt-runtime/src/smp.rs
View file

@ -5041,7 +5041,7 @@ fn build_region_service_trace_report(
"the neighboring atlas-backed restore symmetry already rules the territory side down somewhat too: 0x00487c20 only restores the live territory collection metadata/id family and its current per-entry slots 0x00487670/0x00487680 are still no-op load/save callbacks, so the territory leg now looks less likely than support or a later region-local rebuild to hide [region+0x2a4] or [region+0x310/+0x338/+0x360]".to_string(),
"the atlas-backed support seam is broader than a direct region payload owner too: 0x0040b5d0 sits over collection 0x0062b244, whose grounded live owners maintain goose-entry counters and neighboring world support lanes [world+0x4c9a/+0x4c9e/+0x4ca6/+0x4caa] plus selected support-entry state rather than an obvious per-region acquisition latch family".to_string(),
"the same grounded evidence already narrows the later per-region follow-on too: 0x00444b90 dispatches 0x00420560 over each live region after the broader restore continuation has already advanced".to_string(),
"direct disassembly already rules that per-region follow-on down as a latch owner: 0x00420560 only zeroes and recomputes [region+0x312] from the embedded profile collection [region+0x37f]/[region+0x383] and lazily seeds the year-driven [region+0x317/+0x31b] band through 0x00420350, not [region+0x276/+0x302/+0x316]".to_string(),
"direct disassembly already rules that per-region follow-on down as a latch owner: 0x00420560 only zeroes and recomputes [region+0x312] from the embedded profile collection [region+0x37f]/[region+0x383], revisits the linked placed-structure chain for class-mix terms, and lazily seeds the year-driven [region+0x317/+0x31b] band through 0x00420350, not [region+0x276/+0x302/+0x316]".to_string(),
"that leaves the broader 0x00444887 continuation as the next structured restore seam above the ruled-down 0x00421510 -> 0x0041f5c0 -> 0x00455fc0 path when chasing acquisition-side lanes [region+0x2a4] and [region+0x310/+0x338/+0x360]".to_string(),
],
blockers: vec![

6
docs/rehost-queue.md
View file

@ -613,8 +613,10 @@ Working rule:
checkpoint above the ruled-down `0x00421510 -> 0x0041f5c0 -> 0x00455fc0` path. It immediately
advances into `0x00487c20` territory refresh and `0x0040b5d0` support refresh, then later
re-enters the per-region follow-on loop at `0x00444b90 -> 0x00420560`. Current disassembly keeps
`0x00420560` on `[region+0x312]` / `[region+0x317/+0x31b]` only, so the next region pass should
treat the broader `0x00444887` continuation as the live handoff seam when chasing
`0x00420560` on the profile/class-mix scalar side only: it recomputes `[region+0x312]` from the
embedded profile collection and linked placed-structure class mix, then seeds the year-driven
`[region+0x317/+0x31b]` band through `0x00420350`. So the next region pass should treat the
broader `0x00444887` continuation as the live handoff seam when chasing
`[region+0x2a4]` and `[region+0x310/+0x338/+0x360]`, not as just another generic restore note.
- That same continuation is slightly less symmetric now too: the atlas-backed territory side at
`0x00487c20` currently restores only collection metadata/live ids and still uses no-op per-entry