Preserve Tier2 loader and clone-seed handoff

docs/control-loop-atlas/runtime-roots-camera-and-support-families.md

@@ -1410,11 +1410,15 @@
   layouts, not to the live `0x005c93cc` candidate rows.
   The positive candidate-side bound is tighter now too: `0x004120b0` explicitly declares
   `[candidate+0xba]` and `[candidate+0xbb]` as one-byte parser fields through `0x00531150`, and
-  the later helper `0x00412d70` can clone an already-materialized whole candidate row through
-  `rep movsl`, including those bytes, before `0x00412f02` picks the `Port%02d` vs `Warehouse%02d`
-  naming branch from cloned bit `[candidate+0xba]`. So the unresolved Tier-2 seam is no longer
-  “find any direct writer to candidate `+0xba/+0xbb`”; it is “find the earlier seed or projection
-  owner that first makes some source/live rows reach that clone path with nonzero bank bytes.”
+  the later helper `0x00412d70` does not consume them vaguely. It runs two explicit passes, first
+  accepting only nonzero `[candidate+0xba]` rows and then only nonzero `[candidate+0xbb]` rows,
+  remembers one availability-qualified seed row, prefers an existing target whose ordinal field
+  `[candidate+0x794]` already matches, and otherwise clones one whole `0x1f2`-dword candidate body
+  through `rep movsl`, including those bytes, before `0x00412f02` picks the `Port%02d` vs
+  `Warehouse%02d` naming branch from cloned bit `[candidate+0xba]`. So the unresolved Tier-2 seam
+  is no longer “find any direct writer to candidate `+0xba/+0xbb`”; it is “find the earlier seed
+  or projection owner that first makes some source/live rows reach that clone path with nonzero
+  bank bytes.”
   The top-level stock handoff above that clone pass is tighter now too. Direct disassembly of
   `0x004196c0` shows the broader stock `*.bca` rebuild loop formatting the wildcard path rooted at
   `0x005c93fc`, iterating the local `0x005c8190/0x005c8194/0x005c819c` find-first/find-next